{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 6848,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0004999993423087359,
      "loss": 1.3569,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.240234375,
      "learning_rate": 0.000499997369238404,
      "loss": 1.4018,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.28125,
      "learning_rate": 0.0004999940807993858,
      "loss": 1.1722,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0004999894770089834,
      "loss": 1.2406,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004999835578914198,
      "loss": 1.1451,
      "step": 25
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0004999763234778388,
      "loss": 1.2433,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004999677738063041,
      "loss": 0.9514,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0004999579089218003,
      "loss": 1.1503,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0004999467288762319,
      "loss": 1.0971,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004999342337284227,
      "loss": 1.1548,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0004999204235441167,
      "loss": 1.0912,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0004999052983959764,
      "loss": 1.0526,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.111328125,
      "learning_rate": 0.0004998888583635831,
      "loss": 0.9175,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0004998711035334366,
      "loss": 1.1837,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004998520339989548,
      "loss": 1.1483,
      "step": 75
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004998316498604721,
      "loss": 1.1718,
      "step": 80
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0004998099512252406,
      "loss": 1.1358,
      "step": 85
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004997869382074284,
      "loss": 1.0682,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004997626109281188,
      "loss": 1.202,
      "step": 95
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0004997369695153109,
      "loss": 1.022,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0004997100141039175,
      "loss": 1.1212,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0004996817448357654,
      "loss": 0.9973,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004996521618595943,
      "loss": 1.1896,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0004996212653310557,
      "loss": 1.2285,
      "step": 120
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004995890554127129,
      "loss": 1.0276,
      "step": 125
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0004995555322740391,
      "loss": 1.0489,
      "step": 130
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0004995206960914175,
      "loss": 1.1461,
      "step": 135
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0004994845470481397,
      "loss": 1.1738,
      "step": 140
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0004994470853344047,
      "loss": 1.1,
      "step": 145
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0004994083111473188,
      "loss": 1.0523,
      "step": 150
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0004993682246908933,
      "loss": 1.1993,
      "step": 155
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004993268261760445,
      "loss": 1.2094,
      "step": 160
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004992841158205917,
      "loss": 0.9554,
      "step": 165
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0004992400938492571,
      "loss": 1.3459,
      "step": 170
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0004991947604936632,
      "loss": 1.0545,
      "step": 175
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.000499148115992333,
      "loss": 1.0093,
      "step": 180
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0004991001605906879,
      "loss": 1.0769,
      "step": 185
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004990508945410469,
      "loss": 1.2767,
      "step": 190
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0004990003181026245,
      "loss": 0.9031,
      "step": 195
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0004989484315415304,
      "loss": 1.1937,
      "step": 200
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.130859375,
      "learning_rate": 0.000498895235130767,
      "loss": 1.0573,
      "step": 205
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0004988407291502293,
      "loss": 1.0702,
      "step": 210
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0004987849138867016,
      "loss": 0.9613,
      "step": 215
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0004987277896338581,
      "loss": 1.1099,
      "step": 220
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0004986693566922595,
      "loss": 1.0443,
      "step": 225
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004986096153693524,
      "loss": 1.0332,
      "step": 230
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004985485659794676,
      "loss": 0.9495,
      "step": 235
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0004984862088438187,
      "loss": 0.9367,
      "step": 240
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0004984225442904991,
      "loss": 1.0334,
      "step": 245
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1328125,
      "learning_rate": 0.000498357572654482,
      "loss": 1.042,
      "step": 250
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0004982912942776178,
      "loss": 1.0528,
      "step": 255
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004982237095086319,
      "loss": 0.9818,
      "step": 260
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0004981548187031236,
      "loss": 1.2563,
      "step": 265
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0004980846222235642,
      "loss": 1.2442,
      "step": 270
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0004980131204392943,
      "loss": 1.1043,
      "step": 275
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0004979403137265229,
      "loss": 1.0199,
      "step": 280
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0004978662024683247,
      "loss": 1.0435,
      "step": 285
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0004977907870546382,
      "loss": 0.9776,
      "step": 290
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.000497714067882264,
      "loss": 1.2056,
      "step": 295
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004976360453548621,
      "loss": 1.0263,
      "step": 300
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0004975567198829506,
      "loss": 1.1623,
      "step": 305
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004974760918839028,
      "loss": 1.1521,
      "step": 310
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004973941617819453,
      "loss": 0.9725,
      "step": 315
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0004973109300081559,
      "loss": 0.9627,
      "step": 320
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0004972263970004611,
      "loss": 1.224,
      "step": 325
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0004971405632036337,
      "loss": 1.1421,
      "step": 330
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0004970534290692909,
      "loss": 1.2515,
      "step": 335
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0004969649950558915,
      "loss": 1.2142,
      "step": 340
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0004968752616287339,
      "loss": 1.0142,
      "step": 345
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0004967842292599532,
      "loss": 1.352,
      "step": 350
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004966918984285187,
      "loss": 1.3625,
      "step": 355
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004965982696202321,
      "loss": 1.1559,
      "step": 360
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.000496503343327724,
      "loss": 1.2461,
      "step": 365
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0004964071200504521,
      "loss": 0.9961,
      "step": 370
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0004963096002946982,
      "loss": 0.9184,
      "step": 375
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0004962107845735652,
      "loss": 1.1593,
      "step": 380
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0004961106734069751,
      "loss": 0.9015,
      "step": 385
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0004960092673216656,
      "loss": 0.9922,
      "step": 390
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0004959065668511881,
      "loss": 1.223,
      "step": 395
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0004958025725359043,
      "loss": 0.9188,
      "step": 400
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0004956972849229831,
      "loss": 1.1122,
      "step": 405
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0004955907045663988,
      "loss": 1.2144,
      "step": 410
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0004954828320269268,
      "loss": 0.8657,
      "step": 415
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.154296875,
      "learning_rate": 0.000495373667872142,
      "loss": 1.0499,
      "step": 420
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0004952632126764148,
      "loss": 1.2801,
      "step": 425
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004951514670209086,
      "loss": 1.0664,
      "step": 430
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0004950384314935764,
      "loss": 0.9355,
      "step": 435
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.000494924106689158,
      "loss": 1.0729,
      "step": 440
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13671875,
      "learning_rate": 0.000494808493209177,
      "loss": 0.8503,
      "step": 445
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0004946915916619371,
      "loss": 1.0489,
      "step": 450
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0004945734026625193,
      "loss": 0.9458,
      "step": 455
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0004944539268327786,
      "loss": 1.1663,
      "step": 460
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004943331648013406,
      "loss": 1.0852,
      "step": 465
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0004942111172035984,
      "loss": 1.0142,
      "step": 470
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0004940877846817092,
      "loss": 0.9625,
      "step": 475
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0004939631678845907,
      "loss": 1.1333,
      "step": 480
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.150390625,
      "learning_rate": 0.000493837267467918,
      "loss": 0.851,
      "step": 485
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0004937100840941199,
      "loss": 1.0439,
      "step": 490
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0004935816184323754,
      "loss": 1.1395,
      "step": 495
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0004934518711586106,
      "loss": 1.2047,
      "step": 500
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004933208429554948,
      "loss": 1.006,
      "step": 505
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0004931885345124366,
      "loss": 0.8703,
      "step": 510
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.000493054946525581,
      "loss": 1.0765,
      "step": 515
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.208984375,
      "learning_rate": 0.0004929200796978052,
      "loss": 1.1185,
      "step": 520
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.000492783934738715,
      "loss": 1.0236,
      "step": 525
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004926465123646413,
      "loss": 1.0684,
      "step": 530
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.150390625,
      "learning_rate": 0.000492507813298636,
      "loss": 0.9493,
      "step": 535
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0004923678382704684,
      "loss": 0.9562,
      "step": 540
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.25,
      "learning_rate": 0.0004922265880166214,
      "loss": 1.2393,
      "step": 545
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0004920840632802874,
      "loss": 1.2773,
      "step": 550
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004919402648113646,
      "loss": 0.9897,
      "step": 555
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.15234375,
      "learning_rate": 0.000491795193366453,
      "loss": 1.0711,
      "step": 560
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0004916488497088504,
      "loss": 1.1958,
      "step": 565
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004915012346085483,
      "loss": 1.2852,
      "step": 570
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004913523488422279,
      "loss": 1.2695,
      "step": 575
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004912021931932564,
      "loss": 1.1283,
      "step": 580
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004910507684516821,
      "loss": 0.8647,
      "step": 585
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0004908980754142308,
      "loss": 0.958,
      "step": 590
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0004907441148843016,
      "loss": 0.977,
      "step": 595
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0004905888876719624,
      "loss": 1.2838,
      "step": 600
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14453125,
      "learning_rate": 0.000490432394593946,
      "loss": 1.0568,
      "step": 605
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0004902746364736452,
      "loss": 1.1851,
      "step": 610
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0004901156141411094,
      "loss": 1.3584,
      "step": 615
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0004899553284330392,
      "loss": 1.0006,
      "step": 620
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0004897937801927826,
      "loss": 1.1368,
      "step": 625
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004896309702703308,
      "loss": 1.1571,
      "step": 630
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0004894668995223129,
      "loss": 1.0559,
      "step": 635
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.000489301568811992,
      "loss": 1.0644,
      "step": 640
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0004891349790092609,
      "loss": 1.1105,
      "step": 645
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004889671309906366,
      "loss": 1.3242,
      "step": 650
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0004887980256392567,
      "loss": 1.1187,
      "step": 655
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.000488627663844874,
      "loss": 1.0096,
      "step": 660
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0004884560465038523,
      "loss": 1.0949,
      "step": 665
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004882831745191611,
      "loss": 1.0701,
      "step": 670
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00048810904880037186,
      "loss": 1.141,
      "step": 675
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.404296875,
      "learning_rate": 0.0004879336702636523,
      "loss": 1.0199,
      "step": 680
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00048775703983176176,
      "loss": 1.0905,
      "step": 685
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00048757915843404656,
      "loss": 1.1913,
      "step": 690
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004874000270064351,
      "loss": 1.118,
      "step": 695
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.125,
      "learning_rate": 0.0004872196464914328,
      "loss": 0.9663,
      "step": 700
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00048703801783811707,
      "loss": 0.9864,
      "step": 705
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.00048685514200213256,
      "loss": 1.2736,
      "step": 710
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004866710199456861,
      "loss": 0.9167,
      "step": 715
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0004864856526375413,
      "loss": 1.1975,
      "step": 720
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00048629904105301395,
      "loss": 1.0447,
      "step": 725
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0004861111861739664,
      "loss": 1.0386,
      "step": 730
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0004859220889888028,
      "loss": 1.1379,
      "step": 735
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0004857317504924638,
      "loss": 0.957,
      "step": 740
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004855401716864209,
      "loss": 1.1209,
      "step": 745
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.134765625,
      "learning_rate": 0.000485347353578672,
      "loss": 0.9555,
      "step": 750
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00048515329718373514,
      "loss": 1.3033,
      "step": 755
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00048495800352264405,
      "loss": 0.9557,
      "step": 760
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.130859375,
      "learning_rate": 0.00048476147362294225,
      "loss": 1.2029,
      "step": 765
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00048456370851867757,
      "loss": 1.1479,
      "step": 770
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004843647092503972,
      "loss": 0.9966,
      "step": 775
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0004841644768651417,
      "loss": 0.9547,
      "step": 780
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0004839630124164398,
      "loss": 1.0046,
      "step": 785
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004837603169643028,
      "loss": 1.0157,
      "step": 790
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2265625,
      "learning_rate": 0.00048355639157521887,
      "loss": 1.1056,
      "step": 795
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0004833512373221476,
      "loss": 1.3283,
      "step": 800
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.00048314485528451435,
      "loss": 1.0102,
      "step": 805
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.00048293724654820434,
      "loss": 0.9784,
      "step": 810
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004827284122055572,
      "loss": 1.0426,
      "step": 815
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0004825183533553611,
      "loss": 1.0352,
      "step": 820
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.00048230707110284704,
      "loss": 1.1475,
      "step": 825
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.000482094566559683,
      "loss": 0.9966,
      "step": 830
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00048188084084396787,
      "loss": 1.261,
      "step": 835
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0004816658950802262,
      "loss": 1.0961,
      "step": 840
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00048144973039940134,
      "loss": 1.1438,
      "step": 845
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00048123234793885037,
      "loss": 1.1274,
      "step": 850
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0004810137488423376,
      "loss": 1.0945,
      "step": 855
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0004807939342600289,
      "loss": 0.9427,
      "step": 860
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0004805729053484852,
      "loss": 1.1931,
      "step": 865
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004803506632706569,
      "loss": 1.0598,
      "step": 870
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.00048012720919587715,
      "loss": 1.1676,
      "step": 875
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.00047990254429985653,
      "loss": 1.2744,
      "step": 880
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.000479676669764676,
      "loss": 1.0646,
      "step": 885
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15625,
      "learning_rate": 0.0004794495867787813,
      "loss": 1.1586,
      "step": 890
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00047922129653697634,
      "loss": 1.2567,
      "step": 895
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15625,
      "learning_rate": 0.00047899180024041723,
      "loss": 1.0267,
      "step": 900
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00047876109909660545,
      "loss": 1.0467,
      "step": 905
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004785291943193822,
      "loss": 0.8516,
      "step": 910
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.125,
      "learning_rate": 0.0004782960871289214,
      "loss": 1.2609,
      "step": 915
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00047806177875172343,
      "loss": 0.9976,
      "step": 920
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.000477826270420609,
      "loss": 1.0972,
      "step": 925
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0004775895633747123,
      "loss": 0.9034,
      "step": 930
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00047735165885947446,
      "loss": 0.964,
      "step": 935
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.125,
      "learning_rate": 0.0004771125581266373,
      "loss": 0.9712,
      "step": 940
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.00047687226243423665,
      "loss": 0.9746,
      "step": 945
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0004766307730465954,
      "loss": 1.2146,
      "step": 950
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00047638809123431725,
      "loss": 1.0272,
      "step": 955
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0004761442182742799,
      "loss": 1.1749,
      "step": 960
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00047589915544962827,
      "loss": 1.0741,
      "step": 965
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.123046875,
      "learning_rate": 0.00047565290404976774,
      "loss": 1.0154,
      "step": 970
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0004754054653703575,
      "loss": 1.0645,
      "step": 975
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004751568407133036,
      "loss": 1.1635,
      "step": 980
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0004749070313867522,
      "loss": 0.9108,
      "step": 985
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00047465603870508255,
      "loss": 0.943,
      "step": 990
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0004744038639889002,
      "loss": 1.0041,
      "step": 995
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00047415050856503005,
      "loss": 0.9916,
      "step": 1000
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004738959737665093,
      "loss": 0.967,
      "step": 1005
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00047364026093258045,
      "loss": 1.2214,
      "step": 1010
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004733833714086842,
      "loss": 0.9998,
      "step": 1015
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0004731253065464526,
      "loss": 1.0818,
      "step": 1020
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00047286606770370165,
      "loss": 0.9446,
      "step": 1025
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0004726056562444244,
      "loss": 1.0604,
      "step": 1030
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00047234407353878344,
      "loss": 0.9393,
      "step": 1035
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0004720813209631042,
      "loss": 0.9309,
      "step": 1040
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0004718173998998672,
      "loss": 1.1868,
      "step": 1045
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00047155231173770106,
      "loss": 0.9496,
      "step": 1050
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00047128605787137513,
      "loss": 0.8653,
      "step": 1055
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0004710186397017922,
      "loss": 1.111,
      "step": 1060
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.15234375,
      "learning_rate": 0.000470750058635981,
      "loss": 1.0898,
      "step": 1065
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00047048031608708875,
      "loss": 1.2158,
      "step": 1070
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004702094134743742,
      "loss": 0.8783,
      "step": 1075
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004699373522231996,
      "loss": 0.9491,
      "step": 1080
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.00046966413376502315,
      "loss": 1.0776,
      "step": 1085
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.00046938975953739225,
      "loss": 1.2107,
      "step": 1090
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0004691142309839351,
      "loss": 1.1185,
      "step": 1095
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0004688375495543535,
      "loss": 0.8819,
      "step": 1100
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0004685597167044152,
      "loss": 1.0172,
      "step": 1105
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004682807338959458,
      "loss": 0.8787,
      "step": 1110
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0004680006025968221,
      "loss": 1.0478,
      "step": 1115
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0004677193242809632,
      "loss": 1.0733,
      "step": 1120
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004674369004283234,
      "loss": 1.0968,
      "step": 1125
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11328125,
      "learning_rate": 0.00046715333252488435,
      "loss": 1.1467,
      "step": 1130
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004668686220626471,
      "loss": 1.078,
      "step": 1135
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0004665827705396244,
      "loss": 1.0399,
      "step": 1140
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0004662957794598325,
      "loss": 1.1142,
      "step": 1145
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00046600765033328375,
      "loss": 1.1665,
      "step": 1150
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00046571838467597814,
      "loss": 1.1634,
      "step": 1155
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0004654279840098957,
      "loss": 1.1111,
      "step": 1160
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004651364498629882,
      "loss": 1.0922,
      "step": 1165
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00046484378376917136,
      "loss": 1.1662,
      "step": 1170
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00046454998726831666,
      "loss": 1.1064,
      "step": 1175
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004642550619062432,
      "loss": 1.2852,
      "step": 1180
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004639590092347098,
      "loss": 0.8237,
      "step": 1185
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00046366183081140636,
      "loss": 1.1068,
      "step": 1190
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0004633635281999461,
      "loss": 1.0434,
      "step": 1195
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00046306410296985726,
      "loss": 1.1346,
      "step": 1200
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0004627635566965747,
      "loss": 1.1974,
      "step": 1205
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004624618909614316,
      "loss": 0.9354,
      "step": 1210
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00046215910735165146,
      "loss": 1.1143,
      "step": 1215
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00046185520746033917,
      "loss": 1.036,
      "step": 1220
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004615501928864733,
      "loss": 0.9453,
      "step": 1225
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004612440652348972,
      "loss": 0.9825,
      "step": 1230
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00046093682611631073,
      "loss": 1.3258,
      "step": 1235
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.109375,
      "learning_rate": 0.0004606284771472616,
      "loss": 0.9973,
      "step": 1240
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00046031901995013746,
      "loss": 1.0622,
      "step": 1245
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00046000845615315657,
      "loss": 1.0037,
      "step": 1250
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00045969678739035956,
      "loss": 0.9876,
      "step": 1255
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1875,
      "learning_rate": 0.00045938401530160126,
      "loss": 1.1052,
      "step": 1260
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00045907014153254123,
      "loss": 1.1145,
      "step": 1265
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0004587551677346359,
      "loss": 0.7833,
      "step": 1270
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0004584390955651292,
      "loss": 0.9334,
      "step": 1275
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00045812192668704456,
      "loss": 1.157,
      "step": 1280
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0004578036627691755,
      "loss": 1.0041,
      "step": 1285
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004574843054860772,
      "loss": 0.9626,
      "step": 1290
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0004571638565180577,
      "loss": 1.1321,
      "step": 1295
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004568423175511688,
      "loss": 0.9283,
      "step": 1300
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0004565196902771975,
      "loss": 1.0294,
      "step": 1305
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00045619597639365685,
      "loss": 1.0769,
      "step": 1310
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00045587117760377733,
      "loss": 0.9879,
      "step": 1315
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15625,
      "learning_rate": 0.00045554529561649744,
      "loss": 1.0914,
      "step": 1320
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00045521833214645516,
      "loss": 1.0999,
      "step": 1325
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004548902889139785,
      "loss": 1.0667,
      "step": 1330
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.00045456116764507694,
      "loss": 1.1249,
      "step": 1335
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0004542309700714319,
      "loss": 1.0936,
      "step": 1340
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004538996979303878,
      "loss": 0.9589,
      "step": 1345
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0004535673529649431,
      "loss": 1.088,
      "step": 1350
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0004532339369237408,
      "loss": 1.065,
      "step": 1355
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00045289945156105937,
      "loss": 1.0804,
      "step": 1360
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.189453125,
      "learning_rate": 0.00045256389863680365,
      "loss": 1.1125,
      "step": 1365
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0004522272799164955,
      "loss": 1.0844,
      "step": 1370
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00045188959717126445,
      "loss": 1.0587,
      "step": 1375
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00045155085217783844,
      "loss": 1.0235,
      "step": 1380
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0004512110467185343,
      "loss": 0.9903,
      "step": 1385
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.208984375,
      "learning_rate": 0.0004508701825812489,
      "loss": 1.3046,
      "step": 1390
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0004505282615594491,
      "loss": 1.0024,
      "step": 1395
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0004501852854521626,
      "loss": 1.1726,
      "step": 1400
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0004498412560639685,
      "loss": 1.1269,
      "step": 1405
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00044949617520498785,
      "loss": 0.9801,
      "step": 1410
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00044915004469087404,
      "loss": 1.0223,
      "step": 1415
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00044880286634280304,
      "loss": 0.9374,
      "step": 1420
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00044845464198746426,
      "loss": 1.0407,
      "step": 1425
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0004481053734570507,
      "loss": 1.0165,
      "step": 1430
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.15625,
      "learning_rate": 0.00044775506258924904,
      "loss": 1.2979,
      "step": 1435
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0004474037112272307,
      "loss": 1.0358,
      "step": 1440
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00044705132121964134,
      "loss": 1.0453,
      "step": 1445
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.125,
      "learning_rate": 0.00044669789442059154,
      "loss": 0.7933,
      "step": 1450
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00044634343268964717,
      "loss": 1.1367,
      "step": 1455
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00044598793789181923,
      "loss": 1.0895,
      "step": 1460
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0004456314118975543,
      "loss": 0.9276,
      "step": 1465
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.00044527385658272466,
      "loss": 1.0582,
      "step": 1470
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004449152738286184,
      "loss": 1.0049,
      "step": 1475
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.15625,
      "learning_rate": 0.0004445556655219294,
      "loss": 1.0436,
      "step": 1480
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00044419503355474767,
      "loss": 1.1623,
      "step": 1485
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00044383337982454907,
      "loss": 1.1004,
      "step": 1490
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0004434707062341856,
      "loss": 0.9782,
      "step": 1495
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0004431070146918754,
      "loss": 0.8625,
      "step": 1500
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00044274230711119247,
      "loss": 1.007,
      "step": 1505
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004423765854110565,
      "loss": 1.0016,
      "step": 1510
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0004420098515157235,
      "loss": 1.1458,
      "step": 1515
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.125,
      "learning_rate": 0.0004416421073547747,
      "loss": 1.1517,
      "step": 1520
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004412733548631072,
      "loss": 1.0341,
      "step": 1525
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00044090359598092333,
      "loss": 0.9788,
      "step": 1530
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0004405328326537205,
      "loss": 1.0459,
      "step": 1535
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004401610668322812,
      "loss": 1.021,
      "step": 1540
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00043978830047266247,
      "loss": 0.9662,
      "step": 1545
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004394145355361857,
      "loss": 0.8826,
      "step": 1550
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0004390397739894265,
      "loss": 1.3257,
      "step": 1555
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0004386640178042038,
      "loss": 1.13,
      "step": 1560
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0004382872689575703,
      "loss": 1.1859,
      "step": 1565
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0004379095294318013,
      "loss": 1.0333,
      "step": 1570
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00043753080121438465,
      "loss": 0.9258,
      "step": 1575
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0004371510862980104,
      "loss": 1.008,
      "step": 1580
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00043677038668055987,
      "loss": 1.1719,
      "step": 1585
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00043638870436509567,
      "loss": 1.3385,
      "step": 1590
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0004360060413598507,
      "loss": 0.9583,
      "step": 1595
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1171875,
      "learning_rate": 0.000435622399678218,
      "loss": 0.921,
      "step": 1600
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0004352377813387398,
      "loss": 1.1462,
      "step": 1605
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.000434852188365097,
      "loss": 1.1806,
      "step": 1610
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0004344656227860987,
      "loss": 1.239,
      "step": 1615
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0004340780866356713,
      "loss": 1.1628,
      "step": 1620
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.000433689581952848,
      "loss": 1.1506,
      "step": 1625
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.19140625,
      "learning_rate": 0.0004333001107817578,
      "loss": 1.1992,
      "step": 1630
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00043290967517161505,
      "loss": 1.1323,
      "step": 1635
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00043251827717670846,
      "loss": 1.2102,
      "step": 1640
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0004321259188563904,
      "loss": 1.0335,
      "step": 1645
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.177734375,
      "learning_rate": 0.00043173260227506586,
      "loss": 1.286,
      "step": 1650
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.000431338329502182,
      "loss": 1.155,
      "step": 1655
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0004309431026122169,
      "loss": 1.1915,
      "step": 1660
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004305469236846686,
      "loss": 0.8947,
      "step": 1665
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004301497948040446,
      "loss": 1.0983,
      "step": 1670
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0004297517180598504,
      "loss": 1.1136,
      "step": 1675
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00042935269554657875,
      "loss": 1.1264,
      "step": 1680
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00042895272936369875,
      "loss": 0.9479,
      "step": 1685
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0004285518216156444,
      "loss": 1.1757,
      "step": 1690
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.181640625,
      "learning_rate": 0.00042814997441180386,
      "loss": 1.0151,
      "step": 1695
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0004277471898665084,
      "loss": 0.9314,
      "step": 1700
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00042734347009902097,
      "loss": 1.1818,
      "step": 1705
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004269388172335253,
      "loss": 1.0081,
      "step": 1710
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00042653323339911464,
      "loss": 1.2873,
      "step": 1715
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0004261267207297804,
      "loss": 0.9362,
      "step": 1720
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0004257192813644015,
      "loss": 1.0657,
      "step": 1725
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00042531091744673233,
      "loss": 1.0267,
      "step": 1730
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004249016311253918,
      "loss": 0.9564,
      "step": 1735
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00042449142455385235,
      "loss": 1.1318,
      "step": 1740
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00042408029989042816,
      "loss": 1.1812,
      "step": 1745
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00042366825929826416,
      "loss": 1.166,
      "step": 1750
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.171875,
      "learning_rate": 0.000423255304945324,
      "loss": 0.9768,
      "step": 1755
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00042284143900437986,
      "loss": 0.9451,
      "step": 1760
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.0004224266636529995,
      "loss": 1.0424,
      "step": 1765
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.109375,
      "learning_rate": 0.0004220109810735362,
      "loss": 1.1772,
      "step": 1770
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004215943934531162,
      "loss": 0.9938,
      "step": 1775
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.00042117690298362775,
      "loss": 1.0018,
      "step": 1780
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0004207585118617098,
      "loss": 1.0415,
      "step": 1785
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00042033922228873954,
      "loss": 1.203,
      "step": 1790
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0004199190364708218,
      "loss": 1.0172,
      "step": 1795
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.000419497956618777,
      "loss": 1.0916,
      "step": 1800
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.00041907598494812933,
      "loss": 1.0867,
      "step": 1805
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00041865312367909545,
      "loss": 1.0811,
      "step": 1810
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00041822937503657276,
      "loss": 1.0382,
      "step": 1815
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004178047412501275,
      "loss": 1.0956,
      "step": 1820
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00041737922455398305,
      "loss": 1.0508,
      "step": 1825
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.00041695282718700834,
      "loss": 1.1273,
      "step": 1830
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00041652555139270585,
      "loss": 1.1146,
      "step": 1835
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004160973994192003,
      "loss": 0.9018,
      "step": 1840
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0004156683735192259,
      "loss": 0.973,
      "step": 1845
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0004152384759501156,
      "loss": 0.999,
      "step": 1850
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.181640625,
      "learning_rate": 0.00041480770897378826,
      "loss": 1.1376,
      "step": 1855
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0004143760748567373,
      "loss": 1.0657,
      "step": 1860
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0004139435758700186,
      "loss": 0.9538,
      "step": 1865
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0004135102142892387,
      "loss": 1.0505,
      "step": 1870
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0004130759923945426,
      "loss": 0.8248,
      "step": 1875
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.125,
      "learning_rate": 0.0004126409124706018,
      "loss": 1.1633,
      "step": 1880
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004122049768066025,
      "loss": 1.0844,
      "step": 1885
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.00041176818769623325,
      "loss": 1.1534,
      "step": 1890
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00041133054743767305,
      "loss": 1.004,
      "step": 1895
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.140625,
      "learning_rate": 0.00041089205833357944,
      "loss": 1.1132,
      "step": 1900
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00041045272269107604,
      "loss": 1.2416,
      "step": 1905
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0004100125428217405,
      "loss": 1.0682,
      "step": 1910
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0004095715210415925,
      "loss": 1.2839,
      "step": 1915
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004091296596710812,
      "loss": 1.0805,
      "step": 1920
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00040868696103507385,
      "loss": 1.1538,
      "step": 1925
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0004082434274628425,
      "loss": 1.0307,
      "step": 1930
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00040779906128805235,
      "loss": 1.1501,
      "step": 1935
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0004073538648487495,
      "loss": 0.9575,
      "step": 1940
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0004069078404873483,
      "loss": 0.9019,
      "step": 1945
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004064609905506195,
      "loss": 1.0083,
      "step": 1950
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.140625,
      "learning_rate": 0.00040601331738967736,
      "loss": 1.176,
      "step": 1955
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0004055648233599678,
      "loss": 0.9837,
      "step": 1960
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00040511551082125563,
      "loss": 1.1932,
      "step": 1965
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0004046653821376123,
      "loss": 0.9395,
      "step": 1970
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00040421443967740344,
      "loss": 1.1488,
      "step": 1975
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.00040376268581327645,
      "loss": 1.0614,
      "step": 1980
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.00040331012292214776,
      "loss": 1.0927,
      "step": 1985
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00040285675338519065,
      "loss": 0.9929,
      "step": 1990
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004024025795878228,
      "loss": 1.0894,
      "step": 1995
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00040194760391969305,
      "loss": 1.2229,
      "step": 2000
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.125,
      "learning_rate": 0.00040149182877466974,
      "loss": 1.1571,
      "step": 2005
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00040103525655082746,
      "loss": 0.8849,
      "step": 2010
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.181640625,
      "learning_rate": 0.00040057788965043475,
      "loss": 1.0083,
      "step": 2015
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00040011973047994134,
      "loss": 1.0084,
      "step": 2020
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0003996607814499654,
      "loss": 0.945,
      "step": 2025
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0003992010449752812,
      "loss": 0.9116,
      "step": 2030
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00039874052347480606,
      "loss": 1.1155,
      "step": 2035
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0003982792193715875,
      "loss": 1.0788,
      "step": 2040
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00039781713509279125,
      "loss": 1.1903,
      "step": 2045
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.216796875,
      "learning_rate": 0.00039735427306968746,
      "loss": 1.258,
      "step": 2050
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0003968906357376387,
      "loss": 0.9277,
      "step": 2055
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0003964262255360867,
      "loss": 1.1963,
      "step": 2060
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0003959610449085397,
      "loss": 0.841,
      "step": 2065
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0003954950963025595,
      "loss": 1.0462,
      "step": 2070
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1640625,
      "learning_rate": 0.000395028382169749,
      "loss": 1.0065,
      "step": 2075
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.00039456090496573843,
      "loss": 0.9813,
      "step": 2080
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00039409266715017323,
      "loss": 1.0473,
      "step": 2085
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00039362367118670084,
      "loss": 1.1596,
      "step": 2090
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00039315391954295755,
      "loss": 1.1368,
      "step": 2095
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0003926834146905558,
      "loss": 1.2268,
      "step": 2100
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.00039221215910507104,
      "loss": 1.0425,
      "step": 2105
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0003917401552660287,
      "loss": 1.1053,
      "step": 2110
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00039126740565689126,
      "loss": 1.053,
      "step": 2115
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.15625,
      "learning_rate": 0.00039079391276504504,
      "loss": 1.2397,
      "step": 2120
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.00039031967908178703,
      "loss": 1.067,
      "step": 2125
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.00038984470710231217,
      "loss": 1.1106,
      "step": 2130
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0003893689993256997,
      "loss": 1.0861,
      "step": 2135
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00038889255825490053,
      "loss": 0.9505,
      "step": 2140
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00038841538639672367,
      "loss": 1.1636,
      "step": 2145
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.375,
      "learning_rate": 0.00038793748626182316,
      "loss": 1.0212,
      "step": 2150
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.000387458860364685,
      "loss": 1.0993,
      "step": 2155
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.140625,
      "learning_rate": 0.00038697951122361376,
      "loss": 1.1088,
      "step": 2160
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0003864994413607194,
      "loss": 1.1727,
      "step": 2165
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0003860186533019039,
      "loss": 0.9805,
      "step": 2170
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0003855371495768482,
      "loss": 1.0571,
      "step": 2175
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0003850549327189985,
      "loss": 1.2412,
      "step": 2180
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00038457200526555347,
      "loss": 1.0324,
      "step": 2185
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00038408836975745034,
      "loss": 1.1788,
      "step": 2190
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00038360402873935194,
      "loss": 1.0882,
      "step": 2195
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00038311898475963316,
      "loss": 0.904,
      "step": 2200
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0003826332403703675,
      "loss": 1.0482,
      "step": 2205
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0003821467981273136,
      "loss": 1.176,
      "step": 2210
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.177734375,
      "learning_rate": 0.00038165966058990213,
      "loss": 1.1139,
      "step": 2215
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0003811718303212218,
      "loss": 0.9585,
      "step": 2220
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00038068330988800625,
      "loss": 1.1465,
      "step": 2225
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0003801941018606204,
      "loss": 0.9928,
      "step": 2230
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0003797042088130472,
      "loss": 0.9144,
      "step": 2235
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0003792136333228735,
      "loss": 0.8392,
      "step": 2240
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.00037872237797127717,
      "loss": 1.096,
      "step": 2245
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00037823044534301286,
      "loss": 1.0438,
      "step": 2250
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.00037773783802639904,
      "loss": 1.0796,
      "step": 2255
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0003772445586133039,
      "loss": 1.013,
      "step": 2260
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00037675060969913186,
      "loss": 1.03,
      "step": 2265
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0003762559938828101,
      "loss": 0.8575,
      "step": 2270
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.00037576071376677463,
      "loss": 1.0536,
      "step": 2275
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00037526477195695663,
      "loss": 0.9726,
      "step": 2280
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.140625,
      "learning_rate": 0.0003747681710627689,
      "loss": 1.0853,
      "step": 2285
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.000374270913697092,
      "loss": 1.1967,
      "step": 2290
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00037377300247626056,
      "loss": 0.995,
      "step": 2295
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0003732744400200494,
      "loss": 1.1294,
      "step": 2300
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0003727752289516599,
      "loss": 0.9954,
      "step": 2305
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.140625,
      "learning_rate": 0.00037227537189770624,
      "loss": 1.0589,
      "step": 2310
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0003717748714882012,
      "loss": 1.0241,
      "step": 2315
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.00037127373035654286,
      "loss": 1.0478,
      "step": 2320
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0003707719511395004,
      "loss": 1.0938,
      "step": 2325
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0003702695364772002,
      "loss": 1.1013,
      "step": 2330
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0003697664890131121,
      "loss": 0.9581,
      "step": 2335
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0003692628113940356,
      "loss": 1.0603,
      "step": 2340
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0003687585062700856,
      "loss": 0.9877,
      "step": 2345
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0003682535762946788,
      "loss": 1.1015,
      "step": 2350
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00036774802412451927,
      "loss": 0.834,
      "step": 2355
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2197265625,
      "learning_rate": 0.0003672418524195851,
      "loss": 1.2594,
      "step": 2360
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0003667350638431139,
      "loss": 1.1488,
      "step": 2365
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0003662276610615891,
      "loss": 1.3167,
      "step": 2370
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.00036571964674472567,
      "loss": 0.9609,
      "step": 2375
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0003652110235654563,
      "loss": 1.2101,
      "step": 2380
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00036470179419991703,
      "loss": 1.0423,
      "step": 2385
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0003641919613274336,
      "loss": 1.0078,
      "step": 2390
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00036368152763050697,
      "loss": 0.9718,
      "step": 2395
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00036317049579479954,
      "loss": 1.115,
      "step": 2400
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.177734375,
      "learning_rate": 0.00036265886850912044,
      "loss": 1.1202,
      "step": 2405
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0003621466484654123,
      "loss": 0.9148,
      "step": 2410
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0003616338383587362,
      "loss": 1.1323,
      "step": 2415
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00036112044088725806,
      "loss": 1.0034,
      "step": 2420
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00036060645875223387,
      "loss": 1.0439,
      "step": 2425
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00036009189465799633,
      "loss": 0.9872,
      "step": 2430
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0003595767513119399,
      "loss": 1.0703,
      "step": 2435
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.00035906103142450675,
      "loss": 1.1819,
      "step": 2440
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00035854473770917264,
      "loss": 1.106,
      "step": 2445
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.171875,
      "learning_rate": 0.0003580278728824324,
      "loss": 1.0991,
      "step": 2450
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00035751043966378613,
      "loss": 1.0253,
      "step": 2455
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.125,
      "learning_rate": 0.000356992440775724,
      "loss": 1.0299,
      "step": 2460
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00035647387894371284,
      "loss": 0.9544,
      "step": 2465
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0003559547568961815,
      "loss": 1.0262,
      "step": 2470
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00035543507736450616,
      "loss": 1.0388,
      "step": 2475
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.140625,
      "learning_rate": 0.0003549148430829962,
      "loss": 1.2653,
      "step": 2480
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15625,
      "learning_rate": 0.00035439405678888014,
      "loss": 1.045,
      "step": 2485
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.17578125,
      "learning_rate": 0.00035387272122229063,
      "loss": 1.2579,
      "step": 2490
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.173828125,
      "learning_rate": 0.00035335083912625063,
      "loss": 1.4965,
      "step": 2495
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.19921875,
      "learning_rate": 0.00035282841324665826,
      "loss": 1.2086,
      "step": 2500
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0003523054463322732,
      "loss": 1.0837,
      "step": 2505
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0003517819411347015,
      "loss": 0.969,
      "step": 2510
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00035125790040838164,
      "loss": 0.9484,
      "step": 2515
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0003507333269105695,
      "loss": 1.0408,
      "step": 2520
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0003502082234013246,
      "loss": 0.8761,
      "step": 2525
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0003496825926434947,
      "loss": 1.0746,
      "step": 2530
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0003491564374027018,
      "loss": 1.0468,
      "step": 2535
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.40234375,
      "learning_rate": 0.00034862976044732765,
      "loss": 1.0049,
      "step": 2540
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00034810256454849886,
      "loss": 1.2078,
      "step": 2545
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0003475748524800726,
      "loss": 0.9254,
      "step": 2550
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00034704662701862165,
      "loss": 1.0607,
      "step": 2555
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0003465178909434204,
      "loss": 1.216,
      "step": 2560
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0003459886470364295,
      "loss": 1.1468,
      "step": 2565
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0003454588980822817,
      "loss": 1.1476,
      "step": 2570
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00034492864686826716,
      "loss": 0.9272,
      "step": 2575
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0003443978961843185,
      "loss": 1.2118,
      "step": 2580
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00034386664882299643,
      "loss": 0.8843,
      "step": 2585
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0003433349075794751,
      "loss": 1.1248,
      "step": 2590
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0003428026752515269,
      "loss": 1.0207,
      "step": 2595
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00034226995463950837,
      "loss": 0.9798,
      "step": 2600
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.158203125,
      "learning_rate": 0.00034173674854634495,
      "loss": 0.8772,
      "step": 2605
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00034120305977751653,
      "loss": 1.0818,
      "step": 2610
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0003406688911410428,
      "loss": 1.0153,
      "step": 2615
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.000340134245447468,
      "loss": 1.0314,
      "step": 2620
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00033959912550984655,
      "loss": 0.9562,
      "step": 2625
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0003390635341437282,
      "loss": 0.9481,
      "step": 2630
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0003385274741671429,
      "loss": 1.224,
      "step": 2635
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0003379909484005865,
      "loss": 1.1988,
      "step": 2640
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2216796875,
      "learning_rate": 0.0003374539596670055,
      "loss": 0.9781,
      "step": 2645
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00033691651079178223,
      "loss": 0.9193,
      "step": 2650
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00033637860460272,
      "loss": 1.043,
      "step": 2655
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00033584024393002864,
      "loss": 1.0526,
      "step": 2660
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.171875,
      "learning_rate": 0.00033530143160630894,
      "loss": 1.1236,
      "step": 2665
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00033476217046653813,
      "loss": 1.185,
      "step": 2670
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00033422246334805503,
      "loss": 1.1898,
      "step": 2675
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0003336823130905449,
      "loss": 0.9399,
      "step": 2680
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.00033314172253602435,
      "loss": 1.031,
      "step": 2685
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.00033260069452882725,
      "loss": 0.9471,
      "step": 2690
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.181640625,
      "learning_rate": 0.00033205923191558835,
      "loss": 1.087,
      "step": 2695
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00033151733754522986,
      "loss": 1.3382,
      "step": 2700
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1328125,
      "learning_rate": 0.000330975014268945,
      "loss": 1.048,
      "step": 2705
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0003304322649401843,
      "loss": 0.8484,
      "step": 2710
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0003298890924146395,
      "loss": 0.8403,
      "step": 2715
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0003293454995502293,
      "loss": 1.211,
      "step": 2720
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00032880148920708384,
      "loss": 0.9348,
      "step": 2725
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0003282570642475301,
      "loss": 1.1156,
      "step": 2730
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0003277122275360762,
      "loss": 1.0835,
      "step": 2735
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00032716698193939697,
      "loss": 0.9661,
      "step": 2740
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0003266213303263186,
      "loss": 0.9852,
      "step": 2745
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0003260752755678034,
      "loss": 1.0605,
      "step": 2750
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.00032552882053693495,
      "loss": 1.3394,
      "step": 2755
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00032498196810890284,
      "loss": 1.1218,
      "step": 2760
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00032443472116098766,
      "loss": 1.2519,
      "step": 2765
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0003238870825725456,
      "loss": 0.8018,
      "step": 2770
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.00032333905522499376,
      "loss": 1.1112,
      "step": 2775
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0003227906420017942,
      "loss": 1.0559,
      "step": 2780
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.00032224184578843996,
      "loss": 1.1349,
      "step": 2785
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0003216926694724386,
      "loss": 1.0435,
      "step": 2790
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.142578125,
      "learning_rate": 0.000321143115943298,
      "loss": 0.7837,
      "step": 2795
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.17578125,
      "learning_rate": 0.00032059318809251046,
      "loss": 0.9842,
      "step": 2800
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00032004288881353816,
      "loss": 0.9675,
      "step": 2805
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0003194922210017971,
      "loss": 1.2367,
      "step": 2810
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00031894118755464266,
      "loss": 1.1448,
      "step": 2815
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0003183897913713538,
      "loss": 1.2115,
      "step": 2820
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0003178380353531183,
      "loss": 1.0475,
      "step": 2825
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.00031728592240301697,
      "loss": 1.2608,
      "step": 2830
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.00031673345542600873,
      "loss": 0.9939,
      "step": 2835
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0003161806373289153,
      "loss": 1.1296,
      "step": 2840
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0003156274710204056,
      "loss": 0.9441,
      "step": 2845
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0003150739594109809,
      "loss": 1.1399,
      "step": 2850
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0003145201054129592,
      "loss": 1.0609,
      "step": 2855
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00031396591194046003,
      "loss": 1.0824,
      "step": 2860
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00031341138190938897,
      "loss": 1.0619,
      "step": 2865
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00031285651823742246,
      "loss": 1.1303,
      "step": 2870
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0003123013238439925,
      "loss": 1.2609,
      "step": 2875
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.271484375,
      "learning_rate": 0.00031174580165027106,
      "loss": 1.2319,
      "step": 2880
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0003111899545791549,
      "loss": 0.8276,
      "step": 2885
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00031063378555525007,
      "loss": 1.0235,
      "step": 2890
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0003100772975048567,
      "loss": 0.8105,
      "step": 2895
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00030952049335595345,
      "loss": 1.1349,
      "step": 2900
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0003089633760381821,
      "loss": 1.0796,
      "step": 2905
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0003084059484828321,
      "loss": 0.9593,
      "step": 2910
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00030784821362282537,
      "loss": 1.0396,
      "step": 2915
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00030729017439270074,
      "loss": 0.9951,
      "step": 2920
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.228515625,
      "learning_rate": 0.00030673183372859834,
      "loss": 1.0801,
      "step": 2925
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00030617319456824446,
      "loss": 0.9871,
      "step": 2930
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0003056142598509358,
      "loss": 0.8944,
      "step": 2935
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00030505503251752425,
      "loss": 0.9589,
      "step": 2940
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00030449551551040124,
      "loss": 1.0908,
      "step": 2945
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.158203125,
      "learning_rate": 0.00030393571177348235,
      "loss": 0.9541,
      "step": 2950
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0003033756242521918,
      "loss": 1.004,
      "step": 2955
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.00030281525589344696,
      "loss": 1.2345,
      "step": 2960
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0003022546096456428,
      "loss": 1.3482,
      "step": 2965
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0003016936884586364,
      "loss": 1.1067,
      "step": 2970
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00030113249528373163,
      "loss": 1.0384,
      "step": 2975
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0003005710330736632,
      "loss": 0.8763,
      "step": 2980
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00030000930478258134,
      "loss": 1.0176,
      "step": 2985
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00029944731336603646,
      "loss": 1.1598,
      "step": 2990
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0002988850617809633,
      "loss": 0.9417,
      "step": 2995
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00029832255298566557,
      "loss": 0.8104,
      "step": 3000
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0002977597899398001,
      "loss": 1.0644,
      "step": 3005
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0002971967756043617,
      "loss": 1.1664,
      "step": 3010
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0002966335129416672,
      "loss": 1.0205,
      "step": 3015
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0002960700049153402,
      "loss": 1.1998,
      "step": 3020
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.140625,
      "learning_rate": 0.0002955062544902949,
      "loss": 1.0464,
      "step": 3025
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0002949422646327214,
      "loss": 1.3176,
      "step": 3030
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00029437803831006907,
      "loss": 0.9588,
      "step": 3035
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0002938135784910319,
      "loss": 1.0234,
      "step": 3040
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.00029324888814553213,
      "loss": 1.1289,
      "step": 3045
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00029268397024470507,
      "loss": 1.0552,
      "step": 3050
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0002921188277608832,
      "loss": 1.2105,
      "step": 3055
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1796875,
      "learning_rate": 0.00029155346366758074,
      "loss": 1.0885,
      "step": 3060
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0002909878809394779,
      "loss": 0.8802,
      "step": 3065
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0002904220825524052,
      "loss": 1.1406,
      "step": 3070
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.154296875,
      "learning_rate": 0.000289856071483328,
      "loss": 1.0026,
      "step": 3075
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0002892898507103305,
      "loss": 0.9414,
      "step": 3080
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0002887234232126004,
      "loss": 0.9942,
      "step": 3085
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00028815679197041283,
      "loss": 0.942,
      "step": 3090
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0002875899599651155,
      "loss": 1.0531,
      "step": 3095
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0002870229301791117,
      "loss": 1.0638,
      "step": 3100
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.15625,
      "learning_rate": 0.000286455705595846,
      "loss": 0.9991,
      "step": 3105
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0002858882891997874,
      "loss": 1.2492,
      "step": 3110
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00028532068397641466,
      "loss": 0.8871,
      "step": 3115
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0002847528929121994,
      "loss": 1.0626,
      "step": 3120
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00028418491899459173,
      "loss": 1.2234,
      "step": 3125
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00028361676521200324,
      "loss": 1.254,
      "step": 3130
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00028304843455379236,
      "loss": 1.267,
      "step": 3135
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.19921875,
      "learning_rate": 0.00028247993001024793,
      "loss": 1.1695,
      "step": 3140
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0002819112545725736,
      "loss": 1.1273,
      "step": 3145
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0002813424112328724,
      "loss": 0.9973,
      "step": 3150
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0002807734029841307,
      "loss": 1.049,
      "step": 3155
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00028020423282020256,
      "loss": 0.904,
      "step": 3160
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1796875,
      "learning_rate": 0.00027963490373579384,
      "loss": 1.1445,
      "step": 3165
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0002790654187264468,
      "loss": 1.2037,
      "step": 3170
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0002784957807885238,
      "loss": 0.8471,
      "step": 3175
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00027792599291919217,
      "loss": 1.1505,
      "step": 3180
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0002773560581164078,
      "loss": 1.1143,
      "step": 3185
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0002767859793789,
      "loss": 0.9737,
      "step": 3190
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.000276215759706155,
      "loss": 1.0739,
      "step": 3195
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0002756454020984009,
      "loss": 1.0384,
      "step": 3200
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0002750749095565914,
      "loss": 1.2285,
      "step": 3205
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0002745042850823902,
      "loss": 1.2136,
      "step": 3210
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0002739335316781551,
      "loss": 0.9552,
      "step": 3215
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.173828125,
      "learning_rate": 0.00027336265234692244,
      "loss": 1.1343,
      "step": 3220
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.171875,
      "learning_rate": 0.0002727916500923909,
      "loss": 0.8818,
      "step": 3225
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.189453125,
      "learning_rate": 0.00027222052791890607,
      "loss": 0.9985,
      "step": 3230
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.00027164928883144444,
      "loss": 0.8941,
      "step": 3235
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00027107793583559776,
      "loss": 1.0627,
      "step": 3240
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00027050647193755696,
      "loss": 1.1494,
      "step": 3245
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0002699349001440965,
      "loss": 0.956,
      "step": 3250
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0002693632234625586,
      "loss": 0.9758,
      "step": 3255
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0002687914449008375,
      "loss": 0.9468,
      "step": 3260
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0002682195674673631,
      "loss": 1.0775,
      "step": 3265
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00026764759417108585,
      "loss": 1.0727,
      "step": 3270
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0002670755280214605,
      "loss": 1.2369,
      "step": 3275
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0002665033720284302,
      "loss": 1.0324,
      "step": 3280
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.00026593112920241104,
      "loss": 1.0534,
      "step": 3285
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0002653588025542759,
      "loss": 1.025,
      "step": 3290
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00026478639509533853,
      "loss": 1.1727,
      "step": 3295
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0002642139098373382,
      "loss": 1.1025,
      "step": 3300
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0002636413497924231,
      "loss": 0.936,
      "step": 3305
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0002630687179731354,
      "loss": 1.075,
      "step": 3310
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.00026249601739239447,
      "loss": 0.7971,
      "step": 3315
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.15625,
      "learning_rate": 0.00026192325106348186,
      "loss": 1.0379,
      "step": 3320
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00026135042200002464,
      "loss": 0.8698,
      "step": 3325
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00026077753321598034,
      "loss": 0.9574,
      "step": 3330
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0002602045877256204,
      "loss": 0.9652,
      "step": 3335
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00025963158854351494,
      "loss": 1.0468,
      "step": 3340
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0002590585386845162,
      "loss": 1.0618,
      "step": 3345
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00025848544116374353,
      "loss": 1.0876,
      "step": 3350
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.140625,
      "learning_rate": 0.0002579122989965666,
      "loss": 1.1683,
      "step": 3355
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.00025733911519859023,
      "loss": 1.0189,
      "step": 3360
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.140625,
      "learning_rate": 0.0002567658927856382,
      "loss": 1.0906,
      "step": 3365
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0002561926347737376,
      "loss": 1.2646,
      "step": 3370
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00025561934417910265,
      "loss": 1.1042,
      "step": 3375
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00025504602401811913,
      "loss": 0.8099,
      "step": 3380
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0002544726773073282,
      "loss": 0.9141,
      "step": 3385
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.17578125,
      "learning_rate": 0.00025389930706341097,
      "loss": 1.2311,
      "step": 3390
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.00025332591630317227,
      "loss": 0.9525,
      "step": 3395
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.00025275250804352464,
      "loss": 1.1637,
      "step": 3400
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0002521790853014732,
      "loss": 1.1291,
      "step": 3405
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.125,
      "learning_rate": 0.00025160565109409873,
      "loss": 0.9412,
      "step": 3410
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00025103220843854273,
      "loss": 0.9922,
      "step": 3415
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002504587603519909,
      "loss": 0.9436,
      "step": 3420
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0002498853098516577,
      "loss": 1.463,
      "step": 3425
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0002493118599547702,
      "loss": 0.9141,
      "step": 3430
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00024873841367855216,
      "loss": 0.9547,
      "step": 3435
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0002481649740402086,
      "loss": 1.0718,
      "step": 3440
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00024759154405690946,
      "loss": 0.9502,
      "step": 3445
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.00024701812674577376,
      "loss": 1.0506,
      "step": 3450
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.000246444725123854,
      "loss": 0.9621,
      "step": 3455
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0002458713422081199,
      "loss": 1.2094,
      "step": 3460
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00024529798101544326,
      "loss": 1.1523,
      "step": 3465
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15234375,
      "learning_rate": 0.000244724644562581,
      "loss": 1.1936,
      "step": 3470
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00024415133586616036,
      "loss": 1.1543,
      "step": 3475
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00024357805794266205,
      "loss": 1.1209,
      "step": 3480
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.00024300481380840527,
      "loss": 1.0296,
      "step": 3485
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00024243160647953132,
      "loss": 0.9974,
      "step": 3490
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.00024185843897198784,
      "loss": 1.1089,
      "step": 3495
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.140625,
      "learning_rate": 0.00024128531430151276,
      "loss": 0.8318,
      "step": 3500
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00024071223548361895,
      "loss": 1.0249,
      "step": 3505
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00024013920553357784,
      "loss": 0.9857,
      "step": 3510
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0002395662274664038,
      "loss": 1.117,
      "step": 3515
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15625,
      "learning_rate": 0.0002389933042968381,
      "loss": 1.0439,
      "step": 3520
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0002384204390393332,
      "loss": 1.0804,
      "step": 3525
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00023784763470803712,
      "loss": 1.1855,
      "step": 3530
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.140625,
      "learning_rate": 0.00023727489431677688,
      "loss": 1.0133,
      "step": 3535
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.15625,
      "learning_rate": 0.00023670222087904342,
      "loss": 1.3227,
      "step": 3540
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.177734375,
      "learning_rate": 0.00023612961740797516,
      "loss": 1.3998,
      "step": 3545
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0002355570869163427,
      "loss": 1.0318,
      "step": 3550
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.171875,
      "learning_rate": 0.0002349846324165323,
      "loss": 1.0445,
      "step": 3555
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0002344122569205306,
      "loss": 1.1273,
      "step": 3560
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0002338399634399084,
      "loss": 1.0658,
      "step": 3565
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00023326775498580535,
      "loss": 1.0565,
      "step": 3570
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00023269563456891322,
      "loss": 0.9532,
      "step": 3575
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.00023212360519946095,
      "loss": 1.1475,
      "step": 3580
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00023155166988719805,
      "loss": 1.128,
      "step": 3585
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00023097983164137976,
      "loss": 0.8851,
      "step": 3590
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00023040809347074997,
      "loss": 1.0787,
      "step": 3595
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00022983645838352642,
      "loss": 1.0876,
      "step": 3600
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0002292649293873842,
      "loss": 0.948,
      "step": 3605
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00022869350948944038,
      "loss": 1.1694,
      "step": 3610
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.140625,
      "learning_rate": 0.00022812220169623798,
      "loss": 1.1419,
      "step": 3615
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00022755100901373026,
      "loss": 1.0157,
      "step": 3620
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.00022697993444726456,
      "loss": 1.11,
      "step": 3625
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00022640898100156686,
      "loss": 1.25,
      "step": 3630
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00022583815168072613,
      "loss": 0.9598,
      "step": 3635
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.171875,
      "learning_rate": 0.00022526744948817793,
      "loss": 1.3431,
      "step": 3640
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00022469687742668898,
      "loss": 0.9564,
      "step": 3645
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00022412643849834132,
      "loss": 1.142,
      "step": 3650
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00022355613570451677,
      "loss": 1.0124,
      "step": 3655
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00022298597204588043,
      "loss": 1.013,
      "step": 3660
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.140625,
      "learning_rate": 0.00022241595052236566,
      "loss": 1.0421,
      "step": 3665
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00022184607413315788,
      "loss": 0.9687,
      "step": 3670
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.00022127634587667874,
      "loss": 0.9082,
      "step": 3675
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0002207067687505708,
      "loss": 1.0893,
      "step": 3680
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00022013734575168116,
      "loss": 1.2716,
      "step": 3685
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00021956807987604616,
      "loss": 1.1467,
      "step": 3690
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00021899897411887517,
      "loss": 1.1248,
      "step": 3695
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.171875,
      "learning_rate": 0.00021843003147453544,
      "loss": 1.0846,
      "step": 3700
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.00021786125493653582,
      "loss": 0.9875,
      "step": 3705
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0002172926474975111,
      "loss": 1.1556,
      "step": 3710
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0002167242121492064,
      "loss": 0.9676,
      "step": 3715
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0002161559518824616,
      "loss": 1.0359,
      "step": 3720
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00021558786968719514,
      "loss": 1.0299,
      "step": 3725
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00021501996855238852,
      "loss": 1.1488,
      "step": 3730
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00021445225146607064,
      "loss": 1.1796,
      "step": 3735
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.00021388472141530218,
      "loss": 1.0853,
      "step": 3740
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00021331738138615958,
      "loss": 0.8584,
      "step": 3745
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0002127502343637194,
      "loss": 0.9085,
      "step": 3750
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.17578125,
      "learning_rate": 0.00021218328333204284,
      "loss": 1.1501,
      "step": 3755
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.00021161653127415968,
      "loss": 1.1757,
      "step": 3760
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00021104998117205316,
      "loss": 1.1097,
      "step": 3765
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0002104836360066436,
      "loss": 1.4605,
      "step": 3770
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00020991749875777315,
      "loss": 0.9118,
      "step": 3775
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0002093515724041899,
      "loss": 1.0906,
      "step": 3780
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00020878585992353252,
      "loss": 0.9324,
      "step": 3785
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00020822036429231426,
      "loss": 0.9894,
      "step": 3790
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00020765508848590738,
      "loss": 0.9968,
      "step": 3795
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0002070900354785275,
      "loss": 0.9817,
      "step": 3800
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00020652520824321813,
      "loss": 1.1311,
      "step": 3805
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00020596060975183473,
      "loss": 0.9224,
      "step": 3810
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00020539624297502933,
      "loss": 1.1655,
      "step": 3815
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00020483211088223458,
      "loss": 1.0407,
      "step": 3820
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00020426821644164858,
      "loss": 1.0116,
      "step": 3825
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00020370456262021897,
      "loss": 1.0392,
      "step": 3830
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00020314115238362712,
      "loss": 0.9167,
      "step": 3835
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00020257798869627308,
      "loss": 1.0841,
      "step": 3840
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0002020150745212595,
      "loss": 0.8587,
      "step": 3845
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.181640625,
      "learning_rate": 0.00020145241282037638,
      "loss": 1.0392,
      "step": 3850
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00020089000655408503,
      "loss": 0.97,
      "step": 3855
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00020032785868150308,
      "loss": 0.989,
      "step": 3860
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00019976597216038835,
      "loss": 0.8304,
      "step": 3865
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00019920434994712393,
      "loss": 1.0367,
      "step": 3870
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0001986429949967018,
      "loss": 1.2286,
      "step": 3875
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00019808191026270805,
      "loss": 1.0935,
      "step": 3880
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0001975210986973067,
      "loss": 1.0016,
      "step": 3885
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1953125,
      "learning_rate": 0.00019696056325122502,
      "loss": 1.0657,
      "step": 3890
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0001964003068737369,
      "loss": 0.9734,
      "step": 3895
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.00019584033251264824,
      "loss": 1.0914,
      "step": 3900
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00019528064311428092,
      "loss": 1.0734,
      "step": 3905
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00019472124162345756,
      "loss": 0.7545,
      "step": 3910
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.158203125,
      "learning_rate": 0.00019416213098348605,
      "loss": 1.0816,
      "step": 3915
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00019360331413614384,
      "loss": 1.0022,
      "step": 3920
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.177734375,
      "learning_rate": 0.00019304479402166256,
      "loss": 1.2418,
      "step": 3925
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00019248657357871256,
      "loss": 1.0155,
      "step": 3930
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0001919286557443876,
      "loss": 0.9751,
      "step": 3935
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0001913710434541892,
      "loss": 0.986,
      "step": 3940
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00019081373964201107,
      "loss": 0.9691,
      "step": 3945
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00019025674724012403,
      "loss": 1.2395,
      "step": 3950
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00018970006917916059,
      "loss": 1.0455,
      "step": 3955
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.158203125,
      "learning_rate": 0.00018914370838809885,
      "loss": 1.0043,
      "step": 3960
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.181640625,
      "learning_rate": 0.00018858766779424803,
      "loss": 1.1731,
      "step": 3965
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2177734375,
      "learning_rate": 0.00018803195032323231,
      "loss": 0.9617,
      "step": 3970
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00018747655889897612,
      "loss": 1.0404,
      "step": 3975
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.000186921496443688,
      "loss": 0.9949,
      "step": 3980
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00018636676587784585,
      "loss": 0.8834,
      "step": 3985
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0001858123701201812,
      "loss": 0.9534,
      "step": 3990
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.00018525831208766397,
      "loss": 0.9737,
      "step": 3995
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00018470459469548727,
      "loss": 0.9685,
      "step": 4000
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00018415122085705188,
      "loss": 1.2834,
      "step": 4005
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0001835981934839508,
      "loss": 0.991,
      "step": 4010
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00018304551548595422,
      "loss": 1.2289,
      "step": 4015
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.140625,
      "learning_rate": 0.0001824931897709941,
      "loss": 1.0199,
      "step": 4020
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00018194121924514892,
      "loss": 1.0528,
      "step": 4025
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15625,
      "learning_rate": 0.00018138960681262807,
      "loss": 0.9923,
      "step": 4030
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00018083835537575693,
      "loss": 1.0761,
      "step": 4035
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00018028746783496175,
      "loss": 1.1735,
      "step": 4040
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00017973694708875383,
      "loss": 0.9852,
      "step": 4045
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.00017918679603371455,
      "loss": 0.9332,
      "step": 4050
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00017863701756448035,
      "loss": 1.1593,
      "step": 4055
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00017808761457372706,
      "loss": 1.1001,
      "step": 4060
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00017753858995215539,
      "loss": 1.0817,
      "step": 4065
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.169921875,
      "learning_rate": 0.00017698994658847455,
      "loss": 1.2082,
      "step": 4070
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00017644168736938843,
      "loss": 1.0919,
      "step": 4075
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00017589381517957902,
      "loss": 0.988,
      "step": 4080
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0001753463329016928,
      "loss": 1.0371,
      "step": 4085
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00017479924341632397,
      "loss": 1.0561,
      "step": 4090
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00017425254960200048,
      "loss": 0.9408,
      "step": 4095
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.00017370625433516814,
      "loss": 1.1421,
      "step": 4100
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.140625,
      "learning_rate": 0.0001731603604901761,
      "loss": 1.1159,
      "step": 4105
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00017261487093926117,
      "loss": 1.1286,
      "step": 4110
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00017206978855253315,
      "loss": 0.9616,
      "step": 4115
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00017152511619795925,
      "loss": 0.9986,
      "step": 4120
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00017098085674134962,
      "loss": 1.0591,
      "step": 4125
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0001704370130463418,
      "loss": 1.1427,
      "step": 4130
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.169921875,
      "learning_rate": 0.00016989358797438563,
      "loss": 0.8409,
      "step": 4135
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00016935058438472862,
      "loss": 1.1067,
      "step": 4140
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.00016880800513440037,
      "loss": 1.2404,
      "step": 4145
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00016826585307819813,
      "loss": 1.1614,
      "step": 4150
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00016772413106867112,
      "loss": 1.0634,
      "step": 4155
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00016718284195610606,
      "loss": 1.0528,
      "step": 4160
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0001666419885885118,
      "loss": 0.9877,
      "step": 4165
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0001661015738116049,
      "loss": 1.1577,
      "step": 4170
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.197265625,
      "learning_rate": 0.00016556160046879377,
      "loss": 1.1811,
      "step": 4175
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00016502207140116446,
      "loss": 0.9547,
      "step": 4180
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00016448298944746526,
      "loss": 1.1406,
      "step": 4185
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00016394435744409238,
      "loss": 1.0288,
      "step": 4190
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00016340617822507416,
      "loss": 1.0023,
      "step": 4195
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.158203125,
      "learning_rate": 0.00016286845462205684,
      "loss": 1.1052,
      "step": 4200
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.140625,
      "learning_rate": 0.00016233118946428927,
      "loss": 0.9259,
      "step": 4205
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.171875,
      "learning_rate": 0.0001617943855786082,
      "loss": 0.9792,
      "step": 4210
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15625,
      "learning_rate": 0.00016125804578942364,
      "loss": 1.2032,
      "step": 4215
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00016072217291870344,
      "loss": 1.0617,
      "step": 4220
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2216796875,
      "learning_rate": 0.0001601867697859588,
      "loss": 1.2913,
      "step": 4225
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0001596518392082294,
      "loss": 0.8958,
      "step": 4230
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00015911738400006874,
      "loss": 0.9306,
      "step": 4235
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00015858340697352892,
      "loss": 0.9702,
      "step": 4240
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00015804991093814607,
      "loss": 0.9372,
      "step": 4245
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00015751689870092565,
      "loss": 1.1663,
      "step": 4250
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00015698437306632768,
      "loss": 0.7672,
      "step": 4255
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.201171875,
      "learning_rate": 0.00015645233683625188,
      "loss": 0.9745,
      "step": 4260
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0001559207928100227,
      "loss": 1.058,
      "step": 4265
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00015538974378437513,
      "loss": 0.9706,
      "step": 4270
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00015485919255343972,
      "loss": 1.1274,
      "step": 4275
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00015432914190872756,
      "loss": 0.9899,
      "step": 4280
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00015379959463911622,
      "loss": 1.0313,
      "step": 4285
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0001532705535308346,
      "loss": 0.9323,
      "step": 4290
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00015274202136744824,
      "loss": 0.8996,
      "step": 4295
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00015221400092984527,
      "loss": 0.8551,
      "step": 4300
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00015168649499622104,
      "loss": 0.9615,
      "step": 4305
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0001511595063420638,
      "loss": 1.0077,
      "step": 4310
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.169921875,
      "learning_rate": 0.00015063303774014036,
      "loss": 0.9879,
      "step": 4315
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00015010709196048105,
      "loss": 1.1628,
      "step": 4320
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.00014958167177036552,
      "loss": 1.1096,
      "step": 4325
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0001490567799343078,
      "loss": 0.9157,
      "step": 4330
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00014853241921404208,
      "loss": 1.311,
      "step": 4335
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.390625,
      "learning_rate": 0.00014800859236850823,
      "loss": 1.1399,
      "step": 4340
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00014748530215383693,
      "loss": 1.2342,
      "step": 4345
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00014696255132333529,
      "loss": 1.196,
      "step": 4350
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00014644034262747255,
      "loss": 0.9654,
      "step": 4355
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0001459186788138656,
      "loss": 1.0736,
      "step": 4360
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00014539756262726435,
      "loss": 1.0585,
      "step": 4365
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0001448769968095372,
      "loss": 0.9221,
      "step": 4370
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00014435698409965674,
      "loss": 1.0557,
      "step": 4375
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0001438375272336857,
      "loss": 1.1002,
      "step": 4380
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00014331862894476205,
      "loss": 1.1207,
      "step": 4385
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00014280029196308473,
      "loss": 0.9124,
      "step": 4390
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0001422825190158992,
      "loss": 1.3247,
      "step": 4395
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.212890625,
      "learning_rate": 0.00014176531282748356,
      "loss": 1.0844,
      "step": 4400
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00014124867611913384,
      "loss": 1.1084,
      "step": 4405
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00014073261160914952,
      "loss": 1.1076,
      "step": 4410
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00014021712201281943,
      "loss": 0.9841,
      "step": 4415
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00013970221004240782,
      "loss": 1.0774,
      "step": 4420
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00013918787840713954,
      "loss": 1.3116,
      "step": 4425
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.00013867412981318584,
      "loss": 1.1418,
      "step": 4430
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00013816096696365042,
      "loss": 0.9842,
      "step": 4435
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1953125,
      "learning_rate": 0.0001376483925585552,
      "loss": 1.1092,
      "step": 4440
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.173828125,
      "learning_rate": 0.00013713640929482562,
      "loss": 0.9737,
      "step": 4445
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00013662501986627732,
      "loss": 1.1233,
      "step": 4450
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00013611422696360094,
      "loss": 1.0444,
      "step": 4455
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0001356040332743489,
      "loss": 1.2116,
      "step": 4460
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0001350944414829205,
      "loss": 1.0156,
      "step": 4465
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00013458545427054842,
      "loss": 1.1135,
      "step": 4470
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0001340770743152842,
      "loss": 1.3632,
      "step": 4475
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0001335693042919841,
      "loss": 1.0433,
      "step": 4480
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00013306214687229558,
      "loss": 0.8539,
      "step": 4485
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0001325556047246427,
      "loss": 1.0244,
      "step": 4490
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00013204968051421214,
      "loss": 0.9696,
      "step": 4495
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0001315443769029393,
      "loss": 1.2288,
      "step": 4500
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.140625,
      "learning_rate": 0.0001310396965494944,
      "loss": 0.9919,
      "step": 4505
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00013053564210926844,
      "loss": 0.9188,
      "step": 4510
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.19921875,
      "learning_rate": 0.00013003221623435895,
      "loss": 1.2513,
      "step": 4515
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.00012952942157355617,
      "loss": 1.0146,
      "step": 4520
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.00012902726077232966,
      "loss": 0.9834,
      "step": 4525
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.130859375,
      "learning_rate": 0.00012852573647281328,
      "loss": 1.067,
      "step": 4530
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0001280248513137925,
      "loss": 0.9357,
      "step": 4535
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0001275246079306895,
      "loss": 0.9427,
      "step": 4540
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00012702500895554996,
      "loss": 1.0688,
      "step": 4545
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00012652605701702892,
      "loss": 1.0269,
      "step": 4550
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0001260277547403771,
      "loss": 1.1006,
      "step": 4555
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00012553010474742674,
      "loss": 0.9569,
      "step": 4560
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.00012503310965657853,
      "loss": 1.0812,
      "step": 4565
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.140625,
      "learning_rate": 0.00012453677208278685,
      "loss": 1.1446,
      "step": 4570
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.00012404109463754704,
      "loss": 0.8849,
      "step": 4575
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00012354607992888064,
      "loss": 1.0462,
      "step": 4580
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0001230517305613227,
      "loss": 1.0097,
      "step": 4585
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00012255804913590722,
      "loss": 1.2159,
      "step": 4590
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00012206503825015373,
      "loss": 1.0315,
      "step": 4595
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.169921875,
      "learning_rate": 0.00012157270049805405,
      "loss": 1.2781,
      "step": 4600
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00012108103847005791,
      "loss": 0.9739,
      "step": 4605
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00012059005475306011,
      "loss": 1.2582,
      "step": 4610
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00012009975193038603,
      "loss": 0.9291,
      "step": 4615
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00011961013258177892,
      "loss": 0.7847,
      "step": 4620
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00011912119928338553,
      "loss": 0.9385,
      "step": 4625
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00011863295460774334,
      "loss": 0.982,
      "step": 4630
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00011814540112376623,
      "loss": 1.2993,
      "step": 4635
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.140625,
      "learning_rate": 0.00011765854139673174,
      "loss": 0.9739,
      "step": 4640
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00011717237798826677,
      "loss": 0.9895,
      "step": 4645
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.00011668691345633492,
      "loss": 1.1416,
      "step": 4650
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.00011620215035522225,
      "loss": 1.1118,
      "step": 4655
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00011571809123552462,
      "loss": 1.258,
      "step": 4660
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15625,
      "learning_rate": 0.0001152347386441335,
      "loss": 1.038,
      "step": 4665
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.00011475209512422324,
      "loss": 1.078,
      "step": 4670
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00011427016321523717,
      "loss": 0.9228,
      "step": 4675
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.203125,
      "learning_rate": 0.0001137889454528745,
      "loss": 1.118,
      "step": 4680
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.16796875,
      "learning_rate": 0.000113308444369077,
      "loss": 1.1574,
      "step": 4685
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00011282866249201574,
      "loss": 1.1618,
      "step": 4690
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00011234960234607739,
      "loss": 1.2177,
      "step": 4695
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.00011187126645185122,
      "loss": 1.225,
      "step": 4700
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00011139365732611611,
      "loss": 1.0852,
      "step": 4705
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00011091677748182688,
      "loss": 1.0386,
      "step": 4710
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00011044062942810113,
      "loss": 0.9717,
      "step": 4715
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00010996521567020607,
      "loss": 0.915,
      "step": 4720
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0001094905387095457,
      "loss": 1.0127,
      "step": 4725
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.00010901660104364697,
      "loss": 0.9142,
      "step": 4730
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00010854340516614736,
      "loss": 1.1216,
      "step": 4735
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.189453125,
      "learning_rate": 0.00010807095356678112,
      "loss": 1.0968,
      "step": 4740
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.169921875,
      "learning_rate": 0.00010759924873136647,
      "loss": 1.0457,
      "step": 4745
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00010712829314179274,
      "loss": 0.9762,
      "step": 4750
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.000106658089276007,
      "loss": 0.971,
      "step": 4755
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00010618863960800096,
      "loss": 0.9612,
      "step": 4760
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0001057199466077981,
      "loss": 0.9232,
      "step": 4765
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00010525201274144087,
      "loss": 1.1631,
      "step": 4770
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00010478484047097745,
      "loss": 1.1005,
      "step": 4775
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.140625,
      "learning_rate": 0.00010431843225444873,
      "loss": 0.9329,
      "step": 4780
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00010385279054587551,
      "loss": 1.0208,
      "step": 4785
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00010338791779524573,
      "loss": 1.0081,
      "step": 4790
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.28515625,
      "learning_rate": 0.00010292381644850149,
      "loss": 1.1968,
      "step": 4795
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.00010246048894752588,
      "loss": 1.2157,
      "step": 4800
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00010199793773013041,
      "loss": 0.9195,
      "step": 4805
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00010153616523004233,
      "loss": 0.959,
      "step": 4810
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00010107517387689166,
      "loss": 1.0178,
      "step": 4815
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0001006149660961982,
      "loss": 0.8926,
      "step": 4820
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.171875,
      "learning_rate": 0.00010015554430935902,
      "loss": 1.151,
      "step": 4825
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12451171875,
      "learning_rate": 9.969691093363553e-05,
      "loss": 0.9893,
      "step": 4830
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1357421875,
      "learning_rate": 9.923906838214138e-05,
      "loss": 1.0514,
      "step": 4835
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1416015625,
      "learning_rate": 9.878201906382881e-05,
      "loss": 1.0175,
      "step": 4840
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1533203125,
      "learning_rate": 9.832576538347652e-05,
      "loss": 1.1843,
      "step": 4845
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.140625,
      "learning_rate": 9.787030974167694e-05,
      "loss": 1.0639,
      "step": 4850
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1640625,
      "learning_rate": 9.741565453482368e-05,
      "loss": 1.0156,
      "step": 4855
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.162109375,
      "learning_rate": 9.696180215509892e-05,
      "loss": 1.1617,
      "step": 4860
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1494140625,
      "learning_rate": 9.650875499046058e-05,
      "loss": 0.9796,
      "step": 4865
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1455078125,
      "learning_rate": 9.605651542462979e-05,
      "loss": 0.7412,
      "step": 4870
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1533203125,
      "learning_rate": 9.560508583707878e-05,
      "loss": 1.1196,
      "step": 4875
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1474609375,
      "learning_rate": 9.515446860301802e-05,
      "loss": 0.7949,
      "step": 4880
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1435546875,
      "learning_rate": 9.470466609338351e-05,
      "loss": 1.1594,
      "step": 4885
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.169921875,
      "learning_rate": 9.425568067482468e-05,
      "loss": 1.0454,
      "step": 4890
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.193359375,
      "learning_rate": 9.380751470969186e-05,
      "loss": 1.2346,
      "step": 4895
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1455078125,
      "learning_rate": 9.33601705560239e-05,
      "loss": 0.9849,
      "step": 4900
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1865234375,
      "learning_rate": 9.291365056753543e-05,
      "loss": 1.1268,
      "step": 4905
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1494140625,
      "learning_rate": 9.246795709360472e-05,
      "loss": 0.9571,
      "step": 4910
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1494140625,
      "learning_rate": 9.20230924792615e-05,
      "loss": 1.2293,
      "step": 4915
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.15234375,
      "learning_rate": 9.157905906517444e-05,
      "loss": 0.9942,
      "step": 4920
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.140625,
      "learning_rate": 9.113585918763858e-05,
      "loss": 1.0211,
      "step": 4925
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1416015625,
      "learning_rate": 9.06934951785634e-05,
      "loss": 1.0026,
      "step": 4930
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1611328125,
      "learning_rate": 9.025196936546051e-05,
      "loss": 1.0391,
      "step": 4935
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1748046875,
      "learning_rate": 8.981128407143136e-05,
      "loss": 1.0796,
      "step": 4940
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.142578125,
      "learning_rate": 8.937144161515481e-05,
      "loss": 1.2019,
      "step": 4945
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1865234375,
      "learning_rate": 8.893244431087516e-05,
      "loss": 1.1303,
      "step": 4950
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1669921875,
      "learning_rate": 8.849429446838983e-05,
      "loss": 1.098,
      "step": 4955
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.142578125,
      "learning_rate": 8.805699439303772e-05,
      "loss": 1.205,
      "step": 4960
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.162109375,
      "learning_rate": 8.762054638568614e-05,
      "loss": 1.1974,
      "step": 4965
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.162109375,
      "learning_rate": 8.718495274271951e-05,
      "loss": 0.8459,
      "step": 4970
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14453125,
      "learning_rate": 8.675021575602674e-05,
      "loss": 1.0461,
      "step": 4975
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1416015625,
      "learning_rate": 8.631633771298969e-05,
      "loss": 1.0227,
      "step": 4980
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13671875,
      "learning_rate": 8.588332089647083e-05,
      "loss": 0.9869,
      "step": 4985
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1474609375,
      "learning_rate": 8.545116758480109e-05,
      "loss": 1.0406,
      "step": 4990
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.12158203125,
      "learning_rate": 8.501988005176805e-05,
      "loss": 0.9332,
      "step": 4995
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1875,
      "learning_rate": 8.458946056660413e-05,
      "loss": 1.1838,
      "step": 5000
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1455078125,
      "learning_rate": 8.415991139397452e-05,
      "loss": 1.1195,
      "step": 5005
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.171875,
      "learning_rate": 8.373123479396505e-05,
      "loss": 1.1509,
      "step": 5010
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.16796875,
      "learning_rate": 8.330343302207042e-05,
      "loss": 1.0687,
      "step": 5015
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.177734375,
      "learning_rate": 8.287650832918267e-05,
      "loss": 1.1246,
      "step": 5020
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.142578125,
      "learning_rate": 8.245046296157898e-05,
      "loss": 1.0336,
      "step": 5025
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1845703125,
      "learning_rate": 8.202529916090983e-05,
      "loss": 1.1219,
      "step": 5030
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.154296875,
      "learning_rate": 8.160101916418728e-05,
      "loss": 1.0273,
      "step": 5035
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1435546875,
      "learning_rate": 8.117762520377334e-05,
      "loss": 0.9658,
      "step": 5040
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1708984375,
      "learning_rate": 8.075511950736819e-05,
      "loss": 1.0894,
      "step": 5045
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.23828125,
      "learning_rate": 8.033350429799821e-05,
      "loss": 1.1085,
      "step": 5050
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1767578125,
      "learning_rate": 7.99127817940044e-05,
      "loss": 0.945,
      "step": 5055
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1455078125,
      "learning_rate": 7.94929542090311e-05,
      "loss": 1.1813,
      "step": 5060
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1396484375,
      "learning_rate": 7.907402375201362e-05,
      "loss": 0.9865,
      "step": 5065
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1572265625,
      "learning_rate": 7.86559926271673e-05,
      "loss": 1.1039,
      "step": 5070
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1650390625,
      "learning_rate": 7.823886303397543e-05,
      "loss": 0.9412,
      "step": 5075
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13671875,
      "learning_rate": 7.782263716717783e-05,
      "loss": 0.8976,
      "step": 5080
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.19921875,
      "learning_rate": 7.740731721675953e-05,
      "loss": 1.0748,
      "step": 5085
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1513671875,
      "learning_rate": 7.699290536793904e-05,
      "loss": 0.9474,
      "step": 5090
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1640625,
      "learning_rate": 7.657940380115668e-05,
      "loss": 1.0187,
      "step": 5095
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.140625,
      "learning_rate": 7.616681469206335e-05,
      "loss": 1.0857,
      "step": 5100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1328125,
      "learning_rate": 7.575514021150915e-05,
      "loss": 0.992,
      "step": 5105
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.16015625,
      "learning_rate": 7.534438252553177e-05,
      "loss": 0.9423,
      "step": 5110
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1689453125,
      "learning_rate": 7.493454379534515e-05,
      "loss": 0.9686,
      "step": 5115
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.134765625,
      "learning_rate": 7.452562617732795e-05,
      "loss": 0.965,
      "step": 5120
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.150390625,
      "learning_rate": 7.411763182301275e-05,
      "loss": 0.9155,
      "step": 5125
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1630859375,
      "learning_rate": 7.371056287907396e-05,
      "loss": 1.1172,
      "step": 5130
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1396484375,
      "learning_rate": 7.330442148731726e-05,
      "loss": 1.1276,
      "step": 5135
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1630859375,
      "learning_rate": 7.289920978466758e-05,
      "loss": 0.887,
      "step": 5140
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.158203125,
      "learning_rate": 7.249492990315873e-05,
      "loss": 1.0794,
      "step": 5145
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1767578125,
      "learning_rate": 7.209158396992133e-05,
      "loss": 1.1647,
      "step": 5150
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.16015625,
      "learning_rate": 7.168917410717224e-05,
      "loss": 1.1844,
      "step": 5155
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13671875,
      "learning_rate": 7.128770243220298e-05,
      "loss": 1.1793,
      "step": 5160
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1982421875,
      "learning_rate": 7.088717105736897e-05,
      "loss": 1.2339,
      "step": 5165
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1650390625,
      "learning_rate": 7.048758209007797e-05,
      "loss": 1.114,
      "step": 5170
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.14453125,
      "learning_rate": 7.008893763277952e-05,
      "loss": 1.1371,
      "step": 5175
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.150390625,
      "learning_rate": 6.969123978295325e-05,
      "loss": 0.9877,
      "step": 5180
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1923828125,
      "learning_rate": 6.929449063309857e-05,
      "loss": 1.1437,
      "step": 5185
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1923828125,
      "learning_rate": 6.88986922707229e-05,
      "loss": 1.1434,
      "step": 5190
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.140625,
      "learning_rate": 6.85038467783314e-05,
      "loss": 0.9752,
      "step": 5195
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1572265625,
      "learning_rate": 6.810995623341546e-05,
      "loss": 0.9277,
      "step": 5200
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1494140625,
      "learning_rate": 6.77170227084419e-05,
      "loss": 1.3186,
      "step": 5205
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.142578125,
      "learning_rate": 6.732504827084255e-05,
      "loss": 1.0487,
      "step": 5210
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.15234375,
      "learning_rate": 6.693403498300249e-05,
      "loss": 0.9536,
      "step": 5215
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1689453125,
      "learning_rate": 6.654398490225008e-05,
      "loss": 1.0277,
      "step": 5220
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1396484375,
      "learning_rate": 6.61549000808454e-05,
      "loss": 1.0484,
      "step": 5225
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1650390625,
      "learning_rate": 6.576678256597016e-05,
      "loss": 1.1071,
      "step": 5230
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.19140625,
      "learning_rate": 6.537963439971612e-05,
      "loss": 1.1738,
      "step": 5235
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.162109375,
      "learning_rate": 6.499345761907521e-05,
      "loss": 1.1875,
      "step": 5240
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2197265625,
      "learning_rate": 6.460825425592807e-05,
      "loss": 1.3482,
      "step": 5245
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.134765625,
      "learning_rate": 6.422402633703389e-05,
      "loss": 0.9738,
      "step": 5250
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1572265625,
      "learning_rate": 6.384077588401932e-05,
      "loss": 1.2898,
      "step": 5255
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.140625,
      "learning_rate": 6.345850491336826e-05,
      "loss": 0.9434,
      "step": 5260
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1787109375,
      "learning_rate": 6.307721543641088e-05,
      "loss": 1.2934,
      "step": 5265
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1337890625,
      "learning_rate": 6.269690945931328e-05,
      "loss": 1.0275,
      "step": 5270
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.15234375,
      "learning_rate": 6.23175889830668e-05,
      "loss": 0.928,
      "step": 5275
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1220703125,
      "learning_rate": 6.19392560034775e-05,
      "loss": 0.6696,
      "step": 5280
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.177734375,
      "learning_rate": 6.156191251115575e-05,
      "loss": 1.0608,
      "step": 5285
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.21875,
      "learning_rate": 6.11855604915058e-05,
      "loss": 1.1805,
      "step": 5290
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.154296875,
      "learning_rate": 6.081020192471509e-05,
      "loss": 0.9473,
      "step": 5295
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1650390625,
      "learning_rate": 6.0435838785743905e-05,
      "loss": 0.8069,
      "step": 5300
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1943359375,
      "learning_rate": 6.006247304431528e-05,
      "loss": 0.8966,
      "step": 5305
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.169921875,
      "learning_rate": 5.9690106664904346e-05,
      "loss": 0.9691,
      "step": 5310
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1591796875,
      "learning_rate": 5.931874160672798e-05,
      "loss": 1.1929,
      "step": 5315
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1455078125,
      "learning_rate": 5.894837982373455e-05,
      "loss": 1.0995,
      "step": 5320
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1591796875,
      "learning_rate": 5.857902326459377e-05,
      "loss": 0.9066,
      "step": 5325
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.15234375,
      "learning_rate": 5.8210673872686385e-05,
      "loss": 1.0044,
      "step": 5330
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1669921875,
      "learning_rate": 5.7843333586093754e-05,
      "loss": 1.0552,
      "step": 5335
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.17578125,
      "learning_rate": 5.747700433758776e-05,
      "loss": 0.9869,
      "step": 5340
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1640625,
      "learning_rate": 5.7111688054620945e-05,
      "loss": 1.1215,
      "step": 5345
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.134765625,
      "learning_rate": 5.6747386659315755e-05,
      "loss": 1.0677,
      "step": 5350
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1689453125,
      "learning_rate": 5.638410206845512e-05,
      "loss": 1.0609,
      "step": 5355
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.134765625,
      "learning_rate": 5.602183619347179e-05,
      "loss": 1.1,
      "step": 5360
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1591796875,
      "learning_rate": 5.566059094043849e-05,
      "loss": 1.0852,
      "step": 5365
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.158203125,
      "learning_rate": 5.5300368210058075e-05,
      "loss": 1.2309,
      "step": 5370
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1640625,
      "learning_rate": 5.4941169897653354e-05,
      "loss": 1.0994,
      "step": 5375
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1865234375,
      "learning_rate": 5.458299789315699e-05,
      "loss": 1.2385,
      "step": 5380
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.146484375,
      "learning_rate": 5.4225854081101697e-05,
      "loss": 0.9504,
      "step": 5385
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.16796875,
      "learning_rate": 5.3869740340610515e-05,
      "loss": 1.1537,
      "step": 5390
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1533203125,
      "learning_rate": 5.351465854538659e-05,
      "loss": 1.3266,
      "step": 5395
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1533203125,
      "learning_rate": 5.31606105637035e-05,
      "loss": 0.9765,
      "step": 5400
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.142578125,
      "learning_rate": 5.2807598258395234e-05,
      "loss": 1.0391,
      "step": 5405
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1865234375,
      "learning_rate": 5.2455623486846765e-05,
      "loss": 1.2316,
      "step": 5410
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.171875,
      "learning_rate": 5.210468810098398e-05,
      "loss": 0.9606,
      "step": 5415
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.171875,
      "learning_rate": 5.1754793947263984e-05,
      "loss": 1.2529,
      "step": 5420
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1416015625,
      "learning_rate": 5.140594286666528e-05,
      "loss": 1.1235,
      "step": 5425
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1650390625,
      "learning_rate": 5.105813669467843e-05,
      "loss": 1.0962,
      "step": 5430
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1708984375,
      "learning_rate": 5.0711377261296126e-05,
      "loss": 0.9264,
      "step": 5435
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.224609375,
      "learning_rate": 5.036566639100351e-05,
      "loss": 1.2677,
      "step": 5440
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15625,
      "learning_rate": 5.0021005902768726e-05,
      "loss": 0.9795,
      "step": 5445
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.17578125,
      "learning_rate": 4.96773976100332e-05,
      "loss": 0.9653,
      "step": 5450
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1708984375,
      "learning_rate": 4.933484332070257e-05,
      "loss": 1.2085,
      "step": 5455
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1845703125,
      "learning_rate": 4.899334483713641e-05,
      "loss": 1.3929,
      "step": 5460
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1337890625,
      "learning_rate": 4.865290395613928e-05,
      "loss": 0.9019,
      "step": 5465
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1337890625,
      "learning_rate": 4.831352246895104e-05,
      "loss": 0.9831,
      "step": 5470
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1953125,
      "learning_rate": 4.797520216123769e-05,
      "loss": 1.0529,
      "step": 5475
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.169921875,
      "learning_rate": 4.763794481308176e-05,
      "loss": 1.1049,
      "step": 5480
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11865234375,
      "learning_rate": 4.730175219897292e-05,
      "loss": 0.9096,
      "step": 5485
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1484375,
      "learning_rate": 4.6966626087798614e-05,
      "loss": 1.3048,
      "step": 5490
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1533203125,
      "learning_rate": 4.66325682428351e-05,
      "loss": 1.1603,
      "step": 5495
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.17578125,
      "learning_rate": 4.629958042173779e-05,
      "loss": 0.9917,
      "step": 5500
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1318359375,
      "learning_rate": 4.5967664376532076e-05,
      "loss": 0.9204,
      "step": 5505
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1455078125,
      "learning_rate": 4.563682185360421e-05,
      "loss": 0.9311,
      "step": 5510
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.138671875,
      "learning_rate": 4.530705459369211e-05,
      "loss": 0.9714,
      "step": 5515
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1982421875,
      "learning_rate": 4.497836433187619e-05,
      "loss": 1.1269,
      "step": 5520
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.162109375,
      "learning_rate": 4.4650752797570165e-05,
      "loss": 1.2484,
      "step": 5525
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14453125,
      "learning_rate": 4.432422171451187e-05,
      "loss": 0.9576,
      "step": 5530
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1376953125,
      "learning_rate": 4.399877280075448e-05,
      "loss": 0.9191,
      "step": 5535
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1591796875,
      "learning_rate": 4.3674407768657384e-05,
      "loss": 1.2496,
      "step": 5540
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1650390625,
      "learning_rate": 4.335112832487684e-05,
      "loss": 0.9345,
      "step": 5545
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.146484375,
      "learning_rate": 4.302893617035741e-05,
      "loss": 0.9389,
      "step": 5550
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1669921875,
      "learning_rate": 4.270783300032269e-05,
      "loss": 1.2902,
      "step": 5555
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1650390625,
      "learning_rate": 4.23878205042669e-05,
      "loss": 0.9465,
      "step": 5560
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1484375,
      "learning_rate": 4.206890036594535e-05,
      "loss": 0.9596,
      "step": 5565
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.154296875,
      "learning_rate": 4.175107426336591e-05,
      "loss": 1.0553,
      "step": 5570
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1923828125,
      "learning_rate": 4.1434343868780086e-05,
      "loss": 1.1287,
      "step": 5575
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1572265625,
      "learning_rate": 4.111871084867461e-05,
      "loss": 1.2026,
      "step": 5580
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.14453125,
      "learning_rate": 4.080417686376203e-05,
      "loss": 1.0681,
      "step": 5585
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.140625,
      "learning_rate": 4.0490743568972366e-05,
      "loss": 1.1642,
      "step": 5590
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.130859375,
      "learning_rate": 4.017841261344427e-05,
      "loss": 1.1061,
      "step": 5595
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.14453125,
      "learning_rate": 3.986718564051653e-05,
      "loss": 0.8276,
      "step": 5600
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.146484375,
      "learning_rate": 3.9557064287719294e-05,
      "loss": 0.9965,
      "step": 5605
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.134765625,
      "learning_rate": 3.924805018676536e-05,
      "loss": 1.0255,
      "step": 5610
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1611328125,
      "learning_rate": 3.894014496354165e-05,
      "loss": 1.0776,
      "step": 5615
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.142578125,
      "learning_rate": 3.863335023810083e-05,
      "loss": 0.9152,
      "step": 5620
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1689453125,
      "learning_rate": 3.8327667624652645e-05,
      "loss": 1.1285,
      "step": 5625
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1484375,
      "learning_rate": 3.802309873155529e-05,
      "loss": 1.2754,
      "step": 5630
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1591796875,
      "learning_rate": 3.771964516130716e-05,
      "loss": 1.0922,
      "step": 5635
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.162109375,
      "learning_rate": 3.741730851053834e-05,
      "loss": 1.0333,
      "step": 5640
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1728515625,
      "learning_rate": 3.7116090370002356e-05,
      "loss": 1.1931,
      "step": 5645
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1630859375,
      "learning_rate": 3.681599232456742e-05,
      "loss": 1.2224,
      "step": 5650
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1787109375,
      "learning_rate": 3.651701595320839e-05,
      "loss": 0.9946,
      "step": 5655
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1142578125,
      "learning_rate": 3.621916282899856e-05,
      "loss": 0.9051,
      "step": 5660
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1669921875,
      "learning_rate": 3.592243451910102e-05,
      "loss": 1.0493,
      "step": 5665
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.16015625,
      "learning_rate": 3.5626832584760766e-05,
      "loss": 1.0392,
      "step": 5670
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1708984375,
      "learning_rate": 3.5332358581296214e-05,
      "loss": 1.0141,
      "step": 5675
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.154296875,
      "learning_rate": 3.503901405809129e-05,
      "loss": 1.1582,
      "step": 5680
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.158203125,
      "learning_rate": 3.474680055858695e-05,
      "loss": 1.1519,
      "step": 5685
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15625,
      "learning_rate": 3.4455719620273386e-05,
      "loss": 1.164,
      "step": 5690
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1591796875,
      "learning_rate": 3.416577277468172e-05,
      "loss": 1.0292,
      "step": 5695
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.162109375,
      "learning_rate": 3.387696154737591e-05,
      "loss": 1.2348,
      "step": 5700
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.138671875,
      "learning_rate": 3.3589287457945014e-05,
      "loss": 1.0566,
      "step": 5705
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1787109375,
      "learning_rate": 3.330275201999494e-05,
      "loss": 1.1678,
      "step": 5710
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.142578125,
      "learning_rate": 3.3017356741140506e-05,
      "loss": 0.9696,
      "step": 5715
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1572265625,
      "learning_rate": 3.2733103122997486e-05,
      "loss": 1.0827,
      "step": 5720
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15234375,
      "learning_rate": 3.2449992661174905e-05,
      "loss": 0.9318,
      "step": 5725
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.154296875,
      "learning_rate": 3.216802684526704e-05,
      "loss": 1.1558,
      "step": 5730
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1474609375,
      "learning_rate": 3.1887207158845465e-05,
      "loss": 1.2175,
      "step": 5735
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.189453125,
      "learning_rate": 3.160753507945138e-05,
      "loss": 1.1071,
      "step": 5740
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1455078125,
      "learning_rate": 3.132901207858796e-05,
      "loss": 1.0666,
      "step": 5745
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.171875,
      "learning_rate": 3.10516396217122e-05,
      "loss": 1.2645,
      "step": 5750
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15625,
      "learning_rate": 3.077541916822782e-05,
      "loss": 0.9798,
      "step": 5755
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1494140625,
      "learning_rate": 3.0500352171476896e-05,
      "loss": 0.7795,
      "step": 5760
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.154296875,
      "learning_rate": 3.0226440078732846e-05,
      "loss": 0.9405,
      "step": 5765
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1669921875,
      "learning_rate": 2.9953684331192278e-05,
      "loss": 1.0061,
      "step": 5770
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1572265625,
      "learning_rate": 2.9682086363967852e-05,
      "loss": 1.1415,
      "step": 5775
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1396484375,
      "learning_rate": 2.9411647606080395e-05,
      "loss": 1.1644,
      "step": 5780
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.16015625,
      "learning_rate": 2.9142369480451652e-05,
      "loss": 1.0313,
      "step": 5785
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16015625,
      "learning_rate": 2.8874253403896487e-05,
      "loss": 1.2192,
      "step": 5790
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.138671875,
      "learning_rate": 2.8607300787115818e-05,
      "loss": 0.9965,
      "step": 5795
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1474609375,
      "learning_rate": 2.8341513034688766e-05,
      "loss": 1.0245,
      "step": 5800
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.142578125,
      "learning_rate": 2.807689154506571e-05,
      "loss": 1.036,
      "step": 5805
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.185546875,
      "learning_rate": 2.7813437710560498e-05,
      "loss": 1.1713,
      "step": 5810
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16796875,
      "learning_rate": 2.755115291734342e-05,
      "loss": 1.2636,
      "step": 5815
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2138671875,
      "learning_rate": 2.729003854543388e-05,
      "loss": 1.0071,
      "step": 5820
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1416015625,
      "learning_rate": 2.7030095968692937e-05,
      "loss": 0.9706,
      "step": 5825
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1416015625,
      "learning_rate": 2.6771326554816345e-05,
      "loss": 1.1242,
      "step": 5830
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1298828125,
      "learning_rate": 2.6513731665327085e-05,
      "loss": 0.9072,
      "step": 5835
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.138671875,
      "learning_rate": 2.6257312655568522e-05,
      "loss": 0.9881,
      "step": 5840
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1337890625,
      "learning_rate": 2.6002070874696915e-05,
      "loss": 1.1109,
      "step": 5845
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1318359375,
      "learning_rate": 2.5748007665674646e-05,
      "loss": 1.0871,
      "step": 5850
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1611328125,
      "learning_rate": 2.5495124365262885e-05,
      "loss": 1.0898,
      "step": 5855
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1494140625,
      "learning_rate": 2.5243422304014802e-05,
      "loss": 1.0403,
      "step": 5860
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1650390625,
      "learning_rate": 2.4992902806268258e-05,
      "loss": 1.0823,
      "step": 5865
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16015625,
      "learning_rate": 2.4743567190139287e-05,
      "loss": 0.9326,
      "step": 5870
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1396484375,
      "learning_rate": 2.44954167675146e-05,
      "loss": 1.0215,
      "step": 5875
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1474609375,
      "learning_rate": 2.42484528440452e-05,
      "loss": 0.9758,
      "step": 5880
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.154296875,
      "learning_rate": 2.4002676719139166e-05,
      "loss": 0.8542,
      "step": 5885
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1884765625,
      "learning_rate": 2.3758089685955025e-05,
      "loss": 1.1913,
      "step": 5890
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16015625,
      "learning_rate": 2.3514693031394806e-05,
      "loss": 0.7058,
      "step": 5895
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1484375,
      "learning_rate": 2.3272488036097273e-05,
      "loss": 0.9125,
      "step": 5900
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1513671875,
      "learning_rate": 2.3031475974431337e-05,
      "loss": 1.0825,
      "step": 5905
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1611328125,
      "learning_rate": 2.27916581144893e-05,
      "loss": 0.8948,
      "step": 5910
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1708984375,
      "learning_rate": 2.2553035718079955e-05,
      "loss": 1.2793,
      "step": 5915
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.166015625,
      "learning_rate": 2.2315610040722245e-05,
      "loss": 1.1075,
      "step": 5920
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.15625,
      "learning_rate": 2.2079382331638503e-05,
      "loss": 1.0355,
      "step": 5925
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.140625,
      "learning_rate": 2.1844353833748027e-05,
      "loss": 0.8773,
      "step": 5930
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.142578125,
      "learning_rate": 2.161052578366024e-05,
      "loss": 1.0663,
      "step": 5935
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1494140625,
      "learning_rate": 2.1377899411668483e-05,
      "loss": 1.1142,
      "step": 5940
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.169921875,
      "learning_rate": 2.1146475941743485e-05,
      "loss": 1.1089,
      "step": 5945
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2001953125,
      "learning_rate": 2.091625659152671e-05,
      "loss": 1.1264,
      "step": 5950
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.16015625,
      "learning_rate": 2.0687242572324268e-05,
      "loss": 1.1726,
      "step": 5955
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1455078125,
      "learning_rate": 2.0459435089100255e-05,
      "loss": 0.9742,
      "step": 5960
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.13671875,
      "learning_rate": 2.0232835340470683e-05,
      "loss": 0.9213,
      "step": 5965
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.13671875,
      "learning_rate": 2.0007444518696876e-05,
      "loss": 1.0536,
      "step": 5970
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1767578125,
      "learning_rate": 1.9783263809679512e-05,
      "loss": 1.0193,
      "step": 5975
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.15234375,
      "learning_rate": 1.9560294392952093e-05,
      "loss": 1.0072,
      "step": 5980
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.16015625,
      "learning_rate": 1.9338537441674853e-05,
      "loss": 1.0281,
      "step": 5985
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.185546875,
      "learning_rate": 1.9117994122628767e-05,
      "loss": 1.0693,
      "step": 5990
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1767578125,
      "learning_rate": 1.8898665596209175e-05,
      "loss": 0.9922,
      "step": 5995
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1806640625,
      "learning_rate": 1.8680553016419632e-05,
      "loss": 1.1259,
      "step": 6000
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1513671875,
      "learning_rate": 1.8463657530866068e-05,
      "loss": 1.0088,
      "step": 6005
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.146484375,
      "learning_rate": 1.8247980280750616e-05,
      "loss": 1.2548,
      "step": 6010
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.15234375,
      "learning_rate": 1.803352240086567e-05,
      "loss": 1.0571,
      "step": 6015
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.150390625,
      "learning_rate": 1.7820285019587795e-05,
      "loss": 1.2816,
      "step": 6020
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.18359375,
      "learning_rate": 1.7608269258871823e-05,
      "loss": 1.1032,
      "step": 6025
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1396484375,
      "learning_rate": 1.739747623424512e-05,
      "loss": 0.9287,
      "step": 6030
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1455078125,
      "learning_rate": 1.718790705480161e-05,
      "loss": 0.8832,
      "step": 6035
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.171875,
      "learning_rate": 1.6979562823195738e-05,
      "loss": 1.0844,
      "step": 6040
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1748046875,
      "learning_rate": 1.6772444635636986e-05,
      "loss": 0.9688,
      "step": 6045
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.126953125,
      "learning_rate": 1.656655358188386e-05,
      "loss": 1.0039,
      "step": 6050
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.6361890745238472e-05,
      "loss": 1.1248,
      "step": 6055
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2197265625,
      "learning_rate": 1.6158457202540472e-05,
      "loss": 0.9911,
      "step": 6060
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.138671875,
      "learning_rate": 1.595625402416151e-05,
      "loss": 1.1771,
      "step": 6065
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1416015625,
      "learning_rate": 1.5755282273999694e-05,
      "loss": 0.9097,
      "step": 6070
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.5555543009473953e-05,
      "loss": 1.3623,
      "step": 6075
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1845703125,
      "learning_rate": 1.5357037281518522e-05,
      "loss": 1.1559,
      "step": 6080
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.16796875,
      "learning_rate": 1.5159766134577174e-05,
      "loss": 1.0551,
      "step": 6085
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.4963730606598003e-05,
      "loss": 1.0872,
      "step": 6090
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.158203125,
      "learning_rate": 1.476893172902785e-05,
      "loss": 1.1479,
      "step": 6095
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1396484375,
      "learning_rate": 1.4575370526806936e-05,
      "loss": 0.9236,
      "step": 6100
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.134765625,
      "learning_rate": 1.4383048018363292e-05,
      "loss": 1.2065,
      "step": 6105
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.162109375,
      "learning_rate": 1.4191965215607561e-05,
      "loss": 0.9892,
      "step": 6110
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2080078125,
      "learning_rate": 1.4002123123927678e-05,
      "loss": 1.0428,
      "step": 6115
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14453125,
      "learning_rate": 1.3813522742183615e-05,
      "loss": 1.0641,
      "step": 6120
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.140625,
      "learning_rate": 1.3626165062701867e-05,
      "loss": 1.1613,
      "step": 6125
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14453125,
      "learning_rate": 1.3440051071270526e-05,
      "loss": 1.176,
      "step": 6130
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.146484375,
      "learning_rate": 1.3255181747133944e-05,
      "loss": 0.9431,
      "step": 6135
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.162109375,
      "learning_rate": 1.307155806298771e-05,
      "loss": 0.8613,
      "step": 6140
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.171875,
      "learning_rate": 1.2889180984973298e-05,
      "loss": 1.0397,
      "step": 6145
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1396484375,
      "learning_rate": 1.2708051472673204e-05,
      "loss": 1.1134,
      "step": 6150
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.150390625,
      "learning_rate": 1.2528170479105816e-05,
      "loss": 1.1562,
      "step": 6155
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.234953895072044e-05,
      "loss": 0.9906,
      "step": 6160
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1552734375,
      "learning_rate": 1.2172157827392228e-05,
      "loss": 0.9779,
      "step": 6165
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1396484375,
      "learning_rate": 1.1996028042417257e-05,
      "loss": 1.2537,
      "step": 6170
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1552734375,
      "learning_rate": 1.1821150522507652e-05,
      "loss": 0.9378,
      "step": 6175
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1318359375,
      "learning_rate": 1.1647526187786889e-05,
      "loss": 1.0642,
      "step": 6180
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.140625,
      "learning_rate": 1.1475155951784582e-05,
      "loss": 1.3321,
      "step": 6185
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1572265625,
      "learning_rate": 1.1304040721431874e-05,
      "loss": 1.1456,
      "step": 6190
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1875,
      "learning_rate": 1.1134181397056686e-05,
      "loss": 1.101,
      "step": 6195
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.154296875,
      "learning_rate": 1.0965578872379035e-05,
      "loss": 1.2256,
      "step": 6200
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1494140625,
      "learning_rate": 1.0798234034506166e-05,
      "loss": 0.918,
      "step": 6205
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2001953125,
      "learning_rate": 1.0632147763927985e-05,
      "loss": 1.3359,
      "step": 6210
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.15625,
      "learning_rate": 1.0467320934512359e-05,
      "loss": 0.8908,
      "step": 6215
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.0303754413500648e-05,
      "loss": 1.0354,
      "step": 6220
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.158203125,
      "learning_rate": 1.014144906150305e-05,
      "loss": 1.2781,
      "step": 6225
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1416015625,
      "learning_rate": 9.980405732494036e-06,
      "loss": 1.2613,
      "step": 6230
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.140625,
      "learning_rate": 9.82062527380792e-06,
      "loss": 1.1093,
      "step": 6235
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1611328125,
      "learning_rate": 9.662108526134417e-06,
      "loss": 1.0236,
      "step": 6240
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1416015625,
      "learning_rate": 9.504856323514165e-06,
      "loss": 0.9299,
      "step": 6245
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1455078125,
      "learning_rate": 9.348869493334384e-06,
      "loss": 0.8213,
      "step": 6250
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1416015625,
      "learning_rate": 9.194148856324413e-06,
      "loss": 1.0443,
      "step": 6255
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.142578125,
      "learning_rate": 9.040695226551599e-06,
      "loss": 0.9715,
      "step": 6260
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1455078125,
      "learning_rate": 8.888509411416861e-06,
      "loss": 0.9815,
      "step": 6265
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1708984375,
      "learning_rate": 8.73759221165038e-06,
      "loss": 1.045,
      "step": 6270
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1689453125,
      "learning_rate": 8.587944421307559e-06,
      "loss": 1.1522,
      "step": 6275
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1533203125,
      "learning_rate": 8.439566827764811e-06,
      "loss": 0.9543,
      "step": 6280
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1875,
      "learning_rate": 8.29246021171523e-06,
      "loss": 1.1236,
      "step": 6285
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.154296875,
      "learning_rate": 8.146625347164788e-06,
      "loss": 1.2259,
      "step": 6290
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.142578125,
      "learning_rate": 8.002063001427974e-06,
      "loss": 1.2055,
      "step": 6295
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1630859375,
      "learning_rate": 7.85877393512388e-06,
      "loss": 1.1735,
      "step": 6300
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1513671875,
      "learning_rate": 7.716758902172328e-06,
      "loss": 1.0167,
      "step": 6305
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1572265625,
      "learning_rate": 7.576018649789657e-06,
      "loss": 1.3354,
      "step": 6310
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1533203125,
      "learning_rate": 7.436553918484912e-06,
      "loss": 1.0759,
      "step": 6315
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1513671875,
      "learning_rate": 7.298365442056004e-06,
      "loss": 1.2152,
      "step": 6320
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1689453125,
      "learning_rate": 7.161453947585772e-06,
      "loss": 1.2261,
      "step": 6325
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14453125,
      "learning_rate": 7.025820155438151e-06,
      "loss": 1.2291,
      "step": 6330
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1357421875,
      "learning_rate": 6.891464779254452e-06,
      "loss": 1.0477,
      "step": 6335
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1767578125,
      "learning_rate": 6.758388525949483e-06,
      "loss": 0.9963,
      "step": 6340
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1689453125,
      "learning_rate": 6.626592095707984e-06,
      "loss": 1.0916,
      "step": 6345
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.169921875,
      "learning_rate": 6.4960761819808365e-06,
      "loss": 0.8691,
      "step": 6350
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1455078125,
      "learning_rate": 6.366841471481477e-06,
      "loss": 1.0464,
      "step": 6355
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1376953125,
      "learning_rate": 6.238888644182178e-06,
      "loss": 1.017,
      "step": 6360
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.169921875,
      "learning_rate": 6.112218373310635e-06,
      "loss": 0.9035,
      "step": 6365
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1650390625,
      "learning_rate": 5.9868313253462756e-06,
      "loss": 1.1066,
      "step": 6370
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1806640625,
      "learning_rate": 5.862728160016844e-06,
      "loss": 1.1089,
      "step": 6375
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12451171875,
      "learning_rate": 5.739909530294824e-06,
      "loss": 0.9805,
      "step": 6380
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1787109375,
      "learning_rate": 5.6183760823942385e-06,
      "loss": 0.9165,
      "step": 6385
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1640625,
      "learning_rate": 5.49812845576686e-06,
      "loss": 1.0781,
      "step": 6390
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1650390625,
      "learning_rate": 5.379167283099312e-06,
      "loss": 1.2878,
      "step": 6395
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.15625,
      "learning_rate": 5.261493190309302e-06,
      "loss": 0.7727,
      "step": 6400
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1748046875,
      "learning_rate": 5.145106796542676e-06,
      "loss": 1.3831,
      "step": 6405
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1640625,
      "learning_rate": 5.030008714169892e-06,
      "loss": 0.8447,
      "step": 6410
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.16015625,
      "learning_rate": 4.916199548783029e-06,
      "loss": 1.0932,
      "step": 6415
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1416015625,
      "learning_rate": 4.803679899192393e-06,
      "loss": 0.958,
      "step": 6420
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.171875,
      "learning_rate": 4.692450357423522e-06,
      "loss": 1.147,
      "step": 6425
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1630859375,
      "learning_rate": 4.582511508714027e-06,
      "loss": 1.072,
      "step": 6430
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.166015625,
      "learning_rate": 4.473863931510447e-06,
      "loss": 1.0882,
      "step": 6435
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1640625,
      "learning_rate": 4.366508197465285e-06,
      "loss": 1.2075,
      "step": 6440
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.16015625,
      "learning_rate": 4.260444871433927e-06,
      "loss": 0.979,
      "step": 6445
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.146484375,
      "learning_rate": 4.155674511471835e-06,
      "loss": 1.0348,
      "step": 6450
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.052197668831359e-06,
      "loss": 1.0764,
      "step": 6455
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1650390625,
      "learning_rate": 3.950014887959069e-06,
      "loss": 1.0324,
      "step": 6460
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.150390625,
      "learning_rate": 3.849126706492679e-06,
      "loss": 1.1,
      "step": 6465
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1396484375,
      "learning_rate": 3.7495336552584302e-06,
      "loss": 1.0531,
      "step": 6470
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12890625,
      "learning_rate": 3.651236258268159e-06,
      "loss": 0.9983,
      "step": 6475
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14453125,
      "learning_rate": 3.554235032716596e-06,
      "loss": 1.1773,
      "step": 6480
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.171875,
      "learning_rate": 3.4585304889785675e-06,
      "loss": 0.9983,
      "step": 6485
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.16796875,
      "learning_rate": 3.3641231306064666e-06,
      "loss": 1.2404,
      "step": 6490
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1572265625,
      "learning_rate": 3.2710134543273986e-06,
      "loss": 1.0835,
      "step": 6495
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1591796875,
      "learning_rate": 3.1792019500407885e-06,
      "loss": 0.9551,
      "step": 6500
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.150390625,
      "learning_rate": 3.088689100815556e-06,
      "loss": 1.1014,
      "step": 6505
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1484375,
      "learning_rate": 2.9994753828878617e-06,
      "loss": 1.1044,
      "step": 6510
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.146484375,
      "learning_rate": 2.911561265658308e-06,
      "loss": 1.0531,
      "step": 6515
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1357421875,
      "learning_rate": 2.8249472116896878e-06,
      "loss": 0.9011,
      "step": 6520
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1494140625,
      "learning_rate": 2.739633676704462e-06,
      "loss": 0.9713,
      "step": 6525
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.150390625,
      "learning_rate": 2.655621109582368e-06,
      "loss": 1.0419,
      "step": 6530
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1337890625,
      "learning_rate": 2.572909952358066e-06,
      "loss": 1.0271,
      "step": 6535
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1474609375,
      "learning_rate": 2.4915006402187467e-06,
      "loss": 1.0825,
      "step": 6540
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1611328125,
      "learning_rate": 2.4113936015020266e-06,
      "loss": 1.1312,
      "step": 6545
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1748046875,
      "learning_rate": 2.332589257693446e-06,
      "loss": 1.1144,
      "step": 6550
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1650390625,
      "learning_rate": 2.255088023424501e-06,
      "loss": 1.0819,
      "step": 6555
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.146484375,
      "learning_rate": 2.1788903064701717e-06,
      "loss": 1.1108,
      "step": 6560
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.140625,
      "learning_rate": 2.103996507747147e-06,
      "loss": 1.0079,
      "step": 6565
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.146484375,
      "learning_rate": 2.0304070213112978e-06,
      "loss": 0.8256,
      "step": 6570
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1689453125,
      "learning_rate": 1.9581222343560134e-06,
      "loss": 1.0926,
      "step": 6575
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1484375,
      "learning_rate": 1.8871425272098676e-06,
      "loss": 1.0518,
      "step": 6580
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.8174682733346493e-06,
      "loss": 1.0388,
      "step": 6585
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1845703125,
      "learning_rate": 1.7490998393236145e-06,
      "loss": 1.0957,
      "step": 6590
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14453125,
      "learning_rate": 1.6820375848993208e-06,
      "loss": 0.988,
      "step": 6595
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.166015625,
      "learning_rate": 1.6162818629118227e-06,
      "loss": 1.0691,
      "step": 6600
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.551833019336868e-06,
      "loss": 1.3649,
      "step": 6605
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2041015625,
      "learning_rate": 1.4886913932739278e-06,
      "loss": 1.1142,
      "step": 6610
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.15625,
      "learning_rate": 1.4268573169446408e-06,
      "loss": 1.1082,
      "step": 6615
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.16796875,
      "learning_rate": 1.3663311156908164e-06,
      "loss": 0.9187,
      "step": 6620
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.193359375,
      "learning_rate": 1.307113107972907e-06,
      "loss": 1.0999,
      "step": 6625
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1513671875,
      "learning_rate": 1.2492036053682043e-06,
      "loss": 0.9749,
      "step": 6630
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.244140625,
      "learning_rate": 1.1926029125693406e-06,
      "loss": 0.9945,
      "step": 6635
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.21484375,
      "learning_rate": 1.1373113273825675e-06,
      "loss": 1.1462,
      "step": 6640
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1572265625,
      "learning_rate": 1.0833291407262025e-06,
      "loss": 0.9899,
      "step": 6645
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1572265625,
      "learning_rate": 1.0306566366291292e-06,
      "loss": 0.9839,
      "step": 6650
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1669921875,
      "learning_rate": 9.792940922293825e-07,
      "loss": 1.1274,
      "step": 6655
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.173828125,
      "learning_rate": 9.292417777724827e-07,
      "loss": 1.2596,
      "step": 6660
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1201171875,
      "learning_rate": 8.804999566101868e-07,
      "loss": 0.9965,
      "step": 6665
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14453125,
      "learning_rate": 8.330688851990453e-07,
      "loss": 1.2265,
      "step": 6670
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.146484375,
      "learning_rate": 7.869488130991254e-07,
      "loss": 0.9328,
      "step": 6675
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.154296875,
      "learning_rate": 7.421399829725394e-07,
      "loss": 0.9776,
      "step": 6680
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1552734375,
      "learning_rate": 6.98642630582308e-07,
      "loss": 1.2955,
      "step": 6685
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1357421875,
      "learning_rate": 6.564569847910262e-07,
      "loss": 1.0571,
      "step": 6690
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.17578125,
      "learning_rate": 6.15583267559755e-07,
      "loss": 1.1025,
      "step": 6695
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.140625,
      "learning_rate": 5.760216939468266e-07,
      "loss": 1.13,
      "step": 6700
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1279296875,
      "learning_rate": 5.377724721066235e-07,
      "loss": 1.0371,
      "step": 6705
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1435546875,
      "learning_rate": 5.008358032885518e-07,
      "loss": 0.9811,
      "step": 6710
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1455078125,
      "learning_rate": 4.652118818360418e-07,
      "loss": 0.9162,
      "step": 6715
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18359375,
      "learning_rate": 4.309008951854099e-07,
      "loss": 1.0896,
      "step": 6720
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1552734375,
      "learning_rate": 3.9790302386499835e-07,
      "loss": 1.1249,
      "step": 6725
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.130859375,
      "learning_rate": 3.6621844149400954e-07,
      "loss": 0.928,
      "step": 6730
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1640625,
      "learning_rate": 3.3584731478189524e-07,
      "loss": 1.1065,
      "step": 6735
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.154296875,
      "learning_rate": 3.067898035272465e-07,
      "loss": 1.0509,
      "step": 6740
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.142578125,
      "learning_rate": 2.7904606061698865e-07,
      "loss": 0.8672,
      "step": 6745
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1611328125,
      "learning_rate": 2.5261623202571527e-07,
      "loss": 0.9601,
      "step": 6750
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1484375,
      "learning_rate": 2.275004568147443e-07,
      "loss": 1.0955,
      "step": 6755
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2177734375,
      "learning_rate": 2.036988671314799e-07,
      "loss": 1.1564,
      "step": 6760
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1513671875,
      "learning_rate": 1.8121158820869066e-07,
      "loss": 1.0113,
      "step": 6765
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.15625,
      "learning_rate": 1.6003873836389904e-07,
      "loss": 0.9622,
      "step": 6770
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1181640625,
      "learning_rate": 1.4018042899868744e-07,
      "loss": 1.1171,
      "step": 6775
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.17578125,
      "learning_rate": 1.216367645981431e-07,
      "loss": 1.1264,
      "step": 6780
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1484375,
      "learning_rate": 1.0440784273030302e-07,
      "loss": 0.8995,
      "step": 6785
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1552734375,
      "learning_rate": 8.84937540456543e-08,
      "loss": 0.9804,
      "step": 6790
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1435546875,
      "learning_rate": 7.389458227666234e-08,
      "loss": 1.2118,
      "step": 6795
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.146484375,
      "learning_rate": 6.061040423735453e-08,
      "loss": 1.2899,
      "step": 6800
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1689453125,
      "learning_rate": 4.864128982276506e-08,
      "loss": 1.1498,
      "step": 6805
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1279296875,
      "learning_rate": 3.798730200876843e-08,
      "loss": 1.0363,
      "step": 6810
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1513671875,
      "learning_rate": 2.8648496851663152e-08,
      "loss": 1.0867,
      "step": 6815
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.13671875,
      "learning_rate": 2.062492348783862e-08,
      "loss": 0.9538,
      "step": 6820
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1591796875,
      "learning_rate": 1.3916624133580857e-08,
      "loss": 1.1978,
      "step": 6825
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.140625,
      "learning_rate": 8.523634084794951e-09,
      "loss": 0.9851,
      "step": 6830
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1455078125,
      "learning_rate": 4.445981716866276e-09,
      "loss": 1.0124,
      "step": 6835
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.15625,
      "learning_rate": 1.6836884844662059e-09,
      "loss": 1.0845,
      "step": 6840
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1513671875,
      "learning_rate": 2.3676892152435516e-10,
      "loss": 1.007,
      "step": 6845
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.0585945844650269,
      "eval_runtime": 12437.9638,
      "eval_samples_per_second": 0.11,
      "eval_steps_per_second": 0.11,
      "step": 6848
    }
  ],
  "logging_steps": 5,
  "max_steps": 6848,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 5.565477044449444e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}