{
  "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": 1.267115831375122,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.4878,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.086830735206604,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.818,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8537070751190186,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.1887,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 1.0403074026107788,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.6339,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.630695641040802,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.1053,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.9471701383590698,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.6251,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 0.6331472396850586,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.2957,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.1874464750289917,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.0742,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 1.1561050415039062,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.8899,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.7020412087440491,
      "learning_rate": 0.0015,
      "loss": 3.748,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 1.058180570602417,
      "learning_rate": 0.0015,
      "loss": 3.5931,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.8881299495697021,
      "learning_rate": 0.0015,
      "loss": 3.4895,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.7307445406913757,
      "learning_rate": 0.0015,
      "loss": 3.3785,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.9191378355026245,
      "learning_rate": 0.0015,
      "loss": 3.2912,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.7868037819862366,
      "learning_rate": 0.0015,
      "loss": 3.2078,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 1.0002555847167969,
      "learning_rate": 0.0015,
      "loss": 3.1415,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.811924934387207,
      "learning_rate": 0.0015,
      "loss": 3.0762,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.707933783531189,
      "learning_rate": 0.0015,
      "loss": 3.0415,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.7100071907043457,
      "learning_rate": 0.0015,
      "loss": 2.9773,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.7788935303688049,
      "learning_rate": 0.0015,
      "loss": 2.9367,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.7069195508956909,
      "learning_rate": 0.0015,
      "loss": 2.8957,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.9104593396186829,
      "learning_rate": 0.0015,
      "loss": 2.8617,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.830557644367218,
      "learning_rate": 0.0015,
      "loss": 2.8114,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.8437513709068298,
      "learning_rate": 0.0015,
      "loss": 2.7782,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.9658205509185791,
      "learning_rate": 0.0015,
      "loss": 2.7484,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.8535057902336121,
      "learning_rate": 0.0015,
      "loss": 2.7181,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 1.0181001424789429,
      "learning_rate": 0.0015,
      "loss": 2.6857,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.6967226266860962,
      "learning_rate": 0.0015,
      "loss": 2.6492,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 1.1211285591125488,
      "learning_rate": 0.0015,
      "loss": 2.6307,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.9659064412117004,
      "learning_rate": 0.0015,
      "loss": 2.6054,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.737605094909668,
      "learning_rate": 0.0015,
      "loss": 2.582,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 1.2234512567520142,
      "learning_rate": 0.0015,
      "loss": 2.551,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.8145862221717834,
      "learning_rate": 0.0015,
      "loss": 2.5459,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 1.0696101188659668,
      "learning_rate": 0.0015,
      "loss": 2.5195,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.8013420701026917,
      "learning_rate": 0.0015,
      "loss": 2.4817,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.9070076942443848,
      "learning_rate": 0.0015,
      "loss": 2.4644,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.8164359927177429,
      "learning_rate": 0.0015,
      "loss": 2.4497,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 1.052100419998169,
      "learning_rate": 0.0015,
      "loss": 2.4406,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 1.006508231163025,
      "learning_rate": 0.0015,
      "loss": 2.4185,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.6804226636886597,
      "learning_rate": 0.0015,
      "loss": 2.3939,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 1.1502838134765625,
      "learning_rate": 0.0015,
      "loss": 2.3794,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.8598443865776062,
      "learning_rate": 0.0015,
      "loss": 2.3654,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.9380781650543213,
      "learning_rate": 0.0015,
      "loss": 2.3439,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.9013979434967041,
      "learning_rate": 0.0015,
      "loss": 2.3224,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.9391945600509644,
      "learning_rate": 0.0015,
      "loss": 2.3232,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.9694176316261292,
      "learning_rate": 0.0015,
      "loss": 2.305,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 1.1152819395065308,
      "learning_rate": 0.0015,
      "loss": 2.2843,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.8284817337989807,
      "learning_rate": 0.0015,
      "loss": 2.2773,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.8194335699081421,
      "learning_rate": 0.0015,
      "loss": 2.252,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 1.0222529172897339,
      "learning_rate": 0.0015,
      "loss": 2.2524,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.7683922648429871,
      "learning_rate": 0.0015,
      "loss": 2.2369,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.917959988117218,
      "learning_rate": 0.0015,
      "loss": 2.2205,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 1.0505857467651367,
      "learning_rate": 0.0015,
      "loss": 2.2095,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.8172610998153687,
      "learning_rate": 0.0015,
      "loss": 2.1846,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.8242407441139221,
      "learning_rate": 0.0015,
      "loss": 2.1885,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.74329674243927,
      "learning_rate": 0.0015,
      "loss": 2.1555,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.7791258692741394,
      "learning_rate": 0.0015,
      "loss": 2.159,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 1.2360996007919312,
      "learning_rate": 0.0015,
      "loss": 2.1631,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.662500262260437,
      "learning_rate": 0.0015,
      "loss": 2.1354,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.267376184463501,
      "learning_rate": 0.0015,
      "loss": 2.117,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.7871881127357483,
      "learning_rate": 0.0015,
      "loss": 2.1222,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.6757326722145081,
      "learning_rate": 0.0015,
      "loss": 2.1043,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.6942896842956543,
      "learning_rate": 0.0015,
      "loss": 2.0898,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.9564072489738464,
      "learning_rate": 0.0015,
      "loss": 2.1062,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.9712061285972595,
      "learning_rate": 0.0015,
      "loss": 2.0942,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 1.0676331520080566,
      "learning_rate": 0.0015,
      "loss": 2.0844,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.7057194709777832,
      "learning_rate": 0.0015,
      "loss": 2.0658,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.7662429809570312,
      "learning_rate": 0.0015,
      "loss": 2.054,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.8145813941955566,
      "learning_rate": 0.0015,
      "loss": 2.064,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.8595874309539795,
      "learning_rate": 0.0015,
      "loss": 2.0488,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.7613713145256042,
      "learning_rate": 0.0015,
      "loss": 2.0349,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.9556564092636108,
      "learning_rate": 0.0015,
      "loss": 2.017,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.8757444620132446,
      "learning_rate": 0.0015,
      "loss": 2.0155,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.8322359919548035,
      "learning_rate": 0.0015,
      "loss": 2.0262,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.8489624261856079,
      "learning_rate": 0.0015,
      "loss": 2.0044,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 0.9894970059394836,
      "learning_rate": 0.0015,
      "loss": 2.0049,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.7704083919525146,
      "learning_rate": 0.0015,
      "loss": 1.9892,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.8131700754165649,
      "learning_rate": 0.0015,
      "loss": 1.9873,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.6630898118019104,
      "learning_rate": 0.0015,
      "loss": 1.993,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 1.0395071506500244,
      "learning_rate": 0.0015,
      "loss": 1.9718,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.8773938417434692,
      "learning_rate": 0.0015,
      "loss": 1.9574,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.9530850648880005,
      "learning_rate": 0.0015,
      "loss": 1.9641,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.7528932094573975,
      "learning_rate": 0.0015,
      "loss": 1.967,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.7244347333908081,
      "learning_rate": 0.0015,
      "loss": 1.9426,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.8388285040855408,
      "learning_rate": 0.0015,
      "loss": 1.9452,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.6546905636787415,
      "learning_rate": 0.0015,
      "loss": 1.9389,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.7514728903770447,
      "learning_rate": 0.0015,
      "loss": 1.9398,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.8813492655754089,
      "learning_rate": 0.0015,
      "loss": 1.9308,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.7346323132514954,
      "learning_rate": 0.0015,
      "loss": 1.9256,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 1.3726916313171387,
      "learning_rate": 0.0015,
      "loss": 1.9269,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.7512115240097046,
      "learning_rate": 0.0015,
      "loss": 1.9277,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.6616615056991577,
      "learning_rate": 0.0015,
      "loss": 1.9149,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.7120899558067322,
      "learning_rate": 0.0015,
      "loss": 1.8987,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.7128753066062927,
      "learning_rate": 0.0015,
      "loss": 1.9079,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 1.0504094362258911,
      "learning_rate": 0.0015,
      "loss": 1.8872,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.6815115213394165,
      "learning_rate": 0.0015,
      "loss": 1.8965,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 1.2939375638961792,
      "learning_rate": 0.0015,
      "loss": 1.8964,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.7509056329727173,
      "learning_rate": 0.0015,
      "loss": 1.8753,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.7060526013374329,
      "learning_rate": 0.0015,
      "loss": 1.8772,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.7411452531814575,
      "learning_rate": 0.0015,
      "loss": 1.8865,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.71030592918396,
      "learning_rate": 0.0015,
      "loss": 1.8803,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.8691297769546509,
      "learning_rate": 0.0015,
      "loss": 1.8729,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.8365240693092346,
      "learning_rate": 0.0015,
      "loss": 1.8733,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 1.012802243232727,
      "learning_rate": 0.0015,
      "loss": 1.8601,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.7322136759757996,
      "learning_rate": 0.0015,
      "loss": 1.8612,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.6655473709106445,
      "learning_rate": 0.0015,
      "loss": 1.8496,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.642715573310852,
      "learning_rate": 0.0015,
      "loss": 1.8397,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.7579887509346008,
      "learning_rate": 0.0015,
      "loss": 1.8407,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 1.0999778509140015,
      "learning_rate": 0.0015,
      "loss": 1.8539,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.7269627451896667,
      "learning_rate": 0.0015,
      "loss": 1.8403,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.6931649446487427,
      "learning_rate": 0.0015,
      "loss": 1.8427,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.8667834997177124,
      "learning_rate": 0.0015,
      "loss": 1.8355,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.7619576454162598,
      "learning_rate": 0.0015,
      "loss": 1.8222,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.7043116092681885,
      "learning_rate": 0.0015,
      "loss": 1.8362,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 0.983506977558136,
      "learning_rate": 0.0015,
      "loss": 1.8228,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.7836828231811523,
      "learning_rate": 0.0015,
      "loss": 1.832,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.6655563116073608,
      "learning_rate": 0.0015,
      "loss": 1.8178,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.8217303156852722,
      "learning_rate": 0.0015,
      "loss": 1.8054,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.7065832614898682,
      "learning_rate": 0.0015,
      "loss": 1.8076,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 1.3760422468185425,
      "learning_rate": 0.0015,
      "loss": 1.8235,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.8185279369354248,
      "learning_rate": 0.0015,
      "loss": 1.822,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.727121114730835,
      "learning_rate": 0.0015,
      "loss": 1.7928,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.7874573469161987,
      "learning_rate": 0.0015,
      "loss": 1.7883,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.6458255052566528,
      "learning_rate": 0.0015,
      "loss": 1.8058,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.7022703886032104,
      "learning_rate": 0.0015,
      "loss": 1.8046,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 1.2606161832809448,
      "learning_rate": 0.0015,
      "loss": 1.7912,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.6535800695419312,
      "learning_rate": 0.0015,
      "loss": 1.7835,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.6609776020050049,
      "learning_rate": 0.0015,
      "loss": 1.7984,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.7237253189086914,
      "learning_rate": 0.0015,
      "loss": 1.7837,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.6983092427253723,
      "learning_rate": 0.0015,
      "loss": 1.7736,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.7950878739356995,
      "learning_rate": 0.0015,
      "loss": 1.7806,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.7377669811248779,
      "learning_rate": 0.0015,
      "loss": 1.7784,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 1.0312517881393433,
      "learning_rate": 0.0015,
      "loss": 1.7716,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.7297598719596863,
      "learning_rate": 0.0015,
      "loss": 1.7748,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.7107803821563721,
      "learning_rate": 0.0015,
      "loss": 1.777,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.6729479432106018,
      "learning_rate": 0.0015,
      "loss": 1.7673,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.6634864807128906,
      "learning_rate": 0.0015,
      "loss": 1.7648,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.6926726698875427,
      "learning_rate": 0.0015,
      "loss": 1.7565,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.8461670279502869,
      "learning_rate": 0.0015,
      "loss": 1.7643,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.9065084457397461,
      "learning_rate": 0.0015,
      "loss": 1.7484,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.68721604347229,
      "learning_rate": 0.0015,
      "loss": 1.7508,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.7580296993255615,
      "learning_rate": 0.0015,
      "loss": 1.758,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.7169002294540405,
      "learning_rate": 0.0015,
      "loss": 1.7543,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.673450231552124,
      "learning_rate": 0.0015,
      "loss": 1.7479,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.8440428376197815,
      "learning_rate": 0.0015,
      "loss": 1.7478,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.6620081663131714,
      "learning_rate": 0.0015,
      "loss": 1.7406,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.8996078968048096,
      "learning_rate": 0.0015,
      "loss": 1.7384,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.7992554306983948,
      "learning_rate": 0.0015,
      "loss": 1.7437,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.6644994616508484,
      "learning_rate": 0.0015,
      "loss": 1.7402,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.6620984077453613,
      "learning_rate": 0.0015,
      "loss": 1.7411,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.7442142367362976,
      "learning_rate": 0.0015,
      "loss": 1.7373,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.728027880191803,
      "learning_rate": 0.0015,
      "loss": 1.7397,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.680416464805603,
      "learning_rate": 0.0015,
      "loss": 1.721,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6266154646873474,
      "learning_rate": 0.0015,
      "loss": 1.7126,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.6954091191291809,
      "learning_rate": 0.0015,
      "loss": 1.7163,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.6842447519302368,
      "learning_rate": 0.0015,
      "loss": 1.7154,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.6557826995849609,
      "learning_rate": 0.0015,
      "loss": 1.7081,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.7723187804222107,
      "learning_rate": 0.0015,
      "loss": 1.721,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.6312679052352905,
      "learning_rate": 0.0015,
      "loss": 1.7185,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.7498424053192139,
      "learning_rate": 0.0015,
      "loss": 1.7233,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.6825190782546997,
      "learning_rate": 0.0015,
      "loss": 1.7168,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6334119439125061,
      "learning_rate": 0.0015,
      "loss": 1.6864,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.6661301851272583,
      "learning_rate": 0.0015,
      "loss": 1.6991,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.7066164016723633,
      "learning_rate": 0.0015,
      "loss": 1.7071,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 1.2051690816879272,
      "learning_rate": 0.0015,
      "loss": 1.6975,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.6701253056526184,
      "learning_rate": 0.0015,
      "loss": 1.6953,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.7021659016609192,
      "learning_rate": 0.0015,
      "loss": 1.7067,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.9706708192825317,
      "learning_rate": 0.0015,
      "loss": 1.693,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.7173320055007935,
      "learning_rate": 0.0015,
      "loss": 1.6879,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.661887526512146,
      "learning_rate": 0.0015,
      "loss": 1.7034,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.9063208699226379,
      "learning_rate": 0.0015,
      "loss": 1.6858,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.7057472467422485,
      "learning_rate": 0.0015,
      "loss": 1.6827,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.6843087077140808,
      "learning_rate": 0.0015,
      "loss": 1.6858,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.8571721911430359,
      "learning_rate": 0.0015,
      "loss": 1.6897,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.632314920425415,
      "learning_rate": 0.0015,
      "loss": 1.7077,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.6517012119293213,
      "learning_rate": 0.0015,
      "loss": 1.6838,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6460869908332825,
      "learning_rate": 0.0015,
      "loss": 1.6888,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.746279776096344,
      "learning_rate": 0.0015,
      "loss": 1.6894,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 1.0219115018844604,
      "learning_rate": 0.0015,
      "loss": 1.6931,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 1.011742115020752,
      "learning_rate": 0.0015,
      "loss": 1.7003,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.6796727180480957,
      "learning_rate": 0.0015,
      "loss": 1.6699,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.6485866904258728,
      "learning_rate": 0.0015,
      "loss": 1.665,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.6299443244934082,
      "learning_rate": 0.0015,
      "loss": 1.667,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.707040011882782,
      "learning_rate": 0.0015,
      "loss": 1.6646,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.6451976895332336,
      "learning_rate": 0.0015,
      "loss": 1.6718,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.6749624609947205,
      "learning_rate": 0.0015,
      "loss": 1.6782,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.6485472917556763,
      "learning_rate": 0.0015,
      "loss": 1.6741,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6271134614944458,
      "learning_rate": 0.0015,
      "loss": 1.6642,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.6472867131233215,
      "learning_rate": 0.0015,
      "loss": 1.6682,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 1.1727449893951416,
      "learning_rate": 0.0015,
      "loss": 1.6693,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 1.0739372968673706,
      "learning_rate": 0.0015,
      "loss": 1.6699,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.6262565851211548,
      "learning_rate": 0.0015,
      "loss": 1.6639,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.9380462765693665,
      "learning_rate": 0.0015,
      "loss": 1.6629,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.6129502058029175,
      "learning_rate": 0.0015,
      "loss": 1.6537,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.610660970211029,
      "learning_rate": 0.0015,
      "loss": 1.6551,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.7108441591262817,
      "learning_rate": 0.0015,
      "loss": 1.6571,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.61849445104599,
      "learning_rate": 0.0015,
      "loss": 1.6689,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.6824235916137695,
      "learning_rate": 0.0015,
      "loss": 1.65,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.7959401607513428,
      "learning_rate": 0.0015,
      "loss": 1.6474,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.7023043632507324,
      "learning_rate": 0.0015,
      "loss": 1.6458,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.9508693218231201,
      "learning_rate": 0.0015,
      "loss": 1.655,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.610762357711792,
      "learning_rate": 0.0015,
      "loss": 1.6665,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.6003895998001099,
      "learning_rate": 0.0015,
      "loss": 1.6529,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6540138125419617,
      "learning_rate": 0.0015,
      "loss": 1.6514,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.6738853454589844,
      "learning_rate": 0.0015,
      "loss": 1.6475,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.6962546706199646,
      "learning_rate": 0.0015,
      "loss": 1.6416,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.6957595944404602,
      "learning_rate": 0.0015,
      "loss": 1.6443,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.7290619015693665,
      "learning_rate": 0.0015,
      "loss": 1.6469,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.6296234726905823,
      "learning_rate": 0.0015,
      "loss": 1.6384,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.5824480652809143,
      "learning_rate": 0.0015,
      "loss": 1.6441,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.975032389163971,
      "learning_rate": 0.0015,
      "loss": 1.6376,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.5698814988136292,
      "learning_rate": 0.0015,
      "loss": 1.633,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.6816406846046448,
      "learning_rate": 0.0015,
      "loss": 1.639,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6882089972496033,
      "learning_rate": 0.0015,
      "loss": 1.6291,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.7746725082397461,
      "learning_rate": 0.0015,
      "loss": 1.6364,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.1132783889770508,
      "learning_rate": 0.0015,
      "loss": 1.6518,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.6515901684761047,
      "learning_rate": 0.0015,
      "loss": 1.6343,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.6279417276382446,
      "learning_rate": 0.0015,
      "loss": 1.6195,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.6505322456359863,
      "learning_rate": 0.0015,
      "loss": 1.6359,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.6479453444480896,
      "learning_rate": 0.0015,
      "loss": 1.6348,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.6634397506713867,
      "learning_rate": 0.0015,
      "loss": 1.6199,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.7231172323226929,
      "learning_rate": 0.0015,
      "loss": 1.6334,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.9623863697052002,
      "learning_rate": 0.0015,
      "loss": 1.6273,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.6702282428741455,
      "learning_rate": 0.0015,
      "loss": 1.6269,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.7227782607078552,
      "learning_rate": 0.0015,
      "loss": 1.6222,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.6325075626373291,
      "learning_rate": 0.0015,
      "loss": 1.6106,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.7028928399085999,
      "learning_rate": 0.0015,
      "loss": 1.6446,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.8957093358039856,
      "learning_rate": 0.0015,
      "loss": 1.6307,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 1.2684556245803833,
      "learning_rate": 0.0015,
      "loss": 1.6194,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.6649700403213501,
      "learning_rate": 0.0015,
      "loss": 1.615,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.6177629828453064,
      "learning_rate": 0.0015,
      "loss": 1.6164,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.62782222032547,
      "learning_rate": 0.0015,
      "loss": 1.6224,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.5993375778198242,
      "learning_rate": 0.0015,
      "loss": 1.6071,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.7031516432762146,
      "learning_rate": 0.0015,
      "loss": 1.6027,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.7956550717353821,
      "learning_rate": 0.0015,
      "loss": 1.6064,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.5608900189399719,
      "learning_rate": 0.0015,
      "loss": 1.6189,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5729573369026184,
      "learning_rate": 0.0015,
      "loss": 1.613,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.6592503190040588,
      "learning_rate": 0.0015,
      "loss": 1.6106,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.6544248461723328,
      "learning_rate": 0.0015,
      "loss": 1.6119,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6607205867767334,
      "learning_rate": 0.0015,
      "loss": 1.6101,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.6016968488693237,
      "learning_rate": 0.0015,
      "loss": 1.6105,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.7552834749221802,
      "learning_rate": 0.0015,
      "loss": 1.6079,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.6352930068969727,
      "learning_rate": 0.0015,
      "loss": 1.6039,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.7600632905960083,
      "learning_rate": 0.0015,
      "loss": 1.6145,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.7010210156440735,
      "learning_rate": 0.0015,
      "loss": 1.6019,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.5926560759544373,
      "learning_rate": 0.0015,
      "loss": 1.6154,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.9674777388572693,
      "learning_rate": 0.0015,
      "loss": 1.6056,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.6830757260322571,
      "learning_rate": 0.0015,
      "loss": 1.611,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.7245091795921326,
      "learning_rate": 0.0015,
      "loss": 1.5945,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.6287095546722412,
      "learning_rate": 0.0015,
      "loss": 1.5999,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.5811812281608582,
      "learning_rate": 0.0015,
      "loss": 1.5997,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.8051828742027283,
      "learning_rate": 0.0015,
      "loss": 1.6027,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.690095841884613,
      "learning_rate": 0.0015,
      "loss": 1.5991,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.6827349066734314,
      "learning_rate": 0.0015,
      "loss": 1.5996,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.5629339218139648,
      "learning_rate": 0.0015,
      "loss": 1.5949,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.8621921539306641,
      "learning_rate": 0.0015,
      "loss": 1.604,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.6353599429130554,
      "learning_rate": 0.0015,
      "loss": 1.5982,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.5872393846511841,
      "learning_rate": 0.0015,
      "loss": 1.5955,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.5922430753707886,
      "learning_rate": 0.0015,
      "loss": 1.6028,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.615271270275116,
      "learning_rate": 0.0015,
      "loss": 1.6015,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6027426719665527,
      "learning_rate": 0.0015,
      "loss": 1.599,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6611328721046448,
      "learning_rate": 0.0015,
      "loss": 1.598,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.6782243847846985,
      "learning_rate": 0.0015,
      "loss": 1.586,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.6558618545532227,
      "learning_rate": 0.0015,
      "loss": 1.5986,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.6752070188522339,
      "learning_rate": 0.0015,
      "loss": 1.5919,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.7462228536605835,
      "learning_rate": 0.0015,
      "loss": 1.5855,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.6451160907745361,
      "learning_rate": 0.0015,
      "loss": 1.5828,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.5997004508972168,
      "learning_rate": 0.0015,
      "loss": 1.5869,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.666940450668335,
      "learning_rate": 0.0015,
      "loss": 1.5865,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6081534624099731,
      "learning_rate": 0.0015,
      "loss": 1.5816,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.6612553000450134,
      "learning_rate": 0.0015,
      "loss": 1.5899,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.6863659620285034,
      "learning_rate": 0.0015,
      "loss": 1.5876,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.7044730186462402,
      "learning_rate": 0.0015,
      "loss": 1.5678,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.9346474409103394,
      "learning_rate": 0.0015,
      "loss": 1.5832,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.7210971117019653,
      "learning_rate": 0.0015,
      "loss": 1.5848,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.8036166429519653,
      "learning_rate": 0.0015,
      "loss": 1.5939,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.5704184174537659,
      "learning_rate": 0.0015,
      "loss": 1.5877,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.6624103784561157,
      "learning_rate": 0.0015,
      "loss": 1.5855,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.6007447838783264,
      "learning_rate": 0.0015,
      "loss": 1.5836,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.5330153703689575,
      "learning_rate": 0.0015,
      "loss": 1.5809,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.6317871809005737,
      "learning_rate": 0.0015,
      "loss": 1.5817,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6370088458061218,
      "learning_rate": 0.0015,
      "loss": 1.5856,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 1.0981173515319824,
      "learning_rate": 0.0015,
      "loss": 1.5637,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.8902475833892822,
      "learning_rate": 0.0015,
      "loss": 1.5675,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.6024956107139587,
      "learning_rate": 0.0015,
      "loss": 1.5698,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.5766766667366028,
      "learning_rate": 0.0015,
      "loss": 1.5725,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.7689202427864075,
      "learning_rate": 0.0015,
      "loss": 1.5739,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.7470451593399048,
      "learning_rate": 0.0015,
      "loss": 1.5627,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.6598032116889954,
      "learning_rate": 0.0015,
      "loss": 1.5685,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.6697877645492554,
      "learning_rate": 0.0015,
      "loss": 1.5751,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.7237564921379089,
      "learning_rate": 0.0015,
      "loss": 1.5744,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.5626104474067688,
      "learning_rate": 0.0015,
      "loss": 1.5679,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.6248372197151184,
      "learning_rate": 0.0015,
      "loss": 1.5649,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.5619105696678162,
      "learning_rate": 0.0015,
      "loss": 1.5715,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.8324016332626343,
      "learning_rate": 0.0015,
      "loss": 1.5809,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.6818839907646179,
      "learning_rate": 0.0015,
      "loss": 1.5626,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.8424079418182373,
      "learning_rate": 0.0015,
      "loss": 1.5555,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.5838778614997864,
      "learning_rate": 0.0015,
      "loss": 1.5597,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.7394194006919861,
      "learning_rate": 0.0015,
      "loss": 1.5782,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.5810507535934448,
      "learning_rate": 0.0015,
      "loss": 1.5541,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.6484794616699219,
      "learning_rate": 0.0015,
      "loss": 1.5625,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.895905077457428,
      "learning_rate": 0.0015,
      "loss": 1.5671,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6598799228668213,
      "learning_rate": 0.0015,
      "loss": 1.5531,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.635413408279419,
      "learning_rate": 0.0015,
      "loss": 1.5617,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.5888679027557373,
      "learning_rate": 0.0015,
      "loss": 1.5576,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.5969088077545166,
      "learning_rate": 0.0015,
      "loss": 1.5694,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.5850529074668884,
      "learning_rate": 0.0015,
      "loss": 1.5567,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.6805583238601685,
      "learning_rate": 0.0015,
      "loss": 1.5465,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.5836696624755859,
      "learning_rate": 0.0015,
      "loss": 1.5649,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.6702815294265747,
      "learning_rate": 0.0015,
      "loss": 1.5541,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6411579847335815,
      "learning_rate": 0.0015,
      "loss": 1.5566,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.5650330781936646,
      "learning_rate": 0.0015,
      "loss": 1.558,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.642634928226471,
      "learning_rate": 0.0015,
      "loss": 1.5713,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 1.084348201751709,
      "learning_rate": 0.0015,
      "loss": 1.547,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5517160892486572,
      "learning_rate": 0.0015,
      "loss": 1.5552,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.6044959425926208,
      "learning_rate": 0.0015,
      "loss": 1.5557,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.5960533618927002,
      "learning_rate": 0.0015,
      "loss": 1.5559,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.6424879431724548,
      "learning_rate": 0.0015,
      "loss": 1.5533,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.6925928592681885,
      "learning_rate": 0.0015,
      "loss": 1.5531,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.8220000863075256,
      "learning_rate": 0.0015,
      "loss": 1.5488,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.7931057214736938,
      "learning_rate": 0.0015,
      "loss": 1.5621,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.6466951966285706,
      "learning_rate": 0.0015,
      "loss": 1.5526,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.6298534274101257,
      "learning_rate": 0.0015,
      "loss": 1.5468,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.7880116105079651,
      "learning_rate": 0.0015,
      "loss": 1.5271,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.7439488172531128,
      "learning_rate": 0.0015,
      "loss": 1.5562,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.5762472748756409,
      "learning_rate": 0.0015,
      "loss": 1.5411,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.6992225646972656,
      "learning_rate": 0.0015,
      "loss": 1.5588,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.6673420071601868,
      "learning_rate": 0.0015,
      "loss": 1.5588,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.6831956505775452,
      "learning_rate": 0.0015,
      "loss": 1.5426,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.5462791919708252,
      "learning_rate": 0.0015,
      "loss": 1.5504,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.6818351149559021,
      "learning_rate": 0.0015,
      "loss": 1.5506,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.8001487851142883,
      "learning_rate": 0.0015,
      "loss": 1.5527,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.5488734245300293,
      "learning_rate": 0.0015,
      "loss": 1.5443,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.7689006924629211,
      "learning_rate": 0.0015,
      "loss": 1.5471,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6005131602287292,
      "learning_rate": 0.0015,
      "loss": 1.5421,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.6102413535118103,
      "learning_rate": 0.0015,
      "loss": 1.5415,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.7920178174972534,
      "learning_rate": 0.0015,
      "loss": 1.5482,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.6190624833106995,
      "learning_rate": 0.0015,
      "loss": 1.5438,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.6588506698608398,
      "learning_rate": 0.0015,
      "loss": 1.5379,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.5450959205627441,
      "learning_rate": 0.0015,
      "loss": 1.5435,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.568746030330658,
      "learning_rate": 0.0015,
      "loss": 1.5418,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.6272512674331665,
      "learning_rate": 0.0015,
      "loss": 1.5255,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.5547628402709961,
      "learning_rate": 0.0015,
      "loss": 1.5395,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.5821362137794495,
      "learning_rate": 0.0015,
      "loss": 1.538,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.6259918212890625,
      "learning_rate": 0.0015,
      "loss": 1.534,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.607772946357727,
      "learning_rate": 0.0015,
      "loss": 1.5313,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.9403669238090515,
      "learning_rate": 0.0015,
      "loss": 1.5476,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.5947582721710205,
      "learning_rate": 0.0015,
      "loss": 1.5519,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.5908080339431763,
      "learning_rate": 0.0015,
      "loss": 1.5425,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.6022931933403015,
      "learning_rate": 0.0015,
      "loss": 1.5232,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.7703867554664612,
      "learning_rate": 0.0015,
      "loss": 1.5456,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.6678274869918823,
      "learning_rate": 0.0015,
      "loss": 1.5387,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.6669195294380188,
      "learning_rate": 0.0015,
      "loss": 1.5499,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 1.0063786506652832,
      "learning_rate": 0.0015,
      "loss": 1.5438,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.916503369808197,
      "learning_rate": 0.0015,
      "loss": 1.5345,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.6114853024482727,
      "learning_rate": 0.0015,
      "loss": 1.5324,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.6031618714332581,
      "learning_rate": 0.0015,
      "loss": 1.5299,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.6233285665512085,
      "learning_rate": 0.0015,
      "loss": 1.5302,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.6138935685157776,
      "learning_rate": 0.0015,
      "loss": 1.5401,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.7321248650550842,
      "learning_rate": 0.0015,
      "loss": 1.5253,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.6078612804412842,
      "learning_rate": 0.0015,
      "loss": 1.5286,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.6599511504173279,
      "learning_rate": 0.0015,
      "loss": 1.5288,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.6486068367958069,
      "learning_rate": 0.0015,
      "loss": 1.5292,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.8735135197639465,
      "learning_rate": 0.0015,
      "loss": 1.5321,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.6769356727600098,
      "learning_rate": 0.0015,
      "loss": 1.5399,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.7743157148361206,
      "learning_rate": 0.0015,
      "loss": 1.5328,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.6049450039863586,
      "learning_rate": 0.0015,
      "loss": 1.5277,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.6719385385513306,
      "learning_rate": 0.0015,
      "loss": 1.5281,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.576313316822052,
      "learning_rate": 0.0015,
      "loss": 1.5382,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.6430057883262634,
      "learning_rate": 0.0015,
      "loss": 1.5422,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.5845285654067993,
      "learning_rate": 0.0015,
      "loss": 1.5302,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.612210214138031,
      "learning_rate": 0.0015,
      "loss": 1.5071,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.55598384141922,
      "learning_rate": 0.0015,
      "loss": 1.5233,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.5805742144584656,
      "learning_rate": 0.0015,
      "loss": 1.5258,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.6504670977592468,
      "learning_rate": 0.0015,
      "loss": 1.531,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.7802721858024597,
      "learning_rate": 0.0015,
      "loss": 1.5204,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.5795727968215942,
      "learning_rate": 0.0015,
      "loss": 1.535,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.6096338629722595,
      "learning_rate": 0.0015,
      "loss": 1.5217,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.63264399766922,
      "learning_rate": 0.0015,
      "loss": 1.5268,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.5819994211196899,
      "learning_rate": 0.0015,
      "loss": 1.5292,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.7298890948295593,
      "learning_rate": 0.0015,
      "loss": 1.5206,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.6244053840637207,
      "learning_rate": 0.0015,
      "loss": 1.5121,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.5711541771888733,
      "learning_rate": 0.0015,
      "loss": 1.5136,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.5333322882652283,
      "learning_rate": 0.0015,
      "loss": 1.5331,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.6488534212112427,
      "learning_rate": 0.0015,
      "loss": 1.5138,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.7239291667938232,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.9742686152458191,
      "learning_rate": 0.0015,
      "loss": 1.5187,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.6343986988067627,
      "learning_rate": 0.0015,
      "loss": 1.5083,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.6439376473426819,
      "learning_rate": 0.0015,
      "loss": 1.5098,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.8655267357826233,
      "learning_rate": 0.0015,
      "loss": 1.5106,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.6127061247825623,
      "learning_rate": 0.0015,
      "loss": 1.5127,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.6228435635566711,
      "learning_rate": 0.0015,
      "loss": 1.5086,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.5865804553031921,
      "learning_rate": 0.0015,
      "loss": 1.5151,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.6816365122795105,
      "learning_rate": 0.0015,
      "loss": 1.5133,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.8221717476844788,
      "learning_rate": 0.0015,
      "loss": 1.5163,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.5298409461975098,
      "learning_rate": 0.0015,
      "loss": 1.5161,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.5596839785575867,
      "learning_rate": 0.0015,
      "loss": 1.509,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.532119870185852,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.5433077216148376,
      "learning_rate": 0.0015,
      "loss": 1.5094,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.5456915497779846,
      "learning_rate": 0.0015,
      "loss": 1.5222,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.572923481464386,
      "learning_rate": 0.0015,
      "loss": 1.5111,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.5117266774177551,
      "learning_rate": 0.0015,
      "loss": 1.5129,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.5024316310882568,
      "learning_rate": 0.0015,
      "loss": 1.5047,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5150533318519592,
      "learning_rate": 0.0015,
      "loss": 1.513,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5249171853065491,
      "learning_rate": 0.0015,
      "loss": 1.5133,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.7801232933998108,
      "learning_rate": 0.0015,
      "loss": 1.5198,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.8610855340957642,
      "learning_rate": 0.0015,
      "loss": 1.5142,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6890126466751099,
      "learning_rate": 0.0015,
      "loss": 1.5201,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.6725890636444092,
      "learning_rate": 0.0015,
      "loss": 1.5125,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5794495940208435,
      "learning_rate": 0.0015,
      "loss": 1.5123,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.6073198914527893,
      "learning_rate": 0.0015,
      "loss": 1.504,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.7326914668083191,
      "learning_rate": 0.0015,
      "loss": 1.5107,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.6833801865577698,
      "learning_rate": 0.0015,
      "loss": 1.5025,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5677995681762695,
      "learning_rate": 0.0015,
      "loss": 1.5048,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.6422021985054016,
      "learning_rate": 0.0015,
      "loss": 1.5124,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.5912719368934631,
      "learning_rate": 0.0015,
      "loss": 1.5048,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.5334175825119019,
      "learning_rate": 0.0015,
      "loss": 1.5044,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.5514172911643982,
      "learning_rate": 0.0015,
      "loss": 1.5006,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.6370390057563782,
      "learning_rate": 0.0015,
      "loss": 1.5065,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.6007202863693237,
      "learning_rate": 0.0015,
      "loss": 1.5082,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.5911210179328918,
      "learning_rate": 0.0015,
      "loss": 1.4974,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.6960971355438232,
      "learning_rate": 0.0015,
      "loss": 1.5068,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 1.0867031812667847,
      "learning_rate": 0.0015,
      "loss": 1.5039,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.7056772112846375,
      "learning_rate": 0.0015,
      "loss": 1.5031,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.6110833883285522,
      "learning_rate": 0.0015,
      "loss": 1.509,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.7566470503807068,
      "learning_rate": 0.0015,
      "loss": 1.5115,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5677741765975952,
      "learning_rate": 0.0015,
      "loss": 1.5004,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.7918340563774109,
      "learning_rate": 0.0015,
      "loss": 1.4942,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.9699915051460266,
      "learning_rate": 0.0015,
      "loss": 1.5053,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.6517788767814636,
      "learning_rate": 0.0015,
      "loss": 1.5048,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.5126602649688721,
      "learning_rate": 0.0015,
      "loss": 1.4994,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.6962641477584839,
      "learning_rate": 0.0015,
      "loss": 1.4972,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.5313942432403564,
      "learning_rate": 0.0015,
      "loss": 1.505,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.6441898941993713,
      "learning_rate": 0.0015,
      "loss": 1.4961,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.80213862657547,
      "learning_rate": 0.0015,
      "loss": 1.4976,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.5837754607200623,
      "learning_rate": 0.0015,
      "loss": 1.5131,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.6574090719223022,
      "learning_rate": 0.0015,
      "loss": 1.4909,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.49875763058662415,
      "learning_rate": 0.0015,
      "loss": 1.4971,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.5440434217453003,
      "learning_rate": 0.0015,
      "loss": 1.4923,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.7480255365371704,
      "learning_rate": 0.0015,
      "loss": 1.4991,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.7502830028533936,
      "learning_rate": 0.0015,
      "loss": 1.5154,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.638038694858551,
      "learning_rate": 0.0015,
      "loss": 1.4985,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.5418035984039307,
      "learning_rate": 0.0015,
      "loss": 1.4996,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.7437496185302734,
      "learning_rate": 0.0015,
      "loss": 1.4974,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.7251719832420349,
      "learning_rate": 0.0015,
      "loss": 1.5027,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5135377049446106,
      "learning_rate": 0.0015,
      "loss": 1.4915,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.546423614025116,
      "learning_rate": 0.0015,
      "loss": 1.5014,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.6129780411720276,
      "learning_rate": 0.0015,
      "loss": 1.5068,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.7669563293457031,
      "learning_rate": 0.0015,
      "loss": 1.4937,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.5725547671318054,
      "learning_rate": 0.0015,
      "loss": 1.4939,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5419220924377441,
      "learning_rate": 0.0015,
      "loss": 1.498,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.6583626866340637,
      "learning_rate": 0.0015,
      "loss": 1.4798,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.6703401207923889,
      "learning_rate": 0.0015,
      "loss": 1.4996,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.7447689771652222,
      "learning_rate": 0.0015,
      "loss": 1.5011,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.6889923214912415,
      "learning_rate": 0.0015,
      "loss": 1.4989,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.5502811074256897,
      "learning_rate": 0.0015,
      "loss": 1.4869,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.59233558177948,
      "learning_rate": 0.0015,
      "loss": 1.4674,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.6495541334152222,
      "learning_rate": 0.0015,
      "loss": 1.4921,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.7146870493888855,
      "learning_rate": 0.0015,
      "loss": 1.4795,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.5922014117240906,
      "learning_rate": 0.0015,
      "loss": 1.4991,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.6273832321166992,
      "learning_rate": 0.0015,
      "loss": 1.503,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.5555224418640137,
      "learning_rate": 0.0015,
      "loss": 1.4766,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.507270097732544,
      "learning_rate": 0.0015,
      "loss": 1.4968,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.5838307738304138,
      "learning_rate": 0.0015,
      "loss": 1.4976,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.6076533198356628,
      "learning_rate": 0.0015,
      "loss": 1.4987,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.5476945638656616,
      "learning_rate": 0.0015,
      "loss": 1.4898,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.6522284150123596,
      "learning_rate": 0.0015,
      "loss": 1.4873,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.6084822416305542,
      "learning_rate": 0.0015,
      "loss": 1.485,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.5807247757911682,
      "learning_rate": 0.0015,
      "loss": 1.479,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.5520457625389099,
      "learning_rate": 0.0015,
      "loss": 1.4878,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.5583330988883972,
      "learning_rate": 0.0015,
      "loss": 1.4911,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.8494092226028442,
      "learning_rate": 0.0015,
      "loss": 1.4865,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.6600103974342346,
      "learning_rate": 0.0015,
      "loss": 1.4843,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.702814519405365,
      "learning_rate": 0.0015,
      "loss": 1.4829,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.6927804350852966,
      "learning_rate": 0.0015,
      "loss": 1.5035,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5378764271736145,
      "learning_rate": 0.0015,
      "loss": 1.4872,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.6337290406227112,
      "learning_rate": 0.0015,
      "loss": 1.4848,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5428599119186401,
      "learning_rate": 0.0015,
      "loss": 1.4964,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.7110328078269958,
      "learning_rate": 0.0015,
      "loss": 1.4783,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.55487060546875,
      "learning_rate": 0.0015,
      "loss": 1.4848,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.5764284729957581,
      "learning_rate": 0.0015,
      "loss": 1.4845,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.8112941384315491,
      "learning_rate": 0.0015,
      "loss": 1.5027,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.4965507984161377,
      "learning_rate": 0.0015,
      "loss": 1.4856,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.5271897912025452,
      "learning_rate": 0.0015,
      "loss": 1.4789,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.5478551983833313,
      "learning_rate": 0.0015,
      "loss": 1.4925,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.5441624522209167,
      "learning_rate": 0.0015,
      "loss": 1.4791,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.8168677687644958,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.9002726078033447,
      "learning_rate": 0.0015,
      "loss": 1.496,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.6249581575393677,
      "learning_rate": 0.0015,
      "loss": 1.4848,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.5859907865524292,
      "learning_rate": 0.0015,
      "loss": 1.4904,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.6127180457115173,
      "learning_rate": 0.0015,
      "loss": 1.4868,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.5924661159515381,
      "learning_rate": 0.0015,
      "loss": 1.4755,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.596203625202179,
      "learning_rate": 0.0015,
      "loss": 1.4851,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.5562626123428345,
      "learning_rate": 0.0015,
      "loss": 1.4821,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.7145407795906067,
      "learning_rate": 0.0015,
      "loss": 1.4906,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.8665297627449036,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.5451953411102295,
      "learning_rate": 0.0015,
      "loss": 1.4794,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.6681820750236511,
      "learning_rate": 0.0015,
      "loss": 1.4744,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.6579897999763489,
      "learning_rate": 0.0015,
      "loss": 1.4787,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.7042908072471619,
      "learning_rate": 0.0015,
      "loss": 1.4877,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.5207177400588989,
      "learning_rate": 0.0015,
      "loss": 1.4814,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.5590533018112183,
      "learning_rate": 0.0015,
      "loss": 1.4736,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.6029673218727112,
      "learning_rate": 0.0015,
      "loss": 1.4594,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.5923275947570801,
      "learning_rate": 0.0015,
      "loss": 1.4841,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.5961737632751465,
      "learning_rate": 0.0015,
      "loss": 1.4741,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.5873507261276245,
      "learning_rate": 0.0015,
      "loss": 1.4896,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.9630367755889893,
      "learning_rate": 0.0015,
      "loss": 1.4768,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.6927589774131775,
      "learning_rate": 0.0015,
      "loss": 1.4899,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.5878579020500183,
      "learning_rate": 0.0015,
      "loss": 1.4808,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.5937055945396423,
      "learning_rate": 0.0015,
      "loss": 1.4612,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.7419557571411133,
      "learning_rate": 0.0015,
      "loss": 1.4776,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6134268641471863,
      "learning_rate": 0.0015,
      "loss": 1.4769,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.636572539806366,
      "learning_rate": 0.0015,
      "loss": 1.4738,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.6202790141105652,
      "learning_rate": 0.0015,
      "loss": 1.4894,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.6438086032867432,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.5364874005317688,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.49875974655151367,
      "learning_rate": 0.0015,
      "loss": 1.4761,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.7342543601989746,
      "learning_rate": 0.0015,
      "loss": 1.4883,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.5707510113716125,
      "learning_rate": 0.0015,
      "loss": 1.4696,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.6253118515014648,
      "learning_rate": 0.0015,
      "loss": 1.4792,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.9086582064628601,
      "learning_rate": 0.0015,
      "loss": 1.4667,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.5875641703605652,
      "learning_rate": 0.0015,
      "loss": 1.4756,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.5278132557868958,
      "learning_rate": 0.0015,
      "loss": 1.469,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.5488489866256714,
      "learning_rate": 0.0015,
      "loss": 1.4764,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.9268010854721069,
      "learning_rate": 0.0015,
      "loss": 1.468,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.582885205745697,
      "learning_rate": 0.0015,
      "loss": 1.4772,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.6753647923469543,
      "learning_rate": 0.0015,
      "loss": 1.4842,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.8944094181060791,
      "learning_rate": 0.0015,
      "loss": 1.4754,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.5675341486930847,
      "learning_rate": 0.0015,
      "loss": 1.4751,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.5383754372596741,
      "learning_rate": 0.0015,
      "loss": 1.4789,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.7768966555595398,
      "learning_rate": 0.0015,
      "loss": 1.4675,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.5521121025085449,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.5920925140380859,
      "learning_rate": 0.0015,
      "loss": 1.4784,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.5520569086074829,
      "learning_rate": 0.0015,
      "loss": 1.4684,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.5458168983459473,
      "learning_rate": 0.0015,
      "loss": 1.4631,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.7084580659866333,
      "learning_rate": 0.0015,
      "loss": 1.4717,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.5631759166717529,
      "learning_rate": 0.0015,
      "loss": 1.4733,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.5456891655921936,
      "learning_rate": 0.0015,
      "loss": 1.4666,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.629676342010498,
      "learning_rate": 0.0015,
      "loss": 1.4623,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.6019992232322693,
      "learning_rate": 0.0015,
      "loss": 1.4642,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5304731130599976,
      "learning_rate": 0.0015,
      "loss": 1.4666,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.5179152488708496,
      "learning_rate": 0.0015,
      "loss": 1.4606,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.603348433971405,
      "learning_rate": 0.0015,
      "loss": 1.4595,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.8422991037368774,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.6259440183639526,
      "learning_rate": 0.0015,
      "loss": 1.4671,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.7041751146316528,
      "learning_rate": 0.0015,
      "loss": 1.4686,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.5876805782318115,
      "learning_rate": 0.0015,
      "loss": 1.4659,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.7383052110671997,
      "learning_rate": 0.0015,
      "loss": 1.4691,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.6773043870925903,
      "learning_rate": 0.0015,
      "loss": 1.4761,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.68573397397995,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.6186241507530212,
      "learning_rate": 0.0015,
      "loss": 1.4729,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.5245220065116882,
      "learning_rate": 0.0015,
      "loss": 1.4644,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.5127766132354736,
      "learning_rate": 0.0015,
      "loss": 1.4555,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.5364044308662415,
      "learning_rate": 0.0015,
      "loss": 1.4537,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.6824843883514404,
      "learning_rate": 0.0015,
      "loss": 1.472,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.562599778175354,
      "learning_rate": 0.0015,
      "loss": 1.4768,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.554816722869873,
      "learning_rate": 0.0015,
      "loss": 1.4608,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.5311549305915833,
      "learning_rate": 0.0015,
      "loss": 1.4594,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.6091548800468445,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5266405344009399,
      "learning_rate": 0.0015,
      "loss": 1.4602,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.5213054418563843,
      "learning_rate": 0.0015,
      "loss": 1.4663,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.5713757276535034,
      "learning_rate": 0.0015,
      "loss": 1.4757,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.7337353825569153,
      "learning_rate": 0.0015,
      "loss": 1.4669,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.5308693647384644,
      "learning_rate": 0.0015,
      "loss": 1.4755,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.6518898606300354,
      "learning_rate": 0.0015,
      "loss": 1.4697,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.6611716747283936,
      "learning_rate": 0.0015,
      "loss": 1.4649,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.5270864367485046,
      "learning_rate": 0.0015,
      "loss": 1.4711,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5916832089424133,
      "learning_rate": 0.0015,
      "loss": 1.4719,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.9487943649291992,
      "learning_rate": 0.0015,
      "loss": 1.451,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.5687475800514221,
      "learning_rate": 0.0015,
      "loss": 1.4591,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.5659775137901306,
      "learning_rate": 0.0015,
      "loss": 1.4611,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5397734045982361,
      "learning_rate": 0.0015,
      "loss": 1.4668,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.6298732757568359,
      "learning_rate": 0.0015,
      "loss": 1.4634,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.48626509308815,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.564589262008667,
      "learning_rate": 0.0015,
      "loss": 1.4363,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5767689347267151,
      "learning_rate": 0.0015,
      "loss": 1.4467,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.5449169278144836,
      "learning_rate": 0.0015,
      "loss": 1.4474,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.6358098387718201,
      "learning_rate": 0.0015,
      "loss": 1.4572,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.645729124546051,
      "learning_rate": 0.0015,
      "loss": 1.4553,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.6094199419021606,
      "learning_rate": 0.0015,
      "loss": 1.4608,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5631193518638611,
      "learning_rate": 0.0015,
      "loss": 1.4626,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.6091004610061646,
      "learning_rate": 0.0015,
      "loss": 1.4735,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.6834111213684082,
      "learning_rate": 0.0015,
      "loss": 1.456,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.7020549774169922,
      "learning_rate": 0.0015,
      "loss": 1.4693,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.6763173341751099,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.6037977337837219,
      "learning_rate": 0.0015,
      "loss": 1.461,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.6056775450706482,
      "learning_rate": 0.0015,
      "loss": 1.4543,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.5456224083900452,
      "learning_rate": 0.0015,
      "loss": 1.4633,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5920342206954956,
      "learning_rate": 0.0015,
      "loss": 1.4566,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.5288448333740234,
      "learning_rate": 0.0015,
      "loss": 1.4568,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.6668508648872375,
      "learning_rate": 0.0015,
      "loss": 1.452,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.6770985722541809,
      "learning_rate": 0.0015,
      "loss": 1.4476,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.8159175515174866,
      "learning_rate": 0.0015,
      "loss": 1.456,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.6664168238639832,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.5339956283569336,
      "learning_rate": 0.0015,
      "loss": 1.4493,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.6226065158843994,
      "learning_rate": 0.0015,
      "loss": 1.4577,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5611864328384399,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.49556663632392883,
      "learning_rate": 0.0015,
      "loss": 1.4398,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.8861162662506104,
      "learning_rate": 0.0015,
      "loss": 1.4598,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.8057578206062317,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.6614024639129639,
      "learning_rate": 0.0015,
      "loss": 1.4549,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.591598629951477,
      "learning_rate": 0.0015,
      "loss": 1.4594,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.5660628080368042,
      "learning_rate": 0.0015,
      "loss": 1.4406,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.5870352983474731,
      "learning_rate": 0.0015,
      "loss": 1.4598,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.5869897603988647,
      "learning_rate": 0.0015,
      "loss": 1.4405,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.5584044456481934,
      "learning_rate": 0.0015,
      "loss": 1.4606,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.6476332545280457,
      "learning_rate": 0.0015,
      "loss": 1.4528,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.5212739706039429,
      "learning_rate": 0.0015,
      "loss": 1.4503,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5253696441650391,
      "learning_rate": 0.0015,
      "loss": 1.4489,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.9250648021697998,
      "learning_rate": 0.0015,
      "loss": 1.4463,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.7041076421737671,
      "learning_rate": 0.0015,
      "loss": 1.4567,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.5747056603431702,
      "learning_rate": 0.0015,
      "loss": 1.4528,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.6028876900672913,
      "learning_rate": 0.0015,
      "loss": 1.4502,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 1.0948182344436646,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.6669796109199524,
      "learning_rate": 0.0015,
      "loss": 1.4637,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.592555046081543,
      "learning_rate": 0.0015,
      "loss": 1.4567,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5710325241088867,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.4339,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.5590702295303345,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.455,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5274255871772766,
      "learning_rate": 0.001444828815847542,
      "loss": 1.4558,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.5450711846351624,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.4476,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5037233233451843,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.4416,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.5511847734451294,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4488,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.8101614117622375,
      "learning_rate": 0.001366801618799797,
      "loss": 1.4366,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5284369587898254,
      "learning_rate": 0.001347962334099052,
      "loss": 1.44,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.5404887795448303,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4459,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.5109414458274841,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4358,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.5763075351715088,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.4448,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5637813210487366,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.442,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.5197626352310181,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.435,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.7691735029220581,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4293,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.6103032827377319,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4279,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5120815634727478,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.4225,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.6098062992095947,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4318,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.6241275668144226,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4225,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.5512659549713135,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.4272,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.6220884919166565,
      "learning_rate": 0.001141156021915355,
      "loss": 1.4261,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.5056501030921936,
      "learning_rate": 0.001125426919103997,
      "loss": 1.4054,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.6238651275634766,
      "learning_rate": 0.001109914618088799,
      "loss": 1.4082,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.7365453839302063,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.4211,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.5034173727035522,
      "learning_rate": 0.001079528509501728,
      "loss": 1.4228,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.5035203695297241,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.4156,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.5203441381454468,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.405,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.5264323353767395,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.4121,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.5054761171340942,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3984,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.566516637802124,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.4024,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.5494766235351562,
      "learning_rate": 0.000993270988848579,
      "loss": 1.413,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5305378437042236,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.4059,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.7373847365379333,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.4106,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.4979548156261444,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.4036,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.5357046723365784,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.4184,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.524194598197937,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.4194,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.547835111618042,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.4028,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.6181050539016724,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3926,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.6838914752006531,
      "learning_rate": 0.000888885722567627,
      "loss": 1.4034,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.5421223044395447,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.3958,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.6814435720443726,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3963,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.4749564528465271,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.3934,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.5062347650527954,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3864,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.5495785474777222,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3922,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5529202222824097,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3978,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.5035123229026794,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.3851,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.4964410066604614,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3936,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.6161483526229858,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.384,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.523698627948761,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.3899,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.49107348918914795,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3831,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.5189265608787537,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3804,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5364611744880676,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.383,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.48587021231651306,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.3859,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.5128830075263977,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.3766,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.46401098370552063,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.3791,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.48515763878822327,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.3777,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.6183516979217529,
      "learning_rate": 0.000692383699170611,
      "loss": 1.3688,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.5306627750396729,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3779,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.49693840742111206,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3816,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.5087738633155823,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.3838,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.48053449392318726,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3748,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.5376818776130676,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3777,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.6019859313964844,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3763,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.5441733002662659,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.3809,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.4988143742084503,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3668,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.9648948907852173,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3717,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.539746105670929,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.37,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.6249383687973022,
      "learning_rate": 0.000594349404202577,
      "loss": 1.3662,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.6611071825027466,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.3574,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.49944445490837097,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.3672,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5754870772361755,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3603,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.6039024591445923,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3578,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.5799662470817566,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.366,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.6918553113937378,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3553,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.5505524277687073,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3556,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5248740315437317,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.3785,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.5648968815803528,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.3593,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.5115235447883606,
      "learning_rate": 0.000517326298154212,
      "loss": 1.365,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.5762827396392822,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3666,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.5095155239105225,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3629,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.4851057827472687,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3539,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.47508159279823303,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.3475,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.49186787009239197,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3528,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.5632398128509521,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.3509,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.5433728098869324,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3636,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.4962522089481354,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.335,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.6877003908157349,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.35,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.5457785129547119,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3536,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.6001309752464294,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3588,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5580538511276245,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3451,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.6439946293830872,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.3511,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.5633798241615295,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.357,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.5077953934669495,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3548,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.4964649975299835,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3518,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.49853572249412537,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.3468,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5805441737174988,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.3572,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.5360079407691956,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3502,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.5109590888023376,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.3585,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.49240773916244507,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.3419,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5305712819099426,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.3392,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.45412856340408325,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3468,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.6645715236663818,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.356,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.548294186592102,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.3334,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.4824378788471222,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3351,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.4731684625148773,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.34,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.4967353940010071,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.3434,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.5875260233879089,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.3572,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.6720730066299438,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3519,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.5609285235404968,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3482,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.4963809847831726,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3437,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.4802923798561096,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.341,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.538068950176239,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.3383,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.6135417819023132,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.3399,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.515723466873169,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.3508,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.5416966080665588,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.355,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.5162220001220703,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.3358,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.568890392780304,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.3303,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5876977443695068,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3368,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.6197674870491028,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3339,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.5474351644515991,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3285,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.5118351578712463,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3379,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.46654650568962097,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3384,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.48345306515693665,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3253,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.5052648186683655,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.3355,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.5024939179420471,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3426,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.4708436131477356,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.3234,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.49298837780952454,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3288,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.48005419969558716,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.3146,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.48186084628105164,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.3179,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.48465681076049805,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.3211,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.46673837304115295,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.3332,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.526908814907074,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3238,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.49608245491981506,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3281,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.466810941696167,
      "learning_rate": 0.000237799979917852,
      "loss": 1.3379,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.502129852771759,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3303,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.5259256362915039,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.3365,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.500847339630127,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.3225,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.4649074077606201,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.323,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.5383433103561401,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.321,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.4654619097709656,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.3161,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.6051566004753113,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.3206,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.5371909737586975,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3315,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.5033254027366638,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.3283,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.48168230056762695,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.3181,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.4620802700519562,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.3242,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.47703951597213745,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.3345,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.4817696213722229,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.3087,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.4722531735897064,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.3298,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.5582346320152283,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.3213,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.5184255838394165,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.3382,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.47420576214790344,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.3152,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.4663408696651459,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.3216,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5505437254905701,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.3372,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.48496976494789124,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.3095,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.47624102234840393,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.3192,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.534982442855835,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.3153,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.505545437335968,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.3211,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.5112576484680176,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3274,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.4805108308792114,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3275,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.511094868183136,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.3116,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.48960891366004944,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.3125,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.4569883346557617,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.3218,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.534317672252655,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.3121,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5055877566337585,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.3301,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4627654254436493,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.3202,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.4758392572402954,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.3154,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.46388524770736694,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3238,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.4874839186668396,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.314,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.481004536151886,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.3004,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.46791186928749084,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.3086,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.4821540117263794,
      "learning_rate": 0.000142293772767289,
      "loss": 1.2983,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.47279614210128784,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.3116,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.5413178205490112,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3313,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.47640782594680786,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.3278,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.4538135528564453,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.3064,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5609554648399353,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.3165,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5607855916023254,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.3146,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.47213301062583923,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3234,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.4935596287250519,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.3181,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.47754672169685364,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.322,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.468411922454834,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.306,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.4971744418144226,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.3064,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.49825602769851685,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.3032,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.5823413729667664,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.3174,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.5061120986938477,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.3104,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.5084331035614014,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.319,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.5041303038597107,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.3127,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.49560263752937317,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.3102,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.5009322166442871,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3212,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.5117639899253845,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.3042,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.4982527196407318,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.3132,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.4800499677658081,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.3155,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4854012429714203,
      "learning_rate": 0.000104851799496331,
      "loss": 1.3173,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.4941917359828949,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.3044,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.4994807541370392,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.2993,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.5245217680931091,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.3183,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.48488539457321167,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.3063,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.5023126006126404,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.3086,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.4720862805843353,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.3082,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.4822885990142822,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.3015,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.4595473110675812,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.3143,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.4833104908466339,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.3058,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.5218775272369385,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.3149,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.5130941271781921,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.3076,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.46893396973609924,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.3165,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.4639524221420288,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.3093,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.4881211817264557,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.3184,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.45782536268234253,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.3116,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.4648894667625427,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.3099,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.46031248569488525,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3229,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.4976094365119934,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.3154,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.5178634524345398,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.3099,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.4591270685195923,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2948,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.5108558535575867,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.295,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.4601665437221527,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.3214,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.47098612785339355,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.3,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.46886900067329407,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.3178,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.48209455609321594,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.3122,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.47015511989593506,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.2993,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.48078078031539917,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.3016,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.48987117409706116,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.299,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.46399369835853577,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.3071,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.4776746332645416,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.3153,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.4886586368083954,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.3037,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.5076460838317871,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.3067,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.48554062843322754,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.322,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4504407048225403,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.3099,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.46726056933403015,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.3061,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.49006912112236023,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2991,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5269207954406738,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.3083,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.5159144997596741,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.3061,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.45548757910728455,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.3071,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.4620956778526306,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.299,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.48572149872779846,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.3064,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.46652519702911377,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2968,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.47046583890914917,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.3085,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.47184139490127563,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.3039,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.45916128158569336,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.3163,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.45790746808052063,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.3125,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.4760609269142151,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.3038,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.46838563680648804,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2975,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5150146484375,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.3137,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.4655349850654602,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.306,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5194929838180542,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.3102,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.4914683699607849,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.304,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.504937469959259,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.3056,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.502181351184845,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2998,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.479855477809906,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.3109,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5619160532951355,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2922,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.48387664556503296,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2974,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.5260951519012451,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.3008,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.4827982187271118,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.3042,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.46398162841796875,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.3161,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4830174148082733,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.3072,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.48006269335746765,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.3048,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.47011861205101013,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.3045,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.49005982279777527,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.297,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.4628673195838928,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.3106,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.4530351459980011,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2989,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.4768437147140503,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.297,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.5122866034507751,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.2865,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4650144577026367,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.2969,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.45484885573387146,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.3043,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.48278769850730896,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2988,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.46038711071014404,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.295,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.45165112614631653,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.3157,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.44949862360954285,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.3034,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.47061365842819214,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.3181,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.4659769833087921,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.3,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.49148568511009216,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.3026,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5233895182609558,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.3015,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.46923792362213135,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.3034,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.4924445152282715,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.306,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.4852265417575836,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.3083,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.46191665530204773,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2987,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.5765952467918396,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2972,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.46600738167762756,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.3027,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.49096518754959106,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.3,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.4755017161369324,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.3009,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.46774840354919434,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.3027,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.47193506360054016,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.3033,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5292438864707947,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.303,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5362447500228882,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.3004,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.4858575463294983,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.3037,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.45525240898132324,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.2989,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.4982556104660034,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.297,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.47312718629837036,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.2975,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.5107970237731934,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.3169,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.46951040625572205,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.2927,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.47983628511428833,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.304,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.4806886315345764,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.3048,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.4659721553325653,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.3048,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.46359288692474365,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.3059,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.46917709708213806,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.3071,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.46824902296066284,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2872,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.4602473974227905,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.2974,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.4930126667022705,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.3093,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.4724462628364563,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.2922,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.5241295099258423,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.3119,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.4791512191295624,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.2982,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.4800410270690918,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.306,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.484261155128479,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.3059,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.47010815143585205,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2997,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.47857558727264404,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2977,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.48201119899749756,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2991,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.48913222551345825,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.3011,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.514324426651001,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.2995,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.47292664647102356,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.29,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.46298399567604065,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.283,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.4443071484565735,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.2966,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.4724883437156677,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.2923,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.4791187345981598,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.3048,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.45725077390670776,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2913,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.4532528817653656,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.2992,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.47902509570121765,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2967,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.48579680919647217,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2981,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.4836713969707489,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2941,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4789362847805023,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.3053,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.45681995153427124,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.3029,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.4653390645980835,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.3135,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.4777088463306427,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.3055,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.4822547137737274,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.3066,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.47279828786849976,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2909,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.48346567153930664,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2946,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.46991166472435,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.3076,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.49248459935188293,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2832,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.45897576212882996,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.3021,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.4570271372795105,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.3033,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.4789440631866455,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.2982,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.4682486653327942,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.3047,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.45759284496307373,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.292,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.46701884269714355,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.3033,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3098622560501099,
      "learning_rate": 1.5020833487815421e-05,
      "loss": 1.2948,
      "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
}