{
  "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": 1.2900944948196411,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.4979,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1720187664031982,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.8904,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.850965678691864,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.2363,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.9059324264526367,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.7415,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.6907213926315308,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.2578,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.988332986831665,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.7521,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.29574453830719,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3703,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.2029248476028442,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.1275,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.6928174495697021,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.9265,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 1.0522255897521973,
      "learning_rate": 0.0015,
      "loss": 3.7737,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.721312403678894,
      "learning_rate": 0.0015,
      "loss": 3.6168,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 1.1202422380447388,
      "learning_rate": 0.0015,
      "loss": 3.5,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.6925983428955078,
      "learning_rate": 0.0015,
      "loss": 3.393,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 1.060168743133545,
      "learning_rate": 0.0015,
      "loss": 3.2877,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.7827395796775818,
      "learning_rate": 0.0015,
      "loss": 3.2081,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.7690454721450806,
      "learning_rate": 0.0015,
      "loss": 3.1508,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.7426619529724121,
      "learning_rate": 0.0015,
      "loss": 3.0775,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.7423863410949707,
      "learning_rate": 0.0015,
      "loss": 3.0442,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.7675831913948059,
      "learning_rate": 0.0015,
      "loss": 2.9757,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.923655092716217,
      "learning_rate": 0.0015,
      "loss": 2.9352,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.9558002352714539,
      "learning_rate": 0.0015,
      "loss": 2.8967,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 1.1525710821151733,
      "learning_rate": 0.0015,
      "loss": 2.864,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 1.045106291770935,
      "learning_rate": 0.0015,
      "loss": 2.8201,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.6765981912612915,
      "learning_rate": 0.0015,
      "loss": 2.7772,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 1.1093599796295166,
      "learning_rate": 0.0015,
      "loss": 2.7435,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.9665632843971252,
      "learning_rate": 0.0015,
      "loss": 2.7121,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.7623193860054016,
      "learning_rate": 0.0015,
      "loss": 2.6824,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.9776089191436768,
      "learning_rate": 0.0015,
      "loss": 2.6463,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.9364262819290161,
      "learning_rate": 0.0015,
      "loss": 2.6235,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 1.1343644857406616,
      "learning_rate": 0.0015,
      "loss": 2.6,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.6810590028762817,
      "learning_rate": 0.0015,
      "loss": 2.5703,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.9176515340805054,
      "learning_rate": 0.0015,
      "loss": 2.5367,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.8231906890869141,
      "learning_rate": 0.0015,
      "loss": 2.5325,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.8864274621009827,
      "learning_rate": 0.0015,
      "loss": 2.5042,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 1.045325517654419,
      "learning_rate": 0.0015,
      "loss": 2.4669,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.8459189534187317,
      "learning_rate": 0.0015,
      "loss": 2.4506,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.6683294177055359,
      "learning_rate": 0.0015,
      "loss": 2.4324,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.8342185616493225,
      "learning_rate": 0.0015,
      "loss": 2.4213,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.8517315983772278,
      "learning_rate": 0.0015,
      "loss": 2.4056,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 1.0736851692199707,
      "learning_rate": 0.0015,
      "loss": 2.3738,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.8296298980712891,
      "learning_rate": 0.0015,
      "loss": 2.3685,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.7404698729515076,
      "learning_rate": 0.0015,
      "loss": 2.3413,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.7430567145347595,
      "learning_rate": 0.0015,
      "loss": 2.3199,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 1.1325000524520874,
      "learning_rate": 0.0015,
      "loss": 2.309,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.7666113376617432,
      "learning_rate": 0.0015,
      "loss": 2.3027,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.7912418246269226,
      "learning_rate": 0.0015,
      "loss": 2.2815,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.7484047412872314,
      "learning_rate": 0.0015,
      "loss": 2.2609,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 1.3634499311447144,
      "learning_rate": 0.0015,
      "loss": 2.2574,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.8003369569778442,
      "learning_rate": 0.0015,
      "loss": 2.2351,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.7614564895629883,
      "learning_rate": 0.0015,
      "loss": 2.2242,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.723198413848877,
      "learning_rate": 0.0015,
      "loss": 2.2108,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.8658565878868103,
      "learning_rate": 0.0015,
      "loss": 2.2,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.774768054485321,
      "learning_rate": 0.0015,
      "loss": 2.186,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 1.2603986263275146,
      "learning_rate": 0.0015,
      "loss": 2.1687,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.9657768607139587,
      "learning_rate": 0.0015,
      "loss": 2.1696,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.8478174805641174,
      "learning_rate": 0.0015,
      "loss": 2.1313,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 1.0692206621170044,
      "learning_rate": 0.0015,
      "loss": 2.1391,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 0.95798259973526,
      "learning_rate": 0.0015,
      "loss": 2.1376,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.80659019947052,
      "learning_rate": 0.0015,
      "loss": 2.1094,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.7846077680587769,
      "learning_rate": 0.0015,
      "loss": 2.0957,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.7823826670646667,
      "learning_rate": 0.0015,
      "loss": 2.0981,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.7547616362571716,
      "learning_rate": 0.0015,
      "loss": 2.0858,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.7252967953681946,
      "learning_rate": 0.0015,
      "loss": 2.0692,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.7492804527282715,
      "learning_rate": 0.0015,
      "loss": 2.0778,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.8598700165748596,
      "learning_rate": 0.0015,
      "loss": 2.0704,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.7392456531524658,
      "learning_rate": 0.0015,
      "loss": 2.0576,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.9330459237098694,
      "learning_rate": 0.0015,
      "loss": 2.0451,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.921111524105072,
      "learning_rate": 0.0015,
      "loss": 2.0362,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7925679087638855,
      "learning_rate": 0.0015,
      "loss": 2.0455,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.8526214957237244,
      "learning_rate": 0.0015,
      "loss": 2.027,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.7173347473144531,
      "learning_rate": 0.0015,
      "loss": 2.0174,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.9150657653808594,
      "learning_rate": 0.0015,
      "loss": 1.9992,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.9017384052276611,
      "learning_rate": 0.0015,
      "loss": 1.9897,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.8591272234916687,
      "learning_rate": 0.0015,
      "loss": 2.0025,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.7839824557304382,
      "learning_rate": 0.0015,
      "loss": 1.981,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.238034725189209,
      "learning_rate": 0.0015,
      "loss": 1.9809,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.8030624985694885,
      "learning_rate": 0.0015,
      "loss": 1.972,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.8890962600708008,
      "learning_rate": 0.0015,
      "loss": 1.9648,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.7944323420524597,
      "learning_rate": 0.0015,
      "loss": 1.9706,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.7370387315750122,
      "learning_rate": 0.0015,
      "loss": 1.9479,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.7489625811576843,
      "learning_rate": 0.0015,
      "loss": 1.9328,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.8259197473526001,
      "learning_rate": 0.0015,
      "loss": 1.945,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.8472774624824524,
      "learning_rate": 0.0015,
      "loss": 1.9469,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.8071711659431458,
      "learning_rate": 0.0015,
      "loss": 1.9238,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.7328027486801147,
      "learning_rate": 0.0015,
      "loss": 1.9245,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.7462289929389954,
      "learning_rate": 0.0015,
      "loss": 1.9253,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.6801846623420715,
      "learning_rate": 0.0015,
      "loss": 1.9206,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.8341561555862427,
      "learning_rate": 0.0015,
      "loss": 1.9129,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.8199224472045898,
      "learning_rate": 0.0015,
      "loss": 1.9029,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.985684871673584,
      "learning_rate": 0.0015,
      "loss": 1.9013,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.7472009658813477,
      "learning_rate": 0.0015,
      "loss": 1.9057,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.6926509737968445,
      "learning_rate": 0.0015,
      "loss": 1.8925,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.9702551960945129,
      "learning_rate": 0.0015,
      "loss": 1.8804,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.7516466975212097,
      "learning_rate": 0.0015,
      "loss": 1.8894,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.7638765573501587,
      "learning_rate": 0.0015,
      "loss": 1.8683,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.8820604681968689,
      "learning_rate": 0.0015,
      "loss": 1.8759,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 1.0677109956741333,
      "learning_rate": 0.0015,
      "loss": 1.8773,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.8687389492988586,
      "learning_rate": 0.0015,
      "loss": 1.8558,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.6791872382164001,
      "learning_rate": 0.0015,
      "loss": 1.8525,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.9034246206283569,
      "learning_rate": 0.0015,
      "loss": 1.8624,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.7845780253410339,
      "learning_rate": 0.0015,
      "loss": 1.8542,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.9520986080169678,
      "learning_rate": 0.0015,
      "loss": 1.8475,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.879350483417511,
      "learning_rate": 0.0015,
      "loss": 1.85,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.6869158744812012,
      "learning_rate": 0.0015,
      "loss": 1.8431,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.8841139674186707,
      "learning_rate": 0.0015,
      "loss": 1.8425,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.7698689699172974,
      "learning_rate": 0.0015,
      "loss": 1.8309,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.781151294708252,
      "learning_rate": 0.0015,
      "loss": 1.8231,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.8235698938369751,
      "learning_rate": 0.0015,
      "loss": 1.8185,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.7147543430328369,
      "learning_rate": 0.0015,
      "loss": 1.83,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.7635578513145447,
      "learning_rate": 0.0015,
      "loss": 1.8162,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.6321555376052856,
      "learning_rate": 0.0015,
      "loss": 1.8206,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.7827277779579163,
      "learning_rate": 0.0015,
      "loss": 1.8149,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.9091566205024719,
      "learning_rate": 0.0015,
      "loss": 1.8002,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.7883399724960327,
      "learning_rate": 0.0015,
      "loss": 1.8142,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.0488635301589966,
      "learning_rate": 0.0015,
      "loss": 1.8,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.7445605993270874,
      "learning_rate": 0.0015,
      "loss": 1.807,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.8629063963890076,
      "learning_rate": 0.0015,
      "loss": 1.7939,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.6626562476158142,
      "learning_rate": 0.0015,
      "loss": 1.7806,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.6432856917381287,
      "learning_rate": 0.0015,
      "loss": 1.7822,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.6916394233703613,
      "learning_rate": 0.0015,
      "loss": 1.7933,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.7310307621955872,
      "learning_rate": 0.0015,
      "loss": 1.7886,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.6500587463378906,
      "learning_rate": 0.0015,
      "loss": 1.7723,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.7295315861701965,
      "learning_rate": 0.0015,
      "loss": 1.7634,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.6992786526679993,
      "learning_rate": 0.0015,
      "loss": 1.7791,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.695243775844574,
      "learning_rate": 0.0015,
      "loss": 1.7849,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.7244482636451721,
      "learning_rate": 0.0015,
      "loss": 1.7663,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.6903680562973022,
      "learning_rate": 0.0015,
      "loss": 1.7628,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.8517889976501465,
      "learning_rate": 0.0015,
      "loss": 1.7763,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.6429932117462158,
      "learning_rate": 0.0015,
      "loss": 1.7625,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.5941759347915649,
      "learning_rate": 0.0015,
      "loss": 1.7492,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.7779079079627991,
      "learning_rate": 0.0015,
      "loss": 1.7535,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.8056157231330872,
      "learning_rate": 0.0015,
      "loss": 1.7587,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 1.0647435188293457,
      "learning_rate": 0.0015,
      "loss": 1.7487,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.664645254611969,
      "learning_rate": 0.0015,
      "loss": 1.7533,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.749737024307251,
      "learning_rate": 0.0015,
      "loss": 1.7556,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.6377522349357605,
      "learning_rate": 0.0015,
      "loss": 1.7437,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.8008196353912354,
      "learning_rate": 0.0015,
      "loss": 1.7412,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.740041971206665,
      "learning_rate": 0.0015,
      "loss": 1.7336,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.6271490454673767,
      "learning_rate": 0.0015,
      "loss": 1.7437,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.8197513818740845,
      "learning_rate": 0.0015,
      "loss": 1.7241,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.7526323795318604,
      "learning_rate": 0.0015,
      "loss": 1.7295,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.6436764597892761,
      "learning_rate": 0.0015,
      "loss": 1.7355,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.7338356971740723,
      "learning_rate": 0.0015,
      "loss": 1.7296,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.6511281728744507,
      "learning_rate": 0.0015,
      "loss": 1.7271,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.7010757923126221,
      "learning_rate": 0.0015,
      "loss": 1.7231,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.6682503819465637,
      "learning_rate": 0.0015,
      "loss": 1.7175,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.8640552759170532,
      "learning_rate": 0.0015,
      "loss": 1.7166,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.6911349296569824,
      "learning_rate": 0.0015,
      "loss": 1.7215,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.6677291989326477,
      "learning_rate": 0.0015,
      "loss": 1.7165,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.6320167779922485,
      "learning_rate": 0.0015,
      "loss": 1.7156,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.7043343782424927,
      "learning_rate": 0.0015,
      "loss": 1.7186,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.642522931098938,
      "learning_rate": 0.0015,
      "loss": 1.7156,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.6582235097885132,
      "learning_rate": 0.0015,
      "loss": 1.6986,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6012140512466431,
      "learning_rate": 0.0015,
      "loss": 1.6899,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.9438639879226685,
      "learning_rate": 0.0015,
      "loss": 1.6913,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.6647754907608032,
      "learning_rate": 0.0015,
      "loss": 1.6956,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.9089834690093994,
      "learning_rate": 0.0015,
      "loss": 1.6913,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.7319679856300354,
      "learning_rate": 0.0015,
      "loss": 1.6994,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.8198645710945129,
      "learning_rate": 0.0015,
      "loss": 1.6968,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.6585257053375244,
      "learning_rate": 0.0015,
      "loss": 1.6978,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.637772262096405,
      "learning_rate": 0.0015,
      "loss": 1.6941,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6837565898895264,
      "learning_rate": 0.0015,
      "loss": 1.6658,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.7456943392753601,
      "learning_rate": 0.0015,
      "loss": 1.6795,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.6619670987129211,
      "learning_rate": 0.0015,
      "loss": 1.6826,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.6077799201011658,
      "learning_rate": 0.0015,
      "loss": 1.6719,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.682167649269104,
      "learning_rate": 0.0015,
      "loss": 1.6726,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.8284013867378235,
      "learning_rate": 0.0015,
      "loss": 1.6829,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.3403370380401611,
      "learning_rate": 0.0015,
      "loss": 1.6716,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.7174342274665833,
      "learning_rate": 0.0015,
      "loss": 1.6671,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.7241065502166748,
      "learning_rate": 0.0015,
      "loss": 1.6831,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.7094260454177856,
      "learning_rate": 0.0015,
      "loss": 1.6632,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.6029664874076843,
      "learning_rate": 0.0015,
      "loss": 1.6576,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.7462047934532166,
      "learning_rate": 0.0015,
      "loss": 1.665,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.7119380235671997,
      "learning_rate": 0.0015,
      "loss": 1.6653,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.6449779868125916,
      "learning_rate": 0.0015,
      "loss": 1.6853,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.6548015475273132,
      "learning_rate": 0.0015,
      "loss": 1.6644,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6180350184440613,
      "learning_rate": 0.0015,
      "loss": 1.6644,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.6398687958717346,
      "learning_rate": 0.0015,
      "loss": 1.6651,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.6970463991165161,
      "learning_rate": 0.0015,
      "loss": 1.6649,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.6194822192192078,
      "learning_rate": 0.0015,
      "loss": 1.6652,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.6553013920783997,
      "learning_rate": 0.0015,
      "loss": 1.6466,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.6812267303466797,
      "learning_rate": 0.0015,
      "loss": 1.6478,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.6811752319335938,
      "learning_rate": 0.0015,
      "loss": 1.6476,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.6852722764015198,
      "learning_rate": 0.0015,
      "loss": 1.6464,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.7010133266448975,
      "learning_rate": 0.0015,
      "loss": 1.6476,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.9136163592338562,
      "learning_rate": 0.0015,
      "loss": 1.6652,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.6500228643417358,
      "learning_rate": 0.0015,
      "loss": 1.6515,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.7408334016799927,
      "learning_rate": 0.0015,
      "loss": 1.6387,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.5994712114334106,
      "learning_rate": 0.0015,
      "loss": 1.6444,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.8974518775939941,
      "learning_rate": 0.0015,
      "loss": 1.6446,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.8792579770088196,
      "learning_rate": 0.0015,
      "loss": 1.6502,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.7742037773132324,
      "learning_rate": 0.0015,
      "loss": 1.6411,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.5866346955299377,
      "learning_rate": 0.0015,
      "loss": 1.6373,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.7438547015190125,
      "learning_rate": 0.0015,
      "loss": 1.6313,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.7660514116287231,
      "learning_rate": 0.0015,
      "loss": 1.6308,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.7325049042701721,
      "learning_rate": 0.0015,
      "loss": 1.6344,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.7959612011909485,
      "learning_rate": 0.0015,
      "loss": 1.6448,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.6661118268966675,
      "learning_rate": 0.0015,
      "loss": 1.6239,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.5984247326850891,
      "learning_rate": 0.0015,
      "loss": 1.6262,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.7870999574661255,
      "learning_rate": 0.0015,
      "loss": 1.6222,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.6280139088630676,
      "learning_rate": 0.0015,
      "loss": 1.6316,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 1.1350449323654175,
      "learning_rate": 0.0015,
      "loss": 1.6424,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.7405164241790771,
      "learning_rate": 0.0015,
      "loss": 1.6258,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6154853701591492,
      "learning_rate": 0.0015,
      "loss": 1.6243,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.7433428764343262,
      "learning_rate": 0.0015,
      "loss": 1.6285,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.6919626593589783,
      "learning_rate": 0.0015,
      "loss": 1.6171,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.8441318273544312,
      "learning_rate": 0.0015,
      "loss": 1.623,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.8494009971618652,
      "learning_rate": 0.0015,
      "loss": 1.6277,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.7551376223564148,
      "learning_rate": 0.0015,
      "loss": 1.6159,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.8248035907745361,
      "learning_rate": 0.0015,
      "loss": 1.621,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.6333810687065125,
      "learning_rate": 0.0015,
      "loss": 1.6122,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6450139880180359,
      "learning_rate": 0.0015,
      "loss": 1.61,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.6970928311347961,
      "learning_rate": 0.0015,
      "loss": 1.6154,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.5637871623039246,
      "learning_rate": 0.0015,
      "loss": 1.6041,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.7619059085845947,
      "learning_rate": 0.0015,
      "loss": 1.6165,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.7117628455162048,
      "learning_rate": 0.0015,
      "loss": 1.6299,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.6122785210609436,
      "learning_rate": 0.0015,
      "loss": 1.6117,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.5940090417861938,
      "learning_rate": 0.0015,
      "loss": 1.5916,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.7180793285369873,
      "learning_rate": 0.0015,
      "loss": 1.6133,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.6815030574798584,
      "learning_rate": 0.0015,
      "loss": 1.6155,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.6356433629989624,
      "learning_rate": 0.0015,
      "loss": 1.5973,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.8878588676452637,
      "learning_rate": 0.0015,
      "loss": 1.6103,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.7717560529708862,
      "learning_rate": 0.0015,
      "loss": 1.6048,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.6561765074729919,
      "learning_rate": 0.0015,
      "loss": 1.6015,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.8063392639160156,
      "learning_rate": 0.0015,
      "loss": 1.6,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.7132135629653931,
      "learning_rate": 0.0015,
      "loss": 1.5886,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.8069459795951843,
      "learning_rate": 0.0015,
      "loss": 1.6219,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.9883179664611816,
      "learning_rate": 0.0015,
      "loss": 1.609,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.6709671020507812,
      "learning_rate": 0.0015,
      "loss": 1.5944,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.558814525604248,
      "learning_rate": 0.0015,
      "loss": 1.5934,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.595003068447113,
      "learning_rate": 0.0015,
      "loss": 1.5911,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.6335365176200867,
      "learning_rate": 0.0015,
      "loss": 1.6012,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.6739267110824585,
      "learning_rate": 0.0015,
      "loss": 1.584,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.6889054775238037,
      "learning_rate": 0.0015,
      "loss": 1.582,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.6644899249076843,
      "learning_rate": 0.0015,
      "loss": 1.586,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.8086755871772766,
      "learning_rate": 0.0015,
      "loss": 1.5933,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7880407571792603,
      "learning_rate": 0.0015,
      "loss": 1.59,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.6508652567863464,
      "learning_rate": 0.0015,
      "loss": 1.5898,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.5541046261787415,
      "learning_rate": 0.0015,
      "loss": 1.5875,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6380650997161865,
      "learning_rate": 0.0015,
      "loss": 1.5853,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.7345043420791626,
      "learning_rate": 0.0015,
      "loss": 1.5876,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.77506422996521,
      "learning_rate": 0.0015,
      "loss": 1.5824,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.6960884928703308,
      "learning_rate": 0.0015,
      "loss": 1.5827,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.932823896408081,
      "learning_rate": 0.0015,
      "loss": 1.5919,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.5496547818183899,
      "learning_rate": 0.0015,
      "loss": 1.5747,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.5292431116104126,
      "learning_rate": 0.0015,
      "loss": 1.5903,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.7427460551261902,
      "learning_rate": 0.0015,
      "loss": 1.5819,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.6489607691764832,
      "learning_rate": 0.0015,
      "loss": 1.5887,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.5596098303794861,
      "learning_rate": 0.0015,
      "loss": 1.5714,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.7577886581420898,
      "learning_rate": 0.0015,
      "loss": 1.5779,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.6707363128662109,
      "learning_rate": 0.0015,
      "loss": 1.5778,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.6378468871116638,
      "learning_rate": 0.0015,
      "loss": 1.5822,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.6226636171340942,
      "learning_rate": 0.0015,
      "loss": 1.5768,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 1.0131611824035645,
      "learning_rate": 0.0015,
      "loss": 1.5762,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.7248948812484741,
      "learning_rate": 0.0015,
      "loss": 1.5746,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.6153429746627808,
      "learning_rate": 0.0015,
      "loss": 1.5805,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.5782231688499451,
      "learning_rate": 0.0015,
      "loss": 1.5744,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.587457001209259,
      "learning_rate": 0.0015,
      "loss": 1.5714,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.6506356596946716,
      "learning_rate": 0.0015,
      "loss": 1.5775,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.5758341550827026,
      "learning_rate": 0.0015,
      "loss": 1.5762,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6490718722343445,
      "learning_rate": 0.0015,
      "loss": 1.5749,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6964821815490723,
      "learning_rate": 0.0015,
      "loss": 1.5744,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.7029078006744385,
      "learning_rate": 0.0015,
      "loss": 1.5642,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.6132246851921082,
      "learning_rate": 0.0015,
      "loss": 1.574,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.7094488739967346,
      "learning_rate": 0.0015,
      "loss": 1.5689,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.6512430310249329,
      "learning_rate": 0.0015,
      "loss": 1.5601,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.5545443892478943,
      "learning_rate": 0.0015,
      "loss": 1.5626,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.5843806862831116,
      "learning_rate": 0.0015,
      "loss": 1.5652,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.6593872308731079,
      "learning_rate": 0.0015,
      "loss": 1.5644,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6269451379776001,
      "learning_rate": 0.0015,
      "loss": 1.5579,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.680083692073822,
      "learning_rate": 0.0015,
      "loss": 1.5649,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.586327314376831,
      "learning_rate": 0.0015,
      "loss": 1.5653,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.7322113513946533,
      "learning_rate": 0.0015,
      "loss": 1.5453,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.8342188000679016,
      "learning_rate": 0.0015,
      "loss": 1.5577,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.6292400360107422,
      "learning_rate": 0.0015,
      "loss": 1.5634,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.6687845587730408,
      "learning_rate": 0.0015,
      "loss": 1.572,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.5356091856956482,
      "learning_rate": 0.0015,
      "loss": 1.564,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.5680630803108215,
      "learning_rate": 0.0015,
      "loss": 1.5635,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.589124858379364,
      "learning_rate": 0.0015,
      "loss": 1.5598,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.5229613184928894,
      "learning_rate": 0.0015,
      "loss": 1.5576,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.6444438695907593,
      "learning_rate": 0.0015,
      "loss": 1.5591,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6138685345649719,
      "learning_rate": 0.0015,
      "loss": 1.5648,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.861492395401001,
      "learning_rate": 0.0015,
      "loss": 1.5388,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 1.04345703125,
      "learning_rate": 0.0015,
      "loss": 1.5433,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.6135168075561523,
      "learning_rate": 0.0015,
      "loss": 1.5475,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.6699387431144714,
      "learning_rate": 0.0015,
      "loss": 1.5547,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.8413599729537964,
      "learning_rate": 0.0015,
      "loss": 1.5541,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.8599419593811035,
      "learning_rate": 0.0015,
      "loss": 1.5393,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.6215834021568298,
      "learning_rate": 0.0015,
      "loss": 1.5469,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.6787463426589966,
      "learning_rate": 0.0015,
      "loss": 1.5535,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.8015420436859131,
      "learning_rate": 0.0015,
      "loss": 1.5507,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.6101096272468567,
      "learning_rate": 0.0015,
      "loss": 1.546,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.5448319911956787,
      "learning_rate": 0.0015,
      "loss": 1.5461,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.783310055732727,
      "learning_rate": 0.0015,
      "loss": 1.5509,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.5777860879898071,
      "learning_rate": 0.0015,
      "loss": 1.5591,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.6812126040458679,
      "learning_rate": 0.0015,
      "loss": 1.5378,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.6879600286483765,
      "learning_rate": 0.0015,
      "loss": 1.5318,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.9139813184738159,
      "learning_rate": 0.0015,
      "loss": 1.5376,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.6776876449584961,
      "learning_rate": 0.0015,
      "loss": 1.5528,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.7328812479972839,
      "learning_rate": 0.0015,
      "loss": 1.5356,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.5674942135810852,
      "learning_rate": 0.0015,
      "loss": 1.5426,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.6288122534751892,
      "learning_rate": 0.0015,
      "loss": 1.5406,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.595083475112915,
      "learning_rate": 0.0015,
      "loss": 1.5313,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.6639403700828552,
      "learning_rate": 0.0015,
      "loss": 1.5378,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.7175447940826416,
      "learning_rate": 0.0015,
      "loss": 1.5334,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.5787521004676819,
      "learning_rate": 0.0015,
      "loss": 1.5455,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.5793510675430298,
      "learning_rate": 0.0015,
      "loss": 1.5346,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.8050768971443176,
      "learning_rate": 0.0015,
      "loss": 1.5271,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.6016557812690735,
      "learning_rate": 0.0015,
      "loss": 1.5453,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.5810866951942444,
      "learning_rate": 0.0015,
      "loss": 1.5308,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6229231357574463,
      "learning_rate": 0.0015,
      "loss": 1.5344,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.6157940626144409,
      "learning_rate": 0.0015,
      "loss": 1.5345,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.6028305292129517,
      "learning_rate": 0.0015,
      "loss": 1.5473,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.7928791046142578,
      "learning_rate": 0.0015,
      "loss": 1.527,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.6114389300346375,
      "learning_rate": 0.0015,
      "loss": 1.5322,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5845867991447449,
      "learning_rate": 0.0015,
      "loss": 1.5332,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.5498821139335632,
      "learning_rate": 0.0015,
      "loss": 1.5372,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.5723025798797607,
      "learning_rate": 0.0015,
      "loss": 1.5333,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.7843588590621948,
      "learning_rate": 0.0015,
      "loss": 1.5341,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.7767053842544556,
      "learning_rate": 0.0015,
      "loss": 1.528,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.7209522724151611,
      "learning_rate": 0.0015,
      "loss": 1.5404,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.8024675846099854,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.8624002933502197,
      "learning_rate": 0.0015,
      "loss": 1.5251,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.6364142298698425,
      "learning_rate": 0.0015,
      "loss": 1.505,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.6138675808906555,
      "learning_rate": 0.0015,
      "loss": 1.532,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.5279717445373535,
      "learning_rate": 0.0015,
      "loss": 1.5195,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.5529916882514954,
      "learning_rate": 0.0015,
      "loss": 1.5379,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.6098965406417847,
      "learning_rate": 0.0015,
      "loss": 1.5352,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.9433780312538147,
      "learning_rate": 0.0015,
      "loss": 1.5196,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.6039072871208191,
      "learning_rate": 0.0015,
      "loss": 1.5288,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.5436165928840637,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.7043375372886658,
      "learning_rate": 0.0015,
      "loss": 1.5329,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.5190507173538208,
      "learning_rate": 0.0015,
      "loss": 1.5215,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.5707694292068481,
      "learning_rate": 0.0015,
      "loss": 1.5251,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.5818861722946167,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.6608387231826782,
      "learning_rate": 0.0015,
      "loss": 1.5194,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.7077980041503906,
      "learning_rate": 0.0015,
      "loss": 1.5273,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.6117827892303467,
      "learning_rate": 0.0015,
      "loss": 1.5219,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.5437614917755127,
      "learning_rate": 0.0015,
      "loss": 1.5148,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.6310780048370361,
      "learning_rate": 0.0015,
      "loss": 1.5229,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.5511094927787781,
      "learning_rate": 0.0015,
      "loss": 1.5214,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.634290337562561,
      "learning_rate": 0.0015,
      "loss": 1.5024,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.5730307698249817,
      "learning_rate": 0.0015,
      "loss": 1.52,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.5337285995483398,
      "learning_rate": 0.0015,
      "loss": 1.5158,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.5572408437728882,
      "learning_rate": 0.0015,
      "loss": 1.5108,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.6166617274284363,
      "learning_rate": 0.0015,
      "loss": 1.5066,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.7355265021324158,
      "learning_rate": 0.0015,
      "loss": 1.5229,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.5263220071792603,
      "learning_rate": 0.0015,
      "loss": 1.5267,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.7298345565795898,
      "learning_rate": 0.0015,
      "loss": 1.5176,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.6348539590835571,
      "learning_rate": 0.0015,
      "loss": 1.5018,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.6110532879829407,
      "learning_rate": 0.0015,
      "loss": 1.5244,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.5885331630706787,
      "learning_rate": 0.0015,
      "loss": 1.5159,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.6391881108283997,
      "learning_rate": 0.0015,
      "loss": 1.5262,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 1.229824185371399,
      "learning_rate": 0.0015,
      "loss": 1.522,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.6018309593200684,
      "learning_rate": 0.0015,
      "loss": 1.5086,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.5689561367034912,
      "learning_rate": 0.0015,
      "loss": 1.5091,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.7056559324264526,
      "learning_rate": 0.0015,
      "loss": 1.5062,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.5769897699356079,
      "learning_rate": 0.0015,
      "loss": 1.5088,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.6079292893409729,
      "learning_rate": 0.0015,
      "loss": 1.5193,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.838927686214447,
      "learning_rate": 0.0015,
      "loss": 1.5041,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.5909256339073181,
      "learning_rate": 0.0015,
      "loss": 1.5064,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.5677759647369385,
      "learning_rate": 0.0015,
      "loss": 1.5061,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.6772117018699646,
      "learning_rate": 0.0015,
      "loss": 1.5083,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.6224068999290466,
      "learning_rate": 0.0015,
      "loss": 1.5068,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.8087689280509949,
      "learning_rate": 0.0015,
      "loss": 1.5165,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.819786787033081,
      "learning_rate": 0.0015,
      "loss": 1.5089,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.5673829913139343,
      "learning_rate": 0.0015,
      "loss": 1.5038,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.713397741317749,
      "learning_rate": 0.0015,
      "loss": 1.5059,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.5734171271324158,
      "learning_rate": 0.0015,
      "loss": 1.5153,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.6730978488922119,
      "learning_rate": 0.0015,
      "loss": 1.5218,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.5980945229530334,
      "learning_rate": 0.0015,
      "loss": 1.5097,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.6305373907089233,
      "learning_rate": 0.0015,
      "loss": 1.4865,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.5960437059402466,
      "learning_rate": 0.0015,
      "loss": 1.5088,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.5839282870292664,
      "learning_rate": 0.0015,
      "loss": 1.5076,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.6649579405784607,
      "learning_rate": 0.0015,
      "loss": 1.509,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.9002540707588196,
      "learning_rate": 0.0015,
      "loss": 1.5003,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.5483456254005432,
      "learning_rate": 0.0015,
      "loss": 1.514,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.5238082408905029,
      "learning_rate": 0.0015,
      "loss": 1.4993,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.8858643174171448,
      "learning_rate": 0.0015,
      "loss": 1.5053,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.603560209274292,
      "learning_rate": 0.0015,
      "loss": 1.512,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.5253837704658508,
      "learning_rate": 0.0015,
      "loss": 1.496,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.60108882188797,
      "learning_rate": 0.0015,
      "loss": 1.4908,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.5289162993431091,
      "learning_rate": 0.0015,
      "loss": 1.4933,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.5684453248977661,
      "learning_rate": 0.0015,
      "loss": 1.5096,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.5767807364463806,
      "learning_rate": 0.0015,
      "loss": 1.4914,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.7127223014831543,
      "learning_rate": 0.0015,
      "loss": 1.5065,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.862557053565979,
      "learning_rate": 0.0015,
      "loss": 1.4949,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.6481454968452454,
      "learning_rate": 0.0015,
      "loss": 1.4877,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.5657109022140503,
      "learning_rate": 0.0015,
      "loss": 1.4872,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.8695623278617859,
      "learning_rate": 0.0015,
      "loss": 1.4892,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.576208233833313,
      "learning_rate": 0.0015,
      "loss": 1.4911,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.5760087370872498,
      "learning_rate": 0.0015,
      "loss": 1.4888,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.6440626978874207,
      "learning_rate": 0.0015,
      "loss": 1.493,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.7605494260787964,
      "learning_rate": 0.0015,
      "loss": 1.4894,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.6733703017234802,
      "learning_rate": 0.0015,
      "loss": 1.497,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.5512400269508362,
      "learning_rate": 0.0015,
      "loss": 1.4989,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.7931657433509827,
      "learning_rate": 0.0015,
      "loss": 1.4872,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.5423968434333801,
      "learning_rate": 0.0015,
      "loss": 1.4974,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.5467481017112732,
      "learning_rate": 0.0015,
      "loss": 1.4866,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.5243974328041077,
      "learning_rate": 0.0015,
      "loss": 1.4989,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.5044182538986206,
      "learning_rate": 0.0015,
      "loss": 1.4841,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.5074765682220459,
      "learning_rate": 0.0015,
      "loss": 1.4901,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.5237292051315308,
      "learning_rate": 0.0015,
      "loss": 1.4804,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5365952253341675,
      "learning_rate": 0.0015,
      "loss": 1.4918,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5531340837478638,
      "learning_rate": 0.0015,
      "loss": 1.4934,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.9647518396377563,
      "learning_rate": 0.0015,
      "loss": 1.4984,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.7496301531791687,
      "learning_rate": 0.0015,
      "loss": 1.4945,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.681740403175354,
      "learning_rate": 0.0015,
      "loss": 1.4987,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.6222134232521057,
      "learning_rate": 0.0015,
      "loss": 1.4939,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5539551973342896,
      "learning_rate": 0.0015,
      "loss": 1.488,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.6145344972610474,
      "learning_rate": 0.0015,
      "loss": 1.4845,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.730179488658905,
      "learning_rate": 0.0015,
      "loss": 1.4901,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.6839131712913513,
      "learning_rate": 0.0015,
      "loss": 1.483,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5735271573066711,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.5676992535591125,
      "learning_rate": 0.0015,
      "loss": 1.4878,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.5206440687179565,
      "learning_rate": 0.0015,
      "loss": 1.4827,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.549821674823761,
      "learning_rate": 0.0015,
      "loss": 1.4833,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.49884817004203796,
      "learning_rate": 0.0015,
      "loss": 1.4805,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.5792259573936462,
      "learning_rate": 0.0015,
      "loss": 1.4868,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.6022302508354187,
      "learning_rate": 0.0015,
      "loss": 1.4878,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.6109970808029175,
      "learning_rate": 0.0015,
      "loss": 1.4779,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.7282804250717163,
      "learning_rate": 0.0015,
      "loss": 1.4864,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.8704391121864319,
      "learning_rate": 0.0015,
      "loss": 1.4829,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.7611539363861084,
      "learning_rate": 0.0015,
      "loss": 1.4802,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.5748181939125061,
      "learning_rate": 0.0015,
      "loss": 1.4864,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.9245660305023193,
      "learning_rate": 0.0015,
      "loss": 1.4913,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5048531889915466,
      "learning_rate": 0.0015,
      "loss": 1.4787,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.8192220330238342,
      "learning_rate": 0.0015,
      "loss": 1.4728,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.5813215970993042,
      "learning_rate": 0.0015,
      "loss": 1.4887,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.7224900126457214,
      "learning_rate": 0.0015,
      "loss": 1.4827,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.5357944369316101,
      "learning_rate": 0.0015,
      "loss": 1.4787,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.7337834239006042,
      "learning_rate": 0.0015,
      "loss": 1.4745,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.7391050457954407,
      "learning_rate": 0.0015,
      "loss": 1.4864,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.7839824557304382,
      "learning_rate": 0.0015,
      "loss": 1.4743,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.8363252878189087,
      "learning_rate": 0.0015,
      "loss": 1.4794,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.5706027150154114,
      "learning_rate": 0.0015,
      "loss": 1.4948,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.5642624497413635,
      "learning_rate": 0.0015,
      "loss": 1.4731,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5105932950973511,
      "learning_rate": 0.0015,
      "loss": 1.4782,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.6307752132415771,
      "learning_rate": 0.0015,
      "loss": 1.4722,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.5654680728912354,
      "learning_rate": 0.0015,
      "loss": 1.4771,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.5729908347129822,
      "learning_rate": 0.0015,
      "loss": 1.4931,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.6047531962394714,
      "learning_rate": 0.0015,
      "loss": 1.4753,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.5820277333259583,
      "learning_rate": 0.0015,
      "loss": 1.479,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.5866506695747375,
      "learning_rate": 0.0015,
      "loss": 1.4754,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.698915421962738,
      "learning_rate": 0.0015,
      "loss": 1.4829,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5867848992347717,
      "learning_rate": 0.0015,
      "loss": 1.4696,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.5671797394752502,
      "learning_rate": 0.0015,
      "loss": 1.4765,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.539013147354126,
      "learning_rate": 0.0015,
      "loss": 1.4839,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.6562228202819824,
      "learning_rate": 0.0015,
      "loss": 1.4718,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.5956013202667236,
      "learning_rate": 0.0015,
      "loss": 1.4718,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5919322967529297,
      "learning_rate": 0.0015,
      "loss": 1.4767,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.5776900053024292,
      "learning_rate": 0.0015,
      "loss": 1.4582,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.5107901096343994,
      "learning_rate": 0.0015,
      "loss": 1.4784,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.5599530339241028,
      "learning_rate": 0.0015,
      "loss": 1.48,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.5354790091514587,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.5410628318786621,
      "learning_rate": 0.0015,
      "loss": 1.4665,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.5314594507217407,
      "learning_rate": 0.0015,
      "loss": 1.4488,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.8389555811882019,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.6336533427238464,
      "learning_rate": 0.0015,
      "loss": 1.4605,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.6722411513328552,
      "learning_rate": 0.0015,
      "loss": 1.4758,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.732998251914978,
      "learning_rate": 0.0015,
      "loss": 1.4823,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.5437710285186768,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.5527676939964294,
      "learning_rate": 0.0015,
      "loss": 1.4751,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.6624552011489868,
      "learning_rate": 0.0015,
      "loss": 1.4777,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5752913355827332,
      "learning_rate": 0.0015,
      "loss": 1.4785,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.5568695068359375,
      "learning_rate": 0.0015,
      "loss": 1.4652,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.7825382351875305,
      "learning_rate": 0.0015,
      "loss": 1.4671,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.47440215945243835,
      "learning_rate": 0.0015,
      "loss": 1.4611,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.6024202704429626,
      "learning_rate": 0.0015,
      "loss": 1.459,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.49631550908088684,
      "learning_rate": 0.0015,
      "loss": 1.468,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.5096955299377441,
      "learning_rate": 0.0015,
      "loss": 1.4713,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.6142767071723938,
      "learning_rate": 0.0015,
      "loss": 1.4627,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.6223500370979309,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6369379162788391,
      "learning_rate": 0.0015,
      "loss": 1.4634,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.6030281186103821,
      "learning_rate": 0.0015,
      "loss": 1.4805,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5057412981987,
      "learning_rate": 0.0015,
      "loss": 1.4627,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.5182653069496155,
      "learning_rate": 0.0015,
      "loss": 1.4612,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5908124446868896,
      "learning_rate": 0.0015,
      "loss": 1.4726,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.543255090713501,
      "learning_rate": 0.0015,
      "loss": 1.4563,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5262438058853149,
      "learning_rate": 0.0015,
      "loss": 1.4635,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.5360994338989258,
      "learning_rate": 0.0015,
      "loss": 1.4612,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.5333308577537537,
      "learning_rate": 0.0015,
      "loss": 1.4776,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.8673296570777893,
      "learning_rate": 0.0015,
      "loss": 1.4652,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.7105008363723755,
      "learning_rate": 0.0015,
      "loss": 1.4596,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.534705400466919,
      "learning_rate": 0.0015,
      "loss": 1.473,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.4968915581703186,
      "learning_rate": 0.0015,
      "loss": 1.459,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.5146236419677734,
      "learning_rate": 0.0015,
      "loss": 1.4517,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.5623924136161804,
      "learning_rate": 0.0015,
      "loss": 1.475,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.5382164716720581,
      "learning_rate": 0.0015,
      "loss": 1.4629,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.6249877214431763,
      "learning_rate": 0.0015,
      "loss": 1.4715,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.7355168461799622,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.7187797427177429,
      "learning_rate": 0.0015,
      "loss": 1.4545,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.8246368765830994,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.7025818824768066,
      "learning_rate": 0.0015,
      "loss": 1.4669,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.6499597430229187,
      "learning_rate": 0.0015,
      "loss": 1.4706,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.5627788305282593,
      "learning_rate": 0.0015,
      "loss": 1.4494,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.5110287070274353,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.5370630025863647,
      "learning_rate": 0.0015,
      "loss": 1.4515,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.5351104736328125,
      "learning_rate": 0.0015,
      "loss": 1.4559,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.6204255223274231,
      "learning_rate": 0.0015,
      "loss": 1.4678,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.505325973033905,
      "learning_rate": 0.0015,
      "loss": 1.4581,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.537993848323822,
      "learning_rate": 0.0015,
      "loss": 1.453,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.4995332658290863,
      "learning_rate": 0.0015,
      "loss": 1.4393,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.5203754901885986,
      "learning_rate": 0.0015,
      "loss": 1.4678,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.5425376296043396,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.549592912197113,
      "learning_rate": 0.0015,
      "loss": 1.4676,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.8335903286933899,
      "learning_rate": 0.0015,
      "loss": 1.4571,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.580350399017334,
      "learning_rate": 0.0015,
      "loss": 1.4707,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.5234392881393433,
      "learning_rate": 0.0015,
      "loss": 1.4597,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.5417402982711792,
      "learning_rate": 0.0015,
      "loss": 1.4411,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.6939650177955627,
      "learning_rate": 0.0015,
      "loss": 1.4589,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6127960681915283,
      "learning_rate": 0.0015,
      "loss": 1.4554,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.5292453169822693,
      "learning_rate": 0.0015,
      "loss": 1.4516,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.7645080089569092,
      "learning_rate": 0.0015,
      "loss": 1.4662,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.590851366519928,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.6092954277992249,
      "learning_rate": 0.0015,
      "loss": 1.4569,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5821204781532288,
      "learning_rate": 0.0015,
      "loss": 1.4571,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.7959945797920227,
      "learning_rate": 0.0015,
      "loss": 1.4673,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.5450267195701599,
      "learning_rate": 0.0015,
      "loss": 1.4496,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.5266216397285461,
      "learning_rate": 0.0015,
      "loss": 1.4606,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.6075416803359985,
      "learning_rate": 0.0015,
      "loss": 1.4462,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.5457861423492432,
      "learning_rate": 0.0015,
      "loss": 1.4563,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.57088303565979,
      "learning_rate": 0.0015,
      "loss": 1.4487,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.6412349939346313,
      "learning_rate": 0.0015,
      "loss": 1.4577,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.7487953901290894,
      "learning_rate": 0.0015,
      "loss": 1.4492,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.519920825958252,
      "learning_rate": 0.0015,
      "loss": 1.4514,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.5340611934661865,
      "learning_rate": 0.0015,
      "loss": 1.4623,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.6583961844444275,
      "learning_rate": 0.0015,
      "loss": 1.4508,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.48247647285461426,
      "learning_rate": 0.0015,
      "loss": 1.4534,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.6072560548782349,
      "learning_rate": 0.0015,
      "loss": 1.4563,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.8439222574234009,
      "learning_rate": 0.0015,
      "loss": 1.4462,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.4916550815105438,
      "learning_rate": 0.0015,
      "loss": 1.4441,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.5986077189445496,
      "learning_rate": 0.0015,
      "loss": 1.4559,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.5265570282936096,
      "learning_rate": 0.0015,
      "loss": 1.4487,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.5903233885765076,
      "learning_rate": 0.0015,
      "loss": 1.4431,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.6227239370346069,
      "learning_rate": 0.0015,
      "loss": 1.453,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.6359860897064209,
      "learning_rate": 0.0015,
      "loss": 1.4525,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.5258117914199829,
      "learning_rate": 0.0015,
      "loss": 1.4464,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.6751209497451782,
      "learning_rate": 0.0015,
      "loss": 1.4465,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5402705669403076,
      "learning_rate": 0.0015,
      "loss": 1.4471,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.493479460477829,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.5518337488174438,
      "learning_rate": 0.0015,
      "loss": 1.442,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.8178352117538452,
      "learning_rate": 0.0015,
      "loss": 1.4385,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.4740898609161377,
      "learning_rate": 0.0015,
      "loss": 1.4378,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.5016689896583557,
      "learning_rate": 0.0015,
      "loss": 1.4445,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.6049769520759583,
      "learning_rate": 0.0015,
      "loss": 1.4481,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.49586889147758484,
      "learning_rate": 0.0015,
      "loss": 1.445,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.842395007610321,
      "learning_rate": 0.0015,
      "loss": 1.4504,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 1.0603761672973633,
      "learning_rate": 0.0015,
      "loss": 1.4557,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.6563161611557007,
      "learning_rate": 0.0015,
      "loss": 1.4466,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.811011552810669,
      "learning_rate": 0.0015,
      "loss": 1.45,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.4889761507511139,
      "learning_rate": 0.0015,
      "loss": 1.4422,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.5133126974105835,
      "learning_rate": 0.0015,
      "loss": 1.4377,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.5674935579299927,
      "learning_rate": 0.0015,
      "loss": 1.4356,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.5435529947280884,
      "learning_rate": 0.0015,
      "loss": 1.4516,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.48571109771728516,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.4769778549671173,
      "learning_rate": 0.0015,
      "loss": 1.4423,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.6307300925254822,
      "learning_rate": 0.0015,
      "loss": 1.4402,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.5105009078979492,
      "learning_rate": 0.0015,
      "loss": 1.4466,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5061869025230408,
      "learning_rate": 0.0015,
      "loss": 1.4401,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.5118229389190674,
      "learning_rate": 0.0015,
      "loss": 1.4426,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.5465503334999084,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.7530744075775146,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.5208582878112793,
      "learning_rate": 0.0015,
      "loss": 1.4567,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.6361252069473267,
      "learning_rate": 0.0015,
      "loss": 1.4501,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.7223333716392517,
      "learning_rate": 0.0015,
      "loss": 1.4455,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.6363935470581055,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5155510902404785,
      "learning_rate": 0.0015,
      "loss": 1.4533,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.9333240389823914,
      "learning_rate": 0.0015,
      "loss": 1.4334,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.5917677879333496,
      "learning_rate": 0.0015,
      "loss": 1.4391,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.7863317131996155,
      "learning_rate": 0.0015,
      "loss": 1.4427,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5073190927505493,
      "learning_rate": 0.0015,
      "loss": 1.4454,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.5154165029525757,
      "learning_rate": 0.0015,
      "loss": 1.4429,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.5117877721786499,
      "learning_rate": 0.0015,
      "loss": 1.4399,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.6820557713508606,
      "learning_rate": 0.0015,
      "loss": 1.4158,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5057265758514404,
      "learning_rate": 0.0015,
      "loss": 1.4246,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.6210771799087524,
      "learning_rate": 0.0015,
      "loss": 1.4274,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.4988754093647003,
      "learning_rate": 0.0015,
      "loss": 1.4361,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.5129354596138,
      "learning_rate": 0.0015,
      "loss": 1.4342,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.5080984830856323,
      "learning_rate": 0.0015,
      "loss": 1.437,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5181266665458679,
      "learning_rate": 0.0015,
      "loss": 1.4425,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5294566750526428,
      "learning_rate": 0.0015,
      "loss": 1.4557,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.49302735924720764,
      "learning_rate": 0.0015,
      "loss": 1.4374,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.6732519268989563,
      "learning_rate": 0.0015,
      "loss": 1.4497,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.5781086683273315,
      "learning_rate": 0.0015,
      "loss": 1.4535,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.5936325192451477,
      "learning_rate": 0.0015,
      "loss": 1.4394,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.528407633304596,
      "learning_rate": 0.0015,
      "loss": 1.4366,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.6488287448883057,
      "learning_rate": 0.0015,
      "loss": 1.4436,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5234397053718567,
      "learning_rate": 0.0015,
      "loss": 1.4347,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.6292299628257751,
      "learning_rate": 0.0015,
      "loss": 1.4385,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.6793391108512878,
      "learning_rate": 0.0015,
      "loss": 1.4299,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.5267859101295471,
      "learning_rate": 0.0015,
      "loss": 1.426,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.9769287705421448,
      "learning_rate": 0.0015,
      "loss": 1.4349,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.6040151715278625,
      "learning_rate": 0.0015,
      "loss": 1.4415,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.7830216884613037,
      "learning_rate": 0.0015,
      "loss": 1.4312,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.6196193099021912,
      "learning_rate": 0.0015,
      "loss": 1.4368,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5151229500770569,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.5271943807601929,
      "learning_rate": 0.0015,
      "loss": 1.4242,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.572357714176178,
      "learning_rate": 0.0015,
      "loss": 1.4406,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.49460625648498535,
      "learning_rate": 0.0015,
      "loss": 1.4492,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.6834750771522522,
      "learning_rate": 0.0015,
      "loss": 1.4363,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.5282460451126099,
      "learning_rate": 0.0015,
      "loss": 1.4349,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.6270628571510315,
      "learning_rate": 0.0015,
      "loss": 1.4199,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.5458430051803589,
      "learning_rate": 0.0015,
      "loss": 1.4366,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.4677220582962036,
      "learning_rate": 0.0015,
      "loss": 1.4212,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.6635910272598267,
      "learning_rate": 0.0015,
      "loss": 1.4403,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.6385688781738281,
      "learning_rate": 0.0015,
      "loss": 1.4339,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.5389999151229858,
      "learning_rate": 0.0015,
      "loss": 1.4301,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.6427217125892639,
      "learning_rate": 0.0015,
      "loss": 1.4313,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.5562639236450195,
      "learning_rate": 0.0015,
      "loss": 1.4247,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.5408332943916321,
      "learning_rate": 0.0015,
      "loss": 1.4378,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.6177515983581543,
      "learning_rate": 0.0015,
      "loss": 1.4331,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.7089222073554993,
      "learning_rate": 0.0015,
      "loss": 1.4305,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.924924910068512,
      "learning_rate": 0.0015,
      "loss": 1.4322,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.556202232837677,
      "learning_rate": 0.0015,
      "loss": 1.4416,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.4826835095882416,
      "learning_rate": 0.0015,
      "loss": 1.4353,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5418365001678467,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.413,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.5744474530220032,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.4361,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5739061236381531,
      "learning_rate": 0.001444828815847542,
      "loss": 1.4366,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.5228830575942993,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.4281,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.4765559434890747,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.4247,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.5601405501365662,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4309,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.5145525932312012,
      "learning_rate": 0.001366801618799797,
      "loss": 1.4171,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5060276985168457,
      "learning_rate": 0.001347962334099052,
      "loss": 1.4189,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.4647819995880127,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4251,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.46460989117622375,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4146,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.6334429979324341,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.4247,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5058175325393677,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.4209,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.47807610034942627,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.4145,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.560603678226471,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4096,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.5527917146682739,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4072,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.4986222982406616,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.4043,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.5951113700866699,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4115,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.7368322014808655,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4026,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.5130961537361145,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.4085,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.5715031623840332,
      "learning_rate": 0.001141156021915355,
      "loss": 1.4062,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.48664069175720215,
      "learning_rate": 0.001125426919103997,
      "loss": 1.3849,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.5779435634613037,
      "learning_rate": 0.001109914618088799,
      "loss": 1.3884,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.5953982472419739,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.402,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.529719352722168,
      "learning_rate": 0.001079528509501728,
      "loss": 1.4003,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.48857346177101135,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.3984,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.6054450273513794,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.3836,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.5000371336936951,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.3925,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.5130473971366882,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3763,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5538947582244873,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.3806,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.5086504817008972,
      "learning_rate": 0.000993270988848579,
      "loss": 1.3932,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5263882875442505,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.3865,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.7414752244949341,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.3913,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.46377184987068176,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.386,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.5172143578529358,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.4018,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.48321548104286194,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.3993,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.48881494998931885,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.3815,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.5049849152565002,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3741,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.7140661478042603,
      "learning_rate": 0.000888885722567627,
      "loss": 1.384,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.508492112159729,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.3754,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.637471079826355,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3767,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.5108541250228882,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.3737,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.507684051990509,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3686,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.5356305241584778,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3753,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5399567484855652,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3789,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.49503645300865173,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.3657,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.4790942966938019,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3725,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.4818264842033386,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.3653,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.5164989233016968,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.371,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.4759161174297333,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3621,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.6907697916030884,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3626,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5785808563232422,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.3668,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.4423680305480957,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.3653,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.4949614107608795,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.357,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.4672105014324188,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.3607,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.49704837799072266,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.36,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.4978903830051422,
      "learning_rate": 0.000692383699170611,
      "loss": 1.351,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.48249298334121704,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3599,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.5168253183364868,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3589,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.46104562282562256,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.3625,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.6228384375572205,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3594,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.5275484323501587,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3596,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.48850226402282715,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3555,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.4638110101222992,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.3602,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.512987494468689,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3449,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.5887208580970764,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3496,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.494081050157547,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.3521,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.4984007179737091,
      "learning_rate": 0.000594349404202577,
      "loss": 1.3512,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.5754541754722595,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.338,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.4539002776145935,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.3475,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.4598434567451477,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3412,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.4952813684940338,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3418,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.4592992663383484,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.3454,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.5437552332878113,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3358,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.49555784463882446,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3368,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5283121466636658,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.3601,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.47812119126319885,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.34,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.45417147874832153,
      "learning_rate": 0.000517326298154212,
      "loss": 1.3449,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.5009807348251343,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3484,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.4717162251472473,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3448,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.4565405249595642,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3364,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.48044589161872864,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.331,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.5185188055038452,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3341,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.4750678539276123,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.334,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.5038130283355713,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3443,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.44634950160980225,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.315,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.5554784536361694,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.3295,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.47232624888420105,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3332,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.7059232592582703,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3403,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5471903085708618,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3265,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.5978600978851318,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.3336,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.6368666887283325,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.3376,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.4617348909378052,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3338,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.470568984746933,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3308,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.4789184331893921,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.3269,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5197702646255493,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.3377,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.528596043586731,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3306,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.4621123969554901,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.3386,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.485154390335083,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.3236,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5580912828445435,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.3179,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.4799462556838989,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3273,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.6147610545158386,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.3352,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.5125970244407654,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.3137,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.4622291028499603,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3169,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.4684886932373047,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.324,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.5011488795280457,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.326,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.49282875657081604,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.3368,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.5074600577354431,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3329,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.5084503293037415,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3281,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.44725099205970764,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3223,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.49938493967056274,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.3213,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.48368123173713684,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.3202,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.509196937084198,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.3218,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.4666333794593811,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.3321,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.5396830439567566,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.3348,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.4591995179653168,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.3161,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.5201775431632996,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.3118,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5438663959503174,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3151,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.513739824295044,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3148,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.46607547998428345,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3086,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.49582216143608093,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3192,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.49414581060409546,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3204,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.4685586094856262,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3052,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.5296639204025269,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.3167,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.5685615539550781,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3233,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.4810534715652466,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.303,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.4799834191799164,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3101,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.4580126702785492,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.2965,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.43976375460624695,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.3001,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.45261019468307495,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.3034,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.4689023196697235,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.3143,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.4711395800113678,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3029,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.5190836191177368,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3092,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.5048525333404541,
      "learning_rate": 0.000237799979917852,
      "loss": 1.3179,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.4945871829986572,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3138,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.5310850739479065,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.3185,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.4699118435382843,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.3025,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.48693156242370605,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.3035,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.4786016047000885,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.3004,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.4625243842601776,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.2983,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.5461952686309814,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.3017,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.4864919185638428,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3125,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.4987559914588928,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.3097,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.4474153518676758,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.2983,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.4446052014827728,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.3044,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.46977654099464417,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.3149,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.49264687299728394,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.2884,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.4922375977039337,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.3117,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.5052447319030762,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.3025,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.4612441658973694,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.3187,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.451124906539917,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.2962,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.45290225744247437,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.3029,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.49601346254348755,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.3181,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.48603808879852295,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.2927,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.4748716354370117,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.3005,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.5524786114692688,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.2969,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.48633909225463867,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.3012,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.46730947494506836,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3065,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.4603579342365265,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3094,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.48680615425109863,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.2933,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5896096229553223,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.2948,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.46318498253822327,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.302,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.5044655203819275,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.2939,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.4882090985774994,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.3095,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4377343952655792,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.2977,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.6237848401069641,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.2972,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.44521579146385193,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3057,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.46726253628730774,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.2972,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.4927540421485901,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.2817,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.45653632283210754,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.2896,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.45124268531799316,
      "learning_rate": 0.000142293772767289,
      "loss": 1.2805,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.5420495867729187,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.2925,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.4798906445503235,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3114,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.47891050577163696,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.3098,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.45018500089645386,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.2881,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5154443979263306,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.2974,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5460496544837952,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.2949,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.46702855825424194,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3076,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.4837702512741089,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.2986,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.48372164368629456,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.3023,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.45765557885169983,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.2877,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.4671437442302704,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.2874,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.444602370262146,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.2872,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.5549850463867188,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.2987,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.4987001419067383,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.2928,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.4757237732410431,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.2996,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.4786148965358734,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.2931,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.46024706959724426,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.2921,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.47849148511886597,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3018,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.4562627971172333,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.2835,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.456315815448761,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.2932,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.46576249599456787,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.2979,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4787660539150238,
      "learning_rate": 0.000104851799496331,
      "loss": 1.2995,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.46897056698799133,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.2864,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.5212024450302124,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.2836,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.5523073077201843,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.3002,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.4823736548423767,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.2889,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.483837753534317,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.2893,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.45704153180122375,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.2896,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.4395501911640167,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.2838,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.48120397329330444,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.2965,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.5131252408027649,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.2848,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.462249755859375,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.294,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.4905018210411072,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.2858,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.4609352946281433,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.2979,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.45523637533187866,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.2915,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.4543611705303192,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.2995,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.45446282625198364,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.2938,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.44070711731910706,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.2896,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.460778146982193,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3053,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.461288183927536,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.2969,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.47572875022888184,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.2927,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.45623719692230225,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2762,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.49968743324279785,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.2757,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.46680504083633423,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.3014,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.48668330907821655,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.2822,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.45921820402145386,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.2982,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.46059128642082214,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.2933,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.4603794813156128,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.2785,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.4536525011062622,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.285,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.4565249979496002,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.2823,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.4472019672393799,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.2886,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.49189165234565735,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.2961,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.46698352694511414,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.2848,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.4658431112766266,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.2855,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.4618423581123352,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.3042,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4524480402469635,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.294,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.45932796597480774,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.2888,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.4560113847255707,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2803,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5098149180412292,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.2891,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.48962903022766113,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.2851,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.453960657119751,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.2882,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.4584779143333435,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.2826,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.5073868036270142,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.2869,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.44910764694213867,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2757,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.4433349370956421,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.2893,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.47304221987724304,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.2843,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.4525999426841736,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.2994,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.4573155641555786,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.2938,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.48174911737442017,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.2856,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.46954238414764404,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2776,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5373663306236267,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.2932,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.5043812394142151,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.2862,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5714966654777527,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.2921,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.455990731716156,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.2845,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.5092963576316833,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.285,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.48074761033058167,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2804,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.45509079098701477,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.2914,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5141335725784302,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2732,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.4804074764251709,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2793,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.5113164186477661,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.2788,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.47118645906448364,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.2881,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.43845856189727783,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.295,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4637317955493927,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.2868,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.4513588547706604,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.2873,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.4578799307346344,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.2842,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.4371070861816406,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.2785,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.4726965129375458,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.2929,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.47486382722854614,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2788,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.45456239581108093,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.2795,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.47938287258148193,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.267,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4586459696292877,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.2778,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.47361212968826294,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.2835,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.4621153473854065,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2799,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.4620632529258728,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.2777,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.44179707765579224,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.2967,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.4475201666355133,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.2809,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.46693798899650574,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.2972,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.46138525009155273,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.2823,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.4662035405635834,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.2849,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5317665934562683,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.2836,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.45609912276268005,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.2859,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.5028151869773865,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.2896,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.45612475275993347,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.2882,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.44466960430145264,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2791,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.48849114775657654,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2781,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.4417634904384613,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.2864,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.47263526916503906,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.2816,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.44584980607032776,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.2843,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.4460774064064026,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.2832,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.4497992992401123,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.2877,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5228947401046753,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.2836,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5239676237106323,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.2799,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.4485435485839844,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.2874,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.4365707337856293,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.281,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.4710332453250885,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.2779,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.44824421405792236,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.2774,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.5048888325691223,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.3009,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.4799298644065857,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.2752,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.45537424087524414,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.2851,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.4496147632598877,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.2857,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.45148664712905884,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.288,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.46457111835479736,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.2861,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.47459596395492554,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.2876,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.4512699246406555,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2681,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.43789952993392944,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.2792,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.45485296845436096,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.2893,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.4714756906032562,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.273,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.5224752426147461,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.2941,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.4842609167098999,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.28,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.449357807636261,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.2872,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.4483051598072052,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.2871,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.4558925926685333,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2824,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.46060770750045776,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2778,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.5012911558151245,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2835,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.45284098386764526,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.2795,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.4671890139579773,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.2791,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.45497480034828186,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.271,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.4666505455970764,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.2643,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.455986887216568,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.2788,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.45877569913864136,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.275,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.44860130548477173,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.2868,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.4627366065979004,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2733,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.4672466516494751,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.2797,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.45182323455810547,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2785,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.5160452127456665,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2792,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.45589128136634827,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2766,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4494640529155731,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.2879,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.455919474363327,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.2867,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.45683205127716064,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.2933,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.4401494860649109,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.2841,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.4585829973220825,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.2862,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.4692978858947754,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2685,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.47096991539001465,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2777,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.4629080295562744,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.2904,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.466360479593277,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2625,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.4717717468738556,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.2835,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.45339369773864746,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.2833,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.456876665353775,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.2787,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.45005175471305847,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.292,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.46648889780044556,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.2747,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.44402530789375305,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.2842,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3760031461715698,
      "learning_rate": 1.5020833487815421e-05,
      "loss": 1.2761,
      "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
}