{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010548523206751054,
      "grad_norm": 1.2812775373458862,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.5085,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1593514680862427,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.8939,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8681182861328125,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.2491,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 1.207947850227356,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.7397,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.8585994839668274,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.2381,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.9696254730224609,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.7319,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 0.8094930052757263,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3524,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.9405311346054077,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.1094,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.9403262138366699,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.9086,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.8840659856796265,
      "learning_rate": 0.0015,
      "loss": 3.7682,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.6107629537582397,
      "learning_rate": 0.0015,
      "loss": 3.615,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.6729199290275574,
      "learning_rate": 0.0015,
      "loss": 3.5046,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.9483907222747803,
      "learning_rate": 0.0015,
      "loss": 3.3984,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.6472296118736267,
      "learning_rate": 0.0015,
      "loss": 3.2931,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 1.04964280128479,
      "learning_rate": 0.0015,
      "loss": 3.2087,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 1.2234489917755127,
      "learning_rate": 0.0015,
      "loss": 3.1494,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.7597624659538269,
      "learning_rate": 0.0015,
      "loss": 3.0761,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.8128319382667542,
      "learning_rate": 0.0015,
      "loss": 3.0278,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.7083474397659302,
      "learning_rate": 0.0015,
      "loss": 2.9747,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.931122362613678,
      "learning_rate": 0.0015,
      "loss": 2.922,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.9296771287918091,
      "learning_rate": 0.0015,
      "loss": 2.8801,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.88149094581604,
      "learning_rate": 0.0015,
      "loss": 2.8465,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 1.0161752700805664,
      "learning_rate": 0.0015,
      "loss": 2.7913,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 1.2552896738052368,
      "learning_rate": 0.0015,
      "loss": 2.7604,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.9404460787773132,
      "learning_rate": 0.0015,
      "loss": 2.7247,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.8178855776786804,
      "learning_rate": 0.0015,
      "loss": 2.692,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 1.1399712562561035,
      "learning_rate": 0.0015,
      "loss": 2.6594,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.9524788856506348,
      "learning_rate": 0.0015,
      "loss": 2.6265,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.8410091996192932,
      "learning_rate": 0.0015,
      "loss": 2.6029,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.8256937861442566,
      "learning_rate": 0.0015,
      "loss": 2.5783,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.8652192950248718,
      "learning_rate": 0.0015,
      "loss": 2.5543,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.8378183841705322,
      "learning_rate": 0.0015,
      "loss": 2.5246,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.9703221321105957,
      "learning_rate": 0.0015,
      "loss": 2.5132,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.9964367151260376,
      "learning_rate": 0.0015,
      "loss": 2.491,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.8810935616493225,
      "learning_rate": 0.0015,
      "loss": 2.4549,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 1.1474449634552002,
      "learning_rate": 0.0015,
      "loss": 2.4343,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.806471586227417,
      "learning_rate": 0.0015,
      "loss": 2.4207,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.8217954039573669,
      "learning_rate": 0.0015,
      "loss": 2.4096,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 1.1262099742889404,
      "learning_rate": 0.0015,
      "loss": 2.3877,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.7740164399147034,
      "learning_rate": 0.0015,
      "loss": 2.3658,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 1.1275774240493774,
      "learning_rate": 0.0015,
      "loss": 2.3585,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 1.7523329257965088,
      "learning_rate": 0.0015,
      "loss": 2.3388,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.8717442750930786,
      "learning_rate": 0.0015,
      "loss": 2.3199,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 1.1499181985855103,
      "learning_rate": 0.0015,
      "loss": 2.2914,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.7270553112030029,
      "learning_rate": 0.0015,
      "loss": 2.2933,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.9460956454277039,
      "learning_rate": 0.0015,
      "loss": 2.2776,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.9732592701911926,
      "learning_rate": 0.0015,
      "loss": 2.258,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.8576884269714355,
      "learning_rate": 0.0015,
      "loss": 2.2513,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.7536846995353699,
      "learning_rate": 0.0015,
      "loss": 2.2248,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.7293424606323242,
      "learning_rate": 0.0015,
      "loss": 2.2198,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.8209120035171509,
      "learning_rate": 0.0015,
      "loss": 2.2068,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.9456562995910645,
      "learning_rate": 0.0015,
      "loss": 2.195,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.7237590551376343,
      "learning_rate": 0.0015,
      "loss": 2.183,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.7735239863395691,
      "learning_rate": 0.0015,
      "loss": 2.1598,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.8923589587211609,
      "learning_rate": 0.0015,
      "loss": 2.1625,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.8306816220283508,
      "learning_rate": 0.0015,
      "loss": 2.1284,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.7697585225105286,
      "learning_rate": 0.0015,
      "loss": 2.1323,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 1.1052048206329346,
      "learning_rate": 0.0015,
      "loss": 2.135,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 1.1018104553222656,
      "learning_rate": 0.0015,
      "loss": 2.1067,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.1823917627334595,
      "learning_rate": 0.0015,
      "loss": 2.0901,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.8374204039573669,
      "learning_rate": 0.0015,
      "loss": 2.092,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.8208237886428833,
      "learning_rate": 0.0015,
      "loss": 2.0795,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.8943114280700684,
      "learning_rate": 0.0015,
      "loss": 2.0658,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.9851930141448975,
      "learning_rate": 0.0015,
      "loss": 2.0748,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.810062825679779,
      "learning_rate": 0.0015,
      "loss": 2.06,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.8523632287979126,
      "learning_rate": 0.0015,
      "loss": 2.0514,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 1.0361554622650146,
      "learning_rate": 0.0015,
      "loss": 2.0356,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.7306673526763916,
      "learning_rate": 0.0015,
      "loss": 2.0288,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7139773368835449,
      "learning_rate": 0.0015,
      "loss": 2.0344,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.7125588655471802,
      "learning_rate": 0.0015,
      "loss": 2.0184,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.8224959373474121,
      "learning_rate": 0.0015,
      "loss": 2.0093,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.7942470908164978,
      "learning_rate": 0.0015,
      "loss": 1.9916,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 1.0039414167404175,
      "learning_rate": 0.0015,
      "loss": 1.9782,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.7161281108856201,
      "learning_rate": 0.0015,
      "loss": 1.9929,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.89964359998703,
      "learning_rate": 0.0015,
      "loss": 1.9769,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.3424034118652344,
      "learning_rate": 0.0015,
      "loss": 1.9741,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.7557450532913208,
      "learning_rate": 0.0015,
      "loss": 1.9643,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.8475913405418396,
      "learning_rate": 0.0015,
      "loss": 1.9627,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.7454026341438293,
      "learning_rate": 0.0015,
      "loss": 1.9669,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.7474228143692017,
      "learning_rate": 0.0015,
      "loss": 1.9402,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.6519136428833008,
      "learning_rate": 0.0015,
      "loss": 1.9262,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 1.1171636581420898,
      "learning_rate": 0.0015,
      "loss": 1.9373,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.8100257515907288,
      "learning_rate": 0.0015,
      "loss": 1.9426,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 1.1025477647781372,
      "learning_rate": 0.0015,
      "loss": 1.9219,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.9628086090087891,
      "learning_rate": 0.0015,
      "loss": 1.9181,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.9341379404067993,
      "learning_rate": 0.0015,
      "loss": 1.9154,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.952279269695282,
      "learning_rate": 0.0015,
      "loss": 1.9134,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.8489305377006531,
      "learning_rate": 0.0015,
      "loss": 1.9037,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 1.0164152383804321,
      "learning_rate": 0.0015,
      "loss": 1.9004,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.7298189997673035,
      "learning_rate": 0.0015,
      "loss": 1.8952,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.7593315243721008,
      "learning_rate": 0.0015,
      "loss": 1.9036,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 1.3297529220581055,
      "learning_rate": 0.0015,
      "loss": 1.8938,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.7776719927787781,
      "learning_rate": 0.0015,
      "loss": 1.8775,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.7080864310264587,
      "learning_rate": 0.0015,
      "loss": 1.8765,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 1.080140471458435,
      "learning_rate": 0.0015,
      "loss": 1.8614,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.9091824293136597,
      "learning_rate": 0.0015,
      "loss": 1.8678,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 1.3618862628936768,
      "learning_rate": 0.0015,
      "loss": 1.8728,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.8685083389282227,
      "learning_rate": 0.0015,
      "loss": 1.8522,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.7097916007041931,
      "learning_rate": 0.0015,
      "loss": 1.8522,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 1.032690405845642,
      "learning_rate": 0.0015,
      "loss": 1.8595,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.7951344847679138,
      "learning_rate": 0.0015,
      "loss": 1.8491,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.7099223732948303,
      "learning_rate": 0.0015,
      "loss": 1.8397,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 1.0079152584075928,
      "learning_rate": 0.0015,
      "loss": 1.846,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.8556044101715088,
      "learning_rate": 0.0015,
      "loss": 1.8343,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 1.0511488914489746,
      "learning_rate": 0.0015,
      "loss": 1.8349,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.8263677358627319,
      "learning_rate": 0.0015,
      "loss": 1.8243,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.8897006511688232,
      "learning_rate": 0.0015,
      "loss": 1.8161,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.9446095824241638,
      "learning_rate": 0.0015,
      "loss": 1.8166,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.9363498687744141,
      "learning_rate": 0.0015,
      "loss": 1.826,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.772242546081543,
      "learning_rate": 0.0015,
      "loss": 1.8104,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.6993220448493958,
      "learning_rate": 0.0015,
      "loss": 1.8151,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.6642560362815857,
      "learning_rate": 0.0015,
      "loss": 1.8094,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 1.2585757970809937,
      "learning_rate": 0.0015,
      "loss": 1.799,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 1.1124749183654785,
      "learning_rate": 0.0015,
      "loss": 1.8116,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.1095209121704102,
      "learning_rate": 0.0015,
      "loss": 1.7948,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 1.1040087938308716,
      "learning_rate": 0.0015,
      "loss": 1.8028,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 1.0847810506820679,
      "learning_rate": 0.0015,
      "loss": 1.792,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.7725903987884521,
      "learning_rate": 0.0015,
      "loss": 1.7757,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.7119668126106262,
      "learning_rate": 0.0015,
      "loss": 1.7796,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.9527880549430847,
      "learning_rate": 0.0015,
      "loss": 1.7949,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.7512865662574768,
      "learning_rate": 0.0015,
      "loss": 1.7845,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.6510151028633118,
      "learning_rate": 0.0015,
      "loss": 1.7674,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.7595269680023193,
      "learning_rate": 0.0015,
      "loss": 1.7614,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.7321306467056274,
      "learning_rate": 0.0015,
      "loss": 1.7768,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.966514527797699,
      "learning_rate": 0.0015,
      "loss": 1.7769,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 1.042886734008789,
      "learning_rate": 0.0015,
      "loss": 1.7592,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.7994891405105591,
      "learning_rate": 0.0015,
      "loss": 1.7587,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.8135955333709717,
      "learning_rate": 0.0015,
      "loss": 1.7721,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.8730813264846802,
      "learning_rate": 0.0015,
      "loss": 1.7634,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.6935446858406067,
      "learning_rate": 0.0015,
      "loss": 1.746,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.6731535792350769,
      "learning_rate": 0.0015,
      "loss": 1.7494,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.8686522841453552,
      "learning_rate": 0.0015,
      "loss": 1.7561,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 1.2903318405151367,
      "learning_rate": 0.0015,
      "loss": 1.7437,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.8495928049087524,
      "learning_rate": 0.0015,
      "loss": 1.7481,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.8161746263504028,
      "learning_rate": 0.0015,
      "loss": 1.7456,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.6318230032920837,
      "learning_rate": 0.0015,
      "loss": 1.7396,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.8698726296424866,
      "learning_rate": 0.0015,
      "loss": 1.739,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.6878219246864319,
      "learning_rate": 0.0015,
      "loss": 1.7293,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.7442454099655151,
      "learning_rate": 0.0015,
      "loss": 1.737,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.6869747638702393,
      "learning_rate": 0.0015,
      "loss": 1.7172,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.8279062509536743,
      "learning_rate": 0.0015,
      "loss": 1.7241,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.7597466111183167,
      "learning_rate": 0.0015,
      "loss": 1.7301,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.8928216695785522,
      "learning_rate": 0.0015,
      "loss": 1.7242,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.78310227394104,
      "learning_rate": 0.0015,
      "loss": 1.7211,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.6500014066696167,
      "learning_rate": 0.0015,
      "loss": 1.719,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.8307090401649475,
      "learning_rate": 0.0015,
      "loss": 1.7158,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 1.342738151550293,
      "learning_rate": 0.0015,
      "loss": 1.7165,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.9299972057342529,
      "learning_rate": 0.0015,
      "loss": 1.7232,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.765612781047821,
      "learning_rate": 0.0015,
      "loss": 1.7116,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.6986270546913147,
      "learning_rate": 0.0015,
      "loss": 1.7098,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.6947563290596008,
      "learning_rate": 0.0015,
      "loss": 1.7139,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.828080952167511,
      "learning_rate": 0.0015,
      "loss": 1.7154,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.6334972381591797,
      "learning_rate": 0.0015,
      "loss": 1.6957,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6875629425048828,
      "learning_rate": 0.0015,
      "loss": 1.6861,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.7145254015922546,
      "learning_rate": 0.0015,
      "loss": 1.6917,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.9437435865402222,
      "learning_rate": 0.0015,
      "loss": 1.6912,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.7296016216278076,
      "learning_rate": 0.0015,
      "loss": 1.6833,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.7277224063873291,
      "learning_rate": 0.0015,
      "loss": 1.6956,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.8755956292152405,
      "learning_rate": 0.0015,
      "loss": 1.6914,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.6696412563323975,
      "learning_rate": 0.0015,
      "loss": 1.6954,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.6554185748100281,
      "learning_rate": 0.0015,
      "loss": 1.6938,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6133327484130859,
      "learning_rate": 0.0015,
      "loss": 1.6614,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.6851116418838501,
      "learning_rate": 0.0015,
      "loss": 1.6729,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.7133874297142029,
      "learning_rate": 0.0015,
      "loss": 1.6805,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 1.1855857372283936,
      "learning_rate": 0.0015,
      "loss": 1.6711,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.657446563243866,
      "learning_rate": 0.0015,
      "loss": 1.6701,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.713283360004425,
      "learning_rate": 0.0015,
      "loss": 1.6781,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.720389723777771,
      "learning_rate": 0.0015,
      "loss": 1.668,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.9633141160011292,
      "learning_rate": 0.0015,
      "loss": 1.6713,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.6229538321495056,
      "learning_rate": 0.0015,
      "loss": 1.6784,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.6831251382827759,
      "learning_rate": 0.0015,
      "loss": 1.6523,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.6816312074661255,
      "learning_rate": 0.0015,
      "loss": 1.6506,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.6564023494720459,
      "learning_rate": 0.0015,
      "loss": 1.6548,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.6326689124107361,
      "learning_rate": 0.0015,
      "loss": 1.6613,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.6753053665161133,
      "learning_rate": 0.0015,
      "loss": 1.6806,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.7049668431282043,
      "learning_rate": 0.0015,
      "loss": 1.6607,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6844255328178406,
      "learning_rate": 0.0015,
      "loss": 1.6637,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.7776007652282715,
      "learning_rate": 0.0015,
      "loss": 1.6644,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.9076676964759827,
      "learning_rate": 0.0015,
      "loss": 1.6638,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.7703981399536133,
      "learning_rate": 0.0015,
      "loss": 1.6631,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.7967489361763,
      "learning_rate": 0.0015,
      "loss": 1.6481,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.8075294494628906,
      "learning_rate": 0.0015,
      "loss": 1.6481,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.7406700849533081,
      "learning_rate": 0.0015,
      "loss": 1.6483,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.6987505555152893,
      "learning_rate": 0.0015,
      "loss": 1.6433,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.6239149570465088,
      "learning_rate": 0.0015,
      "loss": 1.6484,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.6112042665481567,
      "learning_rate": 0.0015,
      "loss": 1.6498,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.8930774927139282,
      "learning_rate": 0.0015,
      "loss": 1.6475,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6910252571105957,
      "learning_rate": 0.0015,
      "loss": 1.6368,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.5995741486549377,
      "learning_rate": 0.0015,
      "loss": 1.6444,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 1.0854800939559937,
      "learning_rate": 0.0015,
      "loss": 1.6423,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 1.0601118803024292,
      "learning_rate": 0.0015,
      "loss": 1.6435,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.6452597379684448,
      "learning_rate": 0.0015,
      "loss": 1.6355,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 1.029815673828125,
      "learning_rate": 0.0015,
      "loss": 1.6405,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.6614994406700134,
      "learning_rate": 0.0015,
      "loss": 1.6287,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.7356704473495483,
      "learning_rate": 0.0015,
      "loss": 1.6283,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.7502595782279968,
      "learning_rate": 0.0015,
      "loss": 1.6318,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.6611550450325012,
      "learning_rate": 0.0015,
      "loss": 1.6434,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.9196215867996216,
      "learning_rate": 0.0015,
      "loss": 1.6258,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.6997500658035278,
      "learning_rate": 0.0015,
      "loss": 1.6235,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.6408447027206421,
      "learning_rate": 0.0015,
      "loss": 1.6185,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.810997486114502,
      "learning_rate": 0.0015,
      "loss": 1.6307,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.9732155799865723,
      "learning_rate": 0.0015,
      "loss": 1.6379,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.7736824154853821,
      "learning_rate": 0.0015,
      "loss": 1.6258,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6510221362113953,
      "learning_rate": 0.0015,
      "loss": 1.6236,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.6700575351715088,
      "learning_rate": 0.0015,
      "loss": 1.6242,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.6831074357032776,
      "learning_rate": 0.0015,
      "loss": 1.6154,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.8615200519561768,
      "learning_rate": 0.0015,
      "loss": 1.6176,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.8389783501625061,
      "learning_rate": 0.0015,
      "loss": 1.6262,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.7485233545303345,
      "learning_rate": 0.0015,
      "loss": 1.6142,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.7977454662322998,
      "learning_rate": 0.0015,
      "loss": 1.6177,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 1.111137866973877,
      "learning_rate": 0.0015,
      "loss": 1.6185,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6833589673042297,
      "learning_rate": 0.0015,
      "loss": 1.6099,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.6046237349510193,
      "learning_rate": 0.0015,
      "loss": 1.6052,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.570578396320343,
      "learning_rate": 0.0015,
      "loss": 1.5999,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.818230926990509,
      "learning_rate": 0.0015,
      "loss": 1.6098,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.019039273262024,
      "learning_rate": 0.0015,
      "loss": 1.624,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.8784221410751343,
      "learning_rate": 0.0015,
      "loss": 1.6072,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.6730207800865173,
      "learning_rate": 0.0015,
      "loss": 1.5922,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.9545859098434448,
      "learning_rate": 0.0015,
      "loss": 1.608,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.7341895699501038,
      "learning_rate": 0.0015,
      "loss": 1.6106,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.7155537009239197,
      "learning_rate": 0.0015,
      "loss": 1.5964,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.5885248780250549,
      "learning_rate": 0.0015,
      "loss": 1.6086,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.9014438390731812,
      "learning_rate": 0.0015,
      "loss": 1.6023,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.6999181509017944,
      "learning_rate": 0.0015,
      "loss": 1.5993,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.8914079666137695,
      "learning_rate": 0.0015,
      "loss": 1.5983,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.881064236164093,
      "learning_rate": 0.0015,
      "loss": 1.5878,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.7987886667251587,
      "learning_rate": 0.0015,
      "loss": 1.6197,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.6327575445175171,
      "learning_rate": 0.0015,
      "loss": 1.6093,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.6157426238059998,
      "learning_rate": 0.0015,
      "loss": 1.5916,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.6374841332435608,
      "learning_rate": 0.0015,
      "loss": 1.588,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.6461164355278015,
      "learning_rate": 0.0015,
      "loss": 1.5885,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.6031455397605896,
      "learning_rate": 0.0015,
      "loss": 1.6012,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.5920032858848572,
      "learning_rate": 0.0015,
      "loss": 1.5845,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.8245423436164856,
      "learning_rate": 0.0015,
      "loss": 1.5811,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.682399332523346,
      "learning_rate": 0.0015,
      "loss": 1.5841,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.809495210647583,
      "learning_rate": 0.0015,
      "loss": 1.5943,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.9487093687057495,
      "learning_rate": 0.0015,
      "loss": 1.5883,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.6348952651023865,
      "learning_rate": 0.0015,
      "loss": 1.5879,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.5969370603561401,
      "learning_rate": 0.0015,
      "loss": 1.5845,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6281625628471375,
      "learning_rate": 0.0015,
      "loss": 1.583,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.7266300916671753,
      "learning_rate": 0.0015,
      "loss": 1.5872,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.7617605924606323,
      "learning_rate": 0.0015,
      "loss": 1.5815,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.796266496181488,
      "learning_rate": 0.0015,
      "loss": 1.5806,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.7584456205368042,
      "learning_rate": 0.0015,
      "loss": 1.5909,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.5680705308914185,
      "learning_rate": 0.0015,
      "loss": 1.5747,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.5837324857711792,
      "learning_rate": 0.0015,
      "loss": 1.5897,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 1.034664511680603,
      "learning_rate": 0.0015,
      "loss": 1.5813,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.7929449081420898,
      "learning_rate": 0.0015,
      "loss": 1.5845,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.6136671304702759,
      "learning_rate": 0.0015,
      "loss": 1.5717,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.6934119462966919,
      "learning_rate": 0.0015,
      "loss": 1.5739,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.6323077082633972,
      "learning_rate": 0.0015,
      "loss": 1.5755,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.6266193389892578,
      "learning_rate": 0.0015,
      "loss": 1.5794,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.7616978287696838,
      "learning_rate": 0.0015,
      "loss": 1.5733,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.9609237909317017,
      "learning_rate": 0.0015,
      "loss": 1.5741,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.8360873460769653,
      "learning_rate": 0.0015,
      "loss": 1.5716,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.5792592763900757,
      "learning_rate": 0.0015,
      "loss": 1.5771,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.664145290851593,
      "learning_rate": 0.0015,
      "loss": 1.5731,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.696051836013794,
      "learning_rate": 0.0015,
      "loss": 1.5715,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.6626716256141663,
      "learning_rate": 0.0015,
      "loss": 1.5777,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.6069105267524719,
      "learning_rate": 0.0015,
      "loss": 1.5752,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6481751799583435,
      "learning_rate": 0.0015,
      "loss": 1.5739,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6787488460540771,
      "learning_rate": 0.0015,
      "loss": 1.5742,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.7607318758964539,
      "learning_rate": 0.0015,
      "loss": 1.5593,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.6619844436645508,
      "learning_rate": 0.0015,
      "loss": 1.5746,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.7582201957702637,
      "learning_rate": 0.0015,
      "loss": 1.5663,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.6550025343894958,
      "learning_rate": 0.0015,
      "loss": 1.5612,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.6441704630851746,
      "learning_rate": 0.0015,
      "loss": 1.5659,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.5557734370231628,
      "learning_rate": 0.0015,
      "loss": 1.5623,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.7199389934539795,
      "learning_rate": 0.0015,
      "loss": 1.5612,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6285843253135681,
      "learning_rate": 0.0015,
      "loss": 1.5571,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.7301799058914185,
      "learning_rate": 0.0015,
      "loss": 1.5653,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.5957171320915222,
      "learning_rate": 0.0015,
      "loss": 1.5657,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.8285066485404968,
      "learning_rate": 0.0015,
      "loss": 1.5462,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.7089322209358215,
      "learning_rate": 0.0015,
      "loss": 1.5568,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.5445607900619507,
      "learning_rate": 0.0015,
      "loss": 1.5578,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.9472262263298035,
      "learning_rate": 0.0015,
      "loss": 1.5673,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6500675678253174,
      "learning_rate": 0.0015,
      "loss": 1.5644,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.6101513504981995,
      "learning_rate": 0.0015,
      "loss": 1.5626,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.6570156812667847,
      "learning_rate": 0.0015,
      "loss": 1.5569,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.6476898789405823,
      "learning_rate": 0.0015,
      "loss": 1.5576,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.6236690878868103,
      "learning_rate": 0.0015,
      "loss": 1.557,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6243173480033875,
      "learning_rate": 0.0015,
      "loss": 1.5643,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 1.0588493347167969,
      "learning_rate": 0.0015,
      "loss": 1.5406,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.7831001281738281,
      "learning_rate": 0.0015,
      "loss": 1.5432,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.5851334929466248,
      "learning_rate": 0.0015,
      "loss": 1.5466,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.5993044972419739,
      "learning_rate": 0.0015,
      "loss": 1.5527,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.7337188124656677,
      "learning_rate": 0.0015,
      "loss": 1.554,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.6788395047187805,
      "learning_rate": 0.0015,
      "loss": 1.5374,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.7346245050430298,
      "learning_rate": 0.0015,
      "loss": 1.5409,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.6191657185554504,
      "learning_rate": 0.0015,
      "loss": 1.5497,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.7453780770301819,
      "learning_rate": 0.0015,
      "loss": 1.552,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.6672173142433167,
      "learning_rate": 0.0015,
      "loss": 1.5467,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.5964860320091248,
      "learning_rate": 0.0015,
      "loss": 1.5405,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.6797311305999756,
      "learning_rate": 0.0015,
      "loss": 1.5484,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 1.1362419128417969,
      "learning_rate": 0.0015,
      "loss": 1.5563,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.8725783228874207,
      "learning_rate": 0.0015,
      "loss": 1.5467,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.7338547706604004,
      "learning_rate": 0.0015,
      "loss": 1.5271,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.6057443618774414,
      "learning_rate": 0.0015,
      "loss": 1.535,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.5910390615463257,
      "learning_rate": 0.0015,
      "loss": 1.5496,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.6153573989868164,
      "learning_rate": 0.0015,
      "loss": 1.5312,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.7339136600494385,
      "learning_rate": 0.0015,
      "loss": 1.5414,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.7159757614135742,
      "learning_rate": 0.0015,
      "loss": 1.5436,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6123351454734802,
      "learning_rate": 0.0015,
      "loss": 1.531,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.7141717672348022,
      "learning_rate": 0.0015,
      "loss": 1.5395,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.5910229086875916,
      "learning_rate": 0.0015,
      "loss": 1.5341,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.5639906525611877,
      "learning_rate": 0.0015,
      "loss": 1.5471,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.541513204574585,
      "learning_rate": 0.0015,
      "loss": 1.5344,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.6807689666748047,
      "learning_rate": 0.0015,
      "loss": 1.5269,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.5676193237304688,
      "learning_rate": 0.0015,
      "loss": 1.5419,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.7307770252227783,
      "learning_rate": 0.0015,
      "loss": 1.5291,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6474248766899109,
      "learning_rate": 0.0015,
      "loss": 1.5309,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.6755460500717163,
      "learning_rate": 0.0015,
      "loss": 1.5324,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.6408239006996155,
      "learning_rate": 0.0015,
      "loss": 1.5466,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.6348702311515808,
      "learning_rate": 0.0015,
      "loss": 1.5242,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5707957744598389,
      "learning_rate": 0.0015,
      "loss": 1.5305,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5700860619544983,
      "learning_rate": 0.0015,
      "loss": 1.5326,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.858945369720459,
      "learning_rate": 0.0015,
      "loss": 1.5372,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.7147475481033325,
      "learning_rate": 0.0015,
      "loss": 1.5306,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.8068617582321167,
      "learning_rate": 0.0015,
      "loss": 1.5296,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.6645533442497253,
      "learning_rate": 0.0015,
      "loss": 1.5222,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.589341938495636,
      "learning_rate": 0.0015,
      "loss": 1.5339,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.6254124641418457,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.5573468804359436,
      "learning_rate": 0.0015,
      "loss": 1.5223,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.5843837261199951,
      "learning_rate": 0.0015,
      "loss": 1.5024,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.6093689799308777,
      "learning_rate": 0.0015,
      "loss": 1.527,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.5487991571426392,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.5893562436103821,
      "learning_rate": 0.0015,
      "loss": 1.5365,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.7497988939285278,
      "learning_rate": 0.0015,
      "loss": 1.5355,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.8500696420669556,
      "learning_rate": 0.0015,
      "loss": 1.5176,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.6716281175613403,
      "learning_rate": 0.0015,
      "loss": 1.528,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.5595434904098511,
      "learning_rate": 0.0015,
      "loss": 1.5275,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.6020368933677673,
      "learning_rate": 0.0015,
      "loss": 1.5317,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.5574306845664978,
      "learning_rate": 0.0015,
      "loss": 1.5199,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.5826432108879089,
      "learning_rate": 0.0015,
      "loss": 1.522,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6289241909980774,
      "learning_rate": 0.0015,
      "loss": 1.5146,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.7313932180404663,
      "learning_rate": 0.0015,
      "loss": 1.5181,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.627030074596405,
      "learning_rate": 0.0015,
      "loss": 1.5224,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.5869846940040588,
      "learning_rate": 0.0015,
      "loss": 1.52,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.5843411087989807,
      "learning_rate": 0.0015,
      "loss": 1.5151,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.663027822971344,
      "learning_rate": 0.0015,
      "loss": 1.5233,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.5814419388771057,
      "learning_rate": 0.0015,
      "loss": 1.5199,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.8384828567504883,
      "learning_rate": 0.0015,
      "loss": 1.5013,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.5617758631706238,
      "learning_rate": 0.0015,
      "loss": 1.5179,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.7915652394294739,
      "learning_rate": 0.0015,
      "loss": 1.5136,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.636605441570282,
      "learning_rate": 0.0015,
      "loss": 1.5136,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.8398627638816833,
      "learning_rate": 0.0015,
      "loss": 1.5058,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.6504268050193787,
      "learning_rate": 0.0015,
      "loss": 1.5224,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.6623251438140869,
      "learning_rate": 0.0015,
      "loss": 1.5269,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.5744650363922119,
      "learning_rate": 0.0015,
      "loss": 1.5196,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.6193293333053589,
      "learning_rate": 0.0015,
      "loss": 1.4978,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.6639376878738403,
      "learning_rate": 0.0015,
      "loss": 1.5208,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.6582226157188416,
      "learning_rate": 0.0015,
      "loss": 1.5165,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.6584490537643433,
      "learning_rate": 0.0015,
      "loss": 1.5286,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 1.008704662322998,
      "learning_rate": 0.0015,
      "loss": 1.5216,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.6590273380279541,
      "learning_rate": 0.0015,
      "loss": 1.5095,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.618530809879303,
      "learning_rate": 0.0015,
      "loss": 1.5098,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.6145744919776917,
      "learning_rate": 0.0015,
      "loss": 1.5062,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.6485543251037598,
      "learning_rate": 0.0015,
      "loss": 1.506,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.661384642124176,
      "learning_rate": 0.0015,
      "loss": 1.5181,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.7174956202507019,
      "learning_rate": 0.0015,
      "loss": 1.5004,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.6093496680259705,
      "learning_rate": 0.0015,
      "loss": 1.5045,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.6092379093170166,
      "learning_rate": 0.0015,
      "loss": 1.5073,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.6164376139640808,
      "learning_rate": 0.0015,
      "loss": 1.5077,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.8719568848609924,
      "learning_rate": 0.0015,
      "loss": 1.5079,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.7941826581954956,
      "learning_rate": 0.0015,
      "loss": 1.5177,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.7229615449905396,
      "learning_rate": 0.0015,
      "loss": 1.5096,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.5858650803565979,
      "learning_rate": 0.0015,
      "loss": 1.5015,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.8932380080223083,
      "learning_rate": 0.0015,
      "loss": 1.5024,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.7861613035202026,
      "learning_rate": 0.0015,
      "loss": 1.5146,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.9015659689903259,
      "learning_rate": 0.0015,
      "loss": 1.5205,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.6179216504096985,
      "learning_rate": 0.0015,
      "loss": 1.5076,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.6513644456863403,
      "learning_rate": 0.0015,
      "loss": 1.4843,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.556451678276062,
      "learning_rate": 0.0015,
      "loss": 1.503,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.5759680271148682,
      "learning_rate": 0.0015,
      "loss": 1.5028,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.8100026249885559,
      "learning_rate": 0.0015,
      "loss": 1.5084,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.9177446961402893,
      "learning_rate": 0.0015,
      "loss": 1.499,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.6613394618034363,
      "learning_rate": 0.0015,
      "loss": 1.5096,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.5650438666343689,
      "learning_rate": 0.0015,
      "loss": 1.4972,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.8402285575866699,
      "learning_rate": 0.0015,
      "loss": 1.5026,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.6960042715072632,
      "learning_rate": 0.0015,
      "loss": 1.5076,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.6291213035583496,
      "learning_rate": 0.0015,
      "loss": 1.4961,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.594775378704071,
      "learning_rate": 0.0015,
      "loss": 1.49,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.6108851432800293,
      "learning_rate": 0.0015,
      "loss": 1.4904,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.6478801965713501,
      "learning_rate": 0.0015,
      "loss": 1.5105,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.6294592618942261,
      "learning_rate": 0.0015,
      "loss": 1.4909,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.6587184071540833,
      "learning_rate": 0.0015,
      "loss": 1.502,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.580014169216156,
      "learning_rate": 0.0015,
      "loss": 1.4932,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.5819181799888611,
      "learning_rate": 0.0015,
      "loss": 1.4871,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.7391009330749512,
      "learning_rate": 0.0015,
      "loss": 1.4862,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.7004148960113525,
      "learning_rate": 0.0015,
      "loss": 1.4847,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.5804836750030518,
      "learning_rate": 0.0015,
      "loss": 1.4891,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.6002885699272156,
      "learning_rate": 0.0015,
      "loss": 1.4857,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.6326174139976501,
      "learning_rate": 0.0015,
      "loss": 1.491,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.9303730726242065,
      "learning_rate": 0.0015,
      "loss": 1.4891,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.7025170922279358,
      "learning_rate": 0.0015,
      "loss": 1.4961,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.6459818482398987,
      "learning_rate": 0.0015,
      "loss": 1.4954,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.5743758082389832,
      "learning_rate": 0.0015,
      "loss": 1.4829,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.6225879192352295,
      "learning_rate": 0.0015,
      "loss": 1.4954,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.707120418548584,
      "learning_rate": 0.0015,
      "loss": 1.4878,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.6311498880386353,
      "learning_rate": 0.0015,
      "loss": 1.4981,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.6863111257553101,
      "learning_rate": 0.0015,
      "loss": 1.4839,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.5685213804244995,
      "learning_rate": 0.0015,
      "loss": 1.4901,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.5521535873413086,
      "learning_rate": 0.0015,
      "loss": 1.483,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5836614966392517,
      "learning_rate": 0.0015,
      "loss": 1.4922,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5536697506904602,
      "learning_rate": 0.0015,
      "loss": 1.4923,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.8817647099494934,
      "learning_rate": 0.0015,
      "loss": 1.4953,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.7525816559791565,
      "learning_rate": 0.0015,
      "loss": 1.4896,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6736322641372681,
      "learning_rate": 0.0015,
      "loss": 1.4965,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.7411642670631409,
      "learning_rate": 0.0015,
      "loss": 1.491,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5742962956428528,
      "learning_rate": 0.0015,
      "loss": 1.4894,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.6649968028068542,
      "learning_rate": 0.0015,
      "loss": 1.4824,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.750732958316803,
      "learning_rate": 0.0015,
      "loss": 1.4861,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.6525141596794128,
      "learning_rate": 0.0015,
      "loss": 1.477,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5441020131111145,
      "learning_rate": 0.0015,
      "loss": 1.481,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.6645817756652832,
      "learning_rate": 0.0015,
      "loss": 1.4856,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.6790419220924377,
      "learning_rate": 0.0015,
      "loss": 1.4796,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.6181941628456116,
      "learning_rate": 0.0015,
      "loss": 1.4811,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.5693588256835938,
      "learning_rate": 0.0015,
      "loss": 1.4799,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.6952799558639526,
      "learning_rate": 0.0015,
      "loss": 1.487,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 1.1466480493545532,
      "learning_rate": 0.0015,
      "loss": 1.4913,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.6028270125389099,
      "learning_rate": 0.0015,
      "loss": 1.4765,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.5977668762207031,
      "learning_rate": 0.0015,
      "loss": 1.481,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 1.0622198581695557,
      "learning_rate": 0.0015,
      "loss": 1.479,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.6696746945381165,
      "learning_rate": 0.0015,
      "loss": 1.4798,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.6021679639816284,
      "learning_rate": 0.0015,
      "loss": 1.4834,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.791585385799408,
      "learning_rate": 0.0015,
      "loss": 1.4906,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5709472298622131,
      "learning_rate": 0.0015,
      "loss": 1.4771,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.7294502854347229,
      "learning_rate": 0.0015,
      "loss": 1.4726,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.6758453249931335,
      "learning_rate": 0.0015,
      "loss": 1.4831,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.656216025352478,
      "learning_rate": 0.0015,
      "loss": 1.4775,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.5702846646308899,
      "learning_rate": 0.0015,
      "loss": 1.4768,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.8870499730110168,
      "learning_rate": 0.0015,
      "loss": 1.4748,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.5933520197868347,
      "learning_rate": 0.0015,
      "loss": 1.4825,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.6596553921699524,
      "learning_rate": 0.0015,
      "loss": 1.4745,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.6646473407745361,
      "learning_rate": 0.0015,
      "loss": 1.4762,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.692927360534668,
      "learning_rate": 0.0015,
      "loss": 1.4919,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.6025608777999878,
      "learning_rate": 0.0015,
      "loss": 1.4681,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5913830995559692,
      "learning_rate": 0.0015,
      "loss": 1.4751,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.6132447123527527,
      "learning_rate": 0.0015,
      "loss": 1.4704,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.6659219861030579,
      "learning_rate": 0.0015,
      "loss": 1.4779,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.575726330280304,
      "learning_rate": 0.0015,
      "loss": 1.4908,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.6139013767242432,
      "learning_rate": 0.0015,
      "loss": 1.4732,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.5637103915214539,
      "learning_rate": 0.0015,
      "loss": 1.4783,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.6683405637741089,
      "learning_rate": 0.0015,
      "loss": 1.4752,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.6898501515388489,
      "learning_rate": 0.0015,
      "loss": 1.4806,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5260871052742004,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.6019372940063477,
      "learning_rate": 0.0015,
      "loss": 1.4762,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.6699599027633667,
      "learning_rate": 0.0015,
      "loss": 1.4825,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.6201009154319763,
      "learning_rate": 0.0015,
      "loss": 1.47,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.6440523862838745,
      "learning_rate": 0.0015,
      "loss": 1.4701,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5590440034866333,
      "learning_rate": 0.0015,
      "loss": 1.4759,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.7142961025238037,
      "learning_rate": 0.0015,
      "loss": 1.4569,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.7675514221191406,
      "learning_rate": 0.0015,
      "loss": 1.4769,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.7219570279121399,
      "learning_rate": 0.0015,
      "loss": 1.4793,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.6125030517578125,
      "learning_rate": 0.0015,
      "loss": 1.4745,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.685753583908081,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.5657294988632202,
      "learning_rate": 0.0015,
      "loss": 1.4494,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.6757033467292786,
      "learning_rate": 0.0015,
      "loss": 1.4729,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.6861257553100586,
      "learning_rate": 0.0015,
      "loss": 1.4573,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.5530158877372742,
      "learning_rate": 0.0015,
      "loss": 1.477,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 1.01388418674469,
      "learning_rate": 0.0015,
      "loss": 1.4757,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.5928830504417419,
      "learning_rate": 0.0015,
      "loss": 1.4562,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.5853449702262878,
      "learning_rate": 0.0015,
      "loss": 1.4722,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.6117337942123413,
      "learning_rate": 0.0015,
      "loss": 1.4749,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5882995128631592,
      "learning_rate": 0.0015,
      "loss": 1.4771,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.5885705947875977,
      "learning_rate": 0.0015,
      "loss": 1.4658,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.6743139028549194,
      "learning_rate": 0.0015,
      "loss": 1.4628,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.589303731918335,
      "learning_rate": 0.0015,
      "loss": 1.46,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.8263179063796997,
      "learning_rate": 0.0015,
      "loss": 1.4582,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.5357639193534851,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.6617996096611023,
      "learning_rate": 0.0015,
      "loss": 1.4679,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.7992064952850342,
      "learning_rate": 0.0015,
      "loss": 1.465,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.7020431160926819,
      "learning_rate": 0.0015,
      "loss": 1.4594,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7185554504394531,
      "learning_rate": 0.0015,
      "loss": 1.461,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.8029890656471252,
      "learning_rate": 0.0015,
      "loss": 1.4789,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5433769822120667,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.5386573076248169,
      "learning_rate": 0.0015,
      "loss": 1.4626,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5510920882225037,
      "learning_rate": 0.0015,
      "loss": 1.4749,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.5974557995796204,
      "learning_rate": 0.0015,
      "loss": 1.4547,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5800578594207764,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.5410987734794617,
      "learning_rate": 0.0015,
      "loss": 1.4619,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.6204303503036499,
      "learning_rate": 0.0015,
      "loss": 1.4797,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.8674985766410828,
      "learning_rate": 0.0015,
      "loss": 1.4626,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.606118381023407,
      "learning_rate": 0.0015,
      "loss": 1.4565,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.5624986290931702,
      "learning_rate": 0.0015,
      "loss": 1.468,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.8207271099090576,
      "learning_rate": 0.0015,
      "loss": 1.4569,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.793711245059967,
      "learning_rate": 0.0015,
      "loss": 1.4515,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.5525781512260437,
      "learning_rate": 0.0015,
      "loss": 1.4734,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.5403020977973938,
      "learning_rate": 0.0015,
      "loss": 1.4593,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.554365336894989,
      "learning_rate": 0.0015,
      "loss": 1.4685,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.607978343963623,
      "learning_rate": 0.0015,
      "loss": 1.464,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.5707883834838867,
      "learning_rate": 0.0015,
      "loss": 1.4521,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.6489737033843994,
      "learning_rate": 0.0015,
      "loss": 1.4624,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.694725751876831,
      "learning_rate": 0.0015,
      "loss": 1.4613,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.9797824621200562,
      "learning_rate": 0.0015,
      "loss": 1.4675,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.7563436031341553,
      "learning_rate": 0.0015,
      "loss": 1.4501,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.5978177189826965,
      "learning_rate": 0.0015,
      "loss": 1.4559,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.5603195428848267,
      "learning_rate": 0.0015,
      "loss": 1.4511,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.5828781127929688,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.6030756831169128,
      "learning_rate": 0.0015,
      "loss": 1.4652,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.5839309692382812,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.648682713508606,
      "learning_rate": 0.0015,
      "loss": 1.4495,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.5189370512962341,
      "learning_rate": 0.0015,
      "loss": 1.437,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.638726532459259,
      "learning_rate": 0.0015,
      "loss": 1.4656,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.5538253784179688,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.6002002358436584,
      "learning_rate": 0.0015,
      "loss": 1.4666,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.6380165219306946,
      "learning_rate": 0.0015,
      "loss": 1.4532,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.575442373752594,
      "learning_rate": 0.0015,
      "loss": 1.4684,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.5849810242652893,
      "learning_rate": 0.0015,
      "loss": 1.4609,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.6542881727218628,
      "learning_rate": 0.0015,
      "loss": 1.4416,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.5681846141815186,
      "learning_rate": 0.0015,
      "loss": 1.4558,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.8104299902915955,
      "learning_rate": 0.0015,
      "loss": 1.453,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.6678333282470703,
      "learning_rate": 0.0015,
      "loss": 1.4516,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.8156563639640808,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.5556895136833191,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.5274465680122375,
      "learning_rate": 0.0015,
      "loss": 1.4531,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5245038866996765,
      "learning_rate": 0.0015,
      "loss": 1.4543,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.6143195033073425,
      "learning_rate": 0.0015,
      "loss": 1.4643,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.5971368551254272,
      "learning_rate": 0.0015,
      "loss": 1.4463,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.6224199533462524,
      "learning_rate": 0.0015,
      "loss": 1.4607,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.7090758085250854,
      "learning_rate": 0.0015,
      "loss": 1.4444,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.6592482924461365,
      "learning_rate": 0.0015,
      "loss": 1.4545,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.523442804813385,
      "learning_rate": 0.0015,
      "loss": 1.4475,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.5295791029930115,
      "learning_rate": 0.0015,
      "loss": 1.4549,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.6829699277877808,
      "learning_rate": 0.0015,
      "loss": 1.4462,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.5284391641616821,
      "learning_rate": 0.0015,
      "loss": 1.4536,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.6125273704528809,
      "learning_rate": 0.0015,
      "loss": 1.4591,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.6987561583518982,
      "learning_rate": 0.0015,
      "loss": 1.4474,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.6083083152770996,
      "learning_rate": 0.0015,
      "loss": 1.4521,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.6013514995574951,
      "learning_rate": 0.0015,
      "loss": 1.4563,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 1.1169664859771729,
      "learning_rate": 0.0015,
      "loss": 1.4472,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.6246483325958252,
      "learning_rate": 0.0015,
      "loss": 1.4431,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.7970275282859802,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.5732747316360474,
      "learning_rate": 0.0015,
      "loss": 1.4453,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.6154385805130005,
      "learning_rate": 0.0015,
      "loss": 1.4423,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.6107851266860962,
      "learning_rate": 0.0015,
      "loss": 1.4492,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.553439199924469,
      "learning_rate": 0.0015,
      "loss": 1.4508,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.5437508225440979,
      "learning_rate": 0.0015,
      "loss": 1.4418,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.7134219408035278,
      "learning_rate": 0.0015,
      "loss": 1.4422,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.6530733704566956,
      "learning_rate": 0.0015,
      "loss": 1.4437,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5411928296089172,
      "learning_rate": 0.0015,
      "loss": 1.4451,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.581093966960907,
      "learning_rate": 0.0015,
      "loss": 1.4412,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.8505716919898987,
      "learning_rate": 0.0015,
      "loss": 1.4397,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.5590124130249023,
      "learning_rate": 0.0015,
      "loss": 1.4353,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.6479218006134033,
      "learning_rate": 0.0015,
      "loss": 1.4437,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.6174955368041992,
      "learning_rate": 0.0015,
      "loss": 1.4461,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.6192970275878906,
      "learning_rate": 0.0015,
      "loss": 1.4467,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.6283924579620361,
      "learning_rate": 0.0015,
      "loss": 1.4495,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.5479466319084167,
      "learning_rate": 0.0015,
      "loss": 1.4523,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.6062834858894348,
      "learning_rate": 0.0015,
      "loss": 1.4431,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.7791538834571838,
      "learning_rate": 0.0015,
      "loss": 1.4492,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.5234934091567993,
      "learning_rate": 0.0015,
      "loss": 1.4403,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.7938951849937439,
      "learning_rate": 0.0015,
      "loss": 1.4334,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.6442159414291382,
      "learning_rate": 0.0015,
      "loss": 1.4362,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.5595635771751404,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.5835129022598267,
      "learning_rate": 0.0015,
      "loss": 1.4547,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.5161041021347046,
      "learning_rate": 0.0015,
      "loss": 1.4412,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.5372098088264465,
      "learning_rate": 0.0015,
      "loss": 1.4383,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.5379180908203125,
      "learning_rate": 0.0015,
      "loss": 1.443,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5753316283226013,
      "learning_rate": 0.0015,
      "loss": 1.4377,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.606822669506073,
      "learning_rate": 0.0015,
      "loss": 1.4384,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.534155547618866,
      "learning_rate": 0.0015,
      "loss": 1.4504,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.6653630137443542,
      "learning_rate": 0.0015,
      "loss": 1.445,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.572662889957428,
      "learning_rate": 0.0015,
      "loss": 1.4561,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.5921775698661804,
      "learning_rate": 0.0015,
      "loss": 1.446,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.5696536302566528,
      "learning_rate": 0.0015,
      "loss": 1.4441,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.6740322709083557,
      "learning_rate": 0.0015,
      "loss": 1.4485,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5913907885551453,
      "learning_rate": 0.0015,
      "loss": 1.4505,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.7871809601783752,
      "learning_rate": 0.0015,
      "loss": 1.4299,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.7495800256729126,
      "learning_rate": 0.0015,
      "loss": 1.4374,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.7116259932518005,
      "learning_rate": 0.0015,
      "loss": 1.4395,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5654744505882263,
      "learning_rate": 0.0015,
      "loss": 1.4436,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.6537961363792419,
      "learning_rate": 0.0015,
      "loss": 1.4395,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.5508219599723816,
      "learning_rate": 0.0015,
      "loss": 1.4392,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.748004138469696,
      "learning_rate": 0.0015,
      "loss": 1.4144,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5527656674385071,
      "learning_rate": 0.0015,
      "loss": 1.4259,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.6227961182594299,
      "learning_rate": 0.0015,
      "loss": 1.4259,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.5234470367431641,
      "learning_rate": 0.0015,
      "loss": 1.4345,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.6069667935371399,
      "learning_rate": 0.0015,
      "loss": 1.429,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.5087295174598694,
      "learning_rate": 0.0015,
      "loss": 1.4345,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5701834559440613,
      "learning_rate": 0.0015,
      "loss": 1.4413,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5617038011550903,
      "learning_rate": 0.0015,
      "loss": 1.4505,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.599163830280304,
      "learning_rate": 0.0015,
      "loss": 1.4341,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.816677987575531,
      "learning_rate": 0.0015,
      "loss": 1.4489,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.8053539991378784,
      "learning_rate": 0.0015,
      "loss": 1.455,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.5908822417259216,
      "learning_rate": 0.0015,
      "loss": 1.4392,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.5969804525375366,
      "learning_rate": 0.0015,
      "loss": 1.431,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.6755075454711914,
      "learning_rate": 0.0015,
      "loss": 1.4411,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5911203622817993,
      "learning_rate": 0.0015,
      "loss": 1.4341,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.5493491888046265,
      "learning_rate": 0.0015,
      "loss": 1.4332,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.6947777271270752,
      "learning_rate": 0.0015,
      "loss": 1.429,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.5623560547828674,
      "learning_rate": 0.0015,
      "loss": 1.4257,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.9323939085006714,
      "learning_rate": 0.0015,
      "loss": 1.4341,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.6801377534866333,
      "learning_rate": 0.0015,
      "loss": 1.4411,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.5711943507194519,
      "learning_rate": 0.0015,
      "loss": 1.4261,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.528338611125946,
      "learning_rate": 0.0015,
      "loss": 1.4348,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5535573959350586,
      "learning_rate": 0.0015,
      "loss": 1.435,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.5801048874855042,
      "learning_rate": 0.0015,
      "loss": 1.4207,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.6147446036338806,
      "learning_rate": 0.0015,
      "loss": 1.4389,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.541782796382904,
      "learning_rate": 0.0015,
      "loss": 1.4462,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.6918947100639343,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.6587607264518738,
      "learning_rate": 0.0015,
      "loss": 1.4354,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.5851778984069824,
      "learning_rate": 0.0015,
      "loss": 1.4185,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.6019512414932251,
      "learning_rate": 0.0015,
      "loss": 1.4399,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.5221972465515137,
      "learning_rate": 0.0015,
      "loss": 1.4197,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.5839863419532776,
      "learning_rate": 0.0015,
      "loss": 1.4378,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.6848641037940979,
      "learning_rate": 0.0015,
      "loss": 1.429,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.528335452079773,
      "learning_rate": 0.0015,
      "loss": 1.4277,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5361722707748413,
      "learning_rate": 0.0015,
      "loss": 1.4319,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.46983981132507324,
      "learning_rate": 0.0015,
      "loss": 1.4234,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.5124049186706543,
      "learning_rate": 0.0015,
      "loss": 1.432,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.5647552013397217,
      "learning_rate": 0.0015,
      "loss": 1.4308,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.6874547600746155,
      "learning_rate": 0.0015,
      "loss": 1.4288,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": Infinity,
      "learning_rate": 0.0015,
      "loss": 1.4285,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.574679434299469,
      "learning_rate": 0.0015,
      "loss": 1.4383,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.5606120824813843,
      "learning_rate": 0.0015,
      "loss": 1.4327,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5364856123924255,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.4113,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.5553877353668213,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.4309,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5766550898551941,
      "learning_rate": 0.001444828815847542,
      "loss": 1.435,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.5754528641700745,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.426,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5499405860900879,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.4197,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.5255405902862549,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4269,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.7719149589538574,
      "learning_rate": 0.001366801618799797,
      "loss": 1.4161,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.6151224374771118,
      "learning_rate": 0.001347962334099052,
      "loss": 1.4175,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.5480755567550659,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4252,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.6165157556533813,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4144,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.5327451825141907,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.4212,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5260996222496033,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.4175,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.5738957524299622,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.412,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.6758652925491333,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4071,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.6473555564880371,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4041,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5306556820869446,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.4009,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.664206326007843,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4086,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.7255275249481201,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4025,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.5972916483879089,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.405,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.6715484261512756,
      "learning_rate": 0.001141156021915355,
      "loss": 1.4034,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.4935395121574402,
      "learning_rate": 0.001125426919103997,
      "loss": 1.3856,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.5832482576370239,
      "learning_rate": 0.001109914618088799,
      "loss": 1.3884,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.6867726445198059,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.3999,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.5707427859306335,
      "learning_rate": 0.001079528509501728,
      "loss": 1.3994,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.5250033140182495,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.395,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.5501655340194702,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.3825,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.538463294506073,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.3894,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.5902817845344543,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3756,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5150527358055115,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.3789,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.5589239001274109,
      "learning_rate": 0.000993270988848579,
      "loss": 1.3902,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5444999933242798,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.3853,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.6948215961456299,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.3916,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.5912929773330688,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.3854,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.49268436431884766,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.3981,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.5165446400642395,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.3994,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.47695085406303406,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.3789,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.6241263747215271,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3711,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.5821035504341125,
      "learning_rate": 0.000888885722567627,
      "loss": 1.3828,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.524462878704071,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.3747,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.6472446918487549,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3753,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.5896880626678467,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.3723,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.5659841895103455,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3682,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.6180821657180786,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3714,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5561965703964233,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3779,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.5118362307548523,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.3631,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.4638024568557739,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3692,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.5717614889144897,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.3629,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.6440813541412354,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.3662,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.5368211269378662,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3602,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.5129225850105286,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3607,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5165268778800964,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.3625,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.5030327439308167,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.3635,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.5081734657287598,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.3568,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.4782288670539856,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.3568,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.5021098852157593,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.3591,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.6094057559967041,
      "learning_rate": 0.000692383699170611,
      "loss": 1.3481,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.5212547779083252,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3589,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.5429904460906982,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3594,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.4989031255245209,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.361,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.6769903898239136,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3554,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.5959861874580383,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3562,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.482819527387619,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3548,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.5496542453765869,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.3596,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.5129460096359253,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3433,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.6600692868232727,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3477,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.6072731614112854,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.3504,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.4981010854244232,
      "learning_rate": 0.000594349404202577,
      "loss": 1.348,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.6839028000831604,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.3348,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.4735791087150574,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.3445,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5173911452293396,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3394,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.5153633952140808,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3387,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.5287536978721619,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.3452,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.632926881313324,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3334,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.6215195059776306,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3354,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5091204643249512,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.3583,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.6117019653320312,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.3386,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.48274752497673035,
      "learning_rate": 0.000517326298154212,
      "loss": 1.3444,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.5117499232292175,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3468,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.5102267861366272,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3428,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.5028338432312012,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3352,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.4928154945373535,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.3291,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.4756507873535156,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3327,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.48273026943206787,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.3315,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.5361319184303284,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3426,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.48997265100479126,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.3131,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.6841908693313599,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.3271,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.532696008682251,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3315,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.6385636329650879,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3389,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5205770134925842,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3227,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.54034423828125,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.3302,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.6151791214942932,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.3364,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.4874185025691986,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3322,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.49219995737075806,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3303,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.5832408666610718,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.324,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5040769577026367,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.3332,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.520376980304718,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3282,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.48763787746429443,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.3389,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.5245943069458008,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.3194,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5624497532844543,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.3178,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.47683998942375183,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3261,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.5936027765274048,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.3346,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.5519657731056213,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.3116,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.5041986703872681,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3149,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.47437649965286255,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.3204,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.5577541589736938,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.3237,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.5560529828071594,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.3366,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.538975715637207,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3305,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.5760699510574341,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3268,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.4729076027870178,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3216,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.49490368366241455,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.3196,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.5074576139450073,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.3165,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.4897492527961731,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.3196,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.5229538679122925,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.3275,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.49189868569374084,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.3336,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.48868584632873535,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.3159,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.5069463849067688,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.3091,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5533775091171265,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3154,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.5721158981323242,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3115,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.46923840045928955,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3079,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.47305622696876526,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3167,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.5176447033882141,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3156,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.48857372999191284,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3047,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.4950768053531647,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.3135,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.5579528212547302,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3224,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.4817415177822113,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.3033,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.5814762115478516,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3087,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.46505385637283325,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.2941,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.48983410000801086,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.2987,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.4881570339202881,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.3006,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.45043596625328064,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.3102,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.48906126618385315,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3015,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.49786320328712463,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3094,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.4743672013282776,
      "learning_rate": 0.000237799979917852,
      "loss": 1.3158,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.5006662607192993,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3097,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.5024963617324829,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.3178,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.47887811064720154,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.3003,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.49095380306243896,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.2996,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.570153534412384,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.2998,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.5310852527618408,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.297,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.5419847965240479,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.2986,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.5495837926864624,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3106,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.5267235040664673,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.3072,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.49620896577835083,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.2973,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.5015327334403992,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.3011,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.504551112651825,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.315,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.4976200461387634,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.286,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.5471088886260986,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.3101,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.5165606737136841,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.3029,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.4817538857460022,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.3169,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.4919213056564331,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.294,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.5186462998390198,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.3013,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5643556714057922,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.3162,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.47646427154541016,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.2884,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.48565244674682617,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.2983,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.5267047882080078,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.2964,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.5193334817886353,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.2998,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.5119906067848206,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3066,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.4821706712245941,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3048,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.4766588807106018,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.2912,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5136614441871643,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.2932,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.4924360513687134,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.2976,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.5348549485206604,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.2921,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5341907739639282,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.3092,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4870835244655609,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.2993,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.4815746247768402,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.2965,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.49008122086524963,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3043,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.5038406252861023,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.293,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.46453824639320374,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.2788,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.49911126494407654,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.2878,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.44908633828163147,
      "learning_rate": 0.000142293772767289,
      "loss": 1.2782,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.5436993837356567,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.2894,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.47166144847869873,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3098,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.4969489574432373,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.3068,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.489970862865448,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.2853,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5221937298774719,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.2967,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5233868360519409,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.2933,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.489381343126297,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3043,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.5099634528160095,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.2959,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.5164536237716675,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.3016,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.4713574945926666,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.2834,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.46141287684440613,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.2867,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.49424222111701965,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.2853,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.5870708227157593,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.2983,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.4899046719074249,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.2907,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.5182633996009827,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.2968,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.508344292640686,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.2919,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.5706756711006165,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.288,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.49942389130592346,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3018,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.49767690896987915,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.2816,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.5047425031661987,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.2905,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.46033480763435364,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.2963,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.48442649841308594,
      "learning_rate": 0.000104851799496331,
      "loss": 1.2976,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.49809372425079346,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.2839,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.49856284260749817,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.2797,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.527169942855835,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.2985,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.49904850125312805,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.2868,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.4828746020793915,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.2868,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.4995872974395752,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.2877,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.461446613073349,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.2819,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.49835923314094543,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.294,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.49723586440086365,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.285,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.49502670764923096,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.2927,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.5696793794631958,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.2832,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.4698841869831085,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.297,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.46518006920814514,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.2894,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.4740748703479767,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.2974,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.4718835651874542,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.292,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.46124735474586487,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.2883,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.45440733432769775,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3059,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.5006871223449707,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.2933,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.4942260682582855,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.2908,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.4663679897785187,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2752,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.5059977173805237,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.2756,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.48211827874183655,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.2994,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.4900622069835663,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.2779,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.4712204933166504,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.2971,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.4644489288330078,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.2909,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.4857137203216553,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.2798,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.47446030378341675,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.2809,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.47509390115737915,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.2784,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.46106091141700745,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.2856,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.5065385699272156,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.2947,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.48078155517578125,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.2819,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.47770050168037415,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.2844,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.48455145955085754,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.3026,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.46511292457580566,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.2894,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.48164862394332886,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.2858,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.49383822083473206,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2797,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5019394755363464,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.289,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.5560463666915894,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.2832,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.48706531524658203,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.2865,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.47477051615715027,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.2811,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.5075830817222595,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.2858,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.46826353669166565,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2745,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.48045527935028076,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.286,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.47034719586372375,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.2801,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.4913272559642792,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.2969,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.47024253010749817,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.2918,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.5205992460250854,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.2826,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.47987884283065796,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2747,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5683859586715698,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.2936,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.4900084435939789,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.2847,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5204929113388062,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.2894,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.46514394879341125,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.2827,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.49026528000831604,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.2842,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.47097718715667725,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2807,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.48903873562812805,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.2913,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5587120652198792,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2711,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.48478958010673523,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2778,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.5230157971382141,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.2777,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.47784537076950073,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.2847,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.4680609405040741,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.2931,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4730169177055359,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.2876,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.4942316710948944,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.284,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.4925999641418457,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.2802,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.4492323696613312,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.2774,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.48974716663360596,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.2894,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.46638819575309753,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2799,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.4784853756427765,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.2773,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.47609829902648926,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.2656,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4621826708316803,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.2763,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.49537214636802673,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.2836,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.46112343668937683,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2774,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.4806409180164337,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.2748,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.451741486787796,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.2957,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.4543476998806,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.2797,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.4623158872127533,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.2969,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.4667626917362213,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.2827,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.5451642870903015,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.2805,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5066413283348083,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.2821,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.46192336082458496,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.2833,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.5067285895347595,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.2853,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.46968767046928406,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.2857,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.5083792805671692,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2772,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.4996829330921173,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2748,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.45652031898498535,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.283,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.470613956451416,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.2804,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.45153480768203735,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.2803,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.46365368366241455,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.2828,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.47216668725013733,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.2858,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5004749894142151,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.2825,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5506999492645264,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.2787,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.47063425183296204,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.2821,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.4648600220680237,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.278,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.48788538575172424,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.2758,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.45217838883399963,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.2782,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.47035345435142517,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.2971,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.4552898108959198,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.2727,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.47409144043922424,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.2832,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.44321408867836,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.2852,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.47700709104537964,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.2856,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.48228153586387634,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.2835,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.4789127707481384,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.2856,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.487445592880249,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2657,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.46307191252708435,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.2788,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.517853319644928,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.2872,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.4690190255641937,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.2734,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.5210829973220825,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.2907,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.5002433657646179,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.2775,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.4795536994934082,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.2846,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.4676569998264313,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.2817,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.47357791662216187,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2821,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.4524855315685272,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2764,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.513187825679779,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2776,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.47211483120918274,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.2799,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.4745982587337494,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.2778,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.45773372054100037,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.2697,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.47228479385375977,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.2638,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.46443086862564087,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.275,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.4544282555580139,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.2706,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.47706711292266846,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.2814,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.4694705009460449,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2717,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.4609600901603699,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.2793,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.46797260642051697,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2774,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.4882201552391052,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2767,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.4751724898815155,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2734,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4728923439979553,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.2865,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.48171117901802063,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.2843,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.4628165662288666,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.2943,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.4623371660709381,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.2888,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.4719426929950714,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.2818,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.47426339983940125,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2691,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.47118404507637024,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2717,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.478762149810791,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.2903,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.4964239299297333,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2624,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.4898985028266907,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.2812,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.46895626187324524,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.2823,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.48857542872428894,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.2767,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.4606000483036041,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.2872,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.46462327241897583,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.2721,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.4766453504562378,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.2826,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4233559370040894,
      "learning_rate": 1.5020833487815421e-05,
      "loss": 1.2733,
      "step": 9480
    }
  ],
  "logging_steps": 10,
  "max_steps": 9480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.832308198648013e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}