{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 2860,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006993006993006993,
      "grad_norm": 1.6810442141323327,
      "learning_rate": 3.4965034965034967e-08,
      "loss": 0.8521,
      "step": 1
    },
    {
      "epoch": 0.0034965034965034965,
      "grad_norm": 2.0975461172037795,
      "learning_rate": 1.7482517482517484e-07,
      "loss": 0.8882,
      "step": 5
    },
    {
      "epoch": 0.006993006993006993,
      "grad_norm": 1.6789007038251196,
      "learning_rate": 3.496503496503497e-07,
      "loss": 0.8318,
      "step": 10
    },
    {
      "epoch": 0.01048951048951049,
      "grad_norm": 1.4821851482849189,
      "learning_rate": 5.244755244755246e-07,
      "loss": 0.8318,
      "step": 15
    },
    {
      "epoch": 0.013986013986013986,
      "grad_norm": 0.6234336285116338,
      "learning_rate": 6.993006993006994e-07,
      "loss": 0.7664,
      "step": 20
    },
    {
      "epoch": 0.017482517482517484,
      "grad_norm": 0.46899145436944106,
      "learning_rate": 8.741258741258741e-07,
      "loss": 0.7122,
      "step": 25
    },
    {
      "epoch": 0.02097902097902098,
      "grad_norm": 0.392390148362129,
      "learning_rate": 1.0489510489510491e-06,
      "loss": 0.6534,
      "step": 30
    },
    {
      "epoch": 0.024475524475524476,
      "grad_norm": 0.3350247927290883,
      "learning_rate": 1.2237762237762238e-06,
      "loss": 0.5963,
      "step": 35
    },
    {
      "epoch": 0.027972027972027972,
      "grad_norm": 0.26894335890507826,
      "learning_rate": 1.3986013986013987e-06,
      "loss": 0.5489,
      "step": 40
    },
    {
      "epoch": 0.03146853146853147,
      "grad_norm": 0.2238957490967497,
      "learning_rate": 1.5734265734265736e-06,
      "loss": 0.5003,
      "step": 45
    },
    {
      "epoch": 0.03496503496503497,
      "grad_norm": 0.24251763091157577,
      "learning_rate": 1.7482517482517483e-06,
      "loss": 0.4794,
      "step": 50
    },
    {
      "epoch": 0.038461538461538464,
      "grad_norm": 0.2098939190580422,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 0.4518,
      "step": 55
    },
    {
      "epoch": 0.04195804195804196,
      "grad_norm": 0.17402599508473188,
      "learning_rate": 2.0979020979020983e-06,
      "loss": 0.4491,
      "step": 60
    },
    {
      "epoch": 0.045454545454545456,
      "grad_norm": 0.20318583220999484,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.4666,
      "step": 65
    },
    {
      "epoch": 0.04895104895104895,
      "grad_norm": 0.18554747229883345,
      "learning_rate": 2.4475524475524477e-06,
      "loss": 0.4358,
      "step": 70
    },
    {
      "epoch": 0.05244755244755245,
      "grad_norm": 0.2003347020079377,
      "learning_rate": 2.6223776223776225e-06,
      "loss": 0.4377,
      "step": 75
    },
    {
      "epoch": 0.055944055944055944,
      "grad_norm": 0.17532069480625367,
      "learning_rate": 2.7972027972027974e-06,
      "loss": 0.4144,
      "step": 80
    },
    {
      "epoch": 0.05944055944055944,
      "grad_norm": 0.16132669774185596,
      "learning_rate": 2.972027972027972e-06,
      "loss": 0.4361,
      "step": 85
    },
    {
      "epoch": 0.06293706293706294,
      "grad_norm": 0.18851634778688606,
      "learning_rate": 3.1468531468531472e-06,
      "loss": 0.401,
      "step": 90
    },
    {
      "epoch": 0.06643356643356643,
      "grad_norm": 0.18623129302896582,
      "learning_rate": 3.321678321678322e-06,
      "loss": 0.4045,
      "step": 95
    },
    {
      "epoch": 0.06993006993006994,
      "grad_norm": 0.13440650170364746,
      "learning_rate": 3.4965034965034966e-06,
      "loss": 0.4201,
      "step": 100
    },
    {
      "epoch": 0.07342657342657342,
      "grad_norm": 0.1705089760016885,
      "learning_rate": 3.6713286713286715e-06,
      "loss": 0.3972,
      "step": 105
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 0.1495843537552479,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.4069,
      "step": 110
    },
    {
      "epoch": 0.08041958041958042,
      "grad_norm": 0.14156867249528673,
      "learning_rate": 4.020979020979021e-06,
      "loss": 0.3756,
      "step": 115
    },
    {
      "epoch": 0.08391608391608392,
      "grad_norm": 0.13298597873579593,
      "learning_rate": 4.195804195804197e-06,
      "loss": 0.367,
      "step": 120
    },
    {
      "epoch": 0.08741258741258741,
      "grad_norm": 0.1652443143555161,
      "learning_rate": 4.3706293706293715e-06,
      "loss": 0.3936,
      "step": 125
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 0.14581515723948085,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.3934,
      "step": 130
    },
    {
      "epoch": 0.0944055944055944,
      "grad_norm": 0.16613762962785997,
      "learning_rate": 4.72027972027972e-06,
      "loss": 0.4229,
      "step": 135
    },
    {
      "epoch": 0.0979020979020979,
      "grad_norm": 0.16802882525770493,
      "learning_rate": 4.895104895104895e-06,
      "loss": 0.3842,
      "step": 140
    },
    {
      "epoch": 0.10139860139860139,
      "grad_norm": 0.17368450070810754,
      "learning_rate": 5.06993006993007e-06,
      "loss": 0.3955,
      "step": 145
    },
    {
      "epoch": 0.1048951048951049,
      "grad_norm": 0.17080636635901125,
      "learning_rate": 5.244755244755245e-06,
      "loss": 0.3907,
      "step": 150
    },
    {
      "epoch": 0.10839160839160839,
      "grad_norm": 0.14498613212072356,
      "learning_rate": 5.41958041958042e-06,
      "loss": 0.3734,
      "step": 155
    },
    {
      "epoch": 0.11188811188811189,
      "grad_norm": 0.13432770851676462,
      "learning_rate": 5.594405594405595e-06,
      "loss": 0.4205,
      "step": 160
    },
    {
      "epoch": 0.11538461538461539,
      "grad_norm": 0.13228476949701554,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.3881,
      "step": 165
    },
    {
      "epoch": 0.11888111888111888,
      "grad_norm": 0.19044578741413698,
      "learning_rate": 5.944055944055944e-06,
      "loss": 0.3622,
      "step": 170
    },
    {
      "epoch": 0.12237762237762238,
      "grad_norm": 0.13916568818927608,
      "learning_rate": 6.1188811188811196e-06,
      "loss": 0.3649,
      "step": 175
    },
    {
      "epoch": 0.1258741258741259,
      "grad_norm": 0.15063951671488746,
      "learning_rate": 6.2937062937062944e-06,
      "loss": 0.3821,
      "step": 180
    },
    {
      "epoch": 0.12937062937062938,
      "grad_norm": 0.12392104532299851,
      "learning_rate": 6.468531468531469e-06,
      "loss": 0.3824,
      "step": 185
    },
    {
      "epoch": 0.13286713286713286,
      "grad_norm": 0.11586428398855769,
      "learning_rate": 6.643356643356644e-06,
      "loss": 0.3478,
      "step": 190
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 0.14549736363326018,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.3824,
      "step": 195
    },
    {
      "epoch": 0.13986013986013987,
      "grad_norm": 0.158552396567501,
      "learning_rate": 6.993006993006993e-06,
      "loss": 0.3841,
      "step": 200
    },
    {
      "epoch": 0.14335664335664336,
      "grad_norm": 0.1524916091650796,
      "learning_rate": 7.167832167832168e-06,
      "loss": 0.371,
      "step": 205
    },
    {
      "epoch": 0.14685314685314685,
      "grad_norm": 0.12972459500546188,
      "learning_rate": 7.342657342657343e-06,
      "loss": 0.384,
      "step": 210
    },
    {
      "epoch": 0.15034965034965034,
      "grad_norm": 0.11422883593792324,
      "learning_rate": 7.517482517482519e-06,
      "loss": 0.3884,
      "step": 215
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 0.12707787563469491,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.3771,
      "step": 220
    },
    {
      "epoch": 0.15734265734265734,
      "grad_norm": 0.1544216700654582,
      "learning_rate": 7.867132867132867e-06,
      "loss": 0.391,
      "step": 225
    },
    {
      "epoch": 0.16083916083916083,
      "grad_norm": 0.14454155746316652,
      "learning_rate": 8.041958041958042e-06,
      "loss": 0.3765,
      "step": 230
    },
    {
      "epoch": 0.16433566433566432,
      "grad_norm": 0.13757029987017547,
      "learning_rate": 8.216783216783217e-06,
      "loss": 0.3824,
      "step": 235
    },
    {
      "epoch": 0.16783216783216784,
      "grad_norm": 0.13492667765968142,
      "learning_rate": 8.391608391608393e-06,
      "loss": 0.3797,
      "step": 240
    },
    {
      "epoch": 0.17132867132867133,
      "grad_norm": 0.13092685800146633,
      "learning_rate": 8.566433566433568e-06,
      "loss": 0.3776,
      "step": 245
    },
    {
      "epoch": 0.17482517482517482,
      "grad_norm": 0.13643694605172166,
      "learning_rate": 8.741258741258743e-06,
      "loss": 0.3829,
      "step": 250
    },
    {
      "epoch": 0.17832167832167833,
      "grad_norm": 0.13101604267139608,
      "learning_rate": 8.916083916083916e-06,
      "loss": 0.4047,
      "step": 255
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.11202626470308893,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.3496,
      "step": 260
    },
    {
      "epoch": 0.1853146853146853,
      "grad_norm": 0.1327150355799933,
      "learning_rate": 9.265734265734266e-06,
      "loss": 0.359,
      "step": 265
    },
    {
      "epoch": 0.1888111888111888,
      "grad_norm": 0.11308406562269005,
      "learning_rate": 9.44055944055944e-06,
      "loss": 0.372,
      "step": 270
    },
    {
      "epoch": 0.19230769230769232,
      "grad_norm": 0.12380613230904368,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.3868,
      "step": 275
    },
    {
      "epoch": 0.1958041958041958,
      "grad_norm": 0.1308138619329397,
      "learning_rate": 9.79020979020979e-06,
      "loss": 0.3949,
      "step": 280
    },
    {
      "epoch": 0.1993006993006993,
      "grad_norm": 0.12257237219274138,
      "learning_rate": 9.965034965034966e-06,
      "loss": 0.3754,
      "step": 285
    },
    {
      "epoch": 0.20279720279720279,
      "grad_norm": 0.12394995699972901,
      "learning_rate": 9.999940414335223e-06,
      "loss": 0.3658,
      "step": 290
    },
    {
      "epoch": 0.2062937062937063,
      "grad_norm": 0.12572743531560926,
      "learning_rate": 9.999698350006063e-06,
      "loss": 0.3623,
      "step": 295
    },
    {
      "epoch": 0.2097902097902098,
      "grad_norm": 0.12952652077555094,
      "learning_rate": 9.999270091916259e-06,
      "loss": 0.3714,
      "step": 300
    },
    {
      "epoch": 0.21328671328671328,
      "grad_norm": 0.11751387408871522,
      "learning_rate": 9.998655656014563e-06,
      "loss": 0.3871,
      "step": 305
    },
    {
      "epoch": 0.21678321678321677,
      "grad_norm": 0.12077462902955817,
      "learning_rate": 9.997855065183185e-06,
      "loss": 0.3592,
      "step": 310
    },
    {
      "epoch": 0.2202797202797203,
      "grad_norm": 0.12083600564766793,
      "learning_rate": 9.996868349236927e-06,
      "loss": 0.3502,
      "step": 315
    },
    {
      "epoch": 0.22377622377622378,
      "grad_norm": 0.11120732199233092,
      "learning_rate": 9.995695544922076e-06,
      "loss": 0.3645,
      "step": 320
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.1119589355975915,
      "learning_rate": 9.994336695915041e-06,
      "loss": 0.3678,
      "step": 325
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 0.11640129488127134,
      "learning_rate": 9.992791852820709e-06,
      "loss": 0.3681,
      "step": 330
    },
    {
      "epoch": 0.23426573426573427,
      "grad_norm": 0.09907937484741136,
      "learning_rate": 9.991061073170585e-06,
      "loss": 0.3579,
      "step": 335
    },
    {
      "epoch": 0.23776223776223776,
      "grad_norm": 0.11049603729407835,
      "learning_rate": 9.98914442142063e-06,
      "loss": 0.349,
      "step": 340
    },
    {
      "epoch": 0.24125874125874125,
      "grad_norm": 0.10831326766450866,
      "learning_rate": 9.98704196894887e-06,
      "loss": 0.3768,
      "step": 345
    },
    {
      "epoch": 0.24475524475524477,
      "grad_norm": 0.11312368278029544,
      "learning_rate": 9.984753794052735e-06,
      "loss": 0.3448,
      "step": 350
    },
    {
      "epoch": 0.24825174825174826,
      "grad_norm": 0.12603341829155673,
      "learning_rate": 9.982279981946143e-06,
      "loss": 0.3576,
      "step": 355
    },
    {
      "epoch": 0.2517482517482518,
      "grad_norm": 0.11811477539006092,
      "learning_rate": 9.97962062475633e-06,
      "loss": 0.3782,
      "step": 360
    },
    {
      "epoch": 0.25524475524475526,
      "grad_norm": 0.13917946371117146,
      "learning_rate": 9.976775821520412e-06,
      "loss": 0.3655,
      "step": 365
    },
    {
      "epoch": 0.25874125874125875,
      "grad_norm": 0.12081536318303664,
      "learning_rate": 9.973745678181705e-06,
      "loss": 0.3393,
      "step": 370
    },
    {
      "epoch": 0.26223776223776224,
      "grad_norm": 0.11939046824110922,
      "learning_rate": 9.970530307585774e-06,
      "loss": 0.3504,
      "step": 375
    },
    {
      "epoch": 0.26573426573426573,
      "grad_norm": 0.11612193380529051,
      "learning_rate": 9.967129829476227e-06,
      "loss": 0.3931,
      "step": 380
    },
    {
      "epoch": 0.2692307692307692,
      "grad_norm": 0.10939105627812593,
      "learning_rate": 9.96354437049027e-06,
      "loss": 0.3545,
      "step": 385
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.11397605180188672,
      "learning_rate": 9.959774064153977e-06,
      "loss": 0.3708,
      "step": 390
    },
    {
      "epoch": 0.2762237762237762,
      "grad_norm": 0.09725623960075111,
      "learning_rate": 9.955819050877321e-06,
      "loss": 0.3502,
      "step": 395
    },
    {
      "epoch": 0.27972027972027974,
      "grad_norm": 0.10429437499055316,
      "learning_rate": 9.951679477948946e-06,
      "loss": 0.3785,
      "step": 400
    },
    {
      "epoch": 0.28321678321678323,
      "grad_norm": 0.10450643154921269,
      "learning_rate": 9.947355499530685e-06,
      "loss": 0.3532,
      "step": 405
    },
    {
      "epoch": 0.2867132867132867,
      "grad_norm": 0.11758590783504208,
      "learning_rate": 9.942847276651812e-06,
      "loss": 0.3223,
      "step": 410
    },
    {
      "epoch": 0.2902097902097902,
      "grad_norm": 0.0929966557389189,
      "learning_rate": 9.93815497720305e-06,
      "loss": 0.3407,
      "step": 415
    },
    {
      "epoch": 0.2937062937062937,
      "grad_norm": 0.11254328479901264,
      "learning_rate": 9.933278775930317e-06,
      "loss": 0.349,
      "step": 420
    },
    {
      "epoch": 0.2972027972027972,
      "grad_norm": 0.11007322345211504,
      "learning_rate": 9.928218854428223e-06,
      "loss": 0.355,
      "step": 425
    },
    {
      "epoch": 0.3006993006993007,
      "grad_norm": 0.11509136513939573,
      "learning_rate": 9.922975401133292e-06,
      "loss": 0.3614,
      "step": 430
    },
    {
      "epoch": 0.3041958041958042,
      "grad_norm": 0.10855710661153205,
      "learning_rate": 9.917548611316969e-06,
      "loss": 0.3193,
      "step": 435
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.1003867889772061,
      "learning_rate": 9.911938687078324e-06,
      "loss": 0.3482,
      "step": 440
    },
    {
      "epoch": 0.3111888111888112,
      "grad_norm": 0.10618517266428626,
      "learning_rate": 9.90614583733654e-06,
      "loss": 0.3421,
      "step": 445
    },
    {
      "epoch": 0.3146853146853147,
      "grad_norm": 0.10291042677780514,
      "learning_rate": 9.900170277823129e-06,
      "loss": 0.3426,
      "step": 450
    },
    {
      "epoch": 0.3181818181818182,
      "grad_norm": 0.1024452151160681,
      "learning_rate": 9.894012231073895e-06,
      "loss": 0.3547,
      "step": 455
    },
    {
      "epoch": 0.32167832167832167,
      "grad_norm": 0.11209268298798039,
      "learning_rate": 9.887671926420649e-06,
      "loss": 0.3558,
      "step": 460
    },
    {
      "epoch": 0.32517482517482516,
      "grad_norm": 0.11881577035959273,
      "learning_rate": 9.881149599982671e-06,
      "loss": 0.3673,
      "step": 465
    },
    {
      "epoch": 0.32867132867132864,
      "grad_norm": 0.0976707455150946,
      "learning_rate": 9.874445494657912e-06,
      "loss": 0.3306,
      "step": 470
    },
    {
      "epoch": 0.3321678321678322,
      "grad_norm": 0.12961461872022373,
      "learning_rate": 9.86755986011395e-06,
      "loss": 0.3529,
      "step": 475
    },
    {
      "epoch": 0.3356643356643357,
      "grad_norm": 0.1045711130341524,
      "learning_rate": 9.860492952778695e-06,
      "loss": 0.337,
      "step": 480
    },
    {
      "epoch": 0.33916083916083917,
      "grad_norm": 0.11459102139170876,
      "learning_rate": 9.853245035830834e-06,
      "loss": 0.3577,
      "step": 485
    },
    {
      "epoch": 0.34265734265734266,
      "grad_norm": 0.09410171637748112,
      "learning_rate": 9.845816379190037e-06,
      "loss": 0.3412,
      "step": 490
    },
    {
      "epoch": 0.34615384615384615,
      "grad_norm": 0.1036024829347823,
      "learning_rate": 9.838207259506891e-06,
      "loss": 0.3489,
      "step": 495
    },
    {
      "epoch": 0.34965034965034963,
      "grad_norm": 0.0999873441403134,
      "learning_rate": 9.83041796015262e-06,
      "loss": 0.3737,
      "step": 500
    },
    {
      "epoch": 0.3531468531468531,
      "grad_norm": 0.09945763957696792,
      "learning_rate": 9.82244877120851e-06,
      "loss": 0.3481,
      "step": 505
    },
    {
      "epoch": 0.35664335664335667,
      "grad_norm": 0.10570852887536689,
      "learning_rate": 9.814299989455118e-06,
      "loss": 0.372,
      "step": 510
    },
    {
      "epoch": 0.36013986013986016,
      "grad_norm": 0.11779899500245009,
      "learning_rate": 9.805971918361215e-06,
      "loss": 0.3819,
      "step": 515
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.09363400449170123,
      "learning_rate": 9.797464868072489e-06,
      "loss": 0.3154,
      "step": 520
    },
    {
      "epoch": 0.36713286713286714,
      "grad_norm": 0.11108253738506654,
      "learning_rate": 9.788779155399988e-06,
      "loss": 0.3472,
      "step": 525
    },
    {
      "epoch": 0.3706293706293706,
      "grad_norm": 0.10442522825796895,
      "learning_rate": 9.779915103808328e-06,
      "loss": 0.3423,
      "step": 530
    },
    {
      "epoch": 0.3741258741258741,
      "grad_norm": 0.10361393967804074,
      "learning_rate": 9.770873043403648e-06,
      "loss": 0.3721,
      "step": 535
    },
    {
      "epoch": 0.3776223776223776,
      "grad_norm": 0.09216332201091619,
      "learning_rate": 9.761653310921307e-06,
      "loss": 0.3437,
      "step": 540
    },
    {
      "epoch": 0.3811188811188811,
      "grad_norm": 0.09176246000601548,
      "learning_rate": 9.752256249713352e-06,
      "loss": 0.3374,
      "step": 545
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.10550436648932508,
      "learning_rate": 9.742682209735727e-06,
      "loss": 0.3352,
      "step": 550
    },
    {
      "epoch": 0.3881118881118881,
      "grad_norm": 0.10200282126565531,
      "learning_rate": 9.73293154753525e-06,
      "loss": 0.3357,
      "step": 555
    },
    {
      "epoch": 0.3916083916083916,
      "grad_norm": 0.12406660076628567,
      "learning_rate": 9.723004626236314e-06,
      "loss": 0.3691,
      "step": 560
    },
    {
      "epoch": 0.3951048951048951,
      "grad_norm": 0.09183441698745344,
      "learning_rate": 9.712901815527387e-06,
      "loss": 0.3613,
      "step": 565
    },
    {
      "epoch": 0.3986013986013986,
      "grad_norm": 0.09566399172613213,
      "learning_rate": 9.702623491647232e-06,
      "loss": 0.3532,
      "step": 570
    },
    {
      "epoch": 0.4020979020979021,
      "grad_norm": 0.0963370933234377,
      "learning_rate": 9.692170037370899e-06,
      "loss": 0.3541,
      "step": 575
    },
    {
      "epoch": 0.40559440559440557,
      "grad_norm": 0.09302759265272703,
      "learning_rate": 9.68154184199546e-06,
      "loss": 0.3288,
      "step": 580
    },
    {
      "epoch": 0.4090909090909091,
      "grad_norm": 0.11008699095457149,
      "learning_rate": 9.670739301325534e-06,
      "loss": 0.3644,
      "step": 585
    },
    {
      "epoch": 0.4125874125874126,
      "grad_norm": 0.10421534071446516,
      "learning_rate": 9.659762817658524e-06,
      "loss": 0.3506,
      "step": 590
    },
    {
      "epoch": 0.4160839160839161,
      "grad_norm": 0.11481688994461432,
      "learning_rate": 9.648612799769644e-06,
      "loss": 0.3605,
      "step": 595
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 0.09668173580570356,
      "learning_rate": 9.6372896628967e-06,
      "loss": 0.3436,
      "step": 600
    },
    {
      "epoch": 0.4230769230769231,
      "grad_norm": 0.08699310096824431,
      "learning_rate": 9.62579382872462e-06,
      "loss": 0.3088,
      "step": 605
    },
    {
      "epoch": 0.42657342657342656,
      "grad_norm": 0.09719113972324987,
      "learning_rate": 9.614125725369748e-06,
      "loss": 0.3612,
      "step": 610
    },
    {
      "epoch": 0.43006993006993005,
      "grad_norm": 0.10508674899455878,
      "learning_rate": 9.60228578736391e-06,
      "loss": 0.3277,
      "step": 615
    },
    {
      "epoch": 0.43356643356643354,
      "grad_norm": 0.09503889406720173,
      "learning_rate": 9.590274455638225e-06,
      "loss": 0.3537,
      "step": 620
    },
    {
      "epoch": 0.4370629370629371,
      "grad_norm": 0.11120365222874168,
      "learning_rate": 9.578092177506683e-06,
      "loss": 0.3616,
      "step": 625
    },
    {
      "epoch": 0.4405594405594406,
      "grad_norm": 0.11151160682918018,
      "learning_rate": 9.565739406649492e-06,
      "loss": 0.3547,
      "step": 630
    },
    {
      "epoch": 0.44405594405594406,
      "grad_norm": 0.09295867289306316,
      "learning_rate": 9.553216603096175e-06,
      "loss": 0.331,
      "step": 635
    },
    {
      "epoch": 0.44755244755244755,
      "grad_norm": 0.10478040853034726,
      "learning_rate": 9.540524233208449e-06,
      "loss": 0.3749,
      "step": 640
    },
    {
      "epoch": 0.45104895104895104,
      "grad_norm": 0.10064237088507819,
      "learning_rate": 9.527662769662842e-06,
      "loss": 0.3273,
      "step": 645
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.10623134292322291,
      "learning_rate": 9.514632691433108e-06,
      "loss": 0.3311,
      "step": 650
    },
    {
      "epoch": 0.458041958041958,
      "grad_norm": 0.10975739313893847,
      "learning_rate": 9.501434483772371e-06,
      "loss": 0.3451,
      "step": 655
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.10743243735062692,
      "learning_rate": 9.488068638195072e-06,
      "loss": 0.3731,
      "step": 660
    },
    {
      "epoch": 0.46503496503496505,
      "grad_norm": 0.10220806057623066,
      "learning_rate": 9.474535652458647e-06,
      "loss": 0.3319,
      "step": 665
    },
    {
      "epoch": 0.46853146853146854,
      "grad_norm": 0.08961810168599127,
      "learning_rate": 9.460836030545007e-06,
      "loss": 0.3229,
      "step": 670
    },
    {
      "epoch": 0.47202797202797203,
      "grad_norm": 0.09507031307513555,
      "learning_rate": 9.446970282641754e-06,
      "loss": 0.3274,
      "step": 675
    },
    {
      "epoch": 0.4755244755244755,
      "grad_norm": 0.09539291606924824,
      "learning_rate": 9.43293892512319e-06,
      "loss": 0.319,
      "step": 680
    },
    {
      "epoch": 0.479020979020979,
      "grad_norm": 0.09487862903804278,
      "learning_rate": 9.418742480531086e-06,
      "loss": 0.3432,
      "step": 685
    },
    {
      "epoch": 0.4825174825174825,
      "grad_norm": 0.08519888132503603,
      "learning_rate": 9.404381477555216e-06,
      "loss": 0.3221,
      "step": 690
    },
    {
      "epoch": 0.486013986013986,
      "grad_norm": 0.09519451285333426,
      "learning_rate": 9.38985645101368e-06,
      "loss": 0.3278,
      "step": 695
    },
    {
      "epoch": 0.48951048951048953,
      "grad_norm": 0.10140606823126608,
      "learning_rate": 9.375167941832974e-06,
      "loss": 0.3394,
      "step": 700
    },
    {
      "epoch": 0.493006993006993,
      "grad_norm": 0.08999016510099082,
      "learning_rate": 9.360316497027849e-06,
      "loss": 0.3321,
      "step": 705
    },
    {
      "epoch": 0.4965034965034965,
      "grad_norm": 0.09416548225275408,
      "learning_rate": 9.345302669680947e-06,
      "loss": 0.3368,
      "step": 710
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1068809820493292,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.3519,
      "step": 715
    },
    {
      "epoch": 0.5034965034965035,
      "grad_norm": 0.10619786904410229,
      "learning_rate": 9.314790109907983e-06,
      "loss": 0.3321,
      "step": 720
    },
    {
      "epoch": 0.506993006993007,
      "grad_norm": 0.10578753728315218,
      "learning_rate": 9.299292513800129e-06,
      "loss": 0.3361,
      "step": 725
    },
    {
      "epoch": 0.5104895104895105,
      "grad_norm": 0.10437755694971507,
      "learning_rate": 9.283634807744586e-06,
      "loss": 0.3113,
      "step": 730
    },
    {
      "epoch": 0.513986013986014,
      "grad_norm": 0.09272325063447831,
      "learning_rate": 9.267817574849971e-06,
      "loss": 0.3432,
      "step": 735
    },
    {
      "epoch": 0.5174825174825175,
      "grad_norm": 0.09934909346667362,
      "learning_rate": 9.251841404165835e-06,
      "loss": 0.332,
      "step": 740
    },
    {
      "epoch": 0.5209790209790209,
      "grad_norm": 0.10562545989351961,
      "learning_rate": 9.235706890660735e-06,
      "loss": 0.345,
      "step": 745
    },
    {
      "epoch": 0.5244755244755245,
      "grad_norm": 0.0968746153073039,
      "learning_rate": 9.219414635200065e-06,
      "loss": 0.3365,
      "step": 750
    },
    {
      "epoch": 0.527972027972028,
      "grad_norm": 0.09752153903894524,
      "learning_rate": 9.202965244523696e-06,
      "loss": 0.3356,
      "step": 755
    },
    {
      "epoch": 0.5314685314685315,
      "grad_norm": 0.09992772564256337,
      "learning_rate": 9.18635933122337e-06,
      "loss": 0.3717,
      "step": 760
    },
    {
      "epoch": 0.534965034965035,
      "grad_norm": 0.09637037689623751,
      "learning_rate": 9.169597513719881e-06,
      "loss": 0.3461,
      "step": 765
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 0.10181991528420607,
      "learning_rate": 9.152680416240059e-06,
      "loss": 0.3276,
      "step": 770
    },
    {
      "epoch": 0.541958041958042,
      "grad_norm": 0.0967239603020424,
      "learning_rate": 9.135608668793511e-06,
      "loss": 0.352,
      "step": 775
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.10948912538288452,
      "learning_rate": 9.118382907149164e-06,
      "loss": 0.3411,
      "step": 780
    },
    {
      "epoch": 0.548951048951049,
      "grad_norm": 0.10087193753130455,
      "learning_rate": 9.10100377281159e-06,
      "loss": 0.3469,
      "step": 785
    },
    {
      "epoch": 0.5524475524475524,
      "grad_norm": 0.09817901878682254,
      "learning_rate": 9.08347191299711e-06,
      "loss": 0.3401,
      "step": 790
    },
    {
      "epoch": 0.5559440559440559,
      "grad_norm": 0.0996307379404232,
      "learning_rate": 9.065787980609696e-06,
      "loss": 0.3268,
      "step": 795
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 0.09012542402521335,
      "learning_rate": 9.047952634216652e-06,
      "loss": 0.3142,
      "step": 800
    },
    {
      "epoch": 0.5629370629370629,
      "grad_norm": 0.09876973070500755,
      "learning_rate": 9.029966538024097e-06,
      "loss": 0.328,
      "step": 805
    },
    {
      "epoch": 0.5664335664335665,
      "grad_norm": 0.10300636050922314,
      "learning_rate": 9.011830361852217e-06,
      "loss": 0.2929,
      "step": 810
    },
    {
      "epoch": 0.5699300699300699,
      "grad_norm": 0.09453575507510487,
      "learning_rate": 8.993544781110328e-06,
      "loss": 0.3458,
      "step": 815
    },
    {
      "epoch": 0.5734265734265734,
      "grad_norm": 0.09645704079355306,
      "learning_rate": 8.975110476771724e-06,
      "loss": 0.3265,
      "step": 820
    },
    {
      "epoch": 0.5769230769230769,
      "grad_norm": 0.10659230074391375,
      "learning_rate": 8.95652813534831e-06,
      "loss": 0.3408,
      "step": 825
    },
    {
      "epoch": 0.5804195804195804,
      "grad_norm": 0.10303679318848774,
      "learning_rate": 8.937798448865043e-06,
      "loss": 0.345,
      "step": 830
    },
    {
      "epoch": 0.583916083916084,
      "grad_norm": 0.09708850064577942,
      "learning_rate": 8.918922114834156e-06,
      "loss": 0.3542,
      "step": 835
    },
    {
      "epoch": 0.5874125874125874,
      "grad_norm": 0.09885418371965929,
      "learning_rate": 8.899899836229186e-06,
      "loss": 0.3437,
      "step": 840
    },
    {
      "epoch": 0.5909090909090909,
      "grad_norm": 0.10137363116042968,
      "learning_rate": 8.880732321458785e-06,
      "loss": 0.3419,
      "step": 845
    },
    {
      "epoch": 0.5944055944055944,
      "grad_norm": 0.09269761185654748,
      "learning_rate": 8.861420284340352e-06,
      "loss": 0.324,
      "step": 850
    },
    {
      "epoch": 0.5979020979020979,
      "grad_norm": 0.10554372594782091,
      "learning_rate": 8.841964444073437e-06,
      "loss": 0.3214,
      "step": 855
    },
    {
      "epoch": 0.6013986013986014,
      "grad_norm": 0.08491975086737241,
      "learning_rate": 8.822365525212968e-06,
      "loss": 0.3247,
      "step": 860
    },
    {
      "epoch": 0.6048951048951049,
      "grad_norm": 0.10098499008728264,
      "learning_rate": 8.802624257642262e-06,
      "loss": 0.3188,
      "step": 865
    },
    {
      "epoch": 0.6083916083916084,
      "grad_norm": 0.09815538910569382,
      "learning_rate": 8.782741376545837e-06,
      "loss": 0.3221,
      "step": 870
    },
    {
      "epoch": 0.6118881118881119,
      "grad_norm": 0.08557023327230194,
      "learning_rate": 8.76271762238205e-06,
      "loss": 0.317,
      "step": 875
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.10488906156950217,
      "learning_rate": 8.742553740855507e-06,
      "loss": 0.3289,
      "step": 880
    },
    {
      "epoch": 0.6188811188811189,
      "grad_norm": 0.09919490213339538,
      "learning_rate": 8.722250482889293e-06,
      "loss": 0.3309,
      "step": 885
    },
    {
      "epoch": 0.6223776223776224,
      "grad_norm": 0.10541653894818209,
      "learning_rate": 8.701808604597022e-06,
      "loss": 0.3214,
      "step": 890
    },
    {
      "epoch": 0.6258741258741258,
      "grad_norm": 0.08789508388787895,
      "learning_rate": 8.681228867254657e-06,
      "loss": 0.3163,
      "step": 895
    },
    {
      "epoch": 0.6293706293706294,
      "grad_norm": 0.09322008095706158,
      "learning_rate": 8.660512037272173e-06,
      "loss": 0.3328,
      "step": 900
    },
    {
      "epoch": 0.6328671328671329,
      "grad_norm": 0.08832274612194231,
      "learning_rate": 8.639658886165012e-06,
      "loss": 0.3261,
      "step": 905
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.09800838728002931,
      "learning_rate": 8.61867019052535e-06,
      "loss": 0.3233,
      "step": 910
    },
    {
      "epoch": 0.6398601398601399,
      "grad_norm": 0.09584979890867579,
      "learning_rate": 8.597546731993182e-06,
      "loss": 0.3334,
      "step": 915
    },
    {
      "epoch": 0.6433566433566433,
      "grad_norm": 0.09025651416414181,
      "learning_rate": 8.576289297227201e-06,
      "loss": 0.3011,
      "step": 920
    },
    {
      "epoch": 0.6468531468531469,
      "grad_norm": 0.09894309732175201,
      "learning_rate": 8.554898677875509e-06,
      "loss": 0.318,
      "step": 925
    },
    {
      "epoch": 0.6503496503496503,
      "grad_norm": 0.1075979690627022,
      "learning_rate": 8.533375670546142e-06,
      "loss": 0.3431,
      "step": 930
    },
    {
      "epoch": 0.6538461538461539,
      "grad_norm": 0.09991378975343375,
      "learning_rate": 8.511721076777388e-06,
      "loss": 0.3491,
      "step": 935
    },
    {
      "epoch": 0.6573426573426573,
      "grad_norm": 0.09509906899332317,
      "learning_rate": 8.48993570300795e-06,
      "loss": 0.3236,
      "step": 940
    },
    {
      "epoch": 0.6608391608391608,
      "grad_norm": 0.0925120550424728,
      "learning_rate": 8.46802036054691e-06,
      "loss": 0.3286,
      "step": 945
    },
    {
      "epoch": 0.6643356643356644,
      "grad_norm": 0.09779713093015033,
      "learning_rate": 8.445975865543509e-06,
      "loss": 0.3317,
      "step": 950
    },
    {
      "epoch": 0.6678321678321678,
      "grad_norm": 0.10820894771177475,
      "learning_rate": 8.423803038956763e-06,
      "loss": 0.3397,
      "step": 955
    },
    {
      "epoch": 0.6713286713286714,
      "grad_norm": 0.10837891937167829,
      "learning_rate": 8.401502706524884e-06,
      "loss": 0.3188,
      "step": 960
    },
    {
      "epoch": 0.6748251748251748,
      "grad_norm": 0.10806627523785152,
      "learning_rate": 8.379075698734523e-06,
      "loss": 0.3146,
      "step": 965
    },
    {
      "epoch": 0.6783216783216783,
      "grad_norm": 0.10973597750269369,
      "learning_rate": 8.356522850789852e-06,
      "loss": 0.3129,
      "step": 970
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.08860258388676324,
      "learning_rate": 8.33384500258146e-06,
      "loss": 0.3043,
      "step": 975
    },
    {
      "epoch": 0.6853146853146853,
      "grad_norm": 0.09048043976226512,
      "learning_rate": 8.31104299865506e-06,
      "loss": 0.3164,
      "step": 980
    },
    {
      "epoch": 0.6888111888111889,
      "grad_norm": 0.10902822229969923,
      "learning_rate": 8.288117688180064e-06,
      "loss": 0.3338,
      "step": 985
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.10086469525952368,
      "learning_rate": 8.265069924917925e-06,
      "loss": 0.3096,
      "step": 990
    },
    {
      "epoch": 0.6958041958041958,
      "grad_norm": 0.10018611997857327,
      "learning_rate": 8.241900567190376e-06,
      "loss": 0.3268,
      "step": 995
    },
    {
      "epoch": 0.6993006993006993,
      "grad_norm": 0.1025620986753826,
      "learning_rate": 8.218610477847435e-06,
      "loss": 0.3421,
      "step": 1000
    },
    {
      "epoch": 0.7027972027972028,
      "grad_norm": 0.09312750964973798,
      "learning_rate": 8.1952005242353e-06,
      "loss": 0.3036,
      "step": 1005
    },
    {
      "epoch": 0.7062937062937062,
      "grad_norm": 0.10003495657539234,
      "learning_rate": 8.171671578164023e-06,
      "loss": 0.326,
      "step": 1010
    },
    {
      "epoch": 0.7097902097902098,
      "grad_norm": 0.08739758803456263,
      "learning_rate": 8.148024515875057e-06,
      "loss": 0.3385,
      "step": 1015
    },
    {
      "epoch": 0.7132867132867133,
      "grad_norm": 0.09860060092942846,
      "learning_rate": 8.124260218008627e-06,
      "loss": 0.3327,
      "step": 1020
    },
    {
      "epoch": 0.7167832167832168,
      "grad_norm": 0.08936884639007185,
      "learning_rate": 8.100379569570919e-06,
      "loss": 0.2964,
      "step": 1025
    },
    {
      "epoch": 0.7202797202797203,
      "grad_norm": 0.10077088027776043,
      "learning_rate": 8.076383459901138e-06,
      "loss": 0.3544,
      "step": 1030
    },
    {
      "epoch": 0.7237762237762237,
      "grad_norm": 0.10946264826220208,
      "learning_rate": 8.052272782638375e-06,
      "loss": 0.3626,
      "step": 1035
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.10881197015520062,
      "learning_rate": 8.028048435688333e-06,
      "loss": 0.3362,
      "step": 1040
    },
    {
      "epoch": 0.7307692307692307,
      "grad_norm": 0.09162086485419195,
      "learning_rate": 8.003711321189895e-06,
      "loss": 0.3219,
      "step": 1045
    },
    {
      "epoch": 0.7342657342657343,
      "grad_norm": 0.10225003056884106,
      "learning_rate": 7.979262345481511e-06,
      "loss": 0.3207,
      "step": 1050
    },
    {
      "epoch": 0.7377622377622378,
      "grad_norm": 0.09804273829576925,
      "learning_rate": 7.95470241906746e-06,
      "loss": 0.3214,
      "step": 1055
    },
    {
      "epoch": 0.7412587412587412,
      "grad_norm": 0.09811184755001341,
      "learning_rate": 7.930032456583931e-06,
      "loss": 0.2961,
      "step": 1060
    },
    {
      "epoch": 0.7447552447552448,
      "grad_norm": 0.10078151905469016,
      "learning_rate": 7.905253376764973e-06,
      "loss": 0.324,
      "step": 1065
    },
    {
      "epoch": 0.7482517482517482,
      "grad_norm": 0.09538296555909533,
      "learning_rate": 7.880366102408266e-06,
      "loss": 0.3092,
      "step": 1070
    },
    {
      "epoch": 0.7517482517482518,
      "grad_norm": 0.09881388066923626,
      "learning_rate": 7.85537156034077e-06,
      "loss": 0.3059,
      "step": 1075
    },
    {
      "epoch": 0.7552447552447552,
      "grad_norm": 0.09384506360180256,
      "learning_rate": 7.830270681384196e-06,
      "loss": 0.3221,
      "step": 1080
    },
    {
      "epoch": 0.7587412587412588,
      "grad_norm": 0.10400424434170057,
      "learning_rate": 7.805064400320348e-06,
      "loss": 0.33,
      "step": 1085
    },
    {
      "epoch": 0.7622377622377622,
      "grad_norm": 0.10186295297499602,
      "learning_rate": 7.779753655856313e-06,
      "loss": 0.3088,
      "step": 1090
    },
    {
      "epoch": 0.7657342657342657,
      "grad_norm": 0.10385785062642248,
      "learning_rate": 7.754339390589497e-06,
      "loss": 0.3397,
      "step": 1095
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.08980489141153736,
      "learning_rate": 7.728822550972523e-06,
      "loss": 0.3132,
      "step": 1100
    },
    {
      "epoch": 0.7727272727272727,
      "grad_norm": 0.11113732222360632,
      "learning_rate": 7.703204087277989e-06,
      "loss": 0.328,
      "step": 1105
    },
    {
      "epoch": 0.7762237762237763,
      "grad_norm": 0.09764126860708802,
      "learning_rate": 7.67748495356307e-06,
      "loss": 0.304,
      "step": 1110
    },
    {
      "epoch": 0.7797202797202797,
      "grad_norm": 0.09722225940353538,
      "learning_rate": 7.651666107633996e-06,
      "loss": 0.3072,
      "step": 1115
    },
    {
      "epoch": 0.7832167832167832,
      "grad_norm": 0.10470107374815786,
      "learning_rate": 7.625748511010382e-06,
      "loss": 0.3303,
      "step": 1120
    },
    {
      "epoch": 0.7867132867132867,
      "grad_norm": 0.09996898904388811,
      "learning_rate": 7.599733128889413e-06,
      "loss": 0.3052,
      "step": 1125
    },
    {
      "epoch": 0.7902097902097902,
      "grad_norm": 0.10066263112640013,
      "learning_rate": 7.573620930109906e-06,
      "loss": 0.3217,
      "step": 1130
    },
    {
      "epoch": 0.7937062937062938,
      "grad_norm": 0.09613338639169412,
      "learning_rate": 7.547412887116224e-06,
      "loss": 0.3141,
      "step": 1135
    },
    {
      "epoch": 0.7972027972027972,
      "grad_norm": 0.09874736039723754,
      "learning_rate": 7.521109975922069e-06,
      "loss": 0.326,
      "step": 1140
    },
    {
      "epoch": 0.8006993006993007,
      "grad_norm": 0.09674056322473204,
      "learning_rate": 7.494713176074128e-06,
      "loss": 0.3173,
      "step": 1145
    },
    {
      "epoch": 0.8041958041958042,
      "grad_norm": 0.10992269958755031,
      "learning_rate": 7.468223470615593e-06,
      "loss": 0.3486,
      "step": 1150
    },
    {
      "epoch": 0.8076923076923077,
      "grad_norm": 0.10254380915376603,
      "learning_rate": 7.441641846049557e-06,
      "loss": 0.3164,
      "step": 1155
    },
    {
      "epoch": 0.8111888111888111,
      "grad_norm": 0.10628009968506902,
      "learning_rate": 7.414969292302267e-06,
      "loss": 0.3225,
      "step": 1160
    },
    {
      "epoch": 0.8146853146853147,
      "grad_norm": 0.09700213527412797,
      "learning_rate": 7.388206802686272e-06,
      "loss": 0.3189,
      "step": 1165
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.10807703866655266,
      "learning_rate": 7.361355373863415e-06,
      "loss": 0.3039,
      "step": 1170
    },
    {
      "epoch": 0.8216783216783217,
      "grad_norm": 0.10284550260251323,
      "learning_rate": 7.334416005807726e-06,
      "loss": 0.3153,
      "step": 1175
    },
    {
      "epoch": 0.8251748251748252,
      "grad_norm": 0.12191442002708408,
      "learning_rate": 7.307389701768183e-06,
      "loss": 0.3243,
      "step": 1180
    },
    {
      "epoch": 0.8286713286713286,
      "grad_norm": 0.08592724455868671,
      "learning_rate": 7.280277468231343e-06,
      "loss": 0.3205,
      "step": 1185
    },
    {
      "epoch": 0.8321678321678322,
      "grad_norm": 0.09918852196659654,
      "learning_rate": 7.253080314883863e-06,
      "loss": 0.2841,
      "step": 1190
    },
    {
      "epoch": 0.8356643356643356,
      "grad_norm": 0.09523362546958633,
      "learning_rate": 7.2257992545749045e-06,
      "loss": 0.3168,
      "step": 1195
    },
    {
      "epoch": 0.8391608391608392,
      "grad_norm": 0.09400187815795734,
      "learning_rate": 7.198435303278401e-06,
      "loss": 0.3212,
      "step": 1200
    },
    {
      "epoch": 0.8426573426573427,
      "grad_norm": 0.10200525168094628,
      "learning_rate": 7.170989480055233e-06,
      "loss": 0.3401,
      "step": 1205
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 0.10299521798155448,
      "learning_rate": 7.143462807015271e-06,
      "loss": 0.3329,
      "step": 1210
    },
    {
      "epoch": 0.8496503496503497,
      "grad_norm": 0.09620298567714382,
      "learning_rate": 7.1158563092793165e-06,
      "loss": 0.3033,
      "step": 1215
    },
    {
      "epoch": 0.8531468531468531,
      "grad_norm": 0.10083825621096794,
      "learning_rate": 7.08817101494092e-06,
      "loss": 0.3107,
      "step": 1220
    },
    {
      "epoch": 0.8566433566433567,
      "grad_norm": 0.10331273870293395,
      "learning_rate": 7.060407955028098e-06,
      "loss": 0.3056,
      "step": 1225
    },
    {
      "epoch": 0.8601398601398601,
      "grad_norm": 0.09524481989587572,
      "learning_rate": 7.032568163464932e-06,
      "loss": 0.315,
      "step": 1230
    },
    {
      "epoch": 0.8636363636363636,
      "grad_norm": 0.11236665596957292,
      "learning_rate": 7.004652677033069e-06,
      "loss": 0.3103,
      "step": 1235
    },
    {
      "epoch": 0.8671328671328671,
      "grad_norm": 0.10130841806082222,
      "learning_rate": 6.976662535333107e-06,
      "loss": 0.3077,
      "step": 1240
    },
    {
      "epoch": 0.8706293706293706,
      "grad_norm": 0.10051782592395737,
      "learning_rate": 6.948598780745885e-06,
      "loss": 0.295,
      "step": 1245
    },
    {
      "epoch": 0.8741258741258742,
      "grad_norm": 0.11389848718730608,
      "learning_rate": 6.920462458393652e-06,
      "loss": 0.3359,
      "step": 1250
    },
    {
      "epoch": 0.8776223776223776,
      "grad_norm": 0.1078175959725227,
      "learning_rate": 6.892254616101158e-06,
      "loss": 0.3201,
      "step": 1255
    },
    {
      "epoch": 0.8811188811188811,
      "grad_norm": 0.10829824752019789,
      "learning_rate": 6.86397630435663e-06,
      "loss": 0.2934,
      "step": 1260
    },
    {
      "epoch": 0.8846153846153846,
      "grad_norm": 0.11378406063480054,
      "learning_rate": 6.8356285762726385e-06,
      "loss": 0.3128,
      "step": 1265
    },
    {
      "epoch": 0.8881118881118881,
      "grad_norm": 0.1155802304379582,
      "learning_rate": 6.807212487546897e-06,
      "loss": 0.2941,
      "step": 1270
    },
    {
      "epoch": 0.8916083916083916,
      "grad_norm": 0.09837564394321802,
      "learning_rate": 6.7787290964229325e-06,
      "loss": 0.3077,
      "step": 1275
    },
    {
      "epoch": 0.8951048951048951,
      "grad_norm": 0.1025255748932635,
      "learning_rate": 6.750179463650681e-06,
      "loss": 0.3204,
      "step": 1280
    },
    {
      "epoch": 0.8986013986013986,
      "grad_norm": 0.10639227652882788,
      "learning_rate": 6.721564652446987e-06,
      "loss": 0.3169,
      "step": 1285
    },
    {
      "epoch": 0.9020979020979021,
      "grad_norm": 0.11824484132864124,
      "learning_rate": 6.692885728456001e-06,
      "loss": 0.3267,
      "step": 1290
    },
    {
      "epoch": 0.9055944055944056,
      "grad_norm": 0.10502516014597198,
      "learning_rate": 6.6641437597095005e-06,
      "loss": 0.2983,
      "step": 1295
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.10220450208484791,
      "learning_rate": 6.635339816587109e-06,
      "loss": 0.2895,
      "step": 1300
    },
    {
      "epoch": 0.9125874125874126,
      "grad_norm": 0.10083747573334342,
      "learning_rate": 6.606474971776442e-06,
      "loss": 0.3349,
      "step": 1305
    },
    {
      "epoch": 0.916083916083916,
      "grad_norm": 0.09977510069920313,
      "learning_rate": 6.577550300233151e-06,
      "loss": 0.3165,
      "step": 1310
    },
    {
      "epoch": 0.9195804195804196,
      "grad_norm": 0.1120782507105497,
      "learning_rate": 6.548566879140897e-06,
      "loss": 0.302,
      "step": 1315
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.10220587124302934,
      "learning_rate": 6.519525787871235e-06,
      "loss": 0.3225,
      "step": 1320
    },
    {
      "epoch": 0.9265734265734266,
      "grad_norm": 0.09733851795585291,
      "learning_rate": 6.490428107943409e-06,
      "loss": 0.3052,
      "step": 1325
    },
    {
      "epoch": 0.9300699300699301,
      "grad_norm": 0.12181630950221083,
      "learning_rate": 6.461274922984087e-06,
      "loss": 0.3034,
      "step": 1330
    },
    {
      "epoch": 0.9335664335664335,
      "grad_norm": 0.09985408410518533,
      "learning_rate": 6.4320673186869954e-06,
      "loss": 0.2963,
      "step": 1335
    },
    {
      "epoch": 0.9370629370629371,
      "grad_norm": 0.10053095003817707,
      "learning_rate": 6.402806382772494e-06,
      "loss": 0.3119,
      "step": 1340
    },
    {
      "epoch": 0.9405594405594405,
      "grad_norm": 0.12294363970356818,
      "learning_rate": 6.373493204947065e-06,
      "loss": 0.2977,
      "step": 1345
    },
    {
      "epoch": 0.9440559440559441,
      "grad_norm": 0.08781663953826889,
      "learning_rate": 6.344128876862729e-06,
      "loss": 0.2734,
      "step": 1350
    },
    {
      "epoch": 0.9475524475524476,
      "grad_norm": 0.11355851741329681,
      "learning_rate": 6.3147144920763985e-06,
      "loss": 0.3252,
      "step": 1355
    },
    {
      "epoch": 0.951048951048951,
      "grad_norm": 0.10654773663336697,
      "learning_rate": 6.2852511460091406e-06,
      "loss": 0.2822,
      "step": 1360
    },
    {
      "epoch": 0.9545454545454546,
      "grad_norm": 0.12320855187368913,
      "learning_rate": 6.255739935905396e-06,
      "loss": 0.3056,
      "step": 1365
    },
    {
      "epoch": 0.958041958041958,
      "grad_norm": 0.118993538118343,
      "learning_rate": 6.226181960792107e-06,
      "loss": 0.3005,
      "step": 1370
    },
    {
      "epoch": 0.9615384615384616,
      "grad_norm": 0.1306844890782949,
      "learning_rate": 6.1965783214377895e-06,
      "loss": 0.3034,
      "step": 1375
    },
    {
      "epoch": 0.965034965034965,
      "grad_norm": 0.09291220906399976,
      "learning_rate": 6.166930120311545e-06,
      "loss": 0.2922,
      "step": 1380
    },
    {
      "epoch": 0.9685314685314685,
      "grad_norm": 0.12845778629594434,
      "learning_rate": 6.137238461541998e-06,
      "loss": 0.3026,
      "step": 1385
    },
    {
      "epoch": 0.972027972027972,
      "grad_norm": 0.1054290513425815,
      "learning_rate": 6.1075044508761804e-06,
      "loss": 0.3148,
      "step": 1390
    },
    {
      "epoch": 0.9755244755244755,
      "grad_norm": 0.11291261778659534,
      "learning_rate": 6.077729195638349e-06,
      "loss": 0.2967,
      "step": 1395
    },
    {
      "epoch": 0.9790209790209791,
      "grad_norm": 0.11709723242508564,
      "learning_rate": 6.047913804688751e-06,
      "loss": 0.3139,
      "step": 1400
    },
    {
      "epoch": 0.9825174825174825,
      "grad_norm": 0.10296333711215835,
      "learning_rate": 6.018059388382327e-06,
      "loss": 0.3199,
      "step": 1405
    },
    {
      "epoch": 0.986013986013986,
      "grad_norm": 0.12283451293086087,
      "learning_rate": 5.988167058527361e-06,
      "loss": 0.302,
      "step": 1410
    },
    {
      "epoch": 0.9895104895104895,
      "grad_norm": 0.10500101316339835,
      "learning_rate": 5.958237928344077e-06,
      "loss": 0.3036,
      "step": 1415
    },
    {
      "epoch": 0.993006993006993,
      "grad_norm": 0.10463926702466099,
      "learning_rate": 5.928273112423177e-06,
      "loss": 0.3285,
      "step": 1420
    },
    {
      "epoch": 0.9965034965034965,
      "grad_norm": 0.1081073139740873,
      "learning_rate": 5.898273726684338e-06,
      "loss": 0.3107,
      "step": 1425
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.11081760524407615,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.3149,
      "step": 1430
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.31062406301498413,
      "eval_runtime": 32.6264,
      "eval_samples_per_second": 21.271,
      "eval_steps_per_second": 5.333,
      "step": 1430
    },
    {
      "epoch": 1.0034965034965035,
      "grad_norm": 0.10766057816314371,
      "learning_rate": 5.838175715827015e-06,
      "loss": 0.2744,
      "step": 1435
    },
    {
      "epoch": 1.006993006993007,
      "grad_norm": 0.11009214726577311,
      "learning_rate": 5.808079328818488e-06,
      "loss": 0.2737,
      "step": 1440
    },
    {
      "epoch": 1.0104895104895104,
      "grad_norm": 0.1317061785977455,
      "learning_rate": 5.777952848128578e-06,
      "loss": 0.2959,
      "step": 1445
    },
    {
      "epoch": 1.013986013986014,
      "grad_norm": 0.10172499662912773,
      "learning_rate": 5.747797395697525e-06,
      "loss": 0.2807,
      "step": 1450
    },
    {
      "epoch": 1.0174825174825175,
      "grad_norm": 0.1281707342936968,
      "learning_rate": 5.717614094544497e-06,
      "loss": 0.2871,
      "step": 1455
    },
    {
      "epoch": 1.020979020979021,
      "grad_norm": 0.11017050288756519,
      "learning_rate": 5.687404068725779e-06,
      "loss": 0.2801,
      "step": 1460
    },
    {
      "epoch": 1.0244755244755244,
      "grad_norm": 0.1143774074800733,
      "learning_rate": 5.657168443292909e-06,
      "loss": 0.2768,
      "step": 1465
    },
    {
      "epoch": 1.027972027972028,
      "grad_norm": 0.11510813610035686,
      "learning_rate": 5.626908344250778e-06,
      "loss": 0.29,
      "step": 1470
    },
    {
      "epoch": 1.0314685314685315,
      "grad_norm": 0.11366002894767718,
      "learning_rate": 5.5966248985157e-06,
      "loss": 0.2712,
      "step": 1475
    },
    {
      "epoch": 1.034965034965035,
      "grad_norm": 0.11422066836025106,
      "learning_rate": 5.566319233873446e-06,
      "loss": 0.2946,
      "step": 1480
    },
    {
      "epoch": 1.0384615384615385,
      "grad_norm": 0.1185751830709301,
      "learning_rate": 5.53599247893724e-06,
      "loss": 0.266,
      "step": 1485
    },
    {
      "epoch": 1.0419580419580419,
      "grad_norm": 0.11516228975349556,
      "learning_rate": 5.5056457631057245e-06,
      "loss": 0.2649,
      "step": 1490
    },
    {
      "epoch": 1.0454545454545454,
      "grad_norm": 0.12152666762723717,
      "learning_rate": 5.475280216520913e-06,
      "loss": 0.2723,
      "step": 1495
    },
    {
      "epoch": 1.048951048951049,
      "grad_norm": 0.13384515089024945,
      "learning_rate": 5.444896970026092e-06,
      "loss": 0.2738,
      "step": 1500
    },
    {
      "epoch": 1.0524475524475525,
      "grad_norm": 0.11593977764889435,
      "learning_rate": 5.414497155123708e-06,
      "loss": 0.2567,
      "step": 1505
    },
    {
      "epoch": 1.055944055944056,
      "grad_norm": 0.1617839041181029,
      "learning_rate": 5.384081903933235e-06,
      "loss": 0.2832,
      "step": 1510
    },
    {
      "epoch": 1.0594405594405594,
      "grad_norm": 0.1335680822514584,
      "learning_rate": 5.353652349149007e-06,
      "loss": 0.2676,
      "step": 1515
    },
    {
      "epoch": 1.062937062937063,
      "grad_norm": 0.1322939911029579,
      "learning_rate": 5.323209623998043e-06,
      "loss": 0.2655,
      "step": 1520
    },
    {
      "epoch": 1.0664335664335665,
      "grad_norm": 0.11405919930478965,
      "learning_rate": 5.292754862197831e-06,
      "loss": 0.2672,
      "step": 1525
    },
    {
      "epoch": 1.06993006993007,
      "grad_norm": 0.12155897007205288,
      "learning_rate": 5.2622891979141265e-06,
      "loss": 0.2704,
      "step": 1530
    },
    {
      "epoch": 1.0734265734265733,
      "grad_norm": 0.13843267551292496,
      "learning_rate": 5.231813765718698e-06,
      "loss": 0.2787,
      "step": 1535
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 0.12816302357428316,
      "learning_rate": 5.201329700547077e-06,
      "loss": 0.2522,
      "step": 1540
    },
    {
      "epoch": 1.0804195804195804,
      "grad_norm": 0.1264695722147145,
      "learning_rate": 5.170838137656304e-06,
      "loss": 0.2705,
      "step": 1545
    },
    {
      "epoch": 1.083916083916084,
      "grad_norm": 0.15007226117850633,
      "learning_rate": 5.140340212582639e-06,
      "loss": 0.2711,
      "step": 1550
    },
    {
      "epoch": 1.0874125874125875,
      "grad_norm": 0.11726395920513819,
      "learning_rate": 5.109837061099274e-06,
      "loss": 0.2776,
      "step": 1555
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.11847555566708153,
      "learning_rate": 5.07932981917404e-06,
      "loss": 0.2891,
      "step": 1560
    },
    {
      "epoch": 1.0944055944055944,
      "grad_norm": 0.1310208338558219,
      "learning_rate": 5.0488196229271005e-06,
      "loss": 0.2698,
      "step": 1565
    },
    {
      "epoch": 1.097902097902098,
      "grad_norm": 0.11773031255743507,
      "learning_rate": 5.018307608588637e-06,
      "loss": 0.2553,
      "step": 1570
    },
    {
      "epoch": 1.1013986013986015,
      "grad_norm": 0.1140697610615123,
      "learning_rate": 4.9877949124565415e-06,
      "loss": 0.2691,
      "step": 1575
    },
    {
      "epoch": 1.104895104895105,
      "grad_norm": 0.15402839832086615,
      "learning_rate": 4.957282670854099e-06,
      "loss": 0.2724,
      "step": 1580
    },
    {
      "epoch": 1.1083916083916083,
      "grad_norm": 0.12013471819115372,
      "learning_rate": 4.926772020087663e-06,
      "loss": 0.2942,
      "step": 1585
    },
    {
      "epoch": 1.1118881118881119,
      "grad_norm": 0.13991919878541068,
      "learning_rate": 4.896264096404343e-06,
      "loss": 0.2862,
      "step": 1590
    },
    {
      "epoch": 1.1153846153846154,
      "grad_norm": 0.12556565782941134,
      "learning_rate": 4.865760035949695e-06,
      "loss": 0.2895,
      "step": 1595
    },
    {
      "epoch": 1.118881118881119,
      "grad_norm": 0.11991851334501559,
      "learning_rate": 4.835260974725397e-06,
      "loss": 0.2932,
      "step": 1600
    },
    {
      "epoch": 1.1223776223776223,
      "grad_norm": 0.1124165849676004,
      "learning_rate": 4.804768048546955e-06,
      "loss": 0.2901,
      "step": 1605
    },
    {
      "epoch": 1.1258741258741258,
      "grad_norm": 0.16523502306438984,
      "learning_rate": 4.774282393001397e-06,
      "loss": 0.2706,
      "step": 1610
    },
    {
      "epoch": 1.1293706293706294,
      "grad_norm": 0.13187737235585423,
      "learning_rate": 4.743805143404989e-06,
      "loss": 0.2898,
      "step": 1615
    },
    {
      "epoch": 1.132867132867133,
      "grad_norm": 0.14482445211826303,
      "learning_rate": 4.713337434760949e-06,
      "loss": 0.2779,
      "step": 1620
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.13192656384068377,
      "learning_rate": 4.682880401717178e-06,
      "loss": 0.2496,
      "step": 1625
    },
    {
      "epoch": 1.1398601398601398,
      "grad_norm": 0.1272822534838911,
      "learning_rate": 4.652435178524013e-06,
      "loss": 0.2764,
      "step": 1630
    },
    {
      "epoch": 1.1433566433566433,
      "grad_norm": 0.15754779698519564,
      "learning_rate": 4.622002898991972e-06,
      "loss": 0.2513,
      "step": 1635
    },
    {
      "epoch": 1.1468531468531469,
      "grad_norm": 0.1273601924896046,
      "learning_rate": 4.591584696449549e-06,
      "loss": 0.2663,
      "step": 1640
    },
    {
      "epoch": 1.1503496503496504,
      "grad_norm": 0.13923316833140018,
      "learning_rate": 4.561181703700986e-06,
      "loss": 0.2724,
      "step": 1645
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 0.15408414037242385,
      "learning_rate": 4.530795052984104e-06,
      "loss": 0.3037,
      "step": 1650
    },
    {
      "epoch": 1.1573426573426573,
      "grad_norm": 0.10713473116276952,
      "learning_rate": 4.500425875928129e-06,
      "loss": 0.2396,
      "step": 1655
    },
    {
      "epoch": 1.1608391608391608,
      "grad_norm": 0.14619455258206673,
      "learning_rate": 4.470075303511546e-06,
      "loss": 0.287,
      "step": 1660
    },
    {
      "epoch": 1.1643356643356644,
      "grad_norm": 0.1295335870751147,
      "learning_rate": 4.439744466019993e-06,
      "loss": 0.2868,
      "step": 1665
    },
    {
      "epoch": 1.167832167832168,
      "grad_norm": 0.15230769733669453,
      "learning_rate": 4.4094344930041515e-06,
      "loss": 0.2704,
      "step": 1670
    },
    {
      "epoch": 1.1713286713286712,
      "grad_norm": 0.1475235624098936,
      "learning_rate": 4.379146513237695e-06,
      "loss": 0.2743,
      "step": 1675
    },
    {
      "epoch": 1.1748251748251748,
      "grad_norm": 0.12908423502101304,
      "learning_rate": 4.348881654675241e-06,
      "loss": 0.2552,
      "step": 1680
    },
    {
      "epoch": 1.1783216783216783,
      "grad_norm": 0.13781236488427137,
      "learning_rate": 4.318641044410358e-06,
      "loss": 0.2579,
      "step": 1685
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.11673049554002457,
      "learning_rate": 4.2884258086335755e-06,
      "loss": 0.2498,
      "step": 1690
    },
    {
      "epoch": 1.1853146853146854,
      "grad_norm": 0.15745822517159433,
      "learning_rate": 4.258237072590455e-06,
      "loss": 0.2575,
      "step": 1695
    },
    {
      "epoch": 1.1888111888111887,
      "grad_norm": 0.13232033798128737,
      "learning_rate": 4.228075960539684e-06,
      "loss": 0.2838,
      "step": 1700
    },
    {
      "epoch": 1.1923076923076923,
      "grad_norm": 0.21177422958610773,
      "learning_rate": 4.1979435957111984e-06,
      "loss": 0.2835,
      "step": 1705
    },
    {
      "epoch": 1.1958041958041958,
      "grad_norm": 0.1416944965118804,
      "learning_rate": 4.167841100264369e-06,
      "loss": 0.2611,
      "step": 1710
    },
    {
      "epoch": 1.1993006993006994,
      "grad_norm": 0.14549768570810312,
      "learning_rate": 4.1377695952461905e-06,
      "loss": 0.2462,
      "step": 1715
    },
    {
      "epoch": 1.2027972027972027,
      "grad_norm": 0.1331292507113191,
      "learning_rate": 4.107730200549549e-06,
      "loss": 0.2546,
      "step": 1720
    },
    {
      "epoch": 1.2062937062937062,
      "grad_norm": 0.1408385313729075,
      "learning_rate": 4.0777240348715085e-06,
      "loss": 0.2829,
      "step": 1725
    },
    {
      "epoch": 1.2097902097902098,
      "grad_norm": 0.15912204855524056,
      "learning_rate": 4.04775221567165e-06,
      "loss": 0.2597,
      "step": 1730
    },
    {
      "epoch": 1.2132867132867133,
      "grad_norm": 0.12361692499348212,
      "learning_rate": 4.017815859130462e-06,
      "loss": 0.2659,
      "step": 1735
    },
    {
      "epoch": 1.2167832167832167,
      "grad_norm": 0.14028932029187688,
      "learning_rate": 3.987916080107762e-06,
      "loss": 0.2472,
      "step": 1740
    },
    {
      "epoch": 1.2202797202797202,
      "grad_norm": 0.15019199750221818,
      "learning_rate": 3.958053992101189e-06,
      "loss": 0.2571,
      "step": 1745
    },
    {
      "epoch": 1.2237762237762237,
      "grad_norm": 0.15686203265871243,
      "learning_rate": 3.928230707204729e-06,
      "loss": 0.2568,
      "step": 1750
    },
    {
      "epoch": 1.2272727272727273,
      "grad_norm": 0.14559379787398902,
      "learning_rate": 3.898447336067297e-06,
      "loss": 0.2543,
      "step": 1755
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.15318431900692528,
      "learning_rate": 3.86870498785139e-06,
      "loss": 0.2615,
      "step": 1760
    },
    {
      "epoch": 1.2342657342657342,
      "grad_norm": 0.16672802213473545,
      "learning_rate": 3.8390047701917625e-06,
      "loss": 0.2649,
      "step": 1765
    },
    {
      "epoch": 1.2377622377622377,
      "grad_norm": 0.1509448379227764,
      "learning_rate": 3.8093477891541864e-06,
      "loss": 0.2789,
      "step": 1770
    },
    {
      "epoch": 1.2412587412587412,
      "grad_norm": 0.15034940835738483,
      "learning_rate": 3.779735149194258e-06,
      "loss": 0.2669,
      "step": 1775
    },
    {
      "epoch": 1.2447552447552448,
      "grad_norm": 0.12742378337252733,
      "learning_rate": 3.750167953116272e-06,
      "loss": 0.269,
      "step": 1780
    },
    {
      "epoch": 1.2482517482517483,
      "grad_norm": 0.1415258425991099,
      "learning_rate": 3.720647302032144e-06,
      "loss": 0.2585,
      "step": 1785
    },
    {
      "epoch": 1.2517482517482517,
      "grad_norm": 0.15283095902130808,
      "learning_rate": 3.6911742953204046e-06,
      "loss": 0.258,
      "step": 1790
    },
    {
      "epoch": 1.2552447552447552,
      "grad_norm": 0.11791438765354938,
      "learning_rate": 3.6617500305852676e-06,
      "loss": 0.2381,
      "step": 1795
    },
    {
      "epoch": 1.2587412587412588,
      "grad_norm": 0.16117533627420655,
      "learning_rate": 3.6323756036157386e-06,
      "loss": 0.2844,
      "step": 1800
    },
    {
      "epoch": 1.2622377622377623,
      "grad_norm": 0.13228578441014932,
      "learning_rate": 3.6030521083448256e-06,
      "loss": 0.2615,
      "step": 1805
    },
    {
      "epoch": 1.2657342657342658,
      "grad_norm": 0.1565938827059873,
      "learning_rate": 3.5737806368087776e-06,
      "loss": 0.2617,
      "step": 1810
    },
    {
      "epoch": 1.2692307692307692,
      "grad_norm": 0.1373116437798962,
      "learning_rate": 3.544562279106436e-06,
      "loss": 0.2426,
      "step": 1815
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.16723522422094853,
      "learning_rate": 3.5153981233586277e-06,
      "loss": 0.2632,
      "step": 1820
    },
    {
      "epoch": 1.2762237762237763,
      "grad_norm": 0.14852170977207135,
      "learning_rate": 3.48628925566764e-06,
      "loss": 0.2591,
      "step": 1825
    },
    {
      "epoch": 1.2797202797202798,
      "grad_norm": 0.1541353657668109,
      "learning_rate": 3.4572367600767874e-06,
      "loss": 0.2654,
      "step": 1830
    },
    {
      "epoch": 1.2832167832167833,
      "grad_norm": 0.14156565315594816,
      "learning_rate": 3.428241718530022e-06,
      "loss": 0.2532,
      "step": 1835
    },
    {
      "epoch": 1.2867132867132867,
      "grad_norm": 0.15293487729058486,
      "learning_rate": 3.3993052108316566e-06,
      "loss": 0.254,
      "step": 1840
    },
    {
      "epoch": 1.2902097902097902,
      "grad_norm": 0.17876617072800777,
      "learning_rate": 3.370428314606138e-06,
      "loss": 0.2676,
      "step": 1845
    },
    {
      "epoch": 1.2937062937062938,
      "grad_norm": 0.15068794961193974,
      "learning_rate": 3.3416121052579297e-06,
      "loss": 0.2498,
      "step": 1850
    },
    {
      "epoch": 1.297202797202797,
      "grad_norm": 0.16739156779667366,
      "learning_rate": 3.3128576559314507e-06,
      "loss": 0.2613,
      "step": 1855
    },
    {
      "epoch": 1.3006993006993006,
      "grad_norm": 0.1508017647294946,
      "learning_rate": 3.2841660374711144e-06,
      "loss": 0.2503,
      "step": 1860
    },
    {
      "epoch": 1.3041958041958042,
      "grad_norm": 0.17513634557881674,
      "learning_rate": 3.2555383183814544e-06,
      "loss": 0.2829,
      "step": 1865
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.14488732005796676,
      "learning_rate": 3.226975564787322e-06,
      "loss": 0.2495,
      "step": 1870
    },
    {
      "epoch": 1.3111888111888113,
      "grad_norm": 0.1716063629974628,
      "learning_rate": 3.198478840394196e-06,
      "loss": 0.2764,
      "step": 1875
    },
    {
      "epoch": 1.3146853146853146,
      "grad_norm": 0.15073411570574674,
      "learning_rate": 3.170049206448553e-06,
      "loss": 0.2634,
      "step": 1880
    },
    {
      "epoch": 1.3181818181818181,
      "grad_norm": 0.1596468462051206,
      "learning_rate": 3.141687721698363e-06,
      "loss": 0.2482,
      "step": 1885
    },
    {
      "epoch": 1.3216783216783217,
      "grad_norm": 0.14657291269571518,
      "learning_rate": 3.113395442353646e-06,
      "loss": 0.2782,
      "step": 1890
    },
    {
      "epoch": 1.3251748251748252,
      "grad_norm": 0.1480375390369424,
      "learning_rate": 3.0851734220471434e-06,
      "loss": 0.2503,
      "step": 1895
    },
    {
      "epoch": 1.3286713286713288,
      "grad_norm": 0.1963359101219646,
      "learning_rate": 3.0570227117950857e-06,
      "loss": 0.2632,
      "step": 1900
    },
    {
      "epoch": 1.332167832167832,
      "grad_norm": 0.18631718176302156,
      "learning_rate": 3.0289443599580387e-06,
      "loss": 0.253,
      "step": 1905
    },
    {
      "epoch": 1.3356643356643356,
      "grad_norm": 0.13612998044918345,
      "learning_rate": 3.0009394122018744e-06,
      "loss": 0.2423,
      "step": 1910
    },
    {
      "epoch": 1.3391608391608392,
      "grad_norm": 0.18381431890855784,
      "learning_rate": 2.973008911458816e-06,
      "loss": 0.2428,
      "step": 1915
    },
    {
      "epoch": 1.3426573426573427,
      "grad_norm": 0.20712048020248328,
      "learning_rate": 2.9451538978886165e-06,
      "loss": 0.2556,
      "step": 1920
    },
    {
      "epoch": 1.3461538461538463,
      "grad_norm": 0.15797982697139096,
      "learning_rate": 2.917375408839803e-06,
      "loss": 0.2527,
      "step": 1925
    },
    {
      "epoch": 1.3496503496503496,
      "grad_norm": 0.14951172236668173,
      "learning_rate": 2.8896744788110497e-06,
      "loss": 0.2395,
      "step": 1930
    },
    {
      "epoch": 1.3531468531468531,
      "grad_norm": 0.1566429755635428,
      "learning_rate": 2.86205213941267e-06,
      "loss": 0.2337,
      "step": 1935
    },
    {
      "epoch": 1.3566433566433567,
      "grad_norm": 0.1504421653967441,
      "learning_rate": 2.8345094193281673e-06,
      "loss": 0.2475,
      "step": 1940
    },
    {
      "epoch": 1.3601398601398602,
      "grad_norm": 0.19049031412292836,
      "learning_rate": 2.8070473442759517e-06,
      "loss": 0.2445,
      "step": 1945
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.14751350458434645,
      "learning_rate": 2.7796669369711294e-06,
      "loss": 0.2566,
      "step": 1950
    },
    {
      "epoch": 1.367132867132867,
      "grad_norm": 0.1566695975047669,
      "learning_rate": 2.752369217087418e-06,
      "loss": 0.2504,
      "step": 1955
    },
    {
      "epoch": 1.3706293706293706,
      "grad_norm": 0.16838913254004315,
      "learning_rate": 2.7251552012191763e-06,
      "loss": 0.2297,
      "step": 1960
    },
    {
      "epoch": 1.3741258741258742,
      "grad_norm": 0.16877638644782916,
      "learning_rate": 2.698025902843533e-06,
      "loss": 0.2511,
      "step": 1965
    },
    {
      "epoch": 1.3776223776223775,
      "grad_norm": 0.1856173331599545,
      "learning_rate": 2.670982332282661e-06,
      "loss": 0.2361,
      "step": 1970
    },
    {
      "epoch": 1.381118881118881,
      "grad_norm": 0.16276875400206292,
      "learning_rate": 2.644025496666143e-06,
      "loss": 0.244,
      "step": 1975
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 0.15664973933004409,
      "learning_rate": 2.6171563998934605e-06,
      "loss": 0.246,
      "step": 1980
    },
    {
      "epoch": 1.3881118881118881,
      "grad_norm": 0.17956864564608555,
      "learning_rate": 2.590376042596621e-06,
      "loss": 0.2454,
      "step": 1985
    },
    {
      "epoch": 1.3916083916083917,
      "grad_norm": 0.18696613986435107,
      "learning_rate": 2.5636854221028764e-06,
      "loss": 0.2417,
      "step": 1990
    },
    {
      "epoch": 1.395104895104895,
      "grad_norm": 0.17696305617226574,
      "learning_rate": 2.537085532397595e-06,
      "loss": 0.2494,
      "step": 1995
    },
    {
      "epoch": 1.3986013986013985,
      "grad_norm": 0.20472847227078472,
      "learning_rate": 2.5105773640872416e-06,
      "loss": 0.2328,
      "step": 2000
    },
    {
      "epoch": 1.402097902097902,
      "grad_norm": 0.18133078891008128,
      "learning_rate": 2.484161904362481e-06,
      "loss": 0.2215,
      "step": 2005
    },
    {
      "epoch": 1.4055944055944056,
      "grad_norm": 0.17026054782110633,
      "learning_rate": 2.4578401369614196e-06,
      "loss": 0.2504,
      "step": 2010
    },
    {
      "epoch": 1.4090909090909092,
      "grad_norm": 0.21857345320736957,
      "learning_rate": 2.4316130421329696e-06,
      "loss": 0.2464,
      "step": 2015
    },
    {
      "epoch": 1.4125874125874125,
      "grad_norm": 0.18564475549064688,
      "learning_rate": 2.4054815966003365e-06,
      "loss": 0.2451,
      "step": 2020
    },
    {
      "epoch": 1.416083916083916,
      "grad_norm": 0.2042740525536836,
      "learning_rate": 2.379446773524656e-06,
      "loss": 0.218,
      "step": 2025
    },
    {
      "epoch": 1.4195804195804196,
      "grad_norm": 0.17834518848018271,
      "learning_rate": 2.3535095424687443e-06,
      "loss": 0.2158,
      "step": 2030
    },
    {
      "epoch": 1.4230769230769231,
      "grad_norm": 0.1510768312655819,
      "learning_rate": 2.3276708693609947e-06,
      "loss": 0.2447,
      "step": 2035
    },
    {
      "epoch": 1.4265734265734267,
      "grad_norm": 0.21239587512990743,
      "learning_rate": 2.301931716459405e-06,
      "loss": 0.2439,
      "step": 2040
    },
    {
      "epoch": 1.43006993006993,
      "grad_norm": 0.19287764601763094,
      "learning_rate": 2.2762930423157346e-06,
      "loss": 0.2382,
      "step": 2045
    },
    {
      "epoch": 1.4335664335664335,
      "grad_norm": 0.17488050902596525,
      "learning_rate": 2.2507558017398263e-06,
      "loss": 0.2301,
      "step": 2050
    },
    {
      "epoch": 1.437062937062937,
      "grad_norm": 0.206988640905013,
      "learning_rate": 2.2253209457640205e-06,
      "loss": 0.2411,
      "step": 2055
    },
    {
      "epoch": 1.4405594405594406,
      "grad_norm": 0.1519008631153711,
      "learning_rate": 2.1999894216077633e-06,
      "loss": 0.2451,
      "step": 2060
    },
    {
      "epoch": 1.4440559440559442,
      "grad_norm": 0.2033962684370646,
      "learning_rate": 2.174762172642319e-06,
      "loss": 0.2231,
      "step": 2065
    },
    {
      "epoch": 1.4475524475524475,
      "grad_norm": 0.17331710207175258,
      "learning_rate": 2.149640138355632e-06,
      "loss": 0.2507,
      "step": 2070
    },
    {
      "epoch": 1.451048951048951,
      "grad_norm": 0.2664458793817291,
      "learning_rate": 2.1246242543173613e-06,
      "loss": 0.2448,
      "step": 2075
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.22399416120591836,
      "learning_rate": 2.09971545214401e-06,
      "loss": 0.2336,
      "step": 2080
    },
    {
      "epoch": 1.458041958041958,
      "grad_norm": 0.18773252851709477,
      "learning_rate": 2.074914659464254e-06,
      "loss": 0.2133,
      "step": 2085
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.13913233681949802,
      "learning_rate": 2.050222799884387e-06,
      "loss": 0.2126,
      "step": 2090
    },
    {
      "epoch": 1.465034965034965,
      "grad_norm": 0.1899816341093674,
      "learning_rate": 2.0256407929539203e-06,
      "loss": 0.2468,
      "step": 2095
    },
    {
      "epoch": 1.4685314685314685,
      "grad_norm": 0.18644187412209914,
      "learning_rate": 2.0011695541313514e-06,
      "loss": 0.241,
      "step": 2100
    },
    {
      "epoch": 1.472027972027972,
      "grad_norm": 0.18941735970963874,
      "learning_rate": 1.9768099947500535e-06,
      "loss": 0.2329,
      "step": 2105
    },
    {
      "epoch": 1.4755244755244754,
      "grad_norm": 0.18269213309717788,
      "learning_rate": 1.9525630219843524e-06,
      "loss": 0.2339,
      "step": 2110
    },
    {
      "epoch": 1.479020979020979,
      "grad_norm": 0.2000987782178619,
      "learning_rate": 1.9284295388157328e-06,
      "loss": 0.2244,
      "step": 2115
    },
    {
      "epoch": 1.4825174825174825,
      "grad_norm": 0.17323876227142432,
      "learning_rate": 1.9044104439992145e-06,
      "loss": 0.2306,
      "step": 2120
    },
    {
      "epoch": 1.486013986013986,
      "grad_norm": 0.18819027585465045,
      "learning_rate": 1.880506632029883e-06,
      "loss": 0.2416,
      "step": 2125
    },
    {
      "epoch": 1.4895104895104896,
      "grad_norm": 0.20268371870233162,
      "learning_rate": 1.8567189931095691e-06,
      "loss": 0.218,
      "step": 2130
    },
    {
      "epoch": 1.493006993006993,
      "grad_norm": 0.1953578318569535,
      "learning_rate": 1.8330484131137093e-06,
      "loss": 0.236,
      "step": 2135
    },
    {
      "epoch": 1.4965034965034965,
      "grad_norm": 0.18322617272387082,
      "learning_rate": 1.8094957735583463e-06,
      "loss": 0.2193,
      "step": 2140
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.19017938327788114,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.2198,
      "step": 2145
    },
    {
      "epoch": 1.5034965034965035,
      "grad_norm": 0.18598970620727007,
      "learning_rate": 1.7627478198395193e-06,
      "loss": 0.2493,
      "step": 2150
    },
    {
      "epoch": 1.506993006993007,
      "grad_norm": 0.19667037078430818,
      "learning_rate": 1.7395542466165488e-06,
      "loss": 0.2457,
      "step": 2155
    },
    {
      "epoch": 1.5104895104895104,
      "grad_norm": 0.18576871588399868,
      "learning_rate": 1.7164820956502237e-06,
      "loss": 0.237,
      "step": 2160
    },
    {
      "epoch": 1.513986013986014,
      "grad_norm": 0.21516455587634092,
      "learning_rate": 1.6935322261704929e-06,
      "loss": 0.2125,
      "step": 2165
    },
    {
      "epoch": 1.5174825174825175,
      "grad_norm": 0.15506871463618144,
      "learning_rate": 1.670705492853421e-06,
      "loss": 0.2254,
      "step": 2170
    },
    {
      "epoch": 1.5209790209790208,
      "grad_norm": 0.1881942062602629,
      "learning_rate": 1.6480027457893572e-06,
      "loss": 0.2135,
      "step": 2175
    },
    {
      "epoch": 1.5244755244755246,
      "grad_norm": 0.17951441787942998,
      "learning_rate": 1.6254248304512816e-06,
      "loss": 0.2285,
      "step": 2180
    },
    {
      "epoch": 1.527972027972028,
      "grad_norm": 0.20142927757336132,
      "learning_rate": 1.6029725876633085e-06,
      "loss": 0.2338,
      "step": 2185
    },
    {
      "epoch": 1.5314685314685315,
      "grad_norm": 0.21515360188845764,
      "learning_rate": 1.580646853569393e-06,
      "loss": 0.2135,
      "step": 2190
    },
    {
      "epoch": 1.534965034965035,
      "grad_norm": 0.18905919710908375,
      "learning_rate": 1.5584484596021694e-06,
      "loss": 0.2144,
      "step": 2195
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.19333144663540974,
      "learning_rate": 1.5363782324520033e-06,
      "loss": 0.2255,
      "step": 2200
    },
    {
      "epoch": 1.541958041958042,
      "grad_norm": 0.17838751714167966,
      "learning_rate": 1.514436994036201e-06,
      "loss": 0.2219,
      "step": 2205
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.21722837584706872,
      "learning_rate": 1.4926255614683931e-06,
      "loss": 0.2261,
      "step": 2210
    },
    {
      "epoch": 1.548951048951049,
      "grad_norm": 0.19249829848261585,
      "learning_rate": 1.4709447470281219e-06,
      "loss": 0.2296,
      "step": 2215
    },
    {
      "epoch": 1.5524475524475525,
      "grad_norm": 0.18312291697482874,
      "learning_rate": 1.4493953581305686e-06,
      "loss": 0.2084,
      "step": 2220
    },
    {
      "epoch": 1.5559440559440558,
      "grad_norm": 0.20306243754671757,
      "learning_rate": 1.427978197296503e-06,
      "loss": 0.2308,
      "step": 2225
    },
    {
      "epoch": 1.5594405594405596,
      "grad_norm": 0.1771655381322369,
      "learning_rate": 1.406694062122389e-06,
      "loss": 0.2143,
      "step": 2230
    },
    {
      "epoch": 1.562937062937063,
      "grad_norm": 0.2048290399421429,
      "learning_rate": 1.3855437452506759e-06,
      "loss": 0.2289,
      "step": 2235
    },
    {
      "epoch": 1.5664335664335665,
      "grad_norm": 0.2129953111928562,
      "learning_rate": 1.3645280343402973e-06,
      "loss": 0.2096,
      "step": 2240
    },
    {
      "epoch": 1.56993006993007,
      "grad_norm": 0.18573423410689366,
      "learning_rate": 1.3436477120373143e-06,
      "loss": 0.217,
      "step": 2245
    },
    {
      "epoch": 1.5734265734265733,
      "grad_norm": 0.19621542489979943,
      "learning_rate": 1.3229035559457887e-06,
      "loss": 0.2033,
      "step": 2250
    },
    {
      "epoch": 1.5769230769230769,
      "grad_norm": 0.24008695571611122,
      "learning_rate": 1.3022963385988153e-06,
      "loss": 0.1959,
      "step": 2255
    },
    {
      "epoch": 1.5804195804195804,
      "grad_norm": 0.21843850294705888,
      "learning_rate": 1.2818268274297474e-06,
      "loss": 0.2116,
      "step": 2260
    },
    {
      "epoch": 1.583916083916084,
      "grad_norm": 0.19460372652656613,
      "learning_rate": 1.2614957847436342e-06,
      "loss": 0.1974,
      "step": 2265
    },
    {
      "epoch": 1.5874125874125875,
      "grad_norm": 0.30934257030487106,
      "learning_rate": 1.2413039676888089e-06,
      "loss": 0.21,
      "step": 2270
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.22212188423905793,
      "learning_rate": 1.2212521282287093e-06,
      "loss": 0.2242,
      "step": 2275
    },
    {
      "epoch": 1.5944055944055944,
      "grad_norm": 0.20029609713926094,
      "learning_rate": 1.2013410131138658e-06,
      "loss": 0.2115,
      "step": 2280
    },
    {
      "epoch": 1.597902097902098,
      "grad_norm": 0.24287237218408447,
      "learning_rate": 1.1815713638540943e-06,
      "loss": 0.2128,
      "step": 2285
    },
    {
      "epoch": 1.6013986013986012,
      "grad_norm": 0.21091462289100518,
      "learning_rate": 1.161943916690883e-06,
      "loss": 0.2295,
      "step": 2290
    },
    {
      "epoch": 1.604895104895105,
      "grad_norm": 0.22381470851402516,
      "learning_rate": 1.1424594025699653e-06,
      "loss": 0.2175,
      "step": 2295
    },
    {
      "epoch": 1.6083916083916083,
      "grad_norm": 0.19979691673132668,
      "learning_rate": 1.1231185471141126e-06,
      "loss": 0.2042,
      "step": 2300
    },
    {
      "epoch": 1.6118881118881119,
      "grad_norm": 0.21513560665711312,
      "learning_rate": 1.1039220705961002e-06,
      "loss": 0.2021,
      "step": 2305
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.23305617978947493,
      "learning_rate": 1.0848706879118893e-06,
      "loss": 0.2127,
      "step": 2310
    },
    {
      "epoch": 1.6188811188811187,
      "grad_norm": 0.23244248721394226,
      "learning_rate": 1.065965108554e-06,
      "loss": 0.2299,
      "step": 2315
    },
    {
      "epoch": 1.6223776223776225,
      "grad_norm": 0.181467970622898,
      "learning_rate": 1.047206036585095e-06,
      "loss": 0.2009,
      "step": 2320
    },
    {
      "epoch": 1.6258741258741258,
      "grad_norm": 0.2635412802860612,
      "learning_rate": 1.0285941706117486e-06,
      "loss": 0.2209,
      "step": 2325
    },
    {
      "epoch": 1.6293706293706294,
      "grad_norm": 0.24616619041237905,
      "learning_rate": 1.0101302037584427e-06,
      "loss": 0.2183,
      "step": 2330
    },
    {
      "epoch": 1.632867132867133,
      "grad_norm": 0.20810334682703485,
      "learning_rate": 9.918148236417453e-07,
      "loss": 0.1809,
      "step": 2335
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.1672014242140621,
      "learning_rate": 9.73648712344707e-07,
      "loss": 0.1917,
      "step": 2340
    },
    {
      "epoch": 1.63986013986014,
      "grad_norm": 0.24131670715487236,
      "learning_rate": 9.5563254639146e-07,
      "loss": 0.1849,
      "step": 2345
    },
    {
      "epoch": 1.6433566433566433,
      "grad_norm": 0.22229836869112876,
      "learning_rate": 9.377669967220166e-07,
      "loss": 0.2098,
      "step": 2350
    },
    {
      "epoch": 1.6468531468531469,
      "grad_norm": 0.22611582762979465,
      "learning_rate": 9.200527286672967e-07,
      "loss": 0.2102,
      "step": 2355
    },
    {
      "epoch": 1.6503496503496504,
      "grad_norm": 0.23807228446885387,
      "learning_rate": 9.02490401924333e-07,
      "loss": 0.2077,
      "step": 2360
    },
    {
      "epoch": 1.6538461538461537,
      "grad_norm": 0.24914654872178474,
      "learning_rate": 8.850806705317183e-07,
      "loss": 0.2167,
      "step": 2365
    },
    {
      "epoch": 1.6573426573426573,
      "grad_norm": 0.25734981157347303,
      "learning_rate": 8.678241828452399e-07,
      "loss": 0.2101,
      "step": 2370
    },
    {
      "epoch": 1.6608391608391608,
      "grad_norm": 0.22641878526200002,
      "learning_rate": 8.507215815137337e-07,
      "loss": 0.2002,
      "step": 2375
    },
    {
      "epoch": 1.6643356643356644,
      "grad_norm": 0.2882342793841189,
      "learning_rate": 8.337735034551614e-07,
      "loss": 0.2191,
      "step": 2380
    },
    {
      "epoch": 1.667832167832168,
      "grad_norm": 0.2116647243381401,
      "learning_rate": 8.169805798328745e-07,
      "loss": 0.2121,
      "step": 2385
    },
    {
      "epoch": 1.6713286713286712,
      "grad_norm": 0.1989762007738208,
      "learning_rate": 8.003434360321249e-07,
      "loss": 0.2151,
      "step": 2390
    },
    {
      "epoch": 1.6748251748251748,
      "grad_norm": 0.19082703376727242,
      "learning_rate": 7.838626916367675e-07,
      "loss": 0.1894,
      "step": 2395
    },
    {
      "epoch": 1.6783216783216783,
      "grad_norm": 0.26560838065253123,
      "learning_rate": 7.675389604061817e-07,
      "loss": 0.2055,
      "step": 2400
    },
    {
      "epoch": 1.6818181818181817,
      "grad_norm": 0.23055679946733693,
      "learning_rate": 7.513728502524286e-07,
      "loss": 0.1938,
      "step": 2405
    },
    {
      "epoch": 1.6853146853146854,
      "grad_norm": 0.19893558756672688,
      "learning_rate": 7.353649632175957e-07,
      "loss": 0.1929,
      "step": 2410
    },
    {
      "epoch": 1.6888111888111887,
      "grad_norm": 0.21139143936318708,
      "learning_rate": 7.195158954513864e-07,
      "loss": 0.1969,
      "step": 2415
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 0.19819284147166077,
      "learning_rate": 7.03826237188916e-07,
      "loss": 0.1956,
      "step": 2420
    },
    {
      "epoch": 1.6958041958041958,
      "grad_norm": 0.24137592993550197,
      "learning_rate": 6.882965727287305e-07,
      "loss": 0.2004,
      "step": 2425
    },
    {
      "epoch": 1.6993006993006992,
      "grad_norm": 0.19697817486577932,
      "learning_rate": 6.72927480411047e-07,
      "loss": 0.1999,
      "step": 2430
    },
    {
      "epoch": 1.702797202797203,
      "grad_norm": 0.2835812827089056,
      "learning_rate": 6.577195325962138e-07,
      "loss": 0.189,
      "step": 2435
    },
    {
      "epoch": 1.7062937062937062,
      "grad_norm": 0.24980787748721664,
      "learning_rate": 6.426732956433989e-07,
      "loss": 0.2098,
      "step": 2440
    },
    {
      "epoch": 1.7097902097902098,
      "grad_norm": 0.19781206699502665,
      "learning_rate": 6.277893298894966e-07,
      "loss": 0.1944,
      "step": 2445
    },
    {
      "epoch": 1.7132867132867133,
      "grad_norm": 0.2224155467817442,
      "learning_rate": 6.130681896282581e-07,
      "loss": 0.216,
      "step": 2450
    },
    {
      "epoch": 1.7167832167832167,
      "grad_norm": 0.26034546579039136,
      "learning_rate": 5.985104230896515e-07,
      "loss": 0.2155,
      "step": 2455
    },
    {
      "epoch": 1.7202797202797204,
      "grad_norm": 0.2361138198171239,
      "learning_rate": 5.841165724194453e-07,
      "loss": 0.2003,
      "step": 2460
    },
    {
      "epoch": 1.7237762237762237,
      "grad_norm": 0.24067117775894944,
      "learning_rate": 5.698871736590145e-07,
      "loss": 0.2067,
      "step": 2465
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.20558184101041518,
      "learning_rate": 5.558227567253832e-07,
      "loss": 0.2059,
      "step": 2470
    },
    {
      "epoch": 1.7307692307692308,
      "grad_norm": 0.2221416526449957,
      "learning_rate": 5.41923845391486e-07,
      "loss": 0.1983,
      "step": 2475
    },
    {
      "epoch": 1.7342657342657342,
      "grad_norm": 0.24759902060603323,
      "learning_rate": 5.281909572666649e-07,
      "loss": 0.2042,
      "step": 2480
    },
    {
      "epoch": 1.737762237762238,
      "grad_norm": 0.23296199321438588,
      "learning_rate": 5.146246037773922e-07,
      "loss": 0.1925,
      "step": 2485
    },
    {
      "epoch": 1.7412587412587412,
      "grad_norm": 0.2171843285033186,
      "learning_rate": 5.012252901482211e-07,
      "loss": 0.1844,
      "step": 2490
    },
    {
      "epoch": 1.7447552447552448,
      "grad_norm": 0.27282372046975095,
      "learning_rate": 4.879935153829757e-07,
      "loss": 0.2155,
      "step": 2495
    },
    {
      "epoch": 1.7482517482517483,
      "grad_norm": 0.2676144982976565,
      "learning_rate": 4.74929772246166e-07,
      "loss": 0.1977,
      "step": 2500
    },
    {
      "epoch": 1.7517482517482517,
      "grad_norm": 0.2420189443106692,
      "learning_rate": 4.6203454724463523e-07,
      "loss": 0.1849,
      "step": 2505
    },
    {
      "epoch": 1.7552447552447552,
      "grad_norm": 0.2474376690863707,
      "learning_rate": 4.4930832060944453e-07,
      "loss": 0.2236,
      "step": 2510
    },
    {
      "epoch": 1.7587412587412588,
      "grad_norm": 0.250710313240247,
      "learning_rate": 4.36751566277982e-07,
      "loss": 0.2121,
      "step": 2515
    },
    {
      "epoch": 1.762237762237762,
      "grad_norm": 0.25444114255732747,
      "learning_rate": 4.243647518763261e-07,
      "loss": 0.2022,
      "step": 2520
    },
    {
      "epoch": 1.7657342657342658,
      "grad_norm": 0.20861908576529287,
      "learning_rate": 4.121483387018149e-07,
      "loss": 0.1767,
      "step": 2525
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.2706493720758507,
      "learning_rate": 4.001027817058789e-07,
      "loss": 0.1892,
      "step": 2530
    },
    {
      "epoch": 1.7727272727272727,
      "grad_norm": 0.24577453949394346,
      "learning_rate": 3.882285294770938e-07,
      "loss": 0.2023,
      "step": 2535
    },
    {
      "epoch": 1.7762237762237763,
      "grad_norm": 0.29431787774388873,
      "learning_rate": 3.765260242244706e-07,
      "loss": 0.2007,
      "step": 2540
    },
    {
      "epoch": 1.7797202797202796,
      "grad_norm": 0.193858764485361,
      "learning_rate": 3.649957017609979e-07,
      "loss": 0.2042,
      "step": 2545
    },
    {
      "epoch": 1.7832167832167833,
      "grad_norm": 0.2702741907418955,
      "learning_rate": 3.536379914873983e-07,
      "loss": 0.1886,
      "step": 2550
    },
    {
      "epoch": 1.7867132867132867,
      "grad_norm": 0.25145193456559023,
      "learning_rate": 3.4245331637614955e-07,
      "loss": 0.1962,
      "step": 2555
    },
    {
      "epoch": 1.7902097902097902,
      "grad_norm": 0.2501484856558932,
      "learning_rate": 3.3144209295572493e-07,
      "loss": 0.2054,
      "step": 2560
    },
    {
      "epoch": 1.7937062937062938,
      "grad_norm": 0.23954350984165065,
      "learning_rate": 3.206047312950811e-07,
      "loss": 0.1831,
      "step": 2565
    },
    {
      "epoch": 1.797202797202797,
      "grad_norm": 0.19874812191977084,
      "learning_rate": 3.099416349883949e-07,
      "loss": 0.1795,
      "step": 2570
    },
    {
      "epoch": 1.8006993006993008,
      "grad_norm": 0.22958298027049634,
      "learning_rate": 2.994532011400214e-07,
      "loss": 0.2006,
      "step": 2575
    },
    {
      "epoch": 1.8041958041958042,
      "grad_norm": 0.2441980250214929,
      "learning_rate": 2.8913982034971475e-07,
      "loss": 0.176,
      "step": 2580
    },
    {
      "epoch": 1.8076923076923077,
      "grad_norm": 0.279715766235606,
      "learning_rate": 2.790018766980773e-07,
      "loss": 0.2044,
      "step": 2585
    },
    {
      "epoch": 1.8111888111888113,
      "grad_norm": 0.21055408991164662,
      "learning_rate": 2.6903974773225703e-07,
      "loss": 0.202,
      "step": 2590
    },
    {
      "epoch": 1.8146853146853146,
      "grad_norm": 0.2479858008132291,
      "learning_rate": 2.592538044518889e-07,
      "loss": 0.1857,
      "step": 2595
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.24408864200471844,
      "learning_rate": 2.4964441129527337e-07,
      "loss": 0.2032,
      "step": 2600
    },
    {
      "epoch": 1.8216783216783217,
      "grad_norm": 0.20733367769835062,
      "learning_rate": 2.4021192612581177e-07,
      "loss": 0.2058,
      "step": 2605
    },
    {
      "epoch": 1.8251748251748252,
      "grad_norm": 0.30941756429405426,
      "learning_rate": 2.30956700218673e-07,
      "loss": 0.1843,
      "step": 2610
    },
    {
      "epoch": 1.8286713286713288,
      "grad_norm": 0.2372993874654309,
      "learning_rate": 2.2187907824771538e-07,
      "loss": 0.198,
      "step": 2615
    },
    {
      "epoch": 1.832167832167832,
      "grad_norm": 0.2543609821385058,
      "learning_rate": 2.1297939827264803e-07,
      "loss": 0.2078,
      "step": 2620
    },
    {
      "epoch": 1.8356643356643356,
      "grad_norm": 0.24490801803560427,
      "learning_rate": 2.0425799172644367e-07,
      "loss": 0.1736,
      "step": 2625
    },
    {
      "epoch": 1.8391608391608392,
      "grad_norm": 0.2268680251314891,
      "learning_rate": 1.9571518340299245e-07,
      "loss": 0.1882,
      "step": 2630
    },
    {
      "epoch": 1.8426573426573427,
      "grad_norm": 0.2930933494648304,
      "learning_rate": 1.873512914450104e-07,
      "loss": 0.1635,
      "step": 2635
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.24498583088228768,
      "learning_rate": 1.7916662733218848e-07,
      "loss": 0.1739,
      "step": 2640
    },
    {
      "epoch": 1.8496503496503496,
      "grad_norm": 0.2858319693898795,
      "learning_rate": 1.711614958695934e-07,
      "loss": 0.1835,
      "step": 2645
    },
    {
      "epoch": 1.8531468531468531,
      "grad_norm": 0.23269358304468626,
      "learning_rate": 1.6333619517631795e-07,
      "loss": 0.1972,
      "step": 2650
    },
    {
      "epoch": 1.8566433566433567,
      "grad_norm": 0.24825741471959645,
      "learning_rate": 1.5569101667437414e-07,
      "loss": 0.195,
      "step": 2655
    },
    {
      "epoch": 1.86013986013986,
      "grad_norm": 0.230066906532166,
      "learning_rate": 1.4822624507784932e-07,
      "loss": 0.1849,
      "step": 2660
    },
    {
      "epoch": 1.8636363636363638,
      "grad_norm": 0.23422929326553746,
      "learning_rate": 1.4094215838229176e-07,
      "loss": 0.1907,
      "step": 2665
    },
    {
      "epoch": 1.867132867132867,
      "grad_norm": 0.22983066527968804,
      "learning_rate": 1.3383902785436664e-07,
      "loss": 0.1837,
      "step": 2670
    },
    {
      "epoch": 1.8706293706293706,
      "grad_norm": 0.22073314464887625,
      "learning_rate": 1.2691711802175043e-07,
      "loss": 0.1705,
      "step": 2675
    },
    {
      "epoch": 1.8741258741258742,
      "grad_norm": 0.2668173429907813,
      "learning_rate": 1.2017668666327752e-07,
      "loss": 0.1889,
      "step": 2680
    },
    {
      "epoch": 1.8776223776223775,
      "grad_norm": 0.33800756354994693,
      "learning_rate": 1.1361798479934737e-07,
      "loss": 0.2034,
      "step": 2685
    },
    {
      "epoch": 1.8811188811188813,
      "grad_norm": 0.18294176538412874,
      "learning_rate": 1.0724125668256647e-07,
      "loss": 0.1692,
      "step": 2690
    },
    {
      "epoch": 1.8846153846153846,
      "grad_norm": 0.2289489292975833,
      "learning_rate": 1.0104673978866164e-07,
      "loss": 0.2062,
      "step": 2695
    },
    {
      "epoch": 1.8881118881118881,
      "grad_norm": 0.18359878620889158,
      "learning_rate": 9.50346648076278e-08,
      "loss": 0.1901,
      "step": 2700
    },
    {
      "epoch": 1.8916083916083917,
      "grad_norm": 0.2744536017506954,
      "learning_rate": 8.920525563514249e-08,
      "loss": 0.1922,
      "step": 2705
    },
    {
      "epoch": 1.895104895104895,
      "grad_norm": 0.2351372841256907,
      "learning_rate": 8.355872936422759e-08,
      "loss": 0.1875,
      "step": 2710
    },
    {
      "epoch": 1.8986013986013988,
      "grad_norm": 0.27882590800765195,
      "learning_rate": 7.809529627716028e-08,
      "loss": 0.2048,
      "step": 2715
    },
    {
      "epoch": 1.902097902097902,
      "grad_norm": 0.1987595110979819,
      "learning_rate": 7.281515983764698e-08,
      "loss": 0.1783,
      "step": 2720
    },
    {
      "epoch": 1.9055944055944056,
      "grad_norm": 0.2654329193657635,
      "learning_rate": 6.771851668324226e-08,
      "loss": 0.2158,
      "step": 2725
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.31410010413546313,
      "learning_rate": 6.280555661802857e-08,
      "loss": 0.1707,
      "step": 2730
    },
    {
      "epoch": 1.9125874125874125,
      "grad_norm": 0.30442270101870156,
      "learning_rate": 5.807646260554578e-08,
      "loss": 0.179,
      "step": 2735
    },
    {
      "epoch": 1.916083916083916,
      "grad_norm": 0.26428618741624893,
      "learning_rate": 5.353141076197888e-08,
      "loss": 0.181,
      "step": 2740
    },
    {
      "epoch": 1.9195804195804196,
      "grad_norm": 0.23072970375988636,
      "learning_rate": 4.9170570349598713e-08,
      "loss": 0.2098,
      "step": 2745
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.26605905211277536,
      "learning_rate": 4.499410377045765e-08,
      "loss": 0.1957,
      "step": 2750
    },
    {
      "epoch": 1.9265734265734267,
      "grad_norm": 0.2668770241834831,
      "learning_rate": 4.100216656034328e-08,
      "loss": 0.1941,
      "step": 2755
    },
    {
      "epoch": 1.93006993006993,
      "grad_norm": 0.2524573188005038,
      "learning_rate": 3.7194907382985256e-08,
      "loss": 0.1897,
      "step": 2760
    },
    {
      "epoch": 1.9335664335664335,
      "grad_norm": 0.2163808135197771,
      "learning_rate": 3.357246802451752e-08,
      "loss": 0.1905,
      "step": 2765
    },
    {
      "epoch": 1.937062937062937,
      "grad_norm": 0.26447973666575614,
      "learning_rate": 3.013498338820031e-08,
      "loss": 0.1689,
      "step": 2770
    },
    {
      "epoch": 1.9405594405594404,
      "grad_norm": 0.24986425291376463,
      "learning_rate": 2.6882581489395264e-08,
      "loss": 0.1806,
      "step": 2775
    },
    {
      "epoch": 1.9440559440559442,
      "grad_norm": 0.23106111194917012,
      "learning_rate": 2.381538345079759e-08,
      "loss": 0.1685,
      "step": 2780
    },
    {
      "epoch": 1.9475524475524475,
      "grad_norm": 0.23946489411883592,
      "learning_rate": 2.0933503497926886e-08,
      "loss": 0.2041,
      "step": 2785
    },
    {
      "epoch": 1.951048951048951,
      "grad_norm": 0.24084426183673927,
      "learning_rate": 1.8237048954869997e-08,
      "loss": 0.2032,
      "step": 2790
    },
    {
      "epoch": 1.9545454545454546,
      "grad_norm": 0.23026654450677114,
      "learning_rate": 1.5726120240288632e-08,
      "loss": 0.1955,
      "step": 2795
    },
    {
      "epoch": 1.958041958041958,
      "grad_norm": 0.23770880668006122,
      "learning_rate": 1.3400810863675174e-08,
      "loss": 0.1861,
      "step": 2800
    },
    {
      "epoch": 1.9615384615384617,
      "grad_norm": 0.24841763664984406,
      "learning_rate": 1.1261207421874309e-08,
      "loss": 0.1801,
      "step": 2805
    },
    {
      "epoch": 1.965034965034965,
      "grad_norm": 0.2279471977795885,
      "learning_rate": 9.307389595855088e-09,
      "loss": 0.1861,
      "step": 2810
    },
    {
      "epoch": 1.9685314685314685,
      "grad_norm": 0.20027941897627238,
      "learning_rate": 7.539430147745497e-09,
      "loss": 0.1872,
      "step": 2815
    },
    {
      "epoch": 1.972027972027972,
      "grad_norm": 0.22396064031327859,
      "learning_rate": 5.957394918120196e-09,
      "loss": 0.2027,
      "step": 2820
    },
    {
      "epoch": 1.9755244755244754,
      "grad_norm": 0.255023663918409,
      "learning_rate": 4.561342823553028e-09,
      "loss": 0.1863,
      "step": 2825
    },
    {
      "epoch": 1.9790209790209792,
      "grad_norm": 0.23827665650616262,
      "learning_rate": 3.351325854417109e-09,
      "loss": 0.2057,
      "step": 2830
    },
    {
      "epoch": 1.9825174825174825,
      "grad_norm": 0.24307636575035899,
      "learning_rate": 2.327389072953601e-09,
      "loss": 0.1915,
      "step": 2835
    },
    {
      "epoch": 1.986013986013986,
      "grad_norm": 0.2570377514482702,
      "learning_rate": 1.4895706115908293e-09,
      "loss": 0.1796,
      "step": 2840
    },
    {
      "epoch": 1.9895104895104896,
      "grad_norm": 0.2020392723559334,
      "learning_rate": 8.379016715254207e-10,
      "loss": 0.1829,
      "step": 2845
    },
    {
      "epoch": 1.993006993006993,
      "grad_norm": 0.23552175857949906,
      "learning_rate": 3.724065215593431e-10,
      "loss": 0.1901,
      "step": 2850
    },
    {
      "epoch": 1.9965034965034965,
      "grad_norm": 0.26888660196202563,
      "learning_rate": 9.310249719729491e-11,
      "loss": 0.1817,
      "step": 2855
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.23179960875620506,
      "learning_rate": 0.0,
      "loss": 0.1932,
      "step": 2860
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.21646016836166382,
      "eval_runtime": 32.4443,
      "eval_samples_per_second": 21.391,
      "eval_steps_per_second": 5.363,
      "step": 2860
    },
    {
      "epoch": 2.0,
      "step": 2860,
      "total_flos": 6.695403983227322e+18,
      "train_loss": 0.29170730747543017,
      "train_runtime": 19929.9263,
      "train_samples_per_second": 6.888,
      "train_steps_per_second": 0.144
    }
  ],
  "logging_steps": 5,
  "max_steps": 2860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.695403983227322e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}