{
  "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": 1.3124364614486694,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.5106,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1659833192825317,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.8744,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8274556994438171,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.2127,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 2.096945285797119,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.7132,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.9831845760345459,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.2337,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 1.2277730703353882,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.7354,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.048064947128296,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3617,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.9260877370834351,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.1292,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 1.345253348350525,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.9271,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 1.2292276620864868,
      "learning_rate": 0.0015,
      "loss": 3.7717,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.6336931586265564,
      "learning_rate": 0.0015,
      "loss": 3.616,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.738336980342865,
      "learning_rate": 0.0015,
      "loss": 3.504,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.8369924426078796,
      "learning_rate": 0.0015,
      "loss": 3.3989,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.646684467792511,
      "learning_rate": 0.0015,
      "loss": 3.2992,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.9423049688339233,
      "learning_rate": 0.0015,
      "loss": 3.2291,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 1.4281316995620728,
      "learning_rate": 0.0015,
      "loss": 3.1643,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.7731781601905823,
      "learning_rate": 0.0015,
      "loss": 3.1024,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.860463559627533,
      "learning_rate": 0.0015,
      "loss": 3.0604,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 1.0924943685531616,
      "learning_rate": 0.0015,
      "loss": 3.0025,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 1.1869707107543945,
      "learning_rate": 0.0015,
      "loss": 2.9583,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.9566922187805176,
      "learning_rate": 0.0015,
      "loss": 2.9292,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.8435981869697571,
      "learning_rate": 0.0015,
      "loss": 2.8961,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.7190911769866943,
      "learning_rate": 0.0015,
      "loss": 2.8459,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.8628165125846863,
      "learning_rate": 0.0015,
      "loss": 2.8162,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.9168346524238586,
      "learning_rate": 0.0015,
      "loss": 2.7849,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.9622735977172852,
      "learning_rate": 0.0015,
      "loss": 2.7517,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 1.2251378297805786,
      "learning_rate": 0.0015,
      "loss": 2.7257,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 1.2179166078567505,
      "learning_rate": 0.0015,
      "loss": 2.6938,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.8709646463394165,
      "learning_rate": 0.0015,
      "loss": 2.6663,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.8570034503936768,
      "learning_rate": 0.0015,
      "loss": 2.6402,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.9226229786872864,
      "learning_rate": 0.0015,
      "loss": 2.6211,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.6741663813591003,
      "learning_rate": 0.0015,
      "loss": 2.5844,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 1.1187233924865723,
      "learning_rate": 0.0015,
      "loss": 2.5831,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.7607849836349487,
      "learning_rate": 0.0015,
      "loss": 2.5591,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 1.4700108766555786,
      "learning_rate": 0.0015,
      "loss": 2.5197,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.8260409832000732,
      "learning_rate": 0.0015,
      "loss": 2.4953,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.9084756374359131,
      "learning_rate": 0.0015,
      "loss": 2.4791,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.797562301158905,
      "learning_rate": 0.0015,
      "loss": 2.4715,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 1.210136890411377,
      "learning_rate": 0.0015,
      "loss": 2.448,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.7420892715454102,
      "learning_rate": 0.0015,
      "loss": 2.4179,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.9167479872703552,
      "learning_rate": 0.0015,
      "loss": 2.4091,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.8478865623474121,
      "learning_rate": 0.0015,
      "loss": 2.3941,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.9713917374610901,
      "learning_rate": 0.0015,
      "loss": 2.3615,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.8634635806083679,
      "learning_rate": 0.0015,
      "loss": 2.3476,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.828904390335083,
      "learning_rate": 0.0015,
      "loss": 2.3529,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.777154803276062,
      "learning_rate": 0.0015,
      "loss": 2.3204,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.7510536909103394,
      "learning_rate": 0.0015,
      "loss": 2.3013,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 1.222251534461975,
      "learning_rate": 0.0015,
      "loss": 2.2974,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.7576515078544617,
      "learning_rate": 0.0015,
      "loss": 2.2676,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.8541025519371033,
      "learning_rate": 0.0015,
      "loss": 2.2628,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.9027695655822754,
      "learning_rate": 0.0015,
      "loss": 2.2526,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.7879486680030823,
      "learning_rate": 0.0015,
      "loss": 2.2404,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.7479010224342346,
      "learning_rate": 0.0015,
      "loss": 2.2215,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.9469801783561707,
      "learning_rate": 0.0015,
      "loss": 2.2036,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.7802208662033081,
      "learning_rate": 0.0015,
      "loss": 2.1999,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.8081716299057007,
      "learning_rate": 0.0015,
      "loss": 2.1678,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.7798997759819031,
      "learning_rate": 0.0015,
      "loss": 2.1736,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 0.8570801019668579,
      "learning_rate": 0.0015,
      "loss": 2.1711,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.9379898905754089,
      "learning_rate": 0.0015,
      "loss": 2.1417,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.5215798616409302,
      "learning_rate": 0.0015,
      "loss": 2.1266,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.9008197784423828,
      "learning_rate": 0.0015,
      "loss": 2.1324,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.6915869116783142,
      "learning_rate": 0.0015,
      "loss": 2.1085,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.8122121095657349,
      "learning_rate": 0.0015,
      "loss": 2.0962,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.7248497605323792,
      "learning_rate": 0.0015,
      "loss": 2.1083,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.7885815501213074,
      "learning_rate": 0.0015,
      "loss": 2.096,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 1.0940124988555908,
      "learning_rate": 0.0015,
      "loss": 2.0864,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.9496759176254272,
      "learning_rate": 0.0015,
      "loss": 2.0683,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.7619010806083679,
      "learning_rate": 0.0015,
      "loss": 2.0573,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.6933894753456116,
      "learning_rate": 0.0015,
      "loss": 2.0698,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.7526223659515381,
      "learning_rate": 0.0015,
      "loss": 2.0488,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.6634929180145264,
      "learning_rate": 0.0015,
      "loss": 2.0402,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 1.2680834531784058,
      "learning_rate": 0.0015,
      "loss": 2.0199,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.9700834155082703,
      "learning_rate": 0.0015,
      "loss": 2.0161,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.8383921384811401,
      "learning_rate": 0.0015,
      "loss": 2.0247,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.845635175704956,
      "learning_rate": 0.0015,
      "loss": 2.0064,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.2328712940216064,
      "learning_rate": 0.0015,
      "loss": 2.0011,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 1.0234324932098389,
      "learning_rate": 0.0015,
      "loss": 1.9948,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.9338547587394714,
      "learning_rate": 0.0015,
      "loss": 1.9883,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.7371922135353088,
      "learning_rate": 0.0015,
      "loss": 1.9939,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.7209744453430176,
      "learning_rate": 0.0015,
      "loss": 1.9682,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.7245383262634277,
      "learning_rate": 0.0015,
      "loss": 1.9563,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 1.1245934963226318,
      "learning_rate": 0.0015,
      "loss": 1.9653,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.7008262872695923,
      "learning_rate": 0.0015,
      "loss": 1.9652,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.9483989477157593,
      "learning_rate": 0.0015,
      "loss": 1.9461,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.717894971370697,
      "learning_rate": 0.0015,
      "loss": 1.9419,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.7447081208229065,
      "learning_rate": 0.0015,
      "loss": 1.9442,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.8731544613838196,
      "learning_rate": 0.0015,
      "loss": 1.9419,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.9204733371734619,
      "learning_rate": 0.0015,
      "loss": 1.9304,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 1.242817997932434,
      "learning_rate": 0.0015,
      "loss": 1.9259,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.9278952479362488,
      "learning_rate": 0.0015,
      "loss": 1.9245,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.7062042951583862,
      "learning_rate": 0.0015,
      "loss": 1.926,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.786612868309021,
      "learning_rate": 0.0015,
      "loss": 1.9149,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 1.001168131828308,
      "learning_rate": 0.0015,
      "loss": 1.8969,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.9273200035095215,
      "learning_rate": 0.0015,
      "loss": 1.908,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.8182703852653503,
      "learning_rate": 0.0015,
      "loss": 1.8871,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.8490096926689148,
      "learning_rate": 0.0015,
      "loss": 1.8907,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 1.2461068630218506,
      "learning_rate": 0.0015,
      "loss": 1.8969,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.9289028644561768,
      "learning_rate": 0.0015,
      "loss": 1.8805,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 1.049816608428955,
      "learning_rate": 0.0015,
      "loss": 1.8751,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.8595777750015259,
      "learning_rate": 0.0015,
      "loss": 1.8825,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.7875465750694275,
      "learning_rate": 0.0015,
      "loss": 1.8747,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.9580855965614319,
      "learning_rate": 0.0015,
      "loss": 1.8681,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.7750164866447449,
      "learning_rate": 0.0015,
      "loss": 1.8684,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.7347003817558289,
      "learning_rate": 0.0015,
      "loss": 1.8626,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.9035710096359253,
      "learning_rate": 0.0015,
      "loss": 1.8654,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.6644186973571777,
      "learning_rate": 0.0015,
      "loss": 1.8485,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.6655753254890442,
      "learning_rate": 0.0015,
      "loss": 1.8376,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.9941337704658508,
      "learning_rate": 0.0015,
      "loss": 1.8394,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.935934841632843,
      "learning_rate": 0.0015,
      "loss": 1.8478,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.7573521137237549,
      "learning_rate": 0.0015,
      "loss": 1.8373,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.7465738654136658,
      "learning_rate": 0.0015,
      "loss": 1.839,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.8040273189544678,
      "learning_rate": 0.0015,
      "loss": 1.8344,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 1.0702176094055176,
      "learning_rate": 0.0015,
      "loss": 1.8198,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 1.0176023244857788,
      "learning_rate": 0.0015,
      "loss": 1.8329,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.320648431777954,
      "learning_rate": 0.0015,
      "loss": 1.8165,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.8231999278068542,
      "learning_rate": 0.0015,
      "loss": 1.8279,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.6487593054771423,
      "learning_rate": 0.0015,
      "loss": 1.8084,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.8109908699989319,
      "learning_rate": 0.0015,
      "loss": 1.7991,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.6869083046913147,
      "learning_rate": 0.0015,
      "loss": 1.8032,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 1.4142192602157593,
      "learning_rate": 0.0015,
      "loss": 1.8208,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 1.269906759262085,
      "learning_rate": 0.0015,
      "loss": 1.8178,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.9178128838539124,
      "learning_rate": 0.0015,
      "loss": 1.7924,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.8090775012969971,
      "learning_rate": 0.0015,
      "loss": 1.7765,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.6909911036491394,
      "learning_rate": 0.0015,
      "loss": 1.7947,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.7301704287528992,
      "learning_rate": 0.0015,
      "loss": 1.7976,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.7470296621322632,
      "learning_rate": 0.0015,
      "loss": 1.7845,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.6710943579673767,
      "learning_rate": 0.0015,
      "loss": 1.7799,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.652271568775177,
      "learning_rate": 0.0015,
      "loss": 1.7962,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.7345850467681885,
      "learning_rate": 0.0015,
      "loss": 1.7836,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.7135871052742004,
      "learning_rate": 0.0015,
      "loss": 1.7707,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.643592357635498,
      "learning_rate": 0.0015,
      "loss": 1.772,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.7280349731445312,
      "learning_rate": 0.0015,
      "loss": 1.7779,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 1.4107588529586792,
      "learning_rate": 0.0015,
      "loss": 1.7671,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.709518551826477,
      "learning_rate": 0.0015,
      "loss": 1.7736,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.706581711769104,
      "learning_rate": 0.0015,
      "loss": 1.7738,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.6494163870811462,
      "learning_rate": 0.0015,
      "loss": 1.7618,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.8145359754562378,
      "learning_rate": 0.0015,
      "loss": 1.7601,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.7344682216644287,
      "learning_rate": 0.0015,
      "loss": 1.7516,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 1.0990504026412964,
      "learning_rate": 0.0015,
      "loss": 1.7637,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.7976487278938293,
      "learning_rate": 0.0015,
      "loss": 1.7469,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.7702522277832031,
      "learning_rate": 0.0015,
      "loss": 1.749,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.6728357672691345,
      "learning_rate": 0.0015,
      "loss": 1.7529,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.924201250076294,
      "learning_rate": 0.0015,
      "loss": 1.7487,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.7881402373313904,
      "learning_rate": 0.0015,
      "loss": 1.7482,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.7286288142204285,
      "learning_rate": 0.0015,
      "loss": 1.7428,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.6509220004081726,
      "learning_rate": 0.0015,
      "loss": 1.7408,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 1.1083709001541138,
      "learning_rate": 0.0015,
      "loss": 1.7375,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.6865395307540894,
      "learning_rate": 0.0015,
      "loss": 1.7398,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.7428089380264282,
      "learning_rate": 0.0015,
      "loss": 1.7375,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.745134174823761,
      "learning_rate": 0.0015,
      "loss": 1.7374,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.8267588019371033,
      "learning_rate": 0.0015,
      "loss": 1.7343,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.7420427203178406,
      "learning_rate": 0.0015,
      "loss": 1.735,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.643179178237915,
      "learning_rate": 0.0015,
      "loss": 1.7154,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6770628690719604,
      "learning_rate": 0.0015,
      "loss": 1.7102,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 1.2731250524520874,
      "learning_rate": 0.0015,
      "loss": 1.7121,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 1.104306697845459,
      "learning_rate": 0.0015,
      "loss": 1.7241,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.757106602191925,
      "learning_rate": 0.0015,
      "loss": 1.7131,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.8353426456451416,
      "learning_rate": 0.0015,
      "loss": 1.7125,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 1.0011383295059204,
      "learning_rate": 0.0015,
      "loss": 1.7099,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.7113327980041504,
      "learning_rate": 0.0015,
      "loss": 1.7154,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.723044753074646,
      "learning_rate": 0.0015,
      "loss": 1.7163,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6448427438735962,
      "learning_rate": 0.0015,
      "loss": 1.6848,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.7718586325645447,
      "learning_rate": 0.0015,
      "loss": 1.6985,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.7018836140632629,
      "learning_rate": 0.0015,
      "loss": 1.7001,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.8802307844161987,
      "learning_rate": 0.0015,
      "loss": 1.6923,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.782353937625885,
      "learning_rate": 0.0015,
      "loss": 1.691,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.732453465461731,
      "learning_rate": 0.0015,
      "loss": 1.7032,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.8128688335418701,
      "learning_rate": 0.0015,
      "loss": 1.689,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.7799118757247925,
      "learning_rate": 0.0015,
      "loss": 1.6798,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.6542706489562988,
      "learning_rate": 0.0015,
      "loss": 1.6939,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.6584886908531189,
      "learning_rate": 0.0015,
      "loss": 1.6763,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.7360519766807556,
      "learning_rate": 0.0015,
      "loss": 1.6777,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.868495523929596,
      "learning_rate": 0.0015,
      "loss": 1.6815,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.8044885993003845,
      "learning_rate": 0.0015,
      "loss": 1.684,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.7511466145515442,
      "learning_rate": 0.0015,
      "loss": 1.7041,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.6682915091514587,
      "learning_rate": 0.0015,
      "loss": 1.6836,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6851530075073242,
      "learning_rate": 0.0015,
      "loss": 1.6811,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.8657528758049011,
      "learning_rate": 0.0015,
      "loss": 1.6873,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 1.4020063877105713,
      "learning_rate": 0.0015,
      "loss": 1.6893,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.8835797905921936,
      "learning_rate": 0.0015,
      "loss": 1.6793,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.629051685333252,
      "learning_rate": 0.0015,
      "loss": 1.6651,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.7854632139205933,
      "learning_rate": 0.0015,
      "loss": 1.6655,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.6483657360076904,
      "learning_rate": 0.0015,
      "loss": 1.6634,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.7135815620422363,
      "learning_rate": 0.0015,
      "loss": 1.6644,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.7771666049957275,
      "learning_rate": 0.0015,
      "loss": 1.6689,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.8964939713478088,
      "learning_rate": 0.0015,
      "loss": 1.6729,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.6993046998977661,
      "learning_rate": 0.0015,
      "loss": 1.6697,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6576790809631348,
      "learning_rate": 0.0015,
      "loss": 1.6567,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.7285045981407166,
      "learning_rate": 0.0015,
      "loss": 1.6631,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.9496256113052368,
      "learning_rate": 0.0015,
      "loss": 1.6597,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.7459031939506531,
      "learning_rate": 0.0015,
      "loss": 1.6583,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.8031527996063232,
      "learning_rate": 0.0015,
      "loss": 1.6603,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.801005482673645,
      "learning_rate": 0.0015,
      "loss": 1.6596,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.707309901714325,
      "learning_rate": 0.0015,
      "loss": 1.6475,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.6514493823051453,
      "learning_rate": 0.0015,
      "loss": 1.6458,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.7231051921844482,
      "learning_rate": 0.0015,
      "loss": 1.6509,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.6800807118415833,
      "learning_rate": 0.0015,
      "loss": 1.6629,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.6729750037193298,
      "learning_rate": 0.0015,
      "loss": 1.6425,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.654437243938446,
      "learning_rate": 0.0015,
      "loss": 1.6403,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.8978038430213928,
      "learning_rate": 0.0015,
      "loss": 1.6408,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.6976383924484253,
      "learning_rate": 0.0015,
      "loss": 1.6503,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.8478453755378723,
      "learning_rate": 0.0015,
      "loss": 1.6584,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.719779372215271,
      "learning_rate": 0.0015,
      "loss": 1.6435,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6123868227005005,
      "learning_rate": 0.0015,
      "loss": 1.6438,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.7718461751937866,
      "learning_rate": 0.0015,
      "loss": 1.6403,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.7014973759651184,
      "learning_rate": 0.0015,
      "loss": 1.6331,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 1.0192010402679443,
      "learning_rate": 0.0015,
      "loss": 1.6352,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.7457175254821777,
      "learning_rate": 0.0015,
      "loss": 1.6402,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 1.0028305053710938,
      "learning_rate": 0.0015,
      "loss": 1.6344,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 1.0459880828857422,
      "learning_rate": 0.0015,
      "loss": 1.6409,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.7769726514816284,
      "learning_rate": 0.0015,
      "loss": 1.6245,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6035235524177551,
      "learning_rate": 0.0015,
      "loss": 1.6222,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.7195689082145691,
      "learning_rate": 0.0015,
      "loss": 1.6295,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6837499141693115,
      "learning_rate": 0.0015,
      "loss": 1.6224,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.9146136045455933,
      "learning_rate": 0.0015,
      "loss": 1.63,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.9756319522857666,
      "learning_rate": 0.0015,
      "loss": 1.6454,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.8231519460678101,
      "learning_rate": 0.0015,
      "loss": 1.6263,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.8124420046806335,
      "learning_rate": 0.0015,
      "loss": 1.6107,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.8022341132164001,
      "learning_rate": 0.0015,
      "loss": 1.6293,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.6795050501823425,
      "learning_rate": 0.0015,
      "loss": 1.63,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.783323347568512,
      "learning_rate": 0.0015,
      "loss": 1.6145,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.6651756763458252,
      "learning_rate": 0.0015,
      "loss": 1.6263,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 1.179243803024292,
      "learning_rate": 0.0015,
      "loss": 1.6212,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.8220842480659485,
      "learning_rate": 0.0015,
      "loss": 1.6209,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 1.0973544120788574,
      "learning_rate": 0.0015,
      "loss": 1.6117,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 1.2406624555587769,
      "learning_rate": 0.0015,
      "loss": 1.6039,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.7092892527580261,
      "learning_rate": 0.0015,
      "loss": 1.6396,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.7969619631767273,
      "learning_rate": 0.0015,
      "loss": 1.6216,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.8276784420013428,
      "learning_rate": 0.0015,
      "loss": 1.608,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.9074499607086182,
      "learning_rate": 0.0015,
      "loss": 1.606,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.6614257097244263,
      "learning_rate": 0.0015,
      "loss": 1.6058,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.6448475122451782,
      "learning_rate": 0.0015,
      "loss": 1.6141,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.7352731227874756,
      "learning_rate": 0.0015,
      "loss": 1.5993,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.8879994750022888,
      "learning_rate": 0.0015,
      "loss": 1.5969,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.7568689584732056,
      "learning_rate": 0.0015,
      "loss": 1.6009,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.7219222187995911,
      "learning_rate": 0.0015,
      "loss": 1.6043,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.708109438419342,
      "learning_rate": 0.0015,
      "loss": 1.6037,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.6003740429878235,
      "learning_rate": 0.0015,
      "loss": 1.6037,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.6674261093139648,
      "learning_rate": 0.0015,
      "loss": 1.6059,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6381140947341919,
      "learning_rate": 0.0015,
      "loss": 1.6029,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.6798749566078186,
      "learning_rate": 0.0015,
      "loss": 1.6053,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.8452035188674927,
      "learning_rate": 0.0015,
      "loss": 1.5994,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.6483489274978638,
      "learning_rate": 0.0015,
      "loss": 1.597,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.7470886707305908,
      "learning_rate": 0.0015,
      "loss": 1.6056,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.6924394369125366,
      "learning_rate": 0.0015,
      "loss": 1.5921,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.6187343597412109,
      "learning_rate": 0.0015,
      "loss": 1.6053,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.944770097732544,
      "learning_rate": 0.0015,
      "loss": 1.5954,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.760418176651001,
      "learning_rate": 0.0015,
      "loss": 1.6016,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.767612874507904,
      "learning_rate": 0.0015,
      "loss": 1.5883,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.6893068552017212,
      "learning_rate": 0.0015,
      "loss": 1.5943,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.6650967001914978,
      "learning_rate": 0.0015,
      "loss": 1.5914,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.745049774646759,
      "learning_rate": 0.0015,
      "loss": 1.5954,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.6933596134185791,
      "learning_rate": 0.0015,
      "loss": 1.5913,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.6802793741226196,
      "learning_rate": 0.0015,
      "loss": 1.5928,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.6025201082229614,
      "learning_rate": 0.0015,
      "loss": 1.5859,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.680928647518158,
      "learning_rate": 0.0015,
      "loss": 1.5962,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.71792072057724,
      "learning_rate": 0.0015,
      "loss": 1.5904,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.6014297008514404,
      "learning_rate": 0.0015,
      "loss": 1.5882,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.6050705909729004,
      "learning_rate": 0.0015,
      "loss": 1.5924,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.688946545124054,
      "learning_rate": 0.0015,
      "loss": 1.5909,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6885754466056824,
      "learning_rate": 0.0015,
      "loss": 1.5913,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6848039031028748,
      "learning_rate": 0.0015,
      "loss": 1.5892,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.7295860052108765,
      "learning_rate": 0.0015,
      "loss": 1.5763,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.7082112431526184,
      "learning_rate": 0.0015,
      "loss": 1.5896,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.820855975151062,
      "learning_rate": 0.0015,
      "loss": 1.5832,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.653474748134613,
      "learning_rate": 0.0015,
      "loss": 1.5757,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.6177871227264404,
      "learning_rate": 0.0015,
      "loss": 1.574,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.7549741268157959,
      "learning_rate": 0.0015,
      "loss": 1.5763,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.6169193983078003,
      "learning_rate": 0.0015,
      "loss": 1.5787,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.7113834023475647,
      "learning_rate": 0.0015,
      "loss": 1.5706,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.7188537120819092,
      "learning_rate": 0.0015,
      "loss": 1.5774,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.6048180460929871,
      "learning_rate": 0.0015,
      "loss": 1.5795,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.7804025411605835,
      "learning_rate": 0.0015,
      "loss": 1.5591,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 1.2684450149536133,
      "learning_rate": 0.0015,
      "loss": 1.5732,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.6620485782623291,
      "learning_rate": 0.0015,
      "loss": 1.5801,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.6825975775718689,
      "learning_rate": 0.0015,
      "loss": 1.588,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6704723238945007,
      "learning_rate": 0.0015,
      "loss": 1.5789,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.7003452777862549,
      "learning_rate": 0.0015,
      "loss": 1.5759,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.5659245252609253,
      "learning_rate": 0.0015,
      "loss": 1.5726,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.7406479120254517,
      "learning_rate": 0.0015,
      "loss": 1.5701,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.7669352293014526,
      "learning_rate": 0.0015,
      "loss": 1.5706,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.7264388799667358,
      "learning_rate": 0.0015,
      "loss": 1.5775,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 1.3137651681900024,
      "learning_rate": 0.0015,
      "loss": 1.5563,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.7707687616348267,
      "learning_rate": 0.0015,
      "loss": 1.5596,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.674060046672821,
      "learning_rate": 0.0015,
      "loss": 1.5593,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.6680307388305664,
      "learning_rate": 0.0015,
      "loss": 1.5656,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.7418972849845886,
      "learning_rate": 0.0015,
      "loss": 1.5651,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.638434648513794,
      "learning_rate": 0.0015,
      "loss": 1.5542,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.6422763466835022,
      "learning_rate": 0.0015,
      "loss": 1.5595,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.7120733857154846,
      "learning_rate": 0.0015,
      "loss": 1.5635,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.7064751386642456,
      "learning_rate": 0.0015,
      "loss": 1.5665,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.6330904364585876,
      "learning_rate": 0.0015,
      "loss": 1.5613,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.6634082794189453,
      "learning_rate": 0.0015,
      "loss": 1.5545,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.6552626490592957,
      "learning_rate": 0.0015,
      "loss": 1.5622,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.8915601968765259,
      "learning_rate": 0.0015,
      "loss": 1.5715,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.7891660332679749,
      "learning_rate": 0.0015,
      "loss": 1.5549,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 1.112757921218872,
      "learning_rate": 0.0015,
      "loss": 1.5492,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.5727996826171875,
      "learning_rate": 0.0015,
      "loss": 1.5524,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.634543240070343,
      "learning_rate": 0.0015,
      "loss": 1.5638,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.6999568939208984,
      "learning_rate": 0.0015,
      "loss": 1.5467,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.7002508640289307,
      "learning_rate": 0.0015,
      "loss": 1.5568,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.7466462254524231,
      "learning_rate": 0.0015,
      "loss": 1.5604,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6902084350585938,
      "learning_rate": 0.0015,
      "loss": 1.5441,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.7047731876373291,
      "learning_rate": 0.0015,
      "loss": 1.5516,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.7328782677650452,
      "learning_rate": 0.0015,
      "loss": 1.5474,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.6541309952735901,
      "learning_rate": 0.0015,
      "loss": 1.5603,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.6323668360710144,
      "learning_rate": 0.0015,
      "loss": 1.5463,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.6534177660942078,
      "learning_rate": 0.0015,
      "loss": 1.5422,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.5820841193199158,
      "learning_rate": 0.0015,
      "loss": 1.5568,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.6260690689086914,
      "learning_rate": 0.0015,
      "loss": 1.541,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6181133389472961,
      "learning_rate": 0.0015,
      "loss": 1.545,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.6371879577636719,
      "learning_rate": 0.0015,
      "loss": 1.5482,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.7137547135353088,
      "learning_rate": 0.0015,
      "loss": 1.5609,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.7851402759552002,
      "learning_rate": 0.0015,
      "loss": 1.536,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.636054277420044,
      "learning_rate": 0.0015,
      "loss": 1.5477,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.6029794812202454,
      "learning_rate": 0.0015,
      "loss": 1.546,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.7253355383872986,
      "learning_rate": 0.0015,
      "loss": 1.5516,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.676753580570221,
      "learning_rate": 0.0015,
      "loss": 1.544,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.9133493900299072,
      "learning_rate": 0.0015,
      "loss": 1.547,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.7947659492492676,
      "learning_rate": 0.0015,
      "loss": 1.5414,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.7425976991653442,
      "learning_rate": 0.0015,
      "loss": 1.5499,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.7035290002822876,
      "learning_rate": 0.0015,
      "loss": 1.5433,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.7503587007522583,
      "learning_rate": 0.0015,
      "loss": 1.5366,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.6289841532707214,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.6338568925857544,
      "learning_rate": 0.0015,
      "loss": 1.5426,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.6188430190086365,
      "learning_rate": 0.0015,
      "loss": 1.5321,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.7039831876754761,
      "learning_rate": 0.0015,
      "loss": 1.5482,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.869328498840332,
      "learning_rate": 0.0015,
      "loss": 1.547,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.8635297417640686,
      "learning_rate": 0.0015,
      "loss": 1.5339,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.6145782470703125,
      "learning_rate": 0.0015,
      "loss": 1.541,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.5749341249465942,
      "learning_rate": 0.0015,
      "loss": 1.5391,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.6416504383087158,
      "learning_rate": 0.0015,
      "loss": 1.5426,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.6237147450447083,
      "learning_rate": 0.0015,
      "loss": 1.5337,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.667945921421051,
      "learning_rate": 0.0015,
      "loss": 1.5364,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6189672350883484,
      "learning_rate": 0.0015,
      "loss": 1.5291,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.999944269657135,
      "learning_rate": 0.0015,
      "loss": 1.5309,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.7216845750808716,
      "learning_rate": 0.0015,
      "loss": 1.5398,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.8774672150611877,
      "learning_rate": 0.0015,
      "loss": 1.5336,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.581254243850708,
      "learning_rate": 0.0015,
      "loss": 1.526,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.7107481360435486,
      "learning_rate": 0.0015,
      "loss": 1.5338,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.6095128059387207,
      "learning_rate": 0.0015,
      "loss": 1.5324,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.7119369506835938,
      "learning_rate": 0.0015,
      "loss": 1.5123,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.6254273653030396,
      "learning_rate": 0.0015,
      "loss": 1.5314,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.6041594743728638,
      "learning_rate": 0.0015,
      "loss": 1.526,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.6356289982795715,
      "learning_rate": 0.0015,
      "loss": 1.5249,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.648369312286377,
      "learning_rate": 0.0015,
      "loss": 1.5175,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.6955825090408325,
      "learning_rate": 0.0015,
      "loss": 1.534,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.7062328457832336,
      "learning_rate": 0.0015,
      "loss": 1.5402,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.6580941081047058,
      "learning_rate": 0.0015,
      "loss": 1.5321,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.7443981170654297,
      "learning_rate": 0.0015,
      "loss": 1.5156,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.683188259601593,
      "learning_rate": 0.0015,
      "loss": 1.5344,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.602588415145874,
      "learning_rate": 0.0015,
      "loss": 1.5287,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.669110894203186,
      "learning_rate": 0.0015,
      "loss": 1.5363,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.9849790930747986,
      "learning_rate": 0.0015,
      "loss": 1.5338,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.7312303781509399,
      "learning_rate": 0.0015,
      "loss": 1.5244,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.6108510494232178,
      "learning_rate": 0.0015,
      "loss": 1.5204,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.9470191597938538,
      "learning_rate": 0.0015,
      "loss": 1.5208,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.6268520951271057,
      "learning_rate": 0.0015,
      "loss": 1.5218,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.6916161179542542,
      "learning_rate": 0.0015,
      "loss": 1.5334,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.948076069355011,
      "learning_rate": 0.0015,
      "loss": 1.5156,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.6372155547142029,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.7083125114440918,
      "learning_rate": 0.0015,
      "loss": 1.5203,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.5928717851638794,
      "learning_rate": 0.0015,
      "loss": 1.5198,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.8729598522186279,
      "learning_rate": 0.0015,
      "loss": 1.5204,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.7349838614463806,
      "learning_rate": 0.0015,
      "loss": 1.5304,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.776211678981781,
      "learning_rate": 0.0015,
      "loss": 1.5231,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.6093047857284546,
      "learning_rate": 0.0015,
      "loss": 1.5146,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.7260333895683289,
      "learning_rate": 0.0015,
      "loss": 1.5148,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.6650367975234985,
      "learning_rate": 0.0015,
      "loss": 1.5251,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.6198450922966003,
      "learning_rate": 0.0015,
      "loss": 1.5283,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.6064160466194153,
      "learning_rate": 0.0015,
      "loss": 1.5159,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.7401567697525024,
      "learning_rate": 0.0015,
      "loss": 1.4973,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.7242253422737122,
      "learning_rate": 0.0015,
      "loss": 1.5159,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.6363670229911804,
      "learning_rate": 0.0015,
      "loss": 1.5165,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.7001657485961914,
      "learning_rate": 0.0015,
      "loss": 1.5206,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.9348625540733337,
      "learning_rate": 0.0015,
      "loss": 1.5117,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.5924307107925415,
      "learning_rate": 0.0015,
      "loss": 1.5246,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.5479297637939453,
      "learning_rate": 0.0015,
      "loss": 1.5111,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.7290306091308594,
      "learning_rate": 0.0015,
      "loss": 1.5161,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.6529408097267151,
      "learning_rate": 0.0015,
      "loss": 1.5201,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.6077266335487366,
      "learning_rate": 0.0015,
      "loss": 1.5095,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.620369553565979,
      "learning_rate": 0.0015,
      "loss": 1.5027,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.6094712615013123,
      "learning_rate": 0.0015,
      "loss": 1.5033,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.6363013386726379,
      "learning_rate": 0.0015,
      "loss": 1.521,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.7066325545310974,
      "learning_rate": 0.0015,
      "loss": 1.5051,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 1.0054547786712646,
      "learning_rate": 0.0015,
      "loss": 1.5167,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.7523656487464905,
      "learning_rate": 0.0015,
      "loss": 1.5076,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.6852942705154419,
      "learning_rate": 0.0015,
      "loss": 1.5018,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.6333704590797424,
      "learning_rate": 0.0015,
      "loss": 1.4975,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 1.0130631923675537,
      "learning_rate": 0.0015,
      "loss": 1.4979,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.6275448203086853,
      "learning_rate": 0.0015,
      "loss": 1.5046,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.5929350256919861,
      "learning_rate": 0.0015,
      "loss": 1.5006,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.730098307132721,
      "learning_rate": 0.0015,
      "loss": 1.5065,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.6206305027008057,
      "learning_rate": 0.0015,
      "loss": 1.4993,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.7433390021324158,
      "learning_rate": 0.0015,
      "loss": 1.503,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.5319532155990601,
      "learning_rate": 0.0015,
      "loss": 1.5053,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.7162283062934875,
      "learning_rate": 0.0015,
      "loss": 1.497,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.6116654276847839,
      "learning_rate": 0.0015,
      "loss": 1.5076,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.6190109252929688,
      "learning_rate": 0.0015,
      "loss": 1.4991,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.6667494773864746,
      "learning_rate": 0.0015,
      "loss": 1.5121,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.5939257740974426,
      "learning_rate": 0.0015,
      "loss": 1.4995,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.6996036767959595,
      "learning_rate": 0.0015,
      "loss": 1.5032,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.5522063374519348,
      "learning_rate": 0.0015,
      "loss": 1.4918,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.6450867056846619,
      "learning_rate": 0.0015,
      "loss": 1.5028,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5899419784545898,
      "learning_rate": 0.0015,
      "loss": 1.5052,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 1.1095937490463257,
      "learning_rate": 0.0015,
      "loss": 1.5082,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.798089325428009,
      "learning_rate": 0.0015,
      "loss": 1.5045,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6162708401679993,
      "learning_rate": 0.0015,
      "loss": 1.509,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 1.1070821285247803,
      "learning_rate": 0.0015,
      "loss": 1.5054,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5930264592170715,
      "learning_rate": 0.0015,
      "loss": 1.5009,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.6469377279281616,
      "learning_rate": 0.0015,
      "loss": 1.4911,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.9273725152015686,
      "learning_rate": 0.0015,
      "loss": 1.4969,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.6360483765602112,
      "learning_rate": 0.0015,
      "loss": 1.4906,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.720285177230835,
      "learning_rate": 0.0015,
      "loss": 1.4938,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.6374127268791199,
      "learning_rate": 0.0015,
      "loss": 1.4995,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.6002802848815918,
      "learning_rate": 0.0015,
      "loss": 1.4921,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.6103620529174805,
      "learning_rate": 0.0015,
      "loss": 1.4926,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.6310016512870789,
      "learning_rate": 0.0015,
      "loss": 1.4883,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.7795982360839844,
      "learning_rate": 0.0015,
      "loss": 1.4956,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.7573109865188599,
      "learning_rate": 0.0015,
      "loss": 1.4994,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.8623130917549133,
      "learning_rate": 0.0015,
      "loss": 1.4871,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.6567066311836243,
      "learning_rate": 0.0015,
      "loss": 1.4971,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.8647184371948242,
      "learning_rate": 0.0015,
      "loss": 1.4927,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.7369650602340698,
      "learning_rate": 0.0015,
      "loss": 1.487,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.5668179392814636,
      "learning_rate": 0.0015,
      "loss": 1.4929,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.7031323909759521,
      "learning_rate": 0.0015,
      "loss": 1.5007,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5766967535018921,
      "learning_rate": 0.0015,
      "loss": 1.4901,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.826673686504364,
      "learning_rate": 0.0015,
      "loss": 1.4841,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 1.1313567161560059,
      "learning_rate": 0.0015,
      "loss": 1.4965,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.8283265233039856,
      "learning_rate": 0.0015,
      "loss": 1.4943,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.5803491473197937,
      "learning_rate": 0.0015,
      "loss": 1.4915,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.7237555384635925,
      "learning_rate": 0.0015,
      "loss": 1.487,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.6318506002426147,
      "learning_rate": 0.0015,
      "loss": 1.4942,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.6532315015792847,
      "learning_rate": 0.0015,
      "loss": 1.4865,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.7136020064353943,
      "learning_rate": 0.0015,
      "loss": 1.487,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.617569625377655,
      "learning_rate": 0.0015,
      "loss": 1.5038,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.6590544581413269,
      "learning_rate": 0.0015,
      "loss": 1.4819,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5890007019042969,
      "learning_rate": 0.0015,
      "loss": 1.4861,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.6877003312110901,
      "learning_rate": 0.0015,
      "loss": 1.4814,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.6379115581512451,
      "learning_rate": 0.0015,
      "loss": 1.4887,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.6963144540786743,
      "learning_rate": 0.0015,
      "loss": 1.5052,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.5845509171485901,
      "learning_rate": 0.0015,
      "loss": 1.488,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.6566439867019653,
      "learning_rate": 0.0015,
      "loss": 1.4905,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.7040950059890747,
      "learning_rate": 0.0015,
      "loss": 1.4852,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.6772409081459045,
      "learning_rate": 0.0015,
      "loss": 1.4914,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.6106696724891663,
      "learning_rate": 0.0015,
      "loss": 1.4767,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.5537508130073547,
      "learning_rate": 0.0015,
      "loss": 1.4899,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.6199619770050049,
      "learning_rate": 0.0015,
      "loss": 1.4936,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.5953665375709534,
      "learning_rate": 0.0015,
      "loss": 1.4824,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.6078841686248779,
      "learning_rate": 0.0015,
      "loss": 1.4793,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.581908106803894,
      "learning_rate": 0.0015,
      "loss": 1.4864,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.594734251499176,
      "learning_rate": 0.0015,
      "loss": 1.4692,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.5448296666145325,
      "learning_rate": 0.0015,
      "loss": 1.4895,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.5612491369247437,
      "learning_rate": 0.0015,
      "loss": 1.4897,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.8511930108070374,
      "learning_rate": 0.0015,
      "loss": 1.4883,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.5904642939567566,
      "learning_rate": 0.0015,
      "loss": 1.4751,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.6498848795890808,
      "learning_rate": 0.0015,
      "loss": 1.4604,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.7043668627738953,
      "learning_rate": 0.0015,
      "loss": 1.4833,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.6435434222221375,
      "learning_rate": 0.0015,
      "loss": 1.4686,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.5815362334251404,
      "learning_rate": 0.0015,
      "loss": 1.4883,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 1.1824867725372314,
      "learning_rate": 0.0015,
      "loss": 1.4933,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.6740747690200806,
      "learning_rate": 0.0015,
      "loss": 1.4695,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.5841023325920105,
      "learning_rate": 0.0015,
      "loss": 1.4855,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.5732386112213135,
      "learning_rate": 0.0015,
      "loss": 1.486,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.6099458336830139,
      "learning_rate": 0.0015,
      "loss": 1.4881,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.6234198212623596,
      "learning_rate": 0.0015,
      "loss": 1.473,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.718690812587738,
      "learning_rate": 0.0015,
      "loss": 1.4754,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.626200258731842,
      "learning_rate": 0.0015,
      "loss": 1.4715,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.6853267550468445,
      "learning_rate": 0.0015,
      "loss": 1.4684,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.5991837978363037,
      "learning_rate": 0.0015,
      "loss": 1.4787,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.584218442440033,
      "learning_rate": 0.0015,
      "loss": 1.4807,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.8109069466590881,
      "learning_rate": 0.0015,
      "loss": 1.4765,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.7000900506973267,
      "learning_rate": 0.0015,
      "loss": 1.4744,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6348422169685364,
      "learning_rate": 0.0015,
      "loss": 1.471,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.7023735046386719,
      "learning_rate": 0.0015,
      "loss": 1.4909,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5916067361831665,
      "learning_rate": 0.0015,
      "loss": 1.4749,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.5802889466285706,
      "learning_rate": 0.0015,
      "loss": 1.4754,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5390263795852661,
      "learning_rate": 0.0015,
      "loss": 1.4852,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.6481983661651611,
      "learning_rate": 0.0015,
      "loss": 1.4671,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5872080326080322,
      "learning_rate": 0.0015,
      "loss": 1.4736,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.604631781578064,
      "learning_rate": 0.0015,
      "loss": 1.4734,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.5989955067634583,
      "learning_rate": 0.0015,
      "loss": 1.4908,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.5893116593360901,
      "learning_rate": 0.0015,
      "loss": 1.4745,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.6147662401199341,
      "learning_rate": 0.0015,
      "loss": 1.4668,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.7372254729270935,
      "learning_rate": 0.0015,
      "loss": 1.4799,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.7213212847709656,
      "learning_rate": 0.0015,
      "loss": 1.4684,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.6522668600082397,
      "learning_rate": 0.0015,
      "loss": 1.461,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.6512165069580078,
      "learning_rate": 0.0015,
      "loss": 1.4843,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.6414441466331482,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.6639628410339355,
      "learning_rate": 0.0015,
      "loss": 1.4824,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.6600837707519531,
      "learning_rate": 0.0015,
      "loss": 1.4751,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.5745564699172974,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.6705775260925293,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.6552022695541382,
      "learning_rate": 0.0015,
      "loss": 1.4714,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.7936440706253052,
      "learning_rate": 0.0015,
      "loss": 1.4797,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.8443682789802551,
      "learning_rate": 0.0015,
      "loss": 1.4624,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.5906997919082642,
      "learning_rate": 0.0015,
      "loss": 1.4693,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.8045083284378052,
      "learning_rate": 0.0015,
      "loss": 1.4649,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.6744979619979858,
      "learning_rate": 0.0015,
      "loss": 1.4655,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.5991400480270386,
      "learning_rate": 0.0015,
      "loss": 1.4783,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.6100850701332092,
      "learning_rate": 0.0015,
      "loss": 1.4692,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.6038932204246521,
      "learning_rate": 0.0015,
      "loss": 1.4642,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.5445556044578552,
      "learning_rate": 0.0015,
      "loss": 1.45,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.57991623878479,
      "learning_rate": 0.0015,
      "loss": 1.4777,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.5831894874572754,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.6404194831848145,
      "learning_rate": 0.0015,
      "loss": 1.4759,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.9784829616546631,
      "learning_rate": 0.0015,
      "loss": 1.4649,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.6670748591423035,
      "learning_rate": 0.0015,
      "loss": 1.4806,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.6233241558074951,
      "learning_rate": 0.0015,
      "loss": 1.4695,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.574617862701416,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.6161243319511414,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.568108320236206,
      "learning_rate": 0.0015,
      "loss": 1.4642,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.7579213380813599,
      "learning_rate": 0.0015,
      "loss": 1.4634,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.9048457741737366,
      "learning_rate": 0.0015,
      "loss": 1.4753,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.6220145225524902,
      "learning_rate": 0.0015,
      "loss": 1.4716,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.681603193283081,
      "learning_rate": 0.0015,
      "loss": 1.4663,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5756487250328064,
      "learning_rate": 0.0015,
      "loss": 1.4663,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 1.1531461477279663,
      "learning_rate": 0.0015,
      "loss": 1.4779,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.9628408551216125,
      "learning_rate": 0.0015,
      "loss": 1.4595,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.9518396258354187,
      "learning_rate": 0.0015,
      "loss": 1.4728,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.6302949786186218,
      "learning_rate": 0.0015,
      "loss": 1.4565,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.6904149055480957,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.749798059463501,
      "learning_rate": 0.0015,
      "loss": 1.4591,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.7355242967605591,
      "learning_rate": 0.0015,
      "loss": 1.4668,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.6457265019416809,
      "learning_rate": 0.0015,
      "loss": 1.4608,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.6021826267242432,
      "learning_rate": 0.0015,
      "loss": 1.4623,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.5744286179542542,
      "learning_rate": 0.0015,
      "loss": 1.4725,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.6601939797401428,
      "learning_rate": 0.0015,
      "loss": 1.4617,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.6193671822547913,
      "learning_rate": 0.0015,
      "loss": 1.4621,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.571906566619873,
      "learning_rate": 0.0015,
      "loss": 1.467,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.7392724752426147,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.6911517381668091,
      "learning_rate": 0.0015,
      "loss": 1.4545,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.6782168745994568,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.5899258255958557,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.5885708332061768,
      "learning_rate": 0.0015,
      "loss": 1.4513,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.8930160403251648,
      "learning_rate": 0.0015,
      "loss": 1.4609,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.6113539338111877,
      "learning_rate": 0.0015,
      "loss": 1.4618,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.669581949710846,
      "learning_rate": 0.0015,
      "loss": 1.4553,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.6713536977767944,
      "learning_rate": 0.0015,
      "loss": 1.4544,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5978567600250244,
      "learning_rate": 0.0015,
      "loss": 1.4546,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5797361135482788,
      "learning_rate": 0.0015,
      "loss": 1.4573,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.6175971031188965,
      "learning_rate": 0.0015,
      "loss": 1.4535,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.6148843169212341,
      "learning_rate": 0.0015,
      "loss": 1.4503,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.9311176538467407,
      "learning_rate": 0.0015,
      "loss": 1.4473,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.7705535888671875,
      "learning_rate": 0.0015,
      "loss": 1.4575,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.8579835295677185,
      "learning_rate": 0.0015,
      "loss": 1.4583,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.6973420977592468,
      "learning_rate": 0.0015,
      "loss": 1.456,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.7770690321922302,
      "learning_rate": 0.0015,
      "loss": 1.4589,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.554744303226471,
      "learning_rate": 0.0015,
      "loss": 1.4625,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.6639339327812195,
      "learning_rate": 0.0015,
      "loss": 1.4529,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.6472654938697815,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.6235995888710022,
      "learning_rate": 0.0015,
      "loss": 1.4517,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.599632740020752,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.709097683429718,
      "learning_rate": 0.0015,
      "loss": 1.4439,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.6987321972846985,
      "learning_rate": 0.0015,
      "loss": 1.4604,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.6053097248077393,
      "learning_rate": 0.0015,
      "loss": 1.4668,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.5671143531799316,
      "learning_rate": 0.0015,
      "loss": 1.4516,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.7676701545715332,
      "learning_rate": 0.0015,
      "loss": 1.4496,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.6224042773246765,
      "learning_rate": 0.0015,
      "loss": 1.4566,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5750153660774231,
      "learning_rate": 0.0015,
      "loss": 1.4503,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.552209734916687,
      "learning_rate": 0.0015,
      "loss": 1.4529,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.5752233266830444,
      "learning_rate": 0.0015,
      "loss": 1.4632,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.8984862565994263,
      "learning_rate": 0.0015,
      "loss": 1.456,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.5493113994598389,
      "learning_rate": 0.0015,
      "loss": 1.4646,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.6208711862564087,
      "learning_rate": 0.0015,
      "loss": 1.4606,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.7052538394927979,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.5871936082839966,
      "learning_rate": 0.0015,
      "loss": 1.4602,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.6055148839950562,
      "learning_rate": 0.0015,
      "loss": 1.4607,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.8056248426437378,
      "learning_rate": 0.0015,
      "loss": 1.4431,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.6749975681304932,
      "learning_rate": 0.0015,
      "loss": 1.4484,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.6684216260910034,
      "learning_rate": 0.0015,
      "loss": 1.4514,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.571518063545227,
      "learning_rate": 0.0015,
      "loss": 1.4551,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.6335450410842896,
      "learning_rate": 0.0015,
      "loss": 1.4532,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.6443816423416138,
      "learning_rate": 0.0015,
      "loss": 1.4498,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.6893118619918823,
      "learning_rate": 0.0015,
      "loss": 1.426,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.6394256949424744,
      "learning_rate": 0.0015,
      "loss": 1.4362,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.7784489989280701,
      "learning_rate": 0.0015,
      "loss": 1.4374,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.5581979751586914,
      "learning_rate": 0.0015,
      "loss": 1.4459,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.6324455142021179,
      "learning_rate": 0.0015,
      "loss": 1.4438,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.5833821296691895,
      "learning_rate": 0.0015,
      "loss": 1.4466,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5692071914672852,
      "learning_rate": 0.0015,
      "loss": 1.4517,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5617991089820862,
      "learning_rate": 0.0015,
      "loss": 1.4653,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.5704951286315918,
      "learning_rate": 0.0015,
      "loss": 1.4463,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.655717670917511,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.7156020998954773,
      "learning_rate": 0.0015,
      "loss": 1.4613,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.6333968043327332,
      "learning_rate": 0.0015,
      "loss": 1.4486,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.5942566394805908,
      "learning_rate": 0.0015,
      "loss": 1.4431,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.576947808265686,
      "learning_rate": 0.0015,
      "loss": 1.4536,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5468796491622925,
      "learning_rate": 0.0015,
      "loss": 1.4464,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.5699398517608643,
      "learning_rate": 0.0015,
      "loss": 1.4447,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.7965925335884094,
      "learning_rate": 0.0015,
      "loss": 1.44,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.5579686164855957,
      "learning_rate": 0.0015,
      "loss": 1.4365,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 1.3399968147277832,
      "learning_rate": 0.0015,
      "loss": 1.4457,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.7462083101272583,
      "learning_rate": 0.0015,
      "loss": 1.4512,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.5861807465553284,
      "learning_rate": 0.0015,
      "loss": 1.4385,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.6418489217758179,
      "learning_rate": 0.0015,
      "loss": 1.4457,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.556674599647522,
      "learning_rate": 0.0015,
      "loss": 1.445,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.6408023834228516,
      "learning_rate": 0.0015,
      "loss": 1.4308,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.6888174414634705,
      "learning_rate": 0.0015,
      "loss": 1.4525,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.6021018624305725,
      "learning_rate": 0.0015,
      "loss": 1.4571,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.5619707703590393,
      "learning_rate": 0.0015,
      "loss": 1.4435,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.909877598285675,
      "learning_rate": 0.0015,
      "loss": 1.4468,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.5683539509773254,
      "learning_rate": 0.0015,
      "loss": 1.4294,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.7239286303520203,
      "learning_rate": 0.0015,
      "loss": 1.4474,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.6010700464248657,
      "learning_rate": 0.0015,
      "loss": 1.4298,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.6637555956840515,
      "learning_rate": 0.0015,
      "loss": 1.4497,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.8356103301048279,
      "learning_rate": 0.0015,
      "loss": 1.4417,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.535122811794281,
      "learning_rate": 0.0015,
      "loss": 1.4399,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.6084794402122498,
      "learning_rate": 0.0015,
      "loss": 1.441,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.7110099196434021,
      "learning_rate": 0.0015,
      "loss": 1.436,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.8643055558204651,
      "learning_rate": 0.0015,
      "loss": 1.4471,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.6604209542274475,
      "learning_rate": 0.0015,
      "loss": 1.441,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.7761085033416748,
      "learning_rate": 0.0015,
      "loss": 1.4397,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.8981795907020569,
      "learning_rate": 0.0015,
      "loss": 1.441,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.618600606918335,
      "learning_rate": 0.0015,
      "loss": 1.452,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.6228238344192505,
      "learning_rate": 0.0015,
      "loss": 1.4449,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.6530385613441467,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.4227,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.6597455143928528,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.4444,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5946850180625916,
      "learning_rate": 0.001444828815847542,
      "loss": 1.4454,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.5537338852882385,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.4366,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5353566408157349,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.4298,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.6548413634300232,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4369,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.6430940628051758,
      "learning_rate": 0.001366801618799797,
      "loss": 1.4264,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5876358151435852,
      "learning_rate": 0.001347962334099052,
      "loss": 1.4287,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.5104976892471313,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4349,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.6101750731468201,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4252,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.6192573308944702,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.4322,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5790297985076904,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.4305,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.5484774708747864,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.4249,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.7660226225852966,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4179,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.6046931743621826,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4163,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.6635903716087341,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.4123,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.6356135606765747,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4192,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.6040505766868591,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4101,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.5147744417190552,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.416,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.5723498463630676,
      "learning_rate": 0.001141156021915355,
      "loss": 1.413,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.6113776564598083,
      "learning_rate": 0.001125426919103997,
      "loss": 1.3941,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.6992884278297424,
      "learning_rate": 0.001109914618088799,
      "loss": 1.3988,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.7204107642173767,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.4129,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.615075945854187,
      "learning_rate": 0.001079528509501728,
      "loss": 1.4092,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.5387037992477417,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.4039,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.6567428112030029,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.391,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.5510596632957458,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.3992,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.5978795886039734,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3851,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5756472945213318,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.3883,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.5662719011306763,
      "learning_rate": 0.000993270988848579,
      "loss": 1.4016,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.7225199341773987,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.3955,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.7410444021224976,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.4011,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.5279145240783691,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.3929,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.528181254863739,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.405,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.6095656752586365,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.404,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.5902137160301208,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.3896,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.7211180329322815,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3808,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.8163548707962036,
      "learning_rate": 0.000888885722567627,
      "loss": 1.3922,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.6644204258918762,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.386,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.6455613374710083,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3859,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.5053851008415222,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.382,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.5896329879760742,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3773,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.6229731440544128,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3797,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5269219875335693,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3848,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.5342105627059937,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.371,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.517015814781189,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3796,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.5653436779975891,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.3728,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.6157445907592773,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.3777,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.6121566891670227,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3695,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.664849042892456,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3687,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5752630829811096,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.3722,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.5371103286743164,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.3717,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.48518502712249756,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.3626,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.536507785320282,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.3677,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.5858027338981628,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.368,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.6124630570411682,
      "learning_rate": 0.000692383699170611,
      "loss": 1.3582,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.5956146121025085,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3653,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.5350100994110107,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3673,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.5265085101127625,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.3673,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.6775269508361816,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3631,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.5426746010780334,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3622,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.7744086384773254,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3627,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.5614804625511169,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.368,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.5236024856567383,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3531,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.8842533230781555,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3547,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.5572026968002319,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.3566,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.5783604383468628,
      "learning_rate": 0.000594349404202577,
      "loss": 1.3535,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.615972101688385,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.3443,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.5398380160331726,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.3543,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5902853012084961,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3496,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.5017073154449463,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3454,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.6057908535003662,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.3491,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.6111939549446106,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3415,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.6225577592849731,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3422,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5773703455924988,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.366,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.5412142276763916,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.3439,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.5136180520057678,
      "learning_rate": 0.000517326298154212,
      "loss": 1.3519,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.6220201253890991,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3533,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.5175100564956665,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3507,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.5581703186035156,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3394,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.5413421988487244,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.3362,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.5695057511329651,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3374,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.5546701550483704,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.3393,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.6359083652496338,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3502,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.5230035185813904,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.3184,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.7573515176773071,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.335,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.5882290601730347,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3413,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.6509057879447937,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3442,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.6262148022651672,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3302,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.6157950758934021,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.3381,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.609734058380127,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.3424,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.5529407858848572,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3388,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.5082640647888184,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3383,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.5488715767860413,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.3336,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6439418196678162,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.3415,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.5768157839775085,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3344,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.5306650400161743,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.3451,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.5529879331588745,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.3256,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5800896883010864,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.3233,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.5231634974479675,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3324,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.651831328868866,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.3397,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.5566461682319641,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.318,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.5578268766403198,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3212,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.5183534026145935,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.3266,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.570838451385498,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.3278,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.6385667324066162,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.3432,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.6211504936218262,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3395,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.5946642756462097,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3346,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.5230995416641235,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3279,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.48349809646606445,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.3241,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.5506454706192017,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.3225,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.5584042072296143,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.3249,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.5269228219985962,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.3336,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.5732347965240479,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.3391,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.5150391459465027,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.3206,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.4965534210205078,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.317,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5015170574188232,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3197,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.5498356223106384,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3186,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.5097297430038452,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3147,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.5277174711227417,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3217,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.48281005024909973,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3249,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.5234740972518921,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3109,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.5207615494728088,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.3201,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.5451726317405701,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3251,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.5427468419075012,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.3066,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.537834882736206,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3124,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.5034371018409729,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.2992,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.5310587882995605,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.3035,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.49480363726615906,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.3082,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.5269278883934021,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.3152,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.5210197567939758,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3076,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.5343834161758423,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3157,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.5516914129257202,
      "learning_rate": 0.000237799979917852,
      "loss": 1.319,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.5483276844024658,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3158,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.5133262276649475,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.3219,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.473877489566803,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.3069,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.5395793318748474,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.307,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.6155946850776672,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.3048,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.518414318561554,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.3012,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.6070139408111572,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.3049,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.543787956237793,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3144,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.5364104509353638,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.3108,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.4920696020126343,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.3025,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.5035790205001831,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.3055,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.5044794082641602,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.3191,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.5351099967956543,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.2932,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.5535302758216858,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.3155,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.5851619243621826,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.3069,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.5154551863670349,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.3231,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.5179932713508606,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.2996,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.5227841138839722,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.3072,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.6099350452423096,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.32,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.6272170543670654,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.2945,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.5071361660957336,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.303,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.6405959129333496,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.3002,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.5483779311180115,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.3052,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.546625554561615,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3108,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.520866870880127,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3116,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.4994986057281494,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.2978,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5830324292182922,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.2968,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.5204963088035583,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.3057,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.6280856728553772,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.2982,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5684712529182434,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.314,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4693601429462433,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.3037,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.5411034226417542,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.3003,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.5254245400428772,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3112,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.513707160949707,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.2982,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.4985164701938629,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.2859,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.5627548098564148,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.2939,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.47483623027801514,
      "learning_rate": 0.000142293772767289,
      "loss": 1.2839,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.5388797521591187,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.2949,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.5436922907829285,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3166,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.5270851850509644,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.3102,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.518153727054596,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.2925,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5932394862174988,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.2998,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5680364370346069,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.2977,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.503745436668396,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3086,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.5149855017662048,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.3032,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.5323132276535034,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.3041,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.48893091082572937,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.29,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.5255675911903381,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.2893,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.4878615736961365,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.2893,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.6673786640167236,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.3023,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.5215135812759399,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.2973,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.55266934633255,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.3037,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.5418053865432739,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.2964,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.5645815134048462,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.2947,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.5466896295547485,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3066,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.5144866704940796,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.2879,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.49308452010154724,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.2961,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.5420898199081421,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.3015,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.49578407406806946,
      "learning_rate": 0.000104851799496331,
      "loss": 1.3008,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.5164802670478821,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.2892,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.5347152948379517,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.284,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.5997539162635803,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.3024,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.5137956738471985,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.2911,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.5002637505531311,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.2947,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.5070739984512329,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.291,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.49195969104766846,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.2882,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.5105777978897095,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.2999,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.514224112033844,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.2887,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.5406697392463684,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.2959,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.592902421951294,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.2878,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.5043621063232422,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.3015,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.496030330657959,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.2948,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.5479779243469238,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.3008,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.5193496942520142,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.2962,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.5200868248939514,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.2958,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.5099362134933472,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3082,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.48709386587142944,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.2984,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.5332502126693726,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.2949,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.5007814168930054,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2785,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.5404915809631348,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.2791,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.5205032825469971,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.3043,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.507988452911377,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.2843,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.4945039749145508,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.3016,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.49984875321388245,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.2944,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.5130136013031006,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.2817,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.5221125483512878,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.2852,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.5233004689216614,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.286,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.48948031663894653,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.2913,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.5290145874023438,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.2966,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.5159009099006653,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.2893,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.4972204267978668,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.2879,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.5020485520362854,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.3075,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4924856126308441,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.2947,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.5332484245300293,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.292,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.5037619471549988,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2857,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5982730388641357,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.294,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.5343823432922363,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.2898,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.49162301421165466,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.2915,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.5004287362098694,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.2826,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.5593549013137817,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.29,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.5120792388916016,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2808,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.4898166060447693,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.2924,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.5263928174972534,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.2866,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.4969990849494934,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.3002,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.49934616684913635,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.2964,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.5362349152565002,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.2876,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.5168766379356384,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2782,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5859018564224243,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.2994,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.5291493535041809,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.2917,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.61119145154953,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.2935,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.5436091423034668,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.2868,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.5782431364059448,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.2882,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.5377077460289001,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2829,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.5343286991119385,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.2951,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5730398297309875,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2752,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.5420026183128357,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2816,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.5295568108558655,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.2809,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.4965777099132538,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.2902,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.4828559160232544,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.2984,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4930232763290405,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.2883,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.48897314071655273,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.2877,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.511944055557251,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.2873,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.4932258129119873,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.281,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.5023041367530823,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.2946,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.49912768602371216,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2842,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.5335937738418579,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.2798,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.5522021651268005,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.2692,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.5096961855888367,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.2813,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.497565358877182,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.2884,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.4803836941719055,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2827,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.5137074589729309,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.2796,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.48237344622612,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.2984,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.47588059306144714,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.283,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.5075386166572571,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.2997,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.48565760254859924,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.2846,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.5382689237594604,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.2857,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5447915196418762,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.2859,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.5087395310401917,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.2895,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.5213615894317627,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.29,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.482086718082428,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.2906,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.4832416772842407,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2817,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.5172869563102722,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2812,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.5098752975463867,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.2881,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.5164980292320251,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.2829,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.4869752526283264,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.2857,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.5009376406669617,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.2866,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.5258128046989441,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.2884,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5623719096183777,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.2861,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5587673187255859,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.2824,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.5034454464912415,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.2903,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.4845164120197296,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.2824,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.5366633534431458,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.2805,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.5056827068328857,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.282,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.5826355814933777,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.3008,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.5047662258148193,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.277,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.4909191131591797,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.288,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.4818117618560791,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.2892,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.4789429008960724,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.2889,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.4879331588745117,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.289,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.5181798338890076,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.2906,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.4963332712650299,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2691,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.4964265525341034,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.2805,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.5356233716011047,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.292,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.5141054391860962,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.2775,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.5709123611450195,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.2943,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.535489559173584,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.2818,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.4924616515636444,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.2898,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.5021106600761414,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.2873,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.48365768790245056,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2852,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.5007200241088867,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2809,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.49459993839263916,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2835,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.49230140447616577,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.2836,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.5009668469429016,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.283,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.4961964786052704,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.2728,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.5043670535087585,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.2674,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.49852392077445984,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.2813,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.5222915410995483,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.2765,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.4927261173725128,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.2882,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.4864712059497833,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2735,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.489318311214447,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.2807,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.49541473388671875,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2812,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.5180107355117798,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2831,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.5209818482398987,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2766,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.5105024576187134,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.2884,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.4965643882751465,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.2881,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.4937804639339447,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.2987,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.49657437205314636,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.2896,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.5091279745101929,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.2874,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.5247161388397217,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2736,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.5218049883842468,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2784,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.4994415044784546,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.2921,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.5280937552452087,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2688,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.5012044310569763,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.2853,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.48524028062820435,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.286,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.4760825037956238,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.2819,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.5371087193489075,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.2924,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.5069384574890137,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.2753,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.5072888135910034,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.2894,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 1.5041695911245136e-05,
      "loss": 1.2784,
      "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
}