{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.118375,
  "eval_steps": 500,
  "global_step": 2900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00125,
      "grad_norm": 1.0962235927581787,
      "learning_rate": 1.3499999999999998e-05,
      "loss": 4.542208099365235,
      "step": 10
    },
    {
      "epoch": 0.0025,
      "grad_norm": 0.6401466727256775,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 4.2099754333496096,
      "step": 20
    },
    {
      "epoch": 0.00375,
      "grad_norm": 0.5453175902366638,
      "learning_rate": 4.3499999999999993e-05,
      "loss": 4.047785949707031,
      "step": 30
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.48525792360305786,
      "learning_rate": 5.85e-05,
      "loss": 3.964551544189453,
      "step": 40
    },
    {
      "epoch": 0.00625,
      "grad_norm": 0.4704461395740509,
      "learning_rate": 7.35e-05,
      "loss": 3.8913909912109377,
      "step": 50
    },
    {
      "epoch": 0.0075,
      "grad_norm": 0.46138691902160645,
      "learning_rate": 8.849999999999998e-05,
      "loss": 3.830010986328125,
      "step": 60
    },
    {
      "epoch": 0.00875,
      "grad_norm": 0.4441743493080139,
      "learning_rate": 0.00010349999999999998,
      "loss": 3.7685489654541016,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.45079436898231506,
      "learning_rate": 0.0001185,
      "loss": 3.7476680755615233,
      "step": 80
    },
    {
      "epoch": 0.01125,
      "grad_norm": 0.4830528497695923,
      "learning_rate": 0.0001335,
      "loss": 3.6839160919189453,
      "step": 90
    },
    {
      "epoch": 0.0125,
      "grad_norm": 0.44532763957977295,
      "learning_rate": 0.00014849999999999998,
      "loss": 3.6974281311035155,
      "step": 100
    },
    {
      "epoch": 0.01375,
      "grad_norm": 0.4425051808357239,
      "learning_rate": 0.0001635,
      "loss": 3.6414600372314454,
      "step": 110
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.44826197624206543,
      "learning_rate": 0.00017849999999999997,
      "loss": 3.678114318847656,
      "step": 120
    },
    {
      "epoch": 0.01625,
      "grad_norm": 0.4449610412120819,
      "learning_rate": 0.0001935,
      "loss": 3.620401382446289,
      "step": 130
    },
    {
      "epoch": 0.0175,
      "grad_norm": 0.4469984173774719,
      "learning_rate": 0.00020849999999999997,
      "loss": 3.5873218536376954,
      "step": 140
    },
    {
      "epoch": 0.01875,
      "grad_norm": 0.43191787600517273,
      "learning_rate": 0.00022349999999999998,
      "loss": 3.6034412384033203,
      "step": 150
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.42065128684043884,
      "learning_rate": 0.0002385,
      "loss": 3.5601234436035156,
      "step": 160
    },
    {
      "epoch": 0.02125,
      "grad_norm": 0.4359514117240906,
      "learning_rate": 0.0002535,
      "loss": 3.569548797607422,
      "step": 170
    },
    {
      "epoch": 0.0225,
      "grad_norm": 0.4924817681312561,
      "learning_rate": 0.00026849999999999997,
      "loss": 3.5676345825195312,
      "step": 180
    },
    {
      "epoch": 0.02375,
      "grad_norm": 0.45101380348205566,
      "learning_rate": 0.00028349999999999995,
      "loss": 3.548577880859375,
      "step": 190
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.43757978081703186,
      "learning_rate": 0.0002985,
      "loss": 3.542619323730469,
      "step": 200
    },
    {
      "epoch": 0.02625,
      "grad_norm": 0.42400574684143066,
      "learning_rate": 0.00029999901450063963,
      "loss": 3.519863510131836,
      "step": 210
    },
    {
      "epoch": 0.0275,
      "grad_norm": 0.4135052263736725,
      "learning_rate": 0.00029999560785280927,
      "loss": 3.489061737060547,
      "step": 220
    },
    {
      "epoch": 0.02875,
      "grad_norm": 0.42550331354141235,
      "learning_rate": 0.0002999897679451004,
      "loss": 3.5228813171386717,
      "step": 230
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4192274808883667,
      "learning_rate": 0.0002999814948722491,
      "loss": 3.500740814208984,
      "step": 240
    },
    {
      "epoch": 0.03125,
      "grad_norm": 0.4208814799785614,
      "learning_rate": 0.00029997078876846286,
      "loss": 3.4928520202636717,
      "step": 250
    },
    {
      "epoch": 0.0325,
      "grad_norm": 0.4207195043563843,
      "learning_rate": 0.00029995764980741843,
      "loss": 3.455915069580078,
      "step": 260
    },
    {
      "epoch": 0.03375,
      "grad_norm": 0.4257533550262451,
      "learning_rate": 0.00029994207820225867,
      "loss": 3.4429149627685547,
      "step": 270
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.4169400930404663,
      "learning_rate": 0.0002999240742055895,
      "loss": 3.423783874511719,
      "step": 280
    },
    {
      "epoch": 0.03625,
      "grad_norm": 0.4206733703613281,
      "learning_rate": 0.0002999036381094753,
      "loss": 3.463421630859375,
      "step": 290
    },
    {
      "epoch": 0.0375,
      "grad_norm": 0.4171896278858185,
      "learning_rate": 0.0002998807702454349,
      "loss": 3.4295799255371096,
      "step": 300
    },
    {
      "epoch": 0.03875,
      "grad_norm": 0.4365955889225006,
      "learning_rate": 0.00029985547098443534,
      "loss": 3.395406723022461,
      "step": 310
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.405113160610199,
      "learning_rate": 0.00029982774073688656,
      "loss": 3.4092079162597657,
      "step": 320
    },
    {
      "epoch": 0.04125,
      "grad_norm": 0.4027758836746216,
      "learning_rate": 0.0002997975799526344,
      "loss": 3.3900100708007814,
      "step": 330
    },
    {
      "epoch": 0.0425,
      "grad_norm": 0.36929693818092346,
      "learning_rate": 0.0002997649891209534,
      "loss": 3.3978004455566406,
      "step": 340
    },
    {
      "epoch": 0.04375,
      "grad_norm": 0.36560943722724915,
      "learning_rate": 0.00029972996877053866,
      "loss": 3.389325714111328,
      "step": 350
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.381977915763855,
      "learning_rate": 0.0002996925194694977,
      "loss": 3.3937278747558595,
      "step": 360
    },
    {
      "epoch": 0.04625,
      "grad_norm": 0.4034512937068939,
      "learning_rate": 0.0002996526418253408,
      "loss": 3.3755035400390625,
      "step": 370
    },
    {
      "epoch": 0.0475,
      "grad_norm": 0.40442460775375366,
      "learning_rate": 0.00029961033648497136,
      "loss": 3.3367355346679686,
      "step": 380
    },
    {
      "epoch": 0.04875,
      "grad_norm": 0.4017426073551178,
      "learning_rate": 0.00029956560413467545,
      "loss": 3.358184051513672,
      "step": 390
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.39038756489753723,
      "learning_rate": 0.00029951844550011046,
      "loss": 3.343119430541992,
      "step": 400
    },
    {
      "epoch": 0.05125,
      "grad_norm": 0.3888174295425415,
      "learning_rate": 0.00029946886134629366,
      "loss": 3.3283355712890623,
      "step": 410
    },
    {
      "epoch": 0.0525,
      "grad_norm": 0.37037158012390137,
      "learning_rate": 0.0002994168524775894,
      "loss": 3.3299789428710938,
      "step": 420
    },
    {
      "epoch": 0.05375,
      "grad_norm": 0.3683609366416931,
      "learning_rate": 0.0002993624197376964,
      "loss": 3.3183124542236326,
      "step": 430
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.38251495361328125,
      "learning_rate": 0.00029930556400963374,
      "loss": 3.3277027130126955,
      "step": 440
    },
    {
      "epoch": 0.05625,
      "grad_norm": 0.3685538172721863,
      "learning_rate": 0.00029924628621572693,
      "loss": 3.3209506988525392,
      "step": 450
    },
    {
      "epoch": 0.0575,
      "grad_norm": 0.40296629071235657,
      "learning_rate": 0.0002991845873175927,
      "loss": 3.328449249267578,
      "step": 460
    },
    {
      "epoch": 0.05875,
      "grad_norm": 0.38991594314575195,
      "learning_rate": 0.0002991204683161233,
      "loss": 3.3068389892578125,
      "step": 470
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.39050525426864624,
      "learning_rate": 0.00029905393025147044,
      "loss": 3.3071823120117188,
      "step": 480
    },
    {
      "epoch": 0.06125,
      "grad_norm": 0.3846362829208374,
      "learning_rate": 0.00029898497420302855,
      "loss": 3.280712127685547,
      "step": 490
    },
    {
      "epoch": 0.0625,
      "grad_norm": 0.3796476125717163,
      "learning_rate": 0.0002989136012894168,
      "loss": 3.2889366149902344,
      "step": 500
    },
    {
      "epoch": 0.06375,
      "grad_norm": 0.36522796750068665,
      "learning_rate": 0.0002988398126684615,
      "loss": 3.291548156738281,
      "step": 510
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.3771977722644806,
      "learning_rate": 0.000298763609537177,
      "loss": 3.2864120483398436,
      "step": 520
    },
    {
      "epoch": 0.06625,
      "grad_norm": 0.35490748286247253,
      "learning_rate": 0.00029868499313174624,
      "loss": 3.285882568359375,
      "step": 530
    },
    {
      "epoch": 0.0675,
      "grad_norm": 0.37846773862838745,
      "learning_rate": 0.00029860396472750083,
      "loss": 3.279650115966797,
      "step": 540
    },
    {
      "epoch": 0.06875,
      "grad_norm": 0.38366758823394775,
      "learning_rate": 0.0002985205256389005,
      "loss": 3.2765472412109373,
      "step": 550
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.3582383990287781,
      "learning_rate": 0.0002984346772195113,
      "loss": 3.283465576171875,
      "step": 560
    },
    {
      "epoch": 0.07125,
      "grad_norm": 0.35587790608406067,
      "learning_rate": 0.00029834642086198427,
      "loss": 3.271703338623047,
      "step": 570
    },
    {
      "epoch": 0.0725,
      "grad_norm": 0.37970319390296936,
      "learning_rate": 0.0002982557579980322,
      "loss": 3.245616912841797,
      "step": 580
    },
    {
      "epoch": 0.07375,
      "grad_norm": 0.3639717400074005,
      "learning_rate": 0.000298162690098407,
      "loss": 3.258502960205078,
      "step": 590
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.36604151129722595,
      "learning_rate": 0.0002980672186728754,
      "loss": 3.2504783630371095,
      "step": 600
    },
    {
      "epoch": 0.07625,
      "grad_norm": 0.3567405343055725,
      "learning_rate": 0.0002979693452701947,
      "loss": 3.2632408142089844,
      "step": 610
    },
    {
      "epoch": 0.0775,
      "grad_norm": 0.3561687171459198,
      "learning_rate": 0.0002978690714780875,
      "loss": 3.2412956237792967,
      "step": 620
    },
    {
      "epoch": 0.07875,
      "grad_norm": 0.3695826232433319,
      "learning_rate": 0.00029776639892321606,
      "loss": 3.2321800231933593,
      "step": 630
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3597981035709381,
      "learning_rate": 0.000297661329271156,
      "loss": 3.205059814453125,
      "step": 640
    },
    {
      "epoch": 0.08125,
      "grad_norm": 0.3703027367591858,
      "learning_rate": 0.00029755386422636884,
      "loss": 3.2197303771972656,
      "step": 650
    },
    {
      "epoch": 0.0825,
      "grad_norm": 0.34924793243408203,
      "learning_rate": 0.000297444005532175,
      "loss": 3.1991270065307615,
      "step": 660
    },
    {
      "epoch": 0.08375,
      "grad_norm": 0.3698681592941284,
      "learning_rate": 0.0002973317549707249,
      "loss": 3.2529460906982424,
      "step": 670
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.3563082814216614,
      "learning_rate": 0.0002972171143629705,
      "loss": 3.2158493041992187,
      "step": 680
    },
    {
      "epoch": 0.08625,
      "grad_norm": 0.34349241852760315,
      "learning_rate": 0.0002971000855686355,
      "loss": 3.1953773498535156,
      "step": 690
    },
    {
      "epoch": 0.0875,
      "grad_norm": 0.35246145725250244,
      "learning_rate": 0.00029698067048618536,
      "loss": 3.1936809539794924,
      "step": 700
    },
    {
      "epoch": 0.08875,
      "grad_norm": 0.372384637594223,
      "learning_rate": 0.00029685887105279624,
      "loss": 3.1743907928466797,
      "step": 710
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.36444413661956787,
      "learning_rate": 0.0002967346892443239,
      "loss": 3.1915206909179688,
      "step": 720
    },
    {
      "epoch": 0.09125,
      "grad_norm": 0.351936012506485,
      "learning_rate": 0.00029660812707527133,
      "loss": 3.2071929931640626,
      "step": 730
    },
    {
      "epoch": 0.0925,
      "grad_norm": 0.34430351853370667,
      "learning_rate": 0.00029647918659875635,
      "loss": 3.181209182739258,
      "step": 740
    },
    {
      "epoch": 0.09375,
      "grad_norm": 0.35013672709465027,
      "learning_rate": 0.0002963478699064781,
      "loss": 3.1906646728515624,
      "step": 750
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.34257224202156067,
      "learning_rate": 0.00029621417912868323,
      "loss": 3.228814697265625,
      "step": 760
    },
    {
      "epoch": 0.09625,
      "grad_norm": 0.3684031367301941,
      "learning_rate": 0.00029607811643413135,
      "loss": 3.1923717498779296,
      "step": 770
    },
    {
      "epoch": 0.0975,
      "grad_norm": 0.3443935513496399,
      "learning_rate": 0.0002959396840300596,
      "loss": 3.19796085357666,
      "step": 780
    },
    {
      "epoch": 0.09875,
      "grad_norm": 0.35859978199005127,
      "learning_rate": 0.0002957988841621472,
      "loss": 3.2059059143066406,
      "step": 790
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.35764068365097046,
      "learning_rate": 0.00029565571911447893,
      "loss": 3.197886276245117,
      "step": 800
    },
    {
      "epoch": 0.10125,
      "grad_norm": 0.34071439504623413,
      "learning_rate": 0.0002955101912095078,
      "loss": 3.1616336822509767,
      "step": 810
    },
    {
      "epoch": 0.1025,
      "grad_norm": 0.3420080244541168,
      "learning_rate": 0.00029536230280801767,
      "loss": 3.1601337432861327,
      "step": 820
    },
    {
      "epoch": 0.10375,
      "grad_norm": 0.33847373723983765,
      "learning_rate": 0.000295212056309085,
      "loss": 3.1570825576782227,
      "step": 830
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.33834919333457947,
      "learning_rate": 0.00029505945415003954,
      "loss": 3.1625667572021485,
      "step": 840
    },
    {
      "epoch": 0.10625,
      "grad_norm": 0.33340954780578613,
      "learning_rate": 0.0002949044988064253,
      "loss": 3.180934524536133,
      "step": 850
    },
    {
      "epoch": 0.1075,
      "grad_norm": 0.36460039019584656,
      "learning_rate": 0.0002947471927919599,
      "loss": 3.157451629638672,
      "step": 860
    },
    {
      "epoch": 0.10875,
      "grad_norm": 0.3350942134857178,
      "learning_rate": 0.00029458753865849424,
      "loss": 3.16195068359375,
      "step": 870
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.3496924936771393,
      "learning_rate": 0.00029442553899597075,
      "loss": 3.151328468322754,
      "step": 880
    },
    {
      "epoch": 0.11125,
      "grad_norm": 0.33595219254493713,
      "learning_rate": 0.0002942611964323817,
      "loss": 3.1474483489990233,
      "step": 890
    },
    {
      "epoch": 0.1125,
      "grad_norm": 0.34623855352401733,
      "learning_rate": 0.00029409451363372605,
      "loss": 3.151190757751465,
      "step": 900
    },
    {
      "epoch": 0.11375,
      "grad_norm": 0.3318001925945282,
      "learning_rate": 0.00029392549330396696,
      "loss": 3.166637420654297,
      "step": 910
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.3412299156188965,
      "learning_rate": 0.0002937541381849872,
      "loss": 3.1225704193115233,
      "step": 920
    },
    {
      "epoch": 0.11625,
      "grad_norm": 0.3505077064037323,
      "learning_rate": 0.000293580451056545,
      "loss": 3.1353290557861326,
      "step": 930
    },
    {
      "epoch": 0.1175,
      "grad_norm": 0.36463502049446106,
      "learning_rate": 0.00029340443473622915,
      "loss": 3.136840057373047,
      "step": 940
    },
    {
      "epoch": 0.11875,
      "grad_norm": 0.3351852297782898,
      "learning_rate": 0.00029322609207941283,
      "loss": 3.13529109954834,
      "step": 950
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.34219565987586975,
      "learning_rate": 0.00029304542597920766,
      "loss": 3.158438873291016,
      "step": 960
    },
    {
      "epoch": 0.12125,
      "grad_norm": 0.32809194922447205,
      "learning_rate": 0.0002928624393664166,
      "loss": 3.1299543380737305,
      "step": 970
    },
    {
      "epoch": 0.1225,
      "grad_norm": 0.3327496647834778,
      "learning_rate": 0.00029267713520948643,
      "loss": 3.127825164794922,
      "step": 980
    },
    {
      "epoch": 0.12375,
      "grad_norm": 0.34798717498779297,
      "learning_rate": 0.00029248951651445973,
      "loss": 3.1128513336181642,
      "step": 990
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.3409072756767273,
      "learning_rate": 0.0002922995863249258,
      "loss": 3.0872947692871096,
      "step": 1000
    },
    {
      "epoch": 0.12625,
      "grad_norm": 0.3241552412509918,
      "learning_rate": 0.00029210734772197166,
      "loss": 3.135254669189453,
      "step": 1010
    },
    {
      "epoch": 0.1275,
      "grad_norm": 0.33775585889816284,
      "learning_rate": 0.0002919128038241318,
      "loss": 3.1159934997558594,
      "step": 1020
    },
    {
      "epoch": 0.12875,
      "grad_norm": 0.3477814197540283,
      "learning_rate": 0.0002917159577873377,
      "loss": 3.1170124053955077,
      "step": 1030
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.34592533111572266,
      "learning_rate": 0.00029151681280486656,
      "loss": 3.108189010620117,
      "step": 1040
    },
    {
      "epoch": 0.13125,
      "grad_norm": 0.3351367712020874,
      "learning_rate": 0.00029131537210728975,
      "loss": 3.1113187789916994,
      "step": 1050
    },
    {
      "epoch": 0.1325,
      "grad_norm": 0.354574054479599,
      "learning_rate": 0.00029111163896241996,
      "loss": 3.114427947998047,
      "step": 1060
    },
    {
      "epoch": 0.13375,
      "grad_norm": 0.3495411276817322,
      "learning_rate": 0.0002909056166752586,
      "loss": 3.099652862548828,
      "step": 1070
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.32834145426750183,
      "learning_rate": 0.0002906973085879419,
      "loss": 3.1386039733886717,
      "step": 1080
    },
    {
      "epoch": 0.13625,
      "grad_norm": 0.3264714181423187,
      "learning_rate": 0.000290486718079687,
      "loss": 3.0912006378173826,
      "step": 1090
    },
    {
      "epoch": 0.1375,
      "grad_norm": 0.3366532623767853,
      "learning_rate": 0.0002902738485667367,
      "loss": 3.066594123840332,
      "step": 1100
    },
    {
      "epoch": 0.13875,
      "grad_norm": 0.3517299294471741,
      "learning_rate": 0.00029005870350230453,
      "loss": 3.1176836013793947,
      "step": 1110
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3267839550971985,
      "learning_rate": 0.00028984128637651825,
      "loss": 3.0968740463256834,
      "step": 1120
    },
    {
      "epoch": 0.14125,
      "grad_norm": 0.3284844160079956,
      "learning_rate": 0.0002896216007163637,
      "loss": 3.0969438552856445,
      "step": 1130
    },
    {
      "epoch": 0.1425,
      "grad_norm": 0.32528674602508545,
      "learning_rate": 0.0002893996500856272,
      "loss": 3.1029247283935546,
      "step": 1140
    },
    {
      "epoch": 0.14375,
      "grad_norm": 0.3221900463104248,
      "learning_rate": 0.00028917543808483796,
      "loss": 3.1135093688964846,
      "step": 1150
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.33406203985214233,
      "learning_rate": 0.0002889489683512096,
      "loss": 3.0768909454345703,
      "step": 1160
    },
    {
      "epoch": 0.14625,
      "grad_norm": 0.3397602140903473,
      "learning_rate": 0.0002887202445585811,
      "loss": 3.0875574111938477,
      "step": 1170
    },
    {
      "epoch": 0.1475,
      "grad_norm": 0.33265799283981323,
      "learning_rate": 0.0002884892704173573,
      "loss": 3.0731714248657225,
      "step": 1180
    },
    {
      "epoch": 0.14875,
      "grad_norm": 0.32758811116218567,
      "learning_rate": 0.00028825604967444866,
      "loss": 3.12725772857666,
      "step": 1190
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3414333164691925,
      "learning_rate": 0.0002880205861132105,
      "loss": 3.0664628982543944,
      "step": 1200
    },
    {
      "epoch": 0.15125,
      "grad_norm": 0.33994749188423157,
      "learning_rate": 0.00028778288355338144,
      "loss": 3.0672843933105467,
      "step": 1210
    },
    {
      "epoch": 0.1525,
      "grad_norm": 0.33031970262527466,
      "learning_rate": 0.0002875429458510219,
      "loss": 3.1195865631103517,
      "step": 1220
    },
    {
      "epoch": 0.15375,
      "grad_norm": 0.3435764014720917,
      "learning_rate": 0.0002873007768984511,
      "loss": 3.0530136108398436,
      "step": 1230
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.32233837246894836,
      "learning_rate": 0.00028705638062418386,
      "loss": 3.076219177246094,
      "step": 1240
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.32135722041130066,
      "learning_rate": 0.0002868097609928674,
      "loss": 3.051331329345703,
      "step": 1250
    },
    {
      "epoch": 0.1575,
      "grad_norm": 0.3371255099773407,
      "learning_rate": 0.0002865609220052165,
      "loss": 3.060790252685547,
      "step": 1260
    },
    {
      "epoch": 0.15875,
      "grad_norm": 0.344938188791275,
      "learning_rate": 0.000286309867697949,
      "loss": 3.0386356353759765,
      "step": 1270
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3418465256690979,
      "learning_rate": 0.0002860566021437197,
      "loss": 3.0650426864624025,
      "step": 1280
    },
    {
      "epoch": 0.16125,
      "grad_norm": 0.3343369960784912,
      "learning_rate": 0.0002858011294510552,
      "loss": 3.04056396484375,
      "step": 1290
    },
    {
      "epoch": 0.1625,
      "grad_norm": 0.34760987758636475,
      "learning_rate": 0.0002855434537642865,
      "loss": 3.057343292236328,
      "step": 1300
    },
    {
      "epoch": 0.16375,
      "grad_norm": 0.3561476767063141,
      "learning_rate": 0.000285283579263482,
      "loss": 3.0452644348144533,
      "step": 1310
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.3393990993499756,
      "learning_rate": 0.00028502151016437986,
      "loss": 3.0280082702636717,
      "step": 1320
    },
    {
      "epoch": 0.16625,
      "grad_norm": 0.34054502844810486,
      "learning_rate": 0.0002847572507183193,
      "loss": 3.05167293548584,
      "step": 1330
    },
    {
      "epoch": 0.1675,
      "grad_norm": 0.33443859219551086,
      "learning_rate": 0.000284490805212172,
      "loss": 3.090429496765137,
      "step": 1340
    },
    {
      "epoch": 0.16875,
      "grad_norm": 0.33417847752571106,
      "learning_rate": 0.00028422217796827216,
      "loss": 3.064510726928711,
      "step": 1350
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.33783411979675293,
      "learning_rate": 0.00028395137334434676,
      "loss": 3.03808479309082,
      "step": 1360
    },
    {
      "epoch": 0.17125,
      "grad_norm": 0.3298597037792206,
      "learning_rate": 0.00028367839573344454,
      "loss": 3.0556320190429687,
      "step": 1370
    },
    {
      "epoch": 0.1725,
      "grad_norm": 0.3494487702846527,
      "learning_rate": 0.000283403249563865,
      "loss": 3.039119338989258,
      "step": 1380
    },
    {
      "epoch": 0.17375,
      "grad_norm": 0.3314499258995056,
      "learning_rate": 0.0002831259392990864,
      "loss": 3.0571495056152345,
      "step": 1390
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.32631996273994446,
      "learning_rate": 0.00028284646943769337,
      "loss": 3.0531475067138674,
      "step": 1400
    },
    {
      "epoch": 0.17625,
      "grad_norm": 0.32792332768440247,
      "learning_rate": 0.00028256484451330403,
      "loss": 3.026413917541504,
      "step": 1410
    },
    {
      "epoch": 0.1775,
      "grad_norm": 0.31503939628601074,
      "learning_rate": 0.0002822810690944963,
      "loss": 3.052714538574219,
      "step": 1420
    },
    {
      "epoch": 0.17875,
      "grad_norm": 0.33248430490493774,
      "learning_rate": 0.000281995147784734,
      "loss": 3.050276184082031,
      "step": 1430
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3251739740371704,
      "learning_rate": 0.0002817070852222918,
      "loss": 3.057173156738281,
      "step": 1440
    },
    {
      "epoch": 0.18125,
      "grad_norm": 0.32668808102607727,
      "learning_rate": 0.0002814168860801806,
      "loss": 3.0323795318603515,
      "step": 1450
    },
    {
      "epoch": 0.1825,
      "grad_norm": 0.3308984339237213,
      "learning_rate": 0.0002811245550660709,
      "loss": 3.0549407958984376,
      "step": 1460
    },
    {
      "epoch": 0.18375,
      "grad_norm": 0.33381035923957825,
      "learning_rate": 0.0002808300969222172,
      "loss": 3.038156509399414,
      "step": 1470
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.3257412016391754,
      "learning_rate": 0.0002805335164253806,
      "loss": 2.9872032165527345,
      "step": 1480
    },
    {
      "epoch": 0.18625,
      "grad_norm": 0.33056777715682983,
      "learning_rate": 0.0002802348183867514,
      "loss": 3.0133747100830077,
      "step": 1490
    },
    {
      "epoch": 0.1875,
      "grad_norm": 0.3441354036331177,
      "learning_rate": 0.00027993400765187124,
      "loss": 2.9739742279052734,
      "step": 1500
    },
    {
      "epoch": 0.18875,
      "grad_norm": 0.33665338158607483,
      "learning_rate": 0.0002796310891005542,
      "loss": 3.0187236785888674,
      "step": 1510
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3456818163394928,
      "learning_rate": 0.00027932606764680796,
      "loss": 3.000716781616211,
      "step": 1520
    },
    {
      "epoch": 0.19125,
      "grad_norm": 0.31569328904151917,
      "learning_rate": 0.00027901894823875387,
      "loss": 3.003590774536133,
      "step": 1530
    },
    {
      "epoch": 0.1925,
      "grad_norm": 0.33296287059783936,
      "learning_rate": 0.00027870973585854665,
      "loss": 3.010061264038086,
      "step": 1540
    },
    {
      "epoch": 0.19375,
      "grad_norm": 0.3256262242794037,
      "learning_rate": 0.0002783984355222937,
      "loss": 3.009804534912109,
      "step": 1550
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.33614271879196167,
      "learning_rate": 0.0002780850522799737,
      "loss": 3.0124454498291016,
      "step": 1560
    },
    {
      "epoch": 0.19625,
      "grad_norm": 0.3145756721496582,
      "learning_rate": 0.00027776959121535464,
      "loss": 3.0317821502685547,
      "step": 1570
    },
    {
      "epoch": 0.1975,
      "grad_norm": 0.33354440331459045,
      "learning_rate": 0.0002774520574459113,
      "loss": 2.9961336135864256,
      "step": 1580
    },
    {
      "epoch": 0.19875,
      "grad_norm": 0.33014026284217834,
      "learning_rate": 0.00027713245612274247,
      "loss": 2.996552658081055,
      "step": 1590
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3215552866458893,
      "learning_rate": 0.00027681079243048717,
      "loss": 3.0088550567626955,
      "step": 1600
    },
    {
      "epoch": 0.20125,
      "grad_norm": 0.33168256282806396,
      "learning_rate": 0.0002764870715872405,
      "loss": 3.017764663696289,
      "step": 1610
    },
    {
      "epoch": 0.2025,
      "grad_norm": 0.3189021646976471,
      "learning_rate": 0.00027616129884446916,
      "loss": 3.013848876953125,
      "step": 1620
    },
    {
      "epoch": 0.20375,
      "grad_norm": 0.3378245532512665,
      "learning_rate": 0.0002758334794869262,
      "loss": 2.9824432373046874,
      "step": 1630
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.3281719386577606,
      "learning_rate": 0.00027550361883256535,
      "loss": 2.998569679260254,
      "step": 1640
    },
    {
      "epoch": 0.20625,
      "grad_norm": 0.3160180151462555,
      "learning_rate": 0.00027517172223245445,
      "loss": 2.9816024780273436,
      "step": 1650
    },
    {
      "epoch": 0.2075,
      "grad_norm": 0.33755186200141907,
      "learning_rate": 0.00027483779507068913,
      "loss": 2.9858730316162108,
      "step": 1660
    },
    {
      "epoch": 0.20875,
      "grad_norm": 0.3391866683959961,
      "learning_rate": 0.0002745018427643051,
      "loss": 3.0096504211425783,
      "step": 1670
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.31814366579055786,
      "learning_rate": 0.00027416387076319035,
      "loss": 3.0215930938720703,
      "step": 1680
    },
    {
      "epoch": 0.21125,
      "grad_norm": 0.33236604928970337,
      "learning_rate": 0.00027382388454999686,
      "loss": 2.9959911346435546,
      "step": 1690
    },
    {
      "epoch": 0.2125,
      "grad_norm": 0.31699317693710327,
      "learning_rate": 0.00027348188964005147,
      "loss": 2.9976850509643556,
      "step": 1700
    },
    {
      "epoch": 0.21375,
      "grad_norm": 0.3221769630908966,
      "learning_rate": 0.00027313789158126667,
      "loss": 2.992999267578125,
      "step": 1710
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.3359137773513794,
      "learning_rate": 0.00027279189595405036,
      "loss": 3.005726623535156,
      "step": 1720
    },
    {
      "epoch": 0.21625,
      "grad_norm": 0.3230113685131073,
      "learning_rate": 0.0002724439083712153,
      "loss": 2.9472639083862306,
      "step": 1730
    },
    {
      "epoch": 0.2175,
      "grad_norm": 0.3287331461906433,
      "learning_rate": 0.00027209393447788835,
      "loss": 2.979531478881836,
      "step": 1740
    },
    {
      "epoch": 0.21875,
      "grad_norm": 0.33660778403282166,
      "learning_rate": 0.00027174197995141866,
      "loss": 2.972854232788086,
      "step": 1750
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3239246904850006,
      "learning_rate": 0.0002713880505012855,
      "loss": 2.9944473266601563,
      "step": 1760
    },
    {
      "epoch": 0.22125,
      "grad_norm": 0.33757129311561584,
      "learning_rate": 0.00027103215186900597,
      "loss": 2.984470748901367,
      "step": 1770
    },
    {
      "epoch": 0.2225,
      "grad_norm": 0.3297254741191864,
      "learning_rate": 0.0002706742898280415,
      "loss": 2.9779335021972657,
      "step": 1780
    },
    {
      "epoch": 0.22375,
      "grad_norm": 0.3173220753669739,
      "learning_rate": 0.0002703144701837044,
      "loss": 2.989591598510742,
      "step": 1790
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.3255234658718109,
      "learning_rate": 0.00026995269877306356,
      "loss": 2.9630172729492186,
      "step": 1800
    },
    {
      "epoch": 0.22625,
      "grad_norm": 0.31826311349868774,
      "learning_rate": 0.0002695889814648499,
      "loss": 2.949393463134766,
      "step": 1810
    },
    {
      "epoch": 0.2275,
      "grad_norm": 0.3174227774143219,
      "learning_rate": 0.00026922332415936116,
      "loss": 2.962572479248047,
      "step": 1820
    },
    {
      "epoch": 0.22875,
      "grad_norm": 0.3316509425640106,
      "learning_rate": 0.0002688557327883659,
      "loss": 2.940341567993164,
      "step": 1830
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3224658668041229,
      "learning_rate": 0.00026848621331500766,
      "loss": 2.96124267578125,
      "step": 1840
    },
    {
      "epoch": 0.23125,
      "grad_norm": 0.32638388872146606,
      "learning_rate": 0.00026811477173370815,
      "loss": 2.967144775390625,
      "step": 1850
    },
    {
      "epoch": 0.2325,
      "grad_norm": 0.3130776286125183,
      "learning_rate": 0.0002677414140700696,
      "loss": 2.946596145629883,
      "step": 1860
    },
    {
      "epoch": 0.23375,
      "grad_norm": 0.3233889937400818,
      "learning_rate": 0.0002673661463807776,
      "loss": 2.967226028442383,
      "step": 1870
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.32231318950653076,
      "learning_rate": 0.00026698897475350254,
      "loss": 2.9710044860839844,
      "step": 1880
    },
    {
      "epoch": 0.23625,
      "grad_norm": 0.32844141125679016,
      "learning_rate": 0.0002666099053068007,
      "loss": 2.98897762298584,
      "step": 1890
    },
    {
      "epoch": 0.2375,
      "grad_norm": 0.32306230068206787,
      "learning_rate": 0.00026622894419001537,
      "loss": 2.978479766845703,
      "step": 1900
    },
    {
      "epoch": 0.23875,
      "grad_norm": 0.3146902620792389,
      "learning_rate": 0.0002658460975831769,
      "loss": 2.9550348281860352,
      "step": 1910
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3284965753555298,
      "learning_rate": 0.00026546137169690235,
      "loss": 2.954052734375,
      "step": 1920
    },
    {
      "epoch": 0.24125,
      "grad_norm": 0.3169645071029663,
      "learning_rate": 0.00026507477277229496,
      "loss": 2.9700557708740236,
      "step": 1930
    },
    {
      "epoch": 0.2425,
      "grad_norm": 0.32831114530563354,
      "learning_rate": 0.0002646863070808425,
      "loss": 2.965903091430664,
      "step": 1940
    },
    {
      "epoch": 0.24375,
      "grad_norm": 0.3175643980503082,
      "learning_rate": 0.0002642959809243163,
      "loss": 2.9406291961669924,
      "step": 1950
    },
    {
      "epoch": 1.000875,
      "grad_norm": 0.3207930624485016,
      "learning_rate": 0.00026390380063466806,
      "loss": 3.2728614807128906,
      "step": 1960
    },
    {
      "epoch": 1.002125,
      "grad_norm": 0.33733275532722473,
      "learning_rate": 0.000263509772573928,
      "loss": 2.9439430236816406,
      "step": 1970
    },
    {
      "epoch": 1.003375,
      "grad_norm": 0.3284999132156372,
      "learning_rate": 0.00026311390313410097,
      "loss": 2.9343544006347657,
      "step": 1980
    },
    {
      "epoch": 1.004625,
      "grad_norm": 0.31731417775154114,
      "learning_rate": 0.0002627161987370632,
      "loss": 2.9394100189208983,
      "step": 1990
    },
    {
      "epoch": 1.005875,
      "grad_norm": 0.3323110044002533,
      "learning_rate": 0.000262316665834458,
      "loss": 2.932010269165039,
      "step": 2000
    },
    {
      "epoch": 1.007125,
      "grad_norm": 0.3103737533092499,
      "learning_rate": 0.000261915310907591,
      "loss": 2.9386672973632812,
      "step": 2010
    },
    {
      "epoch": 1.008375,
      "grad_norm": 0.311427503824234,
      "learning_rate": 0.0002615121404673251,
      "loss": 2.893377113342285,
      "step": 2020
    },
    {
      "epoch": 1.009625,
      "grad_norm": 0.31239134073257446,
      "learning_rate": 0.00026110716105397485,
      "loss": 2.9058563232421877,
      "step": 2030
    },
    {
      "epoch": 1.010875,
      "grad_norm": 0.31634724140167236,
      "learning_rate": 0.0002607003792372004,
      "loss": 2.9108051300048827,
      "step": 2040
    },
    {
      "epoch": 1.012125,
      "grad_norm": 0.3243110775947571,
      "learning_rate": 0.00026029180161590067,
      "loss": 2.919489288330078,
      "step": 2050
    },
    {
      "epoch": 1.013375,
      "grad_norm": 0.3139530420303345,
      "learning_rate": 0.0002598814348181068,
      "loss": 2.901732635498047,
      "step": 2060
    },
    {
      "epoch": 1.014625,
      "grad_norm": 0.33199912309646606,
      "learning_rate": 0.000259469285500874,
      "loss": 2.9169565200805665,
      "step": 2070
    },
    {
      "epoch": 1.015875,
      "grad_norm": 0.31367039680480957,
      "learning_rate": 0.0002590553603501741,
      "loss": 2.8976207733154298,
      "step": 2080
    },
    {
      "epoch": 1.017125,
      "grad_norm": 0.3230586349964142,
      "learning_rate": 0.00025863966608078673,
      "loss": 2.882274055480957,
      "step": 2090
    },
    {
      "epoch": 1.018375,
      "grad_norm": 0.313128262758255,
      "learning_rate": 0.0002582222094361907,
      "loss": 2.8793052673339843,
      "step": 2100
    },
    {
      "epoch": 1.019625,
      "grad_norm": 0.31868046522140503,
      "learning_rate": 0.00025780299718845416,
      "loss": 2.8837726593017576,
      "step": 2110
    },
    {
      "epoch": 1.020875,
      "grad_norm": 0.32915839552879333,
      "learning_rate": 0.00025738203613812543,
      "loss": 2.8813737869262694,
      "step": 2120
    },
    {
      "epoch": 1.022125,
      "grad_norm": 0.3112785220146179,
      "learning_rate": 0.0002569593331141218,
      "loss": 2.8455078125,
      "step": 2130
    },
    {
      "epoch": 1.023375,
      "grad_norm": 0.29719340801239014,
      "learning_rate": 0.0002565348949736196,
      "loss": 2.8642330169677734,
      "step": 2140
    },
    {
      "epoch": 1.024625,
      "grad_norm": 0.3277027904987335,
      "learning_rate": 0.0002561087286019424,
      "loss": 2.8876218795776367,
      "step": 2150
    },
    {
      "epoch": 1.025875,
      "grad_norm": 0.3183266520500183,
      "learning_rate": 0.0002556808409124494,
      "loss": 2.883536720275879,
      "step": 2160
    },
    {
      "epoch": 1.027125,
      "grad_norm": 0.3155848979949951,
      "learning_rate": 0.00025525123884642366,
      "loss": 2.862305450439453,
      "step": 2170
    },
    {
      "epoch": 1.028375,
      "grad_norm": 0.3132951855659485,
      "learning_rate": 0.000254819929372959,
      "loss": 2.856226348876953,
      "step": 2180
    },
    {
      "epoch": 1.029625,
      "grad_norm": 0.3311610221862793,
      "learning_rate": 0.0002543869194888471,
      "loss": 2.854077911376953,
      "step": 2190
    },
    {
      "epoch": 1.030875,
      "grad_norm": 0.31400758028030396,
      "learning_rate": 0.00025395221621846435,
      "loss": 2.8614845275878906,
      "step": 2200
    },
    {
      "epoch": 1.032125,
      "grad_norm": 0.30809491872787476,
      "learning_rate": 0.00025351582661365724,
      "loss": 2.849281883239746,
      "step": 2210
    },
    {
      "epoch": 1.033375,
      "grad_norm": 0.3170127272605896,
      "learning_rate": 0.00025307775775362855,
      "loss": 2.8551118850708006,
      "step": 2220
    },
    {
      "epoch": 1.034625,
      "grad_norm": 0.33024269342422485,
      "learning_rate": 0.0002526380167448223,
      "loss": 2.867789459228516,
      "step": 2230
    },
    {
      "epoch": 1.035875,
      "grad_norm": 0.31417202949523926,
      "learning_rate": 0.0002521966107208084,
      "loss": 2.841164779663086,
      "step": 2240
    },
    {
      "epoch": 1.037125,
      "grad_norm": 0.31379812955856323,
      "learning_rate": 0.0002517535468421669,
      "loss": 2.8253313064575196,
      "step": 2250
    },
    {
      "epoch": 1.038375,
      "grad_norm": 0.32265710830688477,
      "learning_rate": 0.00025130883229637196,
      "loss": 2.7978553771972656,
      "step": 2260
    },
    {
      "epoch": 1.039625,
      "grad_norm": 0.3156164288520813,
      "learning_rate": 0.0002508624742976753,
      "loss": 2.8533695220947264,
      "step": 2270
    },
    {
      "epoch": 1.040875,
      "grad_norm": 0.31064850091934204,
      "learning_rate": 0.000250414480086989,
      "loss": 2.8257659912109374,
      "step": 2280
    },
    {
      "epoch": 1.042125,
      "grad_norm": 0.31829559803009033,
      "learning_rate": 0.00024996485693176815,
      "loss": 2.8458267211914063,
      "step": 2290
    },
    {
      "epoch": 1.043375,
      "grad_norm": 0.3130020499229431,
      "learning_rate": 0.0002495136121258928,
      "loss": 2.823675537109375,
      "step": 2300
    },
    {
      "epoch": 1.044625,
      "grad_norm": 0.31633248925209045,
      "learning_rate": 0.0002490607529895499,
      "loss": 2.831635093688965,
      "step": 2310
    },
    {
      "epoch": 1.045875,
      "grad_norm": 0.3220139741897583,
      "learning_rate": 0.00024860628686911436,
      "loss": 2.8292957305908204,
      "step": 2320
    },
    {
      "epoch": 1.047125,
      "grad_norm": 0.3184780180454254,
      "learning_rate": 0.0002481502211370298,
      "loss": 2.813566207885742,
      "step": 2330
    },
    {
      "epoch": 1.048375,
      "grad_norm": 0.3212302625179291,
      "learning_rate": 0.00024769256319168923,
      "loss": 2.820956993103027,
      "step": 2340
    },
    {
      "epoch": 1.049625,
      "grad_norm": 0.31229448318481445,
      "learning_rate": 0.00024723332045731484,
      "loss": 2.8484216690063477,
      "step": 2350
    },
    {
      "epoch": 1.050875,
      "grad_norm": 0.3293566107749939,
      "learning_rate": 0.0002467725003838375,
      "loss": 2.822405242919922,
      "step": 2360
    },
    {
      "epoch": 1.052125,
      "grad_norm": 0.32199931144714355,
      "learning_rate": 0.00024631011044677615,
      "loss": 2.8012535095214846,
      "step": 2370
    },
    {
      "epoch": 1.053375,
      "grad_norm": 0.3443029820919037,
      "learning_rate": 0.00024584615814711626,
      "loss": 2.8285900115966798,
      "step": 2380
    },
    {
      "epoch": 1.054625,
      "grad_norm": 0.31065329909324646,
      "learning_rate": 0.00024538065101118833,
      "loss": 2.8233985900878906,
      "step": 2390
    },
    {
      "epoch": 1.055875,
      "grad_norm": 0.31999537348747253,
      "learning_rate": 0.0002449135965905457,
      "loss": 2.819304656982422,
      "step": 2400
    },
    {
      "epoch": 1.057125,
      "grad_norm": 0.3119924068450928,
      "learning_rate": 0.0002444450024618422,
      "loss": 2.794045257568359,
      "step": 2410
    },
    {
      "epoch": 1.058375,
      "grad_norm": 0.30880260467529297,
      "learning_rate": 0.00024397487622670894,
      "loss": 2.8043006896972655,
      "step": 2420
    },
    {
      "epoch": 1.059625,
      "grad_norm": 0.33332186937332153,
      "learning_rate": 0.0002435032255116313,
      "loss": 2.795323371887207,
      "step": 2430
    },
    {
      "epoch": 1.060875,
      "grad_norm": 0.313965767621994,
      "learning_rate": 0.00024303005796782508,
      "loss": 2.823124313354492,
      "step": 2440
    },
    {
      "epoch": 1.062125,
      "grad_norm": 0.3164885938167572,
      "learning_rate": 0.0002425553812711123,
      "loss": 2.8160400390625,
      "step": 2450
    },
    {
      "epoch": 1.063375,
      "grad_norm": 0.3221667408943176,
      "learning_rate": 0.00024207920312179686,
      "loss": 2.7901626586914063,
      "step": 2460
    },
    {
      "epoch": 1.064625,
      "grad_norm": 0.31535276770591736,
      "learning_rate": 0.0002416015312445396,
      "loss": 2.7759862899780274,
      "step": 2470
    },
    {
      "epoch": 1.065875,
      "grad_norm": 0.3354146480560303,
      "learning_rate": 0.0002411223733882328,
      "loss": 2.7960500717163086,
      "step": 2480
    },
    {
      "epoch": 1.067125,
      "grad_norm": 0.36151498556137085,
      "learning_rate": 0.0002406417373258746,
      "loss": 2.787026596069336,
      "step": 2490
    },
    {
      "epoch": 1.068375,
      "grad_norm": 0.32125601172447205,
      "learning_rate": 0.00024015963085444297,
      "loss": 2.8015857696533204,
      "step": 2500
    },
    {
      "epoch": 1.069625,
      "grad_norm": 0.31707149744033813,
      "learning_rate": 0.00023967606179476914,
      "loss": 2.7936481475830077,
      "step": 2510
    },
    {
      "epoch": 1.070875,
      "grad_norm": 0.3259861171245575,
      "learning_rate": 0.00023919103799141078,
      "loss": 2.7787807464599608,
      "step": 2520
    },
    {
      "epoch": 1.072125,
      "grad_norm": 0.32442429661750793,
      "learning_rate": 0.00023870456731252466,
      "loss": 2.7599460601806642,
      "step": 2530
    },
    {
      "epoch": 1.073375,
      "grad_norm": 0.3134000599384308,
      "learning_rate": 0.0002382166576497391,
      "loss": 2.802600288391113,
      "step": 2540
    },
    {
      "epoch": 1.074625,
      "grad_norm": 0.31214284896850586,
      "learning_rate": 0.00023772731691802583,
      "loss": 2.783060073852539,
      "step": 2550
    },
    {
      "epoch": 1.075875,
      "grad_norm": 0.31292805075645447,
      "learning_rate": 0.00023723655305557187,
      "loss": 2.766581153869629,
      "step": 2560
    },
    {
      "epoch": 1.077125,
      "grad_norm": 0.3185879588127136,
      "learning_rate": 0.0002367443740236504,
      "loss": 2.7777988433837892,
      "step": 2570
    },
    {
      "epoch": 1.078375,
      "grad_norm": 0.33435672521591187,
      "learning_rate": 0.00023625078780649178,
      "loss": 2.783607292175293,
      "step": 2580
    },
    {
      "epoch": 1.079625,
      "grad_norm": 0.32837480306625366,
      "learning_rate": 0.00023575580241115408,
      "loss": 2.778690719604492,
      "step": 2590
    },
    {
      "epoch": 1.080875,
      "grad_norm": 0.3214893639087677,
      "learning_rate": 0.00023525942586739309,
      "loss": 2.7992782592773438,
      "step": 2600
    },
    {
      "epoch": 1.082125,
      "grad_norm": 0.31656068563461304,
      "learning_rate": 0.00023476166622753212,
      "loss": 2.7780582427978517,
      "step": 2610
    },
    {
      "epoch": 1.083375,
      "grad_norm": 0.32466331124305725,
      "learning_rate": 0.0002342625315663314,
      "loss": 2.756012535095215,
      "step": 2620
    },
    {
      "epoch": 1.084625,
      "grad_norm": 0.33582648634910583,
      "learning_rate": 0.0002337620299808569,
      "loss": 2.791756820678711,
      "step": 2630
    },
    {
      "epoch": 1.085875,
      "grad_norm": 0.320222944021225,
      "learning_rate": 0.00023326016959034922,
      "loss": 2.7860565185546875,
      "step": 2640
    },
    {
      "epoch": 1.087125,
      "grad_norm": 0.30881232023239136,
      "learning_rate": 0.00023275695853609184,
      "loss": 2.765717315673828,
      "step": 2650
    },
    {
      "epoch": 1.088375,
      "grad_norm": 0.32089343667030334,
      "learning_rate": 0.00023225240498127883,
      "loss": 2.798817825317383,
      "step": 2660
    },
    {
      "epoch": 1.089625,
      "grad_norm": 0.3162356913089752,
      "learning_rate": 0.00023174651711088272,
      "loss": 2.7743175506591795,
      "step": 2670
    },
    {
      "epoch": 1.090875,
      "grad_norm": 0.3252812623977661,
      "learning_rate": 0.0002312393031315215,
      "loss": 2.74603271484375,
      "step": 2680
    },
    {
      "epoch": 1.092125,
      "grad_norm": 0.32010287046432495,
      "learning_rate": 0.00023073077127132562,
      "loss": 2.7529132843017576,
      "step": 2690
    },
    {
      "epoch": 1.093375,
      "grad_norm": 0.31325286626815796,
      "learning_rate": 0.00023022092977980442,
      "loss": 2.759819984436035,
      "step": 2700
    },
    {
      "epoch": 1.094625,
      "grad_norm": 0.31815576553344727,
      "learning_rate": 0.00022970978692771242,
      "loss": 2.762770080566406,
      "step": 2710
    },
    {
      "epoch": 1.095875,
      "grad_norm": 0.3144858181476593,
      "learning_rate": 0.00022919735100691504,
      "loss": 2.7560394287109373,
      "step": 2720
    },
    {
      "epoch": 1.097125,
      "grad_norm": 0.3316778540611267,
      "learning_rate": 0.00022868363033025406,
      "loss": 2.7469573974609376,
      "step": 2730
    },
    {
      "epoch": 1.098375,
      "grad_norm": 0.3257002830505371,
      "learning_rate": 0.000228168633231413,
      "loss": 2.743033218383789,
      "step": 2740
    },
    {
      "epoch": 1.099625,
      "grad_norm": 0.31909677386283875,
      "learning_rate": 0.00022765236806478154,
      "loss": 2.7673528671264647,
      "step": 2750
    },
    {
      "epoch": 1.100875,
      "grad_norm": 0.31668204069137573,
      "learning_rate": 0.00022713484320532055,
      "loss": 2.7211772918701174,
      "step": 2760
    },
    {
      "epoch": 1.102125,
      "grad_norm": 0.3331785202026367,
      "learning_rate": 0.00022661606704842558,
      "loss": 2.722859191894531,
      "step": 2770
    },
    {
      "epoch": 1.103375,
      "grad_norm": 0.33396124839782715,
      "learning_rate": 0.00022609604800979111,
      "loss": 2.7449823379516602,
      "step": 2780
    },
    {
      "epoch": 1.104625,
      "grad_norm": 0.31729063391685486,
      "learning_rate": 0.00022557479452527392,
      "loss": 2.725127601623535,
      "step": 2790
    },
    {
      "epoch": 1.105875,
      "grad_norm": 0.3257978558540344,
      "learning_rate": 0.00022505231505075613,
      "loss": 2.7355205535888674,
      "step": 2800
    },
    {
      "epoch": 1.107125,
      "grad_norm": 0.33255940675735474,
      "learning_rate": 0.00022452861806200838,
      "loss": 2.7257015228271486,
      "step": 2810
    },
    {
      "epoch": 1.108375,
      "grad_norm": 0.32612425088882446,
      "learning_rate": 0.00022400371205455176,
      "loss": 2.767654800415039,
      "step": 2820
    },
    {
      "epoch": 1.109625,
      "grad_norm": 0.3252048194408417,
      "learning_rate": 0.0002234776055435205,
      "loss": 2.7518121719360353,
      "step": 2830
    },
    {
      "epoch": 1.110875,
      "grad_norm": 0.31889963150024414,
      "learning_rate": 0.00022295030706352356,
      "loss": 2.7312660217285156,
      "step": 2840
    },
    {
      "epoch": 1.112125,
      "grad_norm": 0.319558709859848,
      "learning_rate": 0.00022242182516850635,
      "loss": 2.740167236328125,
      "step": 2850
    },
    {
      "epoch": 1.113375,
      "grad_norm": 0.3091806173324585,
      "learning_rate": 0.0002218921684316119,
      "loss": 2.748370552062988,
      "step": 2860
    },
    {
      "epoch": 1.114625,
      "grad_norm": 0.33689746260643005,
      "learning_rate": 0.00022136134544504163,
      "loss": 2.787190628051758,
      "step": 2870
    },
    {
      "epoch": 1.115875,
      "grad_norm": 0.31660985946655273,
      "learning_rate": 0.0002208293648199162,
      "loss": 2.7415428161621094,
      "step": 2880
    },
    {
      "epoch": 1.117125,
      "grad_norm": 0.325989305973053,
      "learning_rate": 0.0002202962351861357,
      "loss": 2.7200296401977537,
      "step": 2890
    },
    {
      "epoch": 1.118375,
      "grad_norm": 0.3281361758708954,
      "learning_rate": 0.0002197619651922397,
      "loss": 2.7388254165649415,
      "step": 2900
    }
  ],
  "logging_steps": 10,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.8847708770441626e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}