{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.944055944055943,
  "eval_steps": 500,
  "global_step": 2852,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06993006993006994,
      "grad_norm": 30.581405639648438,
      "learning_rate": 1.3986013986013988e-05,
      "loss": 1.1014,
      "step": 10
    },
    {
      "epoch": 0.13986013986013987,
      "grad_norm": 2.1117727756500244,
      "learning_rate": 2.7972027972027976e-05,
      "loss": 0.693,
      "step": 20
    },
    {
      "epoch": 0.2097902097902098,
      "grad_norm": 1.6114064455032349,
      "learning_rate": 4.195804195804196e-05,
      "loss": 0.3677,
      "step": 30
    },
    {
      "epoch": 0.27972027972027974,
      "grad_norm": 0.6671944260597229,
      "learning_rate": 5.594405594405595e-05,
      "loss": 0.2686,
      "step": 40
    },
    {
      "epoch": 0.34965034965034963,
      "grad_norm": 0.8836167454719543,
      "learning_rate": 6.993006993006993e-05,
      "loss": 0.2072,
      "step": 50
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 0.974694013595581,
      "learning_rate": 8.391608391608392e-05,
      "loss": 0.1783,
      "step": 60
    },
    {
      "epoch": 0.48951048951048953,
      "grad_norm": 0.8012842535972595,
      "learning_rate": 9.790209790209791e-05,
      "loss": 0.1542,
      "step": 70
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 0.5671571493148804,
      "learning_rate": 0.0001118881118881119,
      "loss": 0.1271,
      "step": 80
    },
    {
      "epoch": 0.6293706293706294,
      "grad_norm": 0.6660890579223633,
      "learning_rate": 0.00012587412587412587,
      "loss": 0.1197,
      "step": 90
    },
    {
      "epoch": 0.6993006993006993,
      "grad_norm": 0.4433947503566742,
      "learning_rate": 0.00013986013986013986,
      "loss": 0.1032,
      "step": 100
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.36532077193260193,
      "learning_rate": 0.00015384615384615385,
      "loss": 0.1052,
      "step": 110
    },
    {
      "epoch": 0.8391608391608392,
      "grad_norm": 0.48936668038368225,
      "learning_rate": 0.00016783216783216784,
      "loss": 0.0878,
      "step": 120
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.6362347602844238,
      "learning_rate": 0.00018181818181818183,
      "loss": 0.0865,
      "step": 130
    },
    {
      "epoch": 0.9790209790209791,
      "grad_norm": 0.4699188470840454,
      "learning_rate": 0.00019580419580419583,
      "loss": 0.0904,
      "step": 140
    },
    {
      "epoch": 1.048951048951049,
      "grad_norm": 0.42141956090927124,
      "learning_rate": 0.00019999670507574947,
      "loss": 0.0842,
      "step": 150
    },
    {
      "epoch": 1.118881118881119,
      "grad_norm": 0.3909319043159485,
      "learning_rate": 0.00019998056719395973,
      "loss": 0.0808,
      "step": 160
    },
    {
      "epoch": 1.1888111888111887,
      "grad_norm": 0.5473119616508484,
      "learning_rate": 0.00019995098333206742,
      "loss": 0.0713,
      "step": 170
    },
    {
      "epoch": 1.2587412587412588,
      "grad_norm": 0.37541836500167847,
      "learning_rate": 0.00019990795746868583,
      "loss": 0.0644,
      "step": 180
    },
    {
      "epoch": 1.3286713286713288,
      "grad_norm": 0.34557509422302246,
      "learning_rate": 0.00019985149539018855,
      "loss": 0.0707,
      "step": 190
    },
    {
      "epoch": 1.3986013986013985,
      "grad_norm": 0.29241743683815,
      "learning_rate": 0.00019978160468993094,
      "loss": 0.0612,
      "step": 200
    },
    {
      "epoch": 1.4685314685314685,
      "grad_norm": 0.481886625289917,
      "learning_rate": 0.00019969829476722923,
      "loss": 0.065,
      "step": 210
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.30471473932266235,
      "learning_rate": 0.00019960157682609632,
      "loss": 0.0596,
      "step": 220
    },
    {
      "epoch": 1.6083916083916083,
      "grad_norm": 0.30594855546951294,
      "learning_rate": 0.00019949146387373493,
      "loss": 0.06,
      "step": 230
    },
    {
      "epoch": 1.6783216783216783,
      "grad_norm": 0.31589317321777344,
      "learning_rate": 0.00019936797071878854,
      "loss": 0.0643,
      "step": 240
    },
    {
      "epoch": 1.7482517482517483,
      "grad_norm": 0.293947696685791,
      "learning_rate": 0.00019923111396934957,
      "loss": 0.0568,
      "step": 250
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.25768476724624634,
      "learning_rate": 0.00019908091203072598,
      "loss": 0.0556,
      "step": 260
    },
    {
      "epoch": 1.8881118881118881,
      "grad_norm": 0.258080393075943,
      "learning_rate": 0.00019891738510296602,
      "loss": 0.0543,
      "step": 270
    },
    {
      "epoch": 1.958041958041958,
      "grad_norm": 0.27586156129837036,
      "learning_rate": 0.0001987405551781415,
      "loss": 0.051,
      "step": 280
    },
    {
      "epoch": 2.027972027972028,
      "grad_norm": 0.26486584544181824,
      "learning_rate": 0.0001985504460373903,
      "loss": 0.0521,
      "step": 290
    },
    {
      "epoch": 2.097902097902098,
      "grad_norm": 0.2695034146308899,
      "learning_rate": 0.00019834708324771797,
      "loss": 0.0531,
      "step": 300
    },
    {
      "epoch": 2.167832167832168,
      "grad_norm": 0.30032461881637573,
      "learning_rate": 0.00019813049415855964,
      "loss": 0.0541,
      "step": 310
    },
    {
      "epoch": 2.237762237762238,
      "grad_norm": 0.3316197395324707,
      "learning_rate": 0.00019790070789810145,
      "loss": 0.0458,
      "step": 320
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 0.3239120543003082,
      "learning_rate": 0.00019765775536936367,
      "loss": 0.0533,
      "step": 330
    },
    {
      "epoch": 2.3776223776223775,
      "grad_norm": 0.20758603513240814,
      "learning_rate": 0.00019740166924604431,
      "loss": 0.0527,
      "step": 340
    },
    {
      "epoch": 2.4475524475524475,
      "grad_norm": 0.31676483154296875,
      "learning_rate": 0.00019713248396812524,
      "loss": 0.0521,
      "step": 350
    },
    {
      "epoch": 2.5174825174825175,
      "grad_norm": 0.25248146057128906,
      "learning_rate": 0.00019685023573724037,
      "loss": 0.0518,
      "step": 360
    },
    {
      "epoch": 2.5874125874125875,
      "grad_norm": 0.26571494340896606,
      "learning_rate": 0.0001965549625118071,
      "loss": 0.045,
      "step": 370
    },
    {
      "epoch": 2.6573426573426575,
      "grad_norm": 0.2656267285346985,
      "learning_rate": 0.00019624670400192126,
      "loss": 0.0434,
      "step": 380
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.30044880509376526,
      "learning_rate": 0.00019592550166401695,
      "loss": 0.0443,
      "step": 390
    },
    {
      "epoch": 2.797202797202797,
      "grad_norm": 0.27586933970451355,
      "learning_rate": 0.00019559139869529103,
      "loss": 0.044,
      "step": 400
    },
    {
      "epoch": 2.867132867132867,
      "grad_norm": 0.34137022495269775,
      "learning_rate": 0.00019524444002789383,
      "loss": 0.0473,
      "step": 410
    },
    {
      "epoch": 2.937062937062937,
      "grad_norm": 0.2440134882926941,
      "learning_rate": 0.0001948846723228862,
      "loss": 0.0438,
      "step": 420
    },
    {
      "epoch": 3.006993006993007,
      "grad_norm": 0.23091812431812286,
      "learning_rate": 0.00019451214396396454,
      "loss": 0.0463,
      "step": 430
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.15170010924339294,
      "learning_rate": 0.00019412690505095365,
      "loss": 0.0416,
      "step": 440
    },
    {
      "epoch": 3.1468531468531467,
      "grad_norm": 0.25807687640190125,
      "learning_rate": 0.00019372900739306908,
      "loss": 0.0419,
      "step": 450
    },
    {
      "epoch": 3.2167832167832167,
      "grad_norm": 0.38740265369415283,
      "learning_rate": 0.00019331850450194957,
      "loss": 0.0379,
      "step": 460
    },
    {
      "epoch": 3.2867132867132867,
      "grad_norm": 0.24318642914295197,
      "learning_rate": 0.00019289545158446045,
      "loss": 0.0391,
      "step": 470
    },
    {
      "epoch": 3.3566433566433567,
      "grad_norm": 0.2211398035287857,
      "learning_rate": 0.00019245990553526905,
      "loss": 0.043,
      "step": 480
    },
    {
      "epoch": 3.4265734265734267,
      "grad_norm": 0.3056994378566742,
      "learning_rate": 0.00019201192492919317,
      "loss": 0.0376,
      "step": 490
    },
    {
      "epoch": 3.4965034965034967,
      "grad_norm": 0.18653175234794617,
      "learning_rate": 0.00019155157001332374,
      "loss": 0.0442,
      "step": 500
    },
    {
      "epoch": 3.5664335664335667,
      "grad_norm": 0.21762531995773315,
      "learning_rate": 0.00019107890269892214,
      "loss": 0.0392,
      "step": 510
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.24727760255336761,
      "learning_rate": 0.0001905939865530944,
      "loss": 0.0413,
      "step": 520
    },
    {
      "epoch": 3.7062937062937062,
      "grad_norm": 0.34239208698272705,
      "learning_rate": 0.0001900968867902419,
      "loss": 0.0417,
      "step": 530
    },
    {
      "epoch": 3.7762237762237763,
      "grad_norm": 0.23154211044311523,
      "learning_rate": 0.0001895876702632913,
      "loss": 0.0391,
      "step": 540
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 0.25248488783836365,
      "learning_rate": 0.00018906640545470355,
      "loss": 0.0367,
      "step": 550
    },
    {
      "epoch": 3.916083916083916,
      "grad_norm": 0.2887849807739258,
      "learning_rate": 0.000188533162467264,
      "loss": 0.0394,
      "step": 560
    },
    {
      "epoch": 3.986013986013986,
      "grad_norm": 0.30540716648101807,
      "learning_rate": 0.0001879880130146547,
      "loss": 0.0435,
      "step": 570
    },
    {
      "epoch": 4.055944055944056,
      "grad_norm": 0.2462405562400818,
      "learning_rate": 0.0001874310304118096,
      "loss": 0.0354,
      "step": 580
    },
    {
      "epoch": 4.125874125874126,
      "grad_norm": 0.2431563138961792,
      "learning_rate": 0.00018686228956505516,
      "loss": 0.0409,
      "step": 590
    },
    {
      "epoch": 4.195804195804196,
      "grad_norm": 0.2564241886138916,
      "learning_rate": 0.00018628186696203612,
      "loss": 0.0397,
      "step": 600
    },
    {
      "epoch": 4.265734265734266,
      "grad_norm": 0.24190612137317657,
      "learning_rate": 0.00018568984066142917,
      "loss": 0.0377,
      "step": 610
    },
    {
      "epoch": 4.335664335664336,
      "grad_norm": 0.23087257146835327,
      "learning_rate": 0.00018508629028244519,
      "loss": 0.0412,
      "step": 620
    },
    {
      "epoch": 4.405594405594406,
      "grad_norm": 0.4008491039276123,
      "learning_rate": 0.00018447129699412142,
      "loss": 0.039,
      "step": 630
    },
    {
      "epoch": 4.475524475524476,
      "grad_norm": 0.27394184470176697,
      "learning_rate": 0.00018384494350440553,
      "loss": 0.038,
      "step": 640
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.20550896227359772,
      "learning_rate": 0.0001832073140490325,
      "loss": 0.0404,
      "step": 650
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.2519960403442383,
      "learning_rate": 0.00018255849438019608,
      "loss": 0.0378,
      "step": 660
    },
    {
      "epoch": 4.685314685314685,
      "grad_norm": 0.3346613645553589,
      "learning_rate": 0.00018189857175501635,
      "loss": 0.035,
      "step": 670
    },
    {
      "epoch": 4.755244755244755,
      "grad_norm": 0.32892099022865295,
      "learning_rate": 0.00018122763492380486,
      "loss": 0.0325,
      "step": 680
    },
    {
      "epoch": 4.825174825174825,
      "grad_norm": 0.24375475943088531,
      "learning_rate": 0.00018054577411812895,
      "loss": 0.0318,
      "step": 690
    },
    {
      "epoch": 4.895104895104895,
      "grad_norm": 0.19709643721580505,
      "learning_rate": 0.00017985308103867688,
      "loss": 0.0368,
      "step": 700
    },
    {
      "epoch": 4.965034965034965,
      "grad_norm": 0.25721314549446106,
      "learning_rate": 0.00017914964884292544,
      "loss": 0.0377,
      "step": 710
    },
    {
      "epoch": 5.034965034965035,
      "grad_norm": 0.3517259955406189,
      "learning_rate": 0.00017843557213261142,
      "loss": 0.0359,
      "step": 720
    },
    {
      "epoch": 5.104895104895105,
      "grad_norm": 0.2421354055404663,
      "learning_rate": 0.00017771094694100925,
      "loss": 0.0375,
      "step": 730
    },
    {
      "epoch": 5.174825174825175,
      "grad_norm": 0.22226205468177795,
      "learning_rate": 0.00017697587072001557,
      "loss": 0.0376,
      "step": 740
    },
    {
      "epoch": 5.244755244755245,
      "grad_norm": 0.2880263924598694,
      "learning_rate": 0.0001762304423270436,
      "loss": 0.0377,
      "step": 750
    },
    {
      "epoch": 5.314685314685315,
      "grad_norm": 0.2628781795501709,
      "learning_rate": 0.00017547476201172808,
      "loss": 0.0366,
      "step": 760
    },
    {
      "epoch": 5.384615384615385,
      "grad_norm": 0.21351908147335052,
      "learning_rate": 0.00017470893140244303,
      "loss": 0.0346,
      "step": 770
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.23355896770954132,
      "learning_rate": 0.00017393305349263434,
      "loss": 0.0367,
      "step": 780
    },
    {
      "epoch": 5.524475524475524,
      "grad_norm": 0.17863404750823975,
      "learning_rate": 0.00017314723262696848,
      "loss": 0.0319,
      "step": 790
    },
    {
      "epoch": 5.594405594405594,
      "grad_norm": 0.2492302805185318,
      "learning_rate": 0.00017235157448729967,
      "loss": 0.0345,
      "step": 800
    },
    {
      "epoch": 5.664335664335664,
      "grad_norm": 0.1684318482875824,
      "learning_rate": 0.00017154618607845702,
      "loss": 0.0325,
      "step": 810
    },
    {
      "epoch": 5.734265734265734,
      "grad_norm": 0.19033940136432648,
      "learning_rate": 0.00017073117571385414,
      "loss": 0.0325,
      "step": 820
    },
    {
      "epoch": 5.804195804195804,
      "grad_norm": 0.1792939007282257,
      "learning_rate": 0.00016990665300092224,
      "loss": 0.037,
      "step": 830
    },
    {
      "epoch": 5.874125874125874,
      "grad_norm": 0.17368149757385254,
      "learning_rate": 0.00016907272882636968,
      "loss": 0.0329,
      "step": 840
    },
    {
      "epoch": 5.944055944055944,
      "grad_norm": 0.20785829424858093,
      "learning_rate": 0.0001682295153412691,
      "loss": 0.0316,
      "step": 850
    },
    {
      "epoch": 6.013986013986014,
      "grad_norm": 0.19825145602226257,
      "learning_rate": 0.00016737712594597483,
      "loss": 0.0274,
      "step": 860
    },
    {
      "epoch": 6.083916083916084,
      "grad_norm": 0.15773798525333405,
      "learning_rate": 0.00016651567527487204,
      "loss": 0.0347,
      "step": 870
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 0.26358160376548767,
      "learning_rate": 0.00016564527918096005,
      "loss": 0.0299,
      "step": 880
    },
    {
      "epoch": 6.223776223776224,
      "grad_norm": 0.3937130272388458,
      "learning_rate": 0.00016476605472027172,
      "loss": 0.0309,
      "step": 890
    },
    {
      "epoch": 6.293706293706293,
      "grad_norm": 0.13225628435611725,
      "learning_rate": 0.00016387812013613103,
      "loss": 0.035,
      "step": 900
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.19665835797786713,
      "learning_rate": 0.00016298159484325118,
      "loss": 0.0316,
      "step": 910
    },
    {
      "epoch": 6.433566433566433,
      "grad_norm": 0.21388469636440277,
      "learning_rate": 0.00016207659941167485,
      "loss": 0.0272,
      "step": 920
    },
    {
      "epoch": 6.503496503496503,
      "grad_norm": 0.47336331009864807,
      "learning_rate": 0.00016116325555055915,
      "loss": 0.0293,
      "step": 930
    },
    {
      "epoch": 6.573426573426573,
      "grad_norm": 0.1514803022146225,
      "learning_rate": 0.00016024168609180757,
      "loss": 0.0316,
      "step": 940
    },
    {
      "epoch": 6.643356643356643,
      "grad_norm": 0.20940926671028137,
      "learning_rate": 0.00015931201497355088,
      "loss": 0.0282,
      "step": 950
    },
    {
      "epoch": 6.713286713286713,
      "grad_norm": 0.22871293127536774,
      "learning_rate": 0.000158374367223479,
      "loss": 0.0349,
      "step": 960
    },
    {
      "epoch": 6.783216783216783,
      "grad_norm": 0.18799573183059692,
      "learning_rate": 0.00015742886894202674,
      "loss": 0.0295,
      "step": 970
    },
    {
      "epoch": 6.853146853146853,
      "grad_norm": 0.17343267798423767,
      "learning_rate": 0.00015647564728541485,
      "loss": 0.0299,
      "step": 980
    },
    {
      "epoch": 6.923076923076923,
      "grad_norm": 0.21930481493473053,
      "learning_rate": 0.00015551483044854954,
      "loss": 0.0285,
      "step": 990
    },
    {
      "epoch": 6.993006993006993,
      "grad_norm": 0.15558023750782013,
      "learning_rate": 0.00015454654764778187,
      "loss": 0.03,
      "step": 1000
    },
    {
      "epoch": 7.062937062937063,
      "grad_norm": 0.39069393277168274,
      "learning_rate": 0.00015357092910353001,
      "loss": 0.0263,
      "step": 1010
    },
    {
      "epoch": 7.1328671328671325,
      "grad_norm": 0.22350826859474182,
      "learning_rate": 0.00015258810602276654,
      "loss": 0.0341,
      "step": 1020
    },
    {
      "epoch": 7.2027972027972025,
      "grad_norm": 0.16752642393112183,
      "learning_rate": 0.00015159821058137278,
      "loss": 0.0279,
      "step": 1030
    },
    {
      "epoch": 7.2727272727272725,
      "grad_norm": 0.17457683384418488,
      "learning_rate": 0.00015060137590636318,
      "loss": 0.0268,
      "step": 1040
    },
    {
      "epoch": 7.3426573426573425,
      "grad_norm": 0.1841355264186859,
      "learning_rate": 0.00014959773605798145,
      "loss": 0.0248,
      "step": 1050
    },
    {
      "epoch": 7.4125874125874125,
      "grad_norm": 0.18273039162158966,
      "learning_rate": 0.0001485874260116714,
      "loss": 0.0256,
      "step": 1060
    },
    {
      "epoch": 7.4825174825174825,
      "grad_norm": 0.21696248650550842,
      "learning_rate": 0.00014757058163992464,
      "loss": 0.026,
      "step": 1070
    },
    {
      "epoch": 7.5524475524475525,
      "grad_norm": 0.23004591464996338,
      "learning_rate": 0.0001465473396940078,
      "loss": 0.0305,
      "step": 1080
    },
    {
      "epoch": 7.6223776223776225,
      "grad_norm": 0.17306514084339142,
      "learning_rate": 0.000145517837785571,
      "loss": 0.0247,
      "step": 1090
    },
    {
      "epoch": 7.6923076923076925,
      "grad_norm": 0.2152336984872818,
      "learning_rate": 0.0001444822143681415,
      "loss": 0.0274,
      "step": 1100
    },
    {
      "epoch": 7.7622377622377625,
      "grad_norm": 0.1681831181049347,
      "learning_rate": 0.00014344060871850325,
      "loss": 0.0247,
      "step": 1110
    },
    {
      "epoch": 7.8321678321678325,
      "grad_norm": 0.1374141126871109,
      "learning_rate": 0.00014239316091796647,
      "loss": 0.0285,
      "step": 1120
    },
    {
      "epoch": 7.902097902097902,
      "grad_norm": 0.3081527352333069,
      "learning_rate": 0.00014134001183352832,
      "loss": 0.0246,
      "step": 1130
    },
    {
      "epoch": 7.972027972027972,
      "grad_norm": 0.4359898865222931,
      "learning_rate": 0.0001402813030989286,
      "loss": 0.0284,
      "step": 1140
    },
    {
      "epoch": 8.041958041958042,
      "grad_norm": 0.15682591497898102,
      "learning_rate": 0.00013921717709560182,
      "loss": 0.0261,
      "step": 1150
    },
    {
      "epoch": 8.111888111888112,
      "grad_norm": 0.23213213682174683,
      "learning_rate": 0.000138147776933529,
      "loss": 0.0265,
      "step": 1160
    },
    {
      "epoch": 8.181818181818182,
      "grad_norm": 0.1675710529088974,
      "learning_rate": 0.00013707324643199114,
      "loss": 0.0297,
      "step": 1170
    },
    {
      "epoch": 8.251748251748252,
      "grad_norm": 0.1905348300933838,
      "learning_rate": 0.00013599373010022794,
      "loss": 0.0282,
      "step": 1180
    },
    {
      "epoch": 8.321678321678322,
      "grad_norm": 0.22106771171092987,
      "learning_rate": 0.0001349093731180031,
      "loss": 0.0265,
      "step": 1190
    },
    {
      "epoch": 8.391608391608392,
      "grad_norm": 0.21079835295677185,
      "learning_rate": 0.00013382032131607966,
      "loss": 0.0288,
      "step": 1200
    },
    {
      "epoch": 8.461538461538462,
      "grad_norm": 0.3102055788040161,
      "learning_rate": 0.00013272672115660796,
      "loss": 0.0226,
      "step": 1210
    },
    {
      "epoch": 8.531468531468532,
      "grad_norm": 0.16320601105690002,
      "learning_rate": 0.00013162871971342837,
      "loss": 0.026,
      "step": 1220
    },
    {
      "epoch": 8.601398601398602,
      "grad_norm": 0.2458457201719284,
      "learning_rate": 0.00013052646465229207,
      "loss": 0.0257,
      "step": 1230
    },
    {
      "epoch": 8.671328671328672,
      "grad_norm": 0.21947446465492249,
      "learning_rate": 0.00012942010421100207,
      "loss": 0.0255,
      "step": 1240
    },
    {
      "epoch": 8.741258741258742,
      "grad_norm": 0.1916748285293579,
      "learning_rate": 0.00012830978717947718,
      "loss": 0.0239,
      "step": 1250
    },
    {
      "epoch": 8.811188811188812,
      "grad_norm": 0.14072898030281067,
      "learning_rate": 0.00012719566287974204,
      "loss": 0.0261,
      "step": 1260
    },
    {
      "epoch": 8.881118881118882,
      "grad_norm": 0.1773282140493393,
      "learning_rate": 0.00012607788114584522,
      "loss": 0.023,
      "step": 1270
    },
    {
      "epoch": 8.951048951048952,
      "grad_norm": 0.13442933559417725,
      "learning_rate": 0.0001249565923037088,
      "loss": 0.0251,
      "step": 1280
    },
    {
      "epoch": 9.020979020979022,
      "grad_norm": 0.17046289145946503,
      "learning_rate": 0.00012383194715091163,
      "loss": 0.0253,
      "step": 1290
    },
    {
      "epoch": 9.090909090909092,
      "grad_norm": 0.25058501958847046,
      "learning_rate": 0.00012270409693640905,
      "loss": 0.0262,
      "step": 1300
    },
    {
      "epoch": 9.16083916083916,
      "grad_norm": 0.22269104421138763,
      "learning_rate": 0.00012157319334019219,
      "loss": 0.0235,
      "step": 1310
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 0.25462645292282104,
      "learning_rate": 0.00012043938845288904,
      "loss": 0.0256,
      "step": 1320
    },
    {
      "epoch": 9.3006993006993,
      "grad_norm": 0.18311487138271332,
      "learning_rate": 0.00011930283475531048,
      "loss": 0.0259,
      "step": 1330
    },
    {
      "epoch": 9.37062937062937,
      "grad_norm": 0.16915184259414673,
      "learning_rate": 0.00011816368509794364,
      "loss": 0.0259,
      "step": 1340
    },
    {
      "epoch": 9.44055944055944,
      "grad_norm": 0.16743628680706024,
      "learning_rate": 0.00011702209268039581,
      "loss": 0.0229,
      "step": 1350
    },
    {
      "epoch": 9.51048951048951,
      "grad_norm": 0.16608262062072754,
      "learning_rate": 0.00011587821103079111,
      "loss": 0.0238,
      "step": 1360
    },
    {
      "epoch": 9.58041958041958,
      "grad_norm": 0.14811694622039795,
      "learning_rate": 0.00011473219398512316,
      "loss": 0.0226,
      "step": 1370
    },
    {
      "epoch": 9.65034965034965,
      "grad_norm": 0.1586376130580902,
      "learning_rate": 0.00011358419566656642,
      "loss": 0.0196,
      "step": 1380
    },
    {
      "epoch": 9.72027972027972,
      "grad_norm": 0.13793979585170746,
      "learning_rate": 0.00011243437046474853,
      "loss": 0.0209,
      "step": 1390
    },
    {
      "epoch": 9.79020979020979,
      "grad_norm": 0.207151859998703,
      "learning_rate": 0.00011128287301498739,
      "loss": 0.0235,
      "step": 1400
    },
    {
      "epoch": 9.86013986013986,
      "grad_norm": 0.48499253392219543,
      "learning_rate": 0.00011012985817749463,
      "loss": 0.0217,
      "step": 1410
    },
    {
      "epoch": 9.93006993006993,
      "grad_norm": 0.11960924416780472,
      "learning_rate": 0.00010897548101654926,
      "loss": 0.0217,
      "step": 1420
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.12925837934017181,
      "learning_rate": 0.00010781989677964355,
      "loss": 0.0207,
      "step": 1430
    },
    {
      "epoch": 10.06993006993007,
      "grad_norm": 0.1686684936285019,
      "learning_rate": 0.00010666326087660458,
      "loss": 0.0204,
      "step": 1440
    },
    {
      "epoch": 10.13986013986014,
      "grad_norm": 0.15361668169498444,
      "learning_rate": 0.00010550572885869367,
      "loss": 0.0197,
      "step": 1450
    },
    {
      "epoch": 10.20979020979021,
      "grad_norm": 0.2281733602285385,
      "learning_rate": 0.00010434745639768705,
      "loss": 0.0252,
      "step": 1460
    },
    {
      "epoch": 10.27972027972028,
      "grad_norm": 0.14611436426639557,
      "learning_rate": 0.00010318859926494014,
      "loss": 0.02,
      "step": 1470
    },
    {
      "epoch": 10.34965034965035,
      "grad_norm": 0.3116054832935333,
      "learning_rate": 0.00010202931331043839,
      "loss": 0.0217,
      "step": 1480
    },
    {
      "epoch": 10.41958041958042,
      "grad_norm": 0.19651482999324799,
      "learning_rate": 0.00010086975444183782,
      "loss": 0.0232,
      "step": 1490
    },
    {
      "epoch": 10.48951048951049,
      "grad_norm": 0.18464644253253937,
      "learning_rate": 9.971007860349756e-05,
      "loss": 0.02,
      "step": 1500
    },
    {
      "epoch": 10.55944055944056,
      "grad_norm": 0.18407182395458221,
      "learning_rate": 9.855044175550756e-05,
      "loss": 0.0232,
      "step": 1510
    },
    {
      "epoch": 10.62937062937063,
      "grad_norm": 0.12513527274131775,
      "learning_rate": 9.739099985271394e-05,
      "loss": 0.0219,
      "step": 1520
    },
    {
      "epoch": 10.6993006993007,
      "grad_norm": 0.13158084452152252,
      "learning_rate": 9.623190882374564e-05,
      "loss": 0.0224,
      "step": 1530
    },
    {
      "epoch": 10.76923076923077,
      "grad_norm": 0.17640575766563416,
      "learning_rate": 9.507332455004395e-05,
      "loss": 0.0226,
      "step": 1540
    },
    {
      "epoch": 10.83916083916084,
      "grad_norm": 0.30499282479286194,
      "learning_rate": 9.391540284489862e-05,
      "loss": 0.0216,
      "step": 1550
    },
    {
      "epoch": 10.909090909090908,
      "grad_norm": 0.1940823644399643,
      "learning_rate": 9.275829943249334e-05,
      "loss": 0.0241,
      "step": 1560
    },
    {
      "epoch": 10.979020979020978,
      "grad_norm": 0.159086212515831,
      "learning_rate": 9.160216992696286e-05,
      "loss": 0.022,
      "step": 1570
    },
    {
      "epoch": 11.048951048951048,
      "grad_norm": 0.19389483332633972,
      "learning_rate": 9.044716981146526e-05,
      "loss": 0.022,
      "step": 1580
    },
    {
      "epoch": 11.118881118881118,
      "grad_norm": 0.12246847152709961,
      "learning_rate": 8.929345441727142e-05,
      "loss": 0.0205,
      "step": 1590
    },
    {
      "epoch": 11.188811188811188,
      "grad_norm": 0.16194018721580505,
      "learning_rate": 8.814117890287538e-05,
      "loss": 0.0207,
      "step": 1600
    },
    {
      "epoch": 11.258741258741258,
      "grad_norm": 0.17201407253742218,
      "learning_rate": 8.699049823312748e-05,
      "loss": 0.0216,
      "step": 1610
    },
    {
      "epoch": 11.328671328671328,
      "grad_norm": 0.12679952383041382,
      "learning_rate": 8.584156715839401e-05,
      "loss": 0.0194,
      "step": 1620
    },
    {
      "epoch": 11.398601398601398,
      "grad_norm": 0.19168208539485931,
      "learning_rate": 8.469454019374531e-05,
      "loss": 0.016,
      "step": 1630
    },
    {
      "epoch": 11.468531468531468,
      "grad_norm": 0.20202843844890594,
      "learning_rate": 8.354957159817561e-05,
      "loss": 0.0227,
      "step": 1640
    },
    {
      "epoch": 11.538461538461538,
      "grad_norm": 0.14431844651699066,
      "learning_rate": 8.240681535385757e-05,
      "loss": 0.0177,
      "step": 1650
    },
    {
      "epoch": 11.608391608391608,
      "grad_norm": 0.16581477224826813,
      "learning_rate": 8.126642514543359e-05,
      "loss": 0.017,
      "step": 1660
    },
    {
      "epoch": 11.678321678321678,
      "grad_norm": 0.15920375287532806,
      "learning_rate": 8.012855433934765e-05,
      "loss": 0.02,
      "step": 1670
    },
    {
      "epoch": 11.748251748251748,
      "grad_norm": 0.15303653478622437,
      "learning_rate": 7.899335596321944e-05,
      "loss": 0.0207,
      "step": 1680
    },
    {
      "epoch": 11.818181818181818,
      "grad_norm": 0.15237000584602356,
      "learning_rate": 7.786098268526448e-05,
      "loss": 0.0225,
      "step": 1690
    },
    {
      "epoch": 11.888111888111888,
      "grad_norm": 0.13649679720401764,
      "learning_rate": 7.673158679376234e-05,
      "loss": 0.0204,
      "step": 1700
    },
    {
      "epoch": 11.958041958041958,
      "grad_norm": 0.13087141513824463,
      "learning_rate": 7.560532017657585e-05,
      "loss": 0.0211,
      "step": 1710
    },
    {
      "epoch": 12.027972027972028,
      "grad_norm": 0.13551278412342072,
      "learning_rate": 7.448233430072466e-05,
      "loss": 0.0192,
      "step": 1720
    },
    {
      "epoch": 12.097902097902098,
      "grad_norm": 0.19065268337726593,
      "learning_rate": 7.336278019201462e-05,
      "loss": 0.0174,
      "step": 1730
    },
    {
      "epoch": 12.167832167832168,
      "grad_norm": 0.16062477231025696,
      "learning_rate": 7.224680841472741e-05,
      "loss": 0.0209,
      "step": 1740
    },
    {
      "epoch": 12.237762237762238,
      "grad_norm": 0.3261907696723938,
      "learning_rate": 7.113456905137132e-05,
      "loss": 0.0204,
      "step": 1750
    },
    {
      "epoch": 12.307692307692308,
      "grad_norm": 0.36956819891929626,
      "learning_rate": 7.002621168249759e-05,
      "loss": 0.0235,
      "step": 1760
    },
    {
      "epoch": 12.377622377622378,
      "grad_norm": 0.16811484098434448,
      "learning_rate": 6.892188536658369e-05,
      "loss": 0.0201,
      "step": 1770
    },
    {
      "epoch": 12.447552447552448,
      "grad_norm": 0.14103484153747559,
      "learning_rate": 6.782173861998726e-05,
      "loss": 0.0226,
      "step": 1780
    },
    {
      "epoch": 12.517482517482517,
      "grad_norm": 0.17502908408641815,
      "learning_rate": 6.672591939697261e-05,
      "loss": 0.0204,
      "step": 1790
    },
    {
      "epoch": 12.587412587412587,
      "grad_norm": 0.13043367862701416,
      "learning_rate": 6.563457506981297e-05,
      "loss": 0.0201,
      "step": 1800
    },
    {
      "epoch": 12.657342657342657,
      "grad_norm": 0.13106204569339752,
      "learning_rate": 6.454785240897112e-05,
      "loss": 0.0199,
      "step": 1810
    },
    {
      "epoch": 12.727272727272727,
      "grad_norm": 0.30531424283981323,
      "learning_rate": 6.34658975633605e-05,
      "loss": 0.0184,
      "step": 1820
    },
    {
      "epoch": 12.797202797202797,
      "grad_norm": 0.16103509068489075,
      "learning_rate": 6.238885604069075e-05,
      "loss": 0.0212,
      "step": 1830
    },
    {
      "epoch": 12.867132867132867,
      "grad_norm": 0.12883076071739197,
      "learning_rate": 6.131687268789838e-05,
      "loss": 0.0166,
      "step": 1840
    },
    {
      "epoch": 12.937062937062937,
      "grad_norm": 0.1775168776512146,
      "learning_rate": 6.0250091671667484e-05,
      "loss": 0.022,
      "step": 1850
    },
    {
      "epoch": 13.006993006993007,
      "grad_norm": 0.14147868752479553,
      "learning_rate": 5.9188656459040837e-05,
      "loss": 0.024,
      "step": 1860
    },
    {
      "epoch": 13.076923076923077,
      "grad_norm": 0.10535666346549988,
      "learning_rate": 5.813270979812589e-05,
      "loss": 0.0187,
      "step": 1870
    },
    {
      "epoch": 13.146853146853147,
      "grad_norm": 0.18340010941028595,
      "learning_rate": 5.7082393698897166e-05,
      "loss": 0.0201,
      "step": 1880
    },
    {
      "epoch": 13.216783216783217,
      "grad_norm": 0.12412499636411667,
      "learning_rate": 5.60378494140976e-05,
      "loss": 0.0189,
      "step": 1890
    },
    {
      "epoch": 13.286713286713287,
      "grad_norm": 0.21642570197582245,
      "learning_rate": 5.4999217420242576e-05,
      "loss": 0.0178,
      "step": 1900
    },
    {
      "epoch": 13.356643356643357,
      "grad_norm": 0.17489974200725555,
      "learning_rate": 5.396663739872725e-05,
      "loss": 0.0173,
      "step": 1910
    },
    {
      "epoch": 13.426573426573427,
      "grad_norm": 0.12585236132144928,
      "learning_rate": 5.294024821704172e-05,
      "loss": 0.0206,
      "step": 1920
    },
    {
      "epoch": 13.496503496503497,
      "grad_norm": 0.12454655766487122,
      "learning_rate": 5.192018791009521e-05,
      "loss": 0.0172,
      "step": 1930
    },
    {
      "epoch": 13.566433566433567,
      "grad_norm": 0.1352427452802658,
      "learning_rate": 5.090659366165227e-05,
      "loss": 0.0185,
      "step": 1940
    },
    {
      "epoch": 13.636363636363637,
      "grad_norm": 0.15033438801765442,
      "learning_rate": 4.989960178588357e-05,
      "loss": 0.0177,
      "step": 1950
    },
    {
      "epoch": 13.706293706293707,
      "grad_norm": 0.2707054018974304,
      "learning_rate": 4.889934770903336e-05,
      "loss": 0.0165,
      "step": 1960
    },
    {
      "epoch": 13.776223776223777,
      "grad_norm": 0.19302873313426971,
      "learning_rate": 4.790596595120699e-05,
      "loss": 0.0158,
      "step": 1970
    },
    {
      "epoch": 13.846153846153847,
      "grad_norm": 0.14991451799869537,
      "learning_rate": 4.6919590108279254e-05,
      "loss": 0.0191,
      "step": 1980
    },
    {
      "epoch": 13.916083916083917,
      "grad_norm": 0.1627596765756607,
      "learning_rate": 4.594035283392815e-05,
      "loss": 0.0178,
      "step": 1990
    },
    {
      "epoch": 13.986013986013987,
      "grad_norm": 0.1320677548646927,
      "learning_rate": 4.49683858217944e-05,
      "loss": 0.0169,
      "step": 2000
    },
    {
      "epoch": 14.055944055944057,
      "grad_norm": 0.25264090299606323,
      "learning_rate": 4.4003819787770964e-05,
      "loss": 0.0179,
      "step": 2010
    },
    {
      "epoch": 14.125874125874127,
      "grad_norm": 0.11774495244026184,
      "learning_rate": 4.304678445242309e-05,
      "loss": 0.0152,
      "step": 2020
    },
    {
      "epoch": 14.195804195804195,
      "grad_norm": 0.17087212204933167,
      "learning_rate": 4.209740852354313e-05,
      "loss": 0.017,
      "step": 2030
    },
    {
      "epoch": 14.265734265734265,
      "grad_norm": 0.1586153358221054,
      "learning_rate": 4.115581967884094e-05,
      "loss": 0.0199,
      "step": 2040
    },
    {
      "epoch": 14.335664335664335,
      "grad_norm": 0.15853182971477509,
      "learning_rate": 4.022214454877305e-05,
      "loss": 0.0148,
      "step": 2050
    },
    {
      "epoch": 14.405594405594405,
      "grad_norm": 0.131947860121727,
      "learning_rate": 3.929650869951278e-05,
      "loss": 0.0178,
      "step": 2060
    },
    {
      "epoch": 14.475524475524475,
      "grad_norm": 0.26362699270248413,
      "learning_rate": 3.8379036616063066e-05,
      "loss": 0.0173,
      "step": 2070
    },
    {
      "epoch": 14.545454545454545,
      "grad_norm": 0.12325582653284073,
      "learning_rate": 3.746985168551532e-05,
      "loss": 0.0167,
      "step": 2080
    },
    {
      "epoch": 14.615384615384615,
      "grad_norm": 0.1346844732761383,
      "learning_rate": 3.65690761804554e-05,
      "loss": 0.0161,
      "step": 2090
    },
    {
      "epoch": 14.685314685314685,
      "grad_norm": 0.17594227194786072,
      "learning_rate": 3.567683124251972e-05,
      "loss": 0.0198,
      "step": 2100
    },
    {
      "epoch": 14.755244755244755,
      "grad_norm": 0.11365789920091629,
      "learning_rate": 3.4793236866103294e-05,
      "loss": 0.0138,
      "step": 2110
    },
    {
      "epoch": 14.825174825174825,
      "grad_norm": 0.11247463524341583,
      "learning_rate": 3.391841188222246e-05,
      "loss": 0.0137,
      "step": 2120
    },
    {
      "epoch": 14.895104895104895,
      "grad_norm": 0.17104235291481018,
      "learning_rate": 3.305247394253349e-05,
      "loss": 0.0171,
      "step": 2130
    },
    {
      "epoch": 14.965034965034965,
      "grad_norm": 0.11906550079584122,
      "learning_rate": 3.2195539503510164e-05,
      "loss": 0.0195,
      "step": 2140
    },
    {
      "epoch": 15.034965034965035,
      "grad_norm": 0.13180206716060638,
      "learning_rate": 3.1347723810782134e-05,
      "loss": 0.0147,
      "step": 2150
    },
    {
      "epoch": 15.104895104895105,
      "grad_norm": 0.27480393648147583,
      "learning_rate": 3.05091408836359e-05,
      "loss": 0.0142,
      "step": 2160
    },
    {
      "epoch": 15.174825174825175,
      "grad_norm": 0.19437533617019653,
      "learning_rate": 2.967990349968086e-05,
      "loss": 0.0159,
      "step": 2170
    },
    {
      "epoch": 15.244755244755245,
      "grad_norm": 0.10018286108970642,
      "learning_rate": 2.8860123179682242e-05,
      "loss": 0.0139,
      "step": 2180
    },
    {
      "epoch": 15.314685314685315,
      "grad_norm": 0.10660770535469055,
      "learning_rate": 2.8049910172563188e-05,
      "loss": 0.0138,
      "step": 2190
    },
    {
      "epoch": 15.384615384615385,
      "grad_norm": 0.19103875756263733,
      "learning_rate": 2.7249373440577963e-05,
      "loss": 0.0148,
      "step": 2200
    },
    {
      "epoch": 15.454545454545455,
      "grad_norm": 0.11443736404180527,
      "learning_rate": 2.6458620644657693e-05,
      "loss": 0.0193,
      "step": 2210
    },
    {
      "epoch": 15.524475524475525,
      "grad_norm": 0.1157824918627739,
      "learning_rate": 2.567775812993186e-05,
      "loss": 0.0179,
      "step": 2220
    },
    {
      "epoch": 15.594405594405595,
      "grad_norm": 0.09878715127706528,
      "learning_rate": 2.4906890911426208e-05,
      "loss": 0.0177,
      "step": 2230
    },
    {
      "epoch": 15.664335664335665,
      "grad_norm": 0.18489457666873932,
      "learning_rate": 2.4146122659939686e-05,
      "loss": 0.0127,
      "step": 2240
    },
    {
      "epoch": 15.734265734265735,
      "grad_norm": 0.15683647990226746,
      "learning_rate": 2.339555568810221e-05,
      "loss": 0.0152,
      "step": 2250
    },
    {
      "epoch": 15.804195804195803,
      "grad_norm": 0.13426269590854645,
      "learning_rate": 2.2655290936615093e-05,
      "loss": 0.0163,
      "step": 2260
    },
    {
      "epoch": 15.874125874125873,
      "grad_norm": 0.10421716421842575,
      "learning_rate": 2.1925427960675894e-05,
      "loss": 0.0148,
      "step": 2270
    },
    {
      "epoch": 15.944055944055943,
      "grad_norm": 0.20967915654182434,
      "learning_rate": 2.120606491658966e-05,
      "loss": 0.0128,
      "step": 2280
    },
    {
      "epoch": 16.013986013986013,
      "grad_norm": 0.11200631409883499,
      "learning_rate": 2.049729854856832e-05,
      "loss": 0.0133,
      "step": 2290
    },
    {
      "epoch": 16.083916083916083,
      "grad_norm": 0.10998646169900894,
      "learning_rate": 1.9799224175719767e-05,
      "loss": 0.015,
      "step": 2300
    },
    {
      "epoch": 16.153846153846153,
      "grad_norm": 0.1218709722161293,
      "learning_rate": 1.9111935679229142e-05,
      "loss": 0.0155,
      "step": 2310
    },
    {
      "epoch": 16.223776223776223,
      "grad_norm": 0.11821790039539337,
      "learning_rate": 1.843552548973272e-05,
      "loss": 0.0144,
      "step": 2320
    },
    {
      "epoch": 16.293706293706293,
      "grad_norm": 0.1463775783777237,
      "learning_rate": 1.7770084574887567e-05,
      "loss": 0.0117,
      "step": 2330
    },
    {
      "epoch": 16.363636363636363,
      "grad_norm": 0.15775668621063232,
      "learning_rate": 1.7115702427137616e-05,
      "loss": 0.0174,
      "step": 2340
    },
    {
      "epoch": 16.433566433566433,
      "grad_norm": 0.1515921652317047,
      "learning_rate": 1.647246705167812e-05,
      "loss": 0.0137,
      "step": 2350
    },
    {
      "epoch": 16.503496503496503,
      "grad_norm": 0.16316431760787964,
      "learning_rate": 1.5840464954620206e-05,
      "loss": 0.016,
      "step": 2360
    },
    {
      "epoch": 16.573426573426573,
      "grad_norm": 0.07553193718194962,
      "learning_rate": 1.5219781131357103e-05,
      "loss": 0.0162,
      "step": 2370
    },
    {
      "epoch": 16.643356643356643,
      "grad_norm": 0.1977609395980835,
      "learning_rate": 1.4610499055133375e-05,
      "loss": 0.014,
      "step": 2380
    },
    {
      "epoch": 16.713286713286713,
      "grad_norm": 0.2475104033946991,
      "learning_rate": 1.401270066581899e-05,
      "loss": 0.0157,
      "step": 2390
    },
    {
      "epoch": 16.783216783216783,
      "grad_norm": 0.20461061596870422,
      "learning_rate": 1.3426466358889545e-05,
      "loss": 0.0154,
      "step": 2400
    },
    {
      "epoch": 16.853146853146853,
      "grad_norm": 0.17063836753368378,
      "learning_rate": 1.2851874974614097e-05,
      "loss": 0.0132,
      "step": 2410
    },
    {
      "epoch": 16.923076923076923,
      "grad_norm": 0.16239117085933685,
      "learning_rate": 1.2289003787452557e-05,
      "loss": 0.0142,
      "step": 2420
    },
    {
      "epoch": 16.993006993006993,
      "grad_norm": 0.09241970628499985,
      "learning_rate": 1.1737928495662964e-05,
      "loss": 0.0179,
      "step": 2430
    },
    {
      "epoch": 17.062937062937063,
      "grad_norm": 0.104482501745224,
      "learning_rate": 1.1198723211121442e-05,
      "loss": 0.0142,
      "step": 2440
    },
    {
      "epoch": 17.132867132867133,
      "grad_norm": 0.10794669389724731,
      "learning_rate": 1.0671460449355075e-05,
      "loss": 0.014,
      "step": 2450
    },
    {
      "epoch": 17.202797202797203,
      "grad_norm": 0.11316874623298645,
      "learning_rate": 1.0156211119789583e-05,
      "loss": 0.0152,
      "step": 2460
    },
    {
      "epoch": 17.272727272727273,
      "grad_norm": 0.0924522653222084,
      "learning_rate": 9.65304451621304e-06,
      "loss": 0.0113,
      "step": 2470
    },
    {
      "epoch": 17.342657342657343,
      "grad_norm": 0.08180312067270279,
      "learning_rate": 9.162028307456771e-06,
      "loss": 0.0144,
      "step": 2480
    },
    {
      "epoch": 17.412587412587413,
      "grad_norm": 0.19703391194343567,
      "learning_rate": 8.683228528294929e-06,
      "loss": 0.0151,
      "step": 2490
    },
    {
      "epoch": 17.482517482517483,
      "grad_norm": 0.10454503446817398,
      "learning_rate": 8.216709570563685e-06,
      "loss": 0.0144,
      "step": 2500
    },
    {
      "epoch": 17.552447552447553,
      "grad_norm": 0.12390058487653732,
      "learning_rate": 7.76253417450149e-06,
      "loss": 0.0129,
      "step": 2510
    },
    {
      "epoch": 17.622377622377623,
      "grad_norm": 0.0972568616271019,
      "learning_rate": 7.320763420311261e-06,
      "loss": 0.0153,
      "step": 2520
    },
    {
      "epoch": 17.692307692307693,
      "grad_norm": 0.14459727704524994,
      "learning_rate": 6.891456719946188e-06,
      "loss": 0.0151,
      "step": 2530
    },
    {
      "epoch": 17.762237762237763,
      "grad_norm": 0.15339726209640503,
      "learning_rate": 6.4746718091194254e-06,
      "loss": 0.0125,
      "step": 2540
    },
    {
      "epoch": 17.832167832167833,
      "grad_norm": 0.2148740440607071,
      "learning_rate": 6.07046473953955e-06,
      "loss": 0.0126,
      "step": 2550
    },
    {
      "epoch": 17.902097902097903,
      "grad_norm": 0.08716779202222824,
      "learning_rate": 5.678889871372428e-06,
      "loss": 0.0143,
      "step": 2560
    },
    {
      "epoch": 17.972027972027973,
      "grad_norm": 0.2405886948108673,
      "learning_rate": 5.299999865930505e-06,
      "loss": 0.0159,
      "step": 2570
    },
    {
      "epoch": 18.041958041958043,
      "grad_norm": 0.10744497925043106,
      "learning_rate": 4.933845678590587e-06,
      "loss": 0.0121,
      "step": 2580
    },
    {
      "epoch": 18.111888111888113,
      "grad_norm": 0.24450422823429108,
      "learning_rate": 4.580476551941037e-06,
      "loss": 0.0119,
      "step": 2590
    },
    {
      "epoch": 18.181818181818183,
      "grad_norm": 0.13196663558483124,
      "learning_rate": 4.2399400091594154e-06,
      "loss": 0.0146,
      "step": 2600
    },
    {
      "epoch": 18.251748251748253,
      "grad_norm": 0.0889667496085167,
      "learning_rate": 3.912281847621213e-06,
      "loss": 0.0157,
      "step": 2610
    },
    {
      "epoch": 18.32167832167832,
      "grad_norm": 0.10608735680580139,
      "learning_rate": 3.597546132740792e-06,
      "loss": 0.0143,
      "step": 2620
    },
    {
      "epoch": 18.39160839160839,
      "grad_norm": 0.0715111494064331,
      "learning_rate": 3.295775192045181e-06,
      "loss": 0.012,
      "step": 2630
    },
    {
      "epoch": 18.46153846153846,
      "grad_norm": 0.11866133660078049,
      "learning_rate": 3.0070096094816037e-06,
      "loss": 0.0136,
      "step": 2640
    },
    {
      "epoch": 18.53146853146853,
      "grad_norm": 0.1221456378698349,
      "learning_rate": 2.7312882199595826e-06,
      "loss": 0.0143,
      "step": 2650
    },
    {
      "epoch": 18.6013986013986,
      "grad_norm": 0.11350057274103165,
      "learning_rate": 2.4686481041280574e-06,
      "loss": 0.0119,
      "step": 2660
    },
    {
      "epoch": 18.67132867132867,
      "grad_norm": 0.10500749945640564,
      "learning_rate": 2.2191245833886987e-06,
      "loss": 0.0131,
      "step": 2670
    },
    {
      "epoch": 18.74125874125874,
      "grad_norm": 0.208559051156044,
      "learning_rate": 1.9827512151456173e-06,
      "loss": 0.0137,
      "step": 2680
    },
    {
      "epoch": 18.81118881118881,
      "grad_norm": 0.10356497019529343,
      "learning_rate": 1.7595597882923309e-06,
      "loss": 0.0142,
      "step": 2690
    },
    {
      "epoch": 18.88111888111888,
      "grad_norm": 0.12341023981571198,
      "learning_rate": 1.549580318936672e-06,
      "loss": 0.0125,
      "step": 2700
    },
    {
      "epoch": 18.95104895104895,
      "grad_norm": 0.19302833080291748,
      "learning_rate": 1.3528410463639728e-06,
      "loss": 0.0135,
      "step": 2710
    },
    {
      "epoch": 19.02097902097902,
      "grad_norm": 0.18865394592285156,
      "learning_rate": 1.1693684292393704e-06,
      "loss": 0.0122,
      "step": 2720
    },
    {
      "epoch": 19.09090909090909,
      "grad_norm": 0.08378203958272934,
      "learning_rate": 9.991871420493736e-07,
      "loss": 0.0128,
      "step": 2730
    },
    {
      "epoch": 19.16083916083916,
      "grad_norm": 0.12866666913032532,
      "learning_rate": 8.423200717835977e-07,
      "loss": 0.014,
      "step": 2740
    },
    {
      "epoch": 19.23076923076923,
      "grad_norm": 0.13293299078941345,
      "learning_rate": 6.987883148567131e-07,
      "loss": 0.0112,
      "step": 2750
    },
    {
      "epoch": 19.3006993006993,
      "grad_norm": 0.12522312998771667,
      "learning_rate": 5.686111742713162e-07,
      "loss": 0.0119,
      "step": 2760
    },
    {
      "epoch": 19.37062937062937,
      "grad_norm": 0.10424693673849106,
      "learning_rate": 4.5180615702192783e-07,
      "loss": 0.0119,
      "step": 2770
    },
    {
      "epoch": 19.44055944055944,
      "grad_norm": 0.11369384080171585,
      "learning_rate": 3.4838897174055417e-07,
      "loss": 0.0144,
      "step": 2780
    },
    {
      "epoch": 19.51048951048951,
      "grad_norm": 0.279256671667099,
      "learning_rate": 2.583735265840992e-07,
      "loss": 0.0114,
      "step": 2790
    },
    {
      "epoch": 19.58041958041958,
      "grad_norm": 0.12756973505020142,
      "learning_rate": 1.8177192736390515e-07,
      "loss": 0.0124,
      "step": 2800
    },
    {
      "epoch": 19.65034965034965,
      "grad_norm": 0.13745011389255524,
      "learning_rate": 1.1859447591769934e-07,
      "loss": 0.0163,
      "step": 2810
    },
    {
      "epoch": 19.72027972027972,
      "grad_norm": 0.14598555862903595,
      "learning_rate": 6.884966872412468e-08,
      "loss": 0.0127,
      "step": 2820
    },
    {
      "epoch": 19.79020979020979,
      "grad_norm": 0.08950311690568924,
      "learning_rate": 3.254419576012024e-08,
      "loss": 0.0154,
      "step": 2830
    },
    {
      "epoch": 19.86013986013986,
      "grad_norm": 0.09730339050292969,
      "learning_rate": 9.682939601185492e-09,
      "loss": 0.0126,
      "step": 2840
    },
    {
      "epoch": 19.93006993006993,
      "grad_norm": 0.07458403706550598,
      "learning_rate": 2.689747647166563e-10,
      "loss": 0.0124,
      "step": 2850
    },
    {
      "epoch": 19.944055944055943,
      "step": 2852,
      "total_flos": 3.989050496377632e+17,
      "train_loss": 0.03802360156629376,
      "train_runtime": 3245.0123,
      "train_samples_per_second": 56.249,
      "train_steps_per_second": 0.879
    }
  ],
  "logging_steps": 10,
  "max_steps": 2852,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.989050496377632e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}