{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.977827050997783,
  "eval_steps": 500,
  "global_step": 2250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004434589800443459,
      "grad_norm": 10.587126731872559,
      "learning_rate": 8.88888888888889e-07,
      "loss": 3.0058,
      "step": 1
    },
    {
      "epoch": 0.022172949002217297,
      "grad_norm": 10.391400337219238,
      "learning_rate": 4.444444444444445e-06,
      "loss": 3.0384,
      "step": 5
    },
    {
      "epoch": 0.04434589800443459,
      "grad_norm": 10.358064651489258,
      "learning_rate": 8.88888888888889e-06,
      "loss": 2.9329,
      "step": 10
    },
    {
      "epoch": 0.06651884700665188,
      "grad_norm": 9.945845603942871,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.8718,
      "step": 15
    },
    {
      "epoch": 0.08869179600886919,
      "grad_norm": 7.753511905670166,
      "learning_rate": 1.777777777777778e-05,
      "loss": 2.7228,
      "step": 20
    },
    {
      "epoch": 0.11086474501108648,
      "grad_norm": 3.560798406600952,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 2.5677,
      "step": 25
    },
    {
      "epoch": 0.13303769401330376,
      "grad_norm": 1.3562530279159546,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 2.4311,
      "step": 30
    },
    {
      "epoch": 0.15521064301552107,
      "grad_norm": 0.9185900092124939,
      "learning_rate": 3.111111111111111e-05,
      "loss": 2.335,
      "step": 35
    },
    {
      "epoch": 0.17738359201773837,
      "grad_norm": 0.7675415277481079,
      "learning_rate": 3.555555555555556e-05,
      "loss": 2.2184,
      "step": 40
    },
    {
      "epoch": 0.19955654101995565,
      "grad_norm": 0.6789694428443909,
      "learning_rate": 4e-05,
      "loss": 2.1433,
      "step": 45
    },
    {
      "epoch": 0.22172949002217296,
      "grad_norm": 0.583677351474762,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 2.0729,
      "step": 50
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.5190356969833374,
      "learning_rate": 4.888888888888889e-05,
      "loss": 1.9766,
      "step": 55
    },
    {
      "epoch": 0.2660753880266075,
      "grad_norm": 0.4500294327735901,
      "learning_rate": 5.333333333333333e-05,
      "loss": 1.9329,
      "step": 60
    },
    {
      "epoch": 0.28824833702882485,
      "grad_norm": 0.4370773136615753,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 1.8666,
      "step": 65
    },
    {
      "epoch": 0.31042128603104213,
      "grad_norm": 0.4395270049571991,
      "learning_rate": 6.222222222222222e-05,
      "loss": 1.806,
      "step": 70
    },
    {
      "epoch": 0.3325942350332594,
      "grad_norm": 0.3201618492603302,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.785,
      "step": 75
    },
    {
      "epoch": 0.35476718403547675,
      "grad_norm": 0.36038511991500854,
      "learning_rate": 7.111111111111112e-05,
      "loss": 1.7624,
      "step": 80
    },
    {
      "epoch": 0.376940133037694,
      "grad_norm": 0.32811540365219116,
      "learning_rate": 7.555555555555556e-05,
      "loss": 1.7255,
      "step": 85
    },
    {
      "epoch": 0.3991130820399113,
      "grad_norm": 0.3289714753627777,
      "learning_rate": 8e-05,
      "loss": 1.7171,
      "step": 90
    },
    {
      "epoch": 0.4212860310421286,
      "grad_norm": 0.3439875543117523,
      "learning_rate": 8.444444444444444e-05,
      "loss": 1.69,
      "step": 95
    },
    {
      "epoch": 0.4434589800443459,
      "grad_norm": 0.3265211880207062,
      "learning_rate": 8.888888888888889e-05,
      "loss": 1.6868,
      "step": 100
    },
    {
      "epoch": 0.4656319290465632,
      "grad_norm": 0.3699732720851898,
      "learning_rate": 9.333333333333334e-05,
      "loss": 1.6765,
      "step": 105
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.3702254891395569,
      "learning_rate": 9.777777777777778e-05,
      "loss": 1.6533,
      "step": 110
    },
    {
      "epoch": 0.5099778270509978,
      "grad_norm": 0.3689574599266052,
      "learning_rate": 0.00010222222222222222,
      "loss": 1.6488,
      "step": 115
    },
    {
      "epoch": 0.532150776053215,
      "grad_norm": 0.42295676469802856,
      "learning_rate": 0.00010666666666666667,
      "loss": 1.6136,
      "step": 120
    },
    {
      "epoch": 0.5543237250554324,
      "grad_norm": 0.4599798917770386,
      "learning_rate": 0.00011111111111111112,
      "loss": 1.5929,
      "step": 125
    },
    {
      "epoch": 0.5764966740576497,
      "grad_norm": 0.40791696310043335,
      "learning_rate": 0.00011555555555555555,
      "loss": 1.5782,
      "step": 130
    },
    {
      "epoch": 0.5986696230598669,
      "grad_norm": 0.5501047372817993,
      "learning_rate": 0.00012,
      "loss": 1.5679,
      "step": 135
    },
    {
      "epoch": 0.6208425720620843,
      "grad_norm": 0.4360973536968231,
      "learning_rate": 0.00012444444444444444,
      "loss": 1.5719,
      "step": 140
    },
    {
      "epoch": 0.6430155210643016,
      "grad_norm": 0.6553646922111511,
      "learning_rate": 0.00012888888888888892,
      "loss": 1.5643,
      "step": 145
    },
    {
      "epoch": 0.6651884700665188,
      "grad_norm": 0.6760438680648804,
      "learning_rate": 0.00013333333333333334,
      "loss": 1.5642,
      "step": 150
    },
    {
      "epoch": 0.6873614190687362,
      "grad_norm": 0.6198329329490662,
      "learning_rate": 0.0001377777777777778,
      "loss": 1.5522,
      "step": 155
    },
    {
      "epoch": 0.7095343680709535,
      "grad_norm": 0.5623511672019958,
      "learning_rate": 0.00014222222222222224,
      "loss": 1.5425,
      "step": 160
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.4507710337638855,
      "learning_rate": 0.00014666666666666666,
      "loss": 1.52,
      "step": 165
    },
    {
      "epoch": 0.753880266075388,
      "grad_norm": 0.4668518602848053,
      "learning_rate": 0.0001511111111111111,
      "loss": 1.5275,
      "step": 170
    },
    {
      "epoch": 0.7760532150776053,
      "grad_norm": 0.40650302171707153,
      "learning_rate": 0.00015555555555555556,
      "loss": 1.5356,
      "step": 175
    },
    {
      "epoch": 0.7982261640798226,
      "grad_norm": 0.47852110862731934,
      "learning_rate": 0.00016,
      "loss": 1.5367,
      "step": 180
    },
    {
      "epoch": 0.8203991130820399,
      "grad_norm": 0.41115882992744446,
      "learning_rate": 0.00016444444444444444,
      "loss": 1.5142,
      "step": 185
    },
    {
      "epoch": 0.8425720620842572,
      "grad_norm": 0.48749256134033203,
      "learning_rate": 0.00016888888888888889,
      "loss": 1.5132,
      "step": 190
    },
    {
      "epoch": 0.8647450110864745,
      "grad_norm": 0.45032379031181335,
      "learning_rate": 0.00017333333333333334,
      "loss": 1.4971,
      "step": 195
    },
    {
      "epoch": 0.8869179600886918,
      "grad_norm": 0.4163128435611725,
      "learning_rate": 0.00017777777777777779,
      "loss": 1.511,
      "step": 200
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.43136799335479736,
      "learning_rate": 0.00018222222222222224,
      "loss": 1.4945,
      "step": 205
    },
    {
      "epoch": 0.9312638580931264,
      "grad_norm": 0.4096197783946991,
      "learning_rate": 0.0001866666666666667,
      "loss": 1.4786,
      "step": 210
    },
    {
      "epoch": 0.9534368070953437,
      "grad_norm": 0.5415340065956116,
      "learning_rate": 0.00019111111111111114,
      "loss": 1.4733,
      "step": 215
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.5372570753097534,
      "learning_rate": 0.00019555555555555556,
      "loss": 1.4994,
      "step": 220
    },
    {
      "epoch": 0.9977827050997783,
      "grad_norm": 0.4384077787399292,
      "learning_rate": 0.0002,
      "loss": 1.4961,
      "step": 225
    },
    {
      "epoch": 0.9977827050997783,
      "eval_loss": 1.770818829536438,
      "eval_runtime": 0.3325,
      "eval_samples_per_second": 3.008,
      "eval_steps_per_second": 3.008,
      "step": 225
    },
    {
      "epoch": 1.0199556541019956,
      "grad_norm": 0.592540979385376,
      "learning_rate": 0.00019999699144797678,
      "loss": 1.4664,
      "step": 230
    },
    {
      "epoch": 1.042128603104213,
      "grad_norm": 0.4885026514530182,
      "learning_rate": 0.00019998796597293477,
      "loss": 1.4644,
      "step": 235
    },
    {
      "epoch": 1.06430155210643,
      "grad_norm": 0.3968167006969452,
      "learning_rate": 0.00019997292411794618,
      "loss": 1.4751,
      "step": 240
    },
    {
      "epoch": 1.0864745011086474,
      "grad_norm": 0.37674403190612793,
      "learning_rate": 0.00019995186678809513,
      "loss": 1.4596,
      "step": 245
    },
    {
      "epoch": 1.1086474501108647,
      "grad_norm": 0.4001663029193878,
      "learning_rate": 0.00019992479525042303,
      "loss": 1.4788,
      "step": 250
    },
    {
      "epoch": 1.130820399113082,
      "grad_norm": 0.3921445608139038,
      "learning_rate": 0.0001998917111338525,
      "loss": 1.4625,
      "step": 255
    },
    {
      "epoch": 1.1529933481152994,
      "grad_norm": 0.3944418132305145,
      "learning_rate": 0.00019985261642908918,
      "loss": 1.4488,
      "step": 260
    },
    {
      "epoch": 1.1751662971175167,
      "grad_norm": 0.4203372001647949,
      "learning_rate": 0.00019980751348850219,
      "loss": 1.4565,
      "step": 265
    },
    {
      "epoch": 1.1973392461197339,
      "grad_norm": 0.4535711109638214,
      "learning_rate": 0.00019975640502598244,
      "loss": 1.4549,
      "step": 270
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 0.41691839694976807,
      "learning_rate": 0.0001996992941167792,
      "loss": 1.432,
      "step": 275
    },
    {
      "epoch": 1.2416851441241685,
      "grad_norm": 0.38892751932144165,
      "learning_rate": 0.00019963618419731538,
      "loss": 1.4378,
      "step": 280
    },
    {
      "epoch": 1.2638580931263859,
      "grad_norm": 0.3704243302345276,
      "learning_rate": 0.00019956707906498044,
      "loss": 1.4547,
      "step": 285
    },
    {
      "epoch": 1.2860310421286032,
      "grad_norm": 0.4061298668384552,
      "learning_rate": 0.00019949198287790213,
      "loss": 1.4309,
      "step": 290
    },
    {
      "epoch": 1.3082039911308203,
      "grad_norm": 0.41798922419548035,
      "learning_rate": 0.00019941090015469613,
      "loss": 1.4253,
      "step": 295
    },
    {
      "epoch": 1.3303769401330376,
      "grad_norm": 0.4110168516635895,
      "learning_rate": 0.00019932383577419432,
      "loss": 1.4368,
      "step": 300
    },
    {
      "epoch": 1.352549889135255,
      "grad_norm": 0.43604379892349243,
      "learning_rate": 0.00019923079497515098,
      "loss": 1.429,
      "step": 305
    },
    {
      "epoch": 1.3747228381374723,
      "grad_norm": 0.3890831172466278,
      "learning_rate": 0.00019913178335592783,
      "loss": 1.4165,
      "step": 310
    },
    {
      "epoch": 1.3968957871396896,
      "grad_norm": 0.3936193585395813,
      "learning_rate": 0.00019902680687415705,
      "loss": 1.4299,
      "step": 315
    },
    {
      "epoch": 1.4190687361419068,
      "grad_norm": 0.4455225467681885,
      "learning_rate": 0.00019891587184638272,
      "loss": 1.4098,
      "step": 320
    },
    {
      "epoch": 1.441241685144124,
      "grad_norm": 0.3836926817893982,
      "learning_rate": 0.00019879898494768093,
      "loss": 1.4046,
      "step": 325
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 0.41238486766815186,
      "learning_rate": 0.00019867615321125795,
      "loss": 1.42,
      "step": 330
    },
    {
      "epoch": 1.4855875831485588,
      "grad_norm": 0.3580898940563202,
      "learning_rate": 0.00019854738402802716,
      "loss": 1.4199,
      "step": 335
    },
    {
      "epoch": 1.507760532150776,
      "grad_norm": 0.3761023283004761,
      "learning_rate": 0.00019841268514616433,
      "loss": 1.4222,
      "step": 340
    },
    {
      "epoch": 1.5299334811529932,
      "grad_norm": 0.3899005055427551,
      "learning_rate": 0.00019827206467064133,
      "loss": 1.4159,
      "step": 345
    },
    {
      "epoch": 1.5521064301552108,
      "grad_norm": 0.38497841358184814,
      "learning_rate": 0.00019812553106273847,
      "loss": 1.4089,
      "step": 350
    },
    {
      "epoch": 1.5742793791574279,
      "grad_norm": 0.3658756911754608,
      "learning_rate": 0.00019797309313953541,
      "loss": 1.4181,
      "step": 355
    },
    {
      "epoch": 1.5964523281596452,
      "grad_norm": 0.3782854378223419,
      "learning_rate": 0.00019781476007338058,
      "loss": 1.4143,
      "step": 360
    },
    {
      "epoch": 1.6186252771618626,
      "grad_norm": 0.5063923001289368,
      "learning_rate": 0.0001976505413913393,
      "loss": 1.4154,
      "step": 365
    },
    {
      "epoch": 1.6407982261640797,
      "grad_norm": 0.38294726610183716,
      "learning_rate": 0.00019748044697462058,
      "loss": 1.4268,
      "step": 370
    },
    {
      "epoch": 1.6629711751662972,
      "grad_norm": 0.35021841526031494,
      "learning_rate": 0.00019730448705798239,
      "loss": 1.4065,
      "step": 375
    },
    {
      "epoch": 1.6851441241685143,
      "grad_norm": 0.3741037845611572,
      "learning_rate": 0.00019712267222911606,
      "loss": 1.4079,
      "step": 380
    },
    {
      "epoch": 1.7073170731707317,
      "grad_norm": 0.3381967842578888,
      "learning_rate": 0.00019693501342800892,
      "loss": 1.4137,
      "step": 385
    },
    {
      "epoch": 1.729490022172949,
      "grad_norm": 0.38490810990333557,
      "learning_rate": 0.00019674152194628638,
      "loss": 1.4005,
      "step": 390
    },
    {
      "epoch": 1.7516629711751663,
      "grad_norm": 0.36805257201194763,
      "learning_rate": 0.0001965422094265322,
      "loss": 1.3987,
      "step": 395
    },
    {
      "epoch": 1.7738359201773837,
      "grad_norm": 0.3764079213142395,
      "learning_rate": 0.00019633708786158806,
      "loss": 1.4124,
      "step": 400
    },
    {
      "epoch": 1.7960088691796008,
      "grad_norm": 0.3698968291282654,
      "learning_rate": 0.0001961261695938319,
      "loss": 1.4082,
      "step": 405
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.38398101925849915,
      "learning_rate": 0.0001959094673144354,
      "loss": 1.3864,
      "step": 410
    },
    {
      "epoch": 1.8403547671840355,
      "grad_norm": 0.3575521409511566,
      "learning_rate": 0.00019568699406260015,
      "loss": 1.3982,
      "step": 415
    },
    {
      "epoch": 1.8625277161862528,
      "grad_norm": 0.4014054834842682,
      "learning_rate": 0.0001954587632247732,
      "loss": 1.3862,
      "step": 420
    },
    {
      "epoch": 1.8847006651884701,
      "grad_norm": 0.42844465374946594,
      "learning_rate": 0.00019522478853384155,
      "loss": 1.4102,
      "step": 425
    },
    {
      "epoch": 1.9068736141906872,
      "grad_norm": 0.43658122420310974,
      "learning_rate": 0.00019498508406830576,
      "loss": 1.4125,
      "step": 430
    },
    {
      "epoch": 1.9290465631929048,
      "grad_norm": 0.37285885214805603,
      "learning_rate": 0.00019473966425143292,
      "loss": 1.3878,
      "step": 435
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 0.39875471591949463,
      "learning_rate": 0.0001944885438503888,
      "loss": 1.3991,
      "step": 440
    },
    {
      "epoch": 1.9733924611973392,
      "grad_norm": 0.384593665599823,
      "learning_rate": 0.00019423173797534924,
      "loss": 1.3938,
      "step": 445
    },
    {
      "epoch": 1.9955654101995566,
      "grad_norm": 0.4125775098800659,
      "learning_rate": 0.00019396926207859084,
      "loss": 1.3952,
      "step": 450
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.7769551277160645,
      "eval_runtime": 0.3268,
      "eval_samples_per_second": 3.06,
      "eval_steps_per_second": 3.06,
      "step": 451
    },
    {
      "epoch": 2.0177383592017737,
      "grad_norm": 0.40251436829566956,
      "learning_rate": 0.0001937011319535615,
      "loss": 1.3874,
      "step": 455
    },
    {
      "epoch": 2.0399113082039912,
      "grad_norm": 0.4194943904876709,
      "learning_rate": 0.00019342736373392975,
      "loss": 1.3625,
      "step": 460
    },
    {
      "epoch": 2.0620842572062084,
      "grad_norm": 0.3550240099430084,
      "learning_rate": 0.00019314797389261424,
      "loss": 1.3776,
      "step": 465
    },
    {
      "epoch": 2.084257206208426,
      "grad_norm": 0.360365629196167,
      "learning_rate": 0.00019286297924079243,
      "loss": 1.3611,
      "step": 470
    },
    {
      "epoch": 2.106430155210643,
      "grad_norm": 0.39108407497406006,
      "learning_rate": 0.00019257239692688907,
      "loss": 1.3671,
      "step": 475
    },
    {
      "epoch": 2.12860310421286,
      "grad_norm": 0.4020346701145172,
      "learning_rate": 0.00019227624443554425,
      "loss": 1.3647,
      "step": 480
    },
    {
      "epoch": 2.1507760532150777,
      "grad_norm": 0.3887872099876404,
      "learning_rate": 0.00019197453958656158,
      "loss": 1.346,
      "step": 485
    },
    {
      "epoch": 2.172949002217295,
      "grad_norm": 0.374032199382782,
      "learning_rate": 0.00019166730053383568,
      "loss": 1.3442,
      "step": 490
    },
    {
      "epoch": 2.1951219512195124,
      "grad_norm": 0.3590153157711029,
      "learning_rate": 0.0001913545457642601,
      "loss": 1.3695,
      "step": 495
    },
    {
      "epoch": 2.2172949002217295,
      "grad_norm": 0.3909887969493866,
      "learning_rate": 0.0001910362940966147,
      "loss": 1.3629,
      "step": 500
    },
    {
      "epoch": 2.2394678492239466,
      "grad_norm": 0.35918113589286804,
      "learning_rate": 0.00019071256468043338,
      "loss": 1.3759,
      "step": 505
    },
    {
      "epoch": 2.261640798226164,
      "grad_norm": 0.3407536447048187,
      "learning_rate": 0.00019038337699485208,
      "loss": 1.3631,
      "step": 510
    },
    {
      "epoch": 2.2838137472283813,
      "grad_norm": 0.3984699845314026,
      "learning_rate": 0.00019004875084743624,
      "loss": 1.3561,
      "step": 515
    },
    {
      "epoch": 2.305986696230599,
      "grad_norm": 0.4278836250305176,
      "learning_rate": 0.00018970870637298934,
      "loss": 1.3511,
      "step": 520
    },
    {
      "epoch": 2.328159645232816,
      "grad_norm": 0.3488403260707855,
      "learning_rate": 0.00018936326403234125,
      "loss": 1.3579,
      "step": 525
    },
    {
      "epoch": 2.3503325942350335,
      "grad_norm": 0.3680429756641388,
      "learning_rate": 0.00018901244461111695,
      "loss": 1.3565,
      "step": 530
    },
    {
      "epoch": 2.3725055432372506,
      "grad_norm": 0.34674274921417236,
      "learning_rate": 0.00018865626921848615,
      "loss": 1.363,
      "step": 535
    },
    {
      "epoch": 2.3946784922394677,
      "grad_norm": 0.404840886592865,
      "learning_rate": 0.00018829475928589271,
      "loss": 1.3662,
      "step": 540
    },
    {
      "epoch": 2.4168514412416853,
      "grad_norm": 0.3971501588821411,
      "learning_rate": 0.00018792793656576543,
      "loss": 1.3684,
      "step": 545
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 0.4021252393722534,
      "learning_rate": 0.0001875558231302091,
      "loss": 1.3597,
      "step": 550
    },
    {
      "epoch": 2.4611973392461195,
      "grad_norm": 0.49484843015670776,
      "learning_rate": 0.00018717844136967624,
      "loss": 1.3632,
      "step": 555
    },
    {
      "epoch": 2.483370288248337,
      "grad_norm": 0.425786554813385,
      "learning_rate": 0.00018679581399162008,
      "loss": 1.3433,
      "step": 560
    },
    {
      "epoch": 2.505543237250554,
      "grad_norm": 0.35825788974761963,
      "learning_rate": 0.00018640796401912807,
      "loss": 1.3605,
      "step": 565
    },
    {
      "epoch": 2.5277161862527717,
      "grad_norm": 0.3355163335800171,
      "learning_rate": 0.00018601491478953657,
      "loss": 1.355,
      "step": 570
    },
    {
      "epoch": 2.549889135254989,
      "grad_norm": 0.3719523251056671,
      "learning_rate": 0.00018561668995302667,
      "loss": 1.3596,
      "step": 575
    },
    {
      "epoch": 2.5720620842572064,
      "grad_norm": 0.37250471115112305,
      "learning_rate": 0.00018521331347120114,
      "loss": 1.3569,
      "step": 580
    },
    {
      "epoch": 2.5942350332594235,
      "grad_norm": 0.3406711518764496,
      "learning_rate": 0.0001848048096156426,
      "loss": 1.3493,
      "step": 585
    },
    {
      "epoch": 2.6164079822616406,
      "grad_norm": 0.3547060489654541,
      "learning_rate": 0.00018439120296645308,
      "loss": 1.3702,
      "step": 590
    },
    {
      "epoch": 2.638580931263858,
      "grad_norm": 0.3630983531475067,
      "learning_rate": 0.000183972518410775,
      "loss": 1.3761,
      "step": 595
    },
    {
      "epoch": 2.6607538802660753,
      "grad_norm": 0.36963704228401184,
      "learning_rate": 0.00018354878114129367,
      "loss": 1.3614,
      "step": 600
    },
    {
      "epoch": 2.682926829268293,
      "grad_norm": 0.3835570812225342,
      "learning_rate": 0.00018312001665472146,
      "loss": 1.3464,
      "step": 605
    },
    {
      "epoch": 2.70509977827051,
      "grad_norm": 0.33551275730133057,
      "learning_rate": 0.00018268625075026375,
      "loss": 1.338,
      "step": 610
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.34833574295043945,
      "learning_rate": 0.00018224750952806624,
      "loss": 1.3413,
      "step": 615
    },
    {
      "epoch": 2.7494456762749446,
      "grad_norm": 0.3694487512111664,
      "learning_rate": 0.00018180381938764478,
      "loss": 1.3616,
      "step": 620
    },
    {
      "epoch": 2.7716186252771617,
      "grad_norm": 0.36651611328125,
      "learning_rate": 0.00018135520702629675,
      "loss": 1.3556,
      "step": 625
    },
    {
      "epoch": 2.7937915742793793,
      "grad_norm": 0.36336463689804077,
      "learning_rate": 0.00018090169943749476,
      "loss": 1.3505,
      "step": 630
    },
    {
      "epoch": 2.8159645232815964,
      "grad_norm": 0.3422740399837494,
      "learning_rate": 0.00018044332390926223,
      "loss": 1.3473,
      "step": 635
    },
    {
      "epoch": 2.8381374722838135,
      "grad_norm": 0.32323065400123596,
      "learning_rate": 0.0001799801080225316,
      "loss": 1.3328,
      "step": 640
    },
    {
      "epoch": 2.860310421286031,
      "grad_norm": 0.32829082012176514,
      "learning_rate": 0.0001795120796494848,
      "loss": 1.3483,
      "step": 645
    },
    {
      "epoch": 2.882483370288248,
      "grad_norm": 0.3510291576385498,
      "learning_rate": 0.00017903926695187595,
      "loss": 1.3489,
      "step": 650
    },
    {
      "epoch": 2.9046563192904657,
      "grad_norm": 0.36246925592422485,
      "learning_rate": 0.000178561698379337,
      "loss": 1.3463,
      "step": 655
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 0.3481374680995941,
      "learning_rate": 0.00017807940266766593,
      "loss": 1.358,
      "step": 660
    },
    {
      "epoch": 2.9490022172949004,
      "grad_norm": 0.32869285345077515,
      "learning_rate": 0.00017759240883709744,
      "loss": 1.3644,
      "step": 665
    },
    {
      "epoch": 2.9711751662971175,
      "grad_norm": 0.3213963806629181,
      "learning_rate": 0.00017710074619055707,
      "loss": 1.3428,
      "step": 670
    },
    {
      "epoch": 2.9933481152993346,
      "grad_norm": 0.3256833553314209,
      "learning_rate": 0.0001766044443118978,
      "loss": 1.3491,
      "step": 675
    },
    {
      "epoch": 2.9977827050997785,
      "eval_loss": 1.7484476566314697,
      "eval_runtime": 0.3309,
      "eval_samples_per_second": 3.022,
      "eval_steps_per_second": 3.022,
      "step": 676
    },
    {
      "epoch": 3.015521064301552,
      "grad_norm": 0.34128889441490173,
      "learning_rate": 0.00017610353306412007,
      "loss": 1.3328,
      "step": 680
    },
    {
      "epoch": 3.0376940133037693,
      "grad_norm": 0.34881681203842163,
      "learning_rate": 0.0001755980425875748,
      "loss": 1.3193,
      "step": 685
    },
    {
      "epoch": 3.059866962305987,
      "grad_norm": 0.3570633828639984,
      "learning_rate": 0.00017508800329814995,
      "loss": 1.3245,
      "step": 690
    },
    {
      "epoch": 3.082039911308204,
      "grad_norm": 0.33681634068489075,
      "learning_rate": 0.00017457344588544017,
      "loss": 1.3074,
      "step": 695
    },
    {
      "epoch": 3.104212860310421,
      "grad_norm": 0.34801220893859863,
      "learning_rate": 0.00017405440131090048,
      "loss": 1.3033,
      "step": 700
    },
    {
      "epoch": 3.1263858093126387,
      "grad_norm": 0.34218576550483704,
      "learning_rate": 0.0001735309008059829,
      "loss": 1.3135,
      "step": 705
    },
    {
      "epoch": 3.1485587583148558,
      "grad_norm": 0.37351498007774353,
      "learning_rate": 0.00017300297587025747,
      "loss": 1.3212,
      "step": 710
    },
    {
      "epoch": 3.1707317073170733,
      "grad_norm": 0.3948318064212799,
      "learning_rate": 0.00017247065826951695,
      "loss": 1.3142,
      "step": 715
    },
    {
      "epoch": 3.1929046563192904,
      "grad_norm": 0.35082143545150757,
      "learning_rate": 0.0001719339800338651,
      "loss": 1.3088,
      "step": 720
    },
    {
      "epoch": 3.2150776053215075,
      "grad_norm": 0.34117066860198975,
      "learning_rate": 0.00017139297345578994,
      "loss": 1.3219,
      "step": 725
    },
    {
      "epoch": 3.237250554323725,
      "grad_norm": 0.35672473907470703,
      "learning_rate": 0.00017084767108822,
      "loss": 1.3296,
      "step": 730
    },
    {
      "epoch": 3.259423503325942,
      "grad_norm": 0.37271085381507874,
      "learning_rate": 0.0001702981057425662,
      "loss": 1.3428,
      "step": 735
    },
    {
      "epoch": 3.2815964523281598,
      "grad_norm": 0.38227108120918274,
      "learning_rate": 0.00016974431048674715,
      "loss": 1.327,
      "step": 740
    },
    {
      "epoch": 3.303769401330377,
      "grad_norm": 0.34224027395248413,
      "learning_rate": 0.00016918631864319957,
      "loss": 1.3148,
      "step": 745
    },
    {
      "epoch": 3.3259423503325944,
      "grad_norm": 0.3518960177898407,
      "learning_rate": 0.0001686241637868734,
      "loss": 1.3175,
      "step": 750
    },
    {
      "epoch": 3.3481152993348116,
      "grad_norm": 0.3325146734714508,
      "learning_rate": 0.00016805787974321105,
      "loss": 1.2974,
      "step": 755
    },
    {
      "epoch": 3.3702882483370287,
      "grad_norm": 0.34566330909729004,
      "learning_rate": 0.00016748750058611278,
      "loss": 1.3285,
      "step": 760
    },
    {
      "epoch": 3.3924611973392462,
      "grad_norm": 0.4051843285560608,
      "learning_rate": 0.00016691306063588583,
      "loss": 1.3045,
      "step": 765
    },
    {
      "epoch": 3.4146341463414633,
      "grad_norm": 0.36097583174705505,
      "learning_rate": 0.00016633459445717974,
      "loss": 1.3254,
      "step": 770
    },
    {
      "epoch": 3.436807095343681,
      "grad_norm": 0.36703112721443176,
      "learning_rate": 0.0001657521368569064,
      "loss": 1.3131,
      "step": 775
    },
    {
      "epoch": 3.458980044345898,
      "grad_norm": 0.3478531539440155,
      "learning_rate": 0.00016516572288214552,
      "loss": 1.3242,
      "step": 780
    },
    {
      "epoch": 3.481152993348115,
      "grad_norm": 0.34251999855041504,
      "learning_rate": 0.00016457538781803623,
      "loss": 1.3092,
      "step": 785
    },
    {
      "epoch": 3.5033259423503327,
      "grad_norm": 0.3575889766216278,
      "learning_rate": 0.00016398116718565348,
      "loss": 1.3022,
      "step": 790
    },
    {
      "epoch": 3.52549889135255,
      "grad_norm": 0.36192360520362854,
      "learning_rate": 0.00016338309673987101,
      "loss": 1.316,
      "step": 795
    },
    {
      "epoch": 3.5476718403547673,
      "grad_norm": 0.3578505218029022,
      "learning_rate": 0.00016278121246720987,
      "loss": 1.3237,
      "step": 800
    },
    {
      "epoch": 3.5698447893569845,
      "grad_norm": 0.34649041295051575,
      "learning_rate": 0.0001621755505836729,
      "loss": 1.317,
      "step": 805
    },
    {
      "epoch": 3.5920177383592016,
      "grad_norm": 0.36542415618896484,
      "learning_rate": 0.0001615661475325658,
      "loss": 1.3146,
      "step": 810
    },
    {
      "epoch": 3.614190687361419,
      "grad_norm": 0.352225124835968,
      "learning_rate": 0.00016095303998230433,
      "loss": 1.3146,
      "step": 815
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.336407333612442,
      "learning_rate": 0.00016033626482420758,
      "loss": 1.3344,
      "step": 820
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 0.3517133295536041,
      "learning_rate": 0.00015971585917027862,
      "loss": 1.3175,
      "step": 825
    },
    {
      "epoch": 3.680709534368071,
      "grad_norm": 0.3601805567741394,
      "learning_rate": 0.00015909186035097111,
      "loss": 1.3143,
      "step": 830
    },
    {
      "epoch": 3.7028824833702885,
      "grad_norm": 0.33734479546546936,
      "learning_rate": 0.0001584643059129433,
      "loss": 1.3187,
      "step": 835
    },
    {
      "epoch": 3.7250554323725056,
      "grad_norm": 0.3447887897491455,
      "learning_rate": 0.00015783323361679864,
      "loss": 1.3058,
      "step": 840
    },
    {
      "epoch": 3.7472283813747227,
      "grad_norm": 0.332766056060791,
      "learning_rate": 0.00015719868143481384,
      "loss": 1.3096,
      "step": 845
    },
    {
      "epoch": 3.7694013303769403,
      "grad_norm": 0.348834753036499,
      "learning_rate": 0.00015656068754865387,
      "loss": 1.3168,
      "step": 850
    },
    {
      "epoch": 3.7915742793791574,
      "grad_norm": 0.35416728258132935,
      "learning_rate": 0.0001559192903470747,
      "loss": 1.3226,
      "step": 855
    },
    {
      "epoch": 3.8137472283813745,
      "grad_norm": 0.360474556684494,
      "learning_rate": 0.00015527452842361327,
      "loss": 1.3109,
      "step": 860
    },
    {
      "epoch": 3.835920177383592,
      "grad_norm": 0.33958253264427185,
      "learning_rate": 0.0001546264405742654,
      "loss": 1.3182,
      "step": 865
    },
    {
      "epoch": 3.858093126385809,
      "grad_norm": 0.35495725274086,
      "learning_rate": 0.0001539750657951513,
      "loss": 1.3166,
      "step": 870
    },
    {
      "epoch": 3.8802660753880267,
      "grad_norm": 0.3402771055698395,
      "learning_rate": 0.00015332044328016914,
      "loss": 1.3271,
      "step": 875
    },
    {
      "epoch": 3.902439024390244,
      "grad_norm": 0.3348073661327362,
      "learning_rate": 0.00015266261241863674,
      "loss": 1.3155,
      "step": 880
    },
    {
      "epoch": 3.9246119733924614,
      "grad_norm": 0.3448599576950073,
      "learning_rate": 0.00015200161279292155,
      "loss": 1.3353,
      "step": 885
    },
    {
      "epoch": 3.9467849223946785,
      "grad_norm": 0.34724217653274536,
      "learning_rate": 0.00015133748417605876,
      "loss": 1.3302,
      "step": 890
    },
    {
      "epoch": 3.9689578713968956,
      "grad_norm": 0.3367406725883484,
      "learning_rate": 0.00015067026652935823,
      "loss": 1.3155,
      "step": 895
    },
    {
      "epoch": 3.991130820399113,
      "grad_norm": 0.33979812264442444,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.3025,
      "step": 900
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.7902376651763916,
      "eval_runtime": 0.3268,
      "eval_samples_per_second": 3.06,
      "eval_steps_per_second": 3.06,
      "step": 902
    },
    {
      "epoch": 4.013303769401331,
      "grad_norm": 0.3638916313648224,
      "learning_rate": 0.00014932672491861854,
      "loss": 1.2902,
      "step": 905
    },
    {
      "epoch": 4.035476718403547,
      "grad_norm": 0.33381396532058716,
      "learning_rate": 0.000148650481796876,
      "loss": 1.2739,
      "step": 910
    },
    {
      "epoch": 4.057649667405765,
      "grad_norm": 0.3387119472026825,
      "learning_rate": 0.00014797131132502465,
      "loss": 1.2744,
      "step": 915
    },
    {
      "epoch": 4.0798226164079825,
      "grad_norm": 0.3494380712509155,
      "learning_rate": 0.00014728925436945838,
      "loss": 1.2835,
      "step": 920
    },
    {
      "epoch": 4.101995565410199,
      "grad_norm": 0.3876653015613556,
      "learning_rate": 0.0001466043519702539,
      "loss": 1.294,
      "step": 925
    },
    {
      "epoch": 4.124168514412417,
      "grad_norm": 0.3451765477657318,
      "learning_rate": 0.00014591664533870118,
      "loss": 1.2773,
      "step": 930
    },
    {
      "epoch": 4.146341463414634,
      "grad_norm": 0.34489771723747253,
      "learning_rate": 0.00014522617585482377,
      "loss": 1.2798,
      "step": 935
    },
    {
      "epoch": 4.168514412416852,
      "grad_norm": 0.3645501136779785,
      "learning_rate": 0.00014453298506488896,
      "loss": 1.2822,
      "step": 940
    },
    {
      "epoch": 4.1906873614190685,
      "grad_norm": 0.3511221706867218,
      "learning_rate": 0.00014383711467890774,
      "loss": 1.2959,
      "step": 945
    },
    {
      "epoch": 4.212860310421286,
      "grad_norm": 0.34734046459198,
      "learning_rate": 0.00014313860656812536,
      "loss": 1.2804,
      "step": 950
    },
    {
      "epoch": 4.235033259423504,
      "grad_norm": 0.34553733468055725,
      "learning_rate": 0.00014243750276250153,
      "loss": 1.2938,
      "step": 955
    },
    {
      "epoch": 4.25720620842572,
      "grad_norm": 0.3475314676761627,
      "learning_rate": 0.0001417338454481818,
      "loss": 1.2882,
      "step": 960
    },
    {
      "epoch": 4.279379157427938,
      "grad_norm": 0.3424859941005707,
      "learning_rate": 0.00014102767696495884,
      "loss": 1.2957,
      "step": 965
    },
    {
      "epoch": 4.301552106430155,
      "grad_norm": 0.35711121559143066,
      "learning_rate": 0.00014031903980372504,
      "loss": 1.3019,
      "step": 970
    },
    {
      "epoch": 4.323725055432373,
      "grad_norm": 0.3883892595767975,
      "learning_rate": 0.0001396079766039157,
      "loss": 1.2627,
      "step": 975
    },
    {
      "epoch": 4.34589800443459,
      "grad_norm": 0.3351362347602844,
      "learning_rate": 0.00013889453015094338,
      "loss": 1.2894,
      "step": 980
    },
    {
      "epoch": 4.368070953436807,
      "grad_norm": 0.35356107354164124,
      "learning_rate": 0.0001381787433736235,
      "loss": 1.2972,
      "step": 985
    },
    {
      "epoch": 4.390243902439025,
      "grad_norm": 0.34767022728919983,
      "learning_rate": 0.00013746065934159123,
      "loss": 1.2728,
      "step": 990
    },
    {
      "epoch": 4.412416851441241,
      "grad_norm": 0.35094380378723145,
      "learning_rate": 0.00013674032126270982,
      "loss": 1.2875,
      "step": 995
    },
    {
      "epoch": 4.434589800443459,
      "grad_norm": 0.342939555644989,
      "learning_rate": 0.00013601777248047105,
      "loss": 1.2851,
      "step": 1000
    },
    {
      "epoch": 4.4567627494456765,
      "grad_norm": 0.37955859303474426,
      "learning_rate": 0.00013529305647138687,
      "loss": 1.2848,
      "step": 1005
    },
    {
      "epoch": 4.478935698447893,
      "grad_norm": 0.3524990677833557,
      "learning_rate": 0.00013456621684237367,
      "loss": 1.2889,
      "step": 1010
    },
    {
      "epoch": 4.501108647450111,
      "grad_norm": 0.354220986366272,
      "learning_rate": 0.00013383729732812814,
      "loss": 1.2989,
      "step": 1015
    },
    {
      "epoch": 4.523281596452328,
      "grad_norm": 0.34707173705101013,
      "learning_rate": 0.0001331063417884958,
      "loss": 1.2899,
      "step": 1020
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.358018159866333,
      "learning_rate": 0.00013237339420583212,
      "loss": 1.2926,
      "step": 1025
    },
    {
      "epoch": 4.5676274944567625,
      "grad_norm": 0.3700306713581085,
      "learning_rate": 0.00013163849868235564,
      "loss": 1.2942,
      "step": 1030
    },
    {
      "epoch": 4.58980044345898,
      "grad_norm": 0.38958704471588135,
      "learning_rate": 0.00013090169943749476,
      "loss": 1.2684,
      "step": 1035
    },
    {
      "epoch": 4.611973392461198,
      "grad_norm": 0.34153124690055847,
      "learning_rate": 0.00013016304080522656,
      "loss": 1.2885,
      "step": 1040
    },
    {
      "epoch": 4.634146341463414,
      "grad_norm": 0.3536767363548279,
      "learning_rate": 0.00012942256723140952,
      "loss": 1.299,
      "step": 1045
    },
    {
      "epoch": 4.656319290465632,
      "grad_norm": 0.34350860118865967,
      "learning_rate": 0.00012868032327110904,
      "loss": 1.2869,
      "step": 1050
    },
    {
      "epoch": 4.678492239467849,
      "grad_norm": 0.3380618691444397,
      "learning_rate": 0.00012793635358591645,
      "loss": 1.2935,
      "step": 1055
    },
    {
      "epoch": 4.700665188470067,
      "grad_norm": 0.3540153205394745,
      "learning_rate": 0.00012719070294126182,
      "loss": 1.2875,
      "step": 1060
    },
    {
      "epoch": 4.722838137472284,
      "grad_norm": 0.33930808305740356,
      "learning_rate": 0.00012644341620372023,
      "loss": 1.285,
      "step": 1065
    },
    {
      "epoch": 4.745011086474501,
      "grad_norm": 0.38791534304618835,
      "learning_rate": 0.00012569453833831222,
      "loss": 1.2695,
      "step": 1070
    },
    {
      "epoch": 4.767184035476719,
      "grad_norm": 0.34770122170448303,
      "learning_rate": 0.00012494411440579814,
      "loss": 1.3086,
      "step": 1075
    },
    {
      "epoch": 4.789356984478935,
      "grad_norm": 0.3471126854419708,
      "learning_rate": 0.00012419218955996676,
      "loss": 1.2935,
      "step": 1080
    },
    {
      "epoch": 4.811529933481153,
      "grad_norm": 0.33490991592407227,
      "learning_rate": 0.00012343880904491848,
      "loss": 1.2831,
      "step": 1085
    },
    {
      "epoch": 4.8337028824833705,
      "grad_norm": 0.3470149040222168,
      "learning_rate": 0.0001226840181923427,
      "loss": 1.2775,
      "step": 1090
    },
    {
      "epoch": 4.855875831485587,
      "grad_norm": 0.33513540029525757,
      "learning_rate": 0.00012192786241879033,
      "loss": 1.2868,
      "step": 1095
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 0.3534790873527527,
      "learning_rate": 0.0001211703872229411,
      "loss": 1.3015,
      "step": 1100
    },
    {
      "epoch": 4.900221729490022,
      "grad_norm": 0.42340680956840515,
      "learning_rate": 0.00012041163818286559,
      "loss": 1.2753,
      "step": 1105
    },
    {
      "epoch": 4.922394678492239,
      "grad_norm": 0.343871146440506,
      "learning_rate": 0.00011965166095328301,
      "loss": 1.2776,
      "step": 1110
    },
    {
      "epoch": 4.9445676274944566,
      "grad_norm": 0.3540928363800049,
      "learning_rate": 0.00011889050126281405,
      "loss": 1.3005,
      "step": 1115
    },
    {
      "epoch": 4.966740576496674,
      "grad_norm": 0.3478156626224518,
      "learning_rate": 0.00011812820491122918,
      "loss": 1.2888,
      "step": 1120
    },
    {
      "epoch": 4.988913525498892,
      "grad_norm": 0.35215553641319275,
      "learning_rate": 0.00011736481776669306,
      "loss": 1.2904,
      "step": 1125
    },
    {
      "epoch": 4.9977827050997785,
      "eval_loss": 1.79966139793396,
      "eval_runtime": 0.3309,
      "eval_samples_per_second": 3.022,
      "eval_steps_per_second": 3.022,
      "step": 1127
    },
    {
      "epoch": 5.011086474501108,
      "grad_norm": 0.36425259709358215,
      "learning_rate": 0.00011660038576300443,
      "loss": 1.2833,
      "step": 1130
    },
    {
      "epoch": 5.033259423503326,
      "grad_norm": 0.36249682307243347,
      "learning_rate": 0.00011583495489683229,
      "loss": 1.2641,
      "step": 1135
    },
    {
      "epoch": 5.0554323725055434,
      "grad_norm": 0.35734277963638306,
      "learning_rate": 0.00011506857122494831,
      "loss": 1.2565,
      "step": 1140
    },
    {
      "epoch": 5.07760532150776,
      "grad_norm": 0.3540472686290741,
      "learning_rate": 0.00011430128086145542,
      "loss": 1.2788,
      "step": 1145
    },
    {
      "epoch": 5.099778270509978,
      "grad_norm": 0.3659023940563202,
      "learning_rate": 0.00011353312997501313,
      "loss": 1.2464,
      "step": 1150
    },
    {
      "epoch": 5.121951219512195,
      "grad_norm": 0.3884827494621277,
      "learning_rate": 0.00011276416478605949,
      "loss": 1.2621,
      "step": 1155
    },
    {
      "epoch": 5.144124168514413,
      "grad_norm": 0.35218942165374756,
      "learning_rate": 0.00011199443156402998,
      "loss": 1.2643,
      "step": 1160
    },
    {
      "epoch": 5.1662971175166295,
      "grad_norm": 0.35376182198524475,
      "learning_rate": 0.0001112239766245735,
      "loss": 1.25,
      "step": 1165
    },
    {
      "epoch": 5.188470066518847,
      "grad_norm": 0.3779812455177307,
      "learning_rate": 0.00011045284632676536,
      "loss": 1.2623,
      "step": 1170
    },
    {
      "epoch": 5.210643015521065,
      "grad_norm": 0.40983307361602783,
      "learning_rate": 0.00010968108707031792,
      "loss": 1.2573,
      "step": 1175
    },
    {
      "epoch": 5.232815964523281,
      "grad_norm": 0.36376869678497314,
      "learning_rate": 0.00010890874529278865,
      "loss": 1.2556,
      "step": 1180
    },
    {
      "epoch": 5.254988913525499,
      "grad_norm": 0.36532121896743774,
      "learning_rate": 0.00010813586746678583,
      "loss": 1.2672,
      "step": 1185
    },
    {
      "epoch": 5.277161862527716,
      "grad_norm": 0.3567180633544922,
      "learning_rate": 0.00010736250009717247,
      "loss": 1.2623,
      "step": 1190
    },
    {
      "epoch": 5.299334811529933,
      "grad_norm": 0.3723650276660919,
      "learning_rate": 0.00010658868971826785,
      "loss": 1.2609,
      "step": 1195
    },
    {
      "epoch": 5.321507760532151,
      "grad_norm": 0.36513036489486694,
      "learning_rate": 0.00010581448289104758,
      "loss": 1.2605,
      "step": 1200
    },
    {
      "epoch": 5.343680709534368,
      "grad_norm": 0.34975701570510864,
      "learning_rate": 0.00010503992620034202,
      "loss": 1.258,
      "step": 1205
    },
    {
      "epoch": 5.365853658536586,
      "grad_norm": 0.3734095096588135,
      "learning_rate": 0.00010426506625203307,
      "loss": 1.2722,
      "step": 1210
    },
    {
      "epoch": 5.388026607538802,
      "grad_norm": 0.3856619596481323,
      "learning_rate": 0.00010348994967025012,
      "loss": 1.2637,
      "step": 1215
    },
    {
      "epoch": 5.41019955654102,
      "grad_norm": 0.37565699219703674,
      "learning_rate": 0.0001027146230945643,
      "loss": 1.2559,
      "step": 1220
    },
    {
      "epoch": 5.4323725055432375,
      "grad_norm": 0.37365949153900146,
      "learning_rate": 0.00010193913317718244,
      "loss": 1.2604,
      "step": 1225
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.35709646344184875,
      "learning_rate": 0.00010116352658013973,
      "loss": 1.2677,
      "step": 1230
    },
    {
      "epoch": 5.476718403547672,
      "grad_norm": 0.3517821133136749,
      "learning_rate": 0.00010038784997249205,
      "loss": 1.2496,
      "step": 1235
    },
    {
      "epoch": 5.498891352549889,
      "grad_norm": 0.37304961681365967,
      "learning_rate": 9.961215002750799e-05,
      "loss": 1.2818,
      "step": 1240
    },
    {
      "epoch": 5.521064301552107,
      "grad_norm": 0.3745460510253906,
      "learning_rate": 9.883647341986032e-05,
      "loss": 1.2817,
      "step": 1245
    },
    {
      "epoch": 5.5432372505543235,
      "grad_norm": 0.35062116384506226,
      "learning_rate": 9.806086682281758e-05,
      "loss": 1.2515,
      "step": 1250
    },
    {
      "epoch": 5.565410199556541,
      "grad_norm": 0.3595265746116638,
      "learning_rate": 9.728537690543572e-05,
      "loss": 1.2665,
      "step": 1255
    },
    {
      "epoch": 5.587583148558759,
      "grad_norm": 0.36035382747650146,
      "learning_rate": 9.651005032974994e-05,
      "loss": 1.2578,
      "step": 1260
    },
    {
      "epoch": 5.609756097560975,
      "grad_norm": 0.3572876751422882,
      "learning_rate": 9.573493374796693e-05,
      "loss": 1.2488,
      "step": 1265
    },
    {
      "epoch": 5.631929046563193,
      "grad_norm": 0.3515053391456604,
      "learning_rate": 9.496007379965801e-05,
      "loss": 1.2483,
      "step": 1270
    },
    {
      "epoch": 5.65410199556541,
      "grad_norm": 0.36254850029945374,
      "learning_rate": 9.418551710895243e-05,
      "loss": 1.2709,
      "step": 1275
    },
    {
      "epoch": 5.676274944567627,
      "grad_norm": 0.36927610635757446,
      "learning_rate": 9.341131028173214e-05,
      "loss": 1.2587,
      "step": 1280
    },
    {
      "epoch": 5.698447893569845,
      "grad_norm": 0.38191670179367065,
      "learning_rate": 9.263749990282754e-05,
      "loss": 1.2712,
      "step": 1285
    },
    {
      "epoch": 5.720620842572062,
      "grad_norm": 0.376602441072464,
      "learning_rate": 9.186413253321418e-05,
      "loss": 1.2758,
      "step": 1290
    },
    {
      "epoch": 5.74279379157428,
      "grad_norm": 0.3506106734275818,
      "learning_rate": 9.10912547072114e-05,
      "loss": 1.2704,
      "step": 1295
    },
    {
      "epoch": 5.764966740576496,
      "grad_norm": 0.38899239897727966,
      "learning_rate": 9.03189129296821e-05,
      "loss": 1.2454,
      "step": 1300
    },
    {
      "epoch": 5.787139689578714,
      "grad_norm": 0.34721580147743225,
      "learning_rate": 8.954715367323468e-05,
      "loss": 1.2823,
      "step": 1305
    },
    {
      "epoch": 5.8093126385809315,
      "grad_norm": 0.37420088052749634,
      "learning_rate": 8.877602337542655e-05,
      "loss": 1.2583,
      "step": 1310
    },
    {
      "epoch": 5.831485587583149,
      "grad_norm": 0.3575841784477234,
      "learning_rate": 8.800556843597002e-05,
      "loss": 1.2654,
      "step": 1315
    },
    {
      "epoch": 5.853658536585366,
      "grad_norm": 0.36751991510391235,
      "learning_rate": 8.723583521394054e-05,
      "loss": 1.259,
      "step": 1320
    },
    {
      "epoch": 5.875831485587583,
      "grad_norm": 0.41278401017189026,
      "learning_rate": 8.646687002498692e-05,
      "loss": 1.2639,
      "step": 1325
    },
    {
      "epoch": 5.898004434589801,
      "grad_norm": 0.36573198437690735,
      "learning_rate": 8.569871913854458e-05,
      "loss": 1.2682,
      "step": 1330
    },
    {
      "epoch": 5.9201773835920175,
      "grad_norm": 0.3588218092918396,
      "learning_rate": 8.49314287750517e-05,
      "loss": 1.2461,
      "step": 1335
    },
    {
      "epoch": 5.942350332594235,
      "grad_norm": 0.3589507043361664,
      "learning_rate": 8.416504510316773e-05,
      "loss": 1.2569,
      "step": 1340
    },
    {
      "epoch": 5.964523281596453,
      "grad_norm": 0.3812873959541321,
      "learning_rate": 8.339961423699562e-05,
      "loss": 1.2496,
      "step": 1345
    },
    {
      "epoch": 5.986696230598669,
      "grad_norm": 0.35918959975242615,
      "learning_rate": 8.263518223330697e-05,
      "loss": 1.2729,
      "step": 1350
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.817025065422058,
      "eval_runtime": 0.3269,
      "eval_samples_per_second": 3.059,
      "eval_steps_per_second": 3.059,
      "step": 1353
    },
    {
      "epoch": 6.008869179600887,
      "grad_norm": 0.365013062953949,
      "learning_rate": 8.187179508877085e-05,
      "loss": 1.2526,
      "step": 1355
    },
    {
      "epoch": 6.031042128603104,
      "grad_norm": 0.3869335949420929,
      "learning_rate": 8.1109498737186e-05,
      "loss": 1.2482,
      "step": 1360
    },
    {
      "epoch": 6.053215077605321,
      "grad_norm": 0.3877875804901123,
      "learning_rate": 8.034833904671698e-05,
      "loss": 1.2583,
      "step": 1365
    },
    {
      "epoch": 6.075388026607539,
      "grad_norm": 0.3739064931869507,
      "learning_rate": 7.958836181713445e-05,
      "loss": 1.2355,
      "step": 1370
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 0.37528711557388306,
      "learning_rate": 7.882961277705895e-05,
      "loss": 1.2267,
      "step": 1375
    },
    {
      "epoch": 6.119733924611974,
      "grad_norm": 0.36762019991874695,
      "learning_rate": 7.807213758120966e-05,
      "loss": 1.2174,
      "step": 1380
    },
    {
      "epoch": 6.14190687361419,
      "grad_norm": 0.36748361587524414,
      "learning_rate": 7.731598180765732e-05,
      "loss": 1.2372,
      "step": 1385
    },
    {
      "epoch": 6.164079822616408,
      "grad_norm": 0.3874417245388031,
      "learning_rate": 7.656119095508154e-05,
      "loss": 1.2373,
      "step": 1390
    },
    {
      "epoch": 6.1862527716186255,
      "grad_norm": 0.37527891993522644,
      "learning_rate": 7.580781044003324e-05,
      "loss": 1.2413,
      "step": 1395
    },
    {
      "epoch": 6.208425720620842,
      "grad_norm": 0.35388851165771484,
      "learning_rate": 7.505588559420189e-05,
      "loss": 1.2348,
      "step": 1400
    },
    {
      "epoch": 6.23059866962306,
      "grad_norm": 0.37881141901016235,
      "learning_rate": 7.43054616616878e-05,
      "loss": 1.2617,
      "step": 1405
    },
    {
      "epoch": 6.252771618625277,
      "grad_norm": 0.36428573727607727,
      "learning_rate": 7.35565837962798e-05,
      "loss": 1.2363,
      "step": 1410
    },
    {
      "epoch": 6.274944567627495,
      "grad_norm": 0.373054176568985,
      "learning_rate": 7.280929705873818e-05,
      "loss": 1.2405,
      "step": 1415
    },
    {
      "epoch": 6.2971175166297115,
      "grad_norm": 0.3633226454257965,
      "learning_rate": 7.206364641408357e-05,
      "loss": 1.2453,
      "step": 1420
    },
    {
      "epoch": 6.319290465631929,
      "grad_norm": 0.3570459485054016,
      "learning_rate": 7.131967672889101e-05,
      "loss": 1.2399,
      "step": 1425
    },
    {
      "epoch": 6.341463414634147,
      "grad_norm": 0.3895583748817444,
      "learning_rate": 7.057743276859048e-05,
      "loss": 1.2508,
      "step": 1430
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.3734217882156372,
      "learning_rate": 6.983695919477345e-05,
      "loss": 1.2461,
      "step": 1435
    },
    {
      "epoch": 6.385809312638581,
      "grad_norm": 0.35446175932884216,
      "learning_rate": 6.909830056250527e-05,
      "loss": 1.2571,
      "step": 1440
    },
    {
      "epoch": 6.407982261640798,
      "grad_norm": 0.35295209288597107,
      "learning_rate": 6.836150131764434e-05,
      "loss": 1.2395,
      "step": 1445
    },
    {
      "epoch": 6.430155210643015,
      "grad_norm": 0.38218680024147034,
      "learning_rate": 6.762660579416791e-05,
      "loss": 1.234,
      "step": 1450
    },
    {
      "epoch": 6.452328159645233,
      "grad_norm": 0.38666704297065735,
      "learning_rate": 6.68936582115042e-05,
      "loss": 1.2461,
      "step": 1455
    },
    {
      "epoch": 6.47450110864745,
      "grad_norm": 0.35704880952835083,
      "learning_rate": 6.61627026718719e-05,
      "loss": 1.236,
      "step": 1460
    },
    {
      "epoch": 6.496674057649668,
      "grad_norm": 0.37637725472450256,
      "learning_rate": 6.543378315762634e-05,
      "loss": 1.2354,
      "step": 1465
    },
    {
      "epoch": 6.518847006651884,
      "grad_norm": 0.37539881467819214,
      "learning_rate": 6.470694352861312e-05,
      "loss": 1.2471,
      "step": 1470
    },
    {
      "epoch": 6.541019955654102,
      "grad_norm": 0.3633996844291687,
      "learning_rate": 6.398222751952899e-05,
      "loss": 1.2298,
      "step": 1475
    },
    {
      "epoch": 6.5631929046563195,
      "grad_norm": 0.35865873098373413,
      "learning_rate": 6.325967873729018e-05,
      "loss": 1.2382,
      "step": 1480
    },
    {
      "epoch": 6.585365853658536,
      "grad_norm": 0.37201085686683655,
      "learning_rate": 6.25393406584088e-05,
      "loss": 1.2366,
      "step": 1485
    },
    {
      "epoch": 6.607538802660754,
      "grad_norm": 0.38256916403770447,
      "learning_rate": 6.18212566263765e-05,
      "loss": 1.2415,
      "step": 1490
    },
    {
      "epoch": 6.629711751662971,
      "grad_norm": 0.3845401704311371,
      "learning_rate": 6.110546984905661e-05,
      "loss": 1.2381,
      "step": 1495
    },
    {
      "epoch": 6.651884700665189,
      "grad_norm": 0.36959657073020935,
      "learning_rate": 6.039202339608432e-05,
      "loss": 1.2406,
      "step": 1500
    },
    {
      "epoch": 6.674057649667406,
      "grad_norm": 0.3971165120601654,
      "learning_rate": 5.9680960196274994e-05,
      "loss": 1.2421,
      "step": 1505
    },
    {
      "epoch": 6.696230598669623,
      "grad_norm": 0.3609052002429962,
      "learning_rate": 5.89723230350412e-05,
      "loss": 1.2359,
      "step": 1510
    },
    {
      "epoch": 6.718403547671841,
      "grad_norm": 0.3867206871509552,
      "learning_rate": 5.8266154551818216e-05,
      "loss": 1.2419,
      "step": 1515
    },
    {
      "epoch": 6.740576496674057,
      "grad_norm": 0.35991984605789185,
      "learning_rate": 5.756249723749847e-05,
      "loss": 1.2379,
      "step": 1520
    },
    {
      "epoch": 6.762749445676275,
      "grad_norm": 0.3713914752006531,
      "learning_rate": 5.6861393431874675e-05,
      "loss": 1.2346,
      "step": 1525
    },
    {
      "epoch": 6.7849223946784925,
      "grad_norm": 0.37790387868881226,
      "learning_rate": 5.616288532109225e-05,
      "loss": 1.253,
      "step": 1530
    },
    {
      "epoch": 6.807095343680709,
      "grad_norm": 0.39418914914131165,
      "learning_rate": 5.546701493511106e-05,
      "loss": 1.2572,
      "step": 1535
    },
    {
      "epoch": 6.829268292682927,
      "grad_norm": 0.3760906159877777,
      "learning_rate": 5.477382414517624e-05,
      "loss": 1.2436,
      "step": 1540
    },
    {
      "epoch": 6.851441241685144,
      "grad_norm": 0.37196049094200134,
      "learning_rate": 5.4083354661298814e-05,
      "loss": 1.2525,
      "step": 1545
    },
    {
      "epoch": 6.873614190687362,
      "grad_norm": 0.37352004647254944,
      "learning_rate": 5.339564802974615e-05,
      "loss": 1.2381,
      "step": 1550
    },
    {
      "epoch": 6.8957871396895785,
      "grad_norm": 0.3716587722301483,
      "learning_rate": 5.2710745630541666e-05,
      "loss": 1.2383,
      "step": 1555
    },
    {
      "epoch": 6.917960088691796,
      "grad_norm": 0.3760850131511688,
      "learning_rate": 5.2028688674975415e-05,
      "loss": 1.244,
      "step": 1560
    },
    {
      "epoch": 6.940133037694014,
      "grad_norm": 0.36280307173728943,
      "learning_rate": 5.134951820312401e-05,
      "loss": 1.2391,
      "step": 1565
    },
    {
      "epoch": 6.96230598669623,
      "grad_norm": 0.36244943737983704,
      "learning_rate": 5.0673275081381475e-05,
      "loss": 1.2465,
      "step": 1570
    },
    {
      "epoch": 6.984478935698448,
      "grad_norm": 0.3694084584712982,
      "learning_rate": 5.000000000000002e-05,
      "loss": 1.2451,
      "step": 1575
    },
    {
      "epoch": 6.9977827050997785,
      "eval_loss": 1.8179672956466675,
      "eval_runtime": 0.3324,
      "eval_samples_per_second": 3.008,
      "eval_steps_per_second": 3.008,
      "step": 1578
    },
    {
      "epoch": 7.006651884700665,
      "grad_norm": 0.3736945688724518,
      "learning_rate": 4.932973347064177e-05,
      "loss": 1.2391,
      "step": 1580
    },
    {
      "epoch": 7.028824833702883,
      "grad_norm": 0.40935948491096497,
      "learning_rate": 4.8662515823941255e-05,
      "loss": 1.2378,
      "step": 1585
    },
    {
      "epoch": 7.0509977827051,
      "grad_norm": 0.37913885712623596,
      "learning_rate": 4.799838720707846e-05,
      "loss": 1.2151,
      "step": 1590
    },
    {
      "epoch": 7.073170731707317,
      "grad_norm": 0.3685004413127899,
      "learning_rate": 4.733738758136327e-05,
      "loss": 1.2104,
      "step": 1595
    },
    {
      "epoch": 7.095343680709535,
      "grad_norm": 0.3836255669593811,
      "learning_rate": 4.66795567198309e-05,
      "loss": 1.22,
      "step": 1600
    },
    {
      "epoch": 7.117516629711751,
      "grad_norm": 0.4032094478607178,
      "learning_rate": 4.6024934204848745e-05,
      "loss": 1.2326,
      "step": 1605
    },
    {
      "epoch": 7.139689578713969,
      "grad_norm": 0.39435017108917236,
      "learning_rate": 4.537355942573463e-05,
      "loss": 1.2165,
      "step": 1610
    },
    {
      "epoch": 7.1618625277161865,
      "grad_norm": 0.4004512131214142,
      "learning_rate": 4.4725471576386735e-05,
      "loss": 1.2291,
      "step": 1615
    },
    {
      "epoch": 7.184035476718403,
      "grad_norm": 0.37496936321258545,
      "learning_rate": 4.4080709652925336e-05,
      "loss": 1.221,
      "step": 1620
    },
    {
      "epoch": 7.206208425720621,
      "grad_norm": 0.3921966552734375,
      "learning_rate": 4.343931245134616e-05,
      "loss": 1.2243,
      "step": 1625
    },
    {
      "epoch": 7.228381374722838,
      "grad_norm": 0.41742879152297974,
      "learning_rate": 4.2801318565186165e-05,
      "loss": 1.2299,
      "step": 1630
    },
    {
      "epoch": 7.250554323725056,
      "grad_norm": 0.38385558128356934,
      "learning_rate": 4.216676638320135e-05,
      "loss": 1.2364,
      "step": 1635
    },
    {
      "epoch": 7.2727272727272725,
      "grad_norm": 0.36501455307006836,
      "learning_rate": 4.15356940870567e-05,
      "loss": 1.2195,
      "step": 1640
    },
    {
      "epoch": 7.29490022172949,
      "grad_norm": 0.38054943084716797,
      "learning_rate": 4.090813964902889e-05,
      "loss": 1.2293,
      "step": 1645
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 0.36994317173957825,
      "learning_rate": 4.028414082972141e-05,
      "loss": 1.2198,
      "step": 1650
    },
    {
      "epoch": 7.339246119733924,
      "grad_norm": 0.3893408477306366,
      "learning_rate": 3.966373517579244e-05,
      "loss": 1.2384,
      "step": 1655
    },
    {
      "epoch": 7.361419068736142,
      "grad_norm": 0.3703598082065582,
      "learning_rate": 3.904696001769571e-05,
      "loss": 1.2348,
      "step": 1660
    },
    {
      "epoch": 7.383592017738359,
      "grad_norm": 0.3875352740287781,
      "learning_rate": 3.843385246743417e-05,
      "loss": 1.2402,
      "step": 1665
    },
    {
      "epoch": 7.405764966740577,
      "grad_norm": 0.399080365896225,
      "learning_rate": 3.7824449416327126e-05,
      "loss": 1.2268,
      "step": 1670
    },
    {
      "epoch": 7.427937915742794,
      "grad_norm": 0.37249988317489624,
      "learning_rate": 3.721878753279017e-05,
      "loss": 1.227,
      "step": 1675
    },
    {
      "epoch": 7.450110864745011,
      "grad_norm": 0.3863629996776581,
      "learning_rate": 3.661690326012897e-05,
      "loss": 1.22,
      "step": 1680
    },
    {
      "epoch": 7.472283813747229,
      "grad_norm": 0.3829827904701233,
      "learning_rate": 3.601883281434652e-05,
      "loss": 1.2137,
      "step": 1685
    },
    {
      "epoch": 7.494456762749445,
      "grad_norm": 0.3759503960609436,
      "learning_rate": 3.542461218196379e-05,
      "loss": 1.2077,
      "step": 1690
    },
    {
      "epoch": 7.516629711751663,
      "grad_norm": 0.38627204298973083,
      "learning_rate": 3.483427711785449e-05,
      "loss": 1.2233,
      "step": 1695
    },
    {
      "epoch": 7.5388026607538805,
      "grad_norm": 0.38867396116256714,
      "learning_rate": 3.424786314309365e-05,
      "loss": 1.2226,
      "step": 1700
    },
    {
      "epoch": 7.560975609756097,
      "grad_norm": 0.38329002261161804,
      "learning_rate": 3.366540554282028e-05,
      "loss": 1.2247,
      "step": 1705
    },
    {
      "epoch": 7.583148558758315,
      "grad_norm": 0.3766213059425354,
      "learning_rate": 3.308693936411421e-05,
      "loss": 1.2376,
      "step": 1710
    },
    {
      "epoch": 7.605321507760532,
      "grad_norm": 0.3749763071537018,
      "learning_rate": 3.2512499413887255e-05,
      "loss": 1.2099,
      "step": 1715
    },
    {
      "epoch": 7.627494456762749,
      "grad_norm": 0.3896552324295044,
      "learning_rate": 3.194212025678896e-05,
      "loss": 1.2119,
      "step": 1720
    },
    {
      "epoch": 7.6496674057649665,
      "grad_norm": 0.37435224652290344,
      "learning_rate": 3.137583621312665e-05,
      "loss": 1.229,
      "step": 1725
    },
    {
      "epoch": 7.671840354767184,
      "grad_norm": 0.3764979839324951,
      "learning_rate": 3.0813681356800405e-05,
      "loss": 1.2309,
      "step": 1730
    },
    {
      "epoch": 7.694013303769402,
      "grad_norm": 0.39203503727912903,
      "learning_rate": 3.025568951325287e-05,
      "loss": 1.2204,
      "step": 1735
    },
    {
      "epoch": 7.716186252771618,
      "grad_norm": 0.3817265033721924,
      "learning_rate": 2.9701894257433826e-05,
      "loss": 1.2297,
      "step": 1740
    },
    {
      "epoch": 7.738359201773836,
      "grad_norm": 0.38842707872390747,
      "learning_rate": 2.9152328911780026e-05,
      "loss": 1.2134,
      "step": 1745
    },
    {
      "epoch": 7.760532150776053,
      "grad_norm": 0.38738590478897095,
      "learning_rate": 2.8607026544210114e-05,
      "loss": 1.235,
      "step": 1750
    },
    {
      "epoch": 7.782705099778271,
      "grad_norm": 0.3651287257671356,
      "learning_rate": 2.8066019966134904e-05,
      "loss": 1.2327,
      "step": 1755
    },
    {
      "epoch": 7.804878048780488,
      "grad_norm": 0.3861520290374756,
      "learning_rate": 2.7529341730483117e-05,
      "loss": 1.2271,
      "step": 1760
    },
    {
      "epoch": 7.827050997782705,
      "grad_norm": 0.39386945962905884,
      "learning_rate": 2.6997024129742542e-05,
      "loss": 1.2255,
      "step": 1765
    },
    {
      "epoch": 7.849223946784923,
      "grad_norm": 0.3876512944698334,
      "learning_rate": 2.6469099194017143e-05,
      "loss": 1.2323,
      "step": 1770
    },
    {
      "epoch": 7.871396895787139,
      "grad_norm": 0.38528311252593994,
      "learning_rate": 2.594559868909956e-05,
      "loss": 1.2244,
      "step": 1775
    },
    {
      "epoch": 7.893569844789357,
      "grad_norm": 0.3802613615989685,
      "learning_rate": 2.542655411455982e-05,
      "loss": 1.2207,
      "step": 1780
    },
    {
      "epoch": 7.9157427937915745,
      "grad_norm": 0.3857831060886383,
      "learning_rate": 2.491199670185008e-05,
      "loss": 1.2183,
      "step": 1785
    },
    {
      "epoch": 7.937915742793791,
      "grad_norm": 0.3909415304660797,
      "learning_rate": 2.4401957412425214e-05,
      "loss": 1.2336,
      "step": 1790
    },
    {
      "epoch": 7.960088691796009,
      "grad_norm": 0.3930368423461914,
      "learning_rate": 2.389646693587996e-05,
      "loss": 1.2278,
      "step": 1795
    },
    {
      "epoch": 7.982261640798226,
      "grad_norm": 0.37647300958633423,
      "learning_rate": 2.339555568810221e-05,
      "loss": 1.229,
      "step": 1800
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.8372095823287964,
      "eval_runtime": 0.3265,
      "eval_samples_per_second": 3.063,
      "eval_steps_per_second": 3.063,
      "step": 1804
    },
    {
      "epoch": 8.004434589800443,
      "grad_norm": 0.3868635594844818,
      "learning_rate": 2.2899253809442944e-05,
      "loss": 1.233,
      "step": 1805
    },
    {
      "epoch": 8.026607538802661,
      "grad_norm": 0.37865373492240906,
      "learning_rate": 2.2407591162902573e-05,
      "loss": 1.2044,
      "step": 1810
    },
    {
      "epoch": 8.048780487804878,
      "grad_norm": 0.3961890935897827,
      "learning_rate": 2.192059733233408e-05,
      "loss": 1.2192,
      "step": 1815
    },
    {
      "epoch": 8.070953436807095,
      "grad_norm": 0.3783002197742462,
      "learning_rate": 2.1438301620662993e-05,
      "loss": 1.2196,
      "step": 1820
    },
    {
      "epoch": 8.093126385809313,
      "grad_norm": 0.3858291506767273,
      "learning_rate": 2.0960733048124083e-05,
      "loss": 1.2285,
      "step": 1825
    },
    {
      "epoch": 8.11529933481153,
      "grad_norm": 0.37426653504371643,
      "learning_rate": 2.0487920350515212e-05,
      "loss": 1.2164,
      "step": 1830
    },
    {
      "epoch": 8.137472283813747,
      "grad_norm": 0.3853384554386139,
      "learning_rate": 2.0019891977468408e-05,
      "loss": 1.2119,
      "step": 1835
    },
    {
      "epoch": 8.159645232815965,
      "grad_norm": 0.3728988766670227,
      "learning_rate": 1.95566760907378e-05,
      "loss": 1.2032,
      "step": 1840
    },
    {
      "epoch": 8.181818181818182,
      "grad_norm": 0.38846054673194885,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 1.2232,
      "step": 1845
    },
    {
      "epoch": 8.203991130820398,
      "grad_norm": 0.38317111134529114,
      "learning_rate": 1.864479297370325e-05,
      "loss": 1.218,
      "step": 1850
    },
    {
      "epoch": 8.226164079822617,
      "grad_norm": 0.38290783762931824,
      "learning_rate": 1.819618061235525e-05,
      "loss": 1.2037,
      "step": 1855
    },
    {
      "epoch": 8.248337028824833,
      "grad_norm": 0.38224560022354126,
      "learning_rate": 1.775249047193377e-05,
      "loss": 1.2055,
      "step": 1860
    },
    {
      "epoch": 8.27050997782705,
      "grad_norm": 0.3762098252773285,
      "learning_rate": 1.7313749249736267e-05,
      "loss": 1.2073,
      "step": 1865
    },
    {
      "epoch": 8.292682926829269,
      "grad_norm": 0.38844189047813416,
      "learning_rate": 1.687998334527853e-05,
      "loss": 1.2326,
      "step": 1870
    },
    {
      "epoch": 8.314855875831485,
      "grad_norm": 0.3953261971473694,
      "learning_rate": 1.6451218858706374e-05,
      "loss": 1.2224,
      "step": 1875
    },
    {
      "epoch": 8.337028824833704,
      "grad_norm": 0.37807291746139526,
      "learning_rate": 1.6027481589225026e-05,
      "loss": 1.2009,
      "step": 1880
    },
    {
      "epoch": 8.35920177383592,
      "grad_norm": 0.3885143995285034,
      "learning_rate": 1.560879703354693e-05,
      "loss": 1.224,
      "step": 1885
    },
    {
      "epoch": 8.381374722838137,
      "grad_norm": 0.38399428129196167,
      "learning_rate": 1.5195190384357404e-05,
      "loss": 1.2082,
      "step": 1890
    },
    {
      "epoch": 8.403547671840355,
      "grad_norm": 0.3910374045372009,
      "learning_rate": 1.4786686528798876e-05,
      "loss": 1.2202,
      "step": 1895
    },
    {
      "epoch": 8.425720620842572,
      "grad_norm": 0.3843975067138672,
      "learning_rate": 1.4383310046973365e-05,
      "loss": 1.2202,
      "step": 1900
    },
    {
      "epoch": 8.447893569844789,
      "grad_norm": 0.38334450125694275,
      "learning_rate": 1.3985085210463477e-05,
      "loss": 1.2099,
      "step": 1905
    },
    {
      "epoch": 8.470066518847007,
      "grad_norm": 0.3954315185546875,
      "learning_rate": 1.3592035980871953e-05,
      "loss": 1.2161,
      "step": 1910
    },
    {
      "epoch": 8.492239467849224,
      "grad_norm": 0.37927913665771484,
      "learning_rate": 1.3204186008379927e-05,
      "loss": 1.2153,
      "step": 1915
    },
    {
      "epoch": 8.51441241685144,
      "grad_norm": 0.3803432285785675,
      "learning_rate": 1.2821558630323772e-05,
      "loss": 1.2016,
      "step": 1920
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 0.38970625400543213,
      "learning_rate": 1.2444176869790925e-05,
      "loss": 1.2011,
      "step": 1925
    },
    {
      "epoch": 8.558758314855876,
      "grad_norm": 0.40996474027633667,
      "learning_rate": 1.2072063434234559e-05,
      "loss": 1.2238,
      "step": 1930
    },
    {
      "epoch": 8.580931263858092,
      "grad_norm": 0.3958864212036133,
      "learning_rate": 1.1705240714107302e-05,
      "loss": 1.2321,
      "step": 1935
    },
    {
      "epoch": 8.60310421286031,
      "grad_norm": 0.37584978342056274,
      "learning_rate": 1.1343730781513895e-05,
      "loss": 1.2138,
      "step": 1940
    },
    {
      "epoch": 8.625277161862527,
      "grad_norm": 0.39090994000434875,
      "learning_rate": 1.0987555388883041e-05,
      "loss": 1.2095,
      "step": 1945
    },
    {
      "epoch": 8.647450110864746,
      "grad_norm": 0.38422369956970215,
      "learning_rate": 1.0636735967658784e-05,
      "loss": 1.2115,
      "step": 1950
    },
    {
      "epoch": 8.669623059866963,
      "grad_norm": 0.3975456655025482,
      "learning_rate": 1.029129362701068e-05,
      "loss": 1.2078,
      "step": 1955
    },
    {
      "epoch": 8.69179600886918,
      "grad_norm": 0.3793989419937134,
      "learning_rate": 9.95124915256378e-06,
      "loss": 1.2183,
      "step": 1960
    },
    {
      "epoch": 8.713968957871398,
      "grad_norm": 0.3810500502586365,
      "learning_rate": 9.616623005147951e-06,
      "loss": 1.2218,
      "step": 1965
    },
    {
      "epoch": 8.736141906873614,
      "grad_norm": 0.3820514380931854,
      "learning_rate": 9.287435319566618e-06,
      "loss": 1.2122,
      "step": 1970
    },
    {
      "epoch": 8.758314855875831,
      "grad_norm": 0.3797374963760376,
      "learning_rate": 8.963705903385345e-06,
      "loss": 1.2278,
      "step": 1975
    },
    {
      "epoch": 8.78048780487805,
      "grad_norm": 0.376920223236084,
      "learning_rate": 8.645454235739903e-06,
      "loss": 1.2098,
      "step": 1980
    },
    {
      "epoch": 8.802660753880266,
      "grad_norm": 0.37694764137268066,
      "learning_rate": 8.332699466164306e-06,
      "loss": 1.2194,
      "step": 1985
    },
    {
      "epoch": 8.824833702882483,
      "grad_norm": 0.382021427154541,
      "learning_rate": 8.025460413438457e-06,
      "loss": 1.2077,
      "step": 1990
    },
    {
      "epoch": 8.847006651884701,
      "grad_norm": 0.5127786993980408,
      "learning_rate": 7.72375556445577e-06,
      "loss": 1.203,
      "step": 1995
    },
    {
      "epoch": 8.869179600886918,
      "grad_norm": 0.3939337134361267,
      "learning_rate": 7.427603073110967e-06,
      "loss": 1.2157,
      "step": 2000
    },
    {
      "epoch": 8.891352549889135,
      "grad_norm": 0.39616692066192627,
      "learning_rate": 7.13702075920758e-06,
      "loss": 1.2097,
      "step": 2005
    },
    {
      "epoch": 8.913525498891353,
      "grad_norm": 0.3867185115814209,
      "learning_rate": 6.852026107385756e-06,
      "loss": 1.2186,
      "step": 2010
    },
    {
      "epoch": 8.93569844789357,
      "grad_norm": 0.39439383149147034,
      "learning_rate": 6.572636266070264e-06,
      "loss": 1.2125,
      "step": 2015
    },
    {
      "epoch": 8.957871396895786,
      "grad_norm": 0.40186840295791626,
      "learning_rate": 6.298868046438533e-06,
      "loss": 1.214,
      "step": 2020
    },
    {
      "epoch": 8.980044345898005,
      "grad_norm": 0.45572689175605774,
      "learning_rate": 6.030737921409169e-06,
      "loss": 1.2239,
      "step": 2025
    },
    {
      "epoch": 8.997782705099779,
      "eval_loss": 1.848176121711731,
      "eval_runtime": 0.5386,
      "eval_samples_per_second": 1.857,
      "eval_steps_per_second": 1.857,
      "step": 2029
    },
    {
      "epoch": 9.002217294900221,
      "grad_norm": 0.3748861849308014,
      "learning_rate": 5.768262024650773e-06,
      "loss": 1.2133,
      "step": 2030
    },
    {
      "epoch": 9.024390243902438,
      "grad_norm": 0.37916404008865356,
      "learning_rate": 5.511456149611194e-06,
      "loss": 1.222,
      "step": 2035
    },
    {
      "epoch": 9.046563192904657,
      "grad_norm": 0.3922421634197235,
      "learning_rate": 5.26033574856708e-06,
      "loss": 1.2187,
      "step": 2040
    },
    {
      "epoch": 9.068736141906873,
      "grad_norm": 0.4379100501537323,
      "learning_rate": 5.014915931694253e-06,
      "loss": 1.2046,
      "step": 2045
    },
    {
      "epoch": 9.090909090909092,
      "grad_norm": 0.3986884653568268,
      "learning_rate": 4.775211466158469e-06,
      "loss": 1.2112,
      "step": 2050
    },
    {
      "epoch": 9.113082039911308,
      "grad_norm": 0.3886760473251343,
      "learning_rate": 4.541236775226809e-06,
      "loss": 1.2125,
      "step": 2055
    },
    {
      "epoch": 9.135254988913525,
      "grad_norm": 0.3771812617778778,
      "learning_rate": 4.313005937399861e-06,
      "loss": 1.1871,
      "step": 2060
    },
    {
      "epoch": 9.157427937915743,
      "grad_norm": 0.39025792479515076,
      "learning_rate": 4.0905326855646185e-06,
      "loss": 1.2039,
      "step": 2065
    },
    {
      "epoch": 9.17960088691796,
      "grad_norm": 0.3777172267436981,
      "learning_rate": 3.873830406168111e-06,
      "loss": 1.2145,
      "step": 2070
    },
    {
      "epoch": 9.201773835920177,
      "grad_norm": 0.39793717861175537,
      "learning_rate": 3.662912138411967e-06,
      "loss": 1.1983,
      "step": 2075
    },
    {
      "epoch": 9.223946784922395,
      "grad_norm": 0.3853585124015808,
      "learning_rate": 3.457790573467812e-06,
      "loss": 1.2223,
      "step": 2080
    },
    {
      "epoch": 9.246119733924612,
      "grad_norm": 0.3928294777870178,
      "learning_rate": 3.2584780537136207e-06,
      "loss": 1.2319,
      "step": 2085
    },
    {
      "epoch": 9.268292682926829,
      "grad_norm": 0.39439529180526733,
      "learning_rate": 3.0649865719910798e-06,
      "loss": 1.198,
      "step": 2090
    },
    {
      "epoch": 9.290465631929047,
      "grad_norm": 0.3766527771949768,
      "learning_rate": 2.877327770883964e-06,
      "loss": 1.2137,
      "step": 2095
    },
    {
      "epoch": 9.312638580931264,
      "grad_norm": 0.39711084961891174,
      "learning_rate": 2.6955129420176196e-06,
      "loss": 1.2232,
      "step": 2100
    },
    {
      "epoch": 9.33481152993348,
      "grad_norm": 0.37520113587379456,
      "learning_rate": 2.5195530253794396e-06,
      "loss": 1.2156,
      "step": 2105
    },
    {
      "epoch": 9.356984478935699,
      "grad_norm": 0.3946742117404938,
      "learning_rate": 2.349458608660704e-06,
      "loss": 1.2009,
      "step": 2110
    },
    {
      "epoch": 9.379157427937916,
      "grad_norm": 0.3818342685699463,
      "learning_rate": 2.1852399266194314e-06,
      "loss": 1.2066,
      "step": 2115
    },
    {
      "epoch": 9.401330376940132,
      "grad_norm": 0.3952968716621399,
      "learning_rate": 2.026906860464606e-06,
      "loss": 1.2346,
      "step": 2120
    },
    {
      "epoch": 9.42350332594235,
      "grad_norm": 0.3948782682418823,
      "learning_rate": 1.874468937261531e-06,
      "loss": 1.2154,
      "step": 2125
    },
    {
      "epoch": 9.445676274944567,
      "grad_norm": 0.3829614520072937,
      "learning_rate": 1.7279353293586765e-06,
      "loss": 1.2089,
      "step": 2130
    },
    {
      "epoch": 9.467849223946786,
      "grad_norm": 0.38727471232414246,
      "learning_rate": 1.5873148538356753e-06,
      "loss": 1.205,
      "step": 2135
    },
    {
      "epoch": 9.490022172949002,
      "grad_norm": 0.39005568623542786,
      "learning_rate": 1.4526159719728594e-06,
      "loss": 1.2023,
      "step": 2140
    },
    {
      "epoch": 9.512195121951219,
      "grad_norm": 0.3805971145629883,
      "learning_rate": 1.323846788742078e-06,
      "loss": 1.2022,
      "step": 2145
    },
    {
      "epoch": 9.534368070953438,
      "grad_norm": 0.3740484416484833,
      "learning_rate": 1.201015052319099e-06,
      "loss": 1.2251,
      "step": 2150
    },
    {
      "epoch": 9.556541019955654,
      "grad_norm": 0.3876747190952301,
      "learning_rate": 1.084128153617292e-06,
      "loss": 1.2114,
      "step": 2155
    },
    {
      "epoch": 9.57871396895787,
      "grad_norm": 0.37821733951568604,
      "learning_rate": 9.731931258429638e-07,
      "loss": 1.2099,
      "step": 2160
    },
    {
      "epoch": 9.60088691796009,
      "grad_norm": 0.38440433144569397,
      "learning_rate": 8.682166440721728e-07,
      "loss": 1.2067,
      "step": 2165
    },
    {
      "epoch": 9.623059866962306,
      "grad_norm": 0.3851392865180969,
      "learning_rate": 7.69205024849029e-07,
      "loss": 1.2126,
      "step": 2170
    },
    {
      "epoch": 9.645232815964523,
      "grad_norm": 0.39497315883636475,
      "learning_rate": 6.761642258056978e-07,
      "loss": 1.219,
      "step": 2175
    },
    {
      "epoch": 9.667405764966741,
      "grad_norm": 0.3837352395057678,
      "learning_rate": 5.890998453038644e-07,
      "loss": 1.2034,
      "step": 2180
    },
    {
      "epoch": 9.689578713968958,
      "grad_norm": 0.38280948996543884,
      "learning_rate": 5.080171220978813e-07,
      "loss": 1.218,
      "step": 2185
    },
    {
      "epoch": 9.711751662971174,
      "grad_norm": 0.3924092650413513,
      "learning_rate": 4.329209350195651e-07,
      "loss": 1.1972,
      "step": 2190
    },
    {
      "epoch": 9.733924611973393,
      "grad_norm": 0.38612061738967896,
      "learning_rate": 3.638158026846306e-07,
      "loss": 1.2058,
      "step": 2195
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 0.38143137097358704,
      "learning_rate": 3.007058832207976e-07,
      "loss": 1.1958,
      "step": 2200
    },
    {
      "epoch": 9.778270509977826,
      "grad_norm": 0.39435523748397827,
      "learning_rate": 2.4359497401758024e-07,
      "loss": 1.1962,
      "step": 2205
    },
    {
      "epoch": 9.800443458980045,
      "grad_norm": 0.3891894221305847,
      "learning_rate": 1.924865114978025e-07,
      "loss": 1.2133,
      "step": 2210
    },
    {
      "epoch": 9.822616407982261,
      "grad_norm": 0.38457366824150085,
      "learning_rate": 1.4738357091084176e-07,
      "loss": 1.2151,
      "step": 2215
    },
    {
      "epoch": 9.84478935698448,
      "grad_norm": 0.3849795162677765,
      "learning_rate": 1.0828886614754341e-07,
      "loss": 1.2058,
      "step": 2220
    },
    {
      "epoch": 9.866962305986696,
      "grad_norm": 0.38724285364151,
      "learning_rate": 7.520474957699586e-08,
      "loss": 1.2102,
      "step": 2225
    },
    {
      "epoch": 9.889135254988913,
      "grad_norm": 0.3877987265586853,
      "learning_rate": 4.8133211904888285e-08,
      "loss": 1.2122,
      "step": 2230
    },
    {
      "epoch": 9.911308203991132,
      "grad_norm": 0.3823870122432709,
      "learning_rate": 2.7075882053828605e-08,
      "loss": 1.213,
      "step": 2235
    },
    {
      "epoch": 9.933481152993348,
      "grad_norm": 0.3904678523540497,
      "learning_rate": 1.2034027065255249e-08,
      "loss": 1.2077,
      "step": 2240
    },
    {
      "epoch": 9.955654101995565,
      "grad_norm": 0.37739241123199463,
      "learning_rate": 3.0085520232425722e-09,
      "loss": 1.1993,
      "step": 2245
    },
    {
      "epoch": 9.977827050997783,
      "grad_norm": 0.3910938799381256,
      "learning_rate": 0.0,
      "loss": 1.2051,
      "step": 2250
    },
    {
      "epoch": 9.977827050997783,
      "eval_loss": 1.8520119190216064,
      "eval_runtime": 0.3296,
      "eval_samples_per_second": 3.034,
      "eval_steps_per_second": 3.034,
      "step": 2250
    },
    {
      "epoch": 9.977827050997783,
      "step": 2250,
      "total_flos": 3.3259687719144e+18,
      "train_loss": 1.3362829395929972,
      "train_runtime": 6815.0283,
      "train_samples_per_second": 10.572,
      "train_steps_per_second": 0.33
    }
  ],
  "logging_steps": 5,
  "max_steps": 2250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.3259687719144e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}