{
  "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": 4.37381113540513,
      "learning_rate": 3.4965034965034967e-08,
      "loss": 2.0811,
      "step": 1
    },
    {
      "epoch": 0.0034965034965034965,
      "grad_norm": 4.252217899298764,
      "learning_rate": 1.7482517482517484e-07,
      "loss": 2.1063,
      "step": 5
    },
    {
      "epoch": 0.006993006993006993,
      "grad_norm": 4.404207861857175,
      "learning_rate": 3.496503496503497e-07,
      "loss": 2.1104,
      "step": 10
    },
    {
      "epoch": 0.01048951048951049,
      "grad_norm": 3.3737210251020002,
      "learning_rate": 5.244755244755246e-07,
      "loss": 1.9985,
      "step": 15
    },
    {
      "epoch": 0.013986013986013986,
      "grad_norm": 2.021686734185303,
      "learning_rate": 6.993006993006994e-07,
      "loss": 1.8118,
      "step": 20
    },
    {
      "epoch": 0.017482517482517484,
      "grad_norm": 2.0309163807573354,
      "learning_rate": 8.741258741258741e-07,
      "loss": 1.5367,
      "step": 25
    },
    {
      "epoch": 0.02097902097902098,
      "grad_norm": 2.0173956831253403,
      "learning_rate": 1.0489510489510491e-06,
      "loss": 1.2332,
      "step": 30
    },
    {
      "epoch": 0.024475524475524476,
      "grad_norm": 0.6292284380908406,
      "learning_rate": 1.2237762237762238e-06,
      "loss": 0.8794,
      "step": 35
    },
    {
      "epoch": 0.027972027972027972,
      "grad_norm": 0.5873925928700078,
      "learning_rate": 1.3986013986013987e-06,
      "loss": 0.7875,
      "step": 40
    },
    {
      "epoch": 0.03146853146853147,
      "grad_norm": 0.49151785822031857,
      "learning_rate": 1.5734265734265736e-06,
      "loss": 0.7022,
      "step": 45
    },
    {
      "epoch": 0.03496503496503497,
      "grad_norm": 0.40138283435220545,
      "learning_rate": 1.7482517482517483e-06,
      "loss": 0.6464,
      "step": 50
    },
    {
      "epoch": 0.038461538461538464,
      "grad_norm": 0.32203725166447594,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 0.5814,
      "step": 55
    },
    {
      "epoch": 0.04195804195804196,
      "grad_norm": 0.26055998901566146,
      "learning_rate": 2.0979020979020983e-06,
      "loss": 0.5595,
      "step": 60
    },
    {
      "epoch": 0.045454545454545456,
      "grad_norm": 0.2551008009829932,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.5608,
      "step": 65
    },
    {
      "epoch": 0.04895104895104895,
      "grad_norm": 0.23665791665183983,
      "learning_rate": 2.4475524475524477e-06,
      "loss": 0.5354,
      "step": 70
    },
    {
      "epoch": 0.05244755244755245,
      "grad_norm": 0.22961308429968375,
      "learning_rate": 2.6223776223776225e-06,
      "loss": 0.5105,
      "step": 75
    },
    {
      "epoch": 0.055944055944055944,
      "grad_norm": 0.2543666802898581,
      "learning_rate": 2.7972027972027974e-06,
      "loss": 0.4921,
      "step": 80
    },
    {
      "epoch": 0.05944055944055944,
      "grad_norm": 0.2660361645798006,
      "learning_rate": 2.972027972027972e-06,
      "loss": 0.5034,
      "step": 85
    },
    {
      "epoch": 0.06293706293706294,
      "grad_norm": 0.3014764103123489,
      "learning_rate": 3.1468531468531472e-06,
      "loss": 0.4792,
      "step": 90
    },
    {
      "epoch": 0.06643356643356643,
      "grad_norm": 0.27689109931667516,
      "learning_rate": 3.321678321678322e-06,
      "loss": 0.4603,
      "step": 95
    },
    {
      "epoch": 0.06993006993006994,
      "grad_norm": 0.2625100242281368,
      "learning_rate": 3.4965034965034966e-06,
      "loss": 0.4719,
      "step": 100
    },
    {
      "epoch": 0.07342657342657342,
      "grad_norm": 0.26217583226922436,
      "learning_rate": 3.6713286713286715e-06,
      "loss": 0.4474,
      "step": 105
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 0.20880050238815698,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.45,
      "step": 110
    },
    {
      "epoch": 0.08041958041958042,
      "grad_norm": 0.2607881388634329,
      "learning_rate": 4.020979020979021e-06,
      "loss": 0.4116,
      "step": 115
    },
    {
      "epoch": 0.08391608391608392,
      "grad_norm": 0.19276061867830058,
      "learning_rate": 4.195804195804197e-06,
      "loss": 0.4011,
      "step": 120
    },
    {
      "epoch": 0.08741258741258741,
      "grad_norm": 0.206872497416213,
      "learning_rate": 4.3706293706293715e-06,
      "loss": 0.4227,
      "step": 125
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 0.21844912873919067,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.4244,
      "step": 130
    },
    {
      "epoch": 0.0944055944055944,
      "grad_norm": 0.17251200513626644,
      "learning_rate": 4.72027972027972e-06,
      "loss": 0.4495,
      "step": 135
    },
    {
      "epoch": 0.0979020979020979,
      "grad_norm": 0.18127640733082095,
      "learning_rate": 4.895104895104895e-06,
      "loss": 0.4103,
      "step": 140
    },
    {
      "epoch": 0.10139860139860139,
      "grad_norm": 0.1622267706785198,
      "learning_rate": 5.06993006993007e-06,
      "loss": 0.4178,
      "step": 145
    },
    {
      "epoch": 0.1048951048951049,
      "grad_norm": 0.16311308748836312,
      "learning_rate": 5.244755244755245e-06,
      "loss": 0.4145,
      "step": 150
    },
    {
      "epoch": 0.10839160839160839,
      "grad_norm": 0.15870077394566784,
      "learning_rate": 5.41958041958042e-06,
      "loss": 0.3946,
      "step": 155
    },
    {
      "epoch": 0.11188811188811189,
      "grad_norm": 0.1575025763124978,
      "learning_rate": 5.594405594405595e-06,
      "loss": 0.4431,
      "step": 160
    },
    {
      "epoch": 0.11538461538461539,
      "grad_norm": 0.19464925796584603,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.4082,
      "step": 165
    },
    {
      "epoch": 0.11888111888111888,
      "grad_norm": 0.19779522561895393,
      "learning_rate": 5.944055944055944e-06,
      "loss": 0.3833,
      "step": 170
    },
    {
      "epoch": 0.12237762237762238,
      "grad_norm": 0.15172576025838877,
      "learning_rate": 6.1188811188811196e-06,
      "loss": 0.3827,
      "step": 175
    },
    {
      "epoch": 0.1258741258741259,
      "grad_norm": 0.14304151028864226,
      "learning_rate": 6.2937062937062944e-06,
      "loss": 0.3996,
      "step": 180
    },
    {
      "epoch": 0.12937062937062938,
      "grad_norm": 0.1397289499019953,
      "learning_rate": 6.468531468531469e-06,
      "loss": 0.4001,
      "step": 185
    },
    {
      "epoch": 0.13286713286713286,
      "grad_norm": 0.12356479142674442,
      "learning_rate": 6.643356643356644e-06,
      "loss": 0.3631,
      "step": 190
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 0.1456501947340643,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.399,
      "step": 195
    },
    {
      "epoch": 0.13986013986013987,
      "grad_norm": 0.1777441051019193,
      "learning_rate": 6.993006993006993e-06,
      "loss": 0.4013,
      "step": 200
    },
    {
      "epoch": 0.14335664335664336,
      "grad_norm": 0.15674206290465487,
      "learning_rate": 7.167832167832168e-06,
      "loss": 0.3846,
      "step": 205
    },
    {
      "epoch": 0.14685314685314685,
      "grad_norm": 0.14334403555842584,
      "learning_rate": 7.342657342657343e-06,
      "loss": 0.3978,
      "step": 210
    },
    {
      "epoch": 0.15034965034965034,
      "grad_norm": 0.12195727463513337,
      "learning_rate": 7.517482517482519e-06,
      "loss": 0.4025,
      "step": 215
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 0.13655485264527165,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.3907,
      "step": 220
    },
    {
      "epoch": 0.15734265734265734,
      "grad_norm": 0.1526742197214791,
      "learning_rate": 7.867132867132867e-06,
      "loss": 0.4056,
      "step": 225
    },
    {
      "epoch": 0.16083916083916083,
      "grad_norm": 0.19679006943755223,
      "learning_rate": 8.041958041958042e-06,
      "loss": 0.3913,
      "step": 230
    },
    {
      "epoch": 0.16433566433566432,
      "grad_norm": 0.1374898822284108,
      "learning_rate": 8.216783216783217e-06,
      "loss": 0.395,
      "step": 235
    },
    {
      "epoch": 0.16783216783216784,
      "grad_norm": 0.1316377878353779,
      "learning_rate": 8.391608391608393e-06,
      "loss": 0.3913,
      "step": 240
    },
    {
      "epoch": 0.17132867132867133,
      "grad_norm": 0.1360866720355628,
      "learning_rate": 8.566433566433568e-06,
      "loss": 0.389,
      "step": 245
    },
    {
      "epoch": 0.17482517482517482,
      "grad_norm": 0.12896458401156194,
      "learning_rate": 8.741258741258743e-06,
      "loss": 0.3948,
      "step": 250
    },
    {
      "epoch": 0.17832167832167833,
      "grad_norm": 0.13668138902696825,
      "learning_rate": 8.916083916083916e-06,
      "loss": 0.4162,
      "step": 255
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.11375569205265923,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.3583,
      "step": 260
    },
    {
      "epoch": 0.1853146853146853,
      "grad_norm": 0.13923337751226186,
      "learning_rate": 9.265734265734266e-06,
      "loss": 0.3679,
      "step": 265
    },
    {
      "epoch": 0.1888111888111888,
      "grad_norm": 0.12548306730535067,
      "learning_rate": 9.44055944055944e-06,
      "loss": 0.3809,
      "step": 270
    },
    {
      "epoch": 0.19230769230769232,
      "grad_norm": 0.12606188346945477,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.3971,
      "step": 275
    },
    {
      "epoch": 0.1958041958041958,
      "grad_norm": 0.12354250313388718,
      "learning_rate": 9.79020979020979e-06,
      "loss": 0.4049,
      "step": 280
    },
    {
      "epoch": 0.1993006993006993,
      "grad_norm": 0.13053765656410934,
      "learning_rate": 9.965034965034966e-06,
      "loss": 0.3845,
      "step": 285
    },
    {
      "epoch": 0.20279720279720279,
      "grad_norm": 0.11608842411220202,
      "learning_rate": 9.999940414335223e-06,
      "loss": 0.3743,
      "step": 290
    },
    {
      "epoch": 0.2062937062937063,
      "grad_norm": 0.13765081981314303,
      "learning_rate": 9.999698350006063e-06,
      "loss": 0.3708,
      "step": 295
    },
    {
      "epoch": 0.2097902097902098,
      "grad_norm": 0.12381330340903716,
      "learning_rate": 9.999270091916259e-06,
      "loss": 0.3813,
      "step": 300
    },
    {
      "epoch": 0.21328671328671328,
      "grad_norm": 0.11405435023477081,
      "learning_rate": 9.998655656014563e-06,
      "loss": 0.3962,
      "step": 305
    },
    {
      "epoch": 0.21678321678321677,
      "grad_norm": 0.12904825154350827,
      "learning_rate": 9.997855065183185e-06,
      "loss": 0.3669,
      "step": 310
    },
    {
      "epoch": 0.2202797202797203,
      "grad_norm": 0.11527332078152458,
      "learning_rate": 9.996868349236927e-06,
      "loss": 0.3585,
      "step": 315
    },
    {
      "epoch": 0.22377622377622378,
      "grad_norm": 0.11397307562641748,
      "learning_rate": 9.995695544922076e-06,
      "loss": 0.3722,
      "step": 320
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.12094683417044083,
      "learning_rate": 9.994336695915041e-06,
      "loss": 0.3746,
      "step": 325
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 0.11457644069457708,
      "learning_rate": 9.992791852820709e-06,
      "loss": 0.3755,
      "step": 330
    },
    {
      "epoch": 0.23426573426573427,
      "grad_norm": 0.10010657801032996,
      "learning_rate": 9.991061073170585e-06,
      "loss": 0.3655,
      "step": 335
    },
    {
      "epoch": 0.23776223776223776,
      "grad_norm": 0.11875607366309843,
      "learning_rate": 9.98914442142063e-06,
      "loss": 0.3567,
      "step": 340
    },
    {
      "epoch": 0.24125874125874125,
      "grad_norm": 0.12184450566051146,
      "learning_rate": 9.98704196894887e-06,
      "loss": 0.3843,
      "step": 345
    },
    {
      "epoch": 0.24475524475524477,
      "grad_norm": 0.11503255592868246,
      "learning_rate": 9.984753794052735e-06,
      "loss": 0.353,
      "step": 350
    },
    {
      "epoch": 0.24825174825174826,
      "grad_norm": 0.11359123753881316,
      "learning_rate": 9.982279981946143e-06,
      "loss": 0.3645,
      "step": 355
    },
    {
      "epoch": 0.2517482517482518,
      "grad_norm": 0.12099169961130332,
      "learning_rate": 9.97962062475633e-06,
      "loss": 0.3854,
      "step": 360
    },
    {
      "epoch": 0.25524475524475526,
      "grad_norm": 0.13217619060267752,
      "learning_rate": 9.976775821520412e-06,
      "loss": 0.3721,
      "step": 365
    },
    {
      "epoch": 0.25874125874125875,
      "grad_norm": 0.09612633521304394,
      "learning_rate": 9.973745678181705e-06,
      "loss": 0.3457,
      "step": 370
    },
    {
      "epoch": 0.26223776223776224,
      "grad_norm": 0.10517669260119224,
      "learning_rate": 9.970530307585774e-06,
      "loss": 0.3565,
      "step": 375
    },
    {
      "epoch": 0.26573426573426573,
      "grad_norm": 0.11544943707626291,
      "learning_rate": 9.967129829476227e-06,
      "loss": 0.4004,
      "step": 380
    },
    {
      "epoch": 0.2692307692307692,
      "grad_norm": 0.11395353055045045,
      "learning_rate": 9.96354437049027e-06,
      "loss": 0.3604,
      "step": 385
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.1132567024291244,
      "learning_rate": 9.959774064153977e-06,
      "loss": 0.3779,
      "step": 390
    },
    {
      "epoch": 0.2762237762237762,
      "grad_norm": 0.09873707074062922,
      "learning_rate": 9.955819050877321e-06,
      "loss": 0.3567,
      "step": 395
    },
    {
      "epoch": 0.27972027972027974,
      "grad_norm": 0.10782299582051598,
      "learning_rate": 9.951679477948946e-06,
      "loss": 0.3864,
      "step": 400
    },
    {
      "epoch": 0.28321678321678323,
      "grad_norm": 0.11261432971654586,
      "learning_rate": 9.947355499530685e-06,
      "loss": 0.3599,
      "step": 405
    },
    {
      "epoch": 0.2867132867132867,
      "grad_norm": 0.12074994776023469,
      "learning_rate": 9.942847276651812e-06,
      "loss": 0.3284,
      "step": 410
    },
    {
      "epoch": 0.2902097902097902,
      "grad_norm": 0.09741746253500448,
      "learning_rate": 9.93815497720305e-06,
      "loss": 0.3474,
      "step": 415
    },
    {
      "epoch": 0.2937062937062937,
      "grad_norm": 0.11397076083204073,
      "learning_rate": 9.933278775930317e-06,
      "loss": 0.3556,
      "step": 420
    },
    {
      "epoch": 0.2972027972027972,
      "grad_norm": 0.11555465535329675,
      "learning_rate": 9.928218854428223e-06,
      "loss": 0.3611,
      "step": 425
    },
    {
      "epoch": 0.3006993006993007,
      "grad_norm": 0.11398129108565634,
      "learning_rate": 9.922975401133292e-06,
      "loss": 0.3666,
      "step": 430
    },
    {
      "epoch": 0.3041958041958042,
      "grad_norm": 0.1047587143979841,
      "learning_rate": 9.917548611316969e-06,
      "loss": 0.3255,
      "step": 435
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.09926317278084681,
      "learning_rate": 9.911938687078324e-06,
      "loss": 0.3533,
      "step": 440
    },
    {
      "epoch": 0.3111888111888112,
      "grad_norm": 0.09642382922146973,
      "learning_rate": 9.90614583733654e-06,
      "loss": 0.3468,
      "step": 445
    },
    {
      "epoch": 0.3146853146853147,
      "grad_norm": 0.10269176659026831,
      "learning_rate": 9.900170277823129e-06,
      "loss": 0.3487,
      "step": 450
    },
    {
      "epoch": 0.3181818181818182,
      "grad_norm": 0.10195209542391877,
      "learning_rate": 9.894012231073895e-06,
      "loss": 0.3615,
      "step": 455
    },
    {
      "epoch": 0.32167832167832167,
      "grad_norm": 0.10911945373160642,
      "learning_rate": 9.887671926420649e-06,
      "loss": 0.3624,
      "step": 460
    },
    {
      "epoch": 0.32517482517482516,
      "grad_norm": 0.11633743875381607,
      "learning_rate": 9.881149599982671e-06,
      "loss": 0.3736,
      "step": 465
    },
    {
      "epoch": 0.32867132867132864,
      "grad_norm": 0.0919962699070435,
      "learning_rate": 9.874445494657912e-06,
      "loss": 0.3364,
      "step": 470
    },
    {
      "epoch": 0.3321678321678322,
      "grad_norm": 0.1169385489960078,
      "learning_rate": 9.86755986011395e-06,
      "loss": 0.3585,
      "step": 475
    },
    {
      "epoch": 0.3356643356643357,
      "grad_norm": 0.11072840564933388,
      "learning_rate": 9.860492952778695e-06,
      "loss": 0.3441,
      "step": 480
    },
    {
      "epoch": 0.33916083916083917,
      "grad_norm": 0.09838108948679272,
      "learning_rate": 9.853245035830834e-06,
      "loss": 0.3637,
      "step": 485
    },
    {
      "epoch": 0.34265734265734266,
      "grad_norm": 0.09612232855439369,
      "learning_rate": 9.845816379190037e-06,
      "loss": 0.3473,
      "step": 490
    },
    {
      "epoch": 0.34615384615384615,
      "grad_norm": 0.09998498487369935,
      "learning_rate": 9.838207259506891e-06,
      "loss": 0.3551,
      "step": 495
    },
    {
      "epoch": 0.34965034965034963,
      "grad_norm": 0.10258972737213708,
      "learning_rate": 9.83041796015262e-06,
      "loss": 0.3796,
      "step": 500
    },
    {
      "epoch": 0.3531468531468531,
      "grad_norm": 0.09995067780945936,
      "learning_rate": 9.82244877120851e-06,
      "loss": 0.3538,
      "step": 505
    },
    {
      "epoch": 0.35664335664335667,
      "grad_norm": 0.10479687563865843,
      "learning_rate": 9.814299989455118e-06,
      "loss": 0.3782,
      "step": 510
    },
    {
      "epoch": 0.36013986013986016,
      "grad_norm": 0.11535884576394473,
      "learning_rate": 9.805971918361215e-06,
      "loss": 0.3874,
      "step": 515
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.09319987111020978,
      "learning_rate": 9.797464868072489e-06,
      "loss": 0.3216,
      "step": 520
    },
    {
      "epoch": 0.36713286713286714,
      "grad_norm": 0.11218854149013563,
      "learning_rate": 9.788779155399988e-06,
      "loss": 0.3519,
      "step": 525
    },
    {
      "epoch": 0.3706293706293706,
      "grad_norm": 0.10463389683892751,
      "learning_rate": 9.779915103808328e-06,
      "loss": 0.3487,
      "step": 530
    },
    {
      "epoch": 0.3741258741258741,
      "grad_norm": 0.10312783064956889,
      "learning_rate": 9.770873043403648e-06,
      "loss": 0.3779,
      "step": 535
    },
    {
      "epoch": 0.3776223776223776,
      "grad_norm": 0.09457645559277009,
      "learning_rate": 9.761653310921307e-06,
      "loss": 0.349,
      "step": 540
    },
    {
      "epoch": 0.3811188811188811,
      "grad_norm": 0.10044623941110373,
      "learning_rate": 9.752256249713352e-06,
      "loss": 0.344,
      "step": 545
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.10269026724616583,
      "learning_rate": 9.742682209735727e-06,
      "loss": 0.3409,
      "step": 550
    },
    {
      "epoch": 0.3881118881118881,
      "grad_norm": 0.10855761396744447,
      "learning_rate": 9.73293154753525e-06,
      "loss": 0.3418,
      "step": 555
    },
    {
      "epoch": 0.3916083916083916,
      "grad_norm": 0.11951416935360595,
      "learning_rate": 9.723004626236314e-06,
      "loss": 0.3754,
      "step": 560
    },
    {
      "epoch": 0.3951048951048951,
      "grad_norm": 0.09145664665291538,
      "learning_rate": 9.712901815527387e-06,
      "loss": 0.3667,
      "step": 565
    },
    {
      "epoch": 0.3986013986013986,
      "grad_norm": 0.09322645373153875,
      "learning_rate": 9.702623491647232e-06,
      "loss": 0.3578,
      "step": 570
    },
    {
      "epoch": 0.4020979020979021,
      "grad_norm": 0.09436496121677602,
      "learning_rate": 9.692170037370899e-06,
      "loss": 0.36,
      "step": 575
    },
    {
      "epoch": 0.40559440559440557,
      "grad_norm": 0.08980733979197301,
      "learning_rate": 9.68154184199546e-06,
      "loss": 0.3339,
      "step": 580
    },
    {
      "epoch": 0.4090909090909091,
      "grad_norm": 0.10968764839801004,
      "learning_rate": 9.670739301325534e-06,
      "loss": 0.371,
      "step": 585
    },
    {
      "epoch": 0.4125874125874126,
      "grad_norm": 0.10841544769993913,
      "learning_rate": 9.659762817658524e-06,
      "loss": 0.3561,
      "step": 590
    },
    {
      "epoch": 0.4160839160839161,
      "grad_norm": 0.11208125619377086,
      "learning_rate": 9.648612799769644e-06,
      "loss": 0.3665,
      "step": 595
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 0.09710993092315441,
      "learning_rate": 9.6372896628967e-06,
      "loss": 0.3487,
      "step": 600
    },
    {
      "epoch": 0.4230769230769231,
      "grad_norm": 0.08611220115875796,
      "learning_rate": 9.62579382872462e-06,
      "loss": 0.3149,
      "step": 605
    },
    {
      "epoch": 0.42657342657342656,
      "grad_norm": 0.09422154430787315,
      "learning_rate": 9.614125725369748e-06,
      "loss": 0.367,
      "step": 610
    },
    {
      "epoch": 0.43006993006993005,
      "grad_norm": 0.10419775648300175,
      "learning_rate": 9.60228578736391e-06,
      "loss": 0.3339,
      "step": 615
    },
    {
      "epoch": 0.43356643356643354,
      "grad_norm": 0.09182548432657049,
      "learning_rate": 9.590274455638225e-06,
      "loss": 0.3599,
      "step": 620
    },
    {
      "epoch": 0.4370629370629371,
      "grad_norm": 0.09732070144300518,
      "learning_rate": 9.578092177506683e-06,
      "loss": 0.3677,
      "step": 625
    },
    {
      "epoch": 0.4405594405594406,
      "grad_norm": 0.10356238281777523,
      "learning_rate": 9.565739406649492e-06,
      "loss": 0.3603,
      "step": 630
    },
    {
      "epoch": 0.44405594405594406,
      "grad_norm": 0.09542634755210361,
      "learning_rate": 9.553216603096175e-06,
      "loss": 0.3368,
      "step": 635
    },
    {
      "epoch": 0.44755244755244755,
      "grad_norm": 0.10330899480093415,
      "learning_rate": 9.540524233208449e-06,
      "loss": 0.3809,
      "step": 640
    },
    {
      "epoch": 0.45104895104895104,
      "grad_norm": 0.0990570314447027,
      "learning_rate": 9.527662769662842e-06,
      "loss": 0.3321,
      "step": 645
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.098605126175185,
      "learning_rate": 9.514632691433108e-06,
      "loss": 0.3362,
      "step": 650
    },
    {
      "epoch": 0.458041958041958,
      "grad_norm": 0.11280604058910865,
      "learning_rate": 9.501434483772371e-06,
      "loss": 0.3502,
      "step": 655
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.10690820736424175,
      "learning_rate": 9.488068638195072e-06,
      "loss": 0.3792,
      "step": 660
    },
    {
      "epoch": 0.46503496503496505,
      "grad_norm": 0.10685361016021448,
      "learning_rate": 9.474535652458647e-06,
      "loss": 0.3372,
      "step": 665
    },
    {
      "epoch": 0.46853146853146854,
      "grad_norm": 0.091863754283505,
      "learning_rate": 9.460836030545007e-06,
      "loss": 0.3279,
      "step": 670
    },
    {
      "epoch": 0.47202797202797203,
      "grad_norm": 0.10137019023149173,
      "learning_rate": 9.446970282641754e-06,
      "loss": 0.3326,
      "step": 675
    },
    {
      "epoch": 0.4755244755244755,
      "grad_norm": 0.09644694864886368,
      "learning_rate": 9.43293892512319e-06,
      "loss": 0.3238,
      "step": 680
    },
    {
      "epoch": 0.479020979020979,
      "grad_norm": 0.09447255461234934,
      "learning_rate": 9.418742480531086e-06,
      "loss": 0.3482,
      "step": 685
    },
    {
      "epoch": 0.4825174825174825,
      "grad_norm": 0.08135682687879796,
      "learning_rate": 9.404381477555216e-06,
      "loss": 0.3275,
      "step": 690
    },
    {
      "epoch": 0.486013986013986,
      "grad_norm": 0.10069955308963059,
      "learning_rate": 9.38985645101368e-06,
      "loss": 0.3332,
      "step": 695
    },
    {
      "epoch": 0.48951048951048953,
      "grad_norm": 0.09628893677315106,
      "learning_rate": 9.375167941832974e-06,
      "loss": 0.344,
      "step": 700
    },
    {
      "epoch": 0.493006993006993,
      "grad_norm": 0.08809106076365064,
      "learning_rate": 9.360316497027849e-06,
      "loss": 0.3376,
      "step": 705
    },
    {
      "epoch": 0.4965034965034965,
      "grad_norm": 0.09490092152704185,
      "learning_rate": 9.345302669680947e-06,
      "loss": 0.3419,
      "step": 710
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10279282225217452,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.3574,
      "step": 715
    },
    {
      "epoch": 0.5034965034965035,
      "grad_norm": 0.10554593393741951,
      "learning_rate": 9.314790109907983e-06,
      "loss": 0.3372,
      "step": 720
    },
    {
      "epoch": 0.506993006993007,
      "grad_norm": 0.11231474223944753,
      "learning_rate": 9.299292513800129e-06,
      "loss": 0.3419,
      "step": 725
    },
    {
      "epoch": 0.5104895104895105,
      "grad_norm": 0.09498236350352282,
      "learning_rate": 9.283634807744586e-06,
      "loss": 0.3171,
      "step": 730
    },
    {
      "epoch": 0.513986013986014,
      "grad_norm": 0.09593334219573614,
      "learning_rate": 9.267817574849971e-06,
      "loss": 0.3485,
      "step": 735
    },
    {
      "epoch": 0.5174825174825175,
      "grad_norm": 0.09742681439006996,
      "learning_rate": 9.251841404165835e-06,
      "loss": 0.3378,
      "step": 740
    },
    {
      "epoch": 0.5209790209790209,
      "grad_norm": 0.0983937674421585,
      "learning_rate": 9.235706890660735e-06,
      "loss": 0.3493,
      "step": 745
    },
    {
      "epoch": 0.5244755244755245,
      "grad_norm": 0.0915357872203584,
      "learning_rate": 9.219414635200065e-06,
      "loss": 0.3417,
      "step": 750
    },
    {
      "epoch": 0.527972027972028,
      "grad_norm": 0.09146853221285936,
      "learning_rate": 9.202965244523696e-06,
      "loss": 0.3414,
      "step": 755
    },
    {
      "epoch": 0.5314685314685315,
      "grad_norm": 0.09486109366945525,
      "learning_rate": 9.18635933122337e-06,
      "loss": 0.3777,
      "step": 760
    },
    {
      "epoch": 0.534965034965035,
      "grad_norm": 0.09445667455904146,
      "learning_rate": 9.169597513719881e-06,
      "loss": 0.3517,
      "step": 765
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 0.10056106300174576,
      "learning_rate": 9.152680416240059e-06,
      "loss": 0.3329,
      "step": 770
    },
    {
      "epoch": 0.541958041958042,
      "grad_norm": 0.09747043642900897,
      "learning_rate": 9.135608668793511e-06,
      "loss": 0.3577,
      "step": 775
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.1098420896663471,
      "learning_rate": 9.118382907149164e-06,
      "loss": 0.3471,
      "step": 780
    },
    {
      "epoch": 0.548951048951049,
      "grad_norm": 0.10101868252033282,
      "learning_rate": 9.10100377281159e-06,
      "loss": 0.352,
      "step": 785
    },
    {
      "epoch": 0.5524475524475524,
      "grad_norm": 0.09509262269603147,
      "learning_rate": 9.08347191299711e-06,
      "loss": 0.3453,
      "step": 790
    },
    {
      "epoch": 0.5559440559440559,
      "grad_norm": 0.10223863291064357,
      "learning_rate": 9.065787980609696e-06,
      "loss": 0.3323,
      "step": 795
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 0.09307831695448808,
      "learning_rate": 9.047952634216652e-06,
      "loss": 0.3189,
      "step": 800
    },
    {
      "epoch": 0.5629370629370629,
      "grad_norm": 0.09664775207053762,
      "learning_rate": 9.029966538024097e-06,
      "loss": 0.3335,
      "step": 805
    },
    {
      "epoch": 0.5664335664335665,
      "grad_norm": 0.10330830247866737,
      "learning_rate": 9.011830361852217e-06,
      "loss": 0.298,
      "step": 810
    },
    {
      "epoch": 0.5699300699300699,
      "grad_norm": 0.09238529639878745,
      "learning_rate": 8.993544781110328e-06,
      "loss": 0.3515,
      "step": 815
    },
    {
      "epoch": 0.5734265734265734,
      "grad_norm": 0.09946980186720472,
      "learning_rate": 8.975110476771724e-06,
      "loss": 0.3325,
      "step": 820
    },
    {
      "epoch": 0.5769230769230769,
      "grad_norm": 0.10461426738254295,
      "learning_rate": 8.95652813534831e-06,
      "loss": 0.3461,
      "step": 825
    },
    {
      "epoch": 0.5804195804195804,
      "grad_norm": 0.10521038266819552,
      "learning_rate": 8.937798448865043e-06,
      "loss": 0.3509,
      "step": 830
    },
    {
      "epoch": 0.583916083916084,
      "grad_norm": 0.09296107569017567,
      "learning_rate": 8.918922114834156e-06,
      "loss": 0.3591,
      "step": 835
    },
    {
      "epoch": 0.5874125874125874,
      "grad_norm": 0.09978737012221923,
      "learning_rate": 8.899899836229186e-06,
      "loss": 0.3494,
      "step": 840
    },
    {
      "epoch": 0.5909090909090909,
      "grad_norm": 0.09887611819542517,
      "learning_rate": 8.880732321458785e-06,
      "loss": 0.3474,
      "step": 845
    },
    {
      "epoch": 0.5944055944055944,
      "grad_norm": 0.09194493203548675,
      "learning_rate": 8.861420284340352e-06,
      "loss": 0.3287,
      "step": 850
    },
    {
      "epoch": 0.5979020979020979,
      "grad_norm": 0.10774254756172354,
      "learning_rate": 8.841964444073437e-06,
      "loss": 0.3261,
      "step": 855
    },
    {
      "epoch": 0.6013986013986014,
      "grad_norm": 0.08277889114743786,
      "learning_rate": 8.822365525212968e-06,
      "loss": 0.3298,
      "step": 860
    },
    {
      "epoch": 0.6048951048951049,
      "grad_norm": 0.09403899388431841,
      "learning_rate": 8.802624257642262e-06,
      "loss": 0.3239,
      "step": 865
    },
    {
      "epoch": 0.6083916083916084,
      "grad_norm": 0.09869839895903458,
      "learning_rate": 8.782741376545837e-06,
      "loss": 0.3271,
      "step": 870
    },
    {
      "epoch": 0.6118881118881119,
      "grad_norm": 0.08742562281618353,
      "learning_rate": 8.76271762238205e-06,
      "loss": 0.3222,
      "step": 875
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.10669936980092758,
      "learning_rate": 8.742553740855507e-06,
      "loss": 0.3343,
      "step": 880
    },
    {
      "epoch": 0.6188811188811189,
      "grad_norm": 0.09705466668865588,
      "learning_rate": 8.722250482889293e-06,
      "loss": 0.3365,
      "step": 885
    },
    {
      "epoch": 0.6223776223776224,
      "grad_norm": 0.10026304811604293,
      "learning_rate": 8.701808604597022e-06,
      "loss": 0.3268,
      "step": 890
    },
    {
      "epoch": 0.6258741258741258,
      "grad_norm": 0.09499836148950658,
      "learning_rate": 8.681228867254657e-06,
      "loss": 0.3205,
      "step": 895
    },
    {
      "epoch": 0.6293706293706294,
      "grad_norm": 0.09620440611439918,
      "learning_rate": 8.660512037272173e-06,
      "loss": 0.3377,
      "step": 900
    },
    {
      "epoch": 0.6328671328671329,
      "grad_norm": 0.08591400161657516,
      "learning_rate": 8.639658886165012e-06,
      "loss": 0.3312,
      "step": 905
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.09176587951531126,
      "learning_rate": 8.61867019052535e-06,
      "loss": 0.3288,
      "step": 910
    },
    {
      "epoch": 0.6398601398601399,
      "grad_norm": 0.09488106076865752,
      "learning_rate": 8.597546731993182e-06,
      "loss": 0.3386,
      "step": 915
    },
    {
      "epoch": 0.6433566433566433,
      "grad_norm": 0.08369635401062202,
      "learning_rate": 8.576289297227201e-06,
      "loss": 0.3054,
      "step": 920
    },
    {
      "epoch": 0.6468531468531469,
      "grad_norm": 0.09730501351801213,
      "learning_rate": 8.554898677875509e-06,
      "loss": 0.3227,
      "step": 925
    },
    {
      "epoch": 0.6503496503496503,
      "grad_norm": 0.1001056960078254,
      "learning_rate": 8.533375670546142e-06,
      "loss": 0.3492,
      "step": 930
    },
    {
      "epoch": 0.6538461538461539,
      "grad_norm": 0.08957014382624598,
      "learning_rate": 8.511721076777388e-06,
      "loss": 0.3545,
      "step": 935
    },
    {
      "epoch": 0.6573426573426573,
      "grad_norm": 0.09078278112834585,
      "learning_rate": 8.48993570300795e-06,
      "loss": 0.3284,
      "step": 940
    },
    {
      "epoch": 0.6608391608391608,
      "grad_norm": 0.08656855201647384,
      "learning_rate": 8.46802036054691e-06,
      "loss": 0.3342,
      "step": 945
    },
    {
      "epoch": 0.6643356643356644,
      "grad_norm": 0.09414987082795111,
      "learning_rate": 8.445975865543509e-06,
      "loss": 0.3362,
      "step": 950
    },
    {
      "epoch": 0.6678321678321678,
      "grad_norm": 0.10356006952243683,
      "learning_rate": 8.423803038956763e-06,
      "loss": 0.3452,
      "step": 955
    },
    {
      "epoch": 0.6713286713286714,
      "grad_norm": 0.09745963990707376,
      "learning_rate": 8.401502706524884e-06,
      "loss": 0.3245,
      "step": 960
    },
    {
      "epoch": 0.6748251748251748,
      "grad_norm": 0.12682185311194272,
      "learning_rate": 8.379075698734523e-06,
      "loss": 0.3195,
      "step": 965
    },
    {
      "epoch": 0.6783216783216783,
      "grad_norm": 0.10273803895554465,
      "learning_rate": 8.356522850789852e-06,
      "loss": 0.3183,
      "step": 970
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.08936523805510559,
      "learning_rate": 8.33384500258146e-06,
      "loss": 0.3094,
      "step": 975
    },
    {
      "epoch": 0.6853146853146853,
      "grad_norm": 0.09076443538007455,
      "learning_rate": 8.31104299865506e-06,
      "loss": 0.3209,
      "step": 980
    },
    {
      "epoch": 0.6888111888111889,
      "grad_norm": 0.10360319946526882,
      "learning_rate": 8.288117688180064e-06,
      "loss": 0.3393,
      "step": 985
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.09990142074597995,
      "learning_rate": 8.265069924917925e-06,
      "loss": 0.3143,
      "step": 990
    },
    {
      "epoch": 0.6958041958041958,
      "grad_norm": 0.09863731703774027,
      "learning_rate": 8.241900567190376e-06,
      "loss": 0.3318,
      "step": 995
    },
    {
      "epoch": 0.6993006993006993,
      "grad_norm": 0.0891504924468397,
      "learning_rate": 8.218610477847435e-06,
      "loss": 0.3469,
      "step": 1000
    },
    {
      "epoch": 0.7027972027972028,
      "grad_norm": 0.0903949668682383,
      "learning_rate": 8.1952005242353e-06,
      "loss": 0.3084,
      "step": 1005
    },
    {
      "epoch": 0.7062937062937062,
      "grad_norm": 0.0954433441080251,
      "learning_rate": 8.171671578164023e-06,
      "loss": 0.3315,
      "step": 1010
    },
    {
      "epoch": 0.7097902097902098,
      "grad_norm": 0.08559449666366757,
      "learning_rate": 8.148024515875057e-06,
      "loss": 0.3446,
      "step": 1015
    },
    {
      "epoch": 0.7132867132867133,
      "grad_norm": 0.09175217509532477,
      "learning_rate": 8.124260218008627e-06,
      "loss": 0.3376,
      "step": 1020
    },
    {
      "epoch": 0.7167832167832168,
      "grad_norm": 0.08175979271190216,
      "learning_rate": 8.100379569570919e-06,
      "loss": 0.3018,
      "step": 1025
    },
    {
      "epoch": 0.7202797202797203,
      "grad_norm": 0.10086175825188673,
      "learning_rate": 8.076383459901138e-06,
      "loss": 0.3602,
      "step": 1030
    },
    {
      "epoch": 0.7237762237762237,
      "grad_norm": 0.10549585300259735,
      "learning_rate": 8.052272782638375e-06,
      "loss": 0.3685,
      "step": 1035
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.11673660774686632,
      "learning_rate": 8.028048435688333e-06,
      "loss": 0.3421,
      "step": 1040
    },
    {
      "epoch": 0.7307692307692307,
      "grad_norm": 0.09147779127352527,
      "learning_rate": 8.003711321189895e-06,
      "loss": 0.3272,
      "step": 1045
    },
    {
      "epoch": 0.7342657342657343,
      "grad_norm": 0.095327743964242,
      "learning_rate": 7.979262345481511e-06,
      "loss": 0.326,
      "step": 1050
    },
    {
      "epoch": 0.7377622377622378,
      "grad_norm": 0.09278699522643637,
      "learning_rate": 7.95470241906746e-06,
      "loss": 0.3277,
      "step": 1055
    },
    {
      "epoch": 0.7412587412587412,
      "grad_norm": 0.09014483968822934,
      "learning_rate": 7.930032456583931e-06,
      "loss": 0.301,
      "step": 1060
    },
    {
      "epoch": 0.7447552447552448,
      "grad_norm": 0.10302938884427425,
      "learning_rate": 7.905253376764973e-06,
      "loss": 0.3294,
      "step": 1065
    },
    {
      "epoch": 0.7482517482517482,
      "grad_norm": 0.09555338249314554,
      "learning_rate": 7.880366102408266e-06,
      "loss": 0.3144,
      "step": 1070
    },
    {
      "epoch": 0.7517482517482518,
      "grad_norm": 0.09838205058263788,
      "learning_rate": 7.85537156034077e-06,
      "loss": 0.3116,
      "step": 1075
    },
    {
      "epoch": 0.7552447552447552,
      "grad_norm": 0.0902348464066785,
      "learning_rate": 7.830270681384196e-06,
      "loss": 0.3273,
      "step": 1080
    },
    {
      "epoch": 0.7587412587412588,
      "grad_norm": 0.10197904027133771,
      "learning_rate": 7.805064400320348e-06,
      "loss": 0.337,
      "step": 1085
    },
    {
      "epoch": 0.7622377622377622,
      "grad_norm": 0.09136556755535169,
      "learning_rate": 7.779753655856313e-06,
      "loss": 0.3146,
      "step": 1090
    },
    {
      "epoch": 0.7657342657342657,
      "grad_norm": 0.10390395983392695,
      "learning_rate": 7.754339390589497e-06,
      "loss": 0.3455,
      "step": 1095
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.08873006190996087,
      "learning_rate": 7.728822550972523e-06,
      "loss": 0.3189,
      "step": 1100
    },
    {
      "epoch": 0.7727272727272727,
      "grad_norm": 0.10767216954218523,
      "learning_rate": 7.703204087277989e-06,
      "loss": 0.335,
      "step": 1105
    },
    {
      "epoch": 0.7762237762237763,
      "grad_norm": 0.09226028528791834,
      "learning_rate": 7.67748495356307e-06,
      "loss": 0.309,
      "step": 1110
    },
    {
      "epoch": 0.7797202797202797,
      "grad_norm": 0.09501355043720337,
      "learning_rate": 7.651666107633996e-06,
      "loss": 0.3131,
      "step": 1115
    },
    {
      "epoch": 0.7832167832167832,
      "grad_norm": 0.10030664866053496,
      "learning_rate": 7.625748511010382e-06,
      "loss": 0.3368,
      "step": 1120
    },
    {
      "epoch": 0.7867132867132867,
      "grad_norm": 0.0960730084438704,
      "learning_rate": 7.599733128889413e-06,
      "loss": 0.311,
      "step": 1125
    },
    {
      "epoch": 0.7902097902097902,
      "grad_norm": 0.09631301174932007,
      "learning_rate": 7.573620930109906e-06,
      "loss": 0.3275,
      "step": 1130
    },
    {
      "epoch": 0.7937062937062938,
      "grad_norm": 0.09326550273629058,
      "learning_rate": 7.547412887116224e-06,
      "loss": 0.3193,
      "step": 1135
    },
    {
      "epoch": 0.7972027972027972,
      "grad_norm": 0.09732171619011357,
      "learning_rate": 7.521109975922069e-06,
      "loss": 0.3317,
      "step": 1140
    },
    {
      "epoch": 0.8006993006993007,
      "grad_norm": 0.08975727645012774,
      "learning_rate": 7.494713176074128e-06,
      "loss": 0.3232,
      "step": 1145
    },
    {
      "epoch": 0.8041958041958042,
      "grad_norm": 0.10268943209415242,
      "learning_rate": 7.468223470615593e-06,
      "loss": 0.3549,
      "step": 1150
    },
    {
      "epoch": 0.8076923076923077,
      "grad_norm": 0.09719763483337895,
      "learning_rate": 7.441641846049557e-06,
      "loss": 0.3217,
      "step": 1155
    },
    {
      "epoch": 0.8111888111888111,
      "grad_norm": 0.09866887439065343,
      "learning_rate": 7.414969292302267e-06,
      "loss": 0.3276,
      "step": 1160
    },
    {
      "epoch": 0.8146853146853147,
      "grad_norm": 0.09508520292838865,
      "learning_rate": 7.388206802686272e-06,
      "loss": 0.325,
      "step": 1165
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.10885901013429214,
      "learning_rate": 7.361355373863415e-06,
      "loss": 0.309,
      "step": 1170
    },
    {
      "epoch": 0.8216783216783217,
      "grad_norm": 0.100534618852991,
      "learning_rate": 7.334416005807726e-06,
      "loss": 0.3206,
      "step": 1175
    },
    {
      "epoch": 0.8251748251748252,
      "grad_norm": 0.12043671832245108,
      "learning_rate": 7.307389701768183e-06,
      "loss": 0.3299,
      "step": 1180
    },
    {
      "epoch": 0.8286713286713286,
      "grad_norm": 0.08393550874271685,
      "learning_rate": 7.280277468231343e-06,
      "loss": 0.3262,
      "step": 1185
    },
    {
      "epoch": 0.8321678321678322,
      "grad_norm": 0.11121015663273821,
      "learning_rate": 7.253080314883863e-06,
      "loss": 0.2895,
      "step": 1190
    },
    {
      "epoch": 0.8356643356643356,
      "grad_norm": 0.09185059584253953,
      "learning_rate": 7.2257992545749045e-06,
      "loss": 0.3225,
      "step": 1195
    },
    {
      "epoch": 0.8391608391608392,
      "grad_norm": 0.08881396203220153,
      "learning_rate": 7.198435303278401e-06,
      "loss": 0.3276,
      "step": 1200
    },
    {
      "epoch": 0.8426573426573427,
      "grad_norm": 0.09746576829873795,
      "learning_rate": 7.170989480055233e-06,
      "loss": 0.3455,
      "step": 1205
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 0.09888839565795066,
      "learning_rate": 7.143462807015271e-06,
      "loss": 0.3392,
      "step": 1210
    },
    {
      "epoch": 0.8496503496503497,
      "grad_norm": 0.09592592285144286,
      "learning_rate": 7.1158563092793165e-06,
      "loss": 0.3091,
      "step": 1215
    },
    {
      "epoch": 0.8531468531468531,
      "grad_norm": 0.0977987635493939,
      "learning_rate": 7.08817101494092e-06,
      "loss": 0.3165,
      "step": 1220
    },
    {
      "epoch": 0.8566433566433567,
      "grad_norm": 0.10380850676568182,
      "learning_rate": 7.060407955028098e-06,
      "loss": 0.3115,
      "step": 1225
    },
    {
      "epoch": 0.8601398601398601,
      "grad_norm": 0.09447180359888213,
      "learning_rate": 7.032568163464932e-06,
      "loss": 0.3207,
      "step": 1230
    },
    {
      "epoch": 0.8636363636363636,
      "grad_norm": 0.10061269710408283,
      "learning_rate": 7.004652677033069e-06,
      "loss": 0.3161,
      "step": 1235
    },
    {
      "epoch": 0.8671328671328671,
      "grad_norm": 0.09609425213694053,
      "learning_rate": 6.976662535333107e-06,
      "loss": 0.3136,
      "step": 1240
    },
    {
      "epoch": 0.8706293706293706,
      "grad_norm": 0.09598094845731318,
      "learning_rate": 6.948598780745885e-06,
      "loss": 0.3009,
      "step": 1245
    },
    {
      "epoch": 0.8741258741258742,
      "grad_norm": 0.10427104049050953,
      "learning_rate": 6.920462458393652e-06,
      "loss": 0.3428,
      "step": 1250
    },
    {
      "epoch": 0.8776223776223776,
      "grad_norm": 0.0994920233360674,
      "learning_rate": 6.892254616101158e-06,
      "loss": 0.3263,
      "step": 1255
    },
    {
      "epoch": 0.8811188811188811,
      "grad_norm": 0.10387800775103824,
      "learning_rate": 6.86397630435663e-06,
      "loss": 0.2997,
      "step": 1260
    },
    {
      "epoch": 0.8846153846153846,
      "grad_norm": 0.10508396462654218,
      "learning_rate": 6.8356285762726385e-06,
      "loss": 0.3192,
      "step": 1265
    },
    {
      "epoch": 0.8881118881118881,
      "grad_norm": 0.10589020185350513,
      "learning_rate": 6.807212487546897e-06,
      "loss": 0.2999,
      "step": 1270
    },
    {
      "epoch": 0.8916083916083916,
      "grad_norm": 0.09769265892044267,
      "learning_rate": 6.7787290964229325e-06,
      "loss": 0.3131,
      "step": 1275
    },
    {
      "epoch": 0.8951048951048951,
      "grad_norm": 0.09679683640782744,
      "learning_rate": 6.750179463650681e-06,
      "loss": 0.3268,
      "step": 1280
    },
    {
      "epoch": 0.8986013986013986,
      "grad_norm": 0.09209711149231241,
      "learning_rate": 6.721564652446987e-06,
      "loss": 0.3229,
      "step": 1285
    },
    {
      "epoch": 0.9020979020979021,
      "grad_norm": 0.11288807662970714,
      "learning_rate": 6.692885728456001e-06,
      "loss": 0.3331,
      "step": 1290
    },
    {
      "epoch": 0.9055944055944056,
      "grad_norm": 0.10191677161615505,
      "learning_rate": 6.6641437597095005e-06,
      "loss": 0.3037,
      "step": 1295
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.10997644202751486,
      "learning_rate": 6.635339816587109e-06,
      "loss": 0.2949,
      "step": 1300
    },
    {
      "epoch": 0.9125874125874126,
      "grad_norm": 0.09140651524432652,
      "learning_rate": 6.606474971776442e-06,
      "loss": 0.3409,
      "step": 1305
    },
    {
      "epoch": 0.916083916083916,
      "grad_norm": 0.08471656771823269,
      "learning_rate": 6.577550300233151e-06,
      "loss": 0.3238,
      "step": 1310
    },
    {
      "epoch": 0.9195804195804196,
      "grad_norm": 0.09813671908189145,
      "learning_rate": 6.548566879140897e-06,
      "loss": 0.3082,
      "step": 1315
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.09826741583981999,
      "learning_rate": 6.519525787871235e-06,
      "loss": 0.3289,
      "step": 1320
    },
    {
      "epoch": 0.9265734265734266,
      "grad_norm": 0.08826203150682889,
      "learning_rate": 6.490428107943409e-06,
      "loss": 0.3116,
      "step": 1325
    },
    {
      "epoch": 0.9300699300699301,
      "grad_norm": 0.11047210665505915,
      "learning_rate": 6.461274922984087e-06,
      "loss": 0.3096,
      "step": 1330
    },
    {
      "epoch": 0.9335664335664335,
      "grad_norm": 0.09280424318322683,
      "learning_rate": 6.4320673186869954e-06,
      "loss": 0.3022,
      "step": 1335
    },
    {
      "epoch": 0.9370629370629371,
      "grad_norm": 0.09428713209319824,
      "learning_rate": 6.402806382772494e-06,
      "loss": 0.3194,
      "step": 1340
    },
    {
      "epoch": 0.9405594405594405,
      "grad_norm": 0.09817792403050692,
      "learning_rate": 6.373493204947065e-06,
      "loss": 0.3042,
      "step": 1345
    },
    {
      "epoch": 0.9440559440559441,
      "grad_norm": 0.08292232832861181,
      "learning_rate": 6.344128876862729e-06,
      "loss": 0.2787,
      "step": 1350
    },
    {
      "epoch": 0.9475524475524476,
      "grad_norm": 0.1060764346689841,
      "learning_rate": 6.3147144920763985e-06,
      "loss": 0.3323,
      "step": 1355
    },
    {
      "epoch": 0.951048951048951,
      "grad_norm": 0.0908264916657139,
      "learning_rate": 6.2852511460091406e-06,
      "loss": 0.2884,
      "step": 1360
    },
    {
      "epoch": 0.9545454545454546,
      "grad_norm": 0.11839968846971091,
      "learning_rate": 6.255739935905396e-06,
      "loss": 0.3125,
      "step": 1365
    },
    {
      "epoch": 0.958041958041958,
      "grad_norm": 0.11288513824262601,
      "learning_rate": 6.226181960792107e-06,
      "loss": 0.3063,
      "step": 1370
    },
    {
      "epoch": 0.9615384615384616,
      "grad_norm": 0.10439467637777479,
      "learning_rate": 6.1965783214377895e-06,
      "loss": 0.3115,
      "step": 1375
    },
    {
      "epoch": 0.965034965034965,
      "grad_norm": 0.09149993105772337,
      "learning_rate": 6.166930120311545e-06,
      "loss": 0.2981,
      "step": 1380
    },
    {
      "epoch": 0.9685314685314685,
      "grad_norm": 0.10624923889838721,
      "learning_rate": 6.137238461541998e-06,
      "loss": 0.3083,
      "step": 1385
    },
    {
      "epoch": 0.972027972027972,
      "grad_norm": 0.09916502005960598,
      "learning_rate": 6.1075044508761804e-06,
      "loss": 0.3213,
      "step": 1390
    },
    {
      "epoch": 0.9755244755244755,
      "grad_norm": 0.11294055771283724,
      "learning_rate": 6.077729195638349e-06,
      "loss": 0.3038,
      "step": 1395
    },
    {
      "epoch": 0.9790209790209791,
      "grad_norm": 0.10125490787864704,
      "learning_rate": 6.047913804688751e-06,
      "loss": 0.3204,
      "step": 1400
    },
    {
      "epoch": 0.9825174825174825,
      "grad_norm": 0.1015505902245628,
      "learning_rate": 6.018059388382327e-06,
      "loss": 0.3272,
      "step": 1405
    },
    {
      "epoch": 0.986013986013986,
      "grad_norm": 0.0997218077059797,
      "learning_rate": 5.988167058527361e-06,
      "loss": 0.3095,
      "step": 1410
    },
    {
      "epoch": 0.9895104895104895,
      "grad_norm": 0.11184085378893502,
      "learning_rate": 5.958237928344077e-06,
      "loss": 0.3102,
      "step": 1415
    },
    {
      "epoch": 0.993006993006993,
      "grad_norm": 0.09497844419150701,
      "learning_rate": 5.928273112423177e-06,
      "loss": 0.3357,
      "step": 1420
    },
    {
      "epoch": 0.9965034965034965,
      "grad_norm": 0.09785679757144553,
      "learning_rate": 5.898273726684338e-06,
      "loss": 0.3177,
      "step": 1425
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.10414865345603919,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.3216,
      "step": 1430
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3180939555168152,
      "eval_runtime": 33.8853,
      "eval_samples_per_second": 20.481,
      "eval_steps_per_second": 5.135,
      "step": 1430
    },
    {
      "epoch": 1.0034965034965035,
      "grad_norm": 0.09860050229670009,
      "learning_rate": 5.838175715827015e-06,
      "loss": 0.283,
      "step": 1435
    },
    {
      "epoch": 1.006993006993007,
      "grad_norm": 0.09643250585976647,
      "learning_rate": 5.808079328818488e-06,
      "loss": 0.2831,
      "step": 1440
    },
    {
      "epoch": 1.0104895104895104,
      "grad_norm": 0.11960210700385758,
      "learning_rate": 5.777952848128578e-06,
      "loss": 0.306,
      "step": 1445
    },
    {
      "epoch": 1.013986013986014,
      "grad_norm": 0.09213590484634555,
      "learning_rate": 5.747797395697525e-06,
      "loss": 0.2905,
      "step": 1450
    },
    {
      "epoch": 1.0174825174825175,
      "grad_norm": 0.11181608085903355,
      "learning_rate": 5.717614094544497e-06,
      "loss": 0.2965,
      "step": 1455
    },
    {
      "epoch": 1.020979020979021,
      "grad_norm": 0.11274993170515431,
      "learning_rate": 5.687404068725779e-06,
      "loss": 0.2882,
      "step": 1460
    },
    {
      "epoch": 1.0244755244755244,
      "grad_norm": 0.10367759599697986,
      "learning_rate": 5.657168443292909e-06,
      "loss": 0.2848,
      "step": 1465
    },
    {
      "epoch": 1.027972027972028,
      "grad_norm": 0.0941238300478702,
      "learning_rate": 5.626908344250778e-06,
      "loss": 0.2992,
      "step": 1470
    },
    {
      "epoch": 1.0314685314685315,
      "grad_norm": 0.09619381493436249,
      "learning_rate": 5.5966248985157e-06,
      "loss": 0.2812,
      "step": 1475
    },
    {
      "epoch": 1.034965034965035,
      "grad_norm": 0.10400313409523196,
      "learning_rate": 5.566319233873446e-06,
      "loss": 0.3058,
      "step": 1480
    },
    {
      "epoch": 1.0384615384615385,
      "grad_norm": 0.10574597951326378,
      "learning_rate": 5.53599247893724e-06,
      "loss": 0.2758,
      "step": 1485
    },
    {
      "epoch": 1.0419580419580419,
      "grad_norm": 0.10462360020319139,
      "learning_rate": 5.5056457631057245e-06,
      "loss": 0.2746,
      "step": 1490
    },
    {
      "epoch": 1.0454545454545454,
      "grad_norm": 0.11618910922950017,
      "learning_rate": 5.475280216520913e-06,
      "loss": 0.2823,
      "step": 1495
    },
    {
      "epoch": 1.048951048951049,
      "grad_norm": 0.12712963054282944,
      "learning_rate": 5.444896970026092e-06,
      "loss": 0.2835,
      "step": 1500
    },
    {
      "epoch": 1.0524475524475525,
      "grad_norm": 0.09958988336796508,
      "learning_rate": 5.414497155123708e-06,
      "loss": 0.2653,
      "step": 1505
    },
    {
      "epoch": 1.055944055944056,
      "grad_norm": 0.1080659427330968,
      "learning_rate": 5.384081903933235e-06,
      "loss": 0.2949,
      "step": 1510
    },
    {
      "epoch": 1.0594405594405594,
      "grad_norm": 0.11665708479307332,
      "learning_rate": 5.353652349149007e-06,
      "loss": 0.278,
      "step": 1515
    },
    {
      "epoch": 1.062937062937063,
      "grad_norm": 0.12826080199663958,
      "learning_rate": 5.323209623998043e-06,
      "loss": 0.2768,
      "step": 1520
    },
    {
      "epoch": 1.0664335664335665,
      "grad_norm": 0.0931839219725714,
      "learning_rate": 5.292754862197831e-06,
      "loss": 0.2777,
      "step": 1525
    },
    {
      "epoch": 1.06993006993007,
      "grad_norm": 0.107320551035206,
      "learning_rate": 5.2622891979141265e-06,
      "loss": 0.2816,
      "step": 1530
    },
    {
      "epoch": 1.0734265734265733,
      "grad_norm": 0.1061573386524815,
      "learning_rate": 5.231813765718698e-06,
      "loss": 0.2895,
      "step": 1535
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 0.1579249118505657,
      "learning_rate": 5.201329700547077e-06,
      "loss": 0.2634,
      "step": 1540
    },
    {
      "epoch": 1.0804195804195804,
      "grad_norm": 0.10185189280525993,
      "learning_rate": 5.170838137656304e-06,
      "loss": 0.2808,
      "step": 1545
    },
    {
      "epoch": 1.083916083916084,
      "grad_norm": 0.11469755609644863,
      "learning_rate": 5.140340212582639e-06,
      "loss": 0.2817,
      "step": 1550
    },
    {
      "epoch": 1.0874125874125875,
      "grad_norm": 0.09969775272157483,
      "learning_rate": 5.109837061099274e-06,
      "loss": 0.2905,
      "step": 1555
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.11204511989520233,
      "learning_rate": 5.07932981917404e-06,
      "loss": 0.2983,
      "step": 1560
    },
    {
      "epoch": 1.0944055944055944,
      "grad_norm": 0.10859999114246341,
      "learning_rate": 5.0488196229271005e-06,
      "loss": 0.2817,
      "step": 1565
    },
    {
      "epoch": 1.097902097902098,
      "grad_norm": 0.10124983316705717,
      "learning_rate": 5.018307608588637e-06,
      "loss": 0.2694,
      "step": 1570
    },
    {
      "epoch": 1.1013986013986015,
      "grad_norm": 0.09673028033429812,
      "learning_rate": 4.9877949124565415e-06,
      "loss": 0.2804,
      "step": 1575
    },
    {
      "epoch": 1.104895104895105,
      "grad_norm": 0.12406697294731686,
      "learning_rate": 4.957282670854099e-06,
      "loss": 0.2842,
      "step": 1580
    },
    {
      "epoch": 1.1083916083916083,
      "grad_norm": 0.10389483855300476,
      "learning_rate": 4.926772020087663e-06,
      "loss": 0.3067,
      "step": 1585
    },
    {
      "epoch": 1.1118881118881119,
      "grad_norm": 0.11624832680414443,
      "learning_rate": 4.896264096404343e-06,
      "loss": 0.2989,
      "step": 1590
    },
    {
      "epoch": 1.1153846153846154,
      "grad_norm": 0.10001637455582953,
      "learning_rate": 4.865760035949695e-06,
      "loss": 0.3016,
      "step": 1595
    },
    {
      "epoch": 1.118881118881119,
      "grad_norm": 0.11275995426749594,
      "learning_rate": 4.835260974725397e-06,
      "loss": 0.3062,
      "step": 1600
    },
    {
      "epoch": 1.1223776223776223,
      "grad_norm": 0.09710030416818968,
      "learning_rate": 4.804768048546955e-06,
      "loss": 0.3021,
      "step": 1605
    },
    {
      "epoch": 1.1258741258741258,
      "grad_norm": 0.11969077356400575,
      "learning_rate": 4.774282393001397e-06,
      "loss": 0.2839,
      "step": 1610
    },
    {
      "epoch": 1.1293706293706294,
      "grad_norm": 0.1165475840983693,
      "learning_rate": 4.743805143404989e-06,
      "loss": 0.3033,
      "step": 1615
    },
    {
      "epoch": 1.132867132867133,
      "grad_norm": 0.11943387092333048,
      "learning_rate": 4.713337434760949e-06,
      "loss": 0.2911,
      "step": 1620
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.11663805469555324,
      "learning_rate": 4.682880401717178e-06,
      "loss": 0.2585,
      "step": 1625
    },
    {
      "epoch": 1.1398601398601398,
      "grad_norm": 0.10669609950388007,
      "learning_rate": 4.652435178524013e-06,
      "loss": 0.2907,
      "step": 1630
    },
    {
      "epoch": 1.1433566433566433,
      "grad_norm": 0.11232703578650871,
      "learning_rate": 4.622002898991972e-06,
      "loss": 0.2628,
      "step": 1635
    },
    {
      "epoch": 1.1468531468531469,
      "grad_norm": 0.11335115400681652,
      "learning_rate": 4.591584696449549e-06,
      "loss": 0.2793,
      "step": 1640
    },
    {
      "epoch": 1.1503496503496504,
      "grad_norm": 0.12058621719051137,
      "learning_rate": 4.561181703700986e-06,
      "loss": 0.2874,
      "step": 1645
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 0.1106409571474046,
      "learning_rate": 4.530795052984104e-06,
      "loss": 0.3156,
      "step": 1650
    },
    {
      "epoch": 1.1573426573426573,
      "grad_norm": 0.09597427027532647,
      "learning_rate": 4.500425875928129e-06,
      "loss": 0.2507,
      "step": 1655
    },
    {
      "epoch": 1.1608391608391608,
      "grad_norm": 0.11526214137962942,
      "learning_rate": 4.470075303511546e-06,
      "loss": 0.2992,
      "step": 1660
    },
    {
      "epoch": 1.1643356643356644,
      "grad_norm": 0.10720902756157567,
      "learning_rate": 4.439744466019993e-06,
      "loss": 0.3007,
      "step": 1665
    },
    {
      "epoch": 1.167832167832168,
      "grad_norm": 0.10095686169764254,
      "learning_rate": 4.4094344930041515e-06,
      "loss": 0.2827,
      "step": 1670
    },
    {
      "epoch": 1.1713286713286712,
      "grad_norm": 0.12768377416510107,
      "learning_rate": 4.379146513237695e-06,
      "loss": 0.2874,
      "step": 1675
    },
    {
      "epoch": 1.1748251748251748,
      "grad_norm": 0.098356265123962,
      "learning_rate": 4.348881654675241e-06,
      "loss": 0.2678,
      "step": 1680
    },
    {
      "epoch": 1.1783216783216783,
      "grad_norm": 0.12024264878083898,
      "learning_rate": 4.318641044410358e-06,
      "loss": 0.27,
      "step": 1685
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.10256378318597871,
      "learning_rate": 4.2884258086335755e-06,
      "loss": 0.2621,
      "step": 1690
    },
    {
      "epoch": 1.1853146853146854,
      "grad_norm": 0.123973849351433,
      "learning_rate": 4.258237072590455e-06,
      "loss": 0.2723,
      "step": 1695
    },
    {
      "epoch": 1.1888111888111887,
      "grad_norm": 0.12348063377350953,
      "learning_rate": 4.228075960539684e-06,
      "loss": 0.2971,
      "step": 1700
    },
    {
      "epoch": 1.1923076923076923,
      "grad_norm": 0.13820520258807037,
      "learning_rate": 4.1979435957111984e-06,
      "loss": 0.2959,
      "step": 1705
    },
    {
      "epoch": 1.1958041958041958,
      "grad_norm": 0.10508613597681296,
      "learning_rate": 4.167841100264369e-06,
      "loss": 0.2765,
      "step": 1710
    },
    {
      "epoch": 1.1993006993006994,
      "grad_norm": 0.11352798455372175,
      "learning_rate": 4.1377695952461905e-06,
      "loss": 0.2595,
      "step": 1715
    },
    {
      "epoch": 1.2027972027972027,
      "grad_norm": 0.11589669656742159,
      "learning_rate": 4.107730200549549e-06,
      "loss": 0.2673,
      "step": 1720
    },
    {
      "epoch": 1.2062937062937062,
      "grad_norm": 0.11665313294655853,
      "learning_rate": 4.0777240348715085e-06,
      "loss": 0.2986,
      "step": 1725
    },
    {
      "epoch": 1.2097902097902098,
      "grad_norm": 0.11700009297269698,
      "learning_rate": 4.04775221567165e-06,
      "loss": 0.2735,
      "step": 1730
    },
    {
      "epoch": 1.2132867132867133,
      "grad_norm": 0.10262056570150567,
      "learning_rate": 4.017815859130462e-06,
      "loss": 0.2802,
      "step": 1735
    },
    {
      "epoch": 1.2167832167832167,
      "grad_norm": 0.12088842148152552,
      "learning_rate": 3.987916080107762e-06,
      "loss": 0.2595,
      "step": 1740
    },
    {
      "epoch": 1.2202797202797202,
      "grad_norm": 0.123222073537631,
      "learning_rate": 3.958053992101189e-06,
      "loss": 0.2728,
      "step": 1745
    },
    {
      "epoch": 1.2237762237762237,
      "grad_norm": 0.14415879676636115,
      "learning_rate": 3.928230707204729e-06,
      "loss": 0.2757,
      "step": 1750
    },
    {
      "epoch": 1.2272727272727273,
      "grad_norm": 0.1059023549630983,
      "learning_rate": 3.898447336067297e-06,
      "loss": 0.268,
      "step": 1755
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.12235246365024488,
      "learning_rate": 3.86870498785139e-06,
      "loss": 0.2793,
      "step": 1760
    },
    {
      "epoch": 1.2342657342657342,
      "grad_norm": 0.1266933996658085,
      "learning_rate": 3.8390047701917625e-06,
      "loss": 0.2791,
      "step": 1765
    },
    {
      "epoch": 1.2377622377622377,
      "grad_norm": 0.1230245533106991,
      "learning_rate": 3.8093477891541864e-06,
      "loss": 0.3001,
      "step": 1770
    },
    {
      "epoch": 1.2412587412587412,
      "grad_norm": 0.11654806074283108,
      "learning_rate": 3.779735149194258e-06,
      "loss": 0.2826,
      "step": 1775
    },
    {
      "epoch": 1.2447552447552448,
      "grad_norm": 0.10770540503295759,
      "learning_rate": 3.750167953116272e-06,
      "loss": 0.2865,
      "step": 1780
    },
    {
      "epoch": 1.2482517482517483,
      "grad_norm": 0.11414218881414812,
      "learning_rate": 3.720647302032144e-06,
      "loss": 0.2742,
      "step": 1785
    },
    {
      "epoch": 1.2517482517482517,
      "grad_norm": 0.11830539545059722,
      "learning_rate": 3.6911742953204046e-06,
      "loss": 0.2767,
      "step": 1790
    },
    {
      "epoch": 1.2552447552447552,
      "grad_norm": 0.10011126409724302,
      "learning_rate": 3.6617500305852676e-06,
      "loss": 0.2552,
      "step": 1795
    },
    {
      "epoch": 1.2587412587412588,
      "grad_norm": 0.12742089574311471,
      "learning_rate": 3.6323756036157386e-06,
      "loss": 0.301,
      "step": 1800
    },
    {
      "epoch": 1.2622377622377623,
      "grad_norm": 0.12157261199386625,
      "learning_rate": 3.6030521083448256e-06,
      "loss": 0.275,
      "step": 1805
    },
    {
      "epoch": 1.2657342657342658,
      "grad_norm": 0.13018781979587624,
      "learning_rate": 3.5737806368087776e-06,
      "loss": 0.2791,
      "step": 1810
    },
    {
      "epoch": 1.2692307692307692,
      "grad_norm": 0.11472896479036702,
      "learning_rate": 3.544562279106436e-06,
      "loss": 0.2597,
      "step": 1815
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.13983756485000562,
      "learning_rate": 3.5153981233586277e-06,
      "loss": 0.2793,
      "step": 1820
    },
    {
      "epoch": 1.2762237762237763,
      "grad_norm": 0.12377750647958802,
      "learning_rate": 3.48628925566764e-06,
      "loss": 0.2779,
      "step": 1825
    },
    {
      "epoch": 1.2797202797202798,
      "grad_norm": 0.13033680974681505,
      "learning_rate": 3.4572367600767874e-06,
      "loss": 0.284,
      "step": 1830
    },
    {
      "epoch": 1.2832167832167833,
      "grad_norm": 0.1252056404310726,
      "learning_rate": 3.428241718530022e-06,
      "loss": 0.2704,
      "step": 1835
    },
    {
      "epoch": 1.2867132867132867,
      "grad_norm": 0.118752876893844,
      "learning_rate": 3.3993052108316566e-06,
      "loss": 0.2698,
      "step": 1840
    },
    {
      "epoch": 1.2902097902097902,
      "grad_norm": 0.12876147193567186,
      "learning_rate": 3.370428314606138e-06,
      "loss": 0.2885,
      "step": 1845
    },
    {
      "epoch": 1.2937062937062938,
      "grad_norm": 0.11959862346971185,
      "learning_rate": 3.3416121052579297e-06,
      "loss": 0.2695,
      "step": 1850
    },
    {
      "epoch": 1.297202797202797,
      "grad_norm": 0.13176076976769643,
      "learning_rate": 3.3128576559314507e-06,
      "loss": 0.2777,
      "step": 1855
    },
    {
      "epoch": 1.3006993006993006,
      "grad_norm": 0.1279070658782311,
      "learning_rate": 3.2841660374711144e-06,
      "loss": 0.2697,
      "step": 1860
    },
    {
      "epoch": 1.3041958041958042,
      "grad_norm": 0.13489716472147867,
      "learning_rate": 3.2555383183814544e-06,
      "loss": 0.3036,
      "step": 1865
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.12334889072330245,
      "learning_rate": 3.226975564787322e-06,
      "loss": 0.2705,
      "step": 1870
    },
    {
      "epoch": 1.3111888111888113,
      "grad_norm": 0.13592582265687866,
      "learning_rate": 3.198478840394196e-06,
      "loss": 0.2941,
      "step": 1875
    },
    {
      "epoch": 1.3146853146853146,
      "grad_norm": 0.11390774415602396,
      "learning_rate": 3.170049206448553e-06,
      "loss": 0.2824,
      "step": 1880
    },
    {
      "epoch": 1.3181818181818181,
      "grad_norm": 0.15890655850394841,
      "learning_rate": 3.141687721698363e-06,
      "loss": 0.2653,
      "step": 1885
    },
    {
      "epoch": 1.3216783216783217,
      "grad_norm": 0.13577126735405054,
      "learning_rate": 3.113395442353646e-06,
      "loss": 0.2963,
      "step": 1890
    },
    {
      "epoch": 1.3251748251748252,
      "grad_norm": 0.1271334783456917,
      "learning_rate": 3.0851734220471434e-06,
      "loss": 0.2676,
      "step": 1895
    },
    {
      "epoch": 1.3286713286713288,
      "grad_norm": 0.14583157351563786,
      "learning_rate": 3.0570227117950857e-06,
      "loss": 0.2831,
      "step": 1900
    },
    {
      "epoch": 1.332167832167832,
      "grad_norm": 0.14995769193849226,
      "learning_rate": 3.0289443599580387e-06,
      "loss": 0.2717,
      "step": 1905
    },
    {
      "epoch": 1.3356643356643356,
      "grad_norm": 0.12151760728520503,
      "learning_rate": 3.0009394122018744e-06,
      "loss": 0.2594,
      "step": 1910
    },
    {
      "epoch": 1.3391608391608392,
      "grad_norm": 0.14335474194484363,
      "learning_rate": 2.973008911458816e-06,
      "loss": 0.2658,
      "step": 1915
    },
    {
      "epoch": 1.3426573426573427,
      "grad_norm": 0.17742533944724417,
      "learning_rate": 2.9451538978886165e-06,
      "loss": 0.2745,
      "step": 1920
    },
    {
      "epoch": 1.3461538461538463,
      "grad_norm": 0.12048834268944708,
      "learning_rate": 2.917375408839803e-06,
      "loss": 0.2728,
      "step": 1925
    },
    {
      "epoch": 1.3496503496503496,
      "grad_norm": 0.1309489385175574,
      "learning_rate": 2.8896744788110497e-06,
      "loss": 0.2603,
      "step": 1930
    },
    {
      "epoch": 1.3531468531468531,
      "grad_norm": 0.12100172035944871,
      "learning_rate": 2.86205213941267e-06,
      "loss": 0.2497,
      "step": 1935
    },
    {
      "epoch": 1.3566433566433567,
      "grad_norm": 0.13660442658303193,
      "learning_rate": 2.8345094193281673e-06,
      "loss": 0.2683,
      "step": 1940
    },
    {
      "epoch": 1.3601398601398602,
      "grad_norm": 0.20029948084159513,
      "learning_rate": 2.8070473442759517e-06,
      "loss": 0.265,
      "step": 1945
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.1217718840811214,
      "learning_rate": 2.7796669369711294e-06,
      "loss": 0.2758,
      "step": 1950
    },
    {
      "epoch": 1.367132867132867,
      "grad_norm": 0.1218788812023423,
      "learning_rate": 2.752369217087418e-06,
      "loss": 0.2706,
      "step": 1955
    },
    {
      "epoch": 1.3706293706293706,
      "grad_norm": 0.1681619637715074,
      "learning_rate": 2.7251552012191763e-06,
      "loss": 0.2521,
      "step": 1960
    },
    {
      "epoch": 1.3741258741258742,
      "grad_norm": 0.13215871161118758,
      "learning_rate": 2.698025902843533e-06,
      "loss": 0.2668,
      "step": 1965
    },
    {
      "epoch": 1.3776223776223775,
      "grad_norm": 0.16597320281641273,
      "learning_rate": 2.670982332282661e-06,
      "loss": 0.2601,
      "step": 1970
    },
    {
      "epoch": 1.381118881118881,
      "grad_norm": 0.13286398460876903,
      "learning_rate": 2.644025496666143e-06,
      "loss": 0.2684,
      "step": 1975
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 0.14964743881741074,
      "learning_rate": 2.6171563998934605e-06,
      "loss": 0.2636,
      "step": 1980
    },
    {
      "epoch": 1.3881118881118881,
      "grad_norm": 0.17737672923194353,
      "learning_rate": 2.590376042596621e-06,
      "loss": 0.2667,
      "step": 1985
    },
    {
      "epoch": 1.3916083916083917,
      "grad_norm": 0.13696100404924544,
      "learning_rate": 2.5636854221028764e-06,
      "loss": 0.2692,
      "step": 1990
    },
    {
      "epoch": 1.395104895104895,
      "grad_norm": 0.1359722927391487,
      "learning_rate": 2.537085532397595e-06,
      "loss": 0.2672,
      "step": 1995
    },
    {
      "epoch": 1.3986013986013985,
      "grad_norm": 0.17097022815814872,
      "learning_rate": 2.5105773640872416e-06,
      "loss": 0.2561,
      "step": 2000
    },
    {
      "epoch": 1.402097902097902,
      "grad_norm": 0.1414647308761833,
      "learning_rate": 2.484161904362481e-06,
      "loss": 0.2437,
      "step": 2005
    },
    {
      "epoch": 1.4055944055944056,
      "grad_norm": 0.14581181430999388,
      "learning_rate": 2.4578401369614196e-06,
      "loss": 0.2733,
      "step": 2010
    },
    {
      "epoch": 1.4090909090909092,
      "grad_norm": 0.16322883114094733,
      "learning_rate": 2.4316130421329696e-06,
      "loss": 0.2699,
      "step": 2015
    },
    {
      "epoch": 1.4125874125874125,
      "grad_norm": 0.14215927984521137,
      "learning_rate": 2.4054815966003365e-06,
      "loss": 0.2719,
      "step": 2020
    },
    {
      "epoch": 1.416083916083916,
      "grad_norm": 0.14648779179733568,
      "learning_rate": 2.379446773524656e-06,
      "loss": 0.2416,
      "step": 2025
    },
    {
      "epoch": 1.4195804195804196,
      "grad_norm": 0.1521431992671824,
      "learning_rate": 2.3535095424687443e-06,
      "loss": 0.2365,
      "step": 2030
    },
    {
      "epoch": 1.4230769230769231,
      "grad_norm": 0.13347271495618993,
      "learning_rate": 2.3276708693609947e-06,
      "loss": 0.2695,
      "step": 2035
    },
    {
      "epoch": 1.4265734265734267,
      "grad_norm": 0.1645738689375383,
      "learning_rate": 2.301931716459405e-06,
      "loss": 0.2714,
      "step": 2040
    },
    {
      "epoch": 1.43006993006993,
      "grad_norm": 0.1601133081604159,
      "learning_rate": 2.2762930423157346e-06,
      "loss": 0.2628,
      "step": 2045
    },
    {
      "epoch": 1.4335664335664335,
      "grad_norm": 0.15437084886490815,
      "learning_rate": 2.2507558017398263e-06,
      "loss": 0.2559,
      "step": 2050
    },
    {
      "epoch": 1.437062937062937,
      "grad_norm": 0.15803419436134836,
      "learning_rate": 2.2253209457640205e-06,
      "loss": 0.2643,
      "step": 2055
    },
    {
      "epoch": 1.4405594405594406,
      "grad_norm": 0.12851123172489884,
      "learning_rate": 2.1999894216077633e-06,
      "loss": 0.2664,
      "step": 2060
    },
    {
      "epoch": 1.4440559440559442,
      "grad_norm": 0.19040825534872968,
      "learning_rate": 2.174762172642319e-06,
      "loss": 0.2473,
      "step": 2065
    },
    {
      "epoch": 1.4475524475524475,
      "grad_norm": 0.15065545559977453,
      "learning_rate": 2.149640138355632e-06,
      "loss": 0.2742,
      "step": 2070
    },
    {
      "epoch": 1.451048951048951,
      "grad_norm": 0.1922974353635457,
      "learning_rate": 2.1246242543173613e-06,
      "loss": 0.275,
      "step": 2075
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.17982162397533608,
      "learning_rate": 2.09971545214401e-06,
      "loss": 0.2635,
      "step": 2080
    },
    {
      "epoch": 1.458041958041958,
      "grad_norm": 0.1738289418178212,
      "learning_rate": 2.074914659464254e-06,
      "loss": 0.2381,
      "step": 2085
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.12853571069602446,
      "learning_rate": 2.050222799884387e-06,
      "loss": 0.2352,
      "step": 2090
    },
    {
      "epoch": 1.465034965034965,
      "grad_norm": 0.15868082671286524,
      "learning_rate": 2.0256407929539203e-06,
      "loss": 0.2701,
      "step": 2095
    },
    {
      "epoch": 1.4685314685314685,
      "grad_norm": 0.15024281799339492,
      "learning_rate": 2.0011695541313514e-06,
      "loss": 0.2687,
      "step": 2100
    },
    {
      "epoch": 1.472027972027972,
      "grad_norm": 0.16538424584383676,
      "learning_rate": 1.9768099947500535e-06,
      "loss": 0.2665,
      "step": 2105
    },
    {
      "epoch": 1.4755244755244754,
      "grad_norm": 0.1469847725186328,
      "learning_rate": 1.9525630219843524e-06,
      "loss": 0.2592,
      "step": 2110
    },
    {
      "epoch": 1.479020979020979,
      "grad_norm": 0.19914120331298027,
      "learning_rate": 1.9284295388157328e-06,
      "loss": 0.2526,
      "step": 2115
    },
    {
      "epoch": 1.4825174825174825,
      "grad_norm": 0.14171359694212368,
      "learning_rate": 1.9044104439992145e-06,
      "loss": 0.2571,
      "step": 2120
    },
    {
      "epoch": 1.486013986013986,
      "grad_norm": 0.15468593500001104,
      "learning_rate": 1.880506632029883e-06,
      "loss": 0.2665,
      "step": 2125
    },
    {
      "epoch": 1.4895104895104896,
      "grad_norm": 0.17533297642498313,
      "learning_rate": 1.8567189931095691e-06,
      "loss": 0.2493,
      "step": 2130
    },
    {
      "epoch": 1.493006993006993,
      "grad_norm": 0.158575042035332,
      "learning_rate": 1.8330484131137093e-06,
      "loss": 0.2643,
      "step": 2135
    },
    {
      "epoch": 1.4965034965034965,
      "grad_norm": 0.1699070377257086,
      "learning_rate": 1.8094957735583463e-06,
      "loss": 0.2424,
      "step": 2140
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.16191985547326435,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.2442,
      "step": 2145
    },
    {
      "epoch": 1.5034965034965035,
      "grad_norm": 0.16249677192430212,
      "learning_rate": 1.7627478198395193e-06,
      "loss": 0.2755,
      "step": 2150
    },
    {
      "epoch": 1.506993006993007,
      "grad_norm": 0.16721652302600015,
      "learning_rate": 1.7395542466165488e-06,
      "loss": 0.2743,
      "step": 2155
    },
    {
      "epoch": 1.5104895104895104,
      "grad_norm": 0.1659393589666137,
      "learning_rate": 1.7164820956502237e-06,
      "loss": 0.2637,
      "step": 2160
    },
    {
      "epoch": 1.513986013986014,
      "grad_norm": 0.2015968042608536,
      "learning_rate": 1.6935322261704929e-06,
      "loss": 0.2439,
      "step": 2165
    },
    {
      "epoch": 1.5174825174825175,
      "grad_norm": 0.14206056623724686,
      "learning_rate": 1.670705492853421e-06,
      "loss": 0.2501,
      "step": 2170
    },
    {
      "epoch": 1.5209790209790208,
      "grad_norm": 0.16138600835314076,
      "learning_rate": 1.6480027457893572e-06,
      "loss": 0.24,
      "step": 2175
    },
    {
      "epoch": 1.5244755244755246,
      "grad_norm": 0.16457265389043566,
      "learning_rate": 1.6254248304512816e-06,
      "loss": 0.2563,
      "step": 2180
    },
    {
      "epoch": 1.527972027972028,
      "grad_norm": 0.1795356890102337,
      "learning_rate": 1.6029725876633085e-06,
      "loss": 0.263,
      "step": 2185
    },
    {
      "epoch": 1.5314685314685315,
      "grad_norm": 0.1664397601665864,
      "learning_rate": 1.580646853569393e-06,
      "loss": 0.24,
      "step": 2190
    },
    {
      "epoch": 1.534965034965035,
      "grad_norm": 0.14718176529521387,
      "learning_rate": 1.5584484596021694e-06,
      "loss": 0.2428,
      "step": 2195
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.14279558469667486,
      "learning_rate": 1.5363782324520033e-06,
      "loss": 0.249,
      "step": 2200
    },
    {
      "epoch": 1.541958041958042,
      "grad_norm": 0.15140009314370015,
      "learning_rate": 1.514436994036201e-06,
      "loss": 0.2495,
      "step": 2205
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.1619202860450826,
      "learning_rate": 1.4926255614683931e-06,
      "loss": 0.249,
      "step": 2210
    },
    {
      "epoch": 1.548951048951049,
      "grad_norm": 0.1716893915072177,
      "learning_rate": 1.4709447470281219e-06,
      "loss": 0.2588,
      "step": 2215
    },
    {
      "epoch": 1.5524475524475525,
      "grad_norm": 0.16886404646815958,
      "learning_rate": 1.4493953581305686e-06,
      "loss": 0.238,
      "step": 2220
    },
    {
      "epoch": 1.5559440559440558,
      "grad_norm": 0.16562116632419963,
      "learning_rate": 1.427978197296503e-06,
      "loss": 0.256,
      "step": 2225
    },
    {
      "epoch": 1.5594405594405596,
      "grad_norm": 0.1556634989889314,
      "learning_rate": 1.406694062122389e-06,
      "loss": 0.2378,
      "step": 2230
    },
    {
      "epoch": 1.562937062937063,
      "grad_norm": 0.1855724126302037,
      "learning_rate": 1.3855437452506759e-06,
      "loss": 0.2542,
      "step": 2235
    },
    {
      "epoch": 1.5664335664335665,
      "grad_norm": 0.16184087877019146,
      "learning_rate": 1.3645280343402973e-06,
      "loss": 0.2336,
      "step": 2240
    },
    {
      "epoch": 1.56993006993007,
      "grad_norm": 0.16147656381172418,
      "learning_rate": 1.3436477120373143e-06,
      "loss": 0.2446,
      "step": 2245
    },
    {
      "epoch": 1.5734265734265733,
      "grad_norm": 0.18106165581351266,
      "learning_rate": 1.3229035559457887e-06,
      "loss": 0.2339,
      "step": 2250
    },
    {
      "epoch": 1.5769230769230769,
      "grad_norm": 0.2111788787476443,
      "learning_rate": 1.3022963385988153e-06,
      "loss": 0.2269,
      "step": 2255
    },
    {
      "epoch": 1.5804195804195804,
      "grad_norm": 0.1669229456184738,
      "learning_rate": 1.2818268274297474e-06,
      "loss": 0.2395,
      "step": 2260
    },
    {
      "epoch": 1.583916083916084,
      "grad_norm": 0.17915301565504643,
      "learning_rate": 1.2614957847436342e-06,
      "loss": 0.2284,
      "step": 2265
    },
    {
      "epoch": 1.5874125874125875,
      "grad_norm": 0.1822294325304649,
      "learning_rate": 1.2413039676888089e-06,
      "loss": 0.2354,
      "step": 2270
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.19524684728719174,
      "learning_rate": 1.2212521282287093e-06,
      "loss": 0.2561,
      "step": 2275
    },
    {
      "epoch": 1.5944055944055944,
      "grad_norm": 0.21352770506743224,
      "learning_rate": 1.2013410131138658e-06,
      "loss": 0.2436,
      "step": 2280
    },
    {
      "epoch": 1.597902097902098,
      "grad_norm": 0.18729319265403963,
      "learning_rate": 1.1815713638540943e-06,
      "loss": 0.2489,
      "step": 2285
    },
    {
      "epoch": 1.6013986013986012,
      "grad_norm": 0.19511685104147358,
      "learning_rate": 1.161943916690883e-06,
      "loss": 0.2651,
      "step": 2290
    },
    {
      "epoch": 1.604895104895105,
      "grad_norm": 0.17784961830764232,
      "learning_rate": 1.1424594025699653e-06,
      "loss": 0.2459,
      "step": 2295
    },
    {
      "epoch": 1.6083916083916083,
      "grad_norm": 0.20035091122936416,
      "learning_rate": 1.1231185471141126e-06,
      "loss": 0.2425,
      "step": 2300
    },
    {
      "epoch": 1.6118881118881119,
      "grad_norm": 0.16193565878333754,
      "learning_rate": 1.1039220705961002e-06,
      "loss": 0.2277,
      "step": 2305
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.20673878902479834,
      "learning_rate": 1.0848706879118893e-06,
      "loss": 0.2475,
      "step": 2310
    },
    {
      "epoch": 1.6188811188811187,
      "grad_norm": 0.21816354127208326,
      "learning_rate": 1.065965108554e-06,
      "loss": 0.2578,
      "step": 2315
    },
    {
      "epoch": 1.6223776223776225,
      "grad_norm": 0.16456963901908442,
      "learning_rate": 1.047206036585095e-06,
      "loss": 0.2317,
      "step": 2320
    },
    {
      "epoch": 1.6258741258741258,
      "grad_norm": 0.2578576181305842,
      "learning_rate": 1.0285941706117486e-06,
      "loss": 0.2556,
      "step": 2325
    },
    {
      "epoch": 1.6293706293706294,
      "grad_norm": 0.20771441694982992,
      "learning_rate": 1.0101302037584427e-06,
      "loss": 0.2484,
      "step": 2330
    },
    {
      "epoch": 1.632867132867133,
      "grad_norm": 0.20558365348546553,
      "learning_rate": 9.918148236417453e-07,
      "loss": 0.2132,
      "step": 2335
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.1558966734301512,
      "learning_rate": 9.73648712344707e-07,
      "loss": 0.2203,
      "step": 2340
    },
    {
      "epoch": 1.63986013986014,
      "grad_norm": 0.2055595235565553,
      "learning_rate": 9.5563254639146e-07,
      "loss": 0.2185,
      "step": 2345
    },
    {
      "epoch": 1.6433566433566433,
      "grad_norm": 0.23036116631758255,
      "learning_rate": 9.377669967220166e-07,
      "loss": 0.2427,
      "step": 2350
    },
    {
      "epoch": 1.6468531468531469,
      "grad_norm": 0.2541374406766717,
      "learning_rate": 9.200527286672967e-07,
      "loss": 0.2406,
      "step": 2355
    },
    {
      "epoch": 1.6503496503496504,
      "grad_norm": 0.1965665252309361,
      "learning_rate": 9.02490401924333e-07,
      "loss": 0.237,
      "step": 2360
    },
    {
      "epoch": 1.6538461538461537,
      "grad_norm": 0.22022081232887075,
      "learning_rate": 8.850806705317183e-07,
      "loss": 0.2485,
      "step": 2365
    },
    {
      "epoch": 1.6573426573426573,
      "grad_norm": 0.27501820024516205,
      "learning_rate": 8.678241828452399e-07,
      "loss": 0.2407,
      "step": 2370
    },
    {
      "epoch": 1.6608391608391608,
      "grad_norm": 0.2068268227528754,
      "learning_rate": 8.507215815137337e-07,
      "loss": 0.233,
      "step": 2375
    },
    {
      "epoch": 1.6643356643356644,
      "grad_norm": 0.25620799000888017,
      "learning_rate": 8.337735034551614e-07,
      "loss": 0.2458,
      "step": 2380
    },
    {
      "epoch": 1.667832167832168,
      "grad_norm": 0.20951395283357943,
      "learning_rate": 8.169805798328745e-07,
      "loss": 0.2374,
      "step": 2385
    },
    {
      "epoch": 1.6713286713286712,
      "grad_norm": 0.1872465721165221,
      "learning_rate": 8.003434360321249e-07,
      "loss": 0.2465,
      "step": 2390
    },
    {
      "epoch": 1.6748251748251748,
      "grad_norm": 0.18145979976086574,
      "learning_rate": 7.838626916367675e-07,
      "loss": 0.2148,
      "step": 2395
    },
    {
      "epoch": 1.6783216783216783,
      "grad_norm": 0.24058782422084832,
      "learning_rate": 7.675389604061817e-07,
      "loss": 0.2345,
      "step": 2400
    },
    {
      "epoch": 1.6818181818181817,
      "grad_norm": 0.20087468348190377,
      "learning_rate": 7.513728502524286e-07,
      "loss": 0.2286,
      "step": 2405
    },
    {
      "epoch": 1.6853146853146854,
      "grad_norm": 0.17425447474287672,
      "learning_rate": 7.353649632175957e-07,
      "loss": 0.2256,
      "step": 2410
    },
    {
      "epoch": 1.6888111888111887,
      "grad_norm": 0.2396206296610538,
      "learning_rate": 7.195158954513864e-07,
      "loss": 0.225,
      "step": 2415
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 0.17549097854271423,
      "learning_rate": 7.03826237188916e-07,
      "loss": 0.2216,
      "step": 2420
    },
    {
      "epoch": 1.6958041958041958,
      "grad_norm": 0.22053185548722443,
      "learning_rate": 6.882965727287305e-07,
      "loss": 0.2295,
      "step": 2425
    },
    {
      "epoch": 1.6993006993006992,
      "grad_norm": 0.16904985185465962,
      "learning_rate": 6.72927480411047e-07,
      "loss": 0.2328,
      "step": 2430
    },
    {
      "epoch": 1.702797202797203,
      "grad_norm": 0.24716116115131842,
      "learning_rate": 6.577195325962138e-07,
      "loss": 0.2217,
      "step": 2435
    },
    {
      "epoch": 1.7062937062937062,
      "grad_norm": 0.21805674414937837,
      "learning_rate": 6.426732956433989e-07,
      "loss": 0.2389,
      "step": 2440
    },
    {
      "epoch": 1.7097902097902098,
      "grad_norm": 0.16493951842822327,
      "learning_rate": 6.277893298894966e-07,
      "loss": 0.2262,
      "step": 2445
    },
    {
      "epoch": 1.7132867132867133,
      "grad_norm": 0.20549852112013453,
      "learning_rate": 6.130681896282581e-07,
      "loss": 0.248,
      "step": 2450
    },
    {
      "epoch": 1.7167832167832167,
      "grad_norm": 0.24788701675807667,
      "learning_rate": 5.985104230896515e-07,
      "loss": 0.2464,
      "step": 2455
    },
    {
      "epoch": 1.7202797202797204,
      "grad_norm": 0.2019733596582574,
      "learning_rate": 5.841165724194453e-07,
      "loss": 0.2316,
      "step": 2460
    },
    {
      "epoch": 1.7237762237762237,
      "grad_norm": 0.23427331783210922,
      "learning_rate": 5.698871736590145e-07,
      "loss": 0.2353,
      "step": 2465
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.19882817499305466,
      "learning_rate": 5.558227567253832e-07,
      "loss": 0.2331,
      "step": 2470
    },
    {
      "epoch": 1.7307692307692308,
      "grad_norm": 0.1953244944246295,
      "learning_rate": 5.41923845391486e-07,
      "loss": 0.2249,
      "step": 2475
    },
    {
      "epoch": 1.7342657342657342,
      "grad_norm": 0.20860907438922652,
      "learning_rate": 5.281909572666649e-07,
      "loss": 0.2371,
      "step": 2480
    },
    {
      "epoch": 1.737762237762238,
      "grad_norm": 0.201292017090407,
      "learning_rate": 5.146246037773922e-07,
      "loss": 0.2223,
      "step": 2485
    },
    {
      "epoch": 1.7412587412587412,
      "grad_norm": 0.20531608712854288,
      "learning_rate": 5.012252901482211e-07,
      "loss": 0.2108,
      "step": 2490
    },
    {
      "epoch": 1.7447552447552448,
      "grad_norm": 0.24065289336782011,
      "learning_rate": 4.879935153829757e-07,
      "loss": 0.2454,
      "step": 2495
    },
    {
      "epoch": 1.7482517482517483,
      "grad_norm": 0.2410587912972817,
      "learning_rate": 4.74929772246166e-07,
      "loss": 0.2316,
      "step": 2500
    },
    {
      "epoch": 1.7517482517482517,
      "grad_norm": 0.20361234001274467,
      "learning_rate": 4.6203454724463523e-07,
      "loss": 0.217,
      "step": 2505
    },
    {
      "epoch": 1.7552447552447552,
      "grad_norm": 0.24316077385521911,
      "learning_rate": 4.4930832060944453e-07,
      "loss": 0.2519,
      "step": 2510
    },
    {
      "epoch": 1.7587412587412588,
      "grad_norm": 0.2588824914869855,
      "learning_rate": 4.36751566277982e-07,
      "loss": 0.2404,
      "step": 2515
    },
    {
      "epoch": 1.762237762237762,
      "grad_norm": 0.23339078143613176,
      "learning_rate": 4.243647518763261e-07,
      "loss": 0.2344,
      "step": 2520
    },
    {
      "epoch": 1.7657342657342658,
      "grad_norm": 0.1870709992934403,
      "learning_rate": 4.121483387018149e-07,
      "loss": 0.2027,
      "step": 2525
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.24240712525582014,
      "learning_rate": 4.001027817058789e-07,
      "loss": 0.2242,
      "step": 2530
    },
    {
      "epoch": 1.7727272727272727,
      "grad_norm": 0.1975158666317177,
      "learning_rate": 3.882285294770938e-07,
      "loss": 0.2323,
      "step": 2535
    },
    {
      "epoch": 1.7762237762237763,
      "grad_norm": 0.2848982157758639,
      "learning_rate": 3.765260242244706e-07,
      "loss": 0.2336,
      "step": 2540
    },
    {
      "epoch": 1.7797202797202796,
      "grad_norm": 0.18090559581095472,
      "learning_rate": 3.649957017609979e-07,
      "loss": 0.2309,
      "step": 2545
    },
    {
      "epoch": 1.7832167832167833,
      "grad_norm": 0.23858822129312376,
      "learning_rate": 3.536379914873983e-07,
      "loss": 0.2188,
      "step": 2550
    },
    {
      "epoch": 1.7867132867132867,
      "grad_norm": 0.237417705264017,
      "learning_rate": 3.4245331637614955e-07,
      "loss": 0.2231,
      "step": 2555
    },
    {
      "epoch": 1.7902097902097902,
      "grad_norm": 0.21649790024308277,
      "learning_rate": 3.3144209295572493e-07,
      "loss": 0.2329,
      "step": 2560
    },
    {
      "epoch": 1.7937062937062938,
      "grad_norm": 0.2390035333944304,
      "learning_rate": 3.206047312950811e-07,
      "loss": 0.2134,
      "step": 2565
    },
    {
      "epoch": 1.797202797202797,
      "grad_norm": 0.19422897750452517,
      "learning_rate": 3.099416349883949e-07,
      "loss": 0.2096,
      "step": 2570
    },
    {
      "epoch": 1.8006993006993008,
      "grad_norm": 0.2301108253195982,
      "learning_rate": 2.994532011400214e-07,
      "loss": 0.2298,
      "step": 2575
    },
    {
      "epoch": 1.8041958041958042,
      "grad_norm": 0.22166892323516244,
      "learning_rate": 2.8913982034971475e-07,
      "loss": 0.2089,
      "step": 2580
    },
    {
      "epoch": 1.8076923076923077,
      "grad_norm": 0.2627008526014087,
      "learning_rate": 2.790018766980773e-07,
      "loss": 0.2337,
      "step": 2585
    },
    {
      "epoch": 1.8111888111888113,
      "grad_norm": 0.19527688307645244,
      "learning_rate": 2.6903974773225703e-07,
      "loss": 0.2277,
      "step": 2590
    },
    {
      "epoch": 1.8146853146853146,
      "grad_norm": 0.24342278846709312,
      "learning_rate": 2.592538044518889e-07,
      "loss": 0.2122,
      "step": 2595
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.23354073502276002,
      "learning_rate": 2.4964441129527337e-07,
      "loss": 0.2366,
      "step": 2600
    },
    {
      "epoch": 1.8216783216783217,
      "grad_norm": 0.1762278220188555,
      "learning_rate": 2.4021192612581177e-07,
      "loss": 0.2325,
      "step": 2605
    },
    {
      "epoch": 1.8251748251748252,
      "grad_norm": 0.2630266024192207,
      "learning_rate": 2.30956700218673e-07,
      "loss": 0.2161,
      "step": 2610
    },
    {
      "epoch": 1.8286713286713288,
      "grad_norm": 0.22252967945056348,
      "learning_rate": 2.2187907824771538e-07,
      "loss": 0.2256,
      "step": 2615
    },
    {
      "epoch": 1.832167832167832,
      "grad_norm": 0.2652486743034812,
      "learning_rate": 2.1297939827264803e-07,
      "loss": 0.2344,
      "step": 2620
    },
    {
      "epoch": 1.8356643356643356,
      "grad_norm": 0.24375491611024352,
      "learning_rate": 2.0425799172644367e-07,
      "loss": 0.2071,
      "step": 2625
    },
    {
      "epoch": 1.8391608391608392,
      "grad_norm": 0.2279098129494666,
      "learning_rate": 1.9571518340299245e-07,
      "loss": 0.2189,
      "step": 2630
    },
    {
      "epoch": 1.8426573426573427,
      "grad_norm": 0.2725869066752257,
      "learning_rate": 1.873512914450104e-07,
      "loss": 0.1944,
      "step": 2635
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.23392292954377786,
      "learning_rate": 1.7916662733218848e-07,
      "loss": 0.2054,
      "step": 2640
    },
    {
      "epoch": 1.8496503496503496,
      "grad_norm": 0.23155242190332045,
      "learning_rate": 1.711614958695934e-07,
      "loss": 0.2137,
      "step": 2645
    },
    {
      "epoch": 1.8531468531468531,
      "grad_norm": 0.23009398789106972,
      "learning_rate": 1.6333619517631795e-07,
      "loss": 0.2266,
      "step": 2650
    },
    {
      "epoch": 1.8566433566433567,
      "grad_norm": 0.21611380202153502,
      "learning_rate": 1.5569101667437414e-07,
      "loss": 0.2273,
      "step": 2655
    },
    {
      "epoch": 1.86013986013986,
      "grad_norm": 0.2497907255920255,
      "learning_rate": 1.4822624507784932e-07,
      "loss": 0.2129,
      "step": 2660
    },
    {
      "epoch": 1.8636363636363638,
      "grad_norm": 0.21645263133309026,
      "learning_rate": 1.4094215838229176e-07,
      "loss": 0.2235,
      "step": 2665
    },
    {
      "epoch": 1.867132867132867,
      "grad_norm": 0.21828951546632747,
      "learning_rate": 1.3383902785436664e-07,
      "loss": 0.2149,
      "step": 2670
    },
    {
      "epoch": 1.8706293706293706,
      "grad_norm": 0.24393086475970577,
      "learning_rate": 1.2691711802175043e-07,
      "loss": 0.2008,
      "step": 2675
    },
    {
      "epoch": 1.8741258741258742,
      "grad_norm": 0.3188244652652861,
      "learning_rate": 1.2017668666327752e-07,
      "loss": 0.224,
      "step": 2680
    },
    {
      "epoch": 1.8776223776223775,
      "grad_norm": 0.3213240172913652,
      "learning_rate": 1.1361798479934737e-07,
      "loss": 0.2384,
      "step": 2685
    },
    {
      "epoch": 1.8811188811188813,
      "grad_norm": 0.1812445700799793,
      "learning_rate": 1.0724125668256647e-07,
      "loss": 0.2033,
      "step": 2690
    },
    {
      "epoch": 1.8846153846153846,
      "grad_norm": 0.22488642034657966,
      "learning_rate": 1.0104673978866164e-07,
      "loss": 0.2351,
      "step": 2695
    },
    {
      "epoch": 1.8881118881118881,
      "grad_norm": 0.21308817150324919,
      "learning_rate": 9.50346648076278e-08,
      "loss": 0.2177,
      "step": 2700
    },
    {
      "epoch": 1.8916083916083917,
      "grad_norm": 0.24780476048232253,
      "learning_rate": 8.920525563514249e-08,
      "loss": 0.2228,
      "step": 2705
    },
    {
      "epoch": 1.895104895104895,
      "grad_norm": 0.21720420251139552,
      "learning_rate": 8.355872936422759e-08,
      "loss": 0.2159,
      "step": 2710
    },
    {
      "epoch": 1.8986013986013988,
      "grad_norm": 0.25202284291362476,
      "learning_rate": 7.809529627716028e-08,
      "loss": 0.2367,
      "step": 2715
    },
    {
      "epoch": 1.902097902097902,
      "grad_norm": 0.19775164747072374,
      "learning_rate": 7.281515983764698e-08,
      "loss": 0.2072,
      "step": 2720
    },
    {
      "epoch": 1.9055944055944056,
      "grad_norm": 0.2562434879212032,
      "learning_rate": 6.771851668324226e-08,
      "loss": 0.2482,
      "step": 2725
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.26892430221524966,
      "learning_rate": 6.280555661802857e-08,
      "loss": 0.1994,
      "step": 2730
    },
    {
      "epoch": 1.9125874125874125,
      "grad_norm": 0.2887096340492057,
      "learning_rate": 5.807646260554578e-08,
      "loss": 0.2084,
      "step": 2735
    },
    {
      "epoch": 1.916083916083916,
      "grad_norm": 0.24777131941697597,
      "learning_rate": 5.353141076197888e-08,
      "loss": 0.2178,
      "step": 2740
    },
    {
      "epoch": 1.9195804195804196,
      "grad_norm": 0.22689623187545083,
      "learning_rate": 4.9170570349598713e-08,
      "loss": 0.2419,
      "step": 2745
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.2684480264633364,
      "learning_rate": 4.499410377045765e-08,
      "loss": 0.2267,
      "step": 2750
    },
    {
      "epoch": 1.9265734265734267,
      "grad_norm": 0.24416583221125596,
      "learning_rate": 4.100216656034328e-08,
      "loss": 0.2266,
      "step": 2755
    },
    {
      "epoch": 1.93006993006993,
      "grad_norm": 0.2302231356461665,
      "learning_rate": 3.7194907382985256e-08,
      "loss": 0.2166,
      "step": 2760
    },
    {
      "epoch": 1.9335664335664335,
      "grad_norm": 0.23467133093429907,
      "learning_rate": 3.357246802451752e-08,
      "loss": 0.2209,
      "step": 2765
    },
    {
      "epoch": 1.937062937062937,
      "grad_norm": 0.23392229422089317,
      "learning_rate": 3.013498338820031e-08,
      "loss": 0.1973,
      "step": 2770
    },
    {
      "epoch": 1.9405594405594404,
      "grad_norm": 0.20759708547706576,
      "learning_rate": 2.6882581489395264e-08,
      "loss": 0.2111,
      "step": 2775
    },
    {
      "epoch": 1.9440559440559442,
      "grad_norm": 0.20208456863131452,
      "learning_rate": 2.381538345079759e-08,
      "loss": 0.1976,
      "step": 2780
    },
    {
      "epoch": 1.9475524475524475,
      "grad_norm": 0.22727363631709724,
      "learning_rate": 2.0933503497926886e-08,
      "loss": 0.2331,
      "step": 2785
    },
    {
      "epoch": 1.951048951048951,
      "grad_norm": 0.2581882163809143,
      "learning_rate": 1.8237048954869997e-08,
      "loss": 0.2348,
      "step": 2790
    },
    {
      "epoch": 1.9545454545454546,
      "grad_norm": 0.2146854759288832,
      "learning_rate": 1.5726120240288632e-08,
      "loss": 0.2221,
      "step": 2795
    },
    {
      "epoch": 1.958041958041958,
      "grad_norm": 0.21344973354357963,
      "learning_rate": 1.3400810863675174e-08,
      "loss": 0.2164,
      "step": 2800
    },
    {
      "epoch": 1.9615384615384617,
      "grad_norm": 0.24854344588537242,
      "learning_rate": 1.1261207421874309e-08,
      "loss": 0.214,
      "step": 2805
    },
    {
      "epoch": 1.965034965034965,
      "grad_norm": 0.21448547178069746,
      "learning_rate": 9.307389595855088e-09,
      "loss": 0.2176,
      "step": 2810
    },
    {
      "epoch": 1.9685314685314685,
      "grad_norm": 0.20534250574992022,
      "learning_rate": 7.539430147745497e-09,
      "loss": 0.2149,
      "step": 2815
    },
    {
      "epoch": 1.972027972027972,
      "grad_norm": 0.2305338569267447,
      "learning_rate": 5.957394918120196e-09,
      "loss": 0.2336,
      "step": 2820
    },
    {
      "epoch": 1.9755244755244754,
      "grad_norm": 0.254113356726984,
      "learning_rate": 4.561342823553028e-09,
      "loss": 0.223,
      "step": 2825
    },
    {
      "epoch": 1.9790209790209792,
      "grad_norm": 0.23571564569853407,
      "learning_rate": 3.351325854417109e-09,
      "loss": 0.2357,
      "step": 2830
    },
    {
      "epoch": 1.9825174825174825,
      "grad_norm": 0.2321024664149865,
      "learning_rate": 2.327389072953601e-09,
      "loss": 0.2178,
      "step": 2835
    },
    {
      "epoch": 1.986013986013986,
      "grad_norm": 0.22568993609404986,
      "learning_rate": 1.4895706115908293e-09,
      "loss": 0.2094,
      "step": 2840
    },
    {
      "epoch": 1.9895104895104896,
      "grad_norm": 0.18478258906243253,
      "learning_rate": 8.379016715254207e-10,
      "loss": 0.2142,
      "step": 2845
    },
    {
      "epoch": 1.993006993006993,
      "grad_norm": 0.22552534213497075,
      "learning_rate": 3.724065215593431e-10,
      "loss": 0.2227,
      "step": 2850
    },
    {
      "epoch": 1.9965034965034965,
      "grad_norm": 0.242945516186064,
      "learning_rate": 9.310249719729491e-11,
      "loss": 0.2105,
      "step": 2855
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2576848560938581,
      "learning_rate": 0.0,
      "loss": 0.2218,
      "step": 2860
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.24351981282234192,
      "eval_runtime": 32.4748,
      "eval_samples_per_second": 21.37,
      "eval_steps_per_second": 5.358,
      "step": 2860
    },
    {
      "epoch": 2.0,
      "step": 2860,
      "total_flos": 6.695403983227322e+18,
      "train_loss": 0.32075889967121446,
      "train_runtime": 20338.0941,
      "train_samples_per_second": 6.75,
      "train_steps_per_second": 0.141
    }
  ],
  "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
}