{
  "best_metric": 0.34521484375,
  "best_model_checkpoint": "./results/checkpoint-7418",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 11127,
  "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": 3.7463,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 3.1097,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 300.04090826228907,
      "learning_rate": 4.8e-06,
      "loss": 3.4556,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 151.89737946570972,
      "learning_rate": 1.0799999999999998e-05,
      "loss": 2.1859,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 144.06341123478924,
      "learning_rate": 1.6199999999999997e-05,
      "loss": 1.7061,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 113.48755246440719,
      "learning_rate": 2.2199999999999998e-05,
      "loss": 1.4973,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 67.06629665751738,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.8511,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 29.27148018028355,
      "learning_rate": 3.42e-05,
      "loss": 1.7671,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 60.212733389331035,
      "learning_rate": 4.02e-05,
      "loss": 1.2902,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 52.37862444698052,
      "learning_rate": 4.62e-05,
      "loss": 1.0847,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.298008571728236,
      "learning_rate": 5.2199999999999995e-05,
      "loss": 0.5683,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.723020375882976,
      "learning_rate": 5.82e-05,
      "loss": 0.5352,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 16.56814287433871,
      "learning_rate": 6.419999999999999e-05,
      "loss": 0.5804,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 63.82119413026771,
      "learning_rate": 7.02e-05,
      "loss": 0.7593,
      "step": 140
    },
    {
      "epoch": 0.04,
      "grad_norm": 50.021163600396406,
      "learning_rate": 7.62e-05,
      "loss": 1.0557,
      "step": 150
    },
    {
      "epoch": 0.04,
      "grad_norm": 50.186055961917305,
      "learning_rate": 8.22e-05,
      "loss": 0.9004,
      "step": 160
    },
    {
      "epoch": 0.05,
      "grad_norm": 196.05109537485546,
      "learning_rate": 8.819999999999999e-05,
      "loss": 1.2988,
      "step": 170
    },
    {
      "epoch": 0.05,
      "grad_norm": 47.754947201357446,
      "learning_rate": 9.419999999999999e-05,
      "loss": 0.7852,
      "step": 180
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.778177762866042,
      "learning_rate": 0.0001002,
      "loss": 0.7564,
      "step": 190
    },
    {
      "epoch": 0.05,
      "grad_norm": 109.69969588509676,
      "learning_rate": 0.00010619999999999998,
      "loss": 0.6766,
      "step": 200
    },
    {
      "epoch": 0.06,
      "grad_norm": 68.25127398521032,
      "learning_rate": 0.00011219999999999999,
      "loss": 0.9808,
      "step": 210
    },
    {
      "epoch": 0.06,
      "grad_norm": 70.95983665290264,
      "learning_rate": 0.0001182,
      "loss": 0.9574,
      "step": 220
    },
    {
      "epoch": 0.06,
      "grad_norm": 51.484844694180254,
      "learning_rate": 0.00012419999999999998,
      "loss": 0.7799,
      "step": 230
    },
    {
      "epoch": 0.06,
      "grad_norm": 47.70279317890992,
      "learning_rate": 0.0001302,
      "loss": 0.656,
      "step": 240
    },
    {
      "epoch": 0.07,
      "grad_norm": 5.376240714441166,
      "learning_rate": 0.0001362,
      "loss": 0.6618,
      "step": 250
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.2275602479591425,
      "learning_rate": 0.0001422,
      "loss": 0.655,
      "step": 260
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.127110309069806,
      "learning_rate": 0.0001482,
      "loss": 0.7775,
      "step": 270
    },
    {
      "epoch": 0.08,
      "grad_norm": 32.641025781795946,
      "learning_rate": 0.00015419999999999998,
      "loss": 0.6527,
      "step": 280
    },
    {
      "epoch": 0.08,
      "grad_norm": 19.19501597234053,
      "learning_rate": 0.0001602,
      "loss": 0.6196,
      "step": 290
    },
    {
      "epoch": 0.08,
      "grad_norm": 44.31601085528485,
      "learning_rate": 0.0001662,
      "loss": 0.6697,
      "step": 300
    },
    {
      "epoch": 0.08,
      "grad_norm": 33.392592181834054,
      "learning_rate": 0.00017219999999999998,
      "loss": 0.4901,
      "step": 310
    },
    {
      "epoch": 0.09,
      "grad_norm": 34.41374458096861,
      "learning_rate": 0.00017819999999999997,
      "loss": 0.7389,
      "step": 320
    },
    {
      "epoch": 0.09,
      "grad_norm": 43.544184538336324,
      "learning_rate": 0.00018419999999999998,
      "loss": 0.7061,
      "step": 330
    },
    {
      "epoch": 0.09,
      "grad_norm": 40.998049495185,
      "learning_rate": 0.0001902,
      "loss": 0.6555,
      "step": 340
    },
    {
      "epoch": 0.09,
      "grad_norm": 57.85934913103111,
      "learning_rate": 0.0001962,
      "loss": 0.7142,
      "step": 350
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.682392110286486,
      "learning_rate": 0.0002022,
      "loss": 0.6681,
      "step": 360
    },
    {
      "epoch": 0.1,
      "grad_norm": 43.649047515809656,
      "learning_rate": 0.00020819999999999996,
      "loss": 0.7526,
      "step": 370
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.5823197386435215,
      "learning_rate": 0.00021419999999999998,
      "loss": 0.6786,
      "step": 380
    },
    {
      "epoch": 0.11,
      "grad_norm": 28.557723747320463,
      "learning_rate": 0.00022019999999999999,
      "loss": 0.4845,
      "step": 390
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.425640290516206,
      "learning_rate": 0.00022619999999999997,
      "loss": 0.7593,
      "step": 400
    },
    {
      "epoch": 0.11,
      "grad_norm": 17.72667761954851,
      "learning_rate": 0.00023219999999999998,
      "loss": 0.5938,
      "step": 410
    },
    {
      "epoch": 0.11,
      "grad_norm": 74.72777509847097,
      "learning_rate": 0.0002382,
      "loss": 0.798,
      "step": 420
    },
    {
      "epoch": 0.12,
      "grad_norm": 24.16558829835931,
      "learning_rate": 0.00024419999999999997,
      "loss": 0.6362,
      "step": 430
    },
    {
      "epoch": 0.12,
      "grad_norm": 17.895129033929084,
      "learning_rate": 0.00025019999999999996,
      "loss": 0.4857,
      "step": 440
    },
    {
      "epoch": 0.12,
      "grad_norm": 1287.3065164842492,
      "learning_rate": 0.0002562,
      "loss": 3.325,
      "step": 450
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.942601599313834,
      "learning_rate": 0.0002622,
      "loss": 0.9582,
      "step": 460
    },
    {
      "epoch": 0.13,
      "grad_norm": 69.91513374949557,
      "learning_rate": 0.00026819999999999996,
      "loss": 0.6275,
      "step": 470
    },
    {
      "epoch": 0.13,
      "grad_norm": 46.88539106654641,
      "learning_rate": 0.0002742,
      "loss": 0.98,
      "step": 480
    },
    {
      "epoch": 0.13,
      "grad_norm": 70.9293973009022,
      "learning_rate": 0.0002802,
      "loss": 1.0227,
      "step": 490
    },
    {
      "epoch": 0.13,
      "grad_norm": 20.862997728073683,
      "learning_rate": 0.00028619999999999996,
      "loss": 0.5328,
      "step": 500
    },
    {
      "epoch": 0.14,
      "grad_norm": 22.96406371340567,
      "learning_rate": 0.00029219999999999995,
      "loss": 0.5089,
      "step": 510
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.816845140615143,
      "learning_rate": 0.0002982,
      "loss": 0.8142,
      "step": 520
    },
    {
      "epoch": 0.14,
      "grad_norm": 35.79428480843495,
      "learning_rate": 0.00029980239013832686,
      "loss": 0.6503,
      "step": 530
    },
    {
      "epoch": 0.15,
      "grad_norm": 24.61129458055586,
      "learning_rate": 0.0002995200903359367,
      "loss": 1.6341,
      "step": 540
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.942655190217819,
      "learning_rate": 0.0002992377905335466,
      "loss": 0.5908,
      "step": 550
    },
    {
      "epoch": 0.15,
      "grad_norm": 19.502501331889697,
      "learning_rate": 0.00029895549073115645,
      "loss": 0.7248,
      "step": 560
    },
    {
      "epoch": 0.15,
      "grad_norm": 104.07707423830978,
      "learning_rate": 0.00029867319092876635,
      "loss": 1.3126,
      "step": 570
    },
    {
      "epoch": 0.16,
      "grad_norm": 23.30813855697986,
      "learning_rate": 0.0002983908911263762,
      "loss": 0.6965,
      "step": 580
    },
    {
      "epoch": 0.16,
      "grad_norm": 52.04895306209027,
      "learning_rate": 0.00029810859132398603,
      "loss": 0.599,
      "step": 590
    },
    {
      "epoch": 0.16,
      "grad_norm": 70.98008281100472,
      "learning_rate": 0.00029782629152159593,
      "loss": 0.8273,
      "step": 600
    },
    {
      "epoch": 0.16,
      "grad_norm": 22.853667446414942,
      "learning_rate": 0.0002975439917192058,
      "loss": 0.7611,
      "step": 610
    },
    {
      "epoch": 0.17,
      "grad_norm": 13.227033936173223,
      "learning_rate": 0.0002972616919168156,
      "loss": 1.3222,
      "step": 620
    },
    {
      "epoch": 0.17,
      "grad_norm": 21.79897660990205,
      "learning_rate": 0.0002969793921144255,
      "loss": 0.6837,
      "step": 630
    },
    {
      "epoch": 0.17,
      "grad_norm": 62.15546193829161,
      "learning_rate": 0.00029669709231203536,
      "loss": 0.8359,
      "step": 640
    },
    {
      "epoch": 0.18,
      "grad_norm": 40.661237412964134,
      "learning_rate": 0.0002964147925096452,
      "loss": 0.7055,
      "step": 650
    },
    {
      "epoch": 0.18,
      "grad_norm": 66.50686260402266,
      "learning_rate": 0.0002961324927072551,
      "loss": 0.5653,
      "step": 660
    },
    {
      "epoch": 0.18,
      "grad_norm": 21.77413429490788,
      "learning_rate": 0.00029585019290486494,
      "loss": 0.6121,
      "step": 670
    },
    {
      "epoch": 0.18,
      "grad_norm": 36.732773764989574,
      "learning_rate": 0.00029556789310247484,
      "loss": 0.4467,
      "step": 680
    },
    {
      "epoch": 0.19,
      "grad_norm": 38.75833095241455,
      "learning_rate": 0.00029528559330008463,
      "loss": 0.6546,
      "step": 690
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.608711503185333,
      "learning_rate": 0.0002950032934976945,
      "loss": 0.5722,
      "step": 700
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.5825087230140324,
      "learning_rate": 0.0002947209936953044,
      "loss": 0.5519,
      "step": 710
    },
    {
      "epoch": 0.19,
      "grad_norm": 53.89599045466506,
      "learning_rate": 0.0002944386938929142,
      "loss": 0.6136,
      "step": 720
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.509377443230084,
      "learning_rate": 0.0002941563940905241,
      "loss": 0.6329,
      "step": 730
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.462002401326377,
      "learning_rate": 0.00029387409428813395,
      "loss": 0.5336,
      "step": 740
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.8772206157826044,
      "learning_rate": 0.00029359179448574385,
      "loss": 0.4823,
      "step": 750
    },
    {
      "epoch": 0.2,
      "grad_norm": 40.28931626718375,
      "learning_rate": 0.0002933094946833537,
      "loss": 0.6147,
      "step": 760
    },
    {
      "epoch": 0.21,
      "grad_norm": 54.401110261721996,
      "learning_rate": 0.00029302719488096354,
      "loss": 0.5683,
      "step": 770
    },
    {
      "epoch": 0.21,
      "grad_norm": 19.014247657901926,
      "learning_rate": 0.00029274489507857344,
      "loss": 0.6145,
      "step": 780
    },
    {
      "epoch": 0.21,
      "grad_norm": 19.183372336072868,
      "learning_rate": 0.0002924625952761833,
      "loss": 0.588,
      "step": 790
    },
    {
      "epoch": 0.22,
      "grad_norm": 33.03015823215593,
      "learning_rate": 0.0002921802954737931,
      "loss": 0.5984,
      "step": 800
    },
    {
      "epoch": 0.22,
      "grad_norm": 15.681378328398644,
      "learning_rate": 0.000291897995671403,
      "loss": 0.622,
      "step": 810
    },
    {
      "epoch": 0.22,
      "grad_norm": 42.24357706405069,
      "learning_rate": 0.00029161569586901286,
      "loss": 0.526,
      "step": 820
    },
    {
      "epoch": 0.22,
      "grad_norm": 39.128308187954225,
      "learning_rate": 0.0002913333960666227,
      "loss": 0.6394,
      "step": 830
    },
    {
      "epoch": 0.23,
      "grad_norm": 30.04682221559281,
      "learning_rate": 0.0002910510962642326,
      "loss": 0.596,
      "step": 840
    },
    {
      "epoch": 0.23,
      "grad_norm": 79.25471813690982,
      "learning_rate": 0.00029076879646184245,
      "loss": 0.6248,
      "step": 850
    },
    {
      "epoch": 0.23,
      "grad_norm": 20.96294230550091,
      "learning_rate": 0.00029048649665945234,
      "loss": 0.6063,
      "step": 860
    },
    {
      "epoch": 0.23,
      "grad_norm": 35.17083101934739,
      "learning_rate": 0.0002902041968570622,
      "loss": 0.6492,
      "step": 870
    },
    {
      "epoch": 0.24,
      "grad_norm": 18.223650272893128,
      "learning_rate": 0.00028992189705467203,
      "loss": 0.559,
      "step": 880
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.715660152769105,
      "learning_rate": 0.00028963959725228193,
      "loss": 0.7906,
      "step": 890
    },
    {
      "epoch": 0.24,
      "grad_norm": 70.73613208149847,
      "learning_rate": 0.00028935729744989177,
      "loss": 0.6548,
      "step": 900
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.33701579186007,
      "learning_rate": 0.0002890749976475016,
      "loss": 0.7538,
      "step": 910
    },
    {
      "epoch": 0.25,
      "grad_norm": 29.626361182384734,
      "learning_rate": 0.0002887926978451115,
      "loss": 0.5978,
      "step": 920
    },
    {
      "epoch": 0.25,
      "grad_norm": 58.01553324325624,
      "learning_rate": 0.00028851039804272136,
      "loss": 0.6298,
      "step": 930
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.659591385752485,
      "learning_rate": 0.0002882280982403312,
      "loss": 0.5547,
      "step": 940
    },
    {
      "epoch": 0.26,
      "grad_norm": 26.934672364344866,
      "learning_rate": 0.00028794579843794104,
      "loss": 0.5404,
      "step": 950
    },
    {
      "epoch": 0.26,
      "grad_norm": 70.33972104319736,
      "learning_rate": 0.00028766349863555094,
      "loss": 0.677,
      "step": 960
    },
    {
      "epoch": 0.26,
      "grad_norm": 43.29567088425168,
      "learning_rate": 0.0002873811988331608,
      "loss": 0.5111,
      "step": 970
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.992791338572366,
      "learning_rate": 0.00028709889903077063,
      "loss": 0.5037,
      "step": 980
    },
    {
      "epoch": 0.27,
      "grad_norm": 45.97298343764695,
      "learning_rate": 0.0002868165992283805,
      "loss": 0.7216,
      "step": 990
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.275857535832048,
      "learning_rate": 0.00028653429942599037,
      "loss": 0.3847,
      "step": 1000
    },
    {
      "epoch": 0.27,
      "grad_norm": 15.945900534506368,
      "learning_rate": 0.0002862519996236002,
      "loss": 1.0541,
      "step": 1010
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.918283293645446,
      "learning_rate": 0.0002859696998212101,
      "loss": 0.8582,
      "step": 1020
    },
    {
      "epoch": 0.28,
      "grad_norm": 58.22357384010786,
      "learning_rate": 0.00028568740001881995,
      "loss": 0.5408,
      "step": 1030
    },
    {
      "epoch": 0.28,
      "grad_norm": 34.290390297053385,
      "learning_rate": 0.00028540510021642985,
      "loss": 0.5315,
      "step": 1040
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.46273507283339,
      "learning_rate": 0.0002851510303942787,
      "loss": 0.6619,
      "step": 1050
    },
    {
      "epoch": 0.29,
      "grad_norm": 25.235100211065358,
      "learning_rate": 0.00028486873059188855,
      "loss": 0.5833,
      "step": 1060
    },
    {
      "epoch": 0.29,
      "grad_norm": 28.310209119505107,
      "learning_rate": 0.0002845864307894984,
      "loss": 0.6314,
      "step": 1070
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.454002441026121,
      "learning_rate": 0.0002843041309871083,
      "loss": 0.4932,
      "step": 1080
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.558295891130298,
      "learning_rate": 0.00028402183118471814,
      "loss": 0.5807,
      "step": 1090
    },
    {
      "epoch": 0.3,
      "grad_norm": 37.16490696931356,
      "learning_rate": 0.000283739531382328,
      "loss": 0.5661,
      "step": 1100
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.416926050662129,
      "learning_rate": 0.0002834572315799379,
      "loss": 0.4828,
      "step": 1110
    },
    {
      "epoch": 0.3,
      "grad_norm": 40.84929632057358,
      "learning_rate": 0.0002831749317775477,
      "loss": 0.5759,
      "step": 1120
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.729765389172563,
      "learning_rate": 0.00028289263197515756,
      "loss": 0.5712,
      "step": 1130
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.739340981599153,
      "learning_rate": 0.00028261033217276746,
      "loss": 0.5936,
      "step": 1140
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.6404275687143137,
      "learning_rate": 0.0002823280323703773,
      "loss": 0.6137,
      "step": 1150
    },
    {
      "epoch": 0.31,
      "grad_norm": 68.94924662185701,
      "learning_rate": 0.0002820457325679872,
      "loss": 0.7875,
      "step": 1160
    },
    {
      "epoch": 0.32,
      "grad_norm": 32.70234559092242,
      "learning_rate": 0.00028176343276559705,
      "loss": 0.3881,
      "step": 1170
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.748347597119338,
      "learning_rate": 0.0002814811329632069,
      "loss": 0.4901,
      "step": 1180
    },
    {
      "epoch": 0.32,
      "grad_norm": 18.373813387693776,
      "learning_rate": 0.0002811988331608168,
      "loss": 0.6048,
      "step": 1190
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.5100824456412356,
      "learning_rate": 0.00028091653335842663,
      "loss": 0.5068,
      "step": 1200
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.701419733074948,
      "learning_rate": 0.0002806342335560365,
      "loss": 0.5004,
      "step": 1210
    },
    {
      "epoch": 0.33,
      "grad_norm": 21.967276394713913,
      "learning_rate": 0.00028035193375364637,
      "loss": 0.5339,
      "step": 1220
    },
    {
      "epoch": 0.33,
      "grad_norm": 32.47582792617236,
      "learning_rate": 0.0002800696339512562,
      "loss": 0.5762,
      "step": 1230
    },
    {
      "epoch": 0.33,
      "grad_norm": 43.31451808816033,
      "learning_rate": 0.00027978733414886606,
      "loss": 0.682,
      "step": 1240
    },
    {
      "epoch": 0.34,
      "grad_norm": 54.50955853256639,
      "learning_rate": 0.00027950503434647595,
      "loss": 0.5938,
      "step": 1250
    },
    {
      "epoch": 0.34,
      "grad_norm": 37.70165225384339,
      "learning_rate": 0.0002792227345440858,
      "loss": 0.6582,
      "step": 1260
    },
    {
      "epoch": 0.34,
      "grad_norm": 21.076452229345897,
      "learning_rate": 0.00027894043474169564,
      "loss": 0.6169,
      "step": 1270
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.722549567290084,
      "learning_rate": 0.0002786581349393055,
      "loss": 0.777,
      "step": 1280
    },
    {
      "epoch": 0.35,
      "grad_norm": 40.43201810431064,
      "learning_rate": 0.0002783758351369154,
      "loss": 0.6554,
      "step": 1290
    },
    {
      "epoch": 0.35,
      "grad_norm": 42.954671812115336,
      "learning_rate": 0.0002780935353345252,
      "loss": 0.453,
      "step": 1300
    },
    {
      "epoch": 0.35,
      "grad_norm": 38.30303626729069,
      "learning_rate": 0.00027781123553213507,
      "loss": 0.5006,
      "step": 1310
    },
    {
      "epoch": 0.36,
      "grad_norm": 76.05172144078296,
      "learning_rate": 0.00027752893572974497,
      "loss": 0.7633,
      "step": 1320
    },
    {
      "epoch": 0.36,
      "grad_norm": 29.30031593181108,
      "learning_rate": 0.0002772466359273548,
      "loss": 0.7287,
      "step": 1330
    },
    {
      "epoch": 0.36,
      "grad_norm": 40.63825485441602,
      "learning_rate": 0.0002769643361249647,
      "loss": 0.8624,
      "step": 1340
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.002194315322271,
      "learning_rate": 0.00027668203632257455,
      "loss": 0.5755,
      "step": 1350
    },
    {
      "epoch": 0.37,
      "grad_norm": 44.20739516394715,
      "learning_rate": 0.0002763997365201844,
      "loss": 0.5234,
      "step": 1360
    },
    {
      "epoch": 0.37,
      "grad_norm": 30.526565399058487,
      "learning_rate": 0.0002761174367177943,
      "loss": 0.6273,
      "step": 1370
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.237437233822316,
      "learning_rate": 0.00027583513691540414,
      "loss": 0.4537,
      "step": 1380
    },
    {
      "epoch": 0.37,
      "grad_norm": 13.755900833162736,
      "learning_rate": 0.000275552837113014,
      "loss": 0.7142,
      "step": 1390
    },
    {
      "epoch": 0.38,
      "grad_norm": 35.213190247893735,
      "learning_rate": 0.0002752705373106239,
      "loss": 0.619,
      "step": 1400
    },
    {
      "epoch": 0.38,
      "grad_norm": 14.758426523233009,
      "learning_rate": 0.0002749882375082337,
      "loss": 0.5566,
      "step": 1410
    },
    {
      "epoch": 0.38,
      "grad_norm": 18.098771234867144,
      "learning_rate": 0.00027470593770584356,
      "loss": 0.6159,
      "step": 1420
    },
    {
      "epoch": 0.39,
      "grad_norm": 35.96624592595499,
      "learning_rate": 0.00027442363790345346,
      "loss": 0.6237,
      "step": 1430
    },
    {
      "epoch": 0.39,
      "grad_norm": 45.89391430940064,
      "learning_rate": 0.0002741413381010633,
      "loss": 0.6299,
      "step": 1440
    },
    {
      "epoch": 0.39,
      "grad_norm": 13.020139306899988,
      "learning_rate": 0.0002738590382986732,
      "loss": 0.575,
      "step": 1450
    },
    {
      "epoch": 0.39,
      "grad_norm": 34.0665798954868,
      "learning_rate": 0.00027357673849628304,
      "loss": 0.6157,
      "step": 1460
    },
    {
      "epoch": 0.4,
      "grad_norm": 35.35784222525188,
      "learning_rate": 0.0002732944386938929,
      "loss": 0.6084,
      "step": 1470
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.6825952665819144,
      "learning_rate": 0.0002730121388915028,
      "loss": 0.6866,
      "step": 1480
    },
    {
      "epoch": 0.4,
      "grad_norm": 48.967250824761386,
      "learning_rate": 0.0002727298390891126,
      "loss": 0.5778,
      "step": 1490
    },
    {
      "epoch": 0.4,
      "grad_norm": 27.41415419839072,
      "learning_rate": 0.00027244753928672247,
      "loss": 0.521,
      "step": 1500
    },
    {
      "epoch": 0.41,
      "grad_norm": 14.714522153936564,
      "learning_rate": 0.00027216523948433237,
      "loss": 0.4065,
      "step": 1510
    },
    {
      "epoch": 0.41,
      "grad_norm": 44.1855710811133,
      "learning_rate": 0.0002718829396819422,
      "loss": 0.4864,
      "step": 1520
    },
    {
      "epoch": 0.41,
      "grad_norm": 19.658074778115573,
      "learning_rate": 0.00027160063987955206,
      "loss": 0.4207,
      "step": 1530
    },
    {
      "epoch": 0.42,
      "grad_norm": 65.3437140545411,
      "learning_rate": 0.0002713183400771619,
      "loss": 0.5098,
      "step": 1540
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.978161425197959,
      "learning_rate": 0.0002710360402747718,
      "loss": 0.6814,
      "step": 1550
    },
    {
      "epoch": 0.42,
      "grad_norm": 29.551168017935446,
      "learning_rate": 0.00027075374047238164,
      "loss": 0.5473,
      "step": 1560
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.735744562123976,
      "learning_rate": 0.0002704714406699915,
      "loss": 0.5076,
      "step": 1570
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.5733353693637815,
      "learning_rate": 0.0002701891408676014,
      "loss": 0.4966,
      "step": 1580
    },
    {
      "epoch": 0.43,
      "grad_norm": 17.567836646615046,
      "learning_rate": 0.0002699068410652112,
      "loss": 0.4888,
      "step": 1590
    },
    {
      "epoch": 0.43,
      "grad_norm": 89.88486928165655,
      "learning_rate": 0.00026962454126282107,
      "loss": 0.6339,
      "step": 1600
    },
    {
      "epoch": 0.43,
      "grad_norm": 39.414988439210845,
      "learning_rate": 0.00026934224146043097,
      "loss": 0.4998,
      "step": 1610
    },
    {
      "epoch": 0.44,
      "grad_norm": 57.601573172924894,
      "learning_rate": 0.0002690599416580408,
      "loss": 0.4137,
      "step": 1620
    },
    {
      "epoch": 0.44,
      "grad_norm": 18.115275800302854,
      "learning_rate": 0.0002687776418556507,
      "loss": 0.5152,
      "step": 1630
    },
    {
      "epoch": 0.44,
      "grad_norm": 26.50422470278149,
      "learning_rate": 0.00026849534205326055,
      "loss": 0.5487,
      "step": 1640
    },
    {
      "epoch": 0.44,
      "grad_norm": 45.208664041143116,
      "learning_rate": 0.0002682130422508704,
      "loss": 0.6563,
      "step": 1650
    },
    {
      "epoch": 0.45,
      "grad_norm": 39.88684347407983,
      "learning_rate": 0.0002679307424484803,
      "loss": 0.7426,
      "step": 1660
    },
    {
      "epoch": 0.45,
      "grad_norm": 26.768127312187946,
      "learning_rate": 0.00026764844264609013,
      "loss": 0.4631,
      "step": 1670
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.3762747340768686,
      "learning_rate": 0.0002673661428437,
      "loss": 0.4376,
      "step": 1680
    },
    {
      "epoch": 0.46,
      "grad_norm": 21.337202202627296,
      "learning_rate": 0.0002670838430413099,
      "loss": 0.523,
      "step": 1690
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.747310215630781,
      "learning_rate": 0.0002668015432389197,
      "loss": 0.5522,
      "step": 1700
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.308879623704246,
      "learning_rate": 0.00026651924343652956,
      "loss": 0.6139,
      "step": 1710
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.500672260509509,
      "learning_rate": 0.00026623694363413946,
      "loss": 0.5575,
      "step": 1720
    },
    {
      "epoch": 0.47,
      "grad_norm": 66.13087761533203,
      "learning_rate": 0.0002659546438317493,
      "loss": 0.5881,
      "step": 1730
    },
    {
      "epoch": 0.47,
      "grad_norm": 21.242949867036717,
      "learning_rate": 0.00026567234402935915,
      "loss": 0.4831,
      "step": 1740
    },
    {
      "epoch": 0.47,
      "grad_norm": 38.66795875583099,
      "learning_rate": 0.000265390044226969,
      "loss": 0.4735,
      "step": 1750
    },
    {
      "epoch": 0.47,
      "grad_norm": 52.36083685332038,
      "learning_rate": 0.0002651077444245789,
      "loss": 0.6003,
      "step": 1760
    },
    {
      "epoch": 0.48,
      "grad_norm": 55.9483007403524,
      "learning_rate": 0.00026482544462218873,
      "loss": 0.5463,
      "step": 1770
    },
    {
      "epoch": 0.48,
      "grad_norm": 26.430022391481817,
      "learning_rate": 0.0002645431448197986,
      "loss": 0.55,
      "step": 1780
    },
    {
      "epoch": 0.48,
      "grad_norm": 22.993806310835144,
      "learning_rate": 0.00026426084501740847,
      "loss": 0.5551,
      "step": 1790
    },
    {
      "epoch": 0.49,
      "grad_norm": 219.9516455829629,
      "learning_rate": 0.0002639785452150183,
      "loss": 0.9555,
      "step": 1800
    },
    {
      "epoch": 0.49,
      "grad_norm": 43.05528709341598,
      "learning_rate": 0.0002636962454126282,
      "loss": 0.5683,
      "step": 1810
    },
    {
      "epoch": 0.49,
      "grad_norm": 16.06539185630662,
      "learning_rate": 0.00026341394561023806,
      "loss": 0.5561,
      "step": 1820
    },
    {
      "epoch": 0.49,
      "grad_norm": 19.794346343708504,
      "learning_rate": 0.0002631316458078479,
      "loss": 0.4629,
      "step": 1830
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.846653683897656,
      "learning_rate": 0.0002628493460054578,
      "loss": 0.4551,
      "step": 1840
    },
    {
      "epoch": 0.5,
      "grad_norm": 48.04207537427032,
      "learning_rate": 0.00026256704620306764,
      "loss": 0.6367,
      "step": 1850
    },
    {
      "epoch": 0.5,
      "grad_norm": 21.996732095876606,
      "learning_rate": 0.0002622847464006775,
      "loss": 0.4721,
      "step": 1860
    },
    {
      "epoch": 0.5,
      "grad_norm": 47.300896640277394,
      "learning_rate": 0.0002620024465982874,
      "loss": 0.733,
      "step": 1870
    },
    {
      "epoch": 0.51,
      "grad_norm": 62.87894525246514,
      "learning_rate": 0.0002617201467958972,
      "loss": 0.631,
      "step": 1880
    },
    {
      "epoch": 0.51,
      "grad_norm": 39.98833872269375,
      "learning_rate": 0.00026143784699350707,
      "loss": 0.5335,
      "step": 1890
    },
    {
      "epoch": 0.51,
      "grad_norm": 53.44105630496144,
      "learning_rate": 0.00026115554719111696,
      "loss": 0.5746,
      "step": 1900
    },
    {
      "epoch": 0.51,
      "grad_norm": 31.635744539317592,
      "learning_rate": 0.0002608732473887268,
      "loss": 0.5407,
      "step": 1910
    },
    {
      "epoch": 0.52,
      "grad_norm": 36.98173637530453,
      "learning_rate": 0.00026059094758633665,
      "loss": 0.5856,
      "step": 1920
    },
    {
      "epoch": 0.52,
      "grad_norm": 45.62039990848383,
      "learning_rate": 0.00026030864778394655,
      "loss": 0.5987,
      "step": 1930
    },
    {
      "epoch": 0.52,
      "grad_norm": 14.24680469437125,
      "learning_rate": 0.0002600263479815564,
      "loss": 0.5596,
      "step": 1940
    },
    {
      "epoch": 0.53,
      "grad_norm": 29.939256432650087,
      "learning_rate": 0.0002597440481791663,
      "loss": 0.5826,
      "step": 1950
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.43619439588576,
      "learning_rate": 0.0002594617483767761,
      "loss": 0.5803,
      "step": 1960
    },
    {
      "epoch": 0.53,
      "grad_norm": 20.70533577523271,
      "learning_rate": 0.000259179448574386,
      "loss": 0.5009,
      "step": 1970
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.079440225683143,
      "learning_rate": 0.0002588971487719958,
      "loss": 0.4532,
      "step": 1980
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.146249120905656,
      "learning_rate": 0.0002586148489696057,
      "loss": 0.5114,
      "step": 1990
    },
    {
      "epoch": 0.54,
      "grad_norm": 19.05852118167717,
      "learning_rate": 0.00025833254916721556,
      "loss": 0.5252,
      "step": 2000
    },
    {
      "epoch": 0.54,
      "grad_norm": 20.938001393961944,
      "learning_rate": 0.0002580502493648254,
      "loss": 0.4992,
      "step": 2010
    },
    {
      "epoch": 0.54,
      "grad_norm": 22.266428885420456,
      "learning_rate": 0.0002577679495624353,
      "loss": 0.58,
      "step": 2020
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.083866628110543,
      "learning_rate": 0.00025748564976004514,
      "loss": 0.5257,
      "step": 2030
    },
    {
      "epoch": 0.55,
      "grad_norm": 22.90756039299561,
      "learning_rate": 0.000257203349957655,
      "loss": 0.5205,
      "step": 2040
    },
    {
      "epoch": 0.55,
      "grad_norm": 39.7571318737366,
      "learning_rate": 0.0002569210501552649,
      "loss": 0.6173,
      "step": 2050
    },
    {
      "epoch": 0.56,
      "grad_norm": 15.947362574897587,
      "learning_rate": 0.00025663875035287473,
      "loss": 0.6033,
      "step": 2060
    },
    {
      "epoch": 0.56,
      "grad_norm": 53.86075106280294,
      "learning_rate": 0.00025635645055048457,
      "loss": 0.5627,
      "step": 2070
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.830785646294771,
      "learning_rate": 0.00025607415074809447,
      "loss": 0.5465,
      "step": 2080
    },
    {
      "epoch": 0.56,
      "grad_norm": 59.61472265571748,
      "learning_rate": 0.0002557918509457043,
      "loss": 0.687,
      "step": 2090
    },
    {
      "epoch": 0.57,
      "grad_norm": 42.29133499709177,
      "learning_rate": 0.00025550955114331416,
      "loss": 0.6127,
      "step": 2100
    },
    {
      "epoch": 0.57,
      "grad_norm": 21.40899316029326,
      "learning_rate": 0.00025522725134092405,
      "loss": 0.6594,
      "step": 2110
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.418629701350268,
      "learning_rate": 0.0002549449515385339,
      "loss": 0.509,
      "step": 2120
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.77083728366076,
      "learning_rate": 0.0002546626517361438,
      "loss": 0.4877,
      "step": 2130
    },
    {
      "epoch": 0.58,
      "grad_norm": 47.115826884545186,
      "learning_rate": 0.00025438035193375364,
      "loss": 0.526,
      "step": 2140
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.220656392621686,
      "learning_rate": 0.0002540980521313635,
      "loss": 0.5222,
      "step": 2150
    },
    {
      "epoch": 0.58,
      "grad_norm": 20.571632448848582,
      "learning_rate": 0.0002538157523289734,
      "loss": 0.5796,
      "step": 2160
    },
    {
      "epoch": 0.59,
      "grad_norm": 32.44080764383593,
      "learning_rate": 0.0002535334525265832,
      "loss": 0.6216,
      "step": 2170
    },
    {
      "epoch": 0.59,
      "grad_norm": 39.46650533643509,
      "learning_rate": 0.00025325115272419307,
      "loss": 0.5771,
      "step": 2180
    },
    {
      "epoch": 0.59,
      "grad_norm": 28.74403730502063,
      "learning_rate": 0.0002529688529218029,
      "loss": 0.5434,
      "step": 2190
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.959108093511947,
      "learning_rate": 0.0002526865531194128,
      "loss": 0.4837,
      "step": 2200
    },
    {
      "epoch": 0.6,
      "grad_norm": 91.00874953925927,
      "learning_rate": 0.00025240425331702265,
      "loss": 0.5948,
      "step": 2210
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.897584719629194,
      "learning_rate": 0.0002521219535146325,
      "loss": 0.4972,
      "step": 2220
    },
    {
      "epoch": 0.6,
      "grad_norm": 26.448172430126682,
      "learning_rate": 0.0002518396537122424,
      "loss": 0.5818,
      "step": 2230
    },
    {
      "epoch": 0.6,
      "grad_norm": 13.3293846128825,
      "learning_rate": 0.00025155735390985223,
      "loss": 0.6018,
      "step": 2240
    },
    {
      "epoch": 0.61,
      "grad_norm": 29.141684420480527,
      "learning_rate": 0.0002512750541074621,
      "loss": 0.4841,
      "step": 2250
    },
    {
      "epoch": 0.61,
      "grad_norm": 36.62853639010538,
      "learning_rate": 0.000250992754305072,
      "loss": 0.629,
      "step": 2260
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.855223640020943,
      "learning_rate": 0.0002507104545026818,
      "loss": 0.583,
      "step": 2270
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.516113386443599,
      "learning_rate": 0.00025042815470029166,
      "loss": 0.5286,
      "step": 2280
    },
    {
      "epoch": 0.62,
      "grad_norm": 30.519251628675278,
      "learning_rate": 0.00025014585489790156,
      "loss": 0.5741,
      "step": 2290
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.711440527551325,
      "learning_rate": 0.0002498635550955114,
      "loss": 0.5283,
      "step": 2300
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.443916115372712,
      "learning_rate": 0.0002495812552931213,
      "loss": 0.5528,
      "step": 2310
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.600320421125944,
      "learning_rate": 0.00024929895549073114,
      "loss": 0.4525,
      "step": 2320
    },
    {
      "epoch": 0.63,
      "grad_norm": 19.63566312120888,
      "learning_rate": 0.000249016655688341,
      "loss": 0.4591,
      "step": 2330
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.936703576655736,
      "learning_rate": 0.0002487343558859509,
      "loss": 0.5013,
      "step": 2340
    },
    {
      "epoch": 0.63,
      "grad_norm": 46.3722085517493,
      "learning_rate": 0.00024845205608356073,
      "loss": 0.5901,
      "step": 2350
    },
    {
      "epoch": 0.64,
      "grad_norm": 34.9086060236879,
      "learning_rate": 0.00024816975628117057,
      "loss": 0.4417,
      "step": 2360
    },
    {
      "epoch": 0.64,
      "grad_norm": 20.096315471348607,
      "learning_rate": 0.00024788745647878047,
      "loss": 0.5316,
      "step": 2370
    },
    {
      "epoch": 0.64,
      "grad_norm": 60.70601731360524,
      "learning_rate": 0.0002476051566763903,
      "loss": 0.5035,
      "step": 2380
    },
    {
      "epoch": 0.64,
      "grad_norm": 32.330129190235354,
      "learning_rate": 0.00024732285687400016,
      "loss": 0.5351,
      "step": 2390
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.143002412862536,
      "learning_rate": 0.00024704055707161,
      "loss": 0.6104,
      "step": 2400
    },
    {
      "epoch": 0.65,
      "grad_norm": 44.40440422030029,
      "learning_rate": 0.0002467582572692199,
      "loss": 0.6068,
      "step": 2410
    },
    {
      "epoch": 0.65,
      "grad_norm": 23.112171947355797,
      "learning_rate": 0.00024647595746682974,
      "loss": 0.4759,
      "step": 2420
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.724539286497222,
      "learning_rate": 0.0002461936576644396,
      "loss": 0.4302,
      "step": 2430
    },
    {
      "epoch": 0.66,
      "grad_norm": 64.85791866484311,
      "learning_rate": 0.0002459395878422885,
      "loss": 0.6234,
      "step": 2440
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.49390237056048,
      "learning_rate": 0.00024565728803989834,
      "loss": 0.5557,
      "step": 2450
    },
    {
      "epoch": 0.66,
      "grad_norm": 50.562515404961765,
      "learning_rate": 0.00024537498823750824,
      "loss": 0.4449,
      "step": 2460
    },
    {
      "epoch": 0.67,
      "grad_norm": 15.302744860423378,
      "learning_rate": 0.0002450926884351181,
      "loss": 0.5069,
      "step": 2470
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.493116125344576,
      "learning_rate": 0.0002448103886327279,
      "loss": 0.6081,
      "step": 2480
    },
    {
      "epoch": 0.67,
      "grad_norm": 19.050182771052356,
      "learning_rate": 0.0002445280888303378,
      "loss": 0.5239,
      "step": 2490
    },
    {
      "epoch": 0.67,
      "grad_norm": 13.627296885194582,
      "learning_rate": 0.00024424578902794766,
      "loss": 0.481,
      "step": 2500
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.483322705105648,
      "learning_rate": 0.0002439634892255575,
      "loss": 0.3953,
      "step": 2510
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.768444852192465,
      "learning_rate": 0.00024368118942316738,
      "loss": 0.4733,
      "step": 2520
    },
    {
      "epoch": 0.68,
      "grad_norm": 35.65609400829592,
      "learning_rate": 0.00024339888962077725,
      "loss": 0.6246,
      "step": 2530
    },
    {
      "epoch": 0.68,
      "grad_norm": 40.16798270037108,
      "learning_rate": 0.00024311658981838712,
      "loss": 0.512,
      "step": 2540
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.61798386242861,
      "learning_rate": 0.00024283429001599696,
      "loss": 0.4495,
      "step": 2550
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.7610719088004965,
      "learning_rate": 0.00024255199021360683,
      "loss": 0.5017,
      "step": 2560
    },
    {
      "epoch": 0.69,
      "grad_norm": 15.991254478220595,
      "learning_rate": 0.0002422696904112167,
      "loss": 0.4338,
      "step": 2570
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.072734735032982,
      "learning_rate": 0.00024198739060882655,
      "loss": 0.3493,
      "step": 2580
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.858973490544457,
      "learning_rate": 0.00024170509080643642,
      "loss": 0.4524,
      "step": 2590
    },
    {
      "epoch": 0.7,
      "grad_norm": 31.866877217429344,
      "learning_rate": 0.0002414227910040463,
      "loss": 0.4569,
      "step": 2600
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.281099421825626,
      "learning_rate": 0.00024114049120165616,
      "loss": 0.5973,
      "step": 2610
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.2464272628586,
      "learning_rate": 0.00024085819139926597,
      "loss": 0.4841,
      "step": 2620
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.510441231756808,
      "learning_rate": 0.00024057589159687587,
      "loss": 0.5294,
      "step": 2630
    },
    {
      "epoch": 0.71,
      "grad_norm": 19.07360782080447,
      "learning_rate": 0.00024029359179448574,
      "loss": 0.3945,
      "step": 2640
    },
    {
      "epoch": 0.71,
      "grad_norm": 41.162407170640506,
      "learning_rate": 0.0002400112919920956,
      "loss": 0.5966,
      "step": 2650
    },
    {
      "epoch": 0.72,
      "grad_norm": 27.55974785329981,
      "learning_rate": 0.00023972899218970543,
      "loss": 0.4966,
      "step": 2660
    },
    {
      "epoch": 0.72,
      "grad_norm": 28.88414274391822,
      "learning_rate": 0.0002394466923873153,
      "loss": 0.5603,
      "step": 2670
    },
    {
      "epoch": 0.72,
      "grad_norm": 52.958585988269164,
      "learning_rate": 0.00023916439258492517,
      "loss": 0.3952,
      "step": 2680
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.965132773766056,
      "learning_rate": 0.000238882092782535,
      "loss": 0.4739,
      "step": 2690
    },
    {
      "epoch": 0.73,
      "grad_norm": 35.65250451570077,
      "learning_rate": 0.00023859979298014488,
      "loss": 0.4118,
      "step": 2700
    },
    {
      "epoch": 0.73,
      "grad_norm": 48.0346692761587,
      "learning_rate": 0.00023831749317775475,
      "loss": 0.6258,
      "step": 2710
    },
    {
      "epoch": 0.73,
      "grad_norm": 13.281827436123862,
      "learning_rate": 0.00023803519337536462,
      "loss": 0.5371,
      "step": 2720
    },
    {
      "epoch": 0.74,
      "grad_norm": 61.02607180440632,
      "learning_rate": 0.00023775289357297447,
      "loss": 0.6129,
      "step": 2730
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.13395897869061,
      "learning_rate": 0.00023747059377058434,
      "loss": 0.5587,
      "step": 2740
    },
    {
      "epoch": 0.74,
      "grad_norm": 73.12168536498893,
      "learning_rate": 0.0002371882939681942,
      "loss": 0.453,
      "step": 2750
    },
    {
      "epoch": 0.74,
      "grad_norm": 16.538548108702784,
      "learning_rate": 0.00023690599416580405,
      "loss": 0.4789,
      "step": 2760
    },
    {
      "epoch": 0.75,
      "grad_norm": 28.124938145675458,
      "learning_rate": 0.00023662369436341392,
      "loss": 0.5079,
      "step": 2770
    },
    {
      "epoch": 0.75,
      "grad_norm": 23.4746744989872,
      "learning_rate": 0.0002363413945610238,
      "loss": 0.4047,
      "step": 2780
    },
    {
      "epoch": 0.75,
      "grad_norm": 18.173919605760766,
      "learning_rate": 0.00023605909475863366,
      "loss": 0.6889,
      "step": 2790
    },
    {
      "epoch": 0.75,
      "grad_norm": 32.766708630791726,
      "learning_rate": 0.0002357767949562435,
      "loss": 0.5234,
      "step": 2800
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.771456385962614,
      "learning_rate": 0.00023549449515385338,
      "loss": 0.4433,
      "step": 2810
    },
    {
      "epoch": 0.76,
      "grad_norm": 24.227985900137398,
      "learning_rate": 0.00023521219535146325,
      "loss": 0.5761,
      "step": 2820
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.064205736739243,
      "learning_rate": 0.00023492989554907312,
      "loss": 0.5549,
      "step": 2830
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.197989873334812,
      "learning_rate": 0.00023464759574668296,
      "loss": 0.5464,
      "step": 2840
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.777143089321015,
      "learning_rate": 0.00023436529594429283,
      "loss": 0.5405,
      "step": 2850
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.993093177210862,
      "learning_rate": 0.0002340829961419027,
      "loss": 0.5582,
      "step": 2860
    },
    {
      "epoch": 0.77,
      "grad_norm": 24.49648633598827,
      "learning_rate": 0.00023380069633951252,
      "loss": 0.5138,
      "step": 2870
    },
    {
      "epoch": 0.78,
      "grad_norm": 22.747725947613564,
      "learning_rate": 0.0002335183965371224,
      "loss": 0.5351,
      "step": 2880
    },
    {
      "epoch": 0.78,
      "grad_norm": 26.114281676707943,
      "learning_rate": 0.00023323609673473226,
      "loss": 0.5259,
      "step": 2890
    },
    {
      "epoch": 0.78,
      "grad_norm": 24.485686024254516,
      "learning_rate": 0.00023295379693234213,
      "loss": 0.5822,
      "step": 2900
    },
    {
      "epoch": 0.78,
      "grad_norm": 19.102514950579224,
      "learning_rate": 0.00023267149712995197,
      "loss": 0.4981,
      "step": 2910
    },
    {
      "epoch": 0.79,
      "grad_norm": 27.84303837045006,
      "learning_rate": 0.00023238919732756184,
      "loss": 0.5566,
      "step": 2920
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.463309351436824,
      "learning_rate": 0.00023210689752517171,
      "loss": 0.4877,
      "step": 2930
    },
    {
      "epoch": 0.79,
      "grad_norm": 20.889386187052988,
      "learning_rate": 0.00023182459772278156,
      "loss": 0.571,
      "step": 2940
    },
    {
      "epoch": 0.8,
      "grad_norm": 27.12643907806873,
      "learning_rate": 0.00023154229792039143,
      "loss": 0.3906,
      "step": 2950
    },
    {
      "epoch": 0.8,
      "grad_norm": 31.178049488478973,
      "learning_rate": 0.0002312599981180013,
      "loss": 0.3784,
      "step": 2960
    },
    {
      "epoch": 0.8,
      "grad_norm": 21.205595401056414,
      "learning_rate": 0.00023097769831561117,
      "loss": 0.5068,
      "step": 2970
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.524741356572266,
      "learning_rate": 0.000230695398513221,
      "loss": 0.6264,
      "step": 2980
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.518574301336548,
      "learning_rate": 0.00023041309871083088,
      "loss": 0.2793,
      "step": 2990
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.752892823953365,
      "learning_rate": 0.00023013079890844075,
      "loss": 0.6417,
      "step": 3000
    },
    {
      "epoch": 0.81,
      "grad_norm": 26.875277065911504,
      "learning_rate": 0.00022984849910605062,
      "loss": 0.4734,
      "step": 3010
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.34128278280261,
      "learning_rate": 0.00022956619930366047,
      "loss": 0.4454,
      "step": 3020
    },
    {
      "epoch": 0.82,
      "grad_norm": 37.34550332777604,
      "learning_rate": 0.00022928389950127034,
      "loss": 0.4723,
      "step": 3030
    },
    {
      "epoch": 0.82,
      "grad_norm": 16.836025874620635,
      "learning_rate": 0.0002290015996988802,
      "loss": 0.4903,
      "step": 3040
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.895075722537825,
      "learning_rate": 0.00022871929989649005,
      "loss": 0.4938,
      "step": 3050
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.36322044053127,
      "learning_rate": 0.00022843700009409992,
      "loss": 0.3442,
      "step": 3060
    },
    {
      "epoch": 0.83,
      "grad_norm": 20.92716913799448,
      "learning_rate": 0.0002281547002917098,
      "loss": 0.5994,
      "step": 3070
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.543034656249663,
      "learning_rate": 0.00022787240048931966,
      "loss": 0.5742,
      "step": 3080
    },
    {
      "epoch": 0.83,
      "grad_norm": 15.171831604318992,
      "learning_rate": 0.00022759010068692948,
      "loss": 0.5666,
      "step": 3090
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.518749953986788,
      "learning_rate": 0.00022730780088453935,
      "loss": 0.5751,
      "step": 3100
    },
    {
      "epoch": 0.84,
      "grad_norm": 35.31868287980286,
      "learning_rate": 0.00022702550108214922,
      "loss": 0.4429,
      "step": 3110
    },
    {
      "epoch": 0.84,
      "grad_norm": 40.67908068497924,
      "learning_rate": 0.00022674320127975906,
      "loss": 0.4167,
      "step": 3120
    },
    {
      "epoch": 0.84,
      "grad_norm": 13.955948978130767,
      "learning_rate": 0.00022646090147736893,
      "loss": 0.4989,
      "step": 3130
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.558575176106113,
      "learning_rate": 0.0002261786016749788,
      "loss": 0.4525,
      "step": 3140
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.957971500650416,
      "learning_rate": 0.00022589630187258867,
      "loss": 0.4706,
      "step": 3150
    },
    {
      "epoch": 0.85,
      "grad_norm": 6.657435720682589,
      "learning_rate": 0.00022561400207019852,
      "loss": 0.3944,
      "step": 3160
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.705045715256741,
      "learning_rate": 0.0002253317022678084,
      "loss": 0.3066,
      "step": 3170
    },
    {
      "epoch": 0.86,
      "grad_norm": 23.69751014605326,
      "learning_rate": 0.00022504940246541826,
      "loss": 0.5447,
      "step": 3180
    },
    {
      "epoch": 0.86,
      "grad_norm": 19.684812167589943,
      "learning_rate": 0.00022476710266302813,
      "loss": 0.4136,
      "step": 3190
    },
    {
      "epoch": 0.86,
      "grad_norm": 49.03435946136028,
      "learning_rate": 0.00022448480286063797,
      "loss": 0.4373,
      "step": 3200
    },
    {
      "epoch": 0.87,
      "grad_norm": 31.906403166639613,
      "learning_rate": 0.00022420250305824784,
      "loss": 0.3615,
      "step": 3210
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.81829589202541,
      "learning_rate": 0.0002239202032558577,
      "loss": 0.3685,
      "step": 3220
    },
    {
      "epoch": 0.87,
      "grad_norm": 20.35356426882955,
      "learning_rate": 0.00022363790345346756,
      "loss": 0.4436,
      "step": 3230
    },
    {
      "epoch": 0.87,
      "grad_norm": 41.06448966297969,
      "learning_rate": 0.00022335560365107743,
      "loss": 0.4112,
      "step": 3240
    },
    {
      "epoch": 0.88,
      "grad_norm": 20.853609987184576,
      "learning_rate": 0.0002230733038486873,
      "loss": 0.579,
      "step": 3250
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.893418603930167,
      "learning_rate": 0.00022279100404629717,
      "loss": 0.5,
      "step": 3260
    },
    {
      "epoch": 0.88,
      "grad_norm": 39.96437378108241,
      "learning_rate": 0.000222508704243907,
      "loss": 0.4487,
      "step": 3270
    },
    {
      "epoch": 0.88,
      "grad_norm": 48.18863050703126,
      "learning_rate": 0.00022222640444151688,
      "loss": 0.6208,
      "step": 3280
    },
    {
      "epoch": 0.89,
      "grad_norm": 15.927049097256596,
      "learning_rate": 0.00022194410463912675,
      "loss": 0.4545,
      "step": 3290
    },
    {
      "epoch": 0.89,
      "grad_norm": 34.28118764598674,
      "learning_rate": 0.00022166180483673657,
      "loss": 0.3951,
      "step": 3300
    },
    {
      "epoch": 0.89,
      "grad_norm": 27.041531197429492,
      "learning_rate": 0.00022137950503434644,
      "loss": 0.5157,
      "step": 3310
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.40760696671216,
      "learning_rate": 0.0002210972052319563,
      "loss": 0.5265,
      "step": 3320
    },
    {
      "epoch": 0.9,
      "grad_norm": 48.43448861174829,
      "learning_rate": 0.00022081490542956618,
      "loss": 0.5036,
      "step": 3330
    },
    {
      "epoch": 0.9,
      "grad_norm": 22.37243486863623,
      "learning_rate": 0.00022053260562717602,
      "loss": 0.4536,
      "step": 3340
    },
    {
      "epoch": 0.9,
      "grad_norm": 43.916675619502065,
      "learning_rate": 0.0002202503058247859,
      "loss": 0.6473,
      "step": 3350
    },
    {
      "epoch": 0.91,
      "grad_norm": 25.389478172812115,
      "learning_rate": 0.00021996800602239576,
      "loss": 0.5327,
      "step": 3360
    },
    {
      "epoch": 0.91,
      "grad_norm": 15.905278510646042,
      "learning_rate": 0.00021968570622000563,
      "loss": 0.5316,
      "step": 3370
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.140199700759731,
      "learning_rate": 0.00021940340641761548,
      "loss": 0.4513,
      "step": 3380
    },
    {
      "epoch": 0.91,
      "grad_norm": 18.93781862073494,
      "learning_rate": 0.00021912110661522535,
      "loss": 0.4359,
      "step": 3390
    },
    {
      "epoch": 0.92,
      "grad_norm": 27.379219882282815,
      "learning_rate": 0.00021883880681283522,
      "loss": 0.3302,
      "step": 3400
    },
    {
      "epoch": 0.92,
      "grad_norm": 52.921089927902855,
      "learning_rate": 0.00021855650701044506,
      "loss": 0.4667,
      "step": 3410
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.758973583956582,
      "learning_rate": 0.00021827420720805493,
      "loss": 0.4258,
      "step": 3420
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.244405684031436,
      "learning_rate": 0.0002179919074056648,
      "loss": 0.451,
      "step": 3430
    },
    {
      "epoch": 0.93,
      "grad_norm": 13.187290890276453,
      "learning_rate": 0.00021770960760327467,
      "loss": 0.5306,
      "step": 3440
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.301211124449713,
      "learning_rate": 0.00021742730780088452,
      "loss": 0.4122,
      "step": 3450
    },
    {
      "epoch": 0.93,
      "grad_norm": 16.537461968750883,
      "learning_rate": 0.00021714500799849439,
      "loss": 0.4667,
      "step": 3460
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.36697928133076,
      "learning_rate": 0.00021686270819610426,
      "loss": 0.4962,
      "step": 3470
    },
    {
      "epoch": 0.94,
      "grad_norm": 39.338567481816796,
      "learning_rate": 0.0002165804083937141,
      "loss": 0.7026,
      "step": 3480
    },
    {
      "epoch": 0.94,
      "grad_norm": 18.045014600208976,
      "learning_rate": 0.00021629810859132397,
      "loss": 0.4879,
      "step": 3490
    },
    {
      "epoch": 0.94,
      "grad_norm": 5.914601035649798,
      "learning_rate": 0.00021601580878893384,
      "loss": 0.4656,
      "step": 3500
    },
    {
      "epoch": 0.95,
      "grad_norm": 3.2747671876645117,
      "learning_rate": 0.0002157335089865437,
      "loss": 0.4091,
      "step": 3510
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.369017475341312,
      "learning_rate": 0.00021545120918415353,
      "loss": 0.3972,
      "step": 3520
    },
    {
      "epoch": 0.95,
      "grad_norm": 50.498242062892444,
      "learning_rate": 0.0002151689093817634,
      "loss": 0.4414,
      "step": 3530
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.575171913067804,
      "learning_rate": 0.00021488660957937327,
      "loss": 0.3953,
      "step": 3540
    },
    {
      "epoch": 0.96,
      "grad_norm": 25.416922013213053,
      "learning_rate": 0.00021460430977698314,
      "loss": 0.5057,
      "step": 3550
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.214886929537651,
      "learning_rate": 0.00021432200997459298,
      "loss": 0.4496,
      "step": 3560
    },
    {
      "epoch": 0.96,
      "grad_norm": 3.440556572028814,
      "learning_rate": 0.00021403971017220285,
      "loss": 0.3442,
      "step": 3570
    },
    {
      "epoch": 0.97,
      "grad_norm": 5.7528045411129725,
      "learning_rate": 0.00021375741036981272,
      "loss": 0.4982,
      "step": 3580
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.393543868807688,
      "learning_rate": 0.00021347511056742257,
      "loss": 0.4556,
      "step": 3590
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.852010811157858,
      "learning_rate": 0.00021319281076503244,
      "loss": 0.4537,
      "step": 3600
    },
    {
      "epoch": 0.97,
      "grad_norm": 43.568429501621104,
      "learning_rate": 0.0002129105109626423,
      "loss": 0.5859,
      "step": 3610
    },
    {
      "epoch": 0.98,
      "grad_norm": 26.60773800380206,
      "learning_rate": 0.00021262821116025218,
      "loss": 0.4362,
      "step": 3620
    },
    {
      "epoch": 0.98,
      "grad_norm": 20.572856101276884,
      "learning_rate": 0.00021234591135786202,
      "loss": 0.4832,
      "step": 3630
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.690015588986592,
      "learning_rate": 0.0002120636115554719,
      "loss": 0.3118,
      "step": 3640
    },
    {
      "epoch": 0.98,
      "grad_norm": 13.442740441777506,
      "learning_rate": 0.00021178131175308176,
      "loss": 0.448,
      "step": 3650
    },
    {
      "epoch": 0.99,
      "grad_norm": 16.326270470639226,
      "learning_rate": 0.00021149901195069163,
      "loss": 0.5127,
      "step": 3660
    },
    {
      "epoch": 0.99,
      "grad_norm": 54.671378007773775,
      "learning_rate": 0.00021121671214830148,
      "loss": 0.446,
      "step": 3670
    },
    {
      "epoch": 0.99,
      "grad_norm": 47.169739183643344,
      "learning_rate": 0.00021093441234591135,
      "loss": 0.4233,
      "step": 3680
    },
    {
      "epoch": 0.99,
      "grad_norm": 25.23714400856825,
      "learning_rate": 0.00021065211254352122,
      "loss": 0.434,
      "step": 3690
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.6370358974158163,
      "learning_rate": 0.00021036981274113106,
      "loss": 0.3007,
      "step": 3700
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7791547505560852,
      "eval_f1": 0.8484518098560838,
      "eval_loss": 0.47021484375,
      "eval_precision": 0.8619849357554276,
      "eval_recall": 0.8353370545298411,
      "eval_runtime": 426.3066,
      "eval_samples_per_second": 14.764,
      "eval_steps_per_second": 2.461,
      "step": 3709
    },
    {
      "epoch": 1.0,
      "grad_norm": 14.371648422096674,
      "learning_rate": 0.00021008751293874093,
      "loss": 0.6924,
      "step": 3710
    },
    {
      "epoch": 1.0,
      "grad_norm": 20.060615643426214,
      "learning_rate": 0.0002098052131363508,
      "loss": 0.5005,
      "step": 3720
    },
    {
      "epoch": 1.01,
      "grad_norm": 32.51979250324687,
      "learning_rate": 0.00020952291333396067,
      "loss": 0.4839,
      "step": 3730
    },
    {
      "epoch": 1.01,
      "grad_norm": 13.38387019102551,
      "learning_rate": 0.0002092406135315705,
      "loss": 0.3404,
      "step": 3740
    },
    {
      "epoch": 1.01,
      "grad_norm": 26.230116184390333,
      "learning_rate": 0.00020895831372918036,
      "loss": 0.3164,
      "step": 3750
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.289996088401663,
      "learning_rate": 0.00020867601392679026,
      "loss": 0.2541,
      "step": 3760
    },
    {
      "epoch": 1.02,
      "grad_norm": 13.11053718506525,
      "learning_rate": 0.00020839371412440007,
      "loss": 0.3536,
      "step": 3770
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.864355092419986,
      "learning_rate": 0.00020811141432200994,
      "loss": 0.4036,
      "step": 3780
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.107283075089105,
      "learning_rate": 0.0002078291145196198,
      "loss": 0.3984,
      "step": 3790
    },
    {
      "epoch": 1.02,
      "grad_norm": 5.381512437849745,
      "learning_rate": 0.00020754681471722968,
      "loss": 0.346,
      "step": 3800
    },
    {
      "epoch": 1.03,
      "grad_norm": 17.532003954282274,
      "learning_rate": 0.00020726451491483953,
      "loss": 0.6395,
      "step": 3810
    },
    {
      "epoch": 1.03,
      "grad_norm": 12.3369019096953,
      "learning_rate": 0.0002069822151124494,
      "loss": 0.3291,
      "step": 3820
    },
    {
      "epoch": 1.03,
      "grad_norm": 16.166502649470292,
      "learning_rate": 0.00020669991531005927,
      "loss": 0.5956,
      "step": 3830
    },
    {
      "epoch": 1.04,
      "grad_norm": 31.91509449841551,
      "learning_rate": 0.00020641761550766914,
      "loss": 0.6108,
      "step": 3840
    },
    {
      "epoch": 1.04,
      "grad_norm": 13.205395485901509,
      "learning_rate": 0.00020613531570527898,
      "loss": 0.4119,
      "step": 3850
    },
    {
      "epoch": 1.04,
      "grad_norm": 26.482336411745123,
      "learning_rate": 0.00020585301590288885,
      "loss": 0.454,
      "step": 3860
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.497391794569538,
      "learning_rate": 0.00020557071610049872,
      "loss": 0.541,
      "step": 3870
    },
    {
      "epoch": 1.05,
      "grad_norm": 25.350274518509192,
      "learning_rate": 0.00020528841629810856,
      "loss": 0.404,
      "step": 3880
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.541393152766338,
      "learning_rate": 0.00020500611649571844,
      "loss": 0.3566,
      "step": 3890
    },
    {
      "epoch": 1.05,
      "grad_norm": 18.360766386511038,
      "learning_rate": 0.0002047238166933283,
      "loss": 0.3979,
      "step": 3900
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.769805677066469,
      "learning_rate": 0.00020444151689093818,
      "loss": 0.3482,
      "step": 3910
    },
    {
      "epoch": 1.06,
      "grad_norm": 38.724623371648335,
      "learning_rate": 0.00020415921708854802,
      "loss": 0.5332,
      "step": 3920
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.898633943860567,
      "learning_rate": 0.0002038769172861579,
      "loss": 0.3023,
      "step": 3930
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.614100396855306,
      "learning_rate": 0.00020359461748376776,
      "loss": 0.2848,
      "step": 3940
    },
    {
      "epoch": 1.06,
      "grad_norm": 22.633082871143863,
      "learning_rate": 0.00020331231768137758,
      "loss": 0.2874,
      "step": 3950
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.764963641669773,
      "learning_rate": 0.00020303001787898747,
      "loss": 0.3461,
      "step": 3960
    },
    {
      "epoch": 1.07,
      "grad_norm": 20.85232641432967,
      "learning_rate": 0.00020274771807659734,
      "loss": 0.3602,
      "step": 3970
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.705228938182222,
      "learning_rate": 0.00020246541827420722,
      "loss": 0.296,
      "step": 3980
    },
    {
      "epoch": 1.08,
      "grad_norm": 26.92528030386632,
      "learning_rate": 0.00020218311847181703,
      "loss": 0.3379,
      "step": 3990
    },
    {
      "epoch": 1.08,
      "grad_norm": 24.55220586699356,
      "learning_rate": 0.0002019008186694269,
      "loss": 0.4351,
      "step": 4000
    },
    {
      "epoch": 1.08,
      "grad_norm": 20.689579344891733,
      "learning_rate": 0.00020161851886703677,
      "loss": 0.4861,
      "step": 4010
    },
    {
      "epoch": 1.08,
      "grad_norm": 22.50445804853083,
      "learning_rate": 0.00020133621906464664,
      "loss": 0.405,
      "step": 4020
    },
    {
      "epoch": 1.09,
      "grad_norm": 4.73732515110282,
      "learning_rate": 0.00020105391926225649,
      "loss": 0.2775,
      "step": 4030
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.285187137732595,
      "learning_rate": 0.00020077161945986636,
      "loss": 0.3894,
      "step": 4040
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.464242016670411,
      "learning_rate": 0.00020048931965747623,
      "loss": 0.3608,
      "step": 4050
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.7974856677507993,
      "learning_rate": 0.00020020701985508607,
      "loss": 0.223,
      "step": 4060
    },
    {
      "epoch": 1.1,
      "grad_norm": 13.396799255399001,
      "learning_rate": 0.00019992472005269594,
      "loss": 0.7241,
      "step": 4070
    },
    {
      "epoch": 1.1,
      "grad_norm": 23.7346267162609,
      "learning_rate": 0.0001996424202503058,
      "loss": 0.4137,
      "step": 4080
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.127446556433446,
      "learning_rate": 0.00019936012044791568,
      "loss": 0.4196,
      "step": 4090
    },
    {
      "epoch": 1.11,
      "grad_norm": 18.70765049260556,
      "learning_rate": 0.00019907782064552552,
      "loss": 0.4463,
      "step": 4100
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.924117776990543,
      "learning_rate": 0.0001987955208431354,
      "loss": 0.2398,
      "step": 4110
    },
    {
      "epoch": 1.11,
      "grad_norm": 23.893547770310178,
      "learning_rate": 0.00019851322104074527,
      "loss": 0.4519,
      "step": 4120
    },
    {
      "epoch": 1.11,
      "grad_norm": 23.239519927227455,
      "learning_rate": 0.0001982309212383551,
      "loss": 0.4021,
      "step": 4130
    },
    {
      "epoch": 1.12,
      "grad_norm": 3.237805578914565,
      "learning_rate": 0.00019794862143596498,
      "loss": 0.4012,
      "step": 4140
    },
    {
      "epoch": 1.12,
      "grad_norm": 35.18469175410801,
      "learning_rate": 0.00019766632163357485,
      "loss": 0.3911,
      "step": 4150
    },
    {
      "epoch": 1.12,
      "grad_norm": 36.571739317260835,
      "learning_rate": 0.00019738402183118472,
      "loss": 0.3486,
      "step": 4160
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.18484058053458,
      "learning_rate": 0.00019710172202879456,
      "loss": 0.406,
      "step": 4170
    },
    {
      "epoch": 1.13,
      "grad_norm": 35.44715747315455,
      "learning_rate": 0.00019681942222640443,
      "loss": 0.3687,
      "step": 4180
    },
    {
      "epoch": 1.13,
      "grad_norm": 3.3212732666882783,
      "learning_rate": 0.0001965371224240143,
      "loss": 0.2759,
      "step": 4190
    },
    {
      "epoch": 1.13,
      "grad_norm": 26.49816618587648,
      "learning_rate": 0.00019625482262162418,
      "loss": 0.4258,
      "step": 4200
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.296805121547192,
      "learning_rate": 0.000195972522819234,
      "loss": 0.4145,
      "step": 4210
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.649823380302148,
      "learning_rate": 0.00019569022301684386,
      "loss": 0.4036,
      "step": 4220
    },
    {
      "epoch": 1.14,
      "grad_norm": 23.4716825941928,
      "learning_rate": 0.00019540792321445373,
      "loss": 0.2783,
      "step": 4230
    },
    {
      "epoch": 1.14,
      "grad_norm": 25.818100377140134,
      "learning_rate": 0.00019512562341206358,
      "loss": 0.3295,
      "step": 4240
    },
    {
      "epoch": 1.15,
      "grad_norm": 13.377711591913654,
      "learning_rate": 0.00019484332360967345,
      "loss": 0.3769,
      "step": 4250
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.9366237569646367,
      "learning_rate": 0.00019456102380728332,
      "loss": 0.4161,
      "step": 4260
    },
    {
      "epoch": 1.15,
      "grad_norm": 21.071673893113825,
      "learning_rate": 0.0001942787240048932,
      "loss": 0.2956,
      "step": 4270
    },
    {
      "epoch": 1.15,
      "grad_norm": 7.526667702875879,
      "learning_rate": 0.00019399642420250303,
      "loss": 0.4204,
      "step": 4280
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.388699261113459,
      "learning_rate": 0.0001937141244001129,
      "loss": 0.2555,
      "step": 4290
    },
    {
      "epoch": 1.16,
      "grad_norm": 18.62085102380033,
      "learning_rate": 0.00019343182459772277,
      "loss": 0.4479,
      "step": 4300
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.6157239108667545,
      "learning_rate": 0.00019314952479533261,
      "loss": 0.3801,
      "step": 4310
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.826270241778898,
      "learning_rate": 0.00019286722499294248,
      "loss": 0.4413,
      "step": 4320
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.079464288330672,
      "learning_rate": 0.00019258492519055236,
      "loss": 0.3862,
      "step": 4330
    },
    {
      "epoch": 1.17,
      "grad_norm": 36.279558348502825,
      "learning_rate": 0.00019230262538816223,
      "loss": 0.5484,
      "step": 4340
    },
    {
      "epoch": 1.17,
      "grad_norm": 14.20583811570151,
      "learning_rate": 0.00019202032558577207,
      "loss": 0.4893,
      "step": 4350
    },
    {
      "epoch": 1.18,
      "grad_norm": 13.401648883934785,
      "learning_rate": 0.00019173802578338194,
      "loss": 0.31,
      "step": 4360
    },
    {
      "epoch": 1.18,
      "grad_norm": 7.080540896912264,
      "learning_rate": 0.0001914557259809918,
      "loss": 0.3238,
      "step": 4370
    },
    {
      "epoch": 1.18,
      "grad_norm": 18.2351234516558,
      "learning_rate": 0.00019117342617860168,
      "loss": 0.2408,
      "step": 4380
    },
    {
      "epoch": 1.18,
      "grad_norm": 25.999995633099722,
      "learning_rate": 0.00019089112637621152,
      "loss": 0.5679,
      "step": 4390
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.77314872631193,
      "learning_rate": 0.0001906088265738214,
      "loss": 0.3862,
      "step": 4400
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.429783259777437,
      "learning_rate": 0.00019032652677143126,
      "loss": 0.4151,
      "step": 4410
    },
    {
      "epoch": 1.19,
      "grad_norm": 46.34711557933784,
      "learning_rate": 0.00019004422696904108,
      "loss": 0.6102,
      "step": 4420
    },
    {
      "epoch": 1.19,
      "grad_norm": 33.52872814330087,
      "learning_rate": 0.00018976192716665095,
      "loss": 0.4548,
      "step": 4430
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.703844483062786,
      "learning_rate": 0.00018947962736426082,
      "loss": 0.3616,
      "step": 4440
    },
    {
      "epoch": 1.2,
      "grad_norm": 24.21677862788609,
      "learning_rate": 0.0001892255575421097,
      "loss": 0.4202,
      "step": 4450
    },
    {
      "epoch": 1.2,
      "grad_norm": 44.9650864318285,
      "learning_rate": 0.00018894325773971958,
      "loss": 0.3377,
      "step": 4460
    },
    {
      "epoch": 1.21,
      "grad_norm": 32.58733597861577,
      "learning_rate": 0.00018866095793732942,
      "loss": 0.4558,
      "step": 4470
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.217200799230613,
      "learning_rate": 0.0001883786581349393,
      "loss": 0.4926,
      "step": 4480
    },
    {
      "epoch": 1.21,
      "grad_norm": 14.399950292620698,
      "learning_rate": 0.00018809635833254916,
      "loss": 0.4052,
      "step": 4490
    },
    {
      "epoch": 1.21,
      "grad_norm": 35.643075580293434,
      "learning_rate": 0.00018781405853015903,
      "loss": 0.5095,
      "step": 4500
    },
    {
      "epoch": 1.22,
      "grad_norm": 28.54882761113676,
      "learning_rate": 0.00018753175872776888,
      "loss": 0.3979,
      "step": 4510
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.816459189557854,
      "learning_rate": 0.00018724945892537875,
      "loss": 0.4597,
      "step": 4520
    },
    {
      "epoch": 1.22,
      "grad_norm": 24.826308673730534,
      "learning_rate": 0.00018696715912298862,
      "loss": 0.2722,
      "step": 4530
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.718396005528332,
      "learning_rate": 0.00018668485932059843,
      "loss": 0.2637,
      "step": 4540
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.978326952145644,
      "learning_rate": 0.0001864025595182083,
      "loss": 0.3451,
      "step": 4550
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.63848815746713,
      "learning_rate": 0.00018612025971581817,
      "loss": 0.2449,
      "step": 4560
    },
    {
      "epoch": 1.23,
      "grad_norm": 33.57123035403914,
      "learning_rate": 0.00018583795991342807,
      "loss": 0.4154,
      "step": 4570
    },
    {
      "epoch": 1.23,
      "grad_norm": 22.184179160603396,
      "learning_rate": 0.0001855556601110379,
      "loss": 0.3732,
      "step": 4580
    },
    {
      "epoch": 1.24,
      "grad_norm": 46.18032679521059,
      "learning_rate": 0.00018527336030864776,
      "loss": 0.2738,
      "step": 4590
    },
    {
      "epoch": 1.24,
      "grad_norm": 16.377350849560788,
      "learning_rate": 0.00018499106050625763,
      "loss": 0.3229,
      "step": 4600
    },
    {
      "epoch": 1.24,
      "grad_norm": 31.91092597835325,
      "learning_rate": 0.00018470876070386747,
      "loss": 0.4459,
      "step": 4610
    },
    {
      "epoch": 1.25,
      "grad_norm": 20.40831879809259,
      "learning_rate": 0.00018442646090147734,
      "loss": 0.3928,
      "step": 4620
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.811212492428329,
      "learning_rate": 0.0001841441610990872,
      "loss": 0.4514,
      "step": 4630
    },
    {
      "epoch": 1.25,
      "grad_norm": 14.46580839275718,
      "learning_rate": 0.00018386186129669708,
      "loss": 0.512,
      "step": 4640
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.891630925355765,
      "learning_rate": 0.00018357956149430693,
      "loss": 0.4664,
      "step": 4650
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.548463036511295,
      "learning_rate": 0.0001832972616919168,
      "loss": 0.3487,
      "step": 4660
    },
    {
      "epoch": 1.26,
      "grad_norm": 5.985771603253264,
      "learning_rate": 0.00018301496188952667,
      "loss": 0.427,
      "step": 4670
    },
    {
      "epoch": 1.26,
      "grad_norm": 18.110354087666913,
      "learning_rate": 0.00018273266208713654,
      "loss": 0.3864,
      "step": 4680
    },
    {
      "epoch": 1.26,
      "grad_norm": 16.37360903241179,
      "learning_rate": 0.00018245036228474638,
      "loss": 0.3586,
      "step": 4690
    },
    {
      "epoch": 1.27,
      "grad_norm": 43.587679695336604,
      "learning_rate": 0.00018216806248235625,
      "loss": 0.4306,
      "step": 4700
    },
    {
      "epoch": 1.27,
      "grad_norm": 18.292908364374888,
      "learning_rate": 0.00018188576267996612,
      "loss": 0.5503,
      "step": 4710
    },
    {
      "epoch": 1.27,
      "grad_norm": 13.843813591820725,
      "learning_rate": 0.00018160346287757597,
      "loss": 0.2332,
      "step": 4720
    },
    {
      "epoch": 1.28,
      "grad_norm": 4.349105648845497,
      "learning_rate": 0.00018132116307518584,
      "loss": 0.2422,
      "step": 4730
    },
    {
      "epoch": 1.28,
      "grad_norm": 14.33758083417271,
      "learning_rate": 0.0001810388632727957,
      "loss": 0.3736,
      "step": 4740
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.120845857421083,
      "learning_rate": 0.00018075656347040558,
      "loss": 0.5166,
      "step": 4750
    },
    {
      "epoch": 1.28,
      "grad_norm": 19.90717896059776,
      "learning_rate": 0.0001804742636680154,
      "loss": 0.3587,
      "step": 4760
    },
    {
      "epoch": 1.29,
      "grad_norm": 44.003732766775826,
      "learning_rate": 0.0001801919638656253,
      "loss": 0.3787,
      "step": 4770
    },
    {
      "epoch": 1.29,
      "grad_norm": 43.56234297320444,
      "learning_rate": 0.00017990966406323516,
      "loss": 0.5008,
      "step": 4780
    },
    {
      "epoch": 1.29,
      "grad_norm": 21.209510493179437,
      "learning_rate": 0.00017962736426084498,
      "loss": 0.4537,
      "step": 4790
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.347936326257064,
      "learning_rate": 0.00017934506445845485,
      "loss": 0.3925,
      "step": 4800
    },
    {
      "epoch": 1.3,
      "grad_norm": 30.48363490161901,
      "learning_rate": 0.00017906276465606472,
      "loss": 0.3559,
      "step": 4810
    },
    {
      "epoch": 1.3,
      "grad_norm": 17.25510114375899,
      "learning_rate": 0.0001787804648536746,
      "loss": 0.3293,
      "step": 4820
    },
    {
      "epoch": 1.3,
      "grad_norm": 35.37006437340401,
      "learning_rate": 0.00017849816505128443,
      "loss": 0.3755,
      "step": 4830
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.309306838617072,
      "learning_rate": 0.0001782158652488943,
      "loss": 0.5439,
      "step": 4840
    },
    {
      "epoch": 1.31,
      "grad_norm": 24.80686078620767,
      "learning_rate": 0.00017793356544650417,
      "loss": 0.3078,
      "step": 4850
    },
    {
      "epoch": 1.31,
      "grad_norm": 38.90468574504618,
      "learning_rate": 0.00017765126564411404,
      "loss": 0.4582,
      "step": 4860
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.829488842567992,
      "learning_rate": 0.0001773689658417239,
      "loss": 0.3271,
      "step": 4870
    },
    {
      "epoch": 1.32,
      "grad_norm": 24.755736846620604,
      "learning_rate": 0.00017708666603933376,
      "loss": 0.3677,
      "step": 4880
    },
    {
      "epoch": 1.32,
      "grad_norm": 34.35509407051675,
      "learning_rate": 0.00017680436623694363,
      "loss": 0.3704,
      "step": 4890
    },
    {
      "epoch": 1.32,
      "grad_norm": 20.471141490932293,
      "learning_rate": 0.00017652206643455347,
      "loss": 0.2797,
      "step": 4900
    },
    {
      "epoch": 1.32,
      "grad_norm": 4.217204360921141,
      "learning_rate": 0.00017623976663216334,
      "loss": 0.4381,
      "step": 4910
    },
    {
      "epoch": 1.33,
      "grad_norm": 6.346290799511884,
      "learning_rate": 0.0001759574668297732,
      "loss": 0.5076,
      "step": 4920
    },
    {
      "epoch": 1.33,
      "grad_norm": 25.523968835559796,
      "learning_rate": 0.00017567516702738308,
      "loss": 0.3612,
      "step": 4930
    },
    {
      "epoch": 1.33,
      "grad_norm": 13.927791728542356,
      "learning_rate": 0.00017539286722499293,
      "loss": 0.4505,
      "step": 4940
    },
    {
      "epoch": 1.33,
      "grad_norm": 22.213464412055487,
      "learning_rate": 0.0001751105674226028,
      "loss": 0.2757,
      "step": 4950
    },
    {
      "epoch": 1.34,
      "grad_norm": 15.447939485086996,
      "learning_rate": 0.00017485649760045166,
      "loss": 0.6925,
      "step": 4960
    },
    {
      "epoch": 1.34,
      "grad_norm": 16.049582512051508,
      "learning_rate": 0.00017457419779806153,
      "loss": 0.5154,
      "step": 4970
    },
    {
      "epoch": 1.34,
      "grad_norm": 27.78750991331512,
      "learning_rate": 0.0001742918979956714,
      "loss": 0.2774,
      "step": 4980
    },
    {
      "epoch": 1.35,
      "grad_norm": 22.385233996092563,
      "learning_rate": 0.00017400959819328124,
      "loss": 0.3651,
      "step": 4990
    },
    {
      "epoch": 1.35,
      "grad_norm": 49.90645738461855,
      "learning_rate": 0.0001737272983908911,
      "loss": 0.4008,
      "step": 5000
    },
    {
      "epoch": 1.35,
      "grad_norm": 23.242128816418333,
      "learning_rate": 0.00017344499858850098,
      "loss": 0.3577,
      "step": 5010
    },
    {
      "epoch": 1.35,
      "grad_norm": 39.72621869976524,
      "learning_rate": 0.00017316269878611082,
      "loss": 0.5516,
      "step": 5020
    },
    {
      "epoch": 1.36,
      "grad_norm": 30.676462413516926,
      "learning_rate": 0.0001728803989837207,
      "loss": 0.3591,
      "step": 5030
    },
    {
      "epoch": 1.36,
      "grad_norm": 4.158902499171461,
      "learning_rate": 0.00017259809918133056,
      "loss": 0.4531,
      "step": 5040
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.120664692907559,
      "learning_rate": 0.00017231579937894043,
      "loss": 0.3,
      "step": 5050
    },
    {
      "epoch": 1.36,
      "grad_norm": 25.631259427897913,
      "learning_rate": 0.00017203349957655028,
      "loss": 0.5833,
      "step": 5060
    },
    {
      "epoch": 1.37,
      "grad_norm": 31.544411671943255,
      "learning_rate": 0.00017175119977416015,
      "loss": 0.2767,
      "step": 5070
    },
    {
      "epoch": 1.37,
      "grad_norm": 41.1791679240627,
      "learning_rate": 0.00017146889997177002,
      "loss": 0.4477,
      "step": 5080
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.185851921881162,
      "learning_rate": 0.00017118660016937986,
      "loss": 0.3113,
      "step": 5090
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.60396492257944,
      "learning_rate": 0.00017090430036698973,
      "loss": 0.3154,
      "step": 5100
    },
    {
      "epoch": 1.38,
      "grad_norm": 28.576434110442438,
      "learning_rate": 0.0001706220005645996,
      "loss": 0.2589,
      "step": 5110
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.85968019827676,
      "learning_rate": 0.00017033970076220947,
      "loss": 0.2302,
      "step": 5120
    },
    {
      "epoch": 1.38,
      "grad_norm": 55.99816919212105,
      "learning_rate": 0.0001700574009598193,
      "loss": 0.4289,
      "step": 5130
    },
    {
      "epoch": 1.39,
      "grad_norm": 36.89423653044147,
      "learning_rate": 0.00016977510115742916,
      "loss": 0.3095,
      "step": 5140
    },
    {
      "epoch": 1.39,
      "grad_norm": 5.580393848359779,
      "learning_rate": 0.00016949280135503903,
      "loss": 0.5351,
      "step": 5150
    },
    {
      "epoch": 1.39,
      "grad_norm": 4.462892309893868,
      "learning_rate": 0.0001692105015526489,
      "loss": 0.3871,
      "step": 5160
    },
    {
      "epoch": 1.39,
      "grad_norm": 6.300597440119085,
      "learning_rate": 0.00016892820175025874,
      "loss": 0.3467,
      "step": 5170
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.993182464662825,
      "learning_rate": 0.00016864590194786861,
      "loss": 0.421,
      "step": 5180
    },
    {
      "epoch": 1.4,
      "grad_norm": 22.310065030586298,
      "learning_rate": 0.00016836360214547849,
      "loss": 0.4512,
      "step": 5190
    },
    {
      "epoch": 1.4,
      "grad_norm": 23.275643244489054,
      "learning_rate": 0.00016808130234308833,
      "loss": 0.4284,
      "step": 5200
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.180664724635761,
      "learning_rate": 0.0001677990025406982,
      "loss": 0.388,
      "step": 5210
    },
    {
      "epoch": 1.41,
      "grad_norm": 21.348879184577736,
      "learning_rate": 0.00016751670273830807,
      "loss": 0.4848,
      "step": 5220
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.297915190077562,
      "learning_rate": 0.00016723440293591794,
      "loss": 0.3655,
      "step": 5230
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.998633582311555,
      "learning_rate": 0.00016695210313352778,
      "loss": 0.2807,
      "step": 5240
    },
    {
      "epoch": 1.42,
      "grad_norm": 51.850423359797894,
      "learning_rate": 0.00016666980333113765,
      "loss": 0.3715,
      "step": 5250
    },
    {
      "epoch": 1.42,
      "grad_norm": 3.013735827671648,
      "learning_rate": 0.00016638750352874752,
      "loss": 0.5658,
      "step": 5260
    },
    {
      "epoch": 1.42,
      "grad_norm": 14.486828894495519,
      "learning_rate": 0.00016610520372635737,
      "loss": 0.3649,
      "step": 5270
    },
    {
      "epoch": 1.42,
      "grad_norm": 7.391081655383877,
      "learning_rate": 0.00016582290392396724,
      "loss": 0.2949,
      "step": 5280
    },
    {
      "epoch": 1.43,
      "grad_norm": 2.5884286585705385,
      "learning_rate": 0.0001655406041215771,
      "loss": 0.3707,
      "step": 5290
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.091912379642805,
      "learning_rate": 0.00016525830431918698,
      "loss": 0.4874,
      "step": 5300
    },
    {
      "epoch": 1.43,
      "grad_norm": 22.542540511130795,
      "learning_rate": 0.00016497600451679682,
      "loss": 0.3987,
      "step": 5310
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.865960857857342,
      "learning_rate": 0.0001646937047144067,
      "loss": 0.3227,
      "step": 5320
    },
    {
      "epoch": 1.44,
      "grad_norm": 39.31723253430751,
      "learning_rate": 0.00016441140491201656,
      "loss": 0.2917,
      "step": 5330
    },
    {
      "epoch": 1.44,
      "grad_norm": 39.4584632004162,
      "learning_rate": 0.00016412910510962643,
      "loss": 0.3376,
      "step": 5340
    },
    {
      "epoch": 1.44,
      "grad_norm": 15.203037486775214,
      "learning_rate": 0.00016384680530723625,
      "loss": 0.4042,
      "step": 5350
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.743109893886675,
      "learning_rate": 0.00016356450550484612,
      "loss": 0.314,
      "step": 5360
    },
    {
      "epoch": 1.45,
      "grad_norm": 40.258371746798936,
      "learning_rate": 0.000163282205702456,
      "loss": 0.3904,
      "step": 5370
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.700064336496186,
      "learning_rate": 0.00016299990590006583,
      "loss": 0.4211,
      "step": 5380
    },
    {
      "epoch": 1.45,
      "grad_norm": 33.34116490701284,
      "learning_rate": 0.0001627176060976757,
      "loss": 0.3522,
      "step": 5390
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.372970505754688,
      "learning_rate": 0.00016243530629528557,
      "loss": 0.2914,
      "step": 5400
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.630607846599633,
      "learning_rate": 0.00016215300649289545,
      "loss": 0.446,
      "step": 5410
    },
    {
      "epoch": 1.46,
      "grad_norm": 13.58852989636247,
      "learning_rate": 0.0001618707066905053,
      "loss": 0.3701,
      "step": 5420
    },
    {
      "epoch": 1.46,
      "grad_norm": 34.823022763034245,
      "learning_rate": 0.00016158840688811516,
      "loss": 0.2976,
      "step": 5430
    },
    {
      "epoch": 1.47,
      "grad_norm": 5.993988032681155,
      "learning_rate": 0.00016130610708572503,
      "loss": 0.2634,
      "step": 5440
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.674300084619954,
      "learning_rate": 0.00016102380728333487,
      "loss": 0.2912,
      "step": 5450
    },
    {
      "epoch": 1.47,
      "grad_norm": 32.99811276273762,
      "learning_rate": 0.00016074150748094474,
      "loss": 0.406,
      "step": 5460
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.619332903681824,
      "learning_rate": 0.00016045920767855461,
      "loss": 0.3083,
      "step": 5470
    },
    {
      "epoch": 1.48,
      "grad_norm": 34.895040666155715,
      "learning_rate": 0.00016017690787616448,
      "loss": 0.3672,
      "step": 5480
    },
    {
      "epoch": 1.48,
      "grad_norm": 16.088253859086933,
      "learning_rate": 0.00015989460807377433,
      "loss": 0.3341,
      "step": 5490
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.830945624133527,
      "learning_rate": 0.0001596123082713842,
      "loss": 0.5567,
      "step": 5500
    },
    {
      "epoch": 1.49,
      "grad_norm": 27.223869035614634,
      "learning_rate": 0.00015933000846899407,
      "loss": 0.4043,
      "step": 5510
    },
    {
      "epoch": 1.49,
      "grad_norm": 4.288173539985467,
      "learning_rate": 0.00015904770866660394,
      "loss": 0.3379,
      "step": 5520
    },
    {
      "epoch": 1.49,
      "grad_norm": 15.30555612368673,
      "learning_rate": 0.00015876540886421378,
      "loss": 0.3149,
      "step": 5530
    },
    {
      "epoch": 1.49,
      "grad_norm": 23.50815915377938,
      "learning_rate": 0.00015848310906182365,
      "loss": 0.271,
      "step": 5540
    },
    {
      "epoch": 1.5,
      "grad_norm": 7.094398054851781,
      "learning_rate": 0.00015820080925943352,
      "loss": 0.3532,
      "step": 5550
    },
    {
      "epoch": 1.5,
      "grad_norm": 13.325254550744852,
      "learning_rate": 0.00015791850945704334,
      "loss": 0.2735,
      "step": 5560
    },
    {
      "epoch": 1.5,
      "grad_norm": 22.954673920624085,
      "learning_rate": 0.0001576362096546532,
      "loss": 0.3704,
      "step": 5570
    },
    {
      "epoch": 1.5,
      "grad_norm": 39.31536381946065,
      "learning_rate": 0.00015735390985226308,
      "loss": 0.4479,
      "step": 5580
    },
    {
      "epoch": 1.51,
      "grad_norm": 21.806981359536874,
      "learning_rate": 0.00015707161004987298,
      "loss": 0.3336,
      "step": 5590
    },
    {
      "epoch": 1.51,
      "grad_norm": 34.329554705566764,
      "learning_rate": 0.0001567893102474828,
      "loss": 0.3267,
      "step": 5600
    },
    {
      "epoch": 1.51,
      "grad_norm": 5.354809751847723,
      "learning_rate": 0.00015650701044509266,
      "loss": 0.3454,
      "step": 5610
    },
    {
      "epoch": 1.52,
      "grad_norm": 26.4178265207194,
      "learning_rate": 0.00015622471064270253,
      "loss": 0.5798,
      "step": 5620
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.496689770017137,
      "learning_rate": 0.00015594241084031238,
      "loss": 0.4079,
      "step": 5630
    },
    {
      "epoch": 1.52,
      "grad_norm": 15.515175151162078,
      "learning_rate": 0.00015566011103792225,
      "loss": 0.3532,
      "step": 5640
    },
    {
      "epoch": 1.52,
      "grad_norm": 14.461147889019037,
      "learning_rate": 0.00015537781123553212,
      "loss": 0.3778,
      "step": 5650
    },
    {
      "epoch": 1.53,
      "grad_norm": 17.125261515992566,
      "learning_rate": 0.000155095511433142,
      "loss": 0.2629,
      "step": 5660
    },
    {
      "epoch": 1.53,
      "grad_norm": 16.327341070931993,
      "learning_rate": 0.00015481321163075183,
      "loss": 0.3973,
      "step": 5670
    },
    {
      "epoch": 1.53,
      "grad_norm": 3.298703827952682,
      "learning_rate": 0.0001545309118283617,
      "loss": 0.2724,
      "step": 5680
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.184755138104512,
      "learning_rate": 0.00015424861202597157,
      "loss": 0.4472,
      "step": 5690
    },
    {
      "epoch": 1.54,
      "grad_norm": 19.3499540912622,
      "learning_rate": 0.00015396631222358144,
      "loss": 0.356,
      "step": 5700
    },
    {
      "epoch": 1.54,
      "grad_norm": 68.96711690997579,
      "learning_rate": 0.0001536840124211913,
      "loss": 0.3998,
      "step": 5710
    },
    {
      "epoch": 1.54,
      "grad_norm": 5.76747397914272,
      "learning_rate": 0.00015340171261880116,
      "loss": 0.3517,
      "step": 5720
    },
    {
      "epoch": 1.54,
      "grad_norm": 37.21818937460734,
      "learning_rate": 0.00015311941281641103,
      "loss": 0.4953,
      "step": 5730
    },
    {
      "epoch": 1.55,
      "grad_norm": 7.25240026477868,
      "learning_rate": 0.00015283711301402087,
      "loss": 0.2064,
      "step": 5740
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.79170130227607,
      "learning_rate": 0.00015255481321163074,
      "loss": 0.396,
      "step": 5750
    },
    {
      "epoch": 1.55,
      "grad_norm": 21.333263050105753,
      "learning_rate": 0.0001522725134092406,
      "loss": 0.3649,
      "step": 5760
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.884075494774457,
      "learning_rate": 0.00015199021360685048,
      "loss": 0.3311,
      "step": 5770
    },
    {
      "epoch": 1.56,
      "grad_norm": 19.794068811080198,
      "learning_rate": 0.0001517079138044603,
      "loss": 0.4061,
      "step": 5780
    },
    {
      "epoch": 1.56,
      "grad_norm": 6.537898785460164,
      "learning_rate": 0.0001514256140020702,
      "loss": 0.5371,
      "step": 5790
    },
    {
      "epoch": 1.56,
      "grad_norm": 4.810779675477839,
      "learning_rate": 0.00015114331419968007,
      "loss": 0.2938,
      "step": 5800
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.691549377763073,
      "learning_rate": 0.00015086101439728988,
      "loss": 0.3254,
      "step": 5810
    },
    {
      "epoch": 1.57,
      "grad_norm": 6.2935189093073065,
      "learning_rate": 0.00015057871459489975,
      "loss": 0.2548,
      "step": 5820
    },
    {
      "epoch": 1.57,
      "grad_norm": 14.993671245415825,
      "learning_rate": 0.00015029641479250962,
      "loss": 0.4677,
      "step": 5830
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.608299101622812,
      "learning_rate": 0.0001500141149901195,
      "loss": 0.454,
      "step": 5840
    },
    {
      "epoch": 1.58,
      "grad_norm": 18.48856622595982,
      "learning_rate": 0.00014973181518772937,
      "loss": 0.2249,
      "step": 5850
    },
    {
      "epoch": 1.58,
      "grad_norm": 14.948049137949582,
      "learning_rate": 0.0001494495153853392,
      "loss": 0.2912,
      "step": 5860
    },
    {
      "epoch": 1.58,
      "grad_norm": 22.660059625044443,
      "learning_rate": 0.00014916721558294908,
      "loss": 0.2683,
      "step": 5870
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.103598229462358,
      "learning_rate": 0.00014888491578055895,
      "loss": 0.3692,
      "step": 5880
    },
    {
      "epoch": 1.59,
      "grad_norm": 30.383036889002675,
      "learning_rate": 0.00014860261597816882,
      "loss": 0.451,
      "step": 5890
    },
    {
      "epoch": 1.59,
      "grad_norm": 17.506092846066977,
      "learning_rate": 0.00014832031617577866,
      "loss": 0.4592,
      "step": 5900
    },
    {
      "epoch": 1.59,
      "grad_norm": 12.407317043906799,
      "learning_rate": 0.0001480380163733885,
      "loss": 0.3131,
      "step": 5910
    },
    {
      "epoch": 1.6,
      "grad_norm": 3.6029316344307345,
      "learning_rate": 0.00014775571657099838,
      "loss": 0.2815,
      "step": 5920
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.304127443307513,
      "learning_rate": 0.00014747341676860825,
      "loss": 0.5046,
      "step": 5930
    },
    {
      "epoch": 1.6,
      "grad_norm": 24.330314608133015,
      "learning_rate": 0.00014719111696621812,
      "loss": 0.2557,
      "step": 5940
    },
    {
      "epoch": 1.6,
      "grad_norm": 31.678740057083264,
      "learning_rate": 0.00014690881716382796,
      "loss": 0.3507,
      "step": 5950
    },
    {
      "epoch": 1.61,
      "grad_norm": 25.049077827145965,
      "learning_rate": 0.00014662651736143783,
      "loss": 0.3242,
      "step": 5960
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.651970717501142,
      "learning_rate": 0.0001463442175590477,
      "loss": 0.4066,
      "step": 5970
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.2881747913139,
      "learning_rate": 0.00014606191775665757,
      "loss": 0.3632,
      "step": 5980
    },
    {
      "epoch": 1.61,
      "grad_norm": 36.4440322168643,
      "learning_rate": 0.00014577961795426742,
      "loss": 0.235,
      "step": 5990
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.236959140085006,
      "learning_rate": 0.00014549731815187729,
      "loss": 0.4342,
      "step": 6000
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.533275077762876,
      "learning_rate": 0.00014521501834948716,
      "loss": 0.5866,
      "step": 6010
    },
    {
      "epoch": 1.62,
      "grad_norm": 16.28346226554514,
      "learning_rate": 0.000144932718547097,
      "loss": 0.3366,
      "step": 6020
    },
    {
      "epoch": 1.63,
      "grad_norm": 17.356315711577516,
      "learning_rate": 0.00014465041874470687,
      "loss": 0.3347,
      "step": 6030
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.9623321790235537,
      "learning_rate": 0.00014436811894231671,
      "loss": 0.413,
      "step": 6040
    },
    {
      "epoch": 1.63,
      "grad_norm": 18.190149204022724,
      "learning_rate": 0.00014408581913992658,
      "loss": 0.4737,
      "step": 6050
    },
    {
      "epoch": 1.63,
      "grad_norm": 5.4097060949722735,
      "learning_rate": 0.00014380351933753645,
      "loss": 0.3771,
      "step": 6060
    },
    {
      "epoch": 1.64,
      "grad_norm": 28.692723266465883,
      "learning_rate": 0.00014352121953514633,
      "loss": 0.4013,
      "step": 6070
    },
    {
      "epoch": 1.64,
      "grad_norm": 6.704164980639117,
      "learning_rate": 0.00014323891973275617,
      "loss": 0.2788,
      "step": 6080
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.106720526048127,
      "learning_rate": 0.00014295661993036604,
      "loss": 0.2633,
      "step": 6090
    },
    {
      "epoch": 1.64,
      "grad_norm": 32.86881996491162,
      "learning_rate": 0.0001426743201279759,
      "loss": 0.4318,
      "step": 6100
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.740163348995152,
      "learning_rate": 0.00014239202032558575,
      "loss": 0.3447,
      "step": 6110
    },
    {
      "epoch": 1.65,
      "grad_norm": 22.87827196539711,
      "learning_rate": 0.00014210972052319562,
      "loss": 0.346,
      "step": 6120
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.49042890902758,
      "learning_rate": 0.00014182742072080547,
      "loss": 0.2664,
      "step": 6130
    },
    {
      "epoch": 1.66,
      "grad_norm": 4.609745631935454,
      "learning_rate": 0.00014154512091841534,
      "loss": 0.3143,
      "step": 6140
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.702824515520128,
      "learning_rate": 0.0001412628211160252,
      "loss": 0.3166,
      "step": 6150
    },
    {
      "epoch": 1.66,
      "grad_norm": 26.461138852628583,
      "learning_rate": 0.00014098052131363508,
      "loss": 0.3628,
      "step": 6160
    },
    {
      "epoch": 1.66,
      "grad_norm": 25.54502984654189,
      "learning_rate": 0.00014069822151124492,
      "loss": 0.4041,
      "step": 6170
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.020770562747493,
      "learning_rate": 0.0001404159217088548,
      "loss": 0.3701,
      "step": 6180
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.456892732964649,
      "learning_rate": 0.00014013362190646466,
      "loss": 0.357,
      "step": 6190
    },
    {
      "epoch": 1.67,
      "grad_norm": 16.26098154659414,
      "learning_rate": 0.0001398513221040745,
      "loss": 0.3566,
      "step": 6200
    },
    {
      "epoch": 1.67,
      "grad_norm": 38.88299108176432,
      "learning_rate": 0.00013956902230168438,
      "loss": 0.3563,
      "step": 6210
    },
    {
      "epoch": 1.68,
      "grad_norm": 16.691404089000194,
      "learning_rate": 0.00013928672249929425,
      "loss": 0.3288,
      "step": 6220
    },
    {
      "epoch": 1.68,
      "grad_norm": 27.194423052240403,
      "learning_rate": 0.00013900442269690412,
      "loss": 0.3214,
      "step": 6230
    },
    {
      "epoch": 1.68,
      "grad_norm": 41.289846011895776,
      "learning_rate": 0.00013872212289451396,
      "loss": 0.4334,
      "step": 6240
    },
    {
      "epoch": 1.69,
      "grad_norm": 5.143987184746201,
      "learning_rate": 0.00013843982309212383,
      "loss": 0.3107,
      "step": 6250
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.149484056645836,
      "learning_rate": 0.00013815752328973367,
      "loss": 0.5028,
      "step": 6260
    },
    {
      "epoch": 1.69,
      "grad_norm": 5.325077408157439,
      "learning_rate": 0.00013787522348734354,
      "loss": 0.323,
      "step": 6270
    },
    {
      "epoch": 1.69,
      "grad_norm": 4.877899394738392,
      "learning_rate": 0.00013759292368495341,
      "loss": 0.3351,
      "step": 6280
    },
    {
      "epoch": 1.7,
      "grad_norm": 16.87347550746856,
      "learning_rate": 0.00013731062388256326,
      "loss": 0.2805,
      "step": 6290
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.192405287263968,
      "learning_rate": 0.00013702832408017313,
      "loss": 0.3464,
      "step": 6300
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.107190763059997,
      "learning_rate": 0.000136746024277783,
      "loss": 0.2964,
      "step": 6310
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.2232084526940472,
      "learning_rate": 0.00013646372447539287,
      "loss": 0.3312,
      "step": 6320
    },
    {
      "epoch": 1.71,
      "grad_norm": 30.714328162387332,
      "learning_rate": 0.0001361814246730027,
      "loss": 0.4577,
      "step": 6330
    },
    {
      "epoch": 1.71,
      "grad_norm": 21.59453533910639,
      "learning_rate": 0.00013589912487061258,
      "loss": 0.5453,
      "step": 6340
    },
    {
      "epoch": 1.71,
      "grad_norm": 28.072225375973694,
      "learning_rate": 0.00013561682506822245,
      "loss": 0.3857,
      "step": 6350
    },
    {
      "epoch": 1.71,
      "grad_norm": 16.580317202653177,
      "learning_rate": 0.0001353345252658323,
      "loss": 0.3246,
      "step": 6360
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.929318202237337,
      "learning_rate": 0.00013505222546344217,
      "loss": 0.4445,
      "step": 6370
    },
    {
      "epoch": 1.72,
      "grad_norm": 6.725550647995846,
      "learning_rate": 0.000134769925661052,
      "loss": 0.2732,
      "step": 6380
    },
    {
      "epoch": 1.72,
      "grad_norm": 20.71913726600779,
      "learning_rate": 0.00013448762585866188,
      "loss": 0.4111,
      "step": 6390
    },
    {
      "epoch": 1.73,
      "grad_norm": 35.74751043195964,
      "learning_rate": 0.00013420532605627175,
      "loss": 0.3852,
      "step": 6400
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.858038898679881,
      "learning_rate": 0.00013392302625388162,
      "loss": 0.4292,
      "step": 6410
    },
    {
      "epoch": 1.73,
      "grad_norm": 24.50201175115572,
      "learning_rate": 0.00013364072645149147,
      "loss": 0.4118,
      "step": 6420
    },
    {
      "epoch": 1.73,
      "grad_norm": 17.332956008709065,
      "learning_rate": 0.00013335842664910134,
      "loss": 0.3596,
      "step": 6430
    },
    {
      "epoch": 1.74,
      "grad_norm": 7.741996900524846,
      "learning_rate": 0.0001330761268467112,
      "loss": 0.5228,
      "step": 6440
    },
    {
      "epoch": 1.74,
      "grad_norm": 7.353424107158127,
      "learning_rate": 0.00013279382704432105,
      "loss": 0.301,
      "step": 6450
    },
    {
      "epoch": 1.74,
      "grad_norm": 31.90848173281483,
      "learning_rate": 0.00013251152724193092,
      "loss": 0.3032,
      "step": 6460
    },
    {
      "epoch": 1.74,
      "grad_norm": 24.574275203331343,
      "learning_rate": 0.00013222922743954076,
      "loss": 0.2991,
      "step": 6470
    },
    {
      "epoch": 1.75,
      "grad_norm": 4.553181183245165,
      "learning_rate": 0.00013194692763715063,
      "loss": 0.358,
      "step": 6480
    },
    {
      "epoch": 1.75,
      "grad_norm": 19.601115191993912,
      "learning_rate": 0.0001316646278347605,
      "loss": 0.3265,
      "step": 6490
    },
    {
      "epoch": 1.75,
      "grad_norm": 38.772025142759304,
      "learning_rate": 0.00013138232803237037,
      "loss": 0.4264,
      "step": 6500
    },
    {
      "epoch": 1.76,
      "grad_norm": 43.88525090668065,
      "learning_rate": 0.00013110002822998022,
      "loss": 0.4106,
      "step": 6510
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.934368440291996,
      "learning_rate": 0.0001308177284275901,
      "loss": 0.4272,
      "step": 6520
    },
    {
      "epoch": 1.76,
      "grad_norm": 24.179789499335776,
      "learning_rate": 0.00013053542862519996,
      "loss": 0.3914,
      "step": 6530
    },
    {
      "epoch": 1.76,
      "grad_norm": 4.827209179496258,
      "learning_rate": 0.0001302531288228098,
      "loss": 0.2772,
      "step": 6540
    },
    {
      "epoch": 1.77,
      "grad_norm": 23.889194814961925,
      "learning_rate": 0.00012997082902041967,
      "loss": 0.3611,
      "step": 6550
    },
    {
      "epoch": 1.77,
      "grad_norm": 28.90976277890139,
      "learning_rate": 0.00012968852921802954,
      "loss": 0.3909,
      "step": 6560
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.250733950331073,
      "learning_rate": 0.0001294062294156394,
      "loss": 0.4212,
      "step": 6570
    },
    {
      "epoch": 1.77,
      "grad_norm": 6.814866167442905,
      "learning_rate": 0.00012912392961324926,
      "loss": 0.3503,
      "step": 6580
    },
    {
      "epoch": 1.78,
      "grad_norm": 4.886849115959408,
      "learning_rate": 0.00012884162981085913,
      "loss": 0.4606,
      "step": 6590
    },
    {
      "epoch": 1.78,
      "grad_norm": 59.53592514794029,
      "learning_rate": 0.00012855933000846897,
      "loss": 0.4212,
      "step": 6600
    },
    {
      "epoch": 1.78,
      "grad_norm": 29.399453494063412,
      "learning_rate": 0.00012827703020607884,
      "loss": 0.3291,
      "step": 6610
    },
    {
      "epoch": 1.78,
      "grad_norm": 78.54962841494819,
      "learning_rate": 0.0001279947304036887,
      "loss": 0.3618,
      "step": 6620
    },
    {
      "epoch": 1.79,
      "grad_norm": 17.473265144131208,
      "learning_rate": 0.00012771243060129855,
      "loss": 0.4234,
      "step": 6630
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.499983191649382,
      "learning_rate": 0.00012743013079890843,
      "loss": 0.3405,
      "step": 6640
    },
    {
      "epoch": 1.79,
      "grad_norm": 36.8827925796956,
      "learning_rate": 0.0001271478309965183,
      "loss": 0.4003,
      "step": 6650
    },
    {
      "epoch": 1.8,
      "grad_norm": 44.967606008702894,
      "learning_rate": 0.00012686553119412817,
      "loss": 0.4024,
      "step": 6660
    },
    {
      "epoch": 1.8,
      "grad_norm": 23.807751940333322,
      "learning_rate": 0.000126583231391738,
      "loss": 0.4316,
      "step": 6670
    },
    {
      "epoch": 1.8,
      "grad_norm": 19.49729412627736,
      "learning_rate": 0.00012630093158934788,
      "loss": 0.4558,
      "step": 6680
    },
    {
      "epoch": 1.8,
      "grad_norm": 18.854631350430427,
      "learning_rate": 0.00012601863178695772,
      "loss": 0.2691,
      "step": 6690
    },
    {
      "epoch": 1.81,
      "grad_norm": 49.388241549567844,
      "learning_rate": 0.0001257363319845676,
      "loss": 0.3985,
      "step": 6700
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.83527440870295,
      "learning_rate": 0.00012545403218217746,
      "loss": 0.2902,
      "step": 6710
    },
    {
      "epoch": 1.81,
      "grad_norm": 25.799688603137753,
      "learning_rate": 0.0001251717323797873,
      "loss": 0.2792,
      "step": 6720
    },
    {
      "epoch": 1.81,
      "grad_norm": 16.672487318052635,
      "learning_rate": 0.00012488943257739718,
      "loss": 0.4108,
      "step": 6730
    },
    {
      "epoch": 1.82,
      "grad_norm": 3.186304852555112,
      "learning_rate": 0.00012460713277500705,
      "loss": 0.268,
      "step": 6740
    },
    {
      "epoch": 1.82,
      "grad_norm": 5.308526710973965,
      "learning_rate": 0.00012432483297261692,
      "loss": 0.1926,
      "step": 6750
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.984156623356547,
      "learning_rate": 0.00012404253317022676,
      "loss": 0.4313,
      "step": 6760
    },
    {
      "epoch": 1.83,
      "grad_norm": 17.48252769534504,
      "learning_rate": 0.00012376023336783663,
      "loss": 0.343,
      "step": 6770
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.73191399292905,
      "learning_rate": 0.0001234779335654465,
      "loss": 0.4654,
      "step": 6780
    },
    {
      "epoch": 1.83,
      "grad_norm": 5.969252470658802,
      "learning_rate": 0.00012319563376305637,
      "loss": 0.2658,
      "step": 6790
    },
    {
      "epoch": 1.83,
      "grad_norm": 19.24600518293154,
      "learning_rate": 0.00012291333396066622,
      "loss": 0.3893,
      "step": 6800
    },
    {
      "epoch": 1.84,
      "grad_norm": 24.291588754035516,
      "learning_rate": 0.00012263103415827606,
      "loss": 0.2898,
      "step": 6810
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.553137206234199,
      "learning_rate": 0.00012234873435588593,
      "loss": 0.2407,
      "step": 6820
    },
    {
      "epoch": 1.84,
      "grad_norm": 5.908850912017495,
      "learning_rate": 0.0001220664345534958,
      "loss": 0.1475,
      "step": 6830
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.47052908325656,
      "learning_rate": 0.00012178413475110567,
      "loss": 0.4358,
      "step": 6840
    },
    {
      "epoch": 1.85,
      "grad_norm": 17.070862403868173,
      "learning_rate": 0.00012150183494871551,
      "loss": 0.2846,
      "step": 6850
    },
    {
      "epoch": 1.85,
      "grad_norm": 19.62765383297612,
      "learning_rate": 0.00012121953514632538,
      "loss": 0.4156,
      "step": 6860
    },
    {
      "epoch": 1.85,
      "grad_norm": 23.416994659679506,
      "learning_rate": 0.00012093723534393524,
      "loss": 0.4169,
      "step": 6870
    },
    {
      "epoch": 1.85,
      "grad_norm": 14.044584196230364,
      "learning_rate": 0.00012065493554154511,
      "loss": 0.3994,
      "step": 6880
    },
    {
      "epoch": 1.86,
      "grad_norm": 6.336997600233475,
      "learning_rate": 0.00012037263573915497,
      "loss": 0.3229,
      "step": 6890
    },
    {
      "epoch": 1.86,
      "grad_norm": 17.869630678040608,
      "learning_rate": 0.00012009033593676483,
      "loss": 0.2768,
      "step": 6900
    },
    {
      "epoch": 1.86,
      "grad_norm": 15.213741214063417,
      "learning_rate": 0.0001198080361343747,
      "loss": 0.2625,
      "step": 6910
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.7797230199444007,
      "learning_rate": 0.00011952573633198455,
      "loss": 0.2828,
      "step": 6920
    },
    {
      "epoch": 1.87,
      "grad_norm": 25.07771775934045,
      "learning_rate": 0.00011924343652959442,
      "loss": 0.3148,
      "step": 6930
    },
    {
      "epoch": 1.87,
      "grad_norm": 14.536216177683698,
      "learning_rate": 0.00011896113672720428,
      "loss": 0.3141,
      "step": 6940
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.6206559920847314,
      "learning_rate": 0.00011867883692481415,
      "loss": 0.261,
      "step": 6950
    },
    {
      "epoch": 1.88,
      "grad_norm": 29.158284398968,
      "learning_rate": 0.000118396537122424,
      "loss": 0.2958,
      "step": 6960
    },
    {
      "epoch": 1.88,
      "grad_norm": 5.5940817447872275,
      "learning_rate": 0.00011811423732003388,
      "loss": 0.279,
      "step": 6970
    },
    {
      "epoch": 1.88,
      "grad_norm": 41.57670091228017,
      "learning_rate": 0.00011783193751764372,
      "loss": 0.4383,
      "step": 6980
    },
    {
      "epoch": 1.88,
      "grad_norm": 38.11900012535772,
      "learning_rate": 0.00011754963771525359,
      "loss": 0.3349,
      "step": 6990
    },
    {
      "epoch": 1.89,
      "grad_norm": 13.920668234472188,
      "learning_rate": 0.00011726733791286345,
      "loss": 0.2396,
      "step": 7000
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.418657253086558,
      "learning_rate": 0.0001169850381104733,
      "loss": 0.3686,
      "step": 7010
    },
    {
      "epoch": 1.89,
      "grad_norm": 13.054131127974125,
      "learning_rate": 0.00011673096828832218,
      "loss": 0.4592,
      "step": 7020
    },
    {
      "epoch": 1.9,
      "grad_norm": 23.97883943334597,
      "learning_rate": 0.00011644866848593205,
      "loss": 0.4279,
      "step": 7030
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.730353287307535,
      "learning_rate": 0.0001161663686835419,
      "loss": 0.2918,
      "step": 7040
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.724102355660204,
      "learning_rate": 0.00011588406888115178,
      "loss": 0.2832,
      "step": 7050
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.968254756751215,
      "learning_rate": 0.00011560176907876163,
      "loss": 0.2854,
      "step": 7060
    },
    {
      "epoch": 1.91,
      "grad_norm": 25.52731609795378,
      "learning_rate": 0.0001153194692763715,
      "loss": 0.2437,
      "step": 7070
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.448185220108101,
      "learning_rate": 0.00011503716947398136,
      "loss": 0.3057,
      "step": 7080
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.344334427933555,
      "learning_rate": 0.00011475486967159123,
      "loss": 0.4164,
      "step": 7090
    },
    {
      "epoch": 1.91,
      "grad_norm": 5.6676470822502365,
      "learning_rate": 0.00011447256986920107,
      "loss": 0.2009,
      "step": 7100
    },
    {
      "epoch": 1.92,
      "grad_norm": 6.755283925218195,
      "learning_rate": 0.00011419027006681093,
      "loss": 0.216,
      "step": 7110
    },
    {
      "epoch": 1.92,
      "grad_norm": 4.335738589475496,
      "learning_rate": 0.0001139079702644208,
      "loss": 0.4402,
      "step": 7120
    },
    {
      "epoch": 1.92,
      "grad_norm": 29.885236155639696,
      "learning_rate": 0.00011362567046203066,
      "loss": 0.2527,
      "step": 7130
    },
    {
      "epoch": 1.93,
      "grad_norm": 24.819104333037473,
      "learning_rate": 0.00011334337065964053,
      "loss": 0.2838,
      "step": 7140
    },
    {
      "epoch": 1.93,
      "grad_norm": 19.29574056313953,
      "learning_rate": 0.00011306107085725039,
      "loss": 0.2246,
      "step": 7150
    },
    {
      "epoch": 1.93,
      "grad_norm": 25.55211772581389,
      "learning_rate": 0.00011277877105486026,
      "loss": 0.3288,
      "step": 7160
    },
    {
      "epoch": 1.93,
      "grad_norm": 14.8672158481049,
      "learning_rate": 0.00011249647125247011,
      "loss": 0.2481,
      "step": 7170
    },
    {
      "epoch": 1.94,
      "grad_norm": 34.62791225902285,
      "learning_rate": 0.00011221417145007998,
      "loss": 0.3952,
      "step": 7180
    },
    {
      "epoch": 1.94,
      "grad_norm": 43.171566592774084,
      "learning_rate": 0.00011193187164768984,
      "loss": 0.3453,
      "step": 7190
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.8731009219251185,
      "learning_rate": 0.00011164957184529971,
      "loss": 0.2825,
      "step": 7200
    },
    {
      "epoch": 1.94,
      "grad_norm": 5.045083006418917,
      "learning_rate": 0.00011136727204290957,
      "loss": 0.3124,
      "step": 7210
    },
    {
      "epoch": 1.95,
      "grad_norm": 32.265338792293534,
      "learning_rate": 0.00011108497224051941,
      "loss": 0.306,
      "step": 7220
    },
    {
      "epoch": 1.95,
      "grad_norm": 16.01856842649736,
      "learning_rate": 0.00011080267243812928,
      "loss": 0.3164,
      "step": 7230
    },
    {
      "epoch": 1.95,
      "grad_norm": 6.962096184354874,
      "learning_rate": 0.00011052037263573914,
      "loss": 0.3313,
      "step": 7240
    },
    {
      "epoch": 1.95,
      "grad_norm": 22.768897780243268,
      "learning_rate": 0.00011023807283334901,
      "loss": 0.2166,
      "step": 7250
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.721022347119176,
      "learning_rate": 0.00010995577303095887,
      "loss": 0.3317,
      "step": 7260
    },
    {
      "epoch": 1.96,
      "grad_norm": 5.388162646752015,
      "learning_rate": 0.00010967347322856874,
      "loss": 0.417,
      "step": 7270
    },
    {
      "epoch": 1.96,
      "grad_norm": 5.386294873758332,
      "learning_rate": 0.00010939117342617859,
      "loss": 0.234,
      "step": 7280
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.00479483710244,
      "learning_rate": 0.00010910887362378846,
      "loss": 0.4107,
      "step": 7290
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.154282077216171,
      "learning_rate": 0.00010882657382139832,
      "loss": 0.3908,
      "step": 7300
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.712886479198343,
      "learning_rate": 0.00010854427401900816,
      "loss": 0.2448,
      "step": 7310
    },
    {
      "epoch": 1.97,
      "grad_norm": 48.38391260663932,
      "learning_rate": 0.00010826197421661805,
      "loss": 0.3793,
      "step": 7320
    },
    {
      "epoch": 1.98,
      "grad_norm": 24.863047351620974,
      "learning_rate": 0.00010797967441422789,
      "loss": 0.3495,
      "step": 7330
    },
    {
      "epoch": 1.98,
      "grad_norm": 23.485028882323878,
      "learning_rate": 0.00010769737461183776,
      "loss": 0.2123,
      "step": 7340
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.08205774807799,
      "learning_rate": 0.00010741507480944762,
      "loss": 0.411,
      "step": 7350
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.906563746051484,
      "learning_rate": 0.00010713277500705749,
      "loss": 0.2039,
      "step": 7360
    },
    {
      "epoch": 1.99,
      "grad_norm": 13.472862669136846,
      "learning_rate": 0.00010685047520466735,
      "loss": 0.2325,
      "step": 7370
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.890045916984588,
      "learning_rate": 0.00010656817540227722,
      "loss": 0.2785,
      "step": 7380
    },
    {
      "epoch": 1.99,
      "grad_norm": 24.972681410401044,
      "learning_rate": 0.00010628587559988707,
      "loss": 0.3837,
      "step": 7390
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.948880070760067,
      "learning_rate": 0.00010600357579749693,
      "loss": 0.3439,
      "step": 7400
    },
    {
      "epoch": 2.0,
      "grad_norm": 19.340204441249025,
      "learning_rate": 0.0001057212759951068,
      "loss": 0.3556,
      "step": 7410
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8549412138544645,
      "eval_f1": 0.9081581329846091,
      "eval_loss": 0.34521484375,
      "eval_precision": 0.8544387658527351,
      "eval_recall": 0.9690854443967368,
      "eval_runtime": 427.5228,
      "eval_samples_per_second": 14.722,
      "eval_steps_per_second": 2.454,
      "step": 7418
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.723244274845054,
      "learning_rate": 0.00010543897619271666,
      "loss": 0.3455,
      "step": 7420
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.16853889194249,
      "learning_rate": 0.00010515667639032653,
      "loss": 0.2089,
      "step": 7430
    },
    {
      "epoch": 2.01,
      "grad_norm": 5.422759792081017,
      "learning_rate": 0.00010487437658793637,
      "loss": 0.1977,
      "step": 7440
    },
    {
      "epoch": 2.01,
      "grad_norm": 19.42619061187327,
      "learning_rate": 0.00010459207678554624,
      "loss": 0.2121,
      "step": 7450
    },
    {
      "epoch": 2.01,
      "grad_norm": 17.997869568736395,
      "learning_rate": 0.0001043097769831561,
      "loss": 0.1757,
      "step": 7460
    },
    {
      "epoch": 2.01,
      "grad_norm": 26.31859192584454,
      "learning_rate": 0.00010402747718076597,
      "loss": 0.2035,
      "step": 7470
    },
    {
      "epoch": 2.02,
      "grad_norm": 12.579852814500835,
      "learning_rate": 0.00010374517737837583,
      "loss": 0.256,
      "step": 7480
    },
    {
      "epoch": 2.02,
      "grad_norm": 3.7389506877410303,
      "learning_rate": 0.00010346287757598568,
      "loss": 0.2384,
      "step": 7490
    },
    {
      "epoch": 2.02,
      "grad_norm": 26.777645798864,
      "learning_rate": 0.00010318057777359555,
      "loss": 0.28,
      "step": 7500
    },
    {
      "epoch": 2.02,
      "grad_norm": 9.15273688321073,
      "learning_rate": 0.00010289827797120541,
      "loss": 0.0848,
      "step": 7510
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.68582334554762,
      "learning_rate": 0.00010261597816881528,
      "loss": 0.2435,
      "step": 7520
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.799878716280141,
      "learning_rate": 0.00010233367836642514,
      "loss": 0.2127,
      "step": 7530
    },
    {
      "epoch": 2.03,
      "grad_norm": 6.288007417429413,
      "learning_rate": 0.00010205137856403501,
      "loss": 0.1981,
      "step": 7540
    },
    {
      "epoch": 2.04,
      "grad_norm": 6.454560033693452,
      "learning_rate": 0.00010176907876164485,
      "loss": 0.2222,
      "step": 7550
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.7921625834528974,
      "learning_rate": 0.00010148677895925472,
      "loss": 0.2104,
      "step": 7560
    },
    {
      "epoch": 2.04,
      "grad_norm": 7.286009596122917,
      "learning_rate": 0.00010120447915686458,
      "loss": 0.1387,
      "step": 7570
    },
    {
      "epoch": 2.04,
      "grad_norm": 14.572241977539019,
      "learning_rate": 0.00010092217935447444,
      "loss": 0.1972,
      "step": 7580
    },
    {
      "epoch": 2.05,
      "grad_norm": 25.14607434399571,
      "learning_rate": 0.0001006398795520843,
      "loss": 0.3118,
      "step": 7590
    },
    {
      "epoch": 2.05,
      "grad_norm": 40.80634867810613,
      "learning_rate": 0.00010035757974969416,
      "loss": 0.16,
      "step": 7600
    },
    {
      "epoch": 2.05,
      "grad_norm": 20.651895750542913,
      "learning_rate": 0.00010007527994730403,
      "loss": 0.3167,
      "step": 7610
    },
    {
      "epoch": 2.05,
      "grad_norm": 22.010411340968915,
      "learning_rate": 9.979298014491389e-05,
      "loss": 0.2389,
      "step": 7620
    },
    {
      "epoch": 2.06,
      "grad_norm": 14.427587078494692,
      "learning_rate": 9.951068034252376e-05,
      "loss": 0.2054,
      "step": 7630
    },
    {
      "epoch": 2.06,
      "grad_norm": 11.052323139074652,
      "learning_rate": 9.922838054013362e-05,
      "loss": 0.1148,
      "step": 7640
    },
    {
      "epoch": 2.06,
      "grad_norm": 21.70252188496123,
      "learning_rate": 9.894608073774349e-05,
      "loss": 0.1483,
      "step": 7650
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.8615429333823155,
      "learning_rate": 9.866378093535333e-05,
      "loss": 0.1826,
      "step": 7660
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.41150856470262626,
      "learning_rate": 9.838148113296319e-05,
      "loss": 0.1684,
      "step": 7670
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.361893017883698,
      "learning_rate": 9.809918133057306e-05,
      "loss": 0.1706,
      "step": 7680
    },
    {
      "epoch": 2.07,
      "grad_norm": 16.724504132592877,
      "learning_rate": 9.781688152818292e-05,
      "loss": 0.2716,
      "step": 7690
    },
    {
      "epoch": 2.08,
      "grad_norm": 11.605647746002251,
      "learning_rate": 9.753458172579279e-05,
      "loss": 0.1769,
      "step": 7700
    },
    {
      "epoch": 2.08,
      "grad_norm": 11.95173779740697,
      "learning_rate": 9.725228192340264e-05,
      "loss": 0.223,
      "step": 7710
    },
    {
      "epoch": 2.08,
      "grad_norm": 39.366092445600415,
      "learning_rate": 9.696998212101251e-05,
      "loss": 0.1932,
      "step": 7720
    },
    {
      "epoch": 2.08,
      "grad_norm": 13.034823100468513,
      "learning_rate": 9.668768231862237e-05,
      "loss": 0.183,
      "step": 7730
    },
    {
      "epoch": 2.09,
      "grad_norm": 25.362381286846187,
      "learning_rate": 9.640538251623224e-05,
      "loss": 0.197,
      "step": 7740
    },
    {
      "epoch": 2.09,
      "grad_norm": 12.40273461694631,
      "learning_rate": 9.61230827138421e-05,
      "loss": 0.2344,
      "step": 7750
    },
    {
      "epoch": 2.09,
      "grad_norm": 9.195485514906697,
      "learning_rate": 9.584078291145194e-05,
      "loss": 0.1969,
      "step": 7760
    },
    {
      "epoch": 2.09,
      "grad_norm": 35.871940678399206,
      "learning_rate": 9.555848310906181e-05,
      "loss": 0.1643,
      "step": 7770
    },
    {
      "epoch": 2.1,
      "grad_norm": 9.51946228446132,
      "learning_rate": 9.527618330667167e-05,
      "loss": 0.1745,
      "step": 7780
    },
    {
      "epoch": 2.1,
      "grad_norm": 9.446477468136544,
      "learning_rate": 9.499388350428154e-05,
      "loss": 0.099,
      "step": 7790
    },
    {
      "epoch": 2.1,
      "grad_norm": 13.91590805440452,
      "learning_rate": 9.47115837018914e-05,
      "loss": 0.2781,
      "step": 7800
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.2990579540688123,
      "learning_rate": 9.442928389950127e-05,
      "loss": 0.1253,
      "step": 7810
    },
    {
      "epoch": 2.11,
      "grad_norm": 10.810720668040153,
      "learning_rate": 9.414698409711112e-05,
      "loss": 0.1673,
      "step": 7820
    },
    {
      "epoch": 2.11,
      "grad_norm": 3.281130360511212,
      "learning_rate": 9.386468429472099e-05,
      "loss": 0.16,
      "step": 7830
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.886152880481387,
      "learning_rate": 9.358238449233085e-05,
      "loss": 0.2655,
      "step": 7840
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.0090805743593276,
      "learning_rate": 9.33000846899407e-05,
      "loss": 0.2102,
      "step": 7850
    },
    {
      "epoch": 2.12,
      "grad_norm": 3.961473159995717,
      "learning_rate": 9.301778488755058e-05,
      "loss": 0.2552,
      "step": 7860
    },
    {
      "epoch": 2.12,
      "grad_norm": 14.507109212853333,
      "learning_rate": 9.273548508516042e-05,
      "loss": 0.2163,
      "step": 7870
    },
    {
      "epoch": 2.12,
      "grad_norm": 15.399000128052995,
      "learning_rate": 9.24531852827703e-05,
      "loss": 0.2561,
      "step": 7880
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.6419324303897436,
      "learning_rate": 9.217088548038015e-05,
      "loss": 0.1265,
      "step": 7890
    },
    {
      "epoch": 2.13,
      "grad_norm": 7.378017406838396,
      "learning_rate": 9.188858567799002e-05,
      "loss": 0.2252,
      "step": 7900
    },
    {
      "epoch": 2.13,
      "grad_norm": 18.19499551047343,
      "learning_rate": 9.160628587559988e-05,
      "loss": 0.1946,
      "step": 7910
    },
    {
      "epoch": 2.14,
      "grad_norm": 32.441081222384994,
      "learning_rate": 9.132398607320975e-05,
      "loss": 0.1592,
      "step": 7920
    },
    {
      "epoch": 2.14,
      "grad_norm": 5.5522130640295595,
      "learning_rate": 9.10416862708196e-05,
      "loss": 0.1564,
      "step": 7930
    },
    {
      "epoch": 2.14,
      "grad_norm": 13.58027451657248,
      "learning_rate": 9.075938646842946e-05,
      "loss": 0.2716,
      "step": 7940
    },
    {
      "epoch": 2.14,
      "grad_norm": 26.981624580170763,
      "learning_rate": 9.047708666603933e-05,
      "loss": 0.2039,
      "step": 7950
    },
    {
      "epoch": 2.15,
      "grad_norm": 18.129241953402463,
      "learning_rate": 9.019478686364919e-05,
      "loss": 0.224,
      "step": 7960
    },
    {
      "epoch": 2.15,
      "grad_norm": 12.284683162634781,
      "learning_rate": 8.991248706125906e-05,
      "loss": 0.1508,
      "step": 7970
    },
    {
      "epoch": 2.15,
      "grad_norm": 37.3547301075289,
      "learning_rate": 8.963018725886891e-05,
      "loss": 0.1445,
      "step": 7980
    },
    {
      "epoch": 2.15,
      "grad_norm": 39.6216403848631,
      "learning_rate": 8.937611743671779e-05,
      "loss": 0.6445,
      "step": 7990
    },
    {
      "epoch": 2.16,
      "grad_norm": 49.93745729470844,
      "learning_rate": 8.909381763432766e-05,
      "loss": 0.2535,
      "step": 8000
    },
    {
      "epoch": 2.16,
      "grad_norm": 9.281246719135,
      "learning_rate": 8.88115178319375e-05,
      "loss": 0.2237,
      "step": 8010
    },
    {
      "epoch": 2.16,
      "grad_norm": 3.8630840453198174,
      "learning_rate": 8.852921802954737e-05,
      "loss": 0.1242,
      "step": 8020
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.901497407746327,
      "learning_rate": 8.824691822715723e-05,
      "loss": 0.2972,
      "step": 8030
    },
    {
      "epoch": 2.17,
      "grad_norm": 6.993508046936159,
      "learning_rate": 8.79646184247671e-05,
      "loss": 0.129,
      "step": 8040
    },
    {
      "epoch": 2.17,
      "grad_norm": 11.517563413145163,
      "learning_rate": 8.768231862237695e-05,
      "loss": 0.3421,
      "step": 8050
    },
    {
      "epoch": 2.17,
      "grad_norm": 9.217098876619541,
      "learning_rate": 8.740001881998681e-05,
      "loss": 0.2737,
      "step": 8060
    },
    {
      "epoch": 2.18,
      "grad_norm": 26.078133211654546,
      "learning_rate": 8.711771901759668e-05,
      "loss": 0.2713,
      "step": 8070
    },
    {
      "epoch": 2.18,
      "grad_norm": 22.707191835193587,
      "learning_rate": 8.683541921520654e-05,
      "loss": 0.3101,
      "step": 8080
    },
    {
      "epoch": 2.18,
      "grad_norm": 40.74370380772914,
      "learning_rate": 8.655311941281641e-05,
      "loss": 0.2477,
      "step": 8090
    },
    {
      "epoch": 2.18,
      "grad_norm": 8.082782079234564,
      "learning_rate": 8.627081961042627e-05,
      "loss": 0.2329,
      "step": 8100
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.0152799192875965,
      "learning_rate": 8.598851980803614e-05,
      "loss": 0.2363,
      "step": 8110
    },
    {
      "epoch": 2.19,
      "grad_norm": 12.849478401610147,
      "learning_rate": 8.570622000564598e-05,
      "loss": 0.2192,
      "step": 8120
    },
    {
      "epoch": 2.19,
      "grad_norm": 14.619877627849013,
      "learning_rate": 8.542392020325586e-05,
      "loss": 0.1077,
      "step": 8130
    },
    {
      "epoch": 2.19,
      "grad_norm": 10.43935206964753,
      "learning_rate": 8.514162040086571e-05,
      "loss": 0.2379,
      "step": 8140
    },
    {
      "epoch": 2.2,
      "grad_norm": 5.4256059506373795,
      "learning_rate": 8.485932059847556e-05,
      "loss": 0.2218,
      "step": 8150
    },
    {
      "epoch": 2.2,
      "grad_norm": 12.093211642562217,
      "learning_rate": 8.457702079608543e-05,
      "loss": 0.2457,
      "step": 8160
    },
    {
      "epoch": 2.2,
      "grad_norm": 16.354860348224058,
      "learning_rate": 8.429472099369529e-05,
      "loss": 0.2247,
      "step": 8170
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.5520233692411742,
      "learning_rate": 8.401242119130516e-05,
      "loss": 0.1538,
      "step": 8180
    },
    {
      "epoch": 2.21,
      "grad_norm": 8.585531353130497,
      "learning_rate": 8.373012138891502e-05,
      "loss": 0.1736,
      "step": 8190
    },
    {
      "epoch": 2.21,
      "grad_norm": 12.474459202120673,
      "learning_rate": 8.344782158652489e-05,
      "loss": 0.205,
      "step": 8200
    },
    {
      "epoch": 2.21,
      "grad_norm": 23.939423155160902,
      "learning_rate": 8.316552178413475e-05,
      "loss": 0.134,
      "step": 8210
    },
    {
      "epoch": 2.22,
      "grad_norm": 29.740031177981248,
      "learning_rate": 8.288322198174462e-05,
      "loss": 0.23,
      "step": 8220
    },
    {
      "epoch": 2.22,
      "grad_norm": 16.583875426420988,
      "learning_rate": 8.260092217935447e-05,
      "loss": 0.2579,
      "step": 8230
    },
    {
      "epoch": 2.22,
      "grad_norm": 22.476009842088295,
      "learning_rate": 8.231862237696432e-05,
      "loss": 0.1525,
      "step": 8240
    },
    {
      "epoch": 2.22,
      "grad_norm": 3.0210689295744753,
      "learning_rate": 8.203632257457419e-05,
      "loss": 0.2026,
      "step": 8250
    },
    {
      "epoch": 2.23,
      "grad_norm": 14.594915893796864,
      "learning_rate": 8.175402277218404e-05,
      "loss": 0.1211,
      "step": 8260
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.029259758750742,
      "learning_rate": 8.147172296979391e-05,
      "loss": 0.1186,
      "step": 8270
    },
    {
      "epoch": 2.23,
      "grad_norm": 9.331752354976059,
      "learning_rate": 8.118942316740377e-05,
      "loss": 0.1797,
      "step": 8280
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.1439615113585182,
      "learning_rate": 8.090712336501364e-05,
      "loss": 0.2435,
      "step": 8290
    },
    {
      "epoch": 2.24,
      "grad_norm": 15.36118367364266,
      "learning_rate": 8.06248235626235e-05,
      "loss": 0.3038,
      "step": 8300
    },
    {
      "epoch": 2.24,
      "grad_norm": 15.744595890539959,
      "learning_rate": 8.034252376023337e-05,
      "loss": 0.1655,
      "step": 8310
    },
    {
      "epoch": 2.24,
      "grad_norm": 4.635708818622949,
      "learning_rate": 8.006022395784323e-05,
      "loss": 0.1237,
      "step": 8320
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.271395537434196,
      "learning_rate": 7.977792415545308e-05,
      "loss": 0.1836,
      "step": 8330
    },
    {
      "epoch": 2.25,
      "grad_norm": 3.651028687441684,
      "learning_rate": 7.949562435306295e-05,
      "loss": 0.1302,
      "step": 8340
    },
    {
      "epoch": 2.25,
      "grad_norm": 12.805758102165951,
      "learning_rate": 7.92133245506728e-05,
      "loss": 0.1828,
      "step": 8350
    },
    {
      "epoch": 2.25,
      "grad_norm": 28.030990488813735,
      "learning_rate": 7.893102474828267e-05,
      "loss": 0.2764,
      "step": 8360
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.4671465302039572,
      "learning_rate": 7.864872494589252e-05,
      "loss": 0.196,
      "step": 8370
    },
    {
      "epoch": 2.26,
      "grad_norm": 29.782817515122694,
      "learning_rate": 7.83664251435024e-05,
      "loss": 0.2831,
      "step": 8380
    },
    {
      "epoch": 2.26,
      "grad_norm": 16.623228635747815,
      "learning_rate": 7.808412534111225e-05,
      "loss": 0.1449,
      "step": 8390
    },
    {
      "epoch": 2.26,
      "grad_norm": 4.313442382521624,
      "learning_rate": 7.780182553872212e-05,
      "loss": 0.1477,
      "step": 8400
    },
    {
      "epoch": 2.27,
      "grad_norm": 4.708754105675319,
      "learning_rate": 7.751952573633198e-05,
      "loss": 0.1272,
      "step": 8410
    },
    {
      "epoch": 2.27,
      "grad_norm": 15.836657061270156,
      "learning_rate": 7.723722593394184e-05,
      "loss": 0.3165,
      "step": 8420
    },
    {
      "epoch": 2.27,
      "grad_norm": 11.659146455269608,
      "learning_rate": 7.69549261315517e-05,
      "loss": 0.1292,
      "step": 8430
    },
    {
      "epoch": 2.28,
      "grad_norm": 24.34236147792855,
      "learning_rate": 7.667262632916156e-05,
      "loss": 0.2143,
      "step": 8440
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.6476955422629826,
      "learning_rate": 7.639032652677143e-05,
      "loss": 0.284,
      "step": 8450
    },
    {
      "epoch": 2.28,
      "grad_norm": 8.054019705121016,
      "learning_rate": 7.610802672438128e-05,
      "loss": 0.2698,
      "step": 8460
    },
    {
      "epoch": 2.28,
      "grad_norm": 32.4164668377762,
      "learning_rate": 7.582572692199115e-05,
      "loss": 0.1807,
      "step": 8470
    },
    {
      "epoch": 2.29,
      "grad_norm": 15.519147875390773,
      "learning_rate": 7.5543427119601e-05,
      "loss": 0.1457,
      "step": 8480
    },
    {
      "epoch": 2.29,
      "grad_norm": 19.28565869556721,
      "learning_rate": 7.526112731721087e-05,
      "loss": 0.0889,
      "step": 8490
    },
    {
      "epoch": 2.29,
      "grad_norm": 13.868996924969483,
      "learning_rate": 7.497882751482073e-05,
      "loss": 0.2915,
      "step": 8500
    },
    {
      "epoch": 2.29,
      "grad_norm": 29.25053890762934,
      "learning_rate": 7.46965277124306e-05,
      "loss": 0.281,
      "step": 8510
    },
    {
      "epoch": 2.3,
      "grad_norm": 3.8398062703194342,
      "learning_rate": 7.441422791004046e-05,
      "loss": 0.1904,
      "step": 8520
    },
    {
      "epoch": 2.3,
      "grad_norm": 6.702056107349894,
      "learning_rate": 7.413192810765032e-05,
      "loss": 0.1684,
      "step": 8530
    },
    {
      "epoch": 2.3,
      "grad_norm": 10.665714891500867,
      "learning_rate": 7.384962830526019e-05,
      "loss": 0.1817,
      "step": 8540
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.3785745006196886,
      "learning_rate": 7.356732850287004e-05,
      "loss": 0.1782,
      "step": 8550
    },
    {
      "epoch": 2.31,
      "grad_norm": 23.19263867866707,
      "learning_rate": 7.32850287004799e-05,
      "loss": 0.1045,
      "step": 8560
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.3374503226413847,
      "learning_rate": 7.300272889808976e-05,
      "loss": 0.3311,
      "step": 8570
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.073517991936664,
      "learning_rate": 7.272042909569963e-05,
      "loss": 0.2787,
      "step": 8580
    },
    {
      "epoch": 2.32,
      "grad_norm": 19.483387625714556,
      "learning_rate": 7.243812929330948e-05,
      "loss": 0.2188,
      "step": 8590
    },
    {
      "epoch": 2.32,
      "grad_norm": 5.766091888669735,
      "learning_rate": 7.215582949091935e-05,
      "loss": 0.356,
      "step": 8600
    },
    {
      "epoch": 2.32,
      "grad_norm": 5.845763323748258,
      "learning_rate": 7.187352968852921e-05,
      "loss": 0.226,
      "step": 8610
    },
    {
      "epoch": 2.32,
      "grad_norm": 10.899759273976844,
      "learning_rate": 7.159122988613908e-05,
      "loss": 0.163,
      "step": 8620
    },
    {
      "epoch": 2.33,
      "grad_norm": 10.863244494211727,
      "learning_rate": 7.130893008374894e-05,
      "loss": 0.1371,
      "step": 8630
    },
    {
      "epoch": 2.33,
      "grad_norm": 24.32463909329357,
      "learning_rate": 7.10266302813588e-05,
      "loss": 0.3021,
      "step": 8640
    },
    {
      "epoch": 2.33,
      "grad_norm": 9.192550443674804,
      "learning_rate": 7.074433047896865e-05,
      "loss": 0.2461,
      "step": 8650
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.6913341121147463,
      "learning_rate": 7.046203067657852e-05,
      "loss": 0.192,
      "step": 8660
    },
    {
      "epoch": 2.34,
      "grad_norm": 5.5175612894155845,
      "learning_rate": 7.017973087418838e-05,
      "loss": 0.0732,
      "step": 8670
    },
    {
      "epoch": 2.34,
      "grad_norm": 19.889042961452052,
      "learning_rate": 6.989743107179824e-05,
      "loss": 0.1859,
      "step": 8680
    },
    {
      "epoch": 2.34,
      "grad_norm": 3.9255731544094803,
      "learning_rate": 6.961513126940811e-05,
      "loss": 0.0944,
      "step": 8690
    },
    {
      "epoch": 2.35,
      "grad_norm": 6.7705550624750845,
      "learning_rate": 6.933283146701796e-05,
      "loss": 0.0849,
      "step": 8700
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.8577361896389244,
      "learning_rate": 6.905053166462783e-05,
      "loss": 0.1595,
      "step": 8710
    },
    {
      "epoch": 2.35,
      "grad_norm": 26.937360489919914,
      "learning_rate": 6.876823186223769e-05,
      "loss": 0.295,
      "step": 8720
    },
    {
      "epoch": 2.35,
      "grad_norm": 6.2573601372291066,
      "learning_rate": 6.848593205984755e-05,
      "loss": 0.2483,
      "step": 8730
    },
    {
      "epoch": 2.36,
      "grad_norm": 7.0698577081242915,
      "learning_rate": 6.82036322574574e-05,
      "loss": 0.0702,
      "step": 8740
    },
    {
      "epoch": 2.36,
      "grad_norm": 7.817396010721897,
      "learning_rate": 6.792133245506728e-05,
      "loss": 0.1945,
      "step": 8750
    },
    {
      "epoch": 2.36,
      "grad_norm": 4.573118276585261,
      "learning_rate": 6.763903265267713e-05,
      "loss": 0.3672,
      "step": 8760
    },
    {
      "epoch": 2.36,
      "grad_norm": 5.071169251088683,
      "learning_rate": 6.7356732850287e-05,
      "loss": 0.1727,
      "step": 8770
    },
    {
      "epoch": 2.37,
      "grad_norm": 17.68202468284735,
      "learning_rate": 6.707443304789686e-05,
      "loss": 0.1717,
      "step": 8780
    },
    {
      "epoch": 2.37,
      "grad_norm": 4.7449338864409745,
      "learning_rate": 6.679213324550673e-05,
      "loss": 0.349,
      "step": 8790
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.713403039480962,
      "learning_rate": 6.650983344311659e-05,
      "loss": 0.2866,
      "step": 8800
    },
    {
      "epoch": 2.38,
      "grad_norm": 10.406972988891406,
      "learning_rate": 6.622753364072644e-05,
      "loss": 0.248,
      "step": 8810
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.3713631353787425,
      "learning_rate": 6.59452338383363e-05,
      "loss": 0.1991,
      "step": 8820
    },
    {
      "epoch": 2.38,
      "grad_norm": 17.269832655479554,
      "learning_rate": 6.566293403594617e-05,
      "loss": 0.2524,
      "step": 8830
    },
    {
      "epoch": 2.38,
      "grad_norm": 10.732000712356614,
      "learning_rate": 6.538063423355603e-05,
      "loss": 0.2177,
      "step": 8840
    },
    {
      "epoch": 2.39,
      "grad_norm": 11.817548465221842,
      "learning_rate": 6.509833443116589e-05,
      "loss": 0.2166,
      "step": 8850
    },
    {
      "epoch": 2.39,
      "grad_norm": 12.116283420528262,
      "learning_rate": 6.481603462877576e-05,
      "loss": 0.3095,
      "step": 8860
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.044224812034278,
      "learning_rate": 6.453373482638561e-05,
      "loss": 0.0856,
      "step": 8870
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.7338570804353683,
      "learning_rate": 6.425143502399548e-05,
      "loss": 0.1738,
      "step": 8880
    },
    {
      "epoch": 2.4,
      "grad_norm": 59.29371350571651,
      "learning_rate": 6.396913522160534e-05,
      "loss": 0.2815,
      "step": 8890
    },
    {
      "epoch": 2.4,
      "grad_norm": 12.430986995549578,
      "learning_rate": 6.368683541921521e-05,
      "loss": 0.1338,
      "step": 8900
    },
    {
      "epoch": 2.4,
      "grad_norm": 16.249312643039122,
      "learning_rate": 6.340453561682507e-05,
      "loss": 0.2039,
      "step": 8910
    },
    {
      "epoch": 2.4,
      "grad_norm": 20.615081771365276,
      "learning_rate": 6.312223581443492e-05,
      "loss": 0.1738,
      "step": 8920
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.39293086744558864,
      "learning_rate": 6.283993601204478e-05,
      "loss": 0.2138,
      "step": 8930
    },
    {
      "epoch": 2.41,
      "grad_norm": 8.643827429956966,
      "learning_rate": 6.255763620965465e-05,
      "loss": 0.1346,
      "step": 8940
    },
    {
      "epoch": 2.41,
      "grad_norm": 6.813572272757125,
      "learning_rate": 6.227533640726451e-05,
      "loss": 0.2404,
      "step": 8950
    },
    {
      "epoch": 2.42,
      "grad_norm": 43.575713470142,
      "learning_rate": 6.199303660487437e-05,
      "loss": 0.2759,
      "step": 8960
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.3712824021091958,
      "learning_rate": 6.171073680248424e-05,
      "loss": 0.1192,
      "step": 8970
    },
    {
      "epoch": 2.42,
      "grad_norm": 11.86747338821422,
      "learning_rate": 6.142843700009409e-05,
      "loss": 0.3636,
      "step": 8980
    },
    {
      "epoch": 2.42,
      "grad_norm": 22.59628679285179,
      "learning_rate": 6.114613719770396e-05,
      "loss": 0.1625,
      "step": 8990
    },
    {
      "epoch": 2.43,
      "grad_norm": 16.666426390366286,
      "learning_rate": 6.086383739531382e-05,
      "loss": 0.2393,
      "step": 9000
    },
    {
      "epoch": 2.43,
      "grad_norm": 7.240556201543136,
      "learning_rate": 6.058153759292368e-05,
      "loss": 0.2352,
      "step": 9010
    },
    {
      "epoch": 2.43,
      "grad_norm": 6.931845322566829,
      "learning_rate": 6.029923779053354e-05,
      "loss": 0.2252,
      "step": 9020
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.2742740367700476,
      "learning_rate": 6.0016937988143404e-05,
      "loss": 0.1952,
      "step": 9030
    },
    {
      "epoch": 2.44,
      "grad_norm": 22.612490742477497,
      "learning_rate": 5.973463818575326e-05,
      "loss": 0.306,
      "step": 9040
    },
    {
      "epoch": 2.44,
      "grad_norm": 12.879488471352738,
      "learning_rate": 5.9452338383363125e-05,
      "loss": 0.1551,
      "step": 9050
    },
    {
      "epoch": 2.44,
      "grad_norm": 3.303554844135739,
      "learning_rate": 5.917003858097299e-05,
      "loss": 0.1483,
      "step": 9060
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.9841625163689645,
      "learning_rate": 5.888773877858285e-05,
      "loss": 0.1445,
      "step": 9070
    },
    {
      "epoch": 2.45,
      "grad_norm": 7.823629490159578,
      "learning_rate": 5.8605438976192716e-05,
      "loss": 0.2088,
      "step": 9080
    },
    {
      "epoch": 2.45,
      "grad_norm": 53.16384475966128,
      "learning_rate": 5.832313917380258e-05,
      "loss": 0.2083,
      "step": 9090
    },
    {
      "epoch": 2.45,
      "grad_norm": 3.2785637596533572,
      "learning_rate": 5.804083937141243e-05,
      "loss": 0.3154,
      "step": 9100
    },
    {
      "epoch": 2.46,
      "grad_norm": 4.9872522233721135,
      "learning_rate": 5.775853956902229e-05,
      "loss": 0.1712,
      "step": 9110
    },
    {
      "epoch": 2.46,
      "grad_norm": 10.326205706121568,
      "learning_rate": 5.747623976663216e-05,
      "loss": 0.2558,
      "step": 9120
    },
    {
      "epoch": 2.46,
      "grad_norm": 24.96547257859977,
      "learning_rate": 5.719393996424202e-05,
      "loss": 0.1651,
      "step": 9130
    },
    {
      "epoch": 2.46,
      "grad_norm": 20.243443532186205,
      "learning_rate": 5.6911640161851884e-05,
      "loss": 0.1393,
      "step": 9140
    },
    {
      "epoch": 2.47,
      "grad_norm": 17.950818424109162,
      "learning_rate": 5.662934035946175e-05,
      "loss": 0.1663,
      "step": 9150
    },
    {
      "epoch": 2.47,
      "grad_norm": 15.174970178519834,
      "learning_rate": 5.6347040557071605e-05,
      "loss": 0.1498,
      "step": 9160
    },
    {
      "epoch": 2.47,
      "grad_norm": 17.224709285260754,
      "learning_rate": 5.606474075468147e-05,
      "loss": 0.2015,
      "step": 9170
    },
    {
      "epoch": 2.48,
      "grad_norm": 17.854280982677185,
      "learning_rate": 5.578244095229133e-05,
      "loss": 0.2116,
      "step": 9180
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.356805726056697,
      "learning_rate": 5.550014114990119e-05,
      "loss": 0.1513,
      "step": 9190
    },
    {
      "epoch": 2.48,
      "grad_norm": 15.055360584029824,
      "learning_rate": 5.521784134751105e-05,
      "loss": 0.2827,
      "step": 9200
    },
    {
      "epoch": 2.48,
      "grad_norm": 14.300226445651772,
      "learning_rate": 5.493554154512091e-05,
      "loss": 0.2127,
      "step": 9210
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.5352393665325312,
      "learning_rate": 5.465324174273077e-05,
      "loss": 0.1714,
      "step": 9220
    },
    {
      "epoch": 2.49,
      "grad_norm": 13.680455537560576,
      "learning_rate": 5.437094194034064e-05,
      "loss": 0.1957,
      "step": 9230
    },
    {
      "epoch": 2.49,
      "grad_norm": 7.924459310962841,
      "learning_rate": 5.40886421379505e-05,
      "loss": 0.2543,
      "step": 9240
    },
    {
      "epoch": 2.49,
      "grad_norm": 14.340450440294328,
      "learning_rate": 5.3806342335560364e-05,
      "loss": 0.1964,
      "step": 9250
    },
    {
      "epoch": 2.5,
      "grad_norm": 6.101228180980921,
      "learning_rate": 5.352404253317023e-05,
      "loss": 0.1237,
      "step": 9260
    },
    {
      "epoch": 2.5,
      "grad_norm": 5.962907025801654,
      "learning_rate": 5.3241742730780085e-05,
      "loss": 0.3169,
      "step": 9270
    },
    {
      "epoch": 2.5,
      "grad_norm": 17.53056363004829,
      "learning_rate": 5.295944292838994e-05,
      "loss": 0.2177,
      "step": 9280
    },
    {
      "epoch": 2.5,
      "grad_norm": 6.468750269516649,
      "learning_rate": 5.2677143125999805e-05,
      "loss": 0.1628,
      "step": 9290
    },
    {
      "epoch": 2.51,
      "grad_norm": 15.610624859958747,
      "learning_rate": 5.239484332360967e-05,
      "loss": 0.2929,
      "step": 9300
    },
    {
      "epoch": 2.51,
      "grad_norm": 13.917503167940149,
      "learning_rate": 5.211254352121953e-05,
      "loss": 0.1182,
      "step": 9310
    },
    {
      "epoch": 2.51,
      "grad_norm": 32.249391416973076,
      "learning_rate": 5.183024371882939e-05,
      "loss": 0.3041,
      "step": 9320
    },
    {
      "epoch": 2.52,
      "grad_norm": 16.8745136120248,
      "learning_rate": 5.154794391643925e-05,
      "loss": 0.2309,
      "step": 9330
    },
    {
      "epoch": 2.52,
      "grad_norm": 20.193966876417633,
      "learning_rate": 5.126564411404912e-05,
      "loss": 0.2259,
      "step": 9340
    },
    {
      "epoch": 2.52,
      "grad_norm": 3.390373414293016,
      "learning_rate": 5.098334431165898e-05,
      "loss": 0.1503,
      "step": 9350
    },
    {
      "epoch": 2.52,
      "grad_norm": 6.548050945526262,
      "learning_rate": 5.0701044509268844e-05,
      "loss": 0.1775,
      "step": 9360
    },
    {
      "epoch": 2.53,
      "grad_norm": 40.115292584054636,
      "learning_rate": 5.0418744706878694e-05,
      "loss": 0.2203,
      "step": 9370
    },
    {
      "epoch": 2.53,
      "grad_norm": 3.1985930247469287,
      "learning_rate": 5.013644490448856e-05,
      "loss": 0.1319,
      "step": 9380
    },
    {
      "epoch": 2.53,
      "grad_norm": 28.002153686307533,
      "learning_rate": 4.985414510209842e-05,
      "loss": 0.1722,
      "step": 9390
    },
    {
      "epoch": 2.53,
      "grad_norm": 4.378829418247621,
      "learning_rate": 4.9571845299708285e-05,
      "loss": 0.2389,
      "step": 9400
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.2415966999666233,
      "learning_rate": 4.928954549731815e-05,
      "loss": 0.1313,
      "step": 9410
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.310184060753123,
      "learning_rate": 4.900724569492801e-05,
      "loss": 0.1876,
      "step": 9420
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.6950047972254636,
      "learning_rate": 4.8724945892537876e-05,
      "loss": 0.1243,
      "step": 9430
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.0445742318483668,
      "learning_rate": 4.844264609014773e-05,
      "loss": 0.1165,
      "step": 9440
    },
    {
      "epoch": 2.55,
      "grad_norm": 3.9680162928677793,
      "learning_rate": 4.81603462877576e-05,
      "loss": 0.2511,
      "step": 9450
    },
    {
      "epoch": 2.55,
      "grad_norm": 21.34490624712699,
      "learning_rate": 4.7878046485367454e-05,
      "loss": 0.2838,
      "step": 9460
    },
    {
      "epoch": 2.55,
      "grad_norm": 9.72818389978754,
      "learning_rate": 4.759574668297732e-05,
      "loss": 0.1516,
      "step": 9470
    },
    {
      "epoch": 2.56,
      "grad_norm": 12.456315184485808,
      "learning_rate": 4.731344688058718e-05,
      "loss": 0.2693,
      "step": 9480
    },
    {
      "epoch": 2.56,
      "grad_norm": 22.044174103919907,
      "learning_rate": 4.703114707819704e-05,
      "loss": 0.1897,
      "step": 9490
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.4284959755719109,
      "learning_rate": 4.67488472758069e-05,
      "loss": 0.113,
      "step": 9500
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.269977581174073,
      "learning_rate": 4.6466547473416765e-05,
      "loss": 0.1211,
      "step": 9510
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.572829559701346,
      "learning_rate": 4.618424767102663e-05,
      "loss": 0.1177,
      "step": 9520
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.9570099313402664,
      "learning_rate": 4.590194786863649e-05,
      "loss": 0.2311,
      "step": 9530
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.1860228018685577,
      "learning_rate": 4.5619648066246356e-05,
      "loss": 0.1223,
      "step": 9540
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.6965634582380995,
      "learning_rate": 4.5337348263856206e-05,
      "loss": 0.0797,
      "step": 9550
    },
    {
      "epoch": 2.58,
      "grad_norm": 22.44909251349996,
      "learning_rate": 4.505504846146607e-05,
      "loss": 0.1397,
      "step": 9560
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.9225828038928983,
      "learning_rate": 4.4772748659075933e-05,
      "loss": 0.083,
      "step": 9570
    },
    {
      "epoch": 2.58,
      "grad_norm": 33.48547968998044,
      "learning_rate": 4.44904488566858e-05,
      "loss": 0.1529,
      "step": 9580
    },
    {
      "epoch": 2.59,
      "grad_norm": 20.28770542930981,
      "learning_rate": 4.420814905429566e-05,
      "loss": 0.1181,
      "step": 9590
    },
    {
      "epoch": 2.59,
      "grad_norm": 19.523120753911876,
      "learning_rate": 4.392584925190552e-05,
      "loss": 0.3424,
      "step": 9600
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.9182701209012607,
      "learning_rate": 4.364354944951538e-05,
      "loss": 0.1279,
      "step": 9610
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.32187228644325483,
      "learning_rate": 4.3361249647125245e-05,
      "loss": 0.1535,
      "step": 9620
    },
    {
      "epoch": 2.6,
      "grad_norm": 4.5132577267453575,
      "learning_rate": 4.307894984473511e-05,
      "loss": 0.1281,
      "step": 9630
    },
    {
      "epoch": 2.6,
      "grad_norm": 7.678362540349035,
      "learning_rate": 4.2796650042344966e-05,
      "loss": 0.0488,
      "step": 9640
    },
    {
      "epoch": 2.6,
      "grad_norm": 14.685212696817482,
      "learning_rate": 4.251435023995482e-05,
      "loss": 0.1049,
      "step": 9650
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.211336414473553,
      "learning_rate": 4.2232050437564686e-05,
      "loss": 0.2579,
      "step": 9660
    },
    {
      "epoch": 2.61,
      "grad_norm": 11.479698351944092,
      "learning_rate": 4.194975063517455e-05,
      "loss": 0.2264,
      "step": 9670
    },
    {
      "epoch": 2.61,
      "grad_norm": 12.725442954782324,
      "learning_rate": 4.1667450832784413e-05,
      "loss": 0.1741,
      "step": 9680
    },
    {
      "epoch": 2.61,
      "grad_norm": 10.249446087791966,
      "learning_rate": 4.138515103039428e-05,
      "loss": 0.1901,
      "step": 9690
    },
    {
      "epoch": 2.62,
      "grad_norm": 22.348882282807473,
      "learning_rate": 4.110285122800414e-05,
      "loss": 0.3248,
      "step": 9700
    },
    {
      "epoch": 2.62,
      "grad_norm": 29.80112408675683,
      "learning_rate": 4.0820551425614e-05,
      "loss": 0.2892,
      "step": 9710
    },
    {
      "epoch": 2.62,
      "grad_norm": 25.616892364746988,
      "learning_rate": 4.053825162322386e-05,
      "loss": 0.1461,
      "step": 9720
    },
    {
      "epoch": 2.62,
      "grad_norm": 12.3616307252157,
      "learning_rate": 4.025595182083372e-05,
      "loss": 0.1824,
      "step": 9730
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.6004789095816092,
      "learning_rate": 3.997365201844358e-05,
      "loss": 0.1883,
      "step": 9740
    },
    {
      "epoch": 2.63,
      "grad_norm": 11.35350534309671,
      "learning_rate": 3.9691352216053446e-05,
      "loss": 0.1768,
      "step": 9750
    },
    {
      "epoch": 2.63,
      "grad_norm": 28.635074820663036,
      "learning_rate": 3.94090524136633e-05,
      "loss": 0.1519,
      "step": 9760
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.073113513727795,
      "learning_rate": 3.9126752611273166e-05,
      "loss": 0.0755,
      "step": 9770
    },
    {
      "epoch": 2.64,
      "grad_norm": 12.099013554494524,
      "learning_rate": 3.884445280888303e-05,
      "loss": 0.166,
      "step": 9780
    },
    {
      "epoch": 2.64,
      "grad_norm": 4.549412875060068,
      "learning_rate": 3.8562153006492893e-05,
      "loss": 0.209,
      "step": 9790
    },
    {
      "epoch": 2.64,
      "grad_norm": 40.95773271897725,
      "learning_rate": 3.827985320410276e-05,
      "loss": 0.2037,
      "step": 9800
    },
    {
      "epoch": 2.64,
      "grad_norm": 8.894536931836472,
      "learning_rate": 3.799755340171262e-05,
      "loss": 0.1239,
      "step": 9810
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.6014051661872064,
      "learning_rate": 3.771525359932247e-05,
      "loss": 0.1005,
      "step": 9820
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.451617030455998,
      "learning_rate": 3.743295379693234e-05,
      "loss": 0.1972,
      "step": 9830
    },
    {
      "epoch": 2.65,
      "grad_norm": 14.281214271600886,
      "learning_rate": 3.7150653994542205e-05,
      "loss": 0.2091,
      "step": 9840
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.370990823293876,
      "learning_rate": 3.686835419215206e-05,
      "loss": 0.1303,
      "step": 9850
    },
    {
      "epoch": 2.66,
      "grad_norm": 2.59718633224972,
      "learning_rate": 3.6586054389761925e-05,
      "loss": 0.2327,
      "step": 9860
    },
    {
      "epoch": 2.66,
      "grad_norm": 26.302937617387386,
      "learning_rate": 3.630375458737179e-05,
      "loss": 0.0945,
      "step": 9870
    },
    {
      "epoch": 2.66,
      "grad_norm": 25.898811952119246,
      "learning_rate": 3.6021454784981646e-05,
      "loss": 0.3208,
      "step": 9880
    },
    {
      "epoch": 2.67,
      "grad_norm": 28.40276251578497,
      "learning_rate": 3.573915498259151e-05,
      "loss": 0.1088,
      "step": 9890
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.4609229817152286,
      "learning_rate": 3.5456855180201367e-05,
      "loss": 0.2262,
      "step": 9900
    },
    {
      "epoch": 2.67,
      "grad_norm": 12.083885381443258,
      "learning_rate": 3.517455537781123e-05,
      "loss": 0.145,
      "step": 9910
    },
    {
      "epoch": 2.67,
      "grad_norm": 6.192094578672661,
      "learning_rate": 3.4892255575421094e-05,
      "loss": 0.1743,
      "step": 9920
    },
    {
      "epoch": 2.68,
      "grad_norm": 12.935451755464827,
      "learning_rate": 3.460995577303096e-05,
      "loss": 0.1638,
      "step": 9930
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.9625024974638734,
      "learning_rate": 3.4327655970640814e-05,
      "loss": 0.1538,
      "step": 9940
    },
    {
      "epoch": 2.68,
      "grad_norm": 11.988212943853569,
      "learning_rate": 3.404535616825068e-05,
      "loss": 0.1143,
      "step": 9950
    },
    {
      "epoch": 2.69,
      "grad_norm": 21.930326458598895,
      "learning_rate": 3.376305636586054e-05,
      "loss": 0.2101,
      "step": 9960
    },
    {
      "epoch": 2.69,
      "grad_norm": 4.254936579954893,
      "learning_rate": 3.3480756563470405e-05,
      "loss": 0.2284,
      "step": 9970
    },
    {
      "epoch": 2.69,
      "grad_norm": 23.40792316751564,
      "learning_rate": 3.319845676108026e-05,
      "loss": 0.1495,
      "step": 9980
    },
    {
      "epoch": 2.69,
      "grad_norm": 12.5414736333938,
      "learning_rate": 3.2916156958690126e-05,
      "loss": 0.1755,
      "step": 9990
    },
    {
      "epoch": 2.7,
      "grad_norm": 11.57541906278461,
      "learning_rate": 3.263385715629999e-05,
      "loss": 0.2049,
      "step": 10000
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.8565840694358073,
      "learning_rate": 3.235155735390985e-05,
      "loss": 0.2295,
      "step": 10010
    },
    {
      "epoch": 2.7,
      "grad_norm": 4.104404204372011,
      "learning_rate": 3.206925755151971e-05,
      "loss": 0.2286,
      "step": 10020
    },
    {
      "epoch": 2.7,
      "grad_norm": 13.630363386460491,
      "learning_rate": 3.1786957749129574e-05,
      "loss": 0.2001,
      "step": 10030
    },
    {
      "epoch": 2.71,
      "grad_norm": 24.293814727656738,
      "learning_rate": 3.1532887926978446e-05,
      "loss": 0.256,
      "step": 10040
    },
    {
      "epoch": 2.71,
      "grad_norm": 6.241056950400041,
      "learning_rate": 3.125058812458831e-05,
      "loss": 0.2753,
      "step": 10050
    },
    {
      "epoch": 2.71,
      "grad_norm": 15.255685231195665,
      "learning_rate": 3.096828832219817e-05,
      "loss": 0.1191,
      "step": 10060
    },
    {
      "epoch": 2.72,
      "grad_norm": 23.541379953146336,
      "learning_rate": 3.068598851980803e-05,
      "loss": 0.1802,
      "step": 10070
    },
    {
      "epoch": 2.72,
      "grad_norm": 35.462786609270005,
      "learning_rate": 3.0403688717417894e-05,
      "loss": 0.176,
      "step": 10080
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.067508142093401,
      "learning_rate": 3.0121388915027758e-05,
      "loss": 0.2222,
      "step": 10090
    },
    {
      "epoch": 2.72,
      "grad_norm": 12.384863365929474,
      "learning_rate": 2.9839089112637618e-05,
      "loss": 0.1562,
      "step": 10100
    },
    {
      "epoch": 2.73,
      "grad_norm": 5.663444667531255,
      "learning_rate": 2.955678931024748e-05,
      "loss": 0.1152,
      "step": 10110
    },
    {
      "epoch": 2.73,
      "grad_norm": 7.0382821291396755,
      "learning_rate": 2.9274489507857342e-05,
      "loss": 0.2138,
      "step": 10120
    },
    {
      "epoch": 2.73,
      "grad_norm": 10.421781203492715,
      "learning_rate": 2.8992189705467206e-05,
      "loss": 0.2082,
      "step": 10130
    },
    {
      "epoch": 2.73,
      "grad_norm": 3.0326063590913774,
      "learning_rate": 2.8709889903077066e-05,
      "loss": 0.1482,
      "step": 10140
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.12673258365712695,
      "learning_rate": 2.8427590100686926e-05,
      "loss": 0.0894,
      "step": 10150
    },
    {
      "epoch": 2.74,
      "grad_norm": 23.46284584060824,
      "learning_rate": 2.8145290298296787e-05,
      "loss": 0.1418,
      "step": 10160
    },
    {
      "epoch": 2.74,
      "grad_norm": 24.795864279327116,
      "learning_rate": 2.786299049590665e-05,
      "loss": 0.214,
      "step": 10170
    },
    {
      "epoch": 2.74,
      "grad_norm": 18.05046129882148,
      "learning_rate": 2.7580690693516514e-05,
      "loss": 0.1471,
      "step": 10180
    },
    {
      "epoch": 2.75,
      "grad_norm": 15.892096042060471,
      "learning_rate": 2.729839089112637e-05,
      "loss": 0.3161,
      "step": 10190
    },
    {
      "epoch": 2.75,
      "grad_norm": 2.4521237212211204,
      "learning_rate": 2.7016091088736235e-05,
      "loss": 0.1494,
      "step": 10200
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.9512509324264316,
      "learning_rate": 2.6733791286346098e-05,
      "loss": 0.1728,
      "step": 10210
    },
    {
      "epoch": 2.76,
      "grad_norm": 2.4724120650202446,
      "learning_rate": 2.645149148395596e-05,
      "loss": 0.127,
      "step": 10220
    },
    {
      "epoch": 2.76,
      "grad_norm": 6.537427769381818,
      "learning_rate": 2.6169191681565822e-05,
      "loss": 0.1869,
      "step": 10230
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.3591485822430767,
      "learning_rate": 2.5886891879175682e-05,
      "loss": 0.049,
      "step": 10240
    },
    {
      "epoch": 2.76,
      "grad_norm": 9.664441527610268,
      "learning_rate": 2.5604592076785543e-05,
      "loss": 0.1782,
      "step": 10250
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.8662432911974188,
      "learning_rate": 2.5322292274395406e-05,
      "loss": 0.054,
      "step": 10260
    },
    {
      "epoch": 2.77,
      "grad_norm": 7.911172839408286,
      "learning_rate": 2.503999247200527e-05,
      "loss": 0.1397,
      "step": 10270
    },
    {
      "epoch": 2.77,
      "grad_norm": 4.381457639993246,
      "learning_rate": 2.4757692669615127e-05,
      "loss": 0.1403,
      "step": 10280
    },
    {
      "epoch": 2.77,
      "grad_norm": 5.225460692772398,
      "learning_rate": 2.447539286722499e-05,
      "loss": 0.0879,
      "step": 10290
    },
    {
      "epoch": 2.78,
      "grad_norm": 6.130140843961162,
      "learning_rate": 2.419309306483485e-05,
      "loss": 0.1129,
      "step": 10300
    },
    {
      "epoch": 2.78,
      "grad_norm": 19.38322042845477,
      "learning_rate": 2.3910793262444714e-05,
      "loss": 0.2095,
      "step": 10310
    },
    {
      "epoch": 2.78,
      "grad_norm": 17.947935163492453,
      "learning_rate": 2.3628493460054578e-05,
      "loss": 0.116,
      "step": 10320
    },
    {
      "epoch": 2.79,
      "grad_norm": 19.282747261549655,
      "learning_rate": 2.3346193657664435e-05,
      "loss": 0.1943,
      "step": 10330
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.49554576525912725,
      "learning_rate": 2.30638938552743e-05,
      "loss": 0.1193,
      "step": 10340
    },
    {
      "epoch": 2.79,
      "grad_norm": 26.721444132425745,
      "learning_rate": 2.2781594052884162e-05,
      "loss": 0.4734,
      "step": 10350
    },
    {
      "epoch": 2.79,
      "grad_norm": 10.246825080481248,
      "learning_rate": 2.2499294250494023e-05,
      "loss": 0.2296,
      "step": 10360
    },
    {
      "epoch": 2.8,
      "grad_norm": 33.3374749024257,
      "learning_rate": 2.2216994448103886e-05,
      "loss": 0.2848,
      "step": 10370
    },
    {
      "epoch": 2.8,
      "grad_norm": 33.07153989143948,
      "learning_rate": 2.1934694645713743e-05,
      "loss": 0.2041,
      "step": 10380
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.4533168495991373,
      "learning_rate": 2.1652394843323607e-05,
      "loss": 0.183,
      "step": 10390
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.25881940374862,
      "learning_rate": 2.137009504093347e-05,
      "loss": 0.1263,
      "step": 10400
    },
    {
      "epoch": 2.81,
      "grad_norm": 4.496134825128603,
      "learning_rate": 2.1087795238543334e-05,
      "loss": 0.2176,
      "step": 10410
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.6406543339153323,
      "learning_rate": 2.080549543615319e-05,
      "loss": 0.2758,
      "step": 10420
    },
    {
      "epoch": 2.81,
      "grad_norm": 18.336176093377286,
      "learning_rate": 2.0523195633763055e-05,
      "loss": 0.1777,
      "step": 10430
    },
    {
      "epoch": 2.81,
      "grad_norm": 16.3281241165964,
      "learning_rate": 2.0240895831372915e-05,
      "loss": 0.1851,
      "step": 10440
    },
    {
      "epoch": 2.82,
      "grad_norm": 5.072886034733424,
      "learning_rate": 1.995859602898278e-05,
      "loss": 0.0579,
      "step": 10450
    },
    {
      "epoch": 2.82,
      "grad_norm": 6.906086222294693,
      "learning_rate": 1.9676296226592642e-05,
      "loss": 0.1702,
      "step": 10460
    },
    {
      "epoch": 2.82,
      "grad_norm": 22.288798816349974,
      "learning_rate": 1.93939964242025e-05,
      "loss": 0.0783,
      "step": 10470
    },
    {
      "epoch": 2.83,
      "grad_norm": 18.28379128350265,
      "learning_rate": 1.9111696621812363e-05,
      "loss": 0.2655,
      "step": 10480
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.114633962923008,
      "learning_rate": 1.8829396819422227e-05,
      "loss": 0.0803,
      "step": 10490
    },
    {
      "epoch": 2.83,
      "grad_norm": 15.418474013692522,
      "learning_rate": 1.8547097017032087e-05,
      "loss": 0.1851,
      "step": 10500
    },
    {
      "epoch": 2.83,
      "grad_norm": 2.52579277368463,
      "learning_rate": 1.8264797214641947e-05,
      "loss": 0.2114,
      "step": 10510
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.1893992831700284,
      "learning_rate": 1.7982497412251807e-05,
      "loss": 0.2429,
      "step": 10520
    },
    {
      "epoch": 2.84,
      "grad_norm": 18.36244151052219,
      "learning_rate": 1.770019760986167e-05,
      "loss": 0.315,
      "step": 10530
    },
    {
      "epoch": 2.84,
      "grad_norm": 4.57883053524836,
      "learning_rate": 1.7417897807471535e-05,
      "loss": 0.1885,
      "step": 10540
    },
    {
      "epoch": 2.84,
      "grad_norm": 22.14157883481377,
      "learning_rate": 1.7135598005081395e-05,
      "loss": 0.1844,
      "step": 10550
    },
    {
      "epoch": 2.85,
      "grad_norm": 15.614767745622185,
      "learning_rate": 1.685329820269126e-05,
      "loss": 0.2003,
      "step": 10560
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.7684675034723437,
      "learning_rate": 1.657099840030112e-05,
      "loss": 0.2496,
      "step": 10570
    },
    {
      "epoch": 2.85,
      "grad_norm": 8.975794642718418,
      "learning_rate": 1.628869859791098e-05,
      "loss": 0.1111,
      "step": 10580
    },
    {
      "epoch": 2.86,
      "grad_norm": 29.732792173223608,
      "learning_rate": 1.600639879552084e-05,
      "loss": 0.2099,
      "step": 10590
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.8900755968708858,
      "learning_rate": 1.5724098993130703e-05,
      "loss": 0.1302,
      "step": 10600
    },
    {
      "epoch": 2.86,
      "grad_norm": 7.284451979889495,
      "learning_rate": 1.5441799190740563e-05,
      "loss": 0.1201,
      "step": 10610
    },
    {
      "epoch": 2.86,
      "grad_norm": 3.7803466306177063,
      "learning_rate": 1.5159499388350427e-05,
      "loss": 0.1154,
      "step": 10620
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.9301047978888305,
      "learning_rate": 1.4877199585960289e-05,
      "loss": 0.1154,
      "step": 10630
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.283545994120593,
      "learning_rate": 1.459489978357015e-05,
      "loss": 0.0617,
      "step": 10640
    },
    {
      "epoch": 2.87,
      "grad_norm": 7.377032750947038,
      "learning_rate": 1.4312599981180013e-05,
      "loss": 0.1709,
      "step": 10650
    },
    {
      "epoch": 2.87,
      "grad_norm": 17.43230815474847,
      "learning_rate": 1.4030300178789873e-05,
      "loss": 0.1875,
      "step": 10660
    },
    {
      "epoch": 2.88,
      "grad_norm": 13.192557534022258,
      "learning_rate": 1.3748000376399735e-05,
      "loss": 0.246,
      "step": 10670
    },
    {
      "epoch": 2.88,
      "grad_norm": 31.43882672921191,
      "learning_rate": 1.3465700574009595e-05,
      "loss": 0.1333,
      "step": 10680
    },
    {
      "epoch": 2.88,
      "grad_norm": 4.577658886665074,
      "learning_rate": 1.3183400771619459e-05,
      "loss": 0.1919,
      "step": 10690
    },
    {
      "epoch": 2.88,
      "grad_norm": 13.456038283655007,
      "learning_rate": 1.2901100969229321e-05,
      "loss": 0.114,
      "step": 10700
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.3584541741981806,
      "learning_rate": 1.2618801166839181e-05,
      "loss": 0.1205,
      "step": 10710
    },
    {
      "epoch": 2.89,
      "grad_norm": 19.242534877477368,
      "learning_rate": 1.2336501364449045e-05,
      "loss": 0.1972,
      "step": 10720
    },
    {
      "epoch": 2.89,
      "grad_norm": 8.305705299701168,
      "learning_rate": 1.2054201562058905e-05,
      "loss": 0.2397,
      "step": 10730
    },
    {
      "epoch": 2.9,
      "grad_norm": 8.421684746459949,
      "learning_rate": 1.1771901759668767e-05,
      "loss": 0.1897,
      "step": 10740
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.7270123220243163,
      "learning_rate": 1.1489601957278628e-05,
      "loss": 0.2126,
      "step": 10750
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.813197289644501,
      "learning_rate": 1.1207302154888491e-05,
      "loss": 0.1909,
      "step": 10760
    },
    {
      "epoch": 2.9,
      "grad_norm": 11.396372736209983,
      "learning_rate": 1.0925002352498351e-05,
      "loss": 0.1303,
      "step": 10770
    },
    {
      "epoch": 2.91,
      "grad_norm": 4.229047375340985,
      "learning_rate": 1.0642702550108213e-05,
      "loss": 0.249,
      "step": 10780
    },
    {
      "epoch": 2.91,
      "grad_norm": 12.413137902025088,
      "learning_rate": 1.0360402747718077e-05,
      "loss": 0.1451,
      "step": 10790
    },
    {
      "epoch": 2.91,
      "grad_norm": 2.016504409495685,
      "learning_rate": 1.0078102945327937e-05,
      "loss": 0.1039,
      "step": 10800
    },
    {
      "epoch": 2.91,
      "grad_norm": 28.82203935727066,
      "learning_rate": 9.7958031429378e-06,
      "loss": 0.097,
      "step": 10810
    },
    {
      "epoch": 2.92,
      "grad_norm": 16.95449150748396,
      "learning_rate": 9.51350334054766e-06,
      "loss": 0.2607,
      "step": 10820
    },
    {
      "epoch": 2.92,
      "grad_norm": 50.266130459668716,
      "learning_rate": 9.231203538157523e-06,
      "loss": 0.2367,
      "step": 10830
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.2714447413363867,
      "learning_rate": 8.948903735767385e-06,
      "loss": 0.154,
      "step": 10840
    },
    {
      "epoch": 2.93,
      "grad_norm": 30.248618987968726,
      "learning_rate": 8.666603933377246e-06,
      "loss": 0.3012,
      "step": 10850
    },
    {
      "epoch": 2.93,
      "grad_norm": 8.35957699536102,
      "learning_rate": 8.384304130987108e-06,
      "loss": 0.2535,
      "step": 10860
    },
    {
      "epoch": 2.93,
      "grad_norm": 5.402977238430571,
      "learning_rate": 8.10200432859697e-06,
      "loss": 0.1055,
      "step": 10870
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.7658080411848506,
      "learning_rate": 7.819704526206831e-06,
      "loss": 0.0852,
      "step": 10880
    },
    {
      "epoch": 2.94,
      "grad_norm": 3.4455168348818863,
      "learning_rate": 7.5374047238166926e-06,
      "loss": 0.0918,
      "step": 10890
    },
    {
      "epoch": 2.94,
      "grad_norm": 44.560022513978446,
      "learning_rate": 7.255104921426554e-06,
      "loss": 0.223,
      "step": 10900
    },
    {
      "epoch": 2.94,
      "grad_norm": 13.969874511703313,
      "learning_rate": 6.9728051190364165e-06,
      "loss": 0.1228,
      "step": 10910
    },
    {
      "epoch": 2.94,
      "grad_norm": 20.27901599700173,
      "learning_rate": 6.6905053166462785e-06,
      "loss": 0.1186,
      "step": 10920
    },
    {
      "epoch": 2.95,
      "grad_norm": 4.479367735327111,
      "learning_rate": 6.40820551425614e-06,
      "loss": 0.1047,
      "step": 10930
    },
    {
      "epoch": 2.95,
      "grad_norm": 9.364347789036554,
      "learning_rate": 6.1259057118660016e-06,
      "loss": 0.3191,
      "step": 10940
    },
    {
      "epoch": 2.95,
      "grad_norm": 20.35776964300664,
      "learning_rate": 5.843605909475863e-06,
      "loss": 0.2484,
      "step": 10950
    },
    {
      "epoch": 2.95,
      "grad_norm": 7.717125715462946,
      "learning_rate": 5.561306107085725e-06,
      "loss": 0.1793,
      "step": 10960
    },
    {
      "epoch": 2.96,
      "grad_norm": 5.158116225254854,
      "learning_rate": 5.279006304695586e-06,
      "loss": 0.1196,
      "step": 10970
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.1868054675318116,
      "learning_rate": 4.996706502305448e-06,
      "loss": 0.1555,
      "step": 10980
    },
    {
      "epoch": 2.96,
      "grad_norm": 10.017114261501675,
      "learning_rate": 4.714406699915309e-06,
      "loss": 0.1734,
      "step": 10990
    },
    {
      "epoch": 2.97,
      "grad_norm": 20.507649475902905,
      "learning_rate": 4.432106897525171e-06,
      "loss": 0.1899,
      "step": 11000
    },
    {
      "epoch": 2.97,
      "grad_norm": 23.188110629495977,
      "learning_rate": 4.149807095135033e-06,
      "loss": 0.183,
      "step": 11010
    },
    {
      "epoch": 2.97,
      "grad_norm": 8.661035069847358,
      "learning_rate": 3.867507292744895e-06,
      "loss": 0.2224,
      "step": 11020
    },
    {
      "epoch": 2.97,
      "grad_norm": 5.927613956133262,
      "learning_rate": 3.5852074903547563e-06,
      "loss": 0.1563,
      "step": 11030
    },
    {
      "epoch": 2.98,
      "grad_norm": 5.993685079356996,
      "learning_rate": 3.302907687964618e-06,
      "loss": 0.343,
      "step": 11040
    },
    {
      "epoch": 2.98,
      "grad_norm": 12.660792693760373,
      "learning_rate": 3.02060788557448e-06,
      "loss": 0.1164,
      "step": 11050
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.4311284335669525,
      "learning_rate": 2.7383080831843418e-06,
      "loss": 0.1534,
      "step": 11060
    },
    {
      "epoch": 2.98,
      "grad_norm": 7.580731347788203,
      "learning_rate": 2.4560082807942033e-06,
      "loss": 0.1765,
      "step": 11070
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.9417696744342825,
      "learning_rate": 2.173708478404065e-06,
      "loss": 0.1287,
      "step": 11080
    },
    {
      "epoch": 2.99,
      "grad_norm": 12.618587617870885,
      "learning_rate": 1.8914086760139266e-06,
      "loss": 0.1503,
      "step": 11090
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.579704260829425,
      "learning_rate": 1.6091088736237882e-06,
      "loss": 0.0957,
      "step": 11100
    },
    {
      "epoch": 3.0,
      "grad_norm": 34.948411888938296,
      "learning_rate": 1.3268090712336501e-06,
      "loss": 0.3559,
      "step": 11110
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.5880440760048153,
      "learning_rate": 1.0445092688435117e-06,
      "loss": 0.2672,
      "step": 11120
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8666984429615506,
      "eval_f1": 0.9118790043062704,
      "eval_loss": 0.354736328125,
      "eval_precision": 0.8926588525601481,
      "eval_recall": 0.9319450407900386,
      "eval_runtime": 428.6721,
      "eval_samples_per_second": 14.683,
      "eval_steps_per_second": 2.447,
      "step": 11127
    }
  ],
  "logging_steps": 10,
  "max_steps": 11127,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2.0911014688063488e+17,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}