{
  "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": 1.252522349357605,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.4804,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1000920534133911,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.8247,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8507410883903503,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.1978,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.6792912483215332,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.6539,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.5512610673904419,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.1298,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 1.6829159259796143,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.6431,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.5187345743179321,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3098,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.9871526956558228,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.089,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.8603453040122986,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.9106,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.9857443571090698,
      "learning_rate": 0.0015,
      "loss": 3.7692,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.9845615029335022,
      "learning_rate": 0.0015,
      "loss": 3.6229,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.8077515959739685,
      "learning_rate": 0.0015,
      "loss": 3.5022,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.7223982214927673,
      "learning_rate": 0.0015,
      "loss": 3.3904,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 1.2559893131256104,
      "learning_rate": 0.0015,
      "loss": 3.3051,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.5861561894416809,
      "learning_rate": 0.0015,
      "loss": 3.2205,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 1.3253904581069946,
      "learning_rate": 0.0015,
      "loss": 3.1521,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.8034362196922302,
      "learning_rate": 0.0015,
      "loss": 3.0831,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.641057014465332,
      "learning_rate": 0.0015,
      "loss": 3.0401,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.9403523802757263,
      "learning_rate": 0.0015,
      "loss": 2.9806,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.8479517102241516,
      "learning_rate": 0.0015,
      "loss": 2.9328,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.7839453220367432,
      "learning_rate": 0.0015,
      "loss": 2.8927,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.8618499040603638,
      "learning_rate": 0.0015,
      "loss": 2.8606,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.8639710545539856,
      "learning_rate": 0.0015,
      "loss": 2.815,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.9371837377548218,
      "learning_rate": 0.0015,
      "loss": 2.7749,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 1.2291475534439087,
      "learning_rate": 0.0015,
      "loss": 2.7385,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.789775013923645,
      "learning_rate": 0.0015,
      "loss": 2.7046,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.7895159125328064,
      "learning_rate": 0.0015,
      "loss": 2.6751,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 1.0680145025253296,
      "learning_rate": 0.0015,
      "loss": 2.6383,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.9882407784461975,
      "learning_rate": 0.0015,
      "loss": 2.6158,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.8950514793395996,
      "learning_rate": 0.0015,
      "loss": 2.5936,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 1.250455379486084,
      "learning_rate": 0.0015,
      "loss": 2.5668,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.9550969004631042,
      "learning_rate": 0.0015,
      "loss": 2.5322,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.9048781991004944,
      "learning_rate": 0.0015,
      "loss": 2.531,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.760736882686615,
      "learning_rate": 0.0015,
      "loss": 2.4979,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.8057079315185547,
      "learning_rate": 0.0015,
      "loss": 2.4602,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.8271151185035706,
      "learning_rate": 0.0015,
      "loss": 2.4384,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.7715069651603699,
      "learning_rate": 0.0015,
      "loss": 2.4289,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.8445009589195251,
      "learning_rate": 0.0015,
      "loss": 2.4145,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.9835455417633057,
      "learning_rate": 0.0015,
      "loss": 2.393,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.8982846736907959,
      "learning_rate": 0.0015,
      "loss": 2.3699,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.8850642442703247,
      "learning_rate": 0.0015,
      "loss": 2.3584,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 1.1749838590621948,
      "learning_rate": 0.0015,
      "loss": 2.3433,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.757171630859375,
      "learning_rate": 0.0015,
      "loss": 2.3133,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 1.2953829765319824,
      "learning_rate": 0.0015,
      "loss": 2.2933,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.9513627290725708,
      "learning_rate": 0.0015,
      "loss": 2.2979,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.8289362192153931,
      "learning_rate": 0.0015,
      "loss": 2.2778,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.9080241322517395,
      "learning_rate": 0.0015,
      "loss": 2.2616,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.9738650918006897,
      "learning_rate": 0.0015,
      "loss": 2.2516,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.7926310896873474,
      "learning_rate": 0.0015,
      "loss": 2.2263,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 1.1924057006835938,
      "learning_rate": 0.0015,
      "loss": 2.2217,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.7833127379417419,
      "learning_rate": 0.0015,
      "loss": 2.2064,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.8512415289878845,
      "learning_rate": 0.0015,
      "loss": 2.1885,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.747489869594574,
      "learning_rate": 0.0015,
      "loss": 2.177,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.8559404015541077,
      "learning_rate": 0.0015,
      "loss": 2.1614,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.838374674320221,
      "learning_rate": 0.0015,
      "loss": 2.1638,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.9322778582572937,
      "learning_rate": 0.0015,
      "loss": 2.1268,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.8562554717063904,
      "learning_rate": 0.0015,
      "loss": 2.1294,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 1.2495431900024414,
      "learning_rate": 0.0015,
      "loss": 2.1357,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.8939051628112793,
      "learning_rate": 0.0015,
      "loss": 2.108,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.1370033025741577,
      "learning_rate": 0.0015,
      "loss": 2.0899,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.8906479477882385,
      "learning_rate": 0.0015,
      "loss": 2.0908,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.7592072486877441,
      "learning_rate": 0.0015,
      "loss": 2.08,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.8847842812538147,
      "learning_rate": 0.0015,
      "loss": 2.0626,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.80992192029953,
      "learning_rate": 0.0015,
      "loss": 2.0758,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.7670337557792664,
      "learning_rate": 0.0015,
      "loss": 2.0681,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 1.0644383430480957,
      "learning_rate": 0.0015,
      "loss": 2.0542,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.802211344242096,
      "learning_rate": 0.0015,
      "loss": 2.0361,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.7406486868858337,
      "learning_rate": 0.0015,
      "loss": 2.025,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7359614372253418,
      "learning_rate": 0.0015,
      "loss": 2.0367,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.8403868079185486,
      "learning_rate": 0.0015,
      "loss": 2.0209,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.7695398926734924,
      "learning_rate": 0.0015,
      "loss": 2.012,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.7453827261924744,
      "learning_rate": 0.0015,
      "loss": 1.9922,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.8518917560577393,
      "learning_rate": 0.0015,
      "loss": 1.9842,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.8525162935256958,
      "learning_rate": 0.0015,
      "loss": 1.9939,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.9009069800376892,
      "learning_rate": 0.0015,
      "loss": 1.9782,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.3110953569412231,
      "learning_rate": 0.0015,
      "loss": 1.978,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.8284780979156494,
      "learning_rate": 0.0015,
      "loss": 1.9659,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.8100257515907288,
      "learning_rate": 0.0015,
      "loss": 1.9651,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.7772668600082397,
      "learning_rate": 0.0015,
      "loss": 1.9699,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.6757662892341614,
      "learning_rate": 0.0015,
      "loss": 1.943,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.7380781769752502,
      "learning_rate": 0.0015,
      "loss": 1.9317,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 1.0726897716522217,
      "learning_rate": 0.0015,
      "loss": 1.941,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.9912097454071045,
      "learning_rate": 0.0015,
      "loss": 1.9426,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.8630762100219727,
      "learning_rate": 0.0015,
      "loss": 1.9223,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.7220410704612732,
      "learning_rate": 0.0015,
      "loss": 1.9236,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.8265845775604248,
      "learning_rate": 0.0015,
      "loss": 1.9192,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.8414561748504639,
      "learning_rate": 0.0015,
      "loss": 1.9174,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.8560569286346436,
      "learning_rate": 0.0015,
      "loss": 1.9101,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.9429737329483032,
      "learning_rate": 0.0015,
      "loss": 1.9028,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.9511808753013611,
      "learning_rate": 0.0015,
      "loss": 1.8982,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.7845291495323181,
      "learning_rate": 0.0015,
      "loss": 1.9041,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.7083245515823364,
      "learning_rate": 0.0015,
      "loss": 1.8959,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.695321798324585,
      "learning_rate": 0.0015,
      "loss": 1.8762,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.7678042650222778,
      "learning_rate": 0.0015,
      "loss": 1.8839,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.8387939929962158,
      "learning_rate": 0.0015,
      "loss": 1.8689,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.7231588363647461,
      "learning_rate": 0.0015,
      "loss": 1.8725,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 1.4515342712402344,
      "learning_rate": 0.0015,
      "loss": 1.8773,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 1.2844597101211548,
      "learning_rate": 0.0015,
      "loss": 1.8664,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 1.171708583831787,
      "learning_rate": 0.0015,
      "loss": 1.8612,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.9919615387916565,
      "learning_rate": 0.0015,
      "loss": 1.8619,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 1.0693222284317017,
      "learning_rate": 0.0015,
      "loss": 1.8548,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.8183884620666504,
      "learning_rate": 0.0015,
      "loss": 1.8415,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.7043951153755188,
      "learning_rate": 0.0015,
      "loss": 1.8463,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.7884238958358765,
      "learning_rate": 0.0015,
      "loss": 1.8401,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 1.0709902048110962,
      "learning_rate": 0.0015,
      "loss": 1.8438,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.7990682125091553,
      "learning_rate": 0.0015,
      "loss": 1.8367,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.7052449584007263,
      "learning_rate": 0.0015,
      "loss": 1.8247,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 1.0053858757019043,
      "learning_rate": 0.0015,
      "loss": 1.8223,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.6969329118728638,
      "learning_rate": 0.0015,
      "loss": 1.8329,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.7716851830482483,
      "learning_rate": 0.0015,
      "loss": 1.817,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.6982385516166687,
      "learning_rate": 0.0015,
      "loss": 1.8253,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.7668865919113159,
      "learning_rate": 0.0015,
      "loss": 1.8193,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.9826824069023132,
      "learning_rate": 0.0015,
      "loss": 1.8032,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.783207356929779,
      "learning_rate": 0.0015,
      "loss": 1.8184,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.2691607475280762,
      "learning_rate": 0.0015,
      "loss": 1.802,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.7726699709892273,
      "learning_rate": 0.0015,
      "loss": 1.809,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 1.0477081537246704,
      "learning_rate": 0.0015,
      "loss": 1.7988,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.6980409622192383,
      "learning_rate": 0.0015,
      "loss": 1.7825,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.7194324135780334,
      "learning_rate": 0.0015,
      "loss": 1.7873,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 1.509351372718811,
      "learning_rate": 0.0015,
      "loss": 1.805,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.8106460571289062,
      "learning_rate": 0.0015,
      "loss": 1.802,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.6423959732055664,
      "learning_rate": 0.0015,
      "loss": 1.7724,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.7225367426872253,
      "learning_rate": 0.0015,
      "loss": 1.7629,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.7591999173164368,
      "learning_rate": 0.0015,
      "loss": 1.7846,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.6502033472061157,
      "learning_rate": 0.0015,
      "loss": 1.7852,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.8119486570358276,
      "learning_rate": 0.0015,
      "loss": 1.7717,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.7183853387832642,
      "learning_rate": 0.0015,
      "loss": 1.7686,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.6890498399734497,
      "learning_rate": 0.0015,
      "loss": 1.7822,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.6092765927314758,
      "learning_rate": 0.0015,
      "loss": 1.7677,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.7149849534034729,
      "learning_rate": 0.0015,
      "loss": 1.754,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.6687565445899963,
      "learning_rate": 0.0015,
      "loss": 1.7578,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.8480717539787292,
      "learning_rate": 0.0015,
      "loss": 1.7625,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 0.9097123146057129,
      "learning_rate": 0.0015,
      "loss": 1.7542,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.6868681907653809,
      "learning_rate": 0.0015,
      "loss": 1.758,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.7010665535926819,
      "learning_rate": 0.0015,
      "loss": 1.7611,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.7429423928260803,
      "learning_rate": 0.0015,
      "loss": 1.7483,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.6833866834640503,
      "learning_rate": 0.0015,
      "loss": 1.7467,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.7524899840354919,
      "learning_rate": 0.0015,
      "loss": 1.7376,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.8158894181251526,
      "learning_rate": 0.0015,
      "loss": 1.7483,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.8951417803764343,
      "learning_rate": 0.0015,
      "loss": 1.7334,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.673397421836853,
      "learning_rate": 0.0015,
      "loss": 1.7334,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.7225199341773987,
      "learning_rate": 0.0015,
      "loss": 1.7397,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.6290916204452515,
      "learning_rate": 0.0015,
      "loss": 1.7346,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.7073165774345398,
      "learning_rate": 0.0015,
      "loss": 1.7292,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.7441956996917725,
      "learning_rate": 0.0015,
      "loss": 1.7273,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.8966643214225769,
      "learning_rate": 0.0015,
      "loss": 1.7261,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.635562539100647,
      "learning_rate": 0.0015,
      "loss": 1.7272,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.8800082802772522,
      "learning_rate": 0.0015,
      "loss": 1.7286,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.9613121151924133,
      "learning_rate": 0.0015,
      "loss": 1.7221,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.6964594125747681,
      "learning_rate": 0.0015,
      "loss": 1.7203,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.6780398488044739,
      "learning_rate": 0.0015,
      "loss": 1.7212,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.6919646859169006,
      "learning_rate": 0.0015,
      "loss": 1.7235,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.7676806449890137,
      "learning_rate": 0.0015,
      "loss": 1.7042,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.735536515712738,
      "learning_rate": 0.0015,
      "loss": 1.6988,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 1.2225226163864136,
      "learning_rate": 0.0015,
      "loss": 1.7015,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.678997814655304,
      "learning_rate": 0.0015,
      "loss": 1.701,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.8650702834129333,
      "learning_rate": 0.0015,
      "loss": 1.6948,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.8140895962715149,
      "learning_rate": 0.0015,
      "loss": 1.7059,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.8755738735198975,
      "learning_rate": 0.0015,
      "loss": 1.6993,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.8179135918617249,
      "learning_rate": 0.0015,
      "loss": 1.7067,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.6715182662010193,
      "learning_rate": 0.0015,
      "loss": 1.7054,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6825255751609802,
      "learning_rate": 0.0015,
      "loss": 1.6739,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 1.0979033708572388,
      "learning_rate": 0.0015,
      "loss": 1.6866,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.6127614974975586,
      "learning_rate": 0.0015,
      "loss": 1.6902,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.63741534948349,
      "learning_rate": 0.0015,
      "loss": 1.6777,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.8913732767105103,
      "learning_rate": 0.0015,
      "loss": 1.6787,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.7328352332115173,
      "learning_rate": 0.0015,
      "loss": 1.6912,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.4312113523483276,
      "learning_rate": 0.0015,
      "loss": 1.6781,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 1.2871997356414795,
      "learning_rate": 0.0015,
      "loss": 1.6782,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.7266455888748169,
      "learning_rate": 0.0015,
      "loss": 1.6847,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.7218185663223267,
      "learning_rate": 0.0015,
      "loss": 1.665,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.6649265885353088,
      "learning_rate": 0.0015,
      "loss": 1.6629,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.6617045998573303,
      "learning_rate": 0.0015,
      "loss": 1.6654,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.5968663096427917,
      "learning_rate": 0.0015,
      "loss": 1.6699,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.6206800937652588,
      "learning_rate": 0.0015,
      "loss": 1.6901,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.689408540725708,
      "learning_rate": 0.0015,
      "loss": 1.6696,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.7100616097450256,
      "learning_rate": 0.0015,
      "loss": 1.6755,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.6594061851501465,
      "learning_rate": 0.0015,
      "loss": 1.6732,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.9528148770332336,
      "learning_rate": 0.0015,
      "loss": 1.6732,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.7268603444099426,
      "learning_rate": 0.0015,
      "loss": 1.6679,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.616661012172699,
      "learning_rate": 0.0015,
      "loss": 1.6546,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.6803067326545715,
      "learning_rate": 0.0015,
      "loss": 1.6551,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.6617722511291504,
      "learning_rate": 0.0015,
      "loss": 1.6542,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.6185517311096191,
      "learning_rate": 0.0015,
      "loss": 1.6518,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.6060194373130798,
      "learning_rate": 0.0015,
      "loss": 1.6553,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.6753437519073486,
      "learning_rate": 0.0015,
      "loss": 1.661,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.632571816444397,
      "learning_rate": 0.0015,
      "loss": 1.6577,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6047013998031616,
      "learning_rate": 0.0015,
      "loss": 1.648,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.5847437977790833,
      "learning_rate": 0.0015,
      "loss": 1.6536,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.9869843125343323,
      "learning_rate": 0.0015,
      "loss": 1.6487,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.8873304724693298,
      "learning_rate": 0.0015,
      "loss": 1.6506,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.6672500371932983,
      "learning_rate": 0.0015,
      "loss": 1.6478,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.7043957710266113,
      "learning_rate": 0.0015,
      "loss": 1.6483,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.6475924253463745,
      "learning_rate": 0.0015,
      "loss": 1.6394,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.6054782271385193,
      "learning_rate": 0.0015,
      "loss": 1.638,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.7016323804855347,
      "learning_rate": 0.0015,
      "loss": 1.6435,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.6046034693717957,
      "learning_rate": 0.0015,
      "loss": 1.6532,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.6938050985336304,
      "learning_rate": 0.0015,
      "loss": 1.6316,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.6536348462104797,
      "learning_rate": 0.0015,
      "loss": 1.6318,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.6942468881607056,
      "learning_rate": 0.0015,
      "loss": 1.6292,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.851248025894165,
      "learning_rate": 0.0015,
      "loss": 1.6386,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.8210561275482178,
      "learning_rate": 0.0015,
      "loss": 1.6492,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.6815303564071655,
      "learning_rate": 0.0015,
      "loss": 1.6369,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6996192336082458,
      "learning_rate": 0.0015,
      "loss": 1.6333,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.6134830713272095,
      "learning_rate": 0.0015,
      "loss": 1.63,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.7721474170684814,
      "learning_rate": 0.0015,
      "loss": 1.6224,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.7255933284759521,
      "learning_rate": 0.0015,
      "loss": 1.6244,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.6715564131736755,
      "learning_rate": 0.0015,
      "loss": 1.6314,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.8546271920204163,
      "learning_rate": 0.0015,
      "loss": 1.6219,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 1.7003231048583984,
      "learning_rate": 0.0015,
      "loss": 1.6351,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.7009464502334595,
      "learning_rate": 0.0015,
      "loss": 1.628,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6949298977851868,
      "learning_rate": 0.0015,
      "loss": 1.6109,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.6207330822944641,
      "learning_rate": 0.0015,
      "loss": 1.6177,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6025390625,
      "learning_rate": 0.0015,
      "loss": 1.612,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.9670878648757935,
      "learning_rate": 0.0015,
      "loss": 1.6188,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.022936224937439,
      "learning_rate": 0.0015,
      "loss": 1.6342,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.6394636034965515,
      "learning_rate": 0.0015,
      "loss": 1.6203,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.5926973819732666,
      "learning_rate": 0.0015,
      "loss": 1.6028,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.8295377492904663,
      "learning_rate": 0.0015,
      "loss": 1.6196,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.6857534050941467,
      "learning_rate": 0.0015,
      "loss": 1.621,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.8938616514205933,
      "learning_rate": 0.0015,
      "loss": 1.6086,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.5973939299583435,
      "learning_rate": 0.0015,
      "loss": 1.619,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.7303853631019592,
      "learning_rate": 0.0015,
      "loss": 1.6091,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.6592551469802856,
      "learning_rate": 0.0015,
      "loss": 1.6091,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.8553382158279419,
      "learning_rate": 0.0015,
      "loss": 1.607,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.8838529586791992,
      "learning_rate": 0.0015,
      "loss": 1.5974,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.8019864559173584,
      "learning_rate": 0.0015,
      "loss": 1.6278,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 1.0666486024856567,
      "learning_rate": 0.0015,
      "loss": 1.6178,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.661801815032959,
      "learning_rate": 0.0015,
      "loss": 1.6037,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.6086870431900024,
      "learning_rate": 0.0015,
      "loss": 1.5978,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.6623057126998901,
      "learning_rate": 0.0015,
      "loss": 1.6019,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.7000704407691956,
      "learning_rate": 0.0015,
      "loss": 1.6117,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.6461548209190369,
      "learning_rate": 0.0015,
      "loss": 1.5955,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.6932266354560852,
      "learning_rate": 0.0015,
      "loss": 1.5893,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.6384739279747009,
      "learning_rate": 0.0015,
      "loss": 1.5907,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.907963216304779,
      "learning_rate": 0.0015,
      "loss": 1.6026,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.9056057333946228,
      "learning_rate": 0.0015,
      "loss": 1.5964,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.786928653717041,
      "learning_rate": 0.0015,
      "loss": 1.5951,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.7156978249549866,
      "learning_rate": 0.0015,
      "loss": 1.5949,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6144447922706604,
      "learning_rate": 0.0015,
      "loss": 1.5931,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.677020788192749,
      "learning_rate": 0.0015,
      "loss": 1.5953,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.6892838478088379,
      "learning_rate": 0.0015,
      "loss": 1.5909,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.5918874740600586,
      "learning_rate": 0.0015,
      "loss": 1.5913,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.7120735049247742,
      "learning_rate": 0.0015,
      "loss": 1.5995,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.6630368828773499,
      "learning_rate": 0.0015,
      "loss": 1.5828,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.5932007431983948,
      "learning_rate": 0.0015,
      "loss": 1.5982,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.7960039973258972,
      "learning_rate": 0.0015,
      "loss": 1.5893,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.7096516489982605,
      "learning_rate": 0.0015,
      "loss": 1.5952,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.6117181777954102,
      "learning_rate": 0.0015,
      "loss": 1.5803,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.6103050708770752,
      "learning_rate": 0.0015,
      "loss": 1.5856,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.5880393385887146,
      "learning_rate": 0.0015,
      "loss": 1.5833,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 1.1202528476715088,
      "learning_rate": 0.0015,
      "loss": 1.5885,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.6866726875305176,
      "learning_rate": 0.0015,
      "loss": 1.5851,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.7937163710594177,
      "learning_rate": 0.0015,
      "loss": 1.5867,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.7862516045570374,
      "learning_rate": 0.0015,
      "loss": 1.5784,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.5781798362731934,
      "learning_rate": 0.0015,
      "loss": 1.587,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.6267972588539124,
      "learning_rate": 0.0015,
      "loss": 1.5819,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.5809932351112366,
      "learning_rate": 0.0015,
      "loss": 1.5782,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.5802158713340759,
      "learning_rate": 0.0015,
      "loss": 1.5877,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.6015350222587585,
      "learning_rate": 0.0015,
      "loss": 1.5869,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6029126644134521,
      "learning_rate": 0.0015,
      "loss": 1.5837,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.7008916139602661,
      "learning_rate": 0.0015,
      "loss": 1.5816,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.7247995734214783,
      "learning_rate": 0.0015,
      "loss": 1.5706,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.6409516334533691,
      "learning_rate": 0.0015,
      "loss": 1.5824,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.7362030148506165,
      "learning_rate": 0.0015,
      "loss": 1.5755,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.5998350381851196,
      "learning_rate": 0.0015,
      "loss": 1.5679,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.6767162680625916,
      "learning_rate": 0.0015,
      "loss": 1.57,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.5803014636039734,
      "learning_rate": 0.0015,
      "loss": 1.5669,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.5653269290924072,
      "learning_rate": 0.0015,
      "loss": 1.5688,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6908811926841736,
      "learning_rate": 0.0015,
      "loss": 1.5667,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.6035429239273071,
      "learning_rate": 0.0015,
      "loss": 1.5727,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.5633019804954529,
      "learning_rate": 0.0015,
      "loss": 1.5729,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.5946157574653625,
      "learning_rate": 0.0015,
      "loss": 1.5554,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.6822995543479919,
      "learning_rate": 0.0015,
      "loss": 1.5669,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.6577224731445312,
      "learning_rate": 0.0015,
      "loss": 1.5732,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 1.026064157485962,
      "learning_rate": 0.0015,
      "loss": 1.5786,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6183416843414307,
      "learning_rate": 0.0015,
      "loss": 1.5753,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.6068280339241028,
      "learning_rate": 0.0015,
      "loss": 1.5717,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.5816817283630371,
      "learning_rate": 0.0015,
      "loss": 1.5693,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.5484740138053894,
      "learning_rate": 0.0015,
      "loss": 1.5636,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.7095575332641602,
      "learning_rate": 0.0015,
      "loss": 1.5668,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6714445948600769,
      "learning_rate": 0.0015,
      "loss": 1.5712,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.790642499923706,
      "learning_rate": 0.0015,
      "loss": 1.5461,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 1.5911504030227661,
      "learning_rate": 0.0015,
      "loss": 1.5537,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.6780908107757568,
      "learning_rate": 0.0015,
      "loss": 1.5637,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.5470073223114014,
      "learning_rate": 0.0015,
      "loss": 1.5576,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.8236953616142273,
      "learning_rate": 0.0015,
      "loss": 1.5609,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.6634783148765564,
      "learning_rate": 0.0015,
      "loss": 1.5457,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.6497632265090942,
      "learning_rate": 0.0015,
      "loss": 1.5536,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.8404585123062134,
      "learning_rate": 0.0015,
      "loss": 1.5596,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.8150055408477783,
      "learning_rate": 0.0015,
      "loss": 1.5592,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.6306717991828918,
      "learning_rate": 0.0015,
      "loss": 1.5537,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.5247840285301208,
      "learning_rate": 0.0015,
      "loss": 1.5509,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.6982070207595825,
      "learning_rate": 0.0015,
      "loss": 1.5565,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.695550799369812,
      "learning_rate": 0.0015,
      "loss": 1.5651,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.726571261882782,
      "learning_rate": 0.0015,
      "loss": 1.5493,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.9273520708084106,
      "learning_rate": 0.0015,
      "loss": 1.5405,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.7275087237358093,
      "learning_rate": 0.0015,
      "loss": 1.5462,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.6150531768798828,
      "learning_rate": 0.0015,
      "loss": 1.5628,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.9108219742774963,
      "learning_rate": 0.0015,
      "loss": 1.5418,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.5648165345191956,
      "learning_rate": 0.0015,
      "loss": 1.5507,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.6469804644584656,
      "learning_rate": 0.0015,
      "loss": 1.5505,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6678853034973145,
      "learning_rate": 0.0015,
      "loss": 1.5379,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.8343275189399719,
      "learning_rate": 0.0015,
      "loss": 1.5468,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.6772765517234802,
      "learning_rate": 0.0015,
      "loss": 1.5428,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.6098471879959106,
      "learning_rate": 0.0015,
      "loss": 1.5542,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.5519450902938843,
      "learning_rate": 0.0015,
      "loss": 1.5434,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.6524561643600464,
      "learning_rate": 0.0015,
      "loss": 1.532,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.5557029843330383,
      "learning_rate": 0.0015,
      "loss": 1.5502,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.5645538568496704,
      "learning_rate": 0.0015,
      "loss": 1.5349,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6528288125991821,
      "learning_rate": 0.0015,
      "loss": 1.5418,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.6994026303291321,
      "learning_rate": 0.0015,
      "loss": 1.5416,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.7095736861228943,
      "learning_rate": 0.0015,
      "loss": 1.5555,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.7259683012962341,
      "learning_rate": 0.0015,
      "loss": 1.5329,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5394166111946106,
      "learning_rate": 0.0015,
      "loss": 1.5391,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5862396359443665,
      "learning_rate": 0.0015,
      "loss": 1.5411,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.6594740748405457,
      "learning_rate": 0.0015,
      "loss": 1.5448,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.7697080373764038,
      "learning_rate": 0.0015,
      "loss": 1.536,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.8687046766281128,
      "learning_rate": 0.0015,
      "loss": 1.5373,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.811188280582428,
      "learning_rate": 0.0015,
      "loss": 1.5353,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 1.0547219514846802,
      "learning_rate": 0.0015,
      "loss": 1.5487,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.7948180437088013,
      "learning_rate": 0.0015,
      "loss": 1.539,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.7937295436859131,
      "learning_rate": 0.0015,
      "loss": 1.5334,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.5533401966094971,
      "learning_rate": 0.0015,
      "loss": 1.5143,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.7152353525161743,
      "learning_rate": 0.0015,
      "loss": 1.5397,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.562127947807312,
      "learning_rate": 0.0015,
      "loss": 1.5255,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.5752823352813721,
      "learning_rate": 0.0015,
      "loss": 1.5469,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.6228840947151184,
      "learning_rate": 0.0015,
      "loss": 1.5452,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.5474838614463806,
      "learning_rate": 0.0015,
      "loss": 1.5247,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.5703747868537903,
      "learning_rate": 0.0015,
      "loss": 1.5351,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.5856843590736389,
      "learning_rate": 0.0015,
      "loss": 1.5367,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.6777088046073914,
      "learning_rate": 0.0015,
      "loss": 1.5409,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.5606372952461243,
      "learning_rate": 0.0015,
      "loss": 1.5283,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.6570095419883728,
      "learning_rate": 0.0015,
      "loss": 1.5311,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6104706525802612,
      "learning_rate": 0.0015,
      "loss": 1.5253,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.5908493399620056,
      "learning_rate": 0.0015,
      "loss": 1.5245,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.7394626140594482,
      "learning_rate": 0.0015,
      "loss": 1.5312,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.9077485203742981,
      "learning_rate": 0.0015,
      "loss": 1.5302,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.6131513714790344,
      "learning_rate": 0.0015,
      "loss": 1.5228,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.5888971090316772,
      "learning_rate": 0.0015,
      "loss": 1.5312,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.5609570741653442,
      "learning_rate": 0.0015,
      "loss": 1.5268,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.6785055994987488,
      "learning_rate": 0.0015,
      "loss": 1.5079,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.5453798770904541,
      "learning_rate": 0.0015,
      "loss": 1.528,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.5888906717300415,
      "learning_rate": 0.0015,
      "loss": 1.5229,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.5770800709724426,
      "learning_rate": 0.0015,
      "loss": 1.517,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.6824765801429749,
      "learning_rate": 0.0015,
      "loss": 1.5134,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.6052685379981995,
      "learning_rate": 0.0015,
      "loss": 1.5321,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.5701059103012085,
      "learning_rate": 0.0015,
      "loss": 1.5366,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.5969727039337158,
      "learning_rate": 0.0015,
      "loss": 1.5283,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.5816425085067749,
      "learning_rate": 0.0015,
      "loss": 1.5096,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.8447874784469604,
      "learning_rate": 0.0015,
      "loss": 1.5292,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.6487998366355896,
      "learning_rate": 0.0015,
      "loss": 1.5269,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.666244387626648,
      "learning_rate": 0.0015,
      "loss": 1.5315,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.7151220440864563,
      "learning_rate": 0.0015,
      "loss": 1.5314,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.5529572367668152,
      "learning_rate": 0.0015,
      "loss": 1.5168,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.5668327808380127,
      "learning_rate": 0.0015,
      "loss": 1.5159,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.6934638619422913,
      "learning_rate": 0.0015,
      "loss": 1.5137,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.5961264371871948,
      "learning_rate": 0.0015,
      "loss": 1.5157,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.618583619594574,
      "learning_rate": 0.0015,
      "loss": 1.5261,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.8063673377037048,
      "learning_rate": 0.0015,
      "loss": 1.5092,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.6244291663169861,
      "learning_rate": 0.0015,
      "loss": 1.5123,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.607018232345581,
      "learning_rate": 0.0015,
      "loss": 1.5119,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.5306568145751953,
      "learning_rate": 0.0015,
      "loss": 1.5133,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.7277970314025879,
      "learning_rate": 0.0015,
      "loss": 1.5168,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.8040258288383484,
      "learning_rate": 0.0015,
      "loss": 1.5239,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.6845293641090393,
      "learning_rate": 0.0015,
      "loss": 1.5159,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.5492376685142517,
      "learning_rate": 0.0015,
      "loss": 1.5093,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.7942477464675903,
      "learning_rate": 0.0015,
      "loss": 1.5101,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.5903634428977966,
      "learning_rate": 0.0015,
      "loss": 1.5232,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.6824354529380798,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.5715729594230652,
      "learning_rate": 0.0015,
      "loss": 1.5121,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.6855105757713318,
      "learning_rate": 0.0015,
      "loss": 1.4911,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.5223708152770996,
      "learning_rate": 0.0015,
      "loss": 1.5102,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.6070655584335327,
      "learning_rate": 0.0015,
      "loss": 1.5094,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.6628040075302124,
      "learning_rate": 0.0015,
      "loss": 1.5132,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.8938891887664795,
      "learning_rate": 0.0015,
      "loss": 1.5041,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.599801242351532,
      "learning_rate": 0.0015,
      "loss": 1.5211,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.5351318717002869,
      "learning_rate": 0.0015,
      "loss": 1.5076,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.7043532729148865,
      "learning_rate": 0.0015,
      "loss": 1.5123,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.5856007933616638,
      "learning_rate": 0.0015,
      "loss": 1.5153,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.5364360809326172,
      "learning_rate": 0.0015,
      "loss": 1.5032,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.6079488396644592,
      "learning_rate": 0.0015,
      "loss": 1.4981,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.552480161190033,
      "learning_rate": 0.0015,
      "loss": 1.4991,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.545010507106781,
      "learning_rate": 0.0015,
      "loss": 1.5184,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.5817381739616394,
      "learning_rate": 0.0015,
      "loss": 1.4983,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.7624457478523254,
      "learning_rate": 0.0015,
      "loss": 1.5142,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.8512298464775085,
      "learning_rate": 0.0015,
      "loss": 1.5034,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.5915345549583435,
      "learning_rate": 0.0015,
      "loss": 1.493,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.6914456486701965,
      "learning_rate": 0.0015,
      "loss": 1.4918,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.7696536779403687,
      "learning_rate": 0.0015,
      "loss": 1.4921,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.5909621119499207,
      "learning_rate": 0.0015,
      "loss": 1.4974,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.5885836482048035,
      "learning_rate": 0.0015,
      "loss": 1.4959,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.5148940682411194,
      "learning_rate": 0.0015,
      "loss": 1.5002,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.6414653658866882,
      "learning_rate": 0.0015,
      "loss": 1.4957,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.5845354199409485,
      "learning_rate": 0.0015,
      "loss": 1.5019,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.5267220735549927,
      "learning_rate": 0.0015,
      "loss": 1.5018,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.6976961493492126,
      "learning_rate": 0.0015,
      "loss": 1.4917,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.5018602013587952,
      "learning_rate": 0.0015,
      "loss": 1.5016,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.5722861289978027,
      "learning_rate": 0.0015,
      "loss": 1.4951,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.675983726978302,
      "learning_rate": 0.0015,
      "loss": 1.508,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.587409496307373,
      "learning_rate": 0.0015,
      "loss": 1.4915,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.5497817993164062,
      "learning_rate": 0.0015,
      "loss": 1.4969,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.6104870438575745,
      "learning_rate": 0.0015,
      "loss": 1.4905,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5707203149795532,
      "learning_rate": 0.0015,
      "loss": 1.5004,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5172650814056396,
      "learning_rate": 0.0015,
      "loss": 1.4971,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.8300213813781738,
      "learning_rate": 0.0015,
      "loss": 1.5022,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.6693031191825867,
      "learning_rate": 0.0015,
      "loss": 1.4992,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.72288578748703,
      "learning_rate": 0.0015,
      "loss": 1.5054,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.8750916123390198,
      "learning_rate": 0.0015,
      "loss": 1.4995,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5264197587966919,
      "learning_rate": 0.0015,
      "loss": 1.4956,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.6223256587982178,
      "learning_rate": 0.0015,
      "loss": 1.4905,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.6839354634284973,
      "learning_rate": 0.0015,
      "loss": 1.4953,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.5767725706100464,
      "learning_rate": 0.0015,
      "loss": 1.4839,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5760086178779602,
      "learning_rate": 0.0015,
      "loss": 1.4905,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.7305461168289185,
      "learning_rate": 0.0015,
      "loss": 1.4961,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.6796057820320129,
      "learning_rate": 0.0015,
      "loss": 1.4871,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.5641124248504639,
      "learning_rate": 0.0015,
      "loss": 1.488,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.531761109828949,
      "learning_rate": 0.0015,
      "loss": 1.4857,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.5290331840515137,
      "learning_rate": 0.0015,
      "loss": 1.4902,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.5988660454750061,
      "learning_rate": 0.0015,
      "loss": 1.4927,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.5810370445251465,
      "learning_rate": 0.0015,
      "loss": 1.4819,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.6254767775535583,
      "learning_rate": 0.0015,
      "loss": 1.4928,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.8546009063720703,
      "learning_rate": 0.0015,
      "loss": 1.4907,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.6996669769287109,
      "learning_rate": 0.0015,
      "loss": 1.487,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.6641138792037964,
      "learning_rate": 0.0015,
      "loss": 1.4914,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 1.0102465152740479,
      "learning_rate": 0.0015,
      "loss": 1.4977,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5433333516120911,
      "learning_rate": 0.0015,
      "loss": 1.484,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.9442400336265564,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.6479281783103943,
      "learning_rate": 0.0015,
      "loss": 1.4898,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.6578171849250793,
      "learning_rate": 0.0015,
      "loss": 1.4916,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.569848895072937,
      "learning_rate": 0.0015,
      "loss": 1.4855,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.9030901789665222,
      "learning_rate": 0.0015,
      "loss": 1.4831,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.5632250905036926,
      "learning_rate": 0.0015,
      "loss": 1.4906,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.6786665916442871,
      "learning_rate": 0.0015,
      "loss": 1.482,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.7414134740829468,
      "learning_rate": 0.0015,
      "loss": 1.4839,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.6580634117126465,
      "learning_rate": 0.0015,
      "loss": 1.4998,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.6057085394859314,
      "learning_rate": 0.0015,
      "loss": 1.4765,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5455748438835144,
      "learning_rate": 0.0015,
      "loss": 1.4827,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.6651989817619324,
      "learning_rate": 0.0015,
      "loss": 1.4769,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.6850955486297607,
      "learning_rate": 0.0015,
      "loss": 1.4865,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.5547869801521301,
      "learning_rate": 0.0015,
      "loss": 1.4994,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.5541783571243286,
      "learning_rate": 0.0015,
      "loss": 1.4799,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.6864834427833557,
      "learning_rate": 0.0015,
      "loss": 1.4873,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.5392774343490601,
      "learning_rate": 0.0015,
      "loss": 1.4801,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.5740990042686462,
      "learning_rate": 0.0015,
      "loss": 1.4884,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5771586894989014,
      "learning_rate": 0.0015,
      "loss": 1.4743,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.5462093949317932,
      "learning_rate": 0.0015,
      "loss": 1.4815,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.620675265789032,
      "learning_rate": 0.0015,
      "loss": 1.4879,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.5817893147468567,
      "learning_rate": 0.0015,
      "loss": 1.4774,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.598041296005249,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5842782258987427,
      "learning_rate": 0.0015,
      "loss": 1.4849,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.539628803730011,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.5550809502601624,
      "learning_rate": 0.0015,
      "loss": 1.4826,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.5721598863601685,
      "learning_rate": 0.0015,
      "loss": 1.4845,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.5803705453872681,
      "learning_rate": 0.0015,
      "loss": 1.4833,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.5888044238090515,
      "learning_rate": 0.0015,
      "loss": 1.4745,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.5239693522453308,
      "learning_rate": 0.0015,
      "loss": 1.4571,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.6711458563804626,
      "learning_rate": 0.0015,
      "loss": 1.4765,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.6962656378746033,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.6165156960487366,
      "learning_rate": 0.0015,
      "loss": 1.4824,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.8150234818458557,
      "learning_rate": 0.0015,
      "loss": 1.4879,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.5150010585784912,
      "learning_rate": 0.0015,
      "loss": 1.4617,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.5443220734596252,
      "learning_rate": 0.0015,
      "loss": 1.4788,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.482876181602478,
      "learning_rate": 0.0015,
      "loss": 1.4835,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5590141415596008,
      "learning_rate": 0.0015,
      "loss": 1.4839,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.5331428647041321,
      "learning_rate": 0.0015,
      "loss": 1.4707,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.5391295552253723,
      "learning_rate": 0.0015,
      "loss": 1.4725,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.509565532207489,
      "learning_rate": 0.0015,
      "loss": 1.4676,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.618084728717804,
      "learning_rate": 0.0015,
      "loss": 1.4659,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.5926190614700317,
      "learning_rate": 0.0015,
      "loss": 1.474,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.5865287780761719,
      "learning_rate": 0.0015,
      "loss": 1.4752,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.8216726779937744,
      "learning_rate": 0.0015,
      "loss": 1.4728,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.5702864527702332,
      "learning_rate": 0.0015,
      "loss": 1.4675,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.5554937124252319,
      "learning_rate": 0.0015,
      "loss": 1.4687,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.7813419103622437,
      "learning_rate": 0.0015,
      "loss": 1.485,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5859711170196533,
      "learning_rate": 0.0015,
      "loss": 1.4725,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.49484410881996155,
      "learning_rate": 0.0015,
      "loss": 1.4688,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5832527875900269,
      "learning_rate": 0.0015,
      "loss": 1.4781,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.5195610523223877,
      "learning_rate": 0.0015,
      "loss": 1.4628,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.659978449344635,
      "learning_rate": 0.0015,
      "loss": 1.4691,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.511674165725708,
      "learning_rate": 0.0015,
      "loss": 1.4691,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.5177804827690125,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.7505761384963989,
      "learning_rate": 0.0015,
      "loss": 1.4704,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.539092481136322,
      "learning_rate": 0.0015,
      "loss": 1.463,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.5245592594146729,
      "learning_rate": 0.0015,
      "loss": 1.4779,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.5107757449150085,
      "learning_rate": 0.0015,
      "loss": 1.4631,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.6564355492591858,
      "learning_rate": 0.0015,
      "loss": 1.4568,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.8622680902481079,
      "learning_rate": 0.0015,
      "loss": 1.4787,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.7140923142433167,
      "learning_rate": 0.0015,
      "loss": 1.4696,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.5314764976501465,
      "learning_rate": 0.0015,
      "loss": 1.4769,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.703291118144989,
      "learning_rate": 0.0015,
      "loss": 1.4726,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.5653696656227112,
      "learning_rate": 0.0015,
      "loss": 1.4592,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.7576748728752136,
      "learning_rate": 0.0015,
      "loss": 1.4708,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.5658641457557678,
      "learning_rate": 0.0015,
      "loss": 1.4687,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.6822872161865234,
      "learning_rate": 0.0015,
      "loss": 1.4756,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.7656373977661133,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.6432836055755615,
      "learning_rate": 0.0015,
      "loss": 1.4648,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.6013480424880981,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.5194891095161438,
      "learning_rate": 0.0015,
      "loss": 1.4593,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.5971081256866455,
      "learning_rate": 0.0015,
      "loss": 1.4714,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.5031110048294067,
      "learning_rate": 0.0015,
      "loss": 1.4656,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.6080752015113831,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.5187121033668518,
      "learning_rate": 0.0015,
      "loss": 1.4437,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.6215658187866211,
      "learning_rate": 0.0015,
      "loss": 1.4706,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.5294223427772522,
      "learning_rate": 0.0015,
      "loss": 1.4593,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.5327278971672058,
      "learning_rate": 0.0015,
      "loss": 1.4733,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.7155603766441345,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.5754086971282959,
      "learning_rate": 0.0015,
      "loss": 1.4744,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.5307754278182983,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.5501503348350525,
      "learning_rate": 0.0015,
      "loss": 1.4473,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.6310002207756042,
      "learning_rate": 0.0015,
      "loss": 1.4614,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6732348799705505,
      "learning_rate": 0.0015,
      "loss": 1.4604,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.692845344543457,
      "learning_rate": 0.0015,
      "loss": 1.4593,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.8803333044052124,
      "learning_rate": 0.0015,
      "loss": 1.4755,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.7394497394561768,
      "learning_rate": 0.0015,
      "loss": 1.4682,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.5637319684028625,
      "learning_rate": 0.0015,
      "loss": 1.4623,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5626906156539917,
      "learning_rate": 0.0015,
      "loss": 1.4619,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.674372136592865,
      "learning_rate": 0.0015,
      "loss": 1.4734,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.5766422152519226,
      "learning_rate": 0.0015,
      "loss": 1.4565,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.5909757614135742,
      "learning_rate": 0.0015,
      "loss": 1.4676,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.7312908172607422,
      "learning_rate": 0.0015,
      "loss": 1.452,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.5908947587013245,
      "learning_rate": 0.0015,
      "loss": 1.4638,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.715272068977356,
      "learning_rate": 0.0015,
      "loss": 1.455,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.5803365707397461,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.5664737820625305,
      "learning_rate": 0.0015,
      "loss": 1.4545,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.5839225053787231,
      "learning_rate": 0.0015,
      "loss": 1.4602,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.4953925609588623,
      "learning_rate": 0.0015,
      "loss": 1.4659,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.8335318565368652,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.6270488500595093,
      "learning_rate": 0.0015,
      "loss": 1.4612,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.7120893001556396,
      "learning_rate": 0.0015,
      "loss": 1.4638,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.8138550519943237,
      "learning_rate": 0.0015,
      "loss": 1.4496,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.5401758551597595,
      "learning_rate": 0.0015,
      "loss": 1.4483,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.5224828124046326,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.49380800127983093,
      "learning_rate": 0.0015,
      "loss": 1.4516,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.580145537853241,
      "learning_rate": 0.0015,
      "loss": 1.4487,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.7637959122657776,
      "learning_rate": 0.0015,
      "loss": 1.4569,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.6170578002929688,
      "learning_rate": 0.0015,
      "loss": 1.4573,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.49486005306243896,
      "learning_rate": 0.0015,
      "loss": 1.4484,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.7415298223495483,
      "learning_rate": 0.0015,
      "loss": 1.4481,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5081527829170227,
      "learning_rate": 0.0015,
      "loss": 1.4518,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5197334289550781,
      "learning_rate": 0.0015,
      "loss": 1.4541,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.5770809054374695,
      "learning_rate": 0.0015,
      "loss": 1.447,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.5595590472221375,
      "learning_rate": 0.0015,
      "loss": 1.4439,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.5152465105056763,
      "learning_rate": 0.0015,
      "loss": 1.4415,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.5433465838432312,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.5469236373901367,
      "learning_rate": 0.0015,
      "loss": 1.4529,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.7457453012466431,
      "learning_rate": 0.0015,
      "loss": 1.4515,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.8334279656410217,
      "learning_rate": 0.0015,
      "loss": 1.4557,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.5600117444992065,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.6172137260437012,
      "learning_rate": 0.0015,
      "loss": 1.4514,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.8084834814071655,
      "learning_rate": 0.0015,
      "loss": 1.457,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.4771358072757721,
      "learning_rate": 0.0015,
      "loss": 1.4465,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.7505190968513489,
      "learning_rate": 0.0015,
      "loss": 1.4432,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.5385145545005798,
      "learning_rate": 0.0015,
      "loss": 1.4411,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.5953381657600403,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.5023871064186096,
      "learning_rate": 0.0015,
      "loss": 1.4594,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.46754318475723267,
      "learning_rate": 0.0015,
      "loss": 1.4473,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.4978267550468445,
      "learning_rate": 0.0015,
      "loss": 1.4443,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.5876461863517761,
      "learning_rate": 0.0015,
      "loss": 1.4513,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5194767713546753,
      "learning_rate": 0.0015,
      "loss": 1.4447,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.5830156803131104,
      "learning_rate": 0.0015,
      "loss": 1.4489,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.6493149399757385,
      "learning_rate": 0.0015,
      "loss": 1.4573,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.5640086531639099,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.5697202086448669,
      "learning_rate": 0.0015,
      "loss": 1.4607,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.8615607619285583,
      "learning_rate": 0.0015,
      "loss": 1.4543,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.5999366641044617,
      "learning_rate": 0.0015,
      "loss": 1.4532,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.5394980907440186,
      "learning_rate": 0.0015,
      "loss": 1.4571,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5949744582176208,
      "learning_rate": 0.0015,
      "loss": 1.4587,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.8181208372116089,
      "learning_rate": 0.0015,
      "loss": 1.4367,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.7619207501411438,
      "learning_rate": 0.0015,
      "loss": 1.4429,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.5156919360160828,
      "learning_rate": 0.0015,
      "loss": 1.4457,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5499758720397949,
      "learning_rate": 0.0015,
      "loss": 1.4498,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.5424597859382629,
      "learning_rate": 0.0015,
      "loss": 1.4494,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.4815009832382202,
      "learning_rate": 0.0015,
      "loss": 1.445,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.5966301560401917,
      "learning_rate": 0.0015,
      "loss": 1.4216,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5151605606079102,
      "learning_rate": 0.0015,
      "loss": 1.4326,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.6921638250350952,
      "learning_rate": 0.0015,
      "loss": 1.4337,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.46902838349342346,
      "learning_rate": 0.0015,
      "loss": 1.4425,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.5058550834655762,
      "learning_rate": 0.0015,
      "loss": 1.4372,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.5539354085922241,
      "learning_rate": 0.0015,
      "loss": 1.4394,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5441718101501465,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5080265998840332,
      "learning_rate": 0.0015,
      "loss": 1.4596,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.6123008728027344,
      "learning_rate": 0.0015,
      "loss": 1.4438,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.5770524740219116,
      "learning_rate": 0.0015,
      "loss": 1.4542,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.6080031991004944,
      "learning_rate": 0.0015,
      "loss": 1.4609,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.5636224150657654,
      "learning_rate": 0.0015,
      "loss": 1.4467,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.4880855083465576,
      "learning_rate": 0.0015,
      "loss": 1.439,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.5888167023658752,
      "learning_rate": 0.0015,
      "loss": 1.4492,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5131341218948364,
      "learning_rate": 0.0015,
      "loss": 1.4377,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.5880426168441772,
      "learning_rate": 0.0015,
      "loss": 1.4415,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.5205261707305908,
      "learning_rate": 0.0015,
      "loss": 1.4378,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.5249908566474915,
      "learning_rate": 0.0015,
      "loss": 1.4308,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.955223798751831,
      "learning_rate": 0.0015,
      "loss": 1.4398,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.5612366199493408,
      "learning_rate": 0.0015,
      "loss": 1.4457,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.6971794366836548,
      "learning_rate": 0.0015,
      "loss": 1.4354,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.5463219881057739,
      "learning_rate": 0.0015,
      "loss": 1.4409,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5149267315864563,
      "learning_rate": 0.0015,
      "loss": 1.4432,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.4944988787174225,
      "learning_rate": 0.0015,
      "loss": 1.4253,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 1.1925803422927856,
      "learning_rate": 0.0015,
      "loss": 1.4462,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.5493529438972473,
      "learning_rate": 0.0015,
      "loss": 1.4513,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.5455761551856995,
      "learning_rate": 0.0015,
      "loss": 1.4413,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.7629719376564026,
      "learning_rate": 0.0015,
      "loss": 1.4425,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.5936135053634644,
      "learning_rate": 0.0015,
      "loss": 1.4268,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.7498371601104736,
      "learning_rate": 0.0015,
      "loss": 1.4448,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.5150027871131897,
      "learning_rate": 0.0015,
      "loss": 1.4277,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.592757523059845,
      "learning_rate": 0.0015,
      "loss": 1.4471,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.7083083987236023,
      "learning_rate": 0.0015,
      "loss": 1.4379,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.5336272120475769,
      "learning_rate": 0.0015,
      "loss": 1.439,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5165539383888245,
      "learning_rate": 0.0015,
      "loss": 1.4394,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.5278064012527466,
      "learning_rate": 0.0015,
      "loss": 1.4298,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.5715420246124268,
      "learning_rate": 0.0015,
      "loss": 1.4423,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.5232217311859131,
      "learning_rate": 0.0015,
      "loss": 1.437,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.5091473460197449,
      "learning_rate": 0.0015,
      "loss": 1.436,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.8899511098861694,
      "learning_rate": 0.0015,
      "loss": 1.4385,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.6659099459648132,
      "learning_rate": 0.0015,
      "loss": 1.4502,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.5278364419937134,
      "learning_rate": 0.0015,
      "loss": 1.4388,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5966618657112122,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.4178,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.5871420502662659,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.4373,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.49998924136161804,
      "learning_rate": 0.001444828815847542,
      "loss": 1.4403,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.56219881772995,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.4326,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5544670224189758,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.4308,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.6245346069335938,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4363,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.6795433163642883,
      "learning_rate": 0.001366801618799797,
      "loss": 1.421,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5117453336715698,
      "learning_rate": 0.001347962334099052,
      "loss": 1.4258,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.4724520742893219,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4301,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.5391990542411804,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4227,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.5193755030632019,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.429,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.512161135673523,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.4259,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.515159547328949,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.4211,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.6346011757850647,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4143,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.5715067386627197,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4139,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.49045276641845703,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.4086,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.5555257797241211,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4144,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.6634909510612488,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4088,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.5062224864959717,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.41,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.532721757888794,
      "learning_rate": 0.001141156021915355,
      "loss": 1.4108,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.5790921449661255,
      "learning_rate": 0.001125426919103997,
      "loss": 1.391,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.5995064377784729,
      "learning_rate": 0.001109914618088799,
      "loss": 1.3963,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.7425418496131897,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.4084,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.5397632718086243,
      "learning_rate": 0.001079528509501728,
      "loss": 1.4082,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.4879736602306366,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.4022,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.5063172578811646,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.3913,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.4709848165512085,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.3973,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.615486204624176,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3803,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5633809566497803,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.3856,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.545462429523468,
      "learning_rate": 0.000993270988848579,
      "loss": 1.3971,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.4841838777065277,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.3923,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.6843337416648865,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.3959,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.48485007882118225,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.3897,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.5016741752624512,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.4065,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.4839712381362915,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.4018,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.5387133955955505,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.3874,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.5518836975097656,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3771,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.5541582107543945,
      "learning_rate": 0.000888885722567627,
      "loss": 1.3881,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.5158764123916626,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.3812,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.6276925206184387,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3811,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.481979638338089,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.3799,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.48069658875465393,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3746,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.5420559644699097,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3791,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5278546214103699,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3828,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.4805242419242859,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.3695,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.46373307704925537,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3759,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.5256482362747192,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.3712,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.5568493604660034,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.3731,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.5570749640464783,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3683,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.5028032064437866,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3683,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.49775874614715576,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.3706,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.4870249629020691,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.3702,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.5222643613815308,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.3635,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.5388110280036926,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.3656,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.5282334685325623,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.3646,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.5330915451049805,
      "learning_rate": 0.000692383699170611,
      "loss": 1.3565,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.4744960069656372,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3642,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.44814684987068176,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3652,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.5122812390327454,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.3647,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.5803338289260864,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3608,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.5368360877037048,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3631,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.4857064187526703,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3616,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.47593966126441956,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.3643,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.4959394633769989,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3502,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.7511793375015259,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3555,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.5339522361755371,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.3563,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.5237318873405457,
      "learning_rate": 0.000594349404202577,
      "loss": 1.3539,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.609191358089447,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.3444,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.4444977343082428,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.3515,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5894291400909424,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3457,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.5352116227149963,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3428,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.5821011662483215,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.3498,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.5592907667160034,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3406,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.5212944149971008,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3424,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.4783794581890106,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.3661,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.5923948884010315,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.3444,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.4662396311759949,
      "learning_rate": 0.000517326298154212,
      "loss": 1.3504,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.590071976184845,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3542,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.5062225461006165,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3495,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.47687771916389465,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3394,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.5001543164253235,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.3335,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.45192649960517883,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3389,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.47293058037757874,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.3369,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.5003493428230286,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3496,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.4663845896720886,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.3186,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.5916661620140076,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.3355,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.5229008793830872,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3391,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.7030838131904602,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3452,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5374937057495117,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3328,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.5882540941238403,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.3366,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.6345937848091125,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.3421,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.4727292060852051,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3377,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.45928624272346497,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3352,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.48504891991615295,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.3312,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.483320027589798,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.3411,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.5457221269607544,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3351,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.4729653596878052,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.3437,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.4602673351764679,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.3288,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5650603771209717,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.322,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.49608853459358215,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3319,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.5639001131057739,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.3411,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.567976713180542,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.3181,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.45866331458091736,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3192,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.4711613059043884,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.3235,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.5245905518531799,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.3305,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.4977520704269409,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.3432,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.49634262919425964,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3376,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.4820466935634613,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3336,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.5223116874694824,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3302,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.47719502449035645,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.3263,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.4842276871204376,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.3236,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.4913792014122009,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.3264,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.48811912536621094,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.3342,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.5455788969993591,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.3388,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.4813641607761383,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.3215,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.48815804719924927,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.3166,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.4886752963066101,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3228,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.5872246026992798,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3177,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.5208707451820374,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3129,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.4769377112388611,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3223,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.5148775577545166,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3233,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.49648517370224,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3101,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.4622742831707001,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.3196,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.548630952835083,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3279,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.46991637349128723,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.3063,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.4759906530380249,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3126,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.4717313349246979,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.2991,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.49678078293800354,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.3038,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.46686047315597534,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.307,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.4560640752315521,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.3166,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.47867336869239807,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3081,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.4755975902080536,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3151,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.4790399670600891,
      "learning_rate": 0.000237799979917852,
      "loss": 1.3229,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.5322791934013367,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3155,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.5239179134368896,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.3241,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.4789183735847473,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.3076,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.480156272649765,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.3085,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.5658379197120667,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.3054,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.4744018316268921,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.303,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.5904650092124939,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.3057,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.5047197937965393,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3149,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.4757959842681885,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.3123,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.45877909660339355,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.304,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.46163836121559143,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.308,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.4665073752403259,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.3195,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.46690914034843445,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.2942,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.5057730078697205,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.317,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.4762699007987976,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.3071,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.47378572821617126,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.3245,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.48073524236679077,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.3011,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.4430175721645355,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.3064,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5691013932228088,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.3198,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.4707588255405426,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.2948,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.4431469738483429,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.3042,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.5486978888511658,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.3011,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.4700481593608856,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.3071,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.5135657787322998,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3112,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.4729873836040497,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3126,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.49154117703437805,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.2968,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5540194511413574,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.2966,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.49048787355422974,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.3058,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.5284395813941956,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.299,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5401913523674011,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.3161,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4444427788257599,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.303,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.4994732737541199,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.3025,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.5280525088310242,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3077,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.46356168389320374,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.3015,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.44414862990379333,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.2867,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.4924125671386719,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.2947,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.4559844732284546,
      "learning_rate": 0.000142293772767289,
      "loss": 1.2834,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.501009464263916,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.2962,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.48510465025901794,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3165,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.48433762788772583,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.3144,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.4755710959434509,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.2918,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5553343296051025,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.303,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5241166949272156,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.3005,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.4689406156539917,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3093,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.47847384214401245,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.3033,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.4750259816646576,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.3082,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.4762219786643982,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.2906,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.4800224006175995,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.2935,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.46868622303009033,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.2911,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.5961911082267761,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.303,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.50429368019104,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.2978,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.47088149189949036,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.3036,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.47283172607421875,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.2965,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.4766680598258972,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.2947,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.467848002910614,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3078,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.46810755133628845,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.2909,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.4732544720172882,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.2974,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.46181607246398926,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.3011,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4511297941207886,
      "learning_rate": 0.000104851799496331,
      "loss": 1.3039,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.4711008667945862,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.2897,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.4806259572505951,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.2865,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.4904523193836212,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.305,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.4711662828922272,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.2932,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.49776342511177063,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.2937,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.4537979066371918,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.2922,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.4742378890514374,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.2866,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.4839951992034912,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.3003,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.5033496618270874,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.2904,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.5010844469070435,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.2991,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.5384981036186218,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.2903,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.4689342975616455,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.3002,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.47702717781066895,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.2951,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.46512916684150696,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.3005,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.4505891799926758,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.2973,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.46745553612709045,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.2948,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.455531507730484,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3081,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.5059378147125244,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.3006,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.4747602641582489,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.2965,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.461340069770813,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2813,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.5249297618865967,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.2813,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.4644898474216461,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.3058,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.463178426027298,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.2847,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.47134828567504883,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.3032,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.4616844356060028,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.2968,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.460579514503479,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.284,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.4823109209537506,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.2871,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.4757123589515686,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.285,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.4416925013065338,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.2934,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.4845658540725708,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.299,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.4816345274448395,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.2895,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.46081647276878357,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.29,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.5115830898284912,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.3076,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.448098748922348,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.2972,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.4867488443851471,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.2918,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.4599081575870514,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2861,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5270903706550598,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.2936,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.4865814447402954,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.2892,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.4555686414241791,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.2928,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.4486733376979828,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.2861,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.4728319048881531,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.2913,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.4546971023082733,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2811,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.4582926034927368,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.2932,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.4608096778392792,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.2895,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.5004743337631226,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.3038,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.45553311705589294,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.2974,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.4924297630786896,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.2887,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.46293628215789795,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2814,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5086445212364197,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.3001,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.46548640727996826,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.2901,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5457718968391418,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.2968,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.44450920820236206,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.2893,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.5011110901832581,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.2906,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.48211100697517395,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2858,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.4888244867324829,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.2949,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5099353790283203,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2775,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.4793887436389923,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2835,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.5251502394676208,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.2837,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.463763564825058,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.2918,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.43875372409820557,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.3005,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4757961332798004,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.2926,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.46781784296035767,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.2934,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.4844115376472473,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.2895,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.46159836649894714,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.2835,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.4699864983558655,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.2961,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.4685419201850891,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2837,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.48564183712005615,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.2815,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.5339301228523254,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.2711,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4347716271877289,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.2815,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.4856393337249756,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.2884,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.4603196084499359,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2846,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.4456319212913513,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.28,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.45890623331069946,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.3004,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.44558313488960266,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.2872,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.4705146849155426,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.3021,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.46708041429519653,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.286,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.5216341018676758,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.2865,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5233964323997498,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.2856,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.4696406424045563,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.2879,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.4934413433074951,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.2926,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.47221943736076355,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.2917,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.4490326941013336,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2822,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.4612511098384857,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2827,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.46512556076049805,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.2888,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.47384196519851685,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.2848,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.46062570810317993,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.2874,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.4433894455432892,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.2884,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.48196518421173096,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.29,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5478014349937439,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.2874,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5363498330116272,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.2845,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.46773990988731384,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.2899,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.4549839198589325,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.2838,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.4663122594356537,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.282,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.4672456979751587,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.2819,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.48504841327667236,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.3015,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.45739489793777466,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.2792,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.45663920044898987,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.2884,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.4670173227787018,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.2894,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.46393415331840515,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.2916,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.47957542538642883,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.2908,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.4713861346244812,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.2919,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.4453570246696472,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2718,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.46289587020874023,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.284,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.46898356080055237,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.2941,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.4547545611858368,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.2766,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.5211944580078125,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.298,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.4754527807235718,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.2827,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.46048080921173096,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.2915,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.45804089307785034,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.2913,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.44461312890052795,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2863,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.4593360424041748,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2807,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.48727768659591675,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2865,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.4687292277812958,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.2844,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.48308315873146057,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.2853,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.46080321073532104,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.2757,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.4721165597438812,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.2706,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.4544904828071594,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.2828,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.4630659520626068,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.2791,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.4593624770641327,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.2884,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.45893195271492004,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2763,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.4463610351085663,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.2836,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.4578052759170532,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2837,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.4852105379104614,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2834,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.46768200397491455,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2782,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4796695113182068,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.2912,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.46370983123779297,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.2894,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.45639777183532715,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.2986,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.44589945673942566,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.2904,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.4636940658092499,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.29,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.48380720615386963,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2756,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.4814721643924713,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2798,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.44947874546051025,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.2934,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.4699014723300934,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2672,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.46417900919914246,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.2878,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.4559095799922943,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.2876,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.4631689786911011,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.282,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.45717331767082214,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.2925,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.45846042037010193,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.2773,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.4660402238368988,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.2896,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3379178047180176,
      "learning_rate": 1.5020833487815421e-05,
      "loss": 1.28,
      "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
}