{
  "best_metric": 0.357177734375,
  "best_model_checkpoint": "./results/checkpoint-7906",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 11859,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 6.6436,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 6.098,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 134.1601780588041,
      "learning_rate": 4.2e-06,
      "loss": 4.8583,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 133.05228107823655,
      "learning_rate": 1.02e-05,
      "loss": 4.0425,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 195.91524058669003,
      "learning_rate": 1.6199999999999997e-05,
      "loss": 1.9412,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 181.65188510965584,
      "learning_rate": 2.2199999999999998e-05,
      "loss": 2.3468,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 60.78222784908158,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 1.3293,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 67.96261738087682,
      "learning_rate": 3.42e-05,
      "loss": 1.4593,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 86.79690686353207,
      "learning_rate": 4.02e-05,
      "loss": 1.734,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 129.6077267827689,
      "learning_rate": 4.62e-05,
      "loss": 1.8397,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 128.778169813118,
      "learning_rate": 5.2199999999999995e-05,
      "loss": 2.128,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 66.30008183585016,
      "learning_rate": 5.82e-05,
      "loss": 1.1456,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.090731070809323,
      "learning_rate": 6.419999999999999e-05,
      "loss": 0.6283,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 74.2338848286518,
      "learning_rate": 7.02e-05,
      "loss": 1.0037,
      "step": 140
    },
    {
      "epoch": 0.04,
      "grad_norm": 31.89756478099821,
      "learning_rate": 7.62e-05,
      "loss": 0.7777,
      "step": 150
    },
    {
      "epoch": 0.04,
      "grad_norm": 72.26210302069988,
      "learning_rate": 8.22e-05,
      "loss": 1.4065,
      "step": 160
    },
    {
      "epoch": 0.04,
      "grad_norm": 40.07021523514231,
      "learning_rate": 8.819999999999999e-05,
      "loss": 1.3171,
      "step": 170
    },
    {
      "epoch": 0.05,
      "grad_norm": 42.58218330850966,
      "learning_rate": 9.419999999999999e-05,
      "loss": 1.2867,
      "step": 180
    },
    {
      "epoch": 0.05,
      "grad_norm": 23.51267950170972,
      "learning_rate": 0.0001002,
      "loss": 1.2002,
      "step": 190
    },
    {
      "epoch": 0.05,
      "grad_norm": 63.83972633293598,
      "learning_rate": 0.00010619999999999998,
      "loss": 0.9739,
      "step": 200
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.632685004324832,
      "learning_rate": 0.00011219999999999999,
      "loss": 0.6737,
      "step": 210
    },
    {
      "epoch": 0.06,
      "grad_norm": 23.885371404594686,
      "learning_rate": 0.0001182,
      "loss": 0.6572,
      "step": 220
    },
    {
      "epoch": 0.06,
      "grad_norm": 17.49905029172192,
      "learning_rate": 0.00012419999999999998,
      "loss": 0.7263,
      "step": 230
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.325576955018139,
      "learning_rate": 0.0001302,
      "loss": 0.5531,
      "step": 240
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.680218121561142,
      "learning_rate": 0.0001362,
      "loss": 0.694,
      "step": 250
    },
    {
      "epoch": 0.07,
      "grad_norm": 30.394743554491985,
      "learning_rate": 0.0001422,
      "loss": 0.9192,
      "step": 260
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.645189915067395,
      "learning_rate": 0.0001482,
      "loss": 0.8989,
      "step": 270
    },
    {
      "epoch": 0.07,
      "grad_norm": 40.455222802195145,
      "learning_rate": 0.00015419999999999998,
      "loss": 1.8282,
      "step": 280
    },
    {
      "epoch": 0.07,
      "grad_norm": 40.2149893166574,
      "learning_rate": 0.0001602,
      "loss": 1.455,
      "step": 290
    },
    {
      "epoch": 0.08,
      "grad_norm": 12.86712929272992,
      "learning_rate": 0.0001662,
      "loss": 1.3317,
      "step": 300
    },
    {
      "epoch": 0.08,
      "grad_norm": 38.867456553692456,
      "learning_rate": 0.00017219999999999998,
      "loss": 0.938,
      "step": 310
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.2591410314567515,
      "learning_rate": 0.00017819999999999997,
      "loss": 1.2697,
      "step": 320
    },
    {
      "epoch": 0.08,
      "grad_norm": 25.615561417103653,
      "learning_rate": 0.00018419999999999998,
      "loss": 1.0933,
      "step": 330
    },
    {
      "epoch": 0.09,
      "grad_norm": 25.14845233927464,
      "learning_rate": 0.0001902,
      "loss": 1.0341,
      "step": 340
    },
    {
      "epoch": 0.09,
      "grad_norm": 42.51003360957394,
      "learning_rate": 0.0001962,
      "loss": 0.8678,
      "step": 350
    },
    {
      "epoch": 0.09,
      "grad_norm": 12.128840825001165,
      "learning_rate": 0.0002022,
      "loss": 0.7396,
      "step": 360
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.964848825295658,
      "learning_rate": 0.00020819999999999996,
      "loss": 0.7442,
      "step": 370
    },
    {
      "epoch": 0.1,
      "grad_norm": 39.59298428319672,
      "learning_rate": 0.00021419999999999998,
      "loss": 1.0503,
      "step": 380
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7185329297433249,
      "learning_rate": 0.00022019999999999999,
      "loss": 1.2561,
      "step": 390
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.015303486193389,
      "learning_rate": 0.00022619999999999997,
      "loss": 1.6315,
      "step": 400
    },
    {
      "epoch": 0.1,
      "grad_norm": 39.887835051307775,
      "learning_rate": 0.00023219999999999998,
      "loss": 1.0392,
      "step": 410
    },
    {
      "epoch": 0.11,
      "grad_norm": 17.789030039861967,
      "learning_rate": 0.0002382,
      "loss": 1.0087,
      "step": 420
    },
    {
      "epoch": 0.11,
      "grad_norm": 21.119668675796365,
      "learning_rate": 0.00024419999999999997,
      "loss": 1.1151,
      "step": 430
    },
    {
      "epoch": 0.11,
      "grad_norm": 25.79842512878894,
      "learning_rate": 0.00025019999999999996,
      "loss": 1.088,
      "step": 440
    },
    {
      "epoch": 0.11,
      "grad_norm": 15.848194192167474,
      "learning_rate": 0.0002562,
      "loss": 0.7838,
      "step": 450
    },
    {
      "epoch": 0.12,
      "grad_norm": 15.398937771226997,
      "learning_rate": 0.0002622,
      "loss": 1.0873,
      "step": 460
    },
    {
      "epoch": 0.12,
      "grad_norm": 20.90332391672475,
      "learning_rate": 0.00026819999999999996,
      "loss": 0.9293,
      "step": 470
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.3634772721877364,
      "learning_rate": 0.0002742,
      "loss": 0.8292,
      "step": 480
    },
    {
      "epoch": 0.12,
      "grad_norm": 25.53471193923469,
      "learning_rate": 0.0002802,
      "loss": 0.8938,
      "step": 490
    },
    {
      "epoch": 0.13,
      "grad_norm": 23.763159586392472,
      "learning_rate": 0.00028619999999999996,
      "loss": 1.0155,
      "step": 500
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.7779129544247447,
      "learning_rate": 0.00029219999999999995,
      "loss": 1.1806,
      "step": 510
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.642477580459792,
      "learning_rate": 0.0002982,
      "loss": 0.911,
      "step": 520
    },
    {
      "epoch": 0.13,
      "grad_norm": 27.669745330451978,
      "learning_rate": 0.0002998151245708249,
      "loss": 1.1182,
      "step": 530
    },
    {
      "epoch": 0.14,
      "grad_norm": 19.858690106717045,
      "learning_rate": 0.00029955101681486044,
      "loss": 1.1516,
      "step": 540
    },
    {
      "epoch": 0.14,
      "grad_norm": 15.566308581477797,
      "learning_rate": 0.00029928690905889604,
      "loss": 1.0132,
      "step": 550
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.539866709492637,
      "learning_rate": 0.0002990228013029316,
      "loss": 0.9881,
      "step": 560
    },
    {
      "epoch": 0.14,
      "grad_norm": 33.14180454779081,
      "learning_rate": 0.00029875869354696713,
      "loss": 1.4354,
      "step": 570
    },
    {
      "epoch": 0.15,
      "grad_norm": 18.209134483516564,
      "learning_rate": 0.0002984945857910027,
      "loss": 1.2814,
      "step": 580
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.5280953085313502,
      "learning_rate": 0.0002982304780350383,
      "loss": 0.6949,
      "step": 590
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.662246550170599,
      "learning_rate": 0.00029796637027907383,
      "loss": 0.7094,
      "step": 600
    },
    {
      "epoch": 0.15,
      "grad_norm": 14.747851518871128,
      "learning_rate": 0.0002977022625231094,
      "loss": 0.7794,
      "step": 610
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.859442783782644,
      "learning_rate": 0.000297438154767145,
      "loss": 0.6411,
      "step": 620
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.96927823289309,
      "learning_rate": 0.0002971740470111805,
      "loss": 0.7762,
      "step": 630
    },
    {
      "epoch": 0.16,
      "grad_norm": 38.807838563909925,
      "learning_rate": 0.0002969099392552161,
      "loss": 0.8307,
      "step": 640
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.078824122582168,
      "learning_rate": 0.00029664583149925167,
      "loss": 0.7072,
      "step": 650
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.516433721876162,
      "learning_rate": 0.00029638172374328727,
      "loss": 0.6187,
      "step": 660
    },
    {
      "epoch": 0.17,
      "grad_norm": 17.95247613200742,
      "learning_rate": 0.0002961176159873228,
      "loss": 0.7584,
      "step": 670
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.72122601802889,
      "learning_rate": 0.00029585350823135837,
      "loss": 0.697,
      "step": 680
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.3859041275387476,
      "learning_rate": 0.0002955894004753939,
      "loss": 0.6976,
      "step": 690
    },
    {
      "epoch": 0.18,
      "grad_norm": 39.987188376655034,
      "learning_rate": 0.00029532529271942946,
      "loss": 0.8641,
      "step": 700
    },
    {
      "epoch": 0.18,
      "grad_norm": 20.0944364520948,
      "learning_rate": 0.00029506118496346506,
      "loss": 1.1189,
      "step": 710
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.747621641446118,
      "learning_rate": 0.0002947970772075006,
      "loss": 1.1795,
      "step": 720
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.473321247415091,
      "learning_rate": 0.0002945329694515362,
      "loss": 0.8382,
      "step": 730
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.69289455466116,
      "learning_rate": 0.00029426886169557176,
      "loss": 0.7501,
      "step": 740
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.623860952912251,
      "learning_rate": 0.00029400475393960736,
      "loss": 0.6535,
      "step": 750
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.418967068846029,
      "learning_rate": 0.0002937406461836429,
      "loss": 0.7884,
      "step": 760
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.599297021151376,
      "learning_rate": 0.00029347653842767845,
      "loss": 0.6215,
      "step": 770
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.1212965208019026,
      "learning_rate": 0.000293212430671714,
      "loss": 0.7527,
      "step": 780
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.25996637792451,
      "learning_rate": 0.0002929483229157496,
      "loss": 0.761,
      "step": 790
    },
    {
      "epoch": 0.2,
      "grad_norm": 21.335470985081095,
      "learning_rate": 0.00029268421515978515,
      "loss": 1.0277,
      "step": 800
    },
    {
      "epoch": 0.2,
      "grad_norm": 23.380006601416916,
      "learning_rate": 0.00029242010740382075,
      "loss": 0.8902,
      "step": 810
    },
    {
      "epoch": 0.21,
      "grad_norm": 17.392278392164588,
      "learning_rate": 0.0002921559996478563,
      "loss": 0.7655,
      "step": 820
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.304718429892404,
      "learning_rate": 0.0002918918918918919,
      "loss": 0.7682,
      "step": 830
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.8327331801754974,
      "learning_rate": 0.00029162778413592745,
      "loss": 0.6086,
      "step": 840
    },
    {
      "epoch": 0.22,
      "grad_norm": 32.97517046309539,
      "learning_rate": 0.000291363676379963,
      "loss": 0.7071,
      "step": 850
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.785667285051565,
      "learning_rate": 0.0002910995686239986,
      "loss": 0.8056,
      "step": 860
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.498089410707394,
      "learning_rate": 0.00029083546086803414,
      "loss": 0.7692,
      "step": 870
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.145717003985533,
      "learning_rate": 0.0002905713531120697,
      "loss": 0.7532,
      "step": 880
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.931492489393408,
      "learning_rate": 0.00029030724535610524,
      "loss": 0.6855,
      "step": 890
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.008189140846019,
      "learning_rate": 0.00029004313760014084,
      "loss": 0.5884,
      "step": 900
    },
    {
      "epoch": 0.23,
      "grad_norm": 28.461967837050068,
      "learning_rate": 0.0002897790298441764,
      "loss": 0.7168,
      "step": 910
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.19188899701212,
      "learning_rate": 0.000289514922088212,
      "loss": 0.5666,
      "step": 920
    },
    {
      "epoch": 0.24,
      "grad_norm": 25.557101759286965,
      "learning_rate": 0.00028925081433224753,
      "loss": 0.672,
      "step": 930
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.361742645303556,
      "learning_rate": 0.00028898670657628313,
      "loss": 0.8772,
      "step": 940
    },
    {
      "epoch": 0.24,
      "grad_norm": 27.123395529551424,
      "learning_rate": 0.0002887225988203187,
      "loss": 1.2076,
      "step": 950
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.700071862540531,
      "learning_rate": 0.00028845849106435423,
      "loss": 1.7946,
      "step": 960
    },
    {
      "epoch": 0.25,
      "grad_norm": 23.71993783255748,
      "learning_rate": 0.0002881943833083898,
      "loss": 0.72,
      "step": 970
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.997692229274783,
      "learning_rate": 0.0002879302755524254,
      "loss": 0.7971,
      "step": 980
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.935951663120656,
      "learning_rate": 0.0002876661677964609,
      "loss": 0.6953,
      "step": 990
    },
    {
      "epoch": 0.25,
      "grad_norm": 18.097171459954673,
      "learning_rate": 0.00028740206004049647,
      "loss": 0.67,
      "step": 1000
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.44313735700682,
      "learning_rate": 0.00028713795228453207,
      "loss": 0.7554,
      "step": 1010
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.311122698753216,
      "learning_rate": 0.0002868738445285676,
      "loss": 0.6804,
      "step": 1020
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.703908413146344,
      "learning_rate": 0.0002866097367726032,
      "loss": 0.7243,
      "step": 1030
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.117182879812908,
      "learning_rate": 0.00028634562901663877,
      "loss": 0.7858,
      "step": 1040
    },
    {
      "epoch": 0.27,
      "grad_norm": 73.12653578751251,
      "learning_rate": 0.0002860815212606743,
      "loss": 0.779,
      "step": 1050
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.873978056302795,
      "learning_rate": 0.0002858174135047099,
      "loss": 0.8547,
      "step": 1060
    },
    {
      "epoch": 0.27,
      "grad_norm": 27.29586810442331,
      "learning_rate": 0.00028555330574874546,
      "loss": 0.9191,
      "step": 1070
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.436221863931042,
      "learning_rate": 0.000285289197992781,
      "loss": 0.7295,
      "step": 1080
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.590273107484888,
      "learning_rate": 0.0002850250902368166,
      "loss": 0.6445,
      "step": 1090
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.040340894332966,
      "learning_rate": 0.00028476098248085216,
      "loss": 0.7118,
      "step": 1100
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.765987496408746,
      "learning_rate": 0.0002844968747248877,
      "loss": 0.6206,
      "step": 1110
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.067285304810526,
      "learning_rate": 0.0002842327669689233,
      "loss": 0.6261,
      "step": 1120
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.574834164933826,
      "learning_rate": 0.00028396865921295885,
      "loss": 0.5988,
      "step": 1130
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.639025152901535,
      "learning_rate": 0.00028370455145699445,
      "loss": 0.6358,
      "step": 1140
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.33927811216473,
      "learning_rate": 0.00028344044370103,
      "loss": 0.6151,
      "step": 1150
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.710721165097775,
      "learning_rate": 0.00028317633594506555,
      "loss": 0.5678,
      "step": 1160
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.834141261057005,
      "learning_rate": 0.0002829122281891011,
      "loss": 0.619,
      "step": 1170
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.21024200850564,
      "learning_rate": 0.0002826481204331367,
      "loss": 0.5213,
      "step": 1180
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.05609595603206,
      "learning_rate": 0.00028238401267717224,
      "loss": 0.6939,
      "step": 1190
    },
    {
      "epoch": 0.3,
      "grad_norm": 12.306991788875578,
      "learning_rate": 0.00028211990492120785,
      "loss": 0.6109,
      "step": 1200
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.95186713005476,
      "learning_rate": 0.0002818557971652434,
      "loss": 0.7076,
      "step": 1210
    },
    {
      "epoch": 0.31,
      "grad_norm": 11.767871616866502,
      "learning_rate": 0.000281591689409279,
      "loss": 0.5459,
      "step": 1220
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.2642817789157834,
      "learning_rate": 0.00028132758165331454,
      "loss": 0.7554,
      "step": 1230
    },
    {
      "epoch": 0.31,
      "grad_norm": 11.178081059149552,
      "learning_rate": 0.0002810634738973501,
      "loss": 0.6658,
      "step": 1240
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.247946181517104,
      "learning_rate": 0.00028079936614138564,
      "loss": 0.7952,
      "step": 1250
    },
    {
      "epoch": 0.32,
      "grad_norm": 26.230589608874784,
      "learning_rate": 0.00028053525838542124,
      "loss": 0.646,
      "step": 1260
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.595936896088998,
      "learning_rate": 0.0002802711506294568,
      "loss": 0.6389,
      "step": 1270
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.805550894077348,
      "learning_rate": 0.00028000704287349233,
      "loss": 0.6685,
      "step": 1280
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.854245170446724,
      "learning_rate": 0.00027974293511752793,
      "loss": 0.6371,
      "step": 1290
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.383564421649856,
      "learning_rate": 0.0002794788273615635,
      "loss": 0.6386,
      "step": 1300
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.361896888396116,
      "learning_rate": 0.0002792147196055991,
      "loss": 0.5856,
      "step": 1310
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.331278217830638,
      "learning_rate": 0.00027895061184963463,
      "loss": 0.5423,
      "step": 1320
    },
    {
      "epoch": 0.34,
      "grad_norm": 14.006733190567504,
      "learning_rate": 0.00027868650409367023,
      "loss": 0.6314,
      "step": 1330
    },
    {
      "epoch": 0.34,
      "grad_norm": 22.004723578235705,
      "learning_rate": 0.0002784223963377058,
      "loss": 0.6539,
      "step": 1340
    },
    {
      "epoch": 0.34,
      "grad_norm": 14.007983160388825,
      "learning_rate": 0.0002781582885817413,
      "loss": 0.638,
      "step": 1350
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.926185711207916,
      "learning_rate": 0.00027789418082577687,
      "loss": 0.6127,
      "step": 1360
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.136580587101956,
      "learning_rate": 0.00027763007306981247,
      "loss": 0.5212,
      "step": 1370
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.2192146030344,
      "learning_rate": 0.000277365965313848,
      "loss": 0.8935,
      "step": 1380
    },
    {
      "epoch": 0.35,
      "grad_norm": 15.497566564721284,
      "learning_rate": 0.00027710185755788357,
      "loss": 0.7205,
      "step": 1390
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.8681942113478143,
      "learning_rate": 0.00027683774980191917,
      "loss": 0.6125,
      "step": 1400
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.058902856382517,
      "learning_rate": 0.0002765736420459547,
      "loss": 0.5452,
      "step": 1410
    },
    {
      "epoch": 0.36,
      "grad_norm": 18.525701145994237,
      "learning_rate": 0.0002763095342899903,
      "loss": 0.5644,
      "step": 1420
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.3200261698006592,
      "learning_rate": 0.00027604542653402586,
      "loss": 0.4896,
      "step": 1430
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.419398855450636,
      "learning_rate": 0.0002757813187780614,
      "loss": 0.6601,
      "step": 1440
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.999656150453379,
      "learning_rate": 0.00027551721102209696,
      "loss": 0.6464,
      "step": 1450
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.417685550026515,
      "learning_rate": 0.00027525310326613256,
      "loss": 0.6317,
      "step": 1460
    },
    {
      "epoch": 0.37,
      "grad_norm": 14.045184127020653,
      "learning_rate": 0.0002749889955101681,
      "loss": 0.7528,
      "step": 1470
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.390939440920372,
      "learning_rate": 0.0002747248877542037,
      "loss": 0.7923,
      "step": 1480
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.379366418554961,
      "learning_rate": 0.00027446077999823925,
      "loss": 0.5519,
      "step": 1490
    },
    {
      "epoch": 0.38,
      "grad_norm": 24.109342420534976,
      "learning_rate": 0.00027419667224227486,
      "loss": 0.665,
      "step": 1500
    },
    {
      "epoch": 0.38,
      "grad_norm": 16.033689274285177,
      "learning_rate": 0.0002739325644863104,
      "loss": 0.6613,
      "step": 1510
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.417873320686352,
      "learning_rate": 0.00027366845673034595,
      "loss": 0.5534,
      "step": 1520
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.774105476099281,
      "learning_rate": 0.00027340434897438155,
      "loss": 0.5153,
      "step": 1530
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.156967697263397,
      "learning_rate": 0.0002731402412184171,
      "loss": 0.6953,
      "step": 1540
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.541534147261922,
      "learning_rate": 0.00027287613346245265,
      "loss": 0.627,
      "step": 1550
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.7595278692002374,
      "learning_rate": 0.0002726120257064882,
      "loss": 0.6443,
      "step": 1560
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.442783531090978,
      "learning_rate": 0.0002723479179505238,
      "loss": 0.4679,
      "step": 1570
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.415292857030091,
      "learning_rate": 0.00027208381019455934,
      "loss": 0.6031,
      "step": 1580
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.168910593006703,
      "learning_rate": 0.00027181970243859494,
      "loss": 0.5918,
      "step": 1590
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.97684498547781,
      "learning_rate": 0.0002715555946826305,
      "loss": 0.7227,
      "step": 1600
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.009332020345263,
      "learning_rate": 0.0002712914869266661,
      "loss": 0.6532,
      "step": 1610
    },
    {
      "epoch": 0.41,
      "grad_norm": 15.075405052422697,
      "learning_rate": 0.00027102737917070164,
      "loss": 0.5465,
      "step": 1620
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.557703015855992,
      "learning_rate": 0.0002707632714147372,
      "loss": 0.5378,
      "step": 1630
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.020360863839114,
      "learning_rate": 0.00027049916365877273,
      "loss": 0.3787,
      "step": 1640
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.106509945189808,
      "learning_rate": 0.00027023505590280833,
      "loss": 0.876,
      "step": 1650
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.868486395804283,
      "learning_rate": 0.0002699709481468439,
      "loss": 0.7369,
      "step": 1660
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.651700120327037,
      "learning_rate": 0.00026970684039087943,
      "loss": 0.5749,
      "step": 1670
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.8253064792153335,
      "learning_rate": 0.00026944273263491503,
      "loss": 0.5524,
      "step": 1680
    },
    {
      "epoch": 0.43,
      "grad_norm": 18.10581395417861,
      "learning_rate": 0.0002691786248789506,
      "loss": 0.6028,
      "step": 1690
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.549476686940514,
      "learning_rate": 0.0002689145171229862,
      "loss": 0.5602,
      "step": 1700
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.059658671180737,
      "learning_rate": 0.0002686504093670217,
      "loss": 0.4547,
      "step": 1710
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.638791804793982,
      "learning_rate": 0.00026838630161105727,
      "loss": 0.4888,
      "step": 1720
    },
    {
      "epoch": 0.44,
      "grad_norm": 19.886486264160368,
      "learning_rate": 0.00026812219385509287,
      "loss": 0.5952,
      "step": 1730
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.801635078783133,
      "learning_rate": 0.0002678580860991284,
      "loss": 0.5706,
      "step": 1740
    },
    {
      "epoch": 0.44,
      "grad_norm": 14.710153578070637,
      "learning_rate": 0.00026759397834316397,
      "loss": 0.3925,
      "step": 1750
    },
    {
      "epoch": 0.45,
      "grad_norm": 12.069819724878684,
      "learning_rate": 0.00026732987058719957,
      "loss": 0.5302,
      "step": 1760
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.456332856299659,
      "learning_rate": 0.0002670657628312351,
      "loss": 0.4644,
      "step": 1770
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.691992871490271,
      "learning_rate": 0.00026680165507527066,
      "loss": 0.8107,
      "step": 1780
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.993002963331337,
      "learning_rate": 0.00026653754731930626,
      "loss": 0.7395,
      "step": 1790
    },
    {
      "epoch": 0.46,
      "grad_norm": 16.442630137693712,
      "learning_rate": 0.0002662734395633418,
      "loss": 0.5488,
      "step": 1800
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.58882149775824,
      "learning_rate": 0.0002660093318073774,
      "loss": 0.5637,
      "step": 1810
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.186850399825746,
      "learning_rate": 0.00026574522405141296,
      "loss": 0.5963,
      "step": 1820
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.50607476489273,
      "learning_rate": 0.0002654811162954485,
      "loss": 0.5356,
      "step": 1830
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.555998790501263,
      "learning_rate": 0.00026521700853948405,
      "loss": 0.5437,
      "step": 1840
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.011844232153095,
      "learning_rate": 0.00026495290078351965,
      "loss": 0.4832,
      "step": 1850
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.805639693834114,
      "learning_rate": 0.0002646887930275552,
      "loss": 0.6261,
      "step": 1860
    },
    {
      "epoch": 0.47,
      "grad_norm": 12.964148309393078,
      "learning_rate": 0.0002644246852715908,
      "loss": 0.6932,
      "step": 1870
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.582112479892344,
      "learning_rate": 0.00026416057751562635,
      "loss": 0.4207,
      "step": 1880
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.670421832838283,
      "learning_rate": 0.00026389646975966195,
      "loss": 0.55,
      "step": 1890
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.0688458539036425,
      "learning_rate": 0.0002636323620036975,
      "loss": 0.5457,
      "step": 1900
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.925993593066434,
      "learning_rate": 0.00026336825424773305,
      "loss": 0.5954,
      "step": 1910
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.8754276870701614,
      "learning_rate": 0.0002631041464917686,
      "loss": 0.4933,
      "step": 1920
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.765372679005643,
      "learning_rate": 0.0002628400387358042,
      "loss": 0.4824,
      "step": 1930
    },
    {
      "epoch": 0.49,
      "grad_norm": 19.509189680845587,
      "learning_rate": 0.00026257593097983974,
      "loss": 0.6568,
      "step": 1940
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.062175321616727,
      "learning_rate": 0.0002623118232238753,
      "loss": 0.6939,
      "step": 1950
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.159611499070369,
      "learning_rate": 0.0002620477154679109,
      "loss": 0.6518,
      "step": 1960
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.979648103703146,
      "learning_rate": 0.00026178360771194644,
      "loss": 0.5731,
      "step": 1970
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.908967039411184,
      "learning_rate": 0.00026151949995598204,
      "loss": 0.6586,
      "step": 1980
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.997627786755453,
      "learning_rate": 0.0002612553922000176,
      "loss": 0.5739,
      "step": 1990
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.893499066551819,
      "learning_rate": 0.0002609912844440532,
      "loss": 0.6826,
      "step": 2000
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.580638745855676,
      "learning_rate": 0.00026072717668808873,
      "loss": 0.5423,
      "step": 2010
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.35996751327857,
      "learning_rate": 0.0002604630689321243,
      "loss": 0.665,
      "step": 2020
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.354381241634792,
      "learning_rate": 0.00026019896117615983,
      "loss": 0.5428,
      "step": 2030
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.92752738546648,
      "learning_rate": 0.00025993485342019543,
      "loss": 0.4283,
      "step": 2040
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.551392879437307,
      "learning_rate": 0.000259670745664231,
      "loss": 0.437,
      "step": 2050
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.462155992824948,
      "learning_rate": 0.0002594066379082665,
      "loss": 0.5491,
      "step": 2060
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.8080515656559095,
      "learning_rate": 0.0002591425301523021,
      "loss": 0.5937,
      "step": 2070
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.009135060561086,
      "learning_rate": 0.00025887842239633767,
      "loss": 0.5445,
      "step": 2080
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.376854781509573,
      "learning_rate": 0.00025861431464037327,
      "loss": 0.7632,
      "step": 2090
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.627380987442955,
      "learning_rate": 0.0002583502068844088,
      "loss": 0.5898,
      "step": 2100
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.95687109435049,
      "learning_rate": 0.00025808609912844437,
      "loss": 0.5336,
      "step": 2110
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.68013043671797,
      "learning_rate": 0.00025782199137247997,
      "loss": 0.5598,
      "step": 2120
    },
    {
      "epoch": 0.54,
      "grad_norm": 18.86018463552324,
      "learning_rate": 0.0002575578836165155,
      "loss": 0.6714,
      "step": 2130
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.724843407945365,
      "learning_rate": 0.00025729377586055106,
      "loss": 0.5705,
      "step": 2140
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.817247590572551,
      "learning_rate": 0.00025702966810458666,
      "loss": 0.4574,
      "step": 2150
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.917842170261981,
      "learning_rate": 0.0002567655603486222,
      "loss": 0.587,
      "step": 2160
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.627238820448371,
      "learning_rate": 0.0002565014525926578,
      "loss": 0.67,
      "step": 2170
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.530085992123676,
      "learning_rate": 0.00025623734483669336,
      "loss": 0.6785,
      "step": 2180
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.75173224313274,
      "learning_rate": 0.0002559732370807289,
      "loss": 0.4789,
      "step": 2190
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.833713937208152,
      "learning_rate": 0.0002557091293247645,
      "loss": 0.6027,
      "step": 2200
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.208140165958884,
      "learning_rate": 0.00025544502156880005,
      "loss": 0.5169,
      "step": 2210
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.736988531153909,
      "learning_rate": 0.0002551809138128356,
      "loss": 0.475,
      "step": 2220
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.128800674222159,
      "learning_rate": 0.00025491680605687115,
      "loss": 0.5872,
      "step": 2230
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.849855989751077,
      "learning_rate": 0.00025465269830090675,
      "loss": 0.6424,
      "step": 2240
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.780041585488714,
      "learning_rate": 0.0002543885905449423,
      "loss": 0.5252,
      "step": 2250
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.849384257077089,
      "learning_rate": 0.0002541244827889779,
      "loss": 0.518,
      "step": 2260
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.305548999479876,
      "learning_rate": 0.00025386037503301345,
      "loss": 0.4546,
      "step": 2270
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.507299601068285,
      "learning_rate": 0.00025359626727704905,
      "loss": 0.5043,
      "step": 2280
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.388606168597159,
      "learning_rate": 0.0002533321595210846,
      "loss": 0.3738,
      "step": 2290
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.7450663596711564,
      "learning_rate": 0.00025306805176512014,
      "loss": 0.5475,
      "step": 2300
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.644110820061742,
      "learning_rate": 0.0002528039440091557,
      "loss": 0.5164,
      "step": 2310
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.553493826310751,
      "learning_rate": 0.0002525398362531913,
      "loss": 0.6376,
      "step": 2320
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.538284810132889,
      "learning_rate": 0.00025227572849722684,
      "loss": 0.546,
      "step": 2330
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.255730794622848,
      "learning_rate": 0.0002520116207412624,
      "loss": 0.4938,
      "step": 2340
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.06756441932208,
      "learning_rate": 0.000251747512985298,
      "loss": 0.6297,
      "step": 2350
    },
    {
      "epoch": 0.6,
      "grad_norm": 13.962411434452775,
      "learning_rate": 0.00025148340522933353,
      "loss": 0.4572,
      "step": 2360
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.884013956961077,
      "learning_rate": 0.00025121929747336913,
      "loss": 0.518,
      "step": 2370
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.741167916460428,
      "learning_rate": 0.0002509551897174047,
      "loss": 0.6492,
      "step": 2380
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.5518542127821195,
      "learning_rate": 0.00025069108196144023,
      "loss": 0.4654,
      "step": 2390
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.505009953324827,
      "learning_rate": 0.00025042697420547583,
      "loss": 0.4609,
      "step": 2400
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.971887581229028,
      "learning_rate": 0.0002501628664495114,
      "loss": 0.52,
      "step": 2410
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.265527884403745,
      "learning_rate": 0.0002498987586935469,
      "loss": 0.6493,
      "step": 2420
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.334607656416119,
      "learning_rate": 0.0002496346509375825,
      "loss": 0.6016,
      "step": 2430
    },
    {
      "epoch": 0.62,
      "grad_norm": 22.377156430759506,
      "learning_rate": 0.00024937054318161807,
      "loss": 0.6253,
      "step": 2440
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.514787513790615,
      "learning_rate": 0.0002491064354256537,
      "loss": 0.4283,
      "step": 2450
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.537546355671793,
      "learning_rate": 0.0002488423276696892,
      "loss": 0.4697,
      "step": 2460
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.6373042514427474,
      "learning_rate": 0.00024857821991372477,
      "loss": 0.5043,
      "step": 2470
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.732099378394687,
      "learning_rate": 0.00024831411215776037,
      "loss": 0.3868,
      "step": 2480
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.2724796251734904,
      "learning_rate": 0.0002480500044017959,
      "loss": 0.4179,
      "step": 2490
    },
    {
      "epoch": 0.63,
      "grad_norm": 29.76737431870085,
      "learning_rate": 0.00024778589664583146,
      "loss": 0.7539,
      "step": 2500
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.237457685623285,
      "learning_rate": 0.000247521788889867,
      "loss": 0.5763,
      "step": 2510
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.630218202492239,
      "learning_rate": 0.0002472576811339026,
      "loss": 0.5585,
      "step": 2520
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.10703033449879,
      "learning_rate": 0.00024699357337793816,
      "loss": 0.5287,
      "step": 2530
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.936475845702056,
      "learning_rate": 0.00024672946562197376,
      "loss": 0.5337,
      "step": 2540
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.42630275904361,
      "learning_rate": 0.0002464653578660093,
      "loss": 0.7086,
      "step": 2550
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.118383950588261,
      "learning_rate": 0.0002462012501100449,
      "loss": 0.509,
      "step": 2560
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.663275601075963,
      "learning_rate": 0.00024593714235408046,
      "loss": 0.3979,
      "step": 2570
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.98420224056408,
      "learning_rate": 0.000245673034598116,
      "loss": 0.4386,
      "step": 2580
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.55640251867126,
      "learning_rate": 0.00024540892684215155,
      "loss": 0.2599,
      "step": 2590
    },
    {
      "epoch": 0.66,
      "grad_norm": 19.14016970813088,
      "learning_rate": 0.00024514481908618715,
      "loss": 0.8847,
      "step": 2600
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.560194655648171,
      "learning_rate": 0.0002448807113302227,
      "loss": 0.6883,
      "step": 2610
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.905497694480768,
      "learning_rate": 0.00024461660357425824,
      "loss": 0.5749,
      "step": 2620
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.9729602942440145,
      "learning_rate": 0.00024435249581829385,
      "loss": 0.5574,
      "step": 2630
    },
    {
      "epoch": 0.67,
      "grad_norm": 23.12193222317225,
      "learning_rate": 0.0002440883880623294,
      "loss": 0.568,
      "step": 2640
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.26142704553814,
      "learning_rate": 0.000243824280306365,
      "loss": 0.4193,
      "step": 2650
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.466163231555143,
      "learning_rate": 0.00024356017255040054,
      "loss": 0.5367,
      "step": 2660
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.909453729777092,
      "learning_rate": 0.00024329606479443612,
      "loss": 0.5698,
      "step": 2670
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.3942741383189237,
      "learning_rate": 0.00024303195703847166,
      "loss": 0.4005,
      "step": 2680
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.989998129309498,
      "learning_rate": 0.00024276784928250726,
      "loss": 0.457,
      "step": 2690
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.2061024523183,
      "learning_rate": 0.0002425037415265428,
      "loss": 0.6318,
      "step": 2700
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.2232790663472715,
      "learning_rate": 0.00024223963377057839,
      "loss": 0.435,
      "step": 2710
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.638300708728188,
      "learning_rate": 0.00024197552601461393,
      "loss": 0.5269,
      "step": 2720
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.589974429151266,
      "learning_rate": 0.00024171141825864948,
      "loss": 0.5127,
      "step": 2730
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.244784211013016,
      "learning_rate": 0.00024144731050268508,
      "loss": 0.3722,
      "step": 2740
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.6011074716409204,
      "learning_rate": 0.00024118320274672063,
      "loss": 0.4151,
      "step": 2750
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.474185150099283,
      "learning_rate": 0.0002409190949907562,
      "loss": 0.5032,
      "step": 2760
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.9228510791340483,
      "learning_rate": 0.00024065498723479175,
      "loss": 0.4312,
      "step": 2770
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.2931543769646483,
      "learning_rate": 0.00024039087947882735,
      "loss": 0.4744,
      "step": 2780
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.662618625971549,
      "learning_rate": 0.0002401267717228629,
      "loss": 0.515,
      "step": 2790
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.151796952250688,
      "learning_rate": 0.00023986266396689847,
      "loss": 0.5096,
      "step": 2800
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.031392721473967,
      "learning_rate": 0.00023959855621093405,
      "loss": 0.3746,
      "step": 2810
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.656245662846494,
      "learning_rate": 0.00023933444845496962,
      "loss": 0.6719,
      "step": 2820
    },
    {
      "epoch": 0.72,
      "grad_norm": 6.7056302688614196,
      "learning_rate": 0.00023907034069900517,
      "loss": 0.5397,
      "step": 2830
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.202168572804897,
      "learning_rate": 0.00023880623294304074,
      "loss": 0.4804,
      "step": 2840
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.560007457362698,
      "learning_rate": 0.00023854212518707632,
      "loss": 0.6195,
      "step": 2850
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.3492079377472574,
      "learning_rate": 0.00023827801743111186,
      "loss": 0.4512,
      "step": 2860
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.9189295158775,
      "learning_rate": 0.00023801390967514744,
      "loss": 0.4923,
      "step": 2870
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.341286959201107,
      "learning_rate": 0.00023774980191918298,
      "loss": 0.388,
      "step": 2880
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.293197096548353,
      "learning_rate": 0.00023748569416321859,
      "loss": 0.461,
      "step": 2890
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.323508263958569,
      "learning_rate": 0.00023722158640725413,
      "loss": 0.4534,
      "step": 2900
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.563996619300781,
      "learning_rate": 0.0002369574786512897,
      "loss": 0.6514,
      "step": 2910
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.430706263067635,
      "learning_rate": 0.00023669337089532525,
      "loss": 0.4485,
      "step": 2920
    },
    {
      "epoch": 0.74,
      "grad_norm": 12.991699951667346,
      "learning_rate": 0.00023642926313936086,
      "loss": 0.5424,
      "step": 2930
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.942064030959408,
      "learning_rate": 0.0002361651553833964,
      "loss": 0.6151,
      "step": 2940
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.856411254695191,
      "learning_rate": 0.00023590104762743198,
      "loss": 0.4963,
      "step": 2950
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.680294296841557,
      "learning_rate": 0.00023563693987146752,
      "loss": 0.3779,
      "step": 2960
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.48135962742209,
      "learning_rate": 0.00023537283211550313,
      "loss": 0.4061,
      "step": 2970
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.326470246659994,
      "learning_rate": 0.00023510872435953867,
      "loss": 0.4668,
      "step": 2980
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.91107122608516,
      "learning_rate": 0.00023484461660357425,
      "loss": 0.6635,
      "step": 2990
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.128915816791306,
      "learning_rate": 0.0002345805088476098,
      "loss": 0.6607,
      "step": 3000
    },
    {
      "epoch": 0.76,
      "grad_norm": 14.26004929646375,
      "learning_rate": 0.00023431640109164537,
      "loss": 0.3594,
      "step": 3010
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.5966328597923884,
      "learning_rate": 0.00023405229333568094,
      "loss": 0.518,
      "step": 3020
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.770970776458943,
      "learning_rate": 0.0002337881855797165,
      "loss": 0.5155,
      "step": 3030
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.395416602742529,
      "learning_rate": 0.00023352407782375206,
      "loss": 0.4847,
      "step": 3040
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.664220480982938,
      "learning_rate": 0.00023325997006778764,
      "loss": 0.5428,
      "step": 3050
    },
    {
      "epoch": 0.77,
      "grad_norm": 3.039233661639251,
      "learning_rate": 0.0002329958623118232,
      "loss": 0.3664,
      "step": 3060
    },
    {
      "epoch": 0.78,
      "grad_norm": 16.818377752518863,
      "learning_rate": 0.00023273175455585876,
      "loss": 0.4977,
      "step": 3070
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.947186587998635,
      "learning_rate": 0.00023246764679989436,
      "loss": 0.4966,
      "step": 3080
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.4808526444203745,
      "learning_rate": 0.0002322035390439299,
      "loss": 0.3452,
      "step": 3090
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.817395491877276,
      "learning_rate": 0.00023193943128796548,
      "loss": 0.4869,
      "step": 3100
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.532075035626836,
      "learning_rate": 0.00023167532353200103,
      "loss": 0.6532,
      "step": 3110
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.092901799047013,
      "learning_rate": 0.00023141121577603663,
      "loss": 0.4978,
      "step": 3120
    },
    {
      "epoch": 0.79,
      "grad_norm": 7.522284576872321,
      "learning_rate": 0.00023114710802007218,
      "loss": 0.5905,
      "step": 3130
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.398355036614378,
      "learning_rate": 0.00023088300026410772,
      "loss": 0.5277,
      "step": 3140
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.104585056324945,
      "learning_rate": 0.0002306188925081433,
      "loss": 0.356,
      "step": 3150
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.034537089837142,
      "learning_rate": 0.00023035478475217885,
      "loss": 0.5848,
      "step": 3160
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.857463981250751,
      "learning_rate": 0.00023009067699621445,
      "loss": 0.4997,
      "step": 3170
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.727805428786185,
      "learning_rate": 0.00022982656924025,
      "loss": 0.5272,
      "step": 3180
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.582360842526427,
      "learning_rate": 0.00022956246148428557,
      "loss": 0.4024,
      "step": 3190
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.7027802507330385,
      "learning_rate": 0.00022929835372832112,
      "loss": 0.5709,
      "step": 3200
    },
    {
      "epoch": 0.81,
      "grad_norm": 13.650501833151731,
      "learning_rate": 0.00022903424597235672,
      "loss": 0.6356,
      "step": 3210
    },
    {
      "epoch": 0.81,
      "grad_norm": 3.538410769120061,
      "learning_rate": 0.00022877013821639226,
      "loss": 0.5364,
      "step": 3220
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.421316460582023,
      "learning_rate": 0.00022850603046042784,
      "loss": 0.4789,
      "step": 3230
    },
    {
      "epoch": 0.82,
      "grad_norm": 16.382444105613263,
      "learning_rate": 0.00022824192270446338,
      "loss": 0.4113,
      "step": 3240
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.389514021127093,
      "learning_rate": 0.00022797781494849899,
      "loss": 0.3535,
      "step": 3250
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.79789184703559,
      "learning_rate": 0.00022771370719253453,
      "loss": 0.5936,
      "step": 3260
    },
    {
      "epoch": 0.83,
      "grad_norm": 3.942735153547533,
      "learning_rate": 0.00022744959943657008,
      "loss": 0.3831,
      "step": 3270
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.62777327463562,
      "learning_rate": 0.00022718549168060568,
      "loss": 0.6384,
      "step": 3280
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.950217620162037,
      "learning_rate": 0.00022692138392464123,
      "loss": 0.4969,
      "step": 3290
    },
    {
      "epoch": 0.83,
      "grad_norm": 16.399583992719105,
      "learning_rate": 0.0002266572761686768,
      "loss": 0.5946,
      "step": 3300
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.234209657306621,
      "learning_rate": 0.00022639316841271235,
      "loss": 0.4933,
      "step": 3310
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.6348460554697257,
      "learning_rate": 0.00022612906065674795,
      "loss": 0.4455,
      "step": 3320
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.5205466182331855,
      "learning_rate": 0.0002258649529007835,
      "loss": 0.4379,
      "step": 3330
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.930738370574772,
      "learning_rate": 0.00022560084514481907,
      "loss": 0.3278,
      "step": 3340
    },
    {
      "epoch": 0.85,
      "grad_norm": 3.035596470101852,
      "learning_rate": 0.00022533673738885462,
      "loss": 0.4408,
      "step": 3350
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.63755703750273,
      "learning_rate": 0.00022507262963289022,
      "loss": 0.47,
      "step": 3360
    },
    {
      "epoch": 0.85,
      "grad_norm": 3.706132430365636,
      "learning_rate": 0.00022480852187692577,
      "loss": 0.4023,
      "step": 3370
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.585769097917519,
      "learning_rate": 0.00022454441412096134,
      "loss": 0.5224,
      "step": 3380
    },
    {
      "epoch": 0.86,
      "grad_norm": 18.509301972265284,
      "learning_rate": 0.0002242803063649969,
      "loss": 0.5628,
      "step": 3390
    },
    {
      "epoch": 0.86,
      "grad_norm": 13.854335524550581,
      "learning_rate": 0.00022401619860903244,
      "loss": 0.4414,
      "step": 3400
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.18763451721793,
      "learning_rate": 0.00022375209085306804,
      "loss": 0.5034,
      "step": 3410
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.642950105893305,
      "learning_rate": 0.00022348798309710359,
      "loss": 0.6943,
      "step": 3420
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.3505931032203895,
      "learning_rate": 0.00022322387534113916,
      "loss": 0.3146,
      "step": 3430
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.8894968720438055,
      "learning_rate": 0.00022295976758517473,
      "loss": 0.4646,
      "step": 3440
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.970096250118766,
      "learning_rate": 0.0002226956598292103,
      "loss": 0.4324,
      "step": 3450
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.871263672015693,
      "learning_rate": 0.00022243155207324585,
      "loss": 0.319,
      "step": 3460
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.668433980826158,
      "learning_rate": 0.00022216744431728143,
      "loss": 0.5684,
      "step": 3470
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.257915637753227,
      "learning_rate": 0.000221903336561317,
      "loss": 0.4304,
      "step": 3480
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.170590886198052,
      "learning_rate": 0.00022163922880535258,
      "loss": 0.471,
      "step": 3490
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.015283264716949,
      "learning_rate": 0.00022137512104938812,
      "loss": 0.4489,
      "step": 3500
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.901924055363354,
      "learning_rate": 0.0002211110132934237,
      "loss": 0.4896,
      "step": 3510
    },
    {
      "epoch": 0.89,
      "grad_norm": 6.9521037069629354,
      "learning_rate": 0.00022084690553745927,
      "loss": 0.3511,
      "step": 3520
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.022012302982127,
      "learning_rate": 0.00022058279778149485,
      "loss": 0.4645,
      "step": 3530
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.693923281594277,
      "learning_rate": 0.0002203186900255304,
      "loss": 0.4974,
      "step": 3540
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.93642332608164,
      "learning_rate": 0.00022005458226956594,
      "loss": 0.6888,
      "step": 3550
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.162350094500878,
      "learning_rate": 0.00021979047451360154,
      "loss": 0.4518,
      "step": 3560
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.619365411105759,
      "learning_rate": 0.0002195263667576371,
      "loss": 0.5504,
      "step": 3570
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.025064819537734,
      "learning_rate": 0.00021926225900167266,
      "loss": 0.4704,
      "step": 3580
    },
    {
      "epoch": 0.91,
      "grad_norm": 3.8536752316347793,
      "learning_rate": 0.0002189981512457082,
      "loss": 0.3768,
      "step": 3590
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.003788040997863,
      "learning_rate": 0.0002187340434897438,
      "loss": 0.5497,
      "step": 3600
    },
    {
      "epoch": 0.91,
      "grad_norm": 3.3294110305384472,
      "learning_rate": 0.00021846993573377936,
      "loss": 0.3457,
      "step": 3610
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.7516165448430834,
      "learning_rate": 0.00021820582797781493,
      "loss": 0.4296,
      "step": 3620
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.635906796431671,
      "learning_rate": 0.00021794172022185048,
      "loss": 0.3416,
      "step": 3630
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.679756445448465,
      "learning_rate": 0.00021767761246588608,
      "loss": 0.61,
      "step": 3640
    },
    {
      "epoch": 0.92,
      "grad_norm": 6.295507977784215,
      "learning_rate": 0.00021741350470992163,
      "loss": 0.466,
      "step": 3650
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.383640244060453,
      "learning_rate": 0.0002171493969539572,
      "loss": 0.5255,
      "step": 3660
    },
    {
      "epoch": 0.93,
      "grad_norm": 3.9598013665801877,
      "learning_rate": 0.00021688528919799275,
      "loss": 0.4679,
      "step": 3670
    },
    {
      "epoch": 0.93,
      "grad_norm": 3.194966826534859,
      "learning_rate": 0.00021662118144202832,
      "loss": 0.4265,
      "step": 3680
    },
    {
      "epoch": 0.93,
      "grad_norm": 20.31059851916358,
      "learning_rate": 0.0002163570736860639,
      "loss": 0.465,
      "step": 3690
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.318095175840534,
      "learning_rate": 0.00021609296593009945,
      "loss": 0.3438,
      "step": 3700
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.192871617361268,
      "learning_rate": 0.00021582885817413502,
      "loss": 0.4376,
      "step": 3710
    },
    {
      "epoch": 0.94,
      "grad_norm": 3.3160131090019607,
      "learning_rate": 0.0002155647504181706,
      "loss": 0.2692,
      "step": 3720
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.911077873322694,
      "learning_rate": 0.00021530064266220617,
      "loss": 0.5164,
      "step": 3730
    },
    {
      "epoch": 0.95,
      "grad_norm": 6.800288114027628,
      "learning_rate": 0.00021503653490624172,
      "loss": 0.5284,
      "step": 3740
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.935626744814929,
      "learning_rate": 0.00021477242715027732,
      "loss": 0.4553,
      "step": 3750
    },
    {
      "epoch": 0.95,
      "grad_norm": 16.092871581798644,
      "learning_rate": 0.00021450831939431286,
      "loss": 0.53,
      "step": 3760
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.057776607002434,
      "learning_rate": 0.00021424421163834844,
      "loss": 0.6253,
      "step": 3770
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.551995863164397,
      "learning_rate": 0.00021398010388238399,
      "loss": 0.562,
      "step": 3780
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.456350239855379,
      "learning_rate": 0.0002137159961264196,
      "loss": 0.5066,
      "step": 3790
    },
    {
      "epoch": 0.96,
      "grad_norm": 13.67557205240678,
      "learning_rate": 0.00021345188837045513,
      "loss": 0.4223,
      "step": 3800
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.439422356201387,
      "learning_rate": 0.00021318778061449068,
      "loss": 0.5762,
      "step": 3810
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.289284921464211,
      "learning_rate": 0.00021292367285852626,
      "loss": 0.4309,
      "step": 3820
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5246952543345638,
      "learning_rate": 0.0002126595651025618,
      "loss": 0.3514,
      "step": 3830
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.90597694006036,
      "learning_rate": 0.0002123954573465974,
      "loss": 0.4376,
      "step": 3840
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.264944109699735,
      "learning_rate": 0.00021213134959063295,
      "loss": 0.4439,
      "step": 3850
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.2460797473723435,
      "learning_rate": 0.00021186724183466852,
      "loss": 0.4195,
      "step": 3860
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.676528855829685,
      "learning_rate": 0.00021160313407870407,
      "loss": 0.4114,
      "step": 3870
    },
    {
      "epoch": 0.98,
      "grad_norm": 31.036901944651785,
      "learning_rate": 0.00021133902632273967,
      "loss": 0.6423,
      "step": 3880
    },
    {
      "epoch": 0.98,
      "grad_norm": 4.181578204077942,
      "learning_rate": 0.00021107491856677522,
      "loss": 0.4257,
      "step": 3890
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.478353976688533,
      "learning_rate": 0.0002108108108108108,
      "loss": 0.5036,
      "step": 3900
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.960466961810618,
      "learning_rate": 0.00021054670305484634,
      "loss": 0.4279,
      "step": 3910
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.552220995181842,
      "learning_rate": 0.00021028259529888194,
      "loss": 0.2732,
      "step": 3920
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.502280377970437,
      "learning_rate": 0.0002100184875429175,
      "loss": 0.5032,
      "step": 3930
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.145692369646056,
      "learning_rate": 0.00020975437978695306,
      "loss": 0.3472,
      "step": 3940
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.556295887410977,
      "learning_rate": 0.00020949027203098864,
      "loss": 0.4946,
      "step": 3950
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7051721568042928,
      "eval_f1": 0.6853729055052904,
      "eval_loss": 0.54150390625,
      "eval_precision": 0.7030204500613235,
      "eval_recall": 0.7681594809856258,
      "eval_runtime": 465.1347,
      "eval_samples_per_second": 14.424,
      "eval_steps_per_second": 2.406,
      "step": 3953
    },
    {
      "epoch": 1.0,
      "grad_norm": 14.750715824311628,
      "learning_rate": 0.00020922616427502419,
      "loss": 0.4315,
      "step": 3960
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.8211901469385654,
      "learning_rate": 0.00020896205651905976,
      "loss": 0.3911,
      "step": 3970
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.203098593169805,
      "learning_rate": 0.0002086979487630953,
      "loss": 0.3039,
      "step": 3980
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.102765895975901,
      "learning_rate": 0.0002084338410071309,
      "loss": 0.4994,
      "step": 3990
    },
    {
      "epoch": 1.01,
      "grad_norm": 6.616945269045702,
      "learning_rate": 0.00020816973325116646,
      "loss": 0.2976,
      "step": 4000
    },
    {
      "epoch": 1.01,
      "grad_norm": 3.384475721456983,
      "learning_rate": 0.00020790562549520203,
      "loss": 0.3492,
      "step": 4010
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.9550298854471444,
      "learning_rate": 0.00020764151773923758,
      "loss": 0.3062,
      "step": 4020
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.0433509380121029,
      "learning_rate": 0.00020737740998327318,
      "loss": 0.3137,
      "step": 4030
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.748780651660099,
      "learning_rate": 0.00020711330222730873,
      "loss": 0.3892,
      "step": 4040
    },
    {
      "epoch": 1.02,
      "grad_norm": 3.1310767766394516,
      "learning_rate": 0.00020687560524694076,
      "loss": 0.3759,
      "step": 4050
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.015625322347851,
      "learning_rate": 0.0002066114974909763,
      "loss": 0.3816,
      "step": 4060
    },
    {
      "epoch": 1.03,
      "grad_norm": 3.1539686658190753,
      "learning_rate": 0.00020634738973501188,
      "loss": 0.3234,
      "step": 4070
    },
    {
      "epoch": 1.03,
      "grad_norm": 3.2157031129713913,
      "learning_rate": 0.00020608328197904742,
      "loss": 0.2868,
      "step": 4080
    },
    {
      "epoch": 1.03,
      "grad_norm": 14.193121008750008,
      "learning_rate": 0.00020581917422308297,
      "loss": 0.5939,
      "step": 4090
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.536030144737745,
      "learning_rate": 0.00020555506646711857,
      "loss": 0.2593,
      "step": 4100
    },
    {
      "epoch": 1.04,
      "grad_norm": 6.890194212277803,
      "learning_rate": 0.00020529095871115412,
      "loss": 0.2522,
      "step": 4110
    },
    {
      "epoch": 1.04,
      "grad_norm": 14.02265911651655,
      "learning_rate": 0.0002050268509551897,
      "loss": 0.5109,
      "step": 4120
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.357623765347387,
      "learning_rate": 0.00020476274319922527,
      "loss": 0.3223,
      "step": 4130
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.4581890825932495,
      "learning_rate": 0.00020449863544326084,
      "loss": 0.2692,
      "step": 4140
    },
    {
      "epoch": 1.05,
      "grad_norm": 3.2098633368648017,
      "learning_rate": 0.0002042345276872964,
      "loss": 0.4877,
      "step": 4150
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.526325975589617,
      "learning_rate": 0.00020397041993133196,
      "loss": 0.3352,
      "step": 4160
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.145980382522719,
      "learning_rate": 0.00020370631217536754,
      "loss": 0.2768,
      "step": 4170
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.633811660630051,
      "learning_rate": 0.0002034422044194031,
      "loss": 0.2366,
      "step": 4180
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.7060398651558935,
      "learning_rate": 0.00020317809666343866,
      "loss": 0.3208,
      "step": 4190
    },
    {
      "epoch": 1.06,
      "grad_norm": 13.154069609969191,
      "learning_rate": 0.00020291398890747423,
      "loss": 0.3124,
      "step": 4200
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.2490397915258002,
      "learning_rate": 0.0002026498811515098,
      "loss": 0.3983,
      "step": 4210
    },
    {
      "epoch": 1.07,
      "grad_norm": 4.5074853084169275,
      "learning_rate": 0.00020238577339554538,
      "loss": 0.3634,
      "step": 4220
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.432611350975554,
      "learning_rate": 0.00020212166563958093,
      "loss": 0.24,
      "step": 4230
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.5110530289918245,
      "learning_rate": 0.00020185755788361648,
      "loss": 0.3825,
      "step": 4240
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.6088717012760947,
      "learning_rate": 0.00020159345012765208,
      "loss": 0.2321,
      "step": 4250
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.368818898360172,
      "learning_rate": 0.00020132934237168762,
      "loss": 0.4175,
      "step": 4260
    },
    {
      "epoch": 1.08,
      "grad_norm": 3.0566385971464793,
      "learning_rate": 0.0002010652346157232,
      "loss": 0.366,
      "step": 4270
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.829763632525873,
      "learning_rate": 0.00020080112685975875,
      "loss": 0.2711,
      "step": 4280
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.2340116043827645,
      "learning_rate": 0.00020053701910379435,
      "loss": 0.344,
      "step": 4290
    },
    {
      "epoch": 1.09,
      "grad_norm": 5.149576390393793,
      "learning_rate": 0.0002002729113478299,
      "loss": 0.3403,
      "step": 4300
    },
    {
      "epoch": 1.09,
      "grad_norm": 3.9124613694962838,
      "learning_rate": 0.00020000880359186547,
      "loss": 0.2991,
      "step": 4310
    },
    {
      "epoch": 1.09,
      "grad_norm": 12.488201847486222,
      "learning_rate": 0.00019974469583590101,
      "loss": 0.4423,
      "step": 4320
    },
    {
      "epoch": 1.1,
      "grad_norm": 3.884982728237774,
      "learning_rate": 0.00019948058807993662,
      "loss": 0.1908,
      "step": 4330
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.049768614062495,
      "learning_rate": 0.00019921648032397216,
      "loss": 0.2456,
      "step": 4340
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.859679307035771,
      "learning_rate": 0.00019895237256800774,
      "loss": 0.3236,
      "step": 4350
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.167951613915271,
      "learning_rate": 0.00019868826481204328,
      "loss": 0.6773,
      "step": 4360
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.674904564813565,
      "learning_rate": 0.00019842415705607886,
      "loss": 0.3841,
      "step": 4370
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.491068804945932,
      "learning_rate": 0.00019816004930011443,
      "loss": 0.4128,
      "step": 4380
    },
    {
      "epoch": 1.11,
      "grad_norm": 13.741327205705838,
      "learning_rate": 0.00019789594154414998,
      "loss": 0.4301,
      "step": 4390
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.77802489618066,
      "learning_rate": 0.00019763183378818555,
      "loss": 0.3121,
      "step": 4400
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.893664716507649,
      "learning_rate": 0.00019736772603222113,
      "loss": 0.3624,
      "step": 4410
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.519451727534314,
      "learning_rate": 0.0001971036182762567,
      "loss": 0.2001,
      "step": 4420
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.2537621584939747,
      "learning_rate": 0.00019683951052029225,
      "loss": 0.4367,
      "step": 4430
    },
    {
      "epoch": 1.12,
      "grad_norm": 14.867532748144109,
      "learning_rate": 0.00019657540276432785,
      "loss": 0.354,
      "step": 4440
    },
    {
      "epoch": 1.13,
      "grad_norm": 4.670985068926129,
      "learning_rate": 0.0001963112950083634,
      "loss": 0.2072,
      "step": 4450
    },
    {
      "epoch": 1.13,
      "grad_norm": 7.179932982364767,
      "learning_rate": 0.00019604718725239897,
      "loss": 0.4212,
      "step": 4460
    },
    {
      "epoch": 1.13,
      "grad_norm": 7.238353935521112,
      "learning_rate": 0.00019578307949643452,
      "loss": 0.4491,
      "step": 4470
    },
    {
      "epoch": 1.13,
      "grad_norm": 3.0658773658616707,
      "learning_rate": 0.00019551897174047012,
      "loss": 0.4347,
      "step": 4480
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.72293023841812,
      "learning_rate": 0.00019525486398450567,
      "loss": 0.4087,
      "step": 4490
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.518136419487966,
      "learning_rate": 0.00019499075622854122,
      "loss": 0.2174,
      "step": 4500
    },
    {
      "epoch": 1.14,
      "grad_norm": 5.886714661286067,
      "learning_rate": 0.0001947266484725768,
      "loss": 0.3281,
      "step": 4510
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.445400132301571,
      "learning_rate": 0.00019446254071661234,
      "loss": 0.3407,
      "step": 4520
    },
    {
      "epoch": 1.15,
      "grad_norm": 5.494881015581473,
      "learning_rate": 0.00019419843296064794,
      "loss": 0.3496,
      "step": 4530
    },
    {
      "epoch": 1.15,
      "grad_norm": 3.151665702679958,
      "learning_rate": 0.00019393432520468348,
      "loss": 0.2611,
      "step": 4540
    },
    {
      "epoch": 1.15,
      "grad_norm": 13.532449081098884,
      "learning_rate": 0.00019367021744871906,
      "loss": 0.3941,
      "step": 4550
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.235985317755349,
      "learning_rate": 0.0001934061096927546,
      "loss": 0.3201,
      "step": 4560
    },
    {
      "epoch": 1.16,
      "grad_norm": 6.597392212444202,
      "learning_rate": 0.0001931420019367902,
      "loss": 0.4564,
      "step": 4570
    },
    {
      "epoch": 1.16,
      "grad_norm": 3.140435218562986,
      "learning_rate": 0.00019287789418082575,
      "loss": 0.2876,
      "step": 4580
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.68166804572153,
      "learning_rate": 0.00019261378642486133,
      "loss": 0.4184,
      "step": 4590
    },
    {
      "epoch": 1.16,
      "grad_norm": 5.13946024002739,
      "learning_rate": 0.00019234967866889688,
      "loss": 0.2847,
      "step": 4600
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.145968384230681,
      "learning_rate": 0.00019208557091293248,
      "loss": 0.2438,
      "step": 4610
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.403929411620646,
      "learning_rate": 0.00019182146315696802,
      "loss": 0.3279,
      "step": 4620
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.64016161912627,
      "learning_rate": 0.00019155735540100357,
      "loss": 0.3164,
      "step": 4630
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.668585643160457,
      "learning_rate": 0.00019129324764503917,
      "loss": 0.3796,
      "step": 4640
    },
    {
      "epoch": 1.18,
      "grad_norm": 14.943998906205717,
      "learning_rate": 0.00019102913988907472,
      "loss": 0.3244,
      "step": 4650
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.449038152539126,
      "learning_rate": 0.0001907650321331103,
      "loss": 0.4199,
      "step": 4660
    },
    {
      "epoch": 1.18,
      "grad_norm": 6.588596643185838,
      "learning_rate": 0.00019050092437714584,
      "loss": 0.317,
      "step": 4670
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.2478841853180118,
      "learning_rate": 0.00019023681662118144,
      "loss": 0.2229,
      "step": 4680
    },
    {
      "epoch": 1.19,
      "grad_norm": 15.5335226123234,
      "learning_rate": 0.000189972708865217,
      "loss": 0.6227,
      "step": 4690
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.419431089407398,
      "learning_rate": 0.00018970860110925256,
      "loss": 0.6332,
      "step": 4700
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.236499078980196,
      "learning_rate": 0.0001894444933532881,
      "loss": 0.3134,
      "step": 4710
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.575405335243286,
      "learning_rate": 0.0001891803855973237,
      "loss": 0.2478,
      "step": 4720
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.5276012861681565,
      "learning_rate": 0.00018891627784135926,
      "loss": 0.3107,
      "step": 4730
    },
    {
      "epoch": 1.2,
      "grad_norm": 3.915244686315227,
      "learning_rate": 0.00018865217008539483,
      "loss": 0.2537,
      "step": 4740
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.6314608277899836,
      "learning_rate": 0.00018838806232943038,
      "loss": 0.3013,
      "step": 4750
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.4281299113190515,
      "learning_rate": 0.00018812395457346598,
      "loss": 0.4033,
      "step": 4760
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.3344593707818075,
      "learning_rate": 0.00018785984681750153,
      "loss": 0.276,
      "step": 4770
    },
    {
      "epoch": 1.21,
      "grad_norm": 5.637765503351757,
      "learning_rate": 0.00018759573906153708,
      "loss": 0.4755,
      "step": 4780
    },
    {
      "epoch": 1.21,
      "grad_norm": 6.702704562048585,
      "learning_rate": 0.00018733163130557265,
      "loss": 0.386,
      "step": 4790
    },
    {
      "epoch": 1.21,
      "grad_norm": 3.1625393449402934,
      "learning_rate": 0.00018706752354960822,
      "loss": 0.357,
      "step": 4800
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.4188940831370935,
      "learning_rate": 0.0001868034157936438,
      "loss": 0.3086,
      "step": 4810
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.673434660245706,
      "learning_rate": 0.00018653930803767935,
      "loss": 0.1807,
      "step": 4820
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.738887993235016,
      "learning_rate": 0.00018627520028171492,
      "loss": 0.3299,
      "step": 4830
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.319508688160127,
      "learning_rate": 0.0001860110925257505,
      "loss": 0.4157,
      "step": 4840
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.892943360966015,
      "learning_rate": 0.00018574698476978607,
      "loss": 0.2339,
      "step": 4850
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.942098135606841,
      "learning_rate": 0.00018548287701382162,
      "loss": 0.5181,
      "step": 4860
    },
    {
      "epoch": 1.23,
      "grad_norm": 14.39700673839823,
      "learning_rate": 0.0001852187692578572,
      "loss": 0.2945,
      "step": 4870
    },
    {
      "epoch": 1.23,
      "grad_norm": 3.9132635629855486,
      "learning_rate": 0.00018495466150189276,
      "loss": 0.3401,
      "step": 4880
    },
    {
      "epoch": 1.24,
      "grad_norm": 3.0002697966747713,
      "learning_rate": 0.00018469055374592834,
      "loss": 0.2648,
      "step": 4890
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.21238117057248854,
      "learning_rate": 0.00018442644598996389,
      "loss": 0.2277,
      "step": 4900
    },
    {
      "epoch": 1.24,
      "grad_norm": 16.868097673871954,
      "learning_rate": 0.00018416233823399943,
      "loss": 0.2122,
      "step": 4910
    },
    {
      "epoch": 1.24,
      "grad_norm": 7.430973727847297,
      "learning_rate": 0.00018389823047803503,
      "loss": 0.2273,
      "step": 4920
    },
    {
      "epoch": 1.25,
      "grad_norm": 13.39392838076716,
      "learning_rate": 0.00018363412272207058,
      "loss": 0.4154,
      "step": 4930
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.789150745954093,
      "learning_rate": 0.00018337001496610615,
      "loss": 0.1768,
      "step": 4940
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.174877422869125,
      "learning_rate": 0.0001831059072101417,
      "loss": 0.5684,
      "step": 4950
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.6018630455438325,
      "learning_rate": 0.0001828417994541773,
      "loss": 0.3407,
      "step": 4960
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.953101867756693,
      "learning_rate": 0.00018257769169821285,
      "loss": 0.2901,
      "step": 4970
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.407680005230561,
      "learning_rate": 0.00018231358394224842,
      "loss": 0.4594,
      "step": 4980
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.6390957223682454,
      "learning_rate": 0.00018204947618628397,
      "loss": 0.4566,
      "step": 4990
    },
    {
      "epoch": 1.26,
      "grad_norm": 6.740560085736122,
      "learning_rate": 0.00018178536843031957,
      "loss": 0.3352,
      "step": 5000
    },
    {
      "epoch": 1.27,
      "grad_norm": 28.89110470090379,
      "learning_rate": 0.00018152126067435512,
      "loss": 0.4236,
      "step": 5010
    },
    {
      "epoch": 1.27,
      "grad_norm": 5.894457266781238,
      "learning_rate": 0.0001812571529183907,
      "loss": 0.3826,
      "step": 5020
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.90265790457071,
      "learning_rate": 0.00018099304516242624,
      "loss": 0.2422,
      "step": 5030
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.915714947527184,
      "learning_rate": 0.00018072893740646182,
      "loss": 0.4215,
      "step": 5040
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.593603092345282,
      "learning_rate": 0.0001804648296504974,
      "loss": 0.2984,
      "step": 5050
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.610545572953026,
      "learning_rate": 0.00018020072189453294,
      "loss": 0.3861,
      "step": 5060
    },
    {
      "epoch": 1.28,
      "grad_norm": 3.275694600127651,
      "learning_rate": 0.0001799366141385685,
      "loss": 0.1484,
      "step": 5070
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.9917701980802534,
      "learning_rate": 0.00017967250638260409,
      "loss": 0.354,
      "step": 5080
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.752595829803958,
      "learning_rate": 0.00017940839862663966,
      "loss": 0.2423,
      "step": 5090
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.424507759307914,
      "learning_rate": 0.0001791442908706752,
      "loss": 0.2912,
      "step": 5100
    },
    {
      "epoch": 1.29,
      "grad_norm": 5.150914763339286,
      "learning_rate": 0.0001788801831147108,
      "loss": 0.3863,
      "step": 5110
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.950047718285756,
      "learning_rate": 0.00017861607535874636,
      "loss": 0.3273,
      "step": 5120
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.0619701931886967,
      "learning_rate": 0.00017835196760278193,
      "loss": 0.3099,
      "step": 5130
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.802245354515275,
      "learning_rate": 0.00017808785984681748,
      "loss": 0.4913,
      "step": 5140
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.007925086910582,
      "learning_rate": 0.00017782375209085308,
      "loss": 0.3709,
      "step": 5150
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.125023955683686,
      "learning_rate": 0.00017755964433488862,
      "loss": 0.3352,
      "step": 5160
    },
    {
      "epoch": 1.31,
      "grad_norm": 4.89797690004093,
      "learning_rate": 0.0001772955365789242,
      "loss": 0.3096,
      "step": 5170
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.084564328245744,
      "learning_rate": 0.00017703142882295975,
      "loss": 0.2493,
      "step": 5180
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.445125880603614,
      "learning_rate": 0.0001767673210669953,
      "loss": 0.3003,
      "step": 5190
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.6377729302644695,
      "learning_rate": 0.0001765032133110309,
      "loss": 0.2035,
      "step": 5200
    },
    {
      "epoch": 1.32,
      "grad_norm": 4.028090302576488,
      "learning_rate": 0.00017623910555506644,
      "loss": 0.2136,
      "step": 5210
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.747464129470143,
      "learning_rate": 0.00017600140857469847,
      "loss": 0.4821,
      "step": 5220
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.9907365174614449,
      "learning_rate": 0.00017573730081873402,
      "loss": 0.4276,
      "step": 5230
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.291404090563887,
      "learning_rate": 0.0001754731930627696,
      "loss": 0.4383,
      "step": 5240
    },
    {
      "epoch": 1.33,
      "grad_norm": 5.156181022153768,
      "learning_rate": 0.00017520908530680514,
      "loss": 0.2204,
      "step": 5250
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.088807891623482,
      "learning_rate": 0.00017494497755084074,
      "loss": 0.4143,
      "step": 5260
    },
    {
      "epoch": 1.33,
      "grad_norm": 3.9398518664541005,
      "learning_rate": 0.0001746808697948763,
      "loss": 0.2843,
      "step": 5270
    },
    {
      "epoch": 1.34,
      "grad_norm": 6.861595569917812,
      "learning_rate": 0.00017441676203891186,
      "loss": 0.3798,
      "step": 5280
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.2732346010657967,
      "learning_rate": 0.0001741526542829474,
      "loss": 0.2078,
      "step": 5290
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.4383072436268267,
      "learning_rate": 0.000173888546526983,
      "loss": 0.3779,
      "step": 5300
    },
    {
      "epoch": 1.34,
      "grad_norm": 5.914009405172411,
      "learning_rate": 0.00017362443877101856,
      "loss": 0.2341,
      "step": 5310
    },
    {
      "epoch": 1.35,
      "grad_norm": 2.6422195515529956,
      "learning_rate": 0.0001733603310150541,
      "loss": 0.2333,
      "step": 5320
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.3857533289452737,
      "learning_rate": 0.0001730962232590897,
      "loss": 0.1961,
      "step": 5330
    },
    {
      "epoch": 1.35,
      "grad_norm": 2.059953351439511,
      "learning_rate": 0.00017283211550312525,
      "loss": 0.4881,
      "step": 5340
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.597653032886594,
      "learning_rate": 0.00017256800774716083,
      "loss": 0.424,
      "step": 5350
    },
    {
      "epoch": 1.36,
      "grad_norm": 5.171578842871781,
      "learning_rate": 0.00017230389999119638,
      "loss": 0.3054,
      "step": 5360
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.8072490729454778,
      "learning_rate": 0.00017203979223523198,
      "loss": 0.3361,
      "step": 5370
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.67098938868372,
      "learning_rate": 0.00017177568447926752,
      "loss": 0.199,
      "step": 5380
    },
    {
      "epoch": 1.36,
      "grad_norm": 7.583380258309324,
      "learning_rate": 0.0001715115767233031,
      "loss": 0.2661,
      "step": 5390
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.05363459550282,
      "learning_rate": 0.00017124746896733865,
      "loss": 0.3187,
      "step": 5400
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.7695240656878144,
      "learning_rate": 0.00017098336121137425,
      "loss": 0.3067,
      "step": 5410
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.025416113484515,
      "learning_rate": 0.0001707192534554098,
      "loss": 0.4757,
      "step": 5420
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.730906882565963,
      "learning_rate": 0.00017045514569944537,
      "loss": 0.279,
      "step": 5430
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.933729801673655,
      "learning_rate": 0.00017019103794348091,
      "loss": 0.2164,
      "step": 5440
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.069156267463432,
      "learning_rate": 0.00016992693018751652,
      "loss": 0.2898,
      "step": 5450
    },
    {
      "epoch": 1.38,
      "grad_norm": 3.052632966251752,
      "learning_rate": 0.00016966282243155206,
      "loss": 0.3861,
      "step": 5460
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.351753715632027,
      "learning_rate": 0.0001693987146755876,
      "loss": 0.3692,
      "step": 5470
    },
    {
      "epoch": 1.39,
      "grad_norm": 6.595474867843094,
      "learning_rate": 0.00016913460691962318,
      "loss": 0.4144,
      "step": 5480
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.875854869087618,
      "learning_rate": 0.00016887049916365876,
      "loss": 0.3301,
      "step": 5490
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.1474446238717586,
      "learning_rate": 0.00016860639140769433,
      "loss": 0.2172,
      "step": 5500
    },
    {
      "epoch": 1.39,
      "grad_norm": 22.776417456462937,
      "learning_rate": 0.00016836869442732634,
      "loss": 0.6439,
      "step": 5510
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.498663232395253,
      "learning_rate": 0.0001681045866713619,
      "loss": 0.3548,
      "step": 5520
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.540809246554936,
      "learning_rate": 0.00016784047891539746,
      "loss": 0.3104,
      "step": 5530
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.23205251986895,
      "learning_rate": 0.00016757637115943303,
      "loss": 0.4779,
      "step": 5540
    },
    {
      "epoch": 1.4,
      "grad_norm": 15.50340937459547,
      "learning_rate": 0.0001673122634034686,
      "loss": 0.3544,
      "step": 5550
    },
    {
      "epoch": 1.41,
      "grad_norm": 4.649484749641056,
      "learning_rate": 0.00016704815564750418,
      "loss": 0.3832,
      "step": 5560
    },
    {
      "epoch": 1.41,
      "grad_norm": 6.479502493786179,
      "learning_rate": 0.00016678404789153973,
      "loss": 0.3119,
      "step": 5570
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.2456553372360117,
      "learning_rate": 0.0001665199401355753,
      "loss": 0.2147,
      "step": 5580
    },
    {
      "epoch": 1.41,
      "grad_norm": 4.1843629031152485,
      "learning_rate": 0.00016625583237961088,
      "loss": 0.3104,
      "step": 5590
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.240618772565934,
      "learning_rate": 0.00016599172462364645,
      "loss": 0.3794,
      "step": 5600
    },
    {
      "epoch": 1.42,
      "grad_norm": 6.420744423415216,
      "learning_rate": 0.000165727616867682,
      "loss": 0.3225,
      "step": 5610
    },
    {
      "epoch": 1.42,
      "grad_norm": 3.7783835399126193,
      "learning_rate": 0.00016546350911171754,
      "loss": 0.265,
      "step": 5620
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.771413780307158,
      "learning_rate": 0.00016519940135575315,
      "loss": 0.3462,
      "step": 5630
    },
    {
      "epoch": 1.43,
      "grad_norm": 17.099197597109754,
      "learning_rate": 0.0001649352935997887,
      "loss": 0.3665,
      "step": 5640
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.598065804290718,
      "learning_rate": 0.00016467118584382427,
      "loss": 0.3195,
      "step": 5650
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.230585182301318,
      "learning_rate": 0.00016440707808785981,
      "loss": 0.6228,
      "step": 5660
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.557975044284719,
      "learning_rate": 0.00016414297033189542,
      "loss": 0.2,
      "step": 5670
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.237677686409225,
      "learning_rate": 0.00016387886257593096,
      "loss": 0.2912,
      "step": 5680
    },
    {
      "epoch": 1.44,
      "grad_norm": 3.0272821140167574,
      "learning_rate": 0.00016361475481996654,
      "loss": 0.3627,
      "step": 5690
    },
    {
      "epoch": 1.44,
      "grad_norm": 4.132070094762826,
      "learning_rate": 0.00016335064706400208,
      "loss": 0.2754,
      "step": 5700
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.735196686309314,
      "learning_rate": 0.00016308653930803768,
      "loss": 0.289,
      "step": 5710
    },
    {
      "epoch": 1.45,
      "grad_norm": 5.303597924748066,
      "learning_rate": 0.00016282243155207323,
      "loss": 0.4207,
      "step": 5720
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.900482527326059,
      "learning_rate": 0.0001625583237961088,
      "loss": 0.3014,
      "step": 5730
    },
    {
      "epoch": 1.45,
      "grad_norm": 3.510216989082974,
      "learning_rate": 0.00016229421604014435,
      "loss": 0.3637,
      "step": 5740
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.996542111519144,
      "learning_rate": 0.00016203010828417993,
      "loss": 0.355,
      "step": 5750
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.9960128892632762,
      "learning_rate": 0.0001617660005282155,
      "loss": 0.3615,
      "step": 5760
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.410429494498963,
      "learning_rate": 0.00016150189277225105,
      "loss": 0.4041,
      "step": 5770
    },
    {
      "epoch": 1.46,
      "grad_norm": 5.884751081527833,
      "learning_rate": 0.00016123778501628662,
      "loss": 0.2194,
      "step": 5780
    },
    {
      "epoch": 1.46,
      "grad_norm": 6.892881901273992,
      "learning_rate": 0.0001609736772603222,
      "loss": 0.2815,
      "step": 5790
    },
    {
      "epoch": 1.47,
      "grad_norm": 16.378051622703484,
      "learning_rate": 0.00016070956950435777,
      "loss": 0.44,
      "step": 5800
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.747758726063269,
      "learning_rate": 0.00016044546174839332,
      "loss": 0.3339,
      "step": 5810
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.4881745488949276,
      "learning_rate": 0.00016018135399242892,
      "loss": 0.2822,
      "step": 5820
    },
    {
      "epoch": 1.47,
      "grad_norm": 6.925114288101384,
      "learning_rate": 0.00015991724623646447,
      "loss": 0.1522,
      "step": 5830
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.426697298806934,
      "learning_rate": 0.00015965313848050004,
      "loss": 0.5291,
      "step": 5840
    },
    {
      "epoch": 1.48,
      "grad_norm": 3.0859117775460683,
      "learning_rate": 0.0001593890307245356,
      "loss": 0.179,
      "step": 5850
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.481580739953248,
      "learning_rate": 0.0001591249229685712,
      "loss": 0.2707,
      "step": 5860
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.399421336771333,
      "learning_rate": 0.00015886081521260674,
      "loss": 0.4472,
      "step": 5870
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.775123704587289,
      "learning_rate": 0.00015859670745664228,
      "loss": 0.2163,
      "step": 5880
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.003344532407036,
      "learning_rate": 0.00015833259970067786,
      "loss": 0.3938,
      "step": 5890
    },
    {
      "epoch": 1.49,
      "grad_norm": 4.063169363933348,
      "learning_rate": 0.0001580684919447134,
      "loss": 0.3104,
      "step": 5900
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.021103491372309,
      "learning_rate": 0.000157804384188749,
      "loss": 0.2433,
      "step": 5910
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.127815435528477,
      "learning_rate": 0.00015754027643278455,
      "loss": 0.4188,
      "step": 5920
    },
    {
      "epoch": 1.5,
      "grad_norm": 18.575206074912845,
      "learning_rate": 0.00015727616867682013,
      "loss": 0.3035,
      "step": 5930
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.6707897237666264,
      "learning_rate": 0.00015701206092085567,
      "loss": 0.2629,
      "step": 5940
    },
    {
      "epoch": 1.51,
      "grad_norm": 7.0249289338922525,
      "learning_rate": 0.00015674795316489128,
      "loss": 0.4149,
      "step": 5950
    },
    {
      "epoch": 1.51,
      "grad_norm": 24.315251449678733,
      "learning_rate": 0.00015648384540892682,
      "loss": 0.5126,
      "step": 5960
    },
    {
      "epoch": 1.51,
      "grad_norm": 6.3420314053766145,
      "learning_rate": 0.0001562197376529624,
      "loss": 0.2902,
      "step": 5970
    },
    {
      "epoch": 1.51,
      "grad_norm": 14.2218215768708,
      "learning_rate": 0.00015595562989699794,
      "loss": 0.4114,
      "step": 5980
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.800340107636826,
      "learning_rate": 0.00015569152214103355,
      "loss": 0.284,
      "step": 5990
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.7299155241996056,
      "learning_rate": 0.0001554274143850691,
      "loss": 0.3159,
      "step": 6000
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.518196707309734,
      "learning_rate": 0.00015516330662910464,
      "loss": 0.2959,
      "step": 6010
    },
    {
      "epoch": 1.52,
      "grad_norm": 2.332368023026213,
      "learning_rate": 0.00015489919887314024,
      "loss": 0.3974,
      "step": 6020
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.232975643160248,
      "learning_rate": 0.0001546350911171758,
      "loss": 0.2524,
      "step": 6030
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.74121783142979,
      "learning_rate": 0.00015437098336121136,
      "loss": 0.2416,
      "step": 6040
    },
    {
      "epoch": 1.53,
      "grad_norm": 7.504719616139981,
      "learning_rate": 0.0001541068756052469,
      "loss": 0.3549,
      "step": 6050
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.41647084870661,
      "learning_rate": 0.0001538427678492825,
      "loss": 0.471,
      "step": 6060
    },
    {
      "epoch": 1.54,
      "grad_norm": 16.158188481314436,
      "learning_rate": 0.00015357866009331806,
      "loss": 0.3437,
      "step": 6070
    },
    {
      "epoch": 1.54,
      "grad_norm": 16.32739961007586,
      "learning_rate": 0.00015331455233735363,
      "loss": 0.3165,
      "step": 6080
    },
    {
      "epoch": 1.54,
      "grad_norm": 5.184342314118148,
      "learning_rate": 0.00015305044458138918,
      "loss": 0.2971,
      "step": 6090
    },
    {
      "epoch": 1.54,
      "grad_norm": 7.207726323071327,
      "learning_rate": 0.00015278633682542478,
      "loss": 0.3692,
      "step": 6100
    },
    {
      "epoch": 1.55,
      "grad_norm": 23.297237421564507,
      "learning_rate": 0.00015252222906946033,
      "loss": 0.4152,
      "step": 6110
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.4114255435535004,
      "learning_rate": 0.0001522581213134959,
      "loss": 0.2974,
      "step": 6120
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.690999505219285,
      "learning_rate": 0.00015199401355753145,
      "loss": 0.2765,
      "step": 6130
    },
    {
      "epoch": 1.55,
      "grad_norm": 7.335840951773753,
      "learning_rate": 0.00015172990580156705,
      "loss": 0.3455,
      "step": 6140
    },
    {
      "epoch": 1.56,
      "grad_norm": 15.604891813878552,
      "learning_rate": 0.0001514657980456026,
      "loss": 0.4385,
      "step": 6150
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.628438285923433,
      "learning_rate": 0.00015120169028963814,
      "loss": 0.4003,
      "step": 6160
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.306314054776461,
      "learning_rate": 0.00015093758253367372,
      "loss": 0.424,
      "step": 6170
    },
    {
      "epoch": 1.56,
      "grad_norm": 6.260496743722033,
      "learning_rate": 0.0001506734747777093,
      "loss": 0.2472,
      "step": 6180
    },
    {
      "epoch": 1.57,
      "grad_norm": 7.080198000715199,
      "learning_rate": 0.00015040936702174487,
      "loss": 0.2117,
      "step": 6190
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.181287285553235,
      "learning_rate": 0.00015014525926578041,
      "loss": 0.2691,
      "step": 6200
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.118200463227126,
      "learning_rate": 0.000149881151509816,
      "loss": 0.3818,
      "step": 6210
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.655859369876493,
      "learning_rate": 0.00014961704375385156,
      "loss": 0.364,
      "step": 6220
    },
    {
      "epoch": 1.58,
      "grad_norm": 7.16565316656793,
      "learning_rate": 0.0001493529359978871,
      "loss": 0.3077,
      "step": 6230
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.303193350888552,
      "learning_rate": 0.00014908882824192268,
      "loss": 0.3348,
      "step": 6240
    },
    {
      "epoch": 1.58,
      "grad_norm": 3.3438569678761842,
      "learning_rate": 0.00014882472048595826,
      "loss": 0.2974,
      "step": 6250
    },
    {
      "epoch": 1.58,
      "grad_norm": 4.002626485525544,
      "learning_rate": 0.00014856061272999383,
      "loss": 0.3356,
      "step": 6260
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.4996572519866604,
      "learning_rate": 0.00014829650497402938,
      "loss": 0.3667,
      "step": 6270
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.0521944804779153,
      "learning_rate": 0.00014803239721806495,
      "loss": 0.3562,
      "step": 6280
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.402961094189349,
      "learning_rate": 0.00014776828946210053,
      "loss": 0.3747,
      "step": 6290
    },
    {
      "epoch": 1.59,
      "grad_norm": 4.300432959190631,
      "learning_rate": 0.0001475041817061361,
      "loss": 0.2226,
      "step": 6300
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.2533536238959868,
      "learning_rate": 0.00014724007395017165,
      "loss": 0.2996,
      "step": 6310
    },
    {
      "epoch": 1.6,
      "grad_norm": 17.023102286523546,
      "learning_rate": 0.00014697596619420722,
      "loss": 0.2937,
      "step": 6320
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.496957218839603,
      "learning_rate": 0.0001467118584382428,
      "loss": 0.2281,
      "step": 6330
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.9920070855130945,
      "learning_rate": 0.00014644775068227837,
      "loss": 0.2676,
      "step": 6340
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.9563652862881009,
      "learning_rate": 0.00014618364292631392,
      "loss": 0.2081,
      "step": 6350
    },
    {
      "epoch": 1.61,
      "grad_norm": 20.739397647977388,
      "learning_rate": 0.0001459195351703495,
      "loss": 0.4464,
      "step": 6360
    },
    {
      "epoch": 1.61,
      "grad_norm": 3.583675688967886,
      "learning_rate": 0.00014565542741438504,
      "loss": 0.3721,
      "step": 6370
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.510915177884517,
      "learning_rate": 0.00014539131965842061,
      "loss": 0.2997,
      "step": 6380
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.3261929823378318,
      "learning_rate": 0.0001451272119024562,
      "loss": 0.2083,
      "step": 6390
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.204559478601404,
      "learning_rate": 0.00014486310414649176,
      "loss": 0.4097,
      "step": 6400
    },
    {
      "epoch": 1.62,
      "grad_norm": 18.013955291169616,
      "learning_rate": 0.0001445989963905273,
      "loss": 0.3774,
      "step": 6410
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.610762645480038,
      "learning_rate": 0.00014433488863456288,
      "loss": 0.3484,
      "step": 6420
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.177813470182098,
      "learning_rate": 0.00014407078087859846,
      "loss": 0.3181,
      "step": 6430
    },
    {
      "epoch": 1.63,
      "grad_norm": 17.392831094348985,
      "learning_rate": 0.00014380667312263403,
      "loss": 0.4373,
      "step": 6440
    },
    {
      "epoch": 1.63,
      "grad_norm": 4.860262017846033,
      "learning_rate": 0.00014354256536666958,
      "loss": 0.4033,
      "step": 6450
    },
    {
      "epoch": 1.63,
      "grad_norm": 5.639817435634789,
      "learning_rate": 0.00014327845761070515,
      "loss": 0.3144,
      "step": 6460
    },
    {
      "epoch": 1.64,
      "grad_norm": 5.547212157278192,
      "learning_rate": 0.00014301434985474073,
      "loss": 0.3565,
      "step": 6470
    },
    {
      "epoch": 1.64,
      "grad_norm": 4.216430670026243,
      "learning_rate": 0.0001427502420987763,
      "loss": 0.2645,
      "step": 6480
    },
    {
      "epoch": 1.64,
      "grad_norm": 7.470166734375579,
      "learning_rate": 0.00014248613434281185,
      "loss": 0.2985,
      "step": 6490
    },
    {
      "epoch": 1.64,
      "grad_norm": 6.079270967627962,
      "learning_rate": 0.00014222202658684742,
      "loss": 0.3225,
      "step": 6500
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.82678101506829,
      "learning_rate": 0.00014195791883088297,
      "loss": 0.1106,
      "step": 6510
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.1436834838991876,
      "learning_rate": 0.00014169381107491854,
      "loss": 0.2378,
      "step": 6520
    },
    {
      "epoch": 1.65,
      "grad_norm": 5.032168769837695,
      "learning_rate": 0.00014142970331895412,
      "loss": 0.312,
      "step": 6530
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.845380454289769,
      "learning_rate": 0.0001411655955629897,
      "loss": 0.2321,
      "step": 6540
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.437202119786478,
      "learning_rate": 0.00014090148780702524,
      "loss": 0.3306,
      "step": 6550
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.556663714292608,
      "learning_rate": 0.00014063738005106081,
      "loss": 0.3701,
      "step": 6560
    },
    {
      "epoch": 1.66,
      "grad_norm": 5.9124287744903565,
      "learning_rate": 0.0001403732722950964,
      "loss": 0.2823,
      "step": 6570
    },
    {
      "epoch": 1.66,
      "grad_norm": 6.707735033607613,
      "learning_rate": 0.00014010916453913196,
      "loss": 0.3947,
      "step": 6580
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.9719620881655064,
      "learning_rate": 0.00013984505678316754,
      "loss": 0.2768,
      "step": 6590
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.072632760821824,
      "learning_rate": 0.00013958094902720308,
      "loss": 0.5599,
      "step": 6600
    },
    {
      "epoch": 1.67,
      "grad_norm": 6.358318051518618,
      "learning_rate": 0.00013931684127123866,
      "loss": 0.3604,
      "step": 6610
    },
    {
      "epoch": 1.67,
      "grad_norm": 4.577029869748708,
      "learning_rate": 0.0001390527335152742,
      "loss": 0.3292,
      "step": 6620
    },
    {
      "epoch": 1.68,
      "grad_norm": 4.421913815301032,
      "learning_rate": 0.00013878862575930978,
      "loss": 0.2617,
      "step": 6630
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.246052725873391,
      "learning_rate": 0.00013852451800334535,
      "loss": 0.4331,
      "step": 6640
    },
    {
      "epoch": 1.68,
      "grad_norm": 16.779819219746976,
      "learning_rate": 0.00013826041024738093,
      "loss": 0.2813,
      "step": 6650
    },
    {
      "epoch": 1.68,
      "grad_norm": 13.472895007673747,
      "learning_rate": 0.00013799630249141648,
      "loss": 0.3333,
      "step": 6660
    },
    {
      "epoch": 1.69,
      "grad_norm": 4.090768431286489,
      "learning_rate": 0.00013773219473545205,
      "loss": 0.3463,
      "step": 6670
    },
    {
      "epoch": 1.69,
      "grad_norm": 6.296496050367258,
      "learning_rate": 0.00013746808697948762,
      "loss": 0.3922,
      "step": 6680
    },
    {
      "epoch": 1.69,
      "grad_norm": 7.573282179736087,
      "learning_rate": 0.0001372039792235232,
      "loss": 0.5174,
      "step": 6690
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.476190995295713,
      "learning_rate": 0.00013693987146755875,
      "loss": 0.3445,
      "step": 6700
    },
    {
      "epoch": 1.7,
      "grad_norm": 7.503897944936768,
      "learning_rate": 0.00013667576371159432,
      "loss": 0.2797,
      "step": 6710
    },
    {
      "epoch": 1.7,
      "grad_norm": 4.384492334204901,
      "learning_rate": 0.0001364116559556299,
      "loss": 0.2803,
      "step": 6720
    },
    {
      "epoch": 1.7,
      "grad_norm": 4.550543480911678,
      "learning_rate": 0.00013614754819966547,
      "loss": 0.3624,
      "step": 6730
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.551640844627658,
      "learning_rate": 0.00013588344044370101,
      "loss": 0.2134,
      "step": 6740
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.935569639750181,
      "learning_rate": 0.0001356193326877366,
      "loss": 0.3754,
      "step": 6750
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.051912650212629,
      "learning_rate": 0.00013535522493177214,
      "loss": 0.2394,
      "step": 6760
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.5676283972219329,
      "learning_rate": 0.0001350911171758077,
      "loss": 0.1677,
      "step": 6770
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.399054721547771,
      "learning_rate": 0.00013482700941984328,
      "loss": 0.2642,
      "step": 6780
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.742555406471177,
      "learning_rate": 0.00013456290166387886,
      "loss": 0.3408,
      "step": 6790
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.128611034295828,
      "learning_rate": 0.0001342987939079144,
      "loss": 0.3314,
      "step": 6800
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.3202295788243634,
      "learning_rate": 0.00013403468615194998,
      "loss": 0.3683,
      "step": 6810
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.281040697781666,
      "learning_rate": 0.00013377057839598555,
      "loss": 0.3436,
      "step": 6820
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.179128585007064,
      "learning_rate": 0.00013350647064002113,
      "loss": 0.2195,
      "step": 6830
    },
    {
      "epoch": 1.73,
      "grad_norm": 13.569603018038574,
      "learning_rate": 0.00013324236288405668,
      "loss": 0.399,
      "step": 6840
    },
    {
      "epoch": 1.73,
      "grad_norm": 14.599350326434276,
      "learning_rate": 0.00013297825512809225,
      "loss": 0.3329,
      "step": 6850
    },
    {
      "epoch": 1.74,
      "grad_norm": 4.986639468091575,
      "learning_rate": 0.00013271414737212782,
      "loss": 0.203,
      "step": 6860
    },
    {
      "epoch": 1.74,
      "grad_norm": 14.03862023758173,
      "learning_rate": 0.0001324500396161634,
      "loss": 0.2006,
      "step": 6870
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.793862534328231,
      "learning_rate": 0.00013218593186019895,
      "loss": 0.3644,
      "step": 6880
    },
    {
      "epoch": 1.74,
      "grad_norm": 4.283836370554853,
      "learning_rate": 0.00013192182410423452,
      "loss": 0.3263,
      "step": 6890
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.4170345327733065,
      "learning_rate": 0.00013165771634827007,
      "loss": 0.3041,
      "step": 6900
    },
    {
      "epoch": 1.75,
      "grad_norm": 6.851470265914908,
      "learning_rate": 0.00013139360859230564,
      "loss": 0.3416,
      "step": 6910
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.770427591596483,
      "learning_rate": 0.00013112950083634121,
      "loss": 0.3061,
      "step": 6920
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.024461006147305,
      "learning_rate": 0.0001308653930803768,
      "loss": 0.293,
      "step": 6930
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.902704756256012,
      "learning_rate": 0.00013060128532441234,
      "loss": 0.3741,
      "step": 6940
    },
    {
      "epoch": 1.76,
      "grad_norm": 19.847226252906548,
      "learning_rate": 0.0001303371775684479,
      "loss": 0.4751,
      "step": 6950
    },
    {
      "epoch": 1.76,
      "grad_norm": 4.801241012685623,
      "learning_rate": 0.00013007306981248348,
      "loss": 0.3473,
      "step": 6960
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.293307817958716,
      "learning_rate": 0.00012980896205651906,
      "loss": 0.3203,
      "step": 6970
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.580777822195827,
      "learning_rate": 0.0001295448543005546,
      "loss": 0.4492,
      "step": 6980
    },
    {
      "epoch": 1.77,
      "grad_norm": 7.084767692974912,
      "learning_rate": 0.00012928074654459018,
      "loss": 0.2711,
      "step": 6990
    },
    {
      "epoch": 1.77,
      "grad_norm": 3.806799718919279,
      "learning_rate": 0.00012901663878862575,
      "loss": 0.249,
      "step": 7000
    },
    {
      "epoch": 1.77,
      "grad_norm": 3.9048368014810007,
      "learning_rate": 0.00012875253103266133,
      "loss": 0.3136,
      "step": 7010
    },
    {
      "epoch": 1.78,
      "grad_norm": 15.265715601719572,
      "learning_rate": 0.00012848842327669688,
      "loss": 0.4343,
      "step": 7020
    },
    {
      "epoch": 1.78,
      "grad_norm": 3.149093780791876,
      "learning_rate": 0.00012822431552073245,
      "loss": 0.2987,
      "step": 7030
    },
    {
      "epoch": 1.78,
      "grad_norm": 6.493197182696941,
      "learning_rate": 0.000127960207764768,
      "loss": 0.249,
      "step": 7040
    },
    {
      "epoch": 1.78,
      "grad_norm": 5.249477830603063,
      "learning_rate": 0.00012769610000880357,
      "loss": 0.2025,
      "step": 7050
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.7424442549499177,
      "learning_rate": 0.00012743199225283915,
      "loss": 0.1231,
      "step": 7060
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.217226055770746,
      "learning_rate": 0.00012716788449687472,
      "loss": 0.3337,
      "step": 7070
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.0621751429301365,
      "learning_rate": 0.00012690377674091027,
      "loss": 0.336,
      "step": 7080
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.086797386499311,
      "learning_rate": 0.00012663966898494584,
      "loss": 0.2337,
      "step": 7090
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.579111196044498,
      "learning_rate": 0.00012637556122898142,
      "loss": 0.3374,
      "step": 7100
    },
    {
      "epoch": 1.8,
      "grad_norm": 3.7540516028809328,
      "learning_rate": 0.000126111453473017,
      "loss": 0.1698,
      "step": 7110
    },
    {
      "epoch": 1.8,
      "grad_norm": 7.778067854467908,
      "learning_rate": 0.00012584734571705256,
      "loss": 0.3176,
      "step": 7120
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.28249883211765,
      "learning_rate": 0.0001255832379610881,
      "loss": 0.1805,
      "step": 7130
    },
    {
      "epoch": 1.81,
      "grad_norm": 5.472404912586062,
      "learning_rate": 0.00012531913020512368,
      "loss": 0.1059,
      "step": 7140
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.8872730254681165,
      "learning_rate": 0.00012505502244915926,
      "loss": 0.3318,
      "step": 7150
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.7794265745835744,
      "learning_rate": 0.0001247909146931948,
      "loss": 0.2772,
      "step": 7160
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.5333644033252645,
      "learning_rate": 0.00012452680693723038,
      "loss": 0.4262,
      "step": 7170
    },
    {
      "epoch": 1.82,
      "grad_norm": 16.826885464636717,
      "learning_rate": 0.00012426269918126593,
      "loss": 0.2253,
      "step": 7180
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.622835603930756,
      "learning_rate": 0.0001239985914253015,
      "loss": 0.2077,
      "step": 7190
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.764683064619582,
      "learning_rate": 0.00012373448366933708,
      "loss": 0.3954,
      "step": 7200
    },
    {
      "epoch": 1.82,
      "grad_norm": 6.870091906882638,
      "learning_rate": 0.00012347037591337265,
      "loss": 0.2326,
      "step": 7210
    },
    {
      "epoch": 1.83,
      "grad_norm": 3.118790740070018,
      "learning_rate": 0.00012320626815740822,
      "loss": 0.2159,
      "step": 7220
    },
    {
      "epoch": 1.83,
      "grad_norm": 6.244474003280159,
      "learning_rate": 0.00012294216040144377,
      "loss": 0.4341,
      "step": 7230
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.653487270102085,
      "learning_rate": 0.00012267805264547935,
      "loss": 0.5084,
      "step": 7240
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.14669567974351,
      "learning_rate": 0.00012241394488951492,
      "loss": 0.4201,
      "step": 7250
    },
    {
      "epoch": 1.84,
      "grad_norm": 4.792042813792595,
      "learning_rate": 0.0001221498371335505,
      "loss": 0.2149,
      "step": 7260
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.51542484693792,
      "learning_rate": 0.00012188572937758605,
      "loss": 0.3921,
      "step": 7270
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.762497244565239,
      "learning_rate": 0.00012162162162162162,
      "loss": 0.2177,
      "step": 7280
    },
    {
      "epoch": 1.84,
      "grad_norm": 5.475664304577624,
      "learning_rate": 0.00012135751386565719,
      "loss": 0.304,
      "step": 7290
    },
    {
      "epoch": 1.85,
      "grad_norm": 16.69877452415643,
      "learning_rate": 0.00012109340610969274,
      "loss": 0.372,
      "step": 7300
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.221634614551201,
      "learning_rate": 0.00012082929835372831,
      "loss": 0.2879,
      "step": 7310
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.172009412255388,
      "learning_rate": 0.00012056519059776387,
      "loss": 0.2582,
      "step": 7320
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.509769033707379,
      "learning_rate": 0.00012030108284179945,
      "loss": 0.2183,
      "step": 7330
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.321430562793568,
      "learning_rate": 0.000120036975085835,
      "loss": 0.3057,
      "step": 7340
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.0566355491459625,
      "learning_rate": 0.00011977286732987058,
      "loss": 0.3616,
      "step": 7350
    },
    {
      "epoch": 1.86,
      "grad_norm": 6.649560281056566,
      "learning_rate": 0.00011950875957390614,
      "loss": 0.2277,
      "step": 7360
    },
    {
      "epoch": 1.86,
      "grad_norm": 3.823681392264618,
      "learning_rate": 0.00011924465181794172,
      "loss": 0.2506,
      "step": 7370
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.977585296138626,
      "learning_rate": 0.00011898054406197728,
      "loss": 0.2668,
      "step": 7380
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.18817069677191,
      "learning_rate": 0.00011871643630601285,
      "loss": 0.3338,
      "step": 7390
    },
    {
      "epoch": 1.87,
      "grad_norm": 5.13086843253833,
      "learning_rate": 0.00011845232855004841,
      "loss": 0.1987,
      "step": 7400
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.47912661312829385,
      "learning_rate": 0.00011818822079408399,
      "loss": 0.1843,
      "step": 7410
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.810984382614806,
      "learning_rate": 0.00011792411303811955,
      "loss": 0.2258,
      "step": 7420
    },
    {
      "epoch": 1.88,
      "grad_norm": 20.186757822536684,
      "learning_rate": 0.0001176600052821551,
      "loss": 0.3983,
      "step": 7430
    },
    {
      "epoch": 1.88,
      "grad_norm": 4.6582476219915,
      "learning_rate": 0.00011739589752619067,
      "loss": 0.3271,
      "step": 7440
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.3936916683319542,
      "learning_rate": 0.00011713178977022624,
      "loss": 0.3525,
      "step": 7450
    },
    {
      "epoch": 1.89,
      "grad_norm": 3.0701950430849614,
      "learning_rate": 0.0001168676820142618,
      "loss": 0.2089,
      "step": 7460
    },
    {
      "epoch": 1.89,
      "grad_norm": 18.441688716432523,
      "learning_rate": 0.00011660357425829738,
      "loss": 0.3196,
      "step": 7470
    },
    {
      "epoch": 1.89,
      "grad_norm": 6.3186094662020835,
      "learning_rate": 0.00011633946650233294,
      "loss": 0.3754,
      "step": 7480
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.6445611437174326,
      "learning_rate": 0.00011607535874636851,
      "loss": 0.2543,
      "step": 7490
    },
    {
      "epoch": 1.9,
      "grad_norm": 19.763658077394638,
      "learning_rate": 0.00011581125099040407,
      "loss": 0.2198,
      "step": 7500
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.959144520673648,
      "learning_rate": 0.00011554714323443965,
      "loss": 0.2783,
      "step": 7510
    },
    {
      "epoch": 1.9,
      "grad_norm": 7.784092477796405,
      "learning_rate": 0.0001152830354784752,
      "loss": 0.3151,
      "step": 7520
    },
    {
      "epoch": 1.9,
      "grad_norm": 5.998190142436887,
      "learning_rate": 0.00011501892772251078,
      "loss": 0.1617,
      "step": 7530
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.252694593147776,
      "learning_rate": 0.00011475481996654634,
      "loss": 0.2864,
      "step": 7540
    },
    {
      "epoch": 1.91,
      "grad_norm": 12.244038586138855,
      "learning_rate": 0.00011449071221058192,
      "loss": 0.1699,
      "step": 7550
    },
    {
      "epoch": 1.91,
      "grad_norm": 17.43389471787065,
      "learning_rate": 0.00011422660445461748,
      "loss": 0.2885,
      "step": 7560
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.8670922708223348,
      "learning_rate": 0.00011396249669865304,
      "loss": 0.2298,
      "step": 7570
    },
    {
      "epoch": 1.92,
      "grad_norm": 6.967344925020078,
      "learning_rate": 0.0001136983889426886,
      "loss": 0.2609,
      "step": 7580
    },
    {
      "epoch": 1.92,
      "grad_norm": 15.255839592868696,
      "learning_rate": 0.00011346069196232061,
      "loss": 0.3349,
      "step": 7590
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.997459901332054,
      "learning_rate": 0.00011319658420635618,
      "loss": 0.274,
      "step": 7600
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.936816842767314,
      "learning_rate": 0.00011293247645039175,
      "loss": 0.2767,
      "step": 7610
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.722905002127161,
      "learning_rate": 0.00011266836869442731,
      "loss": 0.2499,
      "step": 7620
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.999166638602913,
      "learning_rate": 0.00011240426093846288,
      "loss": 0.2229,
      "step": 7630
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.8032525020782211,
      "learning_rate": 0.00011214015318249844,
      "loss": 0.1764,
      "step": 7640
    },
    {
      "epoch": 1.94,
      "grad_norm": 27.547154782714138,
      "learning_rate": 0.00011187604542653402,
      "loss": 0.2496,
      "step": 7650
    },
    {
      "epoch": 1.94,
      "grad_norm": 20.106686389398114,
      "learning_rate": 0.00011161193767056958,
      "loss": 0.2334,
      "step": 7660
    },
    {
      "epoch": 1.94,
      "grad_norm": 5.083256278071224,
      "learning_rate": 0.00011134782991460515,
      "loss": 0.3343,
      "step": 7670
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.199029959820093,
      "learning_rate": 0.00011108372215864071,
      "loss": 0.3351,
      "step": 7680
    },
    {
      "epoch": 1.95,
      "grad_norm": 20.144031653830446,
      "learning_rate": 0.00011081961440267629,
      "loss": 0.1828,
      "step": 7690
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.0893760301832116,
      "learning_rate": 0.00011055550664671185,
      "loss": 0.3685,
      "step": 7700
    },
    {
      "epoch": 1.95,
      "grad_norm": 4.585526910204328,
      "learning_rate": 0.00011029139889074742,
      "loss": 0.2942,
      "step": 7710
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.846752410262999,
      "learning_rate": 0.00011002729113478297,
      "loss": 0.3366,
      "step": 7720
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.2964170047983092,
      "learning_rate": 0.00010976318337881854,
      "loss": 0.2765,
      "step": 7730
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.119994789567933,
      "learning_rate": 0.0001094990756228541,
      "loss": 0.3336,
      "step": 7740
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.143299496577857,
      "learning_rate": 0.00010923496786688968,
      "loss": 0.2018,
      "step": 7750
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.03543638740937,
      "learning_rate": 0.00010897086011092524,
      "loss": 0.3145,
      "step": 7760
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.465157864032654,
      "learning_rate": 0.00010870675235496081,
      "loss": 0.2603,
      "step": 7770
    },
    {
      "epoch": 1.97,
      "grad_norm": 4.532296356702725,
      "learning_rate": 0.00010844264459899638,
      "loss": 0.5217,
      "step": 7780
    },
    {
      "epoch": 1.97,
      "grad_norm": 5.290637566404521,
      "learning_rate": 0.00010817853684303195,
      "loss": 0.3109,
      "step": 7790
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.579357609290867,
      "learning_rate": 0.00010791442908706751,
      "loss": 0.3611,
      "step": 7800
    },
    {
      "epoch": 1.98,
      "grad_norm": 6.867193048451062,
      "learning_rate": 0.00010765032133110308,
      "loss": 0.4212,
      "step": 7810
    },
    {
      "epoch": 1.98,
      "grad_norm": 5.3338729699400655,
      "learning_rate": 0.00010738621357513866,
      "loss": 0.222,
      "step": 7820
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.936161017952145,
      "learning_rate": 0.00010712210581917422,
      "loss": 0.3205,
      "step": 7830
    },
    {
      "epoch": 1.98,
      "grad_norm": 20.529054885815693,
      "learning_rate": 0.0001068579980632098,
      "loss": 0.3484,
      "step": 7840
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.440117544284218,
      "learning_rate": 0.00010659389030724534,
      "loss": 0.2661,
      "step": 7850
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.764316456522104,
      "learning_rate": 0.0001063297825512809,
      "loss": 0.1981,
      "step": 7860
    },
    {
      "epoch": 1.99,
      "grad_norm": 5.807864772560368,
      "learning_rate": 0.00010606567479531648,
      "loss": 0.3118,
      "step": 7870
    },
    {
      "epoch": 1.99,
      "grad_norm": 14.28570131675949,
      "learning_rate": 0.00010580156703935204,
      "loss": 0.1987,
      "step": 7880
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.070614046363564,
      "learning_rate": 0.00010553745928338761,
      "loss": 0.1938,
      "step": 7890
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.519446863351419,
      "learning_rate": 0.00010527335152742317,
      "loss": 0.3154,
      "step": 7900
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8391712624832315,
      "eval_f1": 0.8067493314121663,
      "eval_loss": 0.357177734375,
      "eval_precision": 0.7895852602851988,
      "eval_recall": 0.841805198689739,
      "eval_runtime": 464.0655,
      "eval_samples_per_second": 14.457,
      "eval_steps_per_second": 2.411,
      "step": 7906
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.570371337454835,
      "learning_rate": 0.00010500924377145875,
      "loss": 0.3369,
      "step": 7910
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.855296400806096,
      "learning_rate": 0.00010474513601549432,
      "loss": 0.2167,
      "step": 7920
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.8449716786359939,
      "learning_rate": 0.00010448102825952988,
      "loss": 0.0707,
      "step": 7930
    },
    {
      "epoch": 2.01,
      "grad_norm": 39.85705463948524,
      "learning_rate": 0.00010421692050356545,
      "loss": 0.1918,
      "step": 7940
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.2797592787806,
      "learning_rate": 0.00010395281274760101,
      "loss": 0.0892,
      "step": 7950
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.9610650197953539,
      "learning_rate": 0.00010368870499163659,
      "loss": 0.0893,
      "step": 7960
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.9959374584504933,
      "learning_rate": 0.00010342459723567215,
      "loss": 0.1205,
      "step": 7970
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.4738182783177045,
      "learning_rate": 0.00010316048947970772,
      "loss": 0.1759,
      "step": 7980
    },
    {
      "epoch": 2.02,
      "grad_norm": 3.1119985220385673,
      "learning_rate": 0.00010289638172374327,
      "loss": 0.1134,
      "step": 7990
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.1231176982192115,
      "learning_rate": 0.00010263227396777885,
      "loss": 0.1927,
      "step": 8000
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.1678565739543172,
      "learning_rate": 0.0001023681662118144,
      "loss": 0.2502,
      "step": 8010
    },
    {
      "epoch": 2.03,
      "grad_norm": 9.265883828901494,
      "learning_rate": 0.00010210405845584998,
      "loss": 0.2104,
      "step": 8020
    },
    {
      "epoch": 2.03,
      "grad_norm": 7.243874798770191,
      "learning_rate": 0.00010183995069988554,
      "loss": 0.1726,
      "step": 8030
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.40591160969289597,
      "learning_rate": 0.00010157584294392111,
      "loss": 0.0678,
      "step": 8040
    },
    {
      "epoch": 2.04,
      "grad_norm": 3.927362068225369,
      "learning_rate": 0.00010131173518795668,
      "loss": 0.0875,
      "step": 8050
    },
    {
      "epoch": 2.04,
      "grad_norm": 5.009131341604889,
      "learning_rate": 0.00010104762743199225,
      "loss": 0.227,
      "step": 8060
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.7148744429646117,
      "learning_rate": 0.00010078351967602781,
      "loss": 0.1711,
      "step": 8070
    },
    {
      "epoch": 2.04,
      "grad_norm": 3.20496942985561,
      "learning_rate": 0.00010051941192006338,
      "loss": 0.1087,
      "step": 8080
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.9231245687182146,
      "learning_rate": 0.00010025530416409895,
      "loss": 0.1996,
      "step": 8090
    },
    {
      "epoch": 2.05,
      "grad_norm": 22.72875318987384,
      "learning_rate": 9.999119640813452e-05,
      "loss": 0.1851,
      "step": 8100
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.00881260240084,
      "learning_rate": 9.972708865217008e-05,
      "loss": 0.0927,
      "step": 8110
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.209720716655473,
      "learning_rate": 9.946298089620564e-05,
      "loss": 0.0849,
      "step": 8120
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.9529919774532654,
      "learning_rate": 9.91988731402412e-05,
      "loss": 0.0215,
      "step": 8130
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.107865291879719,
      "learning_rate": 9.893476538427678e-05,
      "loss": 0.2368,
      "step": 8140
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.268600282889506,
      "learning_rate": 9.867065762831234e-05,
      "loss": 0.037,
      "step": 8150
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.01050701287250132,
      "learning_rate": 9.840654987234791e-05,
      "loss": 0.112,
      "step": 8160
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.4766594187021542,
      "learning_rate": 9.814244211638347e-05,
      "loss": 0.1619,
      "step": 8170
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.977577908902681,
      "learning_rate": 9.787833436041905e-05,
      "loss": 0.0851,
      "step": 8180
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.716767884791129,
      "learning_rate": 9.76142266044546e-05,
      "loss": 0.128,
      "step": 8190
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.8920174800911701,
      "learning_rate": 9.735011884849018e-05,
      "loss": 0.1015,
      "step": 8200
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.3498389647596418,
      "learning_rate": 9.708601109252574e-05,
      "loss": 0.1739,
      "step": 8210
    },
    {
      "epoch": 2.08,
      "grad_norm": 15.60644366259143,
      "learning_rate": 9.682190333656132e-05,
      "loss": 0.2107,
      "step": 8220
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.8701698283820544,
      "learning_rate": 9.655779558059688e-05,
      "loss": 0.1222,
      "step": 8230
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.9143304968078294,
      "learning_rate": 9.629368782463245e-05,
      "loss": 0.061,
      "step": 8240
    },
    {
      "epoch": 2.09,
      "grad_norm": 3.6884628782455313,
      "learning_rate": 9.602958006866801e-05,
      "loss": 0.1234,
      "step": 8250
    },
    {
      "epoch": 2.09,
      "grad_norm": 15.137930438366073,
      "learning_rate": 9.576547231270357e-05,
      "loss": 0.1065,
      "step": 8260
    },
    {
      "epoch": 2.09,
      "grad_norm": 20.497763755991276,
      "learning_rate": 9.550136455673913e-05,
      "loss": 0.1955,
      "step": 8270
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.6297291192661263,
      "learning_rate": 9.52372568007747e-05,
      "loss": 0.1042,
      "step": 8280
    },
    {
      "epoch": 2.1,
      "grad_norm": 7.5522736122013265,
      "learning_rate": 9.497314904481027e-05,
      "loss": 0.1371,
      "step": 8290
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.1954233455079364,
      "learning_rate": 9.470904128884584e-05,
      "loss": 0.1168,
      "step": 8300
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.8568853988431717,
      "learning_rate": 9.44449335328814e-05,
      "loss": 0.0678,
      "step": 8310
    },
    {
      "epoch": 2.1,
      "grad_norm": 8.058061442543563,
      "learning_rate": 9.418082577691698e-05,
      "loss": 0.0824,
      "step": 8320
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.20800373044439,
      "learning_rate": 9.391671802095254e-05,
      "loss": 0.0912,
      "step": 8330
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.0416560806313,
      "learning_rate": 9.365261026498811e-05,
      "loss": 0.1513,
      "step": 8340
    },
    {
      "epoch": 2.11,
      "grad_norm": 27.525859053820138,
      "learning_rate": 9.338850250902367e-05,
      "loss": 0.1325,
      "step": 8350
    },
    {
      "epoch": 2.11,
      "grad_norm": 3.3353720644746128,
      "learning_rate": 9.312439475305925e-05,
      "loss": 0.1358,
      "step": 8360
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.1884532660672202,
      "learning_rate": 9.28602869970948e-05,
      "loss": 0.1026,
      "step": 8370
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.13764298399856836,
      "learning_rate": 9.259617924113038e-05,
      "loss": 0.1127,
      "step": 8380
    },
    {
      "epoch": 2.12,
      "grad_norm": 8.022496581506283,
      "learning_rate": 9.233207148516593e-05,
      "loss": 0.1256,
      "step": 8390
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.21143244051880208,
      "learning_rate": 9.20679637292015e-05,
      "loss": 0.139,
      "step": 8400
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.6104497392849312,
      "learning_rate": 9.180385597323706e-05,
      "loss": 0.1127,
      "step": 8410
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.3076077939010853,
      "learning_rate": 9.153974821727264e-05,
      "loss": 0.0933,
      "step": 8420
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.8183509445857418,
      "learning_rate": 9.12756404613082e-05,
      "loss": 0.1079,
      "step": 8430
    },
    {
      "epoch": 2.14,
      "grad_norm": 4.096549653818764,
      "learning_rate": 9.101153270534377e-05,
      "loss": 0.07,
      "step": 8440
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.7215206673289605,
      "learning_rate": 9.074742494937933e-05,
      "loss": 0.309,
      "step": 8450
    },
    {
      "epoch": 2.14,
      "grad_norm": 15.718577759407667,
      "learning_rate": 9.04833171934149e-05,
      "loss": 0.2524,
      "step": 8460
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.9256241045762397,
      "learning_rate": 9.021920943745048e-05,
      "loss": 0.1415,
      "step": 8470
    },
    {
      "epoch": 2.15,
      "grad_norm": 16.08379613121137,
      "learning_rate": 8.995510168148604e-05,
      "loss": 0.3066,
      "step": 8480
    },
    {
      "epoch": 2.15,
      "grad_norm": 6.760712726710112,
      "learning_rate": 8.969099392552162e-05,
      "loss": 0.1154,
      "step": 8490
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.3985228336827005,
      "learning_rate": 8.942688616955718e-05,
      "loss": 0.0986,
      "step": 8500
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.1802485482692591,
      "learning_rate": 8.916277841359275e-05,
      "loss": 0.0779,
      "step": 8510
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.6563191930790624,
      "learning_rate": 8.889867065762831e-05,
      "loss": 0.1167,
      "step": 8520
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.6652412854559228,
      "learning_rate": 8.863456290166386e-05,
      "loss": 0.0646,
      "step": 8530
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.431258079582006,
      "learning_rate": 8.837045514569943e-05,
      "loss": 0.1383,
      "step": 8540
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.24832586396252196,
      "learning_rate": 8.810634738973499e-05,
      "loss": 0.0864,
      "step": 8550
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.1699998051681302,
      "learning_rate": 8.784223963377057e-05,
      "loss": 0.2314,
      "step": 8560
    },
    {
      "epoch": 2.17,
      "grad_norm": 11.474808327106688,
      "learning_rate": 8.760454265340257e-05,
      "loss": 0.2557,
      "step": 8570
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.9295228929342434,
      "learning_rate": 8.734043489743814e-05,
      "loss": 0.1622,
      "step": 8580
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.3840033482411934,
      "learning_rate": 8.70763271414737e-05,
      "loss": 0.1265,
      "step": 8590
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.0243901637701598,
      "learning_rate": 8.681221938550928e-05,
      "loss": 0.0309,
      "step": 8600
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.413148766104575,
      "learning_rate": 8.654811162954485e-05,
      "loss": 0.0593,
      "step": 8610
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.06719997457490859,
      "learning_rate": 8.628400387358041e-05,
      "loss": 0.0646,
      "step": 8620
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.8835322138559412,
      "learning_rate": 8.601989611761599e-05,
      "loss": 0.1953,
      "step": 8630
    },
    {
      "epoch": 2.19,
      "grad_norm": 15.041935138560776,
      "learning_rate": 8.575578836165155e-05,
      "loss": 0.1152,
      "step": 8640
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.1218005591033131,
      "learning_rate": 8.549168060568712e-05,
      "loss": 0.1248,
      "step": 8650
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.3773948299156011,
      "learning_rate": 8.522757284972268e-05,
      "loss": 0.2706,
      "step": 8660
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.5557269416973856,
      "learning_rate": 8.496346509375826e-05,
      "loss": 0.1381,
      "step": 8670
    },
    {
      "epoch": 2.2,
      "grad_norm": 7.345987263229923,
      "learning_rate": 8.46993573377938e-05,
      "loss": 0.0965,
      "step": 8680
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.3490779839965095,
      "learning_rate": 8.443524958182938e-05,
      "loss": 0.1977,
      "step": 8690
    },
    {
      "epoch": 2.2,
      "grad_norm": 29.747030469606557,
      "learning_rate": 8.417114182586494e-05,
      "loss": 0.2724,
      "step": 8700
    },
    {
      "epoch": 2.2,
      "grad_norm": 12.08882292534392,
      "learning_rate": 8.390703406990051e-05,
      "loss": 0.2033,
      "step": 8710
    },
    {
      "epoch": 2.21,
      "grad_norm": 5.193636589499085,
      "learning_rate": 8.364292631393607e-05,
      "loss": 0.1297,
      "step": 8720
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.3725218267089253,
      "learning_rate": 8.337881855797165e-05,
      "loss": 0.0448,
      "step": 8730
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.8560258938012986,
      "learning_rate": 8.311471080200721e-05,
      "loss": 0.0929,
      "step": 8740
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.05501975523877587,
      "learning_rate": 8.285060304604278e-05,
      "loss": 0.0303,
      "step": 8750
    },
    {
      "epoch": 2.22,
      "grad_norm": 16.01099719850488,
      "learning_rate": 8.258649529007834e-05,
      "loss": 0.1764,
      "step": 8760
    },
    {
      "epoch": 2.22,
      "grad_norm": 22.398232395548956,
      "learning_rate": 8.232238753411392e-05,
      "loss": 0.3596,
      "step": 8770
    },
    {
      "epoch": 2.22,
      "grad_norm": 8.781541772827923,
      "learning_rate": 8.205827977814948e-05,
      "loss": 0.3248,
      "step": 8780
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.2963466234719823,
      "learning_rate": 8.179417202218505e-05,
      "loss": 0.1504,
      "step": 8790
    },
    {
      "epoch": 2.23,
      "grad_norm": 7.267265645265003,
      "learning_rate": 8.153006426622061e-05,
      "loss": 0.1733,
      "step": 8800
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.1250143324457247,
      "learning_rate": 8.126595651025617e-05,
      "loss": 0.1134,
      "step": 8810
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.24929393121602,
      "learning_rate": 8.100184875429174e-05,
      "loss": 0.1758,
      "step": 8820
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.6685242506781202,
      "learning_rate": 8.073774099832731e-05,
      "loss": 0.0951,
      "step": 8830
    },
    {
      "epoch": 2.24,
      "grad_norm": 13.17646145414902,
      "learning_rate": 8.047363324236287e-05,
      "loss": 0.2249,
      "step": 8840
    },
    {
      "epoch": 2.24,
      "grad_norm": 10.766085447366857,
      "learning_rate": 8.020952548639844e-05,
      "loss": 0.0767,
      "step": 8850
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.437347110388571,
      "learning_rate": 7.9945417730434e-05,
      "loss": 0.058,
      "step": 8860
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.1214332634621527,
      "learning_rate": 7.968130997446958e-05,
      "loss": 0.2229,
      "step": 8870
    },
    {
      "epoch": 2.25,
      "grad_norm": 5.876581897211659,
      "learning_rate": 7.941720221850514e-05,
      "loss": 0.193,
      "step": 8880
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.501289763988718,
      "learning_rate": 7.915309446254071e-05,
      "loss": 0.1414,
      "step": 8890
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.6224887226366147,
      "learning_rate": 7.888898670657628e-05,
      "loss": 0.1095,
      "step": 8900
    },
    {
      "epoch": 2.25,
      "grad_norm": 8.347439493123973,
      "learning_rate": 7.862487895061185e-05,
      "loss": 0.1348,
      "step": 8910
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.4864953010659459,
      "learning_rate": 7.836077119464741e-05,
      "loss": 0.05,
      "step": 8920
    },
    {
      "epoch": 2.26,
      "grad_norm": 5.550479008890894,
      "learning_rate": 7.809666343868298e-05,
      "loss": 0.2035,
      "step": 8930
    },
    {
      "epoch": 2.26,
      "grad_norm": 5.654461158579552,
      "learning_rate": 7.783255568271854e-05,
      "loss": 0.1046,
      "step": 8940
    },
    {
      "epoch": 2.26,
      "grad_norm": 2.4307183255414384,
      "learning_rate": 7.75684479267541e-05,
      "loss": 0.1701,
      "step": 8950
    },
    {
      "epoch": 2.27,
      "grad_norm": 13.608080017284609,
      "learning_rate": 7.730434017078967e-05,
      "loss": 0.272,
      "step": 8960
    },
    {
      "epoch": 2.27,
      "grad_norm": 3.678000896802301,
      "learning_rate": 7.704023241482524e-05,
      "loss": 0.2433,
      "step": 8970
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.370415506731004,
      "learning_rate": 7.67761246588608e-05,
      "loss": 0.2123,
      "step": 8980
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.9083728841668193,
      "learning_rate": 7.651201690289638e-05,
      "loss": 0.2176,
      "step": 8990
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.324331069283942,
      "learning_rate": 7.624790914693194e-05,
      "loss": 0.0449,
      "step": 9000
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.3540280241688483,
      "learning_rate": 7.598380139096751e-05,
      "loss": 0.0776,
      "step": 9010
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.8081383403403553,
      "learning_rate": 7.571969363500307e-05,
      "loss": 0.2757,
      "step": 9020
    },
    {
      "epoch": 2.28,
      "grad_norm": 6.340708587827813,
      "learning_rate": 7.545558587903864e-05,
      "loss": 0.1071,
      "step": 9030
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.4771124752811318,
      "learning_rate": 7.51914781230742e-05,
      "loss": 0.2146,
      "step": 9040
    },
    {
      "epoch": 2.29,
      "grad_norm": 12.328840457379812,
      "learning_rate": 7.492737036710977e-05,
      "loss": 0.1453,
      "step": 9050
    },
    {
      "epoch": 2.29,
      "grad_norm": 7.86559890554053,
      "learning_rate": 7.466326261114534e-05,
      "loss": 0.1442,
      "step": 9060
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.183074026280236,
      "learning_rate": 7.43991548551809e-05,
      "loss": 0.0975,
      "step": 9070
    },
    {
      "epoch": 2.3,
      "grad_norm": 7.588202181180001,
      "learning_rate": 7.413504709921648e-05,
      "loss": 0.1359,
      "step": 9080
    },
    {
      "epoch": 2.3,
      "grad_norm": 11.591111775618687,
      "learning_rate": 7.387093934325204e-05,
      "loss": 0.2605,
      "step": 9090
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.4037183503782767,
      "learning_rate": 7.360683158728761e-05,
      "loss": 0.0965,
      "step": 9100
    },
    {
      "epoch": 2.3,
      "grad_norm": 7.767795756329233,
      "learning_rate": 7.334272383132317e-05,
      "loss": 0.2464,
      "step": 9110
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.8595308365752063,
      "learning_rate": 7.307861607535873e-05,
      "loss": 0.2457,
      "step": 9120
    },
    {
      "epoch": 2.31,
      "grad_norm": 18.84334961938705,
      "learning_rate": 7.28145083193943e-05,
      "loss": 0.1245,
      "step": 9130
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.9422947009381484,
      "learning_rate": 7.255040056342987e-05,
      "loss": 0.1193,
      "step": 9140
    },
    {
      "epoch": 2.31,
      "grad_norm": 21.866163000139704,
      "learning_rate": 7.228629280746544e-05,
      "loss": 0.1464,
      "step": 9150
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.5927815780630423,
      "learning_rate": 7.202218505150101e-05,
      "loss": 0.0346,
      "step": 9160
    },
    {
      "epoch": 2.32,
      "grad_norm": 5.703801860227664,
      "learning_rate": 7.175807729553658e-05,
      "loss": 0.0592,
      "step": 9170
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.36254942866638484,
      "learning_rate": 7.149396953957214e-05,
      "loss": 0.1765,
      "step": 9180
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.418908234330049,
      "learning_rate": 7.12298617836077e-05,
      "loss": 0.0874,
      "step": 9190
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.9506347199912738,
      "learning_rate": 7.096575402764327e-05,
      "loss": 0.065,
      "step": 9200
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.2935718801839875,
      "learning_rate": 7.070164627167885e-05,
      "loss": 0.1703,
      "step": 9210
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.1124870355542493,
      "learning_rate": 7.04375385157144e-05,
      "loss": 0.1708,
      "step": 9220
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.3688631854532768,
      "learning_rate": 7.017343075974998e-05,
      "loss": 0.1014,
      "step": 9230
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.8352408086564822,
      "learning_rate": 6.990932300378554e-05,
      "loss": 0.0566,
      "step": 9240
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.1542874366516385,
      "learning_rate": 6.96452152478211e-05,
      "loss": 0.248,
      "step": 9250
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.1544892653823733,
      "learning_rate": 6.938110749185668e-05,
      "loss": 0.2333,
      "step": 9260
    },
    {
      "epoch": 2.35,
      "grad_norm": 21.56280917070406,
      "learning_rate": 6.911699973589224e-05,
      "loss": 0.1926,
      "step": 9270
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.2520688419834509,
      "learning_rate": 6.885289197992781e-05,
      "loss": 0.0806,
      "step": 9280
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.4781905563970963,
      "learning_rate": 6.858878422396337e-05,
      "loss": 0.1101,
      "step": 9290
    },
    {
      "epoch": 2.35,
      "grad_norm": 8.616392641941344,
      "learning_rate": 6.832467646799895e-05,
      "loss": 0.0729,
      "step": 9300
    },
    {
      "epoch": 2.36,
      "grad_norm": 17.42189647348218,
      "learning_rate": 6.80605687120345e-05,
      "loss": 0.2795,
      "step": 9310
    },
    {
      "epoch": 2.36,
      "grad_norm": 8.297361946902777,
      "learning_rate": 6.779646095607007e-05,
      "loss": 0.162,
      "step": 9320
    },
    {
      "epoch": 2.36,
      "grad_norm": 5.375951555212969,
      "learning_rate": 6.753235320010564e-05,
      "loss": 0.1466,
      "step": 9330
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.540354092448967,
      "learning_rate": 6.72682454441412e-05,
      "loss": 0.1,
      "step": 9340
    },
    {
      "epoch": 2.37,
      "grad_norm": 7.226674986298907,
      "learning_rate": 6.700413768817678e-05,
      "loss": 0.0885,
      "step": 9350
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.662689789979408,
      "learning_rate": 6.674002993221234e-05,
      "loss": 0.3726,
      "step": 9360
    },
    {
      "epoch": 2.37,
      "grad_norm": 9.919696238986528,
      "learning_rate": 6.647592217624791e-05,
      "loss": 0.0673,
      "step": 9370
    },
    {
      "epoch": 2.37,
      "grad_norm": 19.61501206459386,
      "learning_rate": 6.621181442028347e-05,
      "loss": 0.1964,
      "step": 9380
    },
    {
      "epoch": 2.38,
      "grad_norm": 4.034768337943533,
      "learning_rate": 6.594770666431903e-05,
      "loss": 0.1425,
      "step": 9390
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.10466655515715,
      "learning_rate": 6.56835989083546e-05,
      "loss": 0.0877,
      "step": 9400
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.8377626611279765,
      "learning_rate": 6.541949115239017e-05,
      "loss": 0.2897,
      "step": 9410
    },
    {
      "epoch": 2.38,
      "grad_norm": 5.694959836933135,
      "learning_rate": 6.515538339642574e-05,
      "loss": 0.1375,
      "step": 9420
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.7072358352528014,
      "learning_rate": 6.48912756404613e-05,
      "loss": 0.1536,
      "step": 9430
    },
    {
      "epoch": 2.39,
      "grad_norm": 21.957339715071846,
      "learning_rate": 6.462716788449688e-05,
      "loss": 0.2325,
      "step": 9440
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.9973069500028089,
      "learning_rate": 6.436306012853244e-05,
      "loss": 0.0966,
      "step": 9450
    },
    {
      "epoch": 2.39,
      "grad_norm": 5.206448503424523,
      "learning_rate": 6.4098952372568e-05,
      "loss": 0.2192,
      "step": 9460
    },
    {
      "epoch": 2.4,
      "grad_norm": 5.719525784887322,
      "learning_rate": 6.383484461660357e-05,
      "loss": 0.1817,
      "step": 9470
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.553404969881463,
      "learning_rate": 6.357073686063913e-05,
      "loss": 0.1441,
      "step": 9480
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.103013880353259,
      "learning_rate": 6.33066291046747e-05,
      "loss": 0.0399,
      "step": 9490
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.6971411697797874,
      "learning_rate": 6.304252134871027e-05,
      "loss": 0.2328,
      "step": 9500
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.849892129001678,
      "learning_rate": 6.277841359274584e-05,
      "loss": 0.1212,
      "step": 9510
    },
    {
      "epoch": 2.41,
      "grad_norm": 4.163269713173385,
      "learning_rate": 6.25143058367814e-05,
      "loss": 0.1184,
      "step": 9520
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.7794331352302883,
      "learning_rate": 6.225019808081696e-05,
      "loss": 0.0563,
      "step": 9530
    },
    {
      "epoch": 2.41,
      "grad_norm": 18.627705595135957,
      "learning_rate": 6.198609032485254e-05,
      "loss": 0.1419,
      "step": 9540
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.2853968833910283,
      "learning_rate": 6.17219825688881e-05,
      "loss": 0.0906,
      "step": 9550
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.2661103776889737,
      "learning_rate": 6.145787481292367e-05,
      "loss": 0.1242,
      "step": 9560
    },
    {
      "epoch": 2.42,
      "grad_norm": 9.592938650565387,
      "learning_rate": 6.119376705695923e-05,
      "loss": 0.1942,
      "step": 9570
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.3300849423184784,
      "learning_rate": 6.0929659300994806e-05,
      "loss": 0.044,
      "step": 9580
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.2053997379213532,
      "learning_rate": 6.066555154503037e-05,
      "loss": 0.0928,
      "step": 9590
    },
    {
      "epoch": 2.43,
      "grad_norm": 7.19395583303236,
      "learning_rate": 6.0401443789065934e-05,
      "loss": 0.1125,
      "step": 9600
    },
    {
      "epoch": 2.43,
      "grad_norm": 8.269855434350298,
      "learning_rate": 6.01373360331015e-05,
      "loss": 0.2177,
      "step": 9610
    },
    {
      "epoch": 2.43,
      "grad_norm": 10.269506918035669,
      "learning_rate": 5.987322827713707e-05,
      "loss": 0.278,
      "step": 9620
    },
    {
      "epoch": 2.44,
      "grad_norm": 8.66184316610978,
      "learning_rate": 5.9609120521172636e-05,
      "loss": 0.103,
      "step": 9630
    },
    {
      "epoch": 2.44,
      "grad_norm": 7.945938685825119,
      "learning_rate": 5.9345012765208204e-05,
      "loss": 0.1133,
      "step": 9640
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.933574221157829,
      "learning_rate": 5.908090500924377e-05,
      "loss": 0.1305,
      "step": 9650
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.7879011912280413,
      "learning_rate": 5.881679725327933e-05,
      "loss": 0.0597,
      "step": 9660
    },
    {
      "epoch": 2.45,
      "grad_norm": 5.3794371918406965,
      "learning_rate": 5.85526894973149e-05,
      "loss": 0.1443,
      "step": 9670
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.5710056874707381,
      "learning_rate": 5.828858174135047e-05,
      "loss": 0.2218,
      "step": 9680
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.32058848381222327,
      "learning_rate": 5.8024473985386034e-05,
      "loss": 0.1602,
      "step": 9690
    },
    {
      "epoch": 2.45,
      "grad_norm": 6.793952549838277,
      "learning_rate": 5.77603662294216e-05,
      "loss": 0.1137,
      "step": 9700
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.9176053157133593,
      "learning_rate": 5.749625847345717e-05,
      "loss": 0.1521,
      "step": 9710
    },
    {
      "epoch": 2.46,
      "grad_norm": 4.959980234017326,
      "learning_rate": 5.723215071749273e-05,
      "loss": 0.0889,
      "step": 9720
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.0075759280667156,
      "learning_rate": 5.69680429615283e-05,
      "loss": 0.118,
      "step": 9730
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.6838990961744704,
      "learning_rate": 5.6703935205563865e-05,
      "loss": 0.088,
      "step": 9740
    },
    {
      "epoch": 2.47,
      "grad_norm": 6.310064678724968,
      "learning_rate": 5.643982744959943e-05,
      "loss": 0.3464,
      "step": 9750
    },
    {
      "epoch": 2.47,
      "grad_norm": 6.959426556418945,
      "learning_rate": 5.6175719693635e-05,
      "loss": 0.0988,
      "step": 9760
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.178761973866902,
      "learning_rate": 5.591161193767057e-05,
      "loss": 0.0488,
      "step": 9770
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.9189330144782912,
      "learning_rate": 5.5647504181706134e-05,
      "loss": 0.1747,
      "step": 9780
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.6121902809356623,
      "learning_rate": 5.5383396425741695e-05,
      "loss": 0.1073,
      "step": 9790
    },
    {
      "epoch": 2.48,
      "grad_norm": 13.509716413826606,
      "learning_rate": 5.511928866977726e-05,
      "loss": 0.1049,
      "step": 9800
    },
    {
      "epoch": 2.48,
      "grad_norm": 4.845712579474245,
      "learning_rate": 5.488159168940927e-05,
      "loss": 0.144,
      "step": 9810
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4574259294233924,
      "learning_rate": 5.461748393344484e-05,
      "loss": 0.0477,
      "step": 9820
    },
    {
      "epoch": 2.49,
      "grad_norm": 11.276614742678678,
      "learning_rate": 5.435337617748041e-05,
      "loss": 0.076,
      "step": 9830
    },
    {
      "epoch": 2.49,
      "grad_norm": 4.287269515433337,
      "learning_rate": 5.4089268421515975e-05,
      "loss": 0.1456,
      "step": 9840
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.7870008322276636,
      "learning_rate": 5.382516066555154e-05,
      "loss": 0.1846,
      "step": 9850
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.4576468257971762,
      "learning_rate": 5.356105290958711e-05,
      "loss": 0.045,
      "step": 9860
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.314871396726202,
      "learning_rate": 5.329694515362267e-05,
      "loss": 0.3069,
      "step": 9870
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.7718785896414182,
      "learning_rate": 5.303283739765824e-05,
      "loss": 0.0325,
      "step": 9880
    },
    {
      "epoch": 2.5,
      "grad_norm": 8.812536626766736,
      "learning_rate": 5.2768729641693805e-05,
      "loss": 0.0756,
      "step": 9890
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.869108315773467,
      "learning_rate": 5.250462188572937e-05,
      "loss": 0.0774,
      "step": 9900
    },
    {
      "epoch": 2.51,
      "grad_norm": 11.637977359161589,
      "learning_rate": 5.224051412976494e-05,
      "loss": 0.1496,
      "step": 9910
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.3582743400795256,
      "learning_rate": 5.197640637380051e-05,
      "loss": 0.1161,
      "step": 9920
    },
    {
      "epoch": 2.51,
      "grad_norm": 5.422991695341771,
      "learning_rate": 5.1712298617836075e-05,
      "loss": 0.0715,
      "step": 9930
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.9971699482282776,
      "learning_rate": 5.1448190861871635e-05,
      "loss": 0.1128,
      "step": 9940
    },
    {
      "epoch": 2.52,
      "grad_norm": 7.354984493390003,
      "learning_rate": 5.11840831059072e-05,
      "loss": 0.0326,
      "step": 9950
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.5444593746342109,
      "learning_rate": 5.091997534994277e-05,
      "loss": 0.0599,
      "step": 9960
    },
    {
      "epoch": 2.52,
      "grad_norm": 10.570928022813604,
      "learning_rate": 5.065586759397834e-05,
      "loss": 0.083,
      "step": 9970
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.0267858564413939,
      "learning_rate": 5.0391759838013905e-05,
      "loss": 0.2821,
      "step": 9980
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.2826701064869344,
      "learning_rate": 5.012765208204947e-05,
      "loss": 0.1695,
      "step": 9990
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.30108004892294266,
      "learning_rate": 4.986354432608504e-05,
      "loss": 0.0506,
      "step": 10000
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.9049715361211421,
      "learning_rate": 4.95994365701206e-05,
      "loss": 0.0949,
      "step": 10010
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.2007249390683805,
      "learning_rate": 4.933532881415617e-05,
      "loss": 0.1289,
      "step": 10020
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.5888459296090317,
      "learning_rate": 4.9071221058191736e-05,
      "loss": 0.0511,
      "step": 10030
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.9691912006046467,
      "learning_rate": 4.88071133022273e-05,
      "loss": 0.0447,
      "step": 10040
    },
    {
      "epoch": 2.54,
      "grad_norm": 6.5134129085276955,
      "learning_rate": 4.854300554626287e-05,
      "loss": 0.102,
      "step": 10050
    },
    {
      "epoch": 2.54,
      "grad_norm": 4.027136926862574,
      "learning_rate": 4.827889779029844e-05,
      "loss": 0.1908,
      "step": 10060
    },
    {
      "epoch": 2.55,
      "grad_norm": 19.250884994066908,
      "learning_rate": 4.8014790034334005e-05,
      "loss": 0.144,
      "step": 10070
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.8893908721884611,
      "learning_rate": 4.7750682278369566e-05,
      "loss": 0.1049,
      "step": 10080
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.11371244448898017,
      "learning_rate": 4.748657452240513e-05,
      "loss": 0.0542,
      "step": 10090
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.9529994334004073,
      "learning_rate": 4.72224667664407e-05,
      "loss": 0.2115,
      "step": 10100
    },
    {
      "epoch": 2.56,
      "grad_norm": 22.695910694426306,
      "learning_rate": 4.695835901047627e-05,
      "loss": 0.1871,
      "step": 10110
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.6508865368070988,
      "learning_rate": 4.6694251254511836e-05,
      "loss": 0.0771,
      "step": 10120
    },
    {
      "epoch": 2.56,
      "grad_norm": 12.688863533018672,
      "learning_rate": 4.64301434985474e-05,
      "loss": 0.1765,
      "step": 10130
    },
    {
      "epoch": 2.57,
      "grad_norm": 11.248140452930153,
      "learning_rate": 4.6166035742582964e-05,
      "loss": 0.144,
      "step": 10140
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.5056240274213692,
      "learning_rate": 4.590192798661853e-05,
      "loss": 0.1073,
      "step": 10150
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.1525606056414703,
      "learning_rate": 4.56378202306541e-05,
      "loss": 0.0754,
      "step": 10160
    },
    {
      "epoch": 2.57,
      "grad_norm": 4.077649798147871,
      "learning_rate": 4.5373712474689666e-05,
      "loss": 0.1682,
      "step": 10170
    },
    {
      "epoch": 2.58,
      "grad_norm": 7.123075586108915,
      "learning_rate": 4.510960471872524e-05,
      "loss": 0.0769,
      "step": 10180
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.07024333898162324,
      "learning_rate": 4.484549696276081e-05,
      "loss": 0.0514,
      "step": 10190
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.1302876456089308,
      "learning_rate": 4.4581389206796375e-05,
      "loss": 0.0357,
      "step": 10200
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.7017409405729192,
      "learning_rate": 4.431728145083193e-05,
      "loss": 0.1135,
      "step": 10210
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.957512424380371,
      "learning_rate": 4.4053173694867496e-05,
      "loss": 0.1427,
      "step": 10220
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.7624371766373486,
      "learning_rate": 4.378906593890307e-05,
      "loss": 0.0967,
      "step": 10230
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.0410211937203258,
      "learning_rate": 4.352495818293864e-05,
      "loss": 0.4155,
      "step": 10240
    },
    {
      "epoch": 2.59,
      "grad_norm": 7.167808785589543,
      "learning_rate": 4.3260850426974205e-05,
      "loss": 0.162,
      "step": 10250
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.6838180745732223,
      "learning_rate": 4.299674267100977e-05,
      "loss": 0.1394,
      "step": 10260
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.20084241971641661,
      "learning_rate": 4.273263491504534e-05,
      "loss": 0.0629,
      "step": 10270
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.9933652116745235,
      "learning_rate": 4.24685271590809e-05,
      "loss": 0.0847,
      "step": 10280
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.124410357695191,
      "learning_rate": 4.220441940311647e-05,
      "loss": 0.1249,
      "step": 10290
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.0881424374659778,
      "learning_rate": 4.1940311647152036e-05,
      "loss": 0.0532,
      "step": 10300
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.9462406945243451,
      "learning_rate": 4.16762038911876e-05,
      "loss": 0.1408,
      "step": 10310
    },
    {
      "epoch": 2.61,
      "grad_norm": 8.446099459905815,
      "learning_rate": 4.141209613522317e-05,
      "loss": 0.0914,
      "step": 10320
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.8034177016932733,
      "learning_rate": 4.114798837925874e-05,
      "loss": 0.2553,
      "step": 10330
    },
    {
      "epoch": 2.62,
      "grad_norm": 3.138858126459365,
      "learning_rate": 4.0883880623294306e-05,
      "loss": 0.0592,
      "step": 10340
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.6359838426779154,
      "learning_rate": 4.0619772867329866e-05,
      "loss": 0.0668,
      "step": 10350
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.644243083157715,
      "learning_rate": 4.0355665111365434e-05,
      "loss": 0.0462,
      "step": 10360
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.7635382174566347,
      "learning_rate": 4.0091557355401e-05,
      "loss": 0.0579,
      "step": 10370
    },
    {
      "epoch": 2.63,
      "grad_norm": 6.1535950795966405,
      "learning_rate": 3.982744959943657e-05,
      "loss": 0.3327,
      "step": 10380
    },
    {
      "epoch": 2.63,
      "grad_norm": 13.05703208638941,
      "learning_rate": 3.9563341843472136e-05,
      "loss": 0.115,
      "step": 10390
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.4184167895624646,
      "learning_rate": 3.92992340875077e-05,
      "loss": 0.0391,
      "step": 10400
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.42191460797965785,
      "learning_rate": 3.9035126331543264e-05,
      "loss": 0.1597,
      "step": 10410
    },
    {
      "epoch": 2.64,
      "grad_norm": 14.6318480422782,
      "learning_rate": 3.877101857557883e-05,
      "loss": 0.0733,
      "step": 10420
    },
    {
      "epoch": 2.64,
      "grad_norm": 5.671967203531022,
      "learning_rate": 3.85069108196144e-05,
      "loss": 0.1082,
      "step": 10430
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.769743830367795,
      "learning_rate": 3.8242803063649966e-05,
      "loss": 0.0929,
      "step": 10440
    },
    {
      "epoch": 2.64,
      "grad_norm": 9.473401117829091,
      "learning_rate": 3.7978695307685534e-05,
      "loss": 0.0671,
      "step": 10450
    },
    {
      "epoch": 2.65,
      "grad_norm": 3.10818851060489,
      "learning_rate": 3.77145875517211e-05,
      "loss": 0.1062,
      "step": 10460
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.19493193116698393,
      "learning_rate": 3.745047979575666e-05,
      "loss": 0.075,
      "step": 10470
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.24854901716368594,
      "learning_rate": 3.718637203979223e-05,
      "loss": 0.2781,
      "step": 10480
    },
    {
      "epoch": 2.65,
      "grad_norm": 5.057687803468933,
      "learning_rate": 3.6922264283827803e-05,
      "loss": 0.1657,
      "step": 10490
    },
    {
      "epoch": 2.66,
      "grad_norm": 4.141377021767007,
      "learning_rate": 3.6658156527863364e-05,
      "loss": 0.1454,
      "step": 10500
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.9507559207300316,
      "learning_rate": 3.639404877189893e-05,
      "loss": 0.0966,
      "step": 10510
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.5757066011596175,
      "learning_rate": 3.61299410159345e-05,
      "loss": 0.0875,
      "step": 10520
    },
    {
      "epoch": 2.66,
      "grad_norm": 13.83734841586774,
      "learning_rate": 3.586583325997006e-05,
      "loss": 0.108,
      "step": 10530
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.15044361157293365,
      "learning_rate": 3.5601725504005634e-05,
      "loss": 0.042,
      "step": 10540
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.7291444270290097,
      "learning_rate": 3.53376177480412e-05,
      "loss": 0.0845,
      "step": 10550
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.8376508635937641,
      "learning_rate": 3.507350999207677e-05,
      "loss": 0.1072,
      "step": 10560
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.662374821524742,
      "learning_rate": 3.480940223611233e-05,
      "loss": 0.044,
      "step": 10570
    },
    {
      "epoch": 2.68,
      "grad_norm": 4.382422265786289,
      "learning_rate": 3.45452944801479e-05,
      "loss": 0.0257,
      "step": 10580
    },
    {
      "epoch": 2.68,
      "grad_norm": 3.7729275642209026,
      "learning_rate": 3.4281186724183464e-05,
      "loss": 0.0982,
      "step": 10590
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.6167807277164736,
      "learning_rate": 3.401707896821903e-05,
      "loss": 0.1444,
      "step": 10600
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.3390979840123256,
      "learning_rate": 3.37529712122546e-05,
      "loss": 0.2813,
      "step": 10610
    },
    {
      "epoch": 2.69,
      "grad_norm": 11.095709715607736,
      "learning_rate": 3.3488863456290166e-05,
      "loss": 0.1907,
      "step": 10620
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.47254607817072725,
      "learning_rate": 3.322475570032573e-05,
      "loss": 0.0919,
      "step": 10630
    },
    {
      "epoch": 2.69,
      "grad_norm": 2.943883990242144,
      "learning_rate": 3.2960647944361295e-05,
      "loss": 0.0797,
      "step": 10640
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.4137256867290577,
      "learning_rate": 3.269654018839686e-05,
      "loss": 0.1045,
      "step": 10650
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.797400283682247,
      "learning_rate": 3.243243243243243e-05,
      "loss": 0.2231,
      "step": 10660
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.18773776950023852,
      "learning_rate": 3.2168324676468e-05,
      "loss": 0.0312,
      "step": 10670
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.40299221520593376,
      "learning_rate": 3.1904216920503564e-05,
      "loss": 0.1194,
      "step": 10680
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.3289196326435833,
      "learning_rate": 3.164010916453913e-05,
      "loss": 0.0647,
      "step": 10690
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.7116763771003195,
      "learning_rate": 3.137600140857469e-05,
      "loss": 0.0626,
      "step": 10700
    },
    {
      "epoch": 2.71,
      "grad_norm": 8.269537704530125,
      "learning_rate": 3.111189365261026e-05,
      "loss": 0.0452,
      "step": 10710
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.5441648707031814,
      "learning_rate": 3.084778589664583e-05,
      "loss": 0.2587,
      "step": 10720
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.2940876118032927,
      "learning_rate": 3.0583678140681395e-05,
      "loss": 0.1419,
      "step": 10730
    },
    {
      "epoch": 2.72,
      "grad_norm": 7.204826101361029,
      "learning_rate": 3.0319570384716962e-05,
      "loss": 0.0269,
      "step": 10740
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.9970746450068529,
      "learning_rate": 3.005546262875253e-05,
      "loss": 0.1342,
      "step": 10750
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.24876466316797904,
      "learning_rate": 2.9791354872788097e-05,
      "loss": 0.0512,
      "step": 10760
    },
    {
      "epoch": 2.72,
      "grad_norm": 6.37439302168123,
      "learning_rate": 2.952724711682366e-05,
      "loss": 0.0765,
      "step": 10770
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.446435610481913,
      "learning_rate": 2.926313936085923e-05,
      "loss": 0.1737,
      "step": 10780
    },
    {
      "epoch": 2.73,
      "grad_norm": 13.127859094837586,
      "learning_rate": 2.8999031604894796e-05,
      "loss": 0.1934,
      "step": 10790
    },
    {
      "epoch": 2.73,
      "grad_norm": 12.34255516208749,
      "learning_rate": 2.873492384893036e-05,
      "loss": 0.0914,
      "step": 10800
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.932453233603026,
      "learning_rate": 2.8470816092965927e-05,
      "loss": 0.1355,
      "step": 10810
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.3836301473079415,
      "learning_rate": 2.8206708337001495e-05,
      "loss": 0.0612,
      "step": 10820
    },
    {
      "epoch": 2.74,
      "grad_norm": 17.59760420746046,
      "learning_rate": 2.7942600581037062e-05,
      "loss": 0.09,
      "step": 10830
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.14360702340771692,
      "learning_rate": 2.7678492825072626e-05,
      "loss": 0.0667,
      "step": 10840
    },
    {
      "epoch": 2.74,
      "grad_norm": 3.131777089264597,
      "learning_rate": 2.7414385069108194e-05,
      "loss": 0.1506,
      "step": 10850
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.36724519212978596,
      "learning_rate": 2.715027731314376e-05,
      "loss": 0.0541,
      "step": 10860
    },
    {
      "epoch": 2.75,
      "grad_norm": 6.116694134514987,
      "learning_rate": 2.6886169557179325e-05,
      "loss": 0.0685,
      "step": 10870
    },
    {
      "epoch": 2.75,
      "grad_norm": 28.441147635297522,
      "learning_rate": 2.6622061801214893e-05,
      "loss": 0.182,
      "step": 10880
    },
    {
      "epoch": 2.75,
      "grad_norm": 6.207440857908955,
      "learning_rate": 2.635795404525046e-05,
      "loss": 0.0841,
      "step": 10890
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.2660646664649682,
      "learning_rate": 2.6093846289286024e-05,
      "loss": 0.1044,
      "step": 10900
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.6092081949682532,
      "learning_rate": 2.582973853332159e-05,
      "loss": 0.1218,
      "step": 10910
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.26754241443153093,
      "learning_rate": 2.556563077735716e-05,
      "loss": 0.0755,
      "step": 10920
    },
    {
      "epoch": 2.76,
      "grad_norm": 4.614645486521455,
      "learning_rate": 2.5301523021392726e-05,
      "loss": 0.2079,
      "step": 10930
    },
    {
      "epoch": 2.77,
      "grad_norm": 10.500071506881186,
      "learning_rate": 2.503741526542829e-05,
      "loss": 0.12,
      "step": 10940
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.4157532308157768,
      "learning_rate": 2.4773307509463858e-05,
      "loss": 0.0486,
      "step": 10950
    },
    {
      "epoch": 2.77,
      "grad_norm": 3.090458090289476,
      "learning_rate": 2.450919975349943e-05,
      "loss": 0.0922,
      "step": 10960
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.725234178573422,
      "learning_rate": 2.424509199753499e-05,
      "loss": 0.1187,
      "step": 10970
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.883964200616445,
      "learning_rate": 2.3980984241570557e-05,
      "loss": 0.0991,
      "step": 10980
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.8016329522459724,
      "learning_rate": 2.3716876485606127e-05,
      "loss": 0.2729,
      "step": 10990
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.19691039993042744,
      "learning_rate": 2.3452768729641688e-05,
      "loss": 0.0165,
      "step": 11000
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.7931561296721391,
      "learning_rate": 2.318866097367726e-05,
      "loss": 0.0438,
      "step": 11010
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.3666105715233941,
      "learning_rate": 2.2924553217712826e-05,
      "loss": 0.0767,
      "step": 11020
    },
    {
      "epoch": 2.79,
      "grad_norm": 7.346452519095264,
      "learning_rate": 2.2660445461748394e-05,
      "loss": 0.0461,
      "step": 11030
    },
    {
      "epoch": 2.79,
      "grad_norm": 2.0799370688756174,
      "learning_rate": 2.2396337705783958e-05,
      "loss": 0.1117,
      "step": 11040
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.6935165088282154,
      "learning_rate": 2.2132229949819525e-05,
      "loss": 0.1418,
      "step": 11050
    },
    {
      "epoch": 2.8,
      "grad_norm": 11.225662105081213,
      "learning_rate": 2.1868122193855093e-05,
      "loss": 0.1905,
      "step": 11060
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.286056290531877,
      "learning_rate": 2.1604014437890657e-05,
      "loss": 0.1624,
      "step": 11070
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.18686178060435013,
      "learning_rate": 2.1339906681926224e-05,
      "loss": 0.0206,
      "step": 11080
    },
    {
      "epoch": 2.81,
      "grad_norm": 11.939093895301632,
      "learning_rate": 2.107579892596179e-05,
      "loss": 0.1333,
      "step": 11090
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.12096721036928222,
      "learning_rate": 2.081169116999736e-05,
      "loss": 0.031,
      "step": 11100
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.8451614998694545,
      "learning_rate": 2.057399418962937e-05,
      "loss": 0.1967,
      "step": 11110
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.6001248999495915,
      "learning_rate": 2.0309886433664933e-05,
      "loss": 0.0261,
      "step": 11120
    },
    {
      "epoch": 2.82,
      "grad_norm": 7.194025556223317,
      "learning_rate": 2.00457786777005e-05,
      "loss": 0.1301,
      "step": 11130
    },
    {
      "epoch": 2.82,
      "grad_norm": 7.208076215728769,
      "learning_rate": 1.9781670921736068e-05,
      "loss": 0.0994,
      "step": 11140
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.9648443648687712,
      "learning_rate": 1.9517563165771632e-05,
      "loss": 0.1464,
      "step": 11150
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.9057615164332242,
      "learning_rate": 1.92534554098072e-05,
      "loss": 0.1518,
      "step": 11160
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.227054647742425,
      "learning_rate": 1.8989347653842767e-05,
      "loss": 0.1274,
      "step": 11170
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.4994864050705437,
      "learning_rate": 1.872523989787833e-05,
      "loss": 0.0346,
      "step": 11180
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.6845515892006273,
      "learning_rate": 1.8461132141913902e-05,
      "loss": 0.0569,
      "step": 11190
    },
    {
      "epoch": 2.83,
      "grad_norm": 7.306151141443391,
      "learning_rate": 1.8197024385949466e-05,
      "loss": 0.1802,
      "step": 11200
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.579042452582513,
      "learning_rate": 1.793291662998503e-05,
      "loss": 0.2344,
      "step": 11210
    },
    {
      "epoch": 2.84,
      "grad_norm": 10.380965944821517,
      "learning_rate": 1.76688088740206e-05,
      "loss": 0.1569,
      "step": 11220
    },
    {
      "epoch": 2.84,
      "grad_norm": 10.372858906621042,
      "learning_rate": 1.7404701118056165e-05,
      "loss": 0.0742,
      "step": 11230
    },
    {
      "epoch": 2.84,
      "grad_norm": 6.759873407800237,
      "learning_rate": 1.7140593362091732e-05,
      "loss": 0.1816,
      "step": 11240
    },
    {
      "epoch": 2.85,
      "grad_norm": 4.903482124866322,
      "learning_rate": 1.68764856061273e-05,
      "loss": 0.1935,
      "step": 11250
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.4418667018871049,
      "learning_rate": 1.6612377850162864e-05,
      "loss": 0.2005,
      "step": 11260
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.8868905220885367,
      "learning_rate": 1.634827009419843e-05,
      "loss": 0.062,
      "step": 11270
    },
    {
      "epoch": 2.85,
      "grad_norm": 9.782640698777652,
      "learning_rate": 1.6084162338234e-05,
      "loss": 0.2171,
      "step": 11280
    },
    {
      "epoch": 2.86,
      "grad_norm": 5.24892181652971,
      "learning_rate": 1.5820054582269566e-05,
      "loss": 0.082,
      "step": 11290
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.8228753879855804,
      "learning_rate": 1.555594682630513e-05,
      "loss": 0.0686,
      "step": 11300
    },
    {
      "epoch": 2.86,
      "grad_norm": 7.80524687342384,
      "learning_rate": 1.5291839070340697e-05,
      "loss": 0.1381,
      "step": 11310
    },
    {
      "epoch": 2.86,
      "grad_norm": 11.936039568158414,
      "learning_rate": 1.5027731314376265e-05,
      "loss": 0.1963,
      "step": 11320
    },
    {
      "epoch": 2.87,
      "grad_norm": 10.237736806471208,
      "learning_rate": 1.476362355841183e-05,
      "loss": 0.1184,
      "step": 11330
    },
    {
      "epoch": 2.87,
      "grad_norm": 10.116058015555941,
      "learning_rate": 1.4499515802447398e-05,
      "loss": 0.1078,
      "step": 11340
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.0749867720444426,
      "learning_rate": 1.4235408046482964e-05,
      "loss": 0.0955,
      "step": 11350
    },
    {
      "epoch": 2.87,
      "grad_norm": 7.283841186751001,
      "learning_rate": 1.3971300290518531e-05,
      "loss": 0.1663,
      "step": 11360
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.3864174609556974,
      "learning_rate": 1.3707192534554097e-05,
      "loss": 0.0397,
      "step": 11370
    },
    {
      "epoch": 2.88,
      "grad_norm": 6.971133001085124,
      "learning_rate": 1.3443084778589663e-05,
      "loss": 0.0954,
      "step": 11380
    },
    {
      "epoch": 2.88,
      "grad_norm": 21.56014161964891,
      "learning_rate": 1.317897702262523e-05,
      "loss": 0.1602,
      "step": 11390
    },
    {
      "epoch": 2.88,
      "grad_norm": 2.356408551684661,
      "learning_rate": 1.2914869266660796e-05,
      "loss": 0.052,
      "step": 11400
    },
    {
      "epoch": 2.89,
      "grad_norm": 3.9611241524261565,
      "learning_rate": 1.2650761510696363e-05,
      "loss": 0.1536,
      "step": 11410
    },
    {
      "epoch": 2.89,
      "grad_norm": 7.450833839158113,
      "learning_rate": 1.2386653754731929e-05,
      "loss": 0.0765,
      "step": 11420
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.7570334390289082,
      "learning_rate": 1.2122545998767495e-05,
      "loss": 0.0384,
      "step": 11430
    },
    {
      "epoch": 2.89,
      "grad_norm": 4.083706888762959,
      "learning_rate": 1.1858438242803064e-05,
      "loss": 0.0996,
      "step": 11440
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.9197630989003147,
      "learning_rate": 1.159433048683863e-05,
      "loss": 0.1304,
      "step": 11450
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.0271103895558773,
      "learning_rate": 1.1330222730874197e-05,
      "loss": 0.1985,
      "step": 11460
    },
    {
      "epoch": 2.9,
      "grad_norm": 6.322376901933602,
      "learning_rate": 1.1066114974909763e-05,
      "loss": 0.0701,
      "step": 11470
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.7574464728810889,
      "learning_rate": 1.0802007218945328e-05,
      "loss": 0.0632,
      "step": 11480
    },
    {
      "epoch": 2.91,
      "grad_norm": 15.947801577922908,
      "learning_rate": 1.0537899462980896e-05,
      "loss": 0.1954,
      "step": 11490
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.5012380137106978,
      "learning_rate": 1.0273791707016462e-05,
      "loss": 0.1728,
      "step": 11500
    },
    {
      "epoch": 2.91,
      "grad_norm": 5.590052436169391,
      "learning_rate": 1.0009683951052029e-05,
      "loss": 0.0982,
      "step": 11510
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.508898054335977,
      "learning_rate": 9.745576195087595e-06,
      "loss": 0.039,
      "step": 11520
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.9269851244005981,
      "learning_rate": 9.48146843912316e-06,
      "loss": 0.0665,
      "step": 11530
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.10790991945570794,
      "learning_rate": 9.217360683158728e-06,
      "loss": 0.0484,
      "step": 11540
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.3788796784544578,
      "learning_rate": 8.953252927194295e-06,
      "loss": 0.095,
      "step": 11550
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.277637154263989,
      "learning_rate": 8.689145171229861e-06,
      "loss": 0.0394,
      "step": 11560
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5254078523987122,
      "learning_rate": 8.425037415265427e-06,
      "loss": 0.0651,
      "step": 11570
    },
    {
      "epoch": 2.93,
      "grad_norm": 4.139641683707518,
      "learning_rate": 8.160929659300994e-06,
      "loss": 0.173,
      "step": 11580
    },
    {
      "epoch": 2.93,
      "grad_norm": 3.7160176647076018,
      "learning_rate": 7.896821903336562e-06,
      "loss": 0.1085,
      "step": 11590
    },
    {
      "epoch": 2.93,
      "grad_norm": 4.628653376951481,
      "learning_rate": 7.632714147372127e-06,
      "loss": 0.2646,
      "step": 11600
    },
    {
      "epoch": 2.94,
      "grad_norm": 4.630616633787641,
      "learning_rate": 7.368606391407693e-06,
      "loss": 0.1046,
      "step": 11610
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.3096562218736908,
      "learning_rate": 7.10449863544326e-06,
      "loss": 0.1456,
      "step": 11620
    },
    {
      "epoch": 2.94,
      "grad_norm": 21.08824128593129,
      "learning_rate": 6.840390879478826e-06,
      "loss": 0.1656,
      "step": 11630
    },
    {
      "epoch": 2.94,
      "grad_norm": 5.7194402635098305,
      "learning_rate": 6.576283123514394e-06,
      "loss": 0.0433,
      "step": 11640
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.1220445555277994,
      "learning_rate": 6.31217536754996e-06,
      "loss": 0.1068,
      "step": 11650
    },
    {
      "epoch": 2.95,
      "grad_norm": 2.4230288425705027,
      "learning_rate": 6.048067611585527e-06,
      "loss": 0.0179,
      "step": 11660
    },
    {
      "epoch": 2.95,
      "grad_norm": 5.344051738078155,
      "learning_rate": 5.783959855621093e-06,
      "loss": 0.0465,
      "step": 11670
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.4671004615480219,
      "learning_rate": 5.519852099656659e-06,
      "loss": 0.0869,
      "step": 11680
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.3656127093170958,
      "learning_rate": 5.255744343692226e-06,
      "loss": 0.1712,
      "step": 11690
    },
    {
      "epoch": 2.96,
      "grad_norm": 33.16154086138305,
      "learning_rate": 4.991636587727792e-06,
      "loss": 0.2911,
      "step": 11700
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.0260483559873532,
      "learning_rate": 4.72752883176336e-06,
      "loss": 0.0819,
      "step": 11710
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.0503134350842027,
      "learning_rate": 4.4634210757989255e-06,
      "loss": 0.0243,
      "step": 11720
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.4059795773621981,
      "learning_rate": 4.199313319834492e-06,
      "loss": 0.0404,
      "step": 11730
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.6346516212919995,
      "learning_rate": 3.935205563870059e-06,
      "loss": 0.0822,
      "step": 11740
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.7317518527949531,
      "learning_rate": 3.6710978079056253e-06,
      "loss": 0.066,
      "step": 11750
    },
    {
      "epoch": 2.97,
      "grad_norm": 22.93625083972818,
      "learning_rate": 3.4069900519411914e-06,
      "loss": 0.1876,
      "step": 11760
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.49771630803031175,
      "learning_rate": 3.1428822959767584e-06,
      "loss": 0.1169,
      "step": 11770
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.4296662652804666,
      "learning_rate": 2.878774540012325e-06,
      "loss": 0.0554,
      "step": 11780
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.29480468277518473,
      "learning_rate": 2.614666784047891e-06,
      "loss": 0.0501,
      "step": 11790
    },
    {
      "epoch": 2.99,
      "grad_norm": 8.666677806901344,
      "learning_rate": 2.3505590280834578e-06,
      "loss": 0.1742,
      "step": 11800
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.787911757147805,
      "learning_rate": 2.0864512721190243e-06,
      "loss": 0.073,
      "step": 11810
    },
    {
      "epoch": 2.99,
      "grad_norm": 15.07480824018013,
      "learning_rate": 1.822343516154591e-06,
      "loss": 0.1233,
      "step": 11820
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.4945345457744,
      "learning_rate": 1.5582357601901575e-06,
      "loss": 0.2263,
      "step": 11830
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.27772193620873387,
      "learning_rate": 1.294128004225724e-06,
      "loss": 0.1931,
      "step": 11840
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.517522622984517,
      "learning_rate": 1.0300202482612905e-06,
      "loss": 0.0399,
      "step": 11850
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8785213891787151,
      "eval_f1": 0.8438316361697416,
      "eval_loss": 0.41259765625,
      "eval_precision": 0.8355454067673094,
      "eval_recall": 0.8536051899559176,
      "eval_runtime": 462.5898,
      "eval_samples_per_second": 14.503,
      "eval_steps_per_second": 2.419,
      "step": 11859
    }
  ],
  "logging_steps": 10,
  "max_steps": 11859,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2.229067174683607e+17,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}