{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9994747347410442,
  "eval_steps": 500,
  "global_step": 2379,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008404244143292362,
      "grad_norm": 0.7383340001106262,
      "learning_rate": 2e-05,
      "loss": 2.4066,
      "step": 10
    },
    {
      "epoch": 0.016808488286584725,
      "grad_norm": 0.46394309401512146,
      "learning_rate": 4e-05,
      "loss": 2.0375,
      "step": 20
    },
    {
      "epoch": 0.025212732429877087,
      "grad_norm": 0.4739478528499603,
      "learning_rate": 6e-05,
      "loss": 1.5044,
      "step": 30
    },
    {
      "epoch": 0.03361697657316945,
      "grad_norm": 0.20930196344852448,
      "learning_rate": 8e-05,
      "loss": 0.8704,
      "step": 40
    },
    {
      "epoch": 0.04202122071646181,
      "grad_norm": 0.15288038551807404,
      "learning_rate": 0.0001,
      "loss": 0.6533,
      "step": 50
    },
    {
      "epoch": 0.050425464859754174,
      "grad_norm": 0.13073962926864624,
      "learning_rate": 0.00012,
      "loss": 0.586,
      "step": 60
    },
    {
      "epoch": 0.058829709003046536,
      "grad_norm": 0.14555367827415466,
      "learning_rate": 0.00014,
      "loss": 0.5793,
      "step": 70
    },
    {
      "epoch": 0.0672339531463389,
      "grad_norm": 0.12397414445877075,
      "learning_rate": 0.00016,
      "loss": 0.581,
      "step": 80
    },
    {
      "epoch": 0.07563819728963127,
      "grad_norm": 0.13021130859851837,
      "learning_rate": 0.00018,
      "loss": 0.5512,
      "step": 90
    },
    {
      "epoch": 0.08404244143292362,
      "grad_norm": 0.13012883067131042,
      "learning_rate": 0.0002,
      "loss": 0.5403,
      "step": 100
    },
    {
      "epoch": 0.09244668557621599,
      "grad_norm": 0.11942347884178162,
      "learning_rate": 0.00019942313239111625,
      "loss": 0.5247,
      "step": 110
    },
    {
      "epoch": 0.10085092971950835,
      "grad_norm": 0.11690942198038101,
      "learning_rate": 0.0001988462647822325,
      "loss": 0.5417,
      "step": 120
    },
    {
      "epoch": 0.10925517386280072,
      "grad_norm": 0.1355101615190506,
      "learning_rate": 0.00019826939717334873,
      "loss": 0.5273,
      "step": 130
    },
    {
      "epoch": 0.11765941800609307,
      "grad_norm": 0.1345665603876114,
      "learning_rate": 0.00019769252956446497,
      "loss": 0.5243,
      "step": 140
    },
    {
      "epoch": 0.12606366214938544,
      "grad_norm": 0.12515193223953247,
      "learning_rate": 0.0001971156619555812,
      "loss": 0.5344,
      "step": 150
    },
    {
      "epoch": 0.1344679062926778,
      "grad_norm": 0.15686553716659546,
      "learning_rate": 0.00019653879434669745,
      "loss": 0.5118,
      "step": 160
    },
    {
      "epoch": 0.14287215043597015,
      "grad_norm": 0.12068944424390793,
      "learning_rate": 0.0001959619267378137,
      "loss": 0.4979,
      "step": 170
    },
    {
      "epoch": 0.15127639457926254,
      "grad_norm": 0.13319459557533264,
      "learning_rate": 0.00019538505912892993,
      "loss": 0.503,
      "step": 180
    },
    {
      "epoch": 0.1596806387225549,
      "grad_norm": 0.11806949228048325,
      "learning_rate": 0.00019480819152004617,
      "loss": 0.49,
      "step": 190
    },
    {
      "epoch": 0.16808488286584725,
      "grad_norm": 0.12932075560092926,
      "learning_rate": 0.00019423132391116238,
      "loss": 0.514,
      "step": 200
    },
    {
      "epoch": 0.17648912700913963,
      "grad_norm": 0.11743929982185364,
      "learning_rate": 0.00019365445630227862,
      "loss": 0.4788,
      "step": 210
    },
    {
      "epoch": 0.18489337115243198,
      "grad_norm": 0.11788313835859299,
      "learning_rate": 0.00019307758869339486,
      "loss": 0.4891,
      "step": 220
    },
    {
      "epoch": 0.19329761529572434,
      "grad_norm": 0.11414741724729538,
      "learning_rate": 0.0001925007210845111,
      "loss": 0.5033,
      "step": 230
    },
    {
      "epoch": 0.2017018594390167,
      "grad_norm": 0.11419043689966202,
      "learning_rate": 0.00019192385347562737,
      "loss": 0.4844,
      "step": 240
    },
    {
      "epoch": 0.21010610358230908,
      "grad_norm": 0.12788020074367523,
      "learning_rate": 0.0001913469858667436,
      "loss": 0.4697,
      "step": 250
    },
    {
      "epoch": 0.21851034772560143,
      "grad_norm": 0.13661302626132965,
      "learning_rate": 0.00019077011825785982,
      "loss": 0.4627,
      "step": 260
    },
    {
      "epoch": 0.2269145918688938,
      "grad_norm": 0.12041325867176056,
      "learning_rate": 0.00019019325064897606,
      "loss": 0.4964,
      "step": 270
    },
    {
      "epoch": 0.23531883601218614,
      "grad_norm": 0.133742094039917,
      "learning_rate": 0.0001896163830400923,
      "loss": 0.4658,
      "step": 280
    },
    {
      "epoch": 0.24372308015547853,
      "grad_norm": 0.1261977106332779,
      "learning_rate": 0.00018903951543120854,
      "loss": 0.4781,
      "step": 290
    },
    {
      "epoch": 0.2521273242987709,
      "grad_norm": 0.130150705575943,
      "learning_rate": 0.00018846264782232478,
      "loss": 0.4922,
      "step": 300
    },
    {
      "epoch": 0.26053156844206327,
      "grad_norm": 0.13174410164356232,
      "learning_rate": 0.00018788578021344102,
      "loss": 0.4559,
      "step": 310
    },
    {
      "epoch": 0.2689358125853556,
      "grad_norm": 0.1186077669262886,
      "learning_rate": 0.00018730891260455726,
      "loss": 0.4722,
      "step": 320
    },
    {
      "epoch": 0.277340056728648,
      "grad_norm": 0.116569384932518,
      "learning_rate": 0.0001867320449956735,
      "loss": 0.4457,
      "step": 330
    },
    {
      "epoch": 0.2857443008719403,
      "grad_norm": 0.12219471484422684,
      "learning_rate": 0.00018615517738678974,
      "loss": 0.4849,
      "step": 340
    },
    {
      "epoch": 0.2941485450152327,
      "grad_norm": 0.12746909260749817,
      "learning_rate": 0.00018557830977790598,
      "loss": 0.4821,
      "step": 350
    },
    {
      "epoch": 0.30255278915852507,
      "grad_norm": 0.14125944674015045,
      "learning_rate": 0.00018500144216902222,
      "loss": 0.4605,
      "step": 360
    },
    {
      "epoch": 0.3109570333018174,
      "grad_norm": 0.19157269597053528,
      "learning_rate": 0.00018442457456013846,
      "loss": 0.4541,
      "step": 370
    },
    {
      "epoch": 0.3193612774451098,
      "grad_norm": 0.12603330612182617,
      "learning_rate": 0.0001838477069512547,
      "loss": 0.4536,
      "step": 380
    },
    {
      "epoch": 0.32776552158840216,
      "grad_norm": 0.12653909623622894,
      "learning_rate": 0.00018327083934237091,
      "loss": 0.4468,
      "step": 390
    },
    {
      "epoch": 0.3361697657316945,
      "grad_norm": 0.15930472314357758,
      "learning_rate": 0.00018269397173348718,
      "loss": 0.4542,
      "step": 400
    },
    {
      "epoch": 0.3445740098749869,
      "grad_norm": 0.13266988098621368,
      "learning_rate": 0.00018211710412460342,
      "loss": 0.4335,
      "step": 410
    },
    {
      "epoch": 0.35297825401827926,
      "grad_norm": 0.12103667855262756,
      "learning_rate": 0.00018154023651571966,
      "loss": 0.4575,
      "step": 420
    },
    {
      "epoch": 0.3613824981615716,
      "grad_norm": 0.14439740777015686,
      "learning_rate": 0.0001809633689068359,
      "loss": 0.4377,
      "step": 430
    },
    {
      "epoch": 0.36978674230486397,
      "grad_norm": 0.12652407586574554,
      "learning_rate": 0.00018038650129795214,
      "loss": 0.4363,
      "step": 440
    },
    {
      "epoch": 0.3781909864481563,
      "grad_norm": 0.14594405889511108,
      "learning_rate": 0.00017980963368906835,
      "loss": 0.4306,
      "step": 450
    },
    {
      "epoch": 0.3865952305914487,
      "grad_norm": 0.12562687695026398,
      "learning_rate": 0.0001792327660801846,
      "loss": 0.4501,
      "step": 460
    },
    {
      "epoch": 0.39499947473474106,
      "grad_norm": 0.14584492146968842,
      "learning_rate": 0.00017865589847130083,
      "loss": 0.4509,
      "step": 470
    },
    {
      "epoch": 0.4034037188780334,
      "grad_norm": 0.13192500174045563,
      "learning_rate": 0.00017807903086241707,
      "loss": 0.4505,
      "step": 480
    },
    {
      "epoch": 0.4118079630213258,
      "grad_norm": 0.14266645908355713,
      "learning_rate": 0.00017750216325353331,
      "loss": 0.4585,
      "step": 490
    },
    {
      "epoch": 0.42021220716461816,
      "grad_norm": 0.1400412619113922,
      "learning_rate": 0.00017692529564464958,
      "loss": 0.4365,
      "step": 500
    },
    {
      "epoch": 0.4286164513079105,
      "grad_norm": 0.14728468656539917,
      "learning_rate": 0.0001763484280357658,
      "loss": 0.4303,
      "step": 510
    },
    {
      "epoch": 0.43702069545120287,
      "grad_norm": 0.15791365504264832,
      "learning_rate": 0.00017577156042688203,
      "loss": 0.4407,
      "step": 520
    },
    {
      "epoch": 0.4454249395944952,
      "grad_norm": 0.15447258949279785,
      "learning_rate": 0.00017519469281799827,
      "loss": 0.4365,
      "step": 530
    },
    {
      "epoch": 0.4538291837377876,
      "grad_norm": 0.1518252044916153,
      "learning_rate": 0.00017461782520911451,
      "loss": 0.4305,
      "step": 540
    },
    {
      "epoch": 0.46223342788107996,
      "grad_norm": 0.1154065877199173,
      "learning_rate": 0.00017404095760023075,
      "loss": 0.4212,
      "step": 550
    },
    {
      "epoch": 0.4706376720243723,
      "grad_norm": 0.12900012731552124,
      "learning_rate": 0.000173464089991347,
      "loss": 0.4277,
      "step": 560
    },
    {
      "epoch": 0.47904191616766467,
      "grad_norm": 0.1349458247423172,
      "learning_rate": 0.00017288722238246323,
      "loss": 0.4051,
      "step": 570
    },
    {
      "epoch": 0.48744616031095706,
      "grad_norm": 0.16337165236473083,
      "learning_rate": 0.00017231035477357947,
      "loss": 0.407,
      "step": 580
    },
    {
      "epoch": 0.4958504044542494,
      "grad_norm": 0.13420593738555908,
      "learning_rate": 0.0001717334871646957,
      "loss": 0.4138,
      "step": 590
    },
    {
      "epoch": 0.5042546485975418,
      "grad_norm": 0.13840581476688385,
      "learning_rate": 0.00017115661955581195,
      "loss": 0.4099,
      "step": 600
    },
    {
      "epoch": 0.5126588927408341,
      "grad_norm": 0.1378021389245987,
      "learning_rate": 0.0001705797519469282,
      "loss": 0.4254,
      "step": 610
    },
    {
      "epoch": 0.5210631368841265,
      "grad_norm": 0.1607150137424469,
      "learning_rate": 0.00017000288433804443,
      "loss": 0.4353,
      "step": 620
    },
    {
      "epoch": 0.5294673810274189,
      "grad_norm": 0.13462169468402863,
      "learning_rate": 0.00016942601672916067,
      "loss": 0.4267,
      "step": 630
    },
    {
      "epoch": 0.5378716251707112,
      "grad_norm": 0.14311543107032776,
      "learning_rate": 0.00016884914912027689,
      "loss": 0.4301,
      "step": 640
    },
    {
      "epoch": 0.5462758693140036,
      "grad_norm": 0.15559442341327667,
      "learning_rate": 0.00016827228151139313,
      "loss": 0.4102,
      "step": 650
    },
    {
      "epoch": 0.554680113457296,
      "grad_norm": 0.15557149052619934,
      "learning_rate": 0.00016769541390250937,
      "loss": 0.4136,
      "step": 660
    },
    {
      "epoch": 0.5630843576005883,
      "grad_norm": 0.135511115193367,
      "learning_rate": 0.00016711854629362563,
      "loss": 0.4153,
      "step": 670
    },
    {
      "epoch": 0.5714886017438806,
      "grad_norm": 0.13760776817798615,
      "learning_rate": 0.00016654167868474187,
      "loss": 0.4145,
      "step": 680
    },
    {
      "epoch": 0.579892845887173,
      "grad_norm": 0.14971590042114258,
      "learning_rate": 0.0001659648110758581,
      "loss": 0.3875,
      "step": 690
    },
    {
      "epoch": 0.5882970900304654,
      "grad_norm": 0.16005663573741913,
      "learning_rate": 0.00016538794346697433,
      "loss": 0.3938,
      "step": 700
    },
    {
      "epoch": 0.5967013341737577,
      "grad_norm": 0.1625218689441681,
      "learning_rate": 0.00016481107585809057,
      "loss": 0.3871,
      "step": 710
    },
    {
      "epoch": 0.6051055783170501,
      "grad_norm": 0.17047689855098724,
      "learning_rate": 0.0001642342082492068,
      "loss": 0.412,
      "step": 720
    },
    {
      "epoch": 0.6135098224603425,
      "grad_norm": 0.13825903832912445,
      "learning_rate": 0.00016365734064032305,
      "loss": 0.3948,
      "step": 730
    },
    {
      "epoch": 0.6219140666036348,
      "grad_norm": 0.14830929040908813,
      "learning_rate": 0.00016308047303143929,
      "loss": 0.3927,
      "step": 740
    },
    {
      "epoch": 0.6303183107469272,
      "grad_norm": 0.13950933516025543,
      "learning_rate": 0.00016250360542255553,
      "loss": 0.4051,
      "step": 750
    },
    {
      "epoch": 0.6387225548902196,
      "grad_norm": 0.15511371195316315,
      "learning_rate": 0.0001619267378136718,
      "loss": 0.4041,
      "step": 760
    },
    {
      "epoch": 0.6471267990335119,
      "grad_norm": 0.14828190207481384,
      "learning_rate": 0.000161349870204788,
      "loss": 0.3824,
      "step": 770
    },
    {
      "epoch": 0.6555310431768043,
      "grad_norm": 0.144051194190979,
      "learning_rate": 0.00016077300259590425,
      "loss": 0.3829,
      "step": 780
    },
    {
      "epoch": 0.6639352873200967,
      "grad_norm": 0.14780694246292114,
      "learning_rate": 0.00016019613498702049,
      "loss": 0.3814,
      "step": 790
    },
    {
      "epoch": 0.672339531463389,
      "grad_norm": 0.15042325854301453,
      "learning_rate": 0.00015961926737813673,
      "loss": 0.3962,
      "step": 800
    },
    {
      "epoch": 0.6807437756066814,
      "grad_norm": 0.16325107216835022,
      "learning_rate": 0.00015904239976925297,
      "loss": 0.3801,
      "step": 810
    },
    {
      "epoch": 0.6891480197499738,
      "grad_norm": 0.14843328297138214,
      "learning_rate": 0.0001584655321603692,
      "loss": 0.4082,
      "step": 820
    },
    {
      "epoch": 0.6975522638932661,
      "grad_norm": 0.16731064021587372,
      "learning_rate": 0.00015788866455148545,
      "loss": 0.4192,
      "step": 830
    },
    {
      "epoch": 0.7059565080365585,
      "grad_norm": 0.18703435361385345,
      "learning_rate": 0.00015731179694260169,
      "loss": 0.4009,
      "step": 840
    },
    {
      "epoch": 0.7143607521798508,
      "grad_norm": 0.13935630023479462,
      "learning_rate": 0.00015673492933371793,
      "loss": 0.3618,
      "step": 850
    },
    {
      "epoch": 0.7227649963231432,
      "grad_norm": 0.13263636827468872,
      "learning_rate": 0.00015615806172483417,
      "loss": 0.3963,
      "step": 860
    },
    {
      "epoch": 0.7311692404664355,
      "grad_norm": 0.14940643310546875,
      "learning_rate": 0.0001555811941159504,
      "loss": 0.3585,
      "step": 870
    },
    {
      "epoch": 0.7395734846097279,
      "grad_norm": 0.14807912707328796,
      "learning_rate": 0.00015500432650706665,
      "loss": 0.3748,
      "step": 880
    },
    {
      "epoch": 0.7479777287530203,
      "grad_norm": 0.15254080295562744,
      "learning_rate": 0.00015442745889818286,
      "loss": 0.3718,
      "step": 890
    },
    {
      "epoch": 0.7563819728963126,
      "grad_norm": 0.16590768098831177,
      "learning_rate": 0.0001538505912892991,
      "loss": 0.386,
      "step": 900
    },
    {
      "epoch": 0.764786217039605,
      "grad_norm": 0.15733902156352997,
      "learning_rate": 0.00015327372368041534,
      "loss": 0.3756,
      "step": 910
    },
    {
      "epoch": 0.7731904611828974,
      "grad_norm": 0.13757385313510895,
      "learning_rate": 0.00015269685607153158,
      "loss": 0.3843,
      "step": 920
    },
    {
      "epoch": 0.7815947053261897,
      "grad_norm": 0.14952607452869415,
      "learning_rate": 0.00015211998846264784,
      "loss": 0.3634,
      "step": 930
    },
    {
      "epoch": 0.7899989494694821,
      "grad_norm": 0.1516282558441162,
      "learning_rate": 0.00015154312085376408,
      "loss": 0.3798,
      "step": 940
    },
    {
      "epoch": 0.7984031936127745,
      "grad_norm": 0.17785628139972687,
      "learning_rate": 0.00015096625324488032,
      "loss": 0.3681,
      "step": 950
    },
    {
      "epoch": 0.8068074377560668,
      "grad_norm": 0.171351820230484,
      "learning_rate": 0.00015038938563599654,
      "loss": 0.3686,
      "step": 960
    },
    {
      "epoch": 0.8152116818993592,
      "grad_norm": 0.1742231398820877,
      "learning_rate": 0.00014981251802711278,
      "loss": 0.3792,
      "step": 970
    },
    {
      "epoch": 0.8236159260426515,
      "grad_norm": 0.16650599241256714,
      "learning_rate": 0.00014923565041822902,
      "loss": 0.3577,
      "step": 980
    },
    {
      "epoch": 0.8320201701859439,
      "grad_norm": 0.1497887670993805,
      "learning_rate": 0.00014865878280934526,
      "loss": 0.3553,
      "step": 990
    },
    {
      "epoch": 0.8404244143292363,
      "grad_norm": 0.14781557023525238,
      "learning_rate": 0.0001480819152004615,
      "loss": 0.3538,
      "step": 1000
    },
    {
      "epoch": 0.8488286584725286,
      "grad_norm": 0.15724751353263855,
      "learning_rate": 0.00014750504759157774,
      "loss": 0.3597,
      "step": 1010
    },
    {
      "epoch": 0.857232902615821,
      "grad_norm": 0.18635571002960205,
      "learning_rate": 0.00014692817998269398,
      "loss": 0.3615,
      "step": 1020
    },
    {
      "epoch": 0.8656371467591134,
      "grad_norm": 0.17742526531219482,
      "learning_rate": 0.00014635131237381022,
      "loss": 0.348,
      "step": 1030
    },
    {
      "epoch": 0.8740413909024057,
      "grad_norm": 0.20535768568515778,
      "learning_rate": 0.00014577444476492646,
      "loss": 0.3343,
      "step": 1040
    },
    {
      "epoch": 0.8824456350456981,
      "grad_norm": 0.18968522548675537,
      "learning_rate": 0.0001451975771560427,
      "loss": 0.3615,
      "step": 1050
    },
    {
      "epoch": 0.8908498791889904,
      "grad_norm": 0.1528492122888565,
      "learning_rate": 0.00014462070954715894,
      "loss": 0.3786,
      "step": 1060
    },
    {
      "epoch": 0.8992541233322828,
      "grad_norm": 0.15841075778007507,
      "learning_rate": 0.00014404384193827518,
      "loss": 0.3761,
      "step": 1070
    },
    {
      "epoch": 0.9076583674755752,
      "grad_norm": 0.15167982876300812,
      "learning_rate": 0.0001434669743293914,
      "loss": 0.3528,
      "step": 1080
    },
    {
      "epoch": 0.9160626116188675,
      "grad_norm": 0.14096671342849731,
      "learning_rate": 0.00014289010672050766,
      "loss": 0.371,
      "step": 1090
    },
    {
      "epoch": 0.9244668557621599,
      "grad_norm": 0.1579194813966751,
      "learning_rate": 0.0001423132391116239,
      "loss": 0.3491,
      "step": 1100
    },
    {
      "epoch": 0.9328710999054523,
      "grad_norm": 0.16789057850837708,
      "learning_rate": 0.00014173637150274014,
      "loss": 0.3536,
      "step": 1110
    },
    {
      "epoch": 0.9412753440487446,
      "grad_norm": 0.13980717957019806,
      "learning_rate": 0.00014115950389385638,
      "loss": 0.3423,
      "step": 1120
    },
    {
      "epoch": 0.949679588192037,
      "grad_norm": 0.19879643619060516,
      "learning_rate": 0.00014058263628497262,
      "loss": 0.3285,
      "step": 1130
    },
    {
      "epoch": 0.9580838323353293,
      "grad_norm": 0.16574440896511078,
      "learning_rate": 0.00014000576867608886,
      "loss": 0.3568,
      "step": 1140
    },
    {
      "epoch": 0.9664880764786217,
      "grad_norm": 0.15376180410385132,
      "learning_rate": 0.00013942890106720507,
      "loss": 0.3558,
      "step": 1150
    },
    {
      "epoch": 0.9748923206219141,
      "grad_norm": 0.17232170701026917,
      "learning_rate": 0.0001388520334583213,
      "loss": 0.342,
      "step": 1160
    },
    {
      "epoch": 0.9832965647652064,
      "grad_norm": 0.1959993690252304,
      "learning_rate": 0.00013827516584943755,
      "loss": 0.3458,
      "step": 1170
    },
    {
      "epoch": 0.9917008089084988,
      "grad_norm": 0.14029347896575928,
      "learning_rate": 0.0001376982982405538,
      "loss": 0.3297,
      "step": 1180
    },
    {
      "epoch": 1.0002101061035824,
      "grad_norm": 0.20758652687072754,
      "learning_rate": 0.00013712143063167006,
      "loss": 0.3642,
      "step": 1190
    },
    {
      "epoch": 1.0086143502468747,
      "grad_norm": 0.15599438548088074,
      "learning_rate": 0.0001365445630227863,
      "loss": 0.3004,
      "step": 1200
    },
    {
      "epoch": 1.017018594390167,
      "grad_norm": 0.16680683195590973,
      "learning_rate": 0.0001359676954139025,
      "loss": 0.2915,
      "step": 1210
    },
    {
      "epoch": 1.0254228385334594,
      "grad_norm": 0.1668105274438858,
      "learning_rate": 0.00013539082780501875,
      "loss": 0.2963,
      "step": 1220
    },
    {
      "epoch": 1.0338270826767517,
      "grad_norm": 0.16461539268493652,
      "learning_rate": 0.000134813960196135,
      "loss": 0.3041,
      "step": 1230
    },
    {
      "epoch": 1.042231326820044,
      "grad_norm": 0.18869394063949585,
      "learning_rate": 0.00013423709258725123,
      "loss": 0.3046,
      "step": 1240
    },
    {
      "epoch": 1.0506355709633364,
      "grad_norm": 0.16899700462818146,
      "learning_rate": 0.00013366022497836747,
      "loss": 0.2921,
      "step": 1250
    },
    {
      "epoch": 1.059039815106629,
      "grad_norm": 0.1905297338962555,
      "learning_rate": 0.0001330833573694837,
      "loss": 0.2879,
      "step": 1260
    },
    {
      "epoch": 1.0674440592499213,
      "grad_norm": 0.17273731529712677,
      "learning_rate": 0.00013250648976059995,
      "loss": 0.3038,
      "step": 1270
    },
    {
      "epoch": 1.0758483033932136,
      "grad_norm": 0.1947745531797409,
      "learning_rate": 0.0001319296221517162,
      "loss": 0.3029,
      "step": 1280
    },
    {
      "epoch": 1.084252547536506,
      "grad_norm": 0.1741725355386734,
      "learning_rate": 0.00013135275454283243,
      "loss": 0.3073,
      "step": 1290
    },
    {
      "epoch": 1.0926567916797982,
      "grad_norm": 0.18244194984436035,
      "learning_rate": 0.00013077588693394867,
      "loss": 0.287,
      "step": 1300
    },
    {
      "epoch": 1.1010610358230906,
      "grad_norm": 0.18360966444015503,
      "learning_rate": 0.0001301990193250649,
      "loss": 0.307,
      "step": 1310
    },
    {
      "epoch": 1.1094652799663831,
      "grad_norm": 0.16066686809062958,
      "learning_rate": 0.00012962215171618115,
      "loss": 0.2712,
      "step": 1320
    },
    {
      "epoch": 1.1178695241096754,
      "grad_norm": 0.16239213943481445,
      "learning_rate": 0.00012904528410729736,
      "loss": 0.2857,
      "step": 1330
    },
    {
      "epoch": 1.1262737682529678,
      "grad_norm": 0.16966617107391357,
      "learning_rate": 0.0001284684164984136,
      "loss": 0.3087,
      "step": 1340
    },
    {
      "epoch": 1.13467801239626,
      "grad_norm": 0.16753819584846497,
      "learning_rate": 0.00012789154888952984,
      "loss": 0.2852,
      "step": 1350
    },
    {
      "epoch": 1.1430822565395524,
      "grad_norm": 0.19184084236621857,
      "learning_rate": 0.0001273146812806461,
      "loss": 0.3138,
      "step": 1360
    },
    {
      "epoch": 1.1514865006828447,
      "grad_norm": 0.15949766337871552,
      "learning_rate": 0.00012673781367176235,
      "loss": 0.2812,
      "step": 1370
    },
    {
      "epoch": 1.159890744826137,
      "grad_norm": 0.16187496483325958,
      "learning_rate": 0.0001261609460628786,
      "loss": 0.2841,
      "step": 1380
    },
    {
      "epoch": 1.1682949889694296,
      "grad_norm": 0.1778268665075302,
      "learning_rate": 0.00012558407845399483,
      "loss": 0.3181,
      "step": 1390
    },
    {
      "epoch": 1.176699233112722,
      "grad_norm": 0.17179737985134125,
      "learning_rate": 0.00012500721084511104,
      "loss": 0.2904,
      "step": 1400
    },
    {
      "epoch": 1.1851034772560143,
      "grad_norm": 0.16989010572433472,
      "learning_rate": 0.00012443034323622728,
      "loss": 0.2856,
      "step": 1410
    },
    {
      "epoch": 1.1935077213993066,
      "grad_norm": 0.21040703356266022,
      "learning_rate": 0.00012385347562734352,
      "loss": 0.2743,
      "step": 1420
    },
    {
      "epoch": 1.201911965542599,
      "grad_norm": 0.19255656003952026,
      "learning_rate": 0.00012327660801845976,
      "loss": 0.316,
      "step": 1430
    },
    {
      "epoch": 1.2103162096858915,
      "grad_norm": 0.16303245723247528,
      "learning_rate": 0.000122699740409576,
      "loss": 0.2671,
      "step": 1440
    },
    {
      "epoch": 1.2187204538291838,
      "grad_norm": 0.21385671198368073,
      "learning_rate": 0.00012212287280069227,
      "loss": 0.2865,
      "step": 1450
    },
    {
      "epoch": 1.2271246979724761,
      "grad_norm": 0.18770861625671387,
      "learning_rate": 0.00012154600519180848,
      "loss": 0.2795,
      "step": 1460
    },
    {
      "epoch": 1.2355289421157685,
      "grad_norm": 0.20827870070934296,
      "learning_rate": 0.00012096913758292472,
      "loss": 0.2769,
      "step": 1470
    },
    {
      "epoch": 1.2439331862590608,
      "grad_norm": 0.1704486757516861,
      "learning_rate": 0.00012039226997404096,
      "loss": 0.2993,
      "step": 1480
    },
    {
      "epoch": 1.2523374304023531,
      "grad_norm": 0.21233461797237396,
      "learning_rate": 0.0001198154023651572,
      "loss": 0.2912,
      "step": 1490
    },
    {
      "epoch": 1.2607416745456455,
      "grad_norm": 0.1879620999097824,
      "learning_rate": 0.00011923853475627344,
      "loss": 0.2885,
      "step": 1500
    },
    {
      "epoch": 1.2691459186889378,
      "grad_norm": 0.14288674294948578,
      "learning_rate": 0.00011866166714738968,
      "loss": 0.2794,
      "step": 1510
    },
    {
      "epoch": 1.2775501628322303,
      "grad_norm": 0.1654644012451172,
      "learning_rate": 0.00011808479953850591,
      "loss": 0.2762,
      "step": 1520
    },
    {
      "epoch": 1.2859544069755227,
      "grad_norm": 0.15648572146892548,
      "learning_rate": 0.00011750793192962215,
      "loss": 0.2853,
      "step": 1530
    },
    {
      "epoch": 1.294358651118815,
      "grad_norm": 0.14321617782115936,
      "learning_rate": 0.00011693106432073839,
      "loss": 0.2949,
      "step": 1540
    },
    {
      "epoch": 1.3027628952621073,
      "grad_norm": 0.18823479115962982,
      "learning_rate": 0.00011635419671185464,
      "loss": 0.2734,
      "step": 1550
    },
    {
      "epoch": 1.3111671394053999,
      "grad_norm": 0.1524640917778015,
      "learning_rate": 0.00011577732910297088,
      "loss": 0.2668,
      "step": 1560
    },
    {
      "epoch": 1.3195713835486922,
      "grad_norm": 0.1731933057308197,
      "learning_rate": 0.00011520046149408712,
      "loss": 0.2815,
      "step": 1570
    },
    {
      "epoch": 1.3279756276919845,
      "grad_norm": 0.19858598709106445,
      "learning_rate": 0.00011462359388520336,
      "loss": 0.2863,
      "step": 1580
    },
    {
      "epoch": 1.3363798718352768,
      "grad_norm": 0.20350554585456848,
      "learning_rate": 0.00011404672627631959,
      "loss": 0.2974,
      "step": 1590
    },
    {
      "epoch": 1.3447841159785692,
      "grad_norm": 0.16735605895519257,
      "learning_rate": 0.00011346985866743583,
      "loss": 0.2742,
      "step": 1600
    },
    {
      "epoch": 1.3531883601218615,
      "grad_norm": 0.18708328902721405,
      "learning_rate": 0.00011289299105855207,
      "loss": 0.2877,
      "step": 1610
    },
    {
      "epoch": 1.3615926042651538,
      "grad_norm": 0.19334456324577332,
      "learning_rate": 0.00011231612344966831,
      "loss": 0.2735,
      "step": 1620
    },
    {
      "epoch": 1.3699968484084462,
      "grad_norm": 0.20367129147052765,
      "learning_rate": 0.00011173925584078455,
      "loss": 0.2801,
      "step": 1630
    },
    {
      "epoch": 1.3784010925517387,
      "grad_norm": 0.18539854884147644,
      "learning_rate": 0.00011116238823190079,
      "loss": 0.2842,
      "step": 1640
    },
    {
      "epoch": 1.386805336695031,
      "grad_norm": 0.2150140106678009,
      "learning_rate": 0.00011058552062301701,
      "loss": 0.2611,
      "step": 1650
    },
    {
      "epoch": 1.3952095808383234,
      "grad_norm": 0.162113755941391,
      "learning_rate": 0.00011000865301413325,
      "loss": 0.289,
      "step": 1660
    },
    {
      "epoch": 1.4036138249816157,
      "grad_norm": 0.18180853128433228,
      "learning_rate": 0.0001094317854052495,
      "loss": 0.2808,
      "step": 1670
    },
    {
      "epoch": 1.412018069124908,
      "grad_norm": 0.17916476726531982,
      "learning_rate": 0.00010885491779636575,
      "loss": 0.2912,
      "step": 1680
    },
    {
      "epoch": 1.4204223132682006,
      "grad_norm": 0.22721944749355316,
      "learning_rate": 0.00010827805018748199,
      "loss": 0.2611,
      "step": 1690
    },
    {
      "epoch": 1.428826557411493,
      "grad_norm": 0.16184848546981812,
      "learning_rate": 0.00010770118257859823,
      "loss": 0.2722,
      "step": 1700
    },
    {
      "epoch": 1.4372308015547852,
      "grad_norm": 0.19588448107242584,
      "learning_rate": 0.00010712431496971444,
      "loss": 0.2817,
      "step": 1710
    },
    {
      "epoch": 1.4456350456980775,
      "grad_norm": 0.1870766133069992,
      "learning_rate": 0.0001065474473608307,
      "loss": 0.2835,
      "step": 1720
    },
    {
      "epoch": 1.4540392898413699,
      "grad_norm": 0.1768248826265335,
      "learning_rate": 0.00010597057975194693,
      "loss": 0.2643,
      "step": 1730
    },
    {
      "epoch": 1.4624435339846622,
      "grad_norm": 0.1726955771446228,
      "learning_rate": 0.00010539371214306317,
      "loss": 0.2674,
      "step": 1740
    },
    {
      "epoch": 1.4708477781279545,
      "grad_norm": 0.1709883064031601,
      "learning_rate": 0.00010481684453417941,
      "loss": 0.262,
      "step": 1750
    },
    {
      "epoch": 1.4792520222712469,
      "grad_norm": 0.2008083164691925,
      "learning_rate": 0.00010423997692529565,
      "loss": 0.2634,
      "step": 1760
    },
    {
      "epoch": 1.4876562664145394,
      "grad_norm": 0.17773209512233734,
      "learning_rate": 0.0001036631093164119,
      "loss": 0.2805,
      "step": 1770
    },
    {
      "epoch": 1.4960605105578317,
      "grad_norm": 0.18000538647174835,
      "learning_rate": 0.00010308624170752812,
      "loss": 0.2443,
      "step": 1780
    },
    {
      "epoch": 1.504464754701124,
      "grad_norm": 0.2176659256219864,
      "learning_rate": 0.00010250937409864436,
      "loss": 0.2594,
      "step": 1790
    },
    {
      "epoch": 1.5128689988444164,
      "grad_norm": 0.15863171219825745,
      "learning_rate": 0.0001019325064897606,
      "loss": 0.2751,
      "step": 1800
    },
    {
      "epoch": 1.521273242987709,
      "grad_norm": 0.19906319677829742,
      "learning_rate": 0.00010135563888087685,
      "loss": 0.2865,
      "step": 1810
    },
    {
      "epoch": 1.5296774871310013,
      "grad_norm": 0.21247649192810059,
      "learning_rate": 0.00010077877127199309,
      "loss": 0.2892,
      "step": 1820
    },
    {
      "epoch": 1.5380817312742936,
      "grad_norm": 0.21099700033664703,
      "learning_rate": 0.00010020190366310933,
      "loss": 0.3008,
      "step": 1830
    },
    {
      "epoch": 1.546485975417586,
      "grad_norm": 0.15469135344028473,
      "learning_rate": 9.962503605422556e-05,
      "loss": 0.2672,
      "step": 1840
    },
    {
      "epoch": 1.5548902195608783,
      "grad_norm": 0.16477440297603607,
      "learning_rate": 9.90481684453418e-05,
      "loss": 0.2799,
      "step": 1850
    },
    {
      "epoch": 1.5632944637041706,
      "grad_norm": 0.17361459136009216,
      "learning_rate": 9.847130083645804e-05,
      "loss": 0.2756,
      "step": 1860
    },
    {
      "epoch": 1.571698707847463,
      "grad_norm": 0.15138483047485352,
      "learning_rate": 9.789443322757428e-05,
      "loss": 0.2785,
      "step": 1870
    },
    {
      "epoch": 1.5801029519907552,
      "grad_norm": 0.16653598845005035,
      "learning_rate": 9.731756561869052e-05,
      "loss": 0.2814,
      "step": 1880
    },
    {
      "epoch": 1.5885071961340476,
      "grad_norm": 0.16785801947116852,
      "learning_rate": 9.674069800980675e-05,
      "loss": 0.2752,
      "step": 1890
    },
    {
      "epoch": 1.59691144027734,
      "grad_norm": 0.21643054485321045,
      "learning_rate": 9.6163830400923e-05,
      "loss": 0.2623,
      "step": 1900
    },
    {
      "epoch": 1.6053156844206324,
      "grad_norm": 0.15368995070457458,
      "learning_rate": 9.558696279203924e-05,
      "loss": 0.2722,
      "step": 1910
    },
    {
      "epoch": 1.6137199285639248,
      "grad_norm": 0.21962004899978638,
      "learning_rate": 9.501009518315547e-05,
      "loss": 0.2563,
      "step": 1920
    },
    {
      "epoch": 1.622124172707217,
      "grad_norm": 0.14919191598892212,
      "learning_rate": 9.44332275742717e-05,
      "loss": 0.2502,
      "step": 1930
    },
    {
      "epoch": 1.6305284168505096,
      "grad_norm": 0.2036961317062378,
      "learning_rate": 9.385635996538795e-05,
      "loss": 0.2539,
      "step": 1940
    },
    {
      "epoch": 1.638932660993802,
      "grad_norm": 0.19002236425876617,
      "learning_rate": 9.327949235650419e-05,
      "loss": 0.2464,
      "step": 1950
    },
    {
      "epoch": 1.6473369051370943,
      "grad_norm": 0.16677500307559967,
      "learning_rate": 9.270262474762043e-05,
      "loss": 0.2684,
      "step": 1960
    },
    {
      "epoch": 1.6557411492803866,
      "grad_norm": 0.15206314623355865,
      "learning_rate": 9.212575713873667e-05,
      "loss": 0.242,
      "step": 1970
    },
    {
      "epoch": 1.664145393423679,
      "grad_norm": 0.17641034722328186,
      "learning_rate": 9.15488895298529e-05,
      "loss": 0.2604,
      "step": 1980
    },
    {
      "epoch": 1.6725496375669713,
      "grad_norm": 0.17574937641620636,
      "learning_rate": 9.097202192096915e-05,
      "loss": 0.2547,
      "step": 1990
    },
    {
      "epoch": 1.6809538817102636,
      "grad_norm": 0.16344806551933289,
      "learning_rate": 9.039515431208539e-05,
      "loss": 0.2681,
      "step": 2000
    },
    {
      "epoch": 1.689358125853556,
      "grad_norm": 0.18498322367668152,
      "learning_rate": 8.981828670320161e-05,
      "loss": 0.2713,
      "step": 2010
    },
    {
      "epoch": 1.6977623699968483,
      "grad_norm": 0.14767137169837952,
      "learning_rate": 8.924141909431785e-05,
      "loss": 0.2604,
      "step": 2020
    },
    {
      "epoch": 1.7061666141401408,
      "grad_norm": 0.1902410387992859,
      "learning_rate": 8.86645514854341e-05,
      "loss": 0.2516,
      "step": 2030
    },
    {
      "epoch": 1.7145708582834331,
      "grad_norm": 0.1728687733411789,
      "learning_rate": 8.808768387655033e-05,
      "loss": 0.2711,
      "step": 2040
    },
    {
      "epoch": 1.7229751024267255,
      "grad_norm": 0.1836615651845932,
      "learning_rate": 8.751081626766657e-05,
      "loss": 0.2717,
      "step": 2050
    },
    {
      "epoch": 1.731379346570018,
      "grad_norm": 0.1553170531988144,
      "learning_rate": 8.693394865878281e-05,
      "loss": 0.2303,
      "step": 2060
    },
    {
      "epoch": 1.7397835907133103,
      "grad_norm": 0.1942613571882248,
      "learning_rate": 8.635708104989905e-05,
      "loss": 0.2581,
      "step": 2070
    },
    {
      "epoch": 1.7481878348566027,
      "grad_norm": 0.1734922230243683,
      "learning_rate": 8.578021344101529e-05,
      "loss": 0.259,
      "step": 2080
    },
    {
      "epoch": 1.756592078999895,
      "grad_norm": 0.1309240758419037,
      "learning_rate": 8.520334583213153e-05,
      "loss": 0.2381,
      "step": 2090
    },
    {
      "epoch": 1.7649963231431873,
      "grad_norm": 0.17716042697429657,
      "learning_rate": 8.462647822324777e-05,
      "loss": 0.2413,
      "step": 2100
    },
    {
      "epoch": 1.7734005672864797,
      "grad_norm": 0.16437722742557526,
      "learning_rate": 8.404961061436401e-05,
      "loss": 0.2699,
      "step": 2110
    },
    {
      "epoch": 1.781804811429772,
      "grad_norm": 0.15865294635295868,
      "learning_rate": 8.347274300548025e-05,
      "loss": 0.2515,
      "step": 2120
    },
    {
      "epoch": 1.7902090555730643,
      "grad_norm": 0.16365793347358704,
      "learning_rate": 8.289587539659649e-05,
      "loss": 0.2507,
      "step": 2130
    },
    {
      "epoch": 1.7986132997163566,
      "grad_norm": 0.19089579582214355,
      "learning_rate": 8.231900778771272e-05,
      "loss": 0.2572,
      "step": 2140
    },
    {
      "epoch": 1.807017543859649,
      "grad_norm": 0.1750141978263855,
      "learning_rate": 8.174214017882896e-05,
      "loss": 0.2692,
      "step": 2150
    },
    {
      "epoch": 1.8154217880029415,
      "grad_norm": 0.14101552963256836,
      "learning_rate": 8.116527256994521e-05,
      "loss": 0.2658,
      "step": 2160
    },
    {
      "epoch": 1.8238260321462338,
      "grad_norm": 0.14396284520626068,
      "learning_rate": 8.058840496106144e-05,
      "loss": 0.2556,
      "step": 2170
    },
    {
      "epoch": 1.8322302762895262,
      "grad_norm": 0.15593650937080383,
      "learning_rate": 8.001153735217768e-05,
      "loss": 0.2442,
      "step": 2180
    },
    {
      "epoch": 1.8406345204328187,
      "grad_norm": 0.18202078342437744,
      "learning_rate": 7.943466974329392e-05,
      "loss": 0.2509,
      "step": 2190
    },
    {
      "epoch": 1.849038764576111,
      "grad_norm": 0.17855936288833618,
      "learning_rate": 7.885780213441016e-05,
      "loss": 0.2595,
      "step": 2200
    },
    {
      "epoch": 1.8574430087194034,
      "grad_norm": 0.16823212802410126,
      "learning_rate": 7.82809345255264e-05,
      "loss": 0.2469,
      "step": 2210
    },
    {
      "epoch": 1.8658472528626957,
      "grad_norm": 0.15248893201351166,
      "learning_rate": 7.770406691664264e-05,
      "loss": 0.2603,
      "step": 2220
    },
    {
      "epoch": 1.874251497005988,
      "grad_norm": 0.16229604184627533,
      "learning_rate": 7.712719930775886e-05,
      "loss": 0.2434,
      "step": 2230
    },
    {
      "epoch": 1.8826557411492804,
      "grad_norm": 0.18594375252723694,
      "learning_rate": 7.655033169887512e-05,
      "loss": 0.266,
      "step": 2240
    },
    {
      "epoch": 1.8910599852925727,
      "grad_norm": 0.18467053771018982,
      "learning_rate": 7.597346408999136e-05,
      "loss": 0.2535,
      "step": 2250
    },
    {
      "epoch": 1.899464229435865,
      "grad_norm": 0.18451227247714996,
      "learning_rate": 7.539659648110758e-05,
      "loss": 0.2579,
      "step": 2260
    },
    {
      "epoch": 1.9078684735791573,
      "grad_norm": 0.15458305180072784,
      "learning_rate": 7.481972887222382e-05,
      "loss": 0.2506,
      "step": 2270
    },
    {
      "epoch": 1.91627271772245,
      "grad_norm": 0.17949137091636658,
      "learning_rate": 7.424286126334006e-05,
      "loss": 0.2659,
      "step": 2280
    },
    {
      "epoch": 1.9246769618657422,
      "grad_norm": 0.1898379623889923,
      "learning_rate": 7.366599365445632e-05,
      "loss": 0.2882,
      "step": 2290
    },
    {
      "epoch": 1.9330812060090345,
      "grad_norm": 0.14720788598060608,
      "learning_rate": 7.308912604557254e-05,
      "loss": 0.2367,
      "step": 2300
    },
    {
      "epoch": 1.9414854501523269,
      "grad_norm": 0.15253467857837677,
      "learning_rate": 7.251225843668878e-05,
      "loss": 0.256,
      "step": 2310
    },
    {
      "epoch": 1.9498896942956194,
      "grad_norm": 0.1564057618379593,
      "learning_rate": 7.193539082780502e-05,
      "loss": 0.2536,
      "step": 2320
    },
    {
      "epoch": 1.9582939384389118,
      "grad_norm": 0.15893864631652832,
      "learning_rate": 7.135852321892126e-05,
      "loss": 0.2347,
      "step": 2330
    },
    {
      "epoch": 1.966698182582204,
      "grad_norm": 0.20592626929283142,
      "learning_rate": 7.07816556100375e-05,
      "loss": 0.2419,
      "step": 2340
    },
    {
      "epoch": 1.9751024267254964,
      "grad_norm": 0.20137999951839447,
      "learning_rate": 7.020478800115374e-05,
      "loss": 0.2415,
      "step": 2350
    },
    {
      "epoch": 1.9835066708687887,
      "grad_norm": 0.19287312030792236,
      "learning_rate": 6.962792039226997e-05,
      "loss": 0.2484,
      "step": 2360
    },
    {
      "epoch": 1.991910915012081,
      "grad_norm": 0.1620776355266571,
      "learning_rate": 6.905105278338622e-05,
      "loss": 0.2599,
      "step": 2370
    }
  ],
  "logging_steps": 10,
  "max_steps": 3567,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.613824550319268e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}