{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004484304932735426,
      "grad_norm": 4.697020980518476,
      "learning_rate": 4.4843049327354265e-08,
      "loss": 0.9912,
      "step": 1
    },
    {
      "epoch": 0.004484304932735426,
      "grad_norm": 5.023039914107906,
      "learning_rate": 4.484304932735426e-07,
      "loss": 1.0335,
      "step": 10
    },
    {
      "epoch": 0.008968609865470852,
      "grad_norm": 4.618075407201601,
      "learning_rate": 8.968609865470852e-07,
      "loss": 1.0355,
      "step": 20
    },
    {
      "epoch": 0.013452914798206279,
      "grad_norm": 2.181160774619855,
      "learning_rate": 1.345291479820628e-06,
      "loss": 0.9824,
      "step": 30
    },
    {
      "epoch": 0.017937219730941704,
      "grad_norm": 2.008569343268214,
      "learning_rate": 1.7937219730941704e-06,
      "loss": 0.9383,
      "step": 40
    },
    {
      "epoch": 0.02242152466367713,
      "grad_norm": 1.7133805307500916,
      "learning_rate": 2.242152466367713e-06,
      "loss": 0.87,
      "step": 50
    },
    {
      "epoch": 0.026905829596412557,
      "grad_norm": 1.6402285214131669,
      "learning_rate": 2.690582959641256e-06,
      "loss": 0.8608,
      "step": 60
    },
    {
      "epoch": 0.03139013452914798,
      "grad_norm": 1.548076900788033,
      "learning_rate": 3.1390134529147986e-06,
      "loss": 0.8675,
      "step": 70
    },
    {
      "epoch": 0.03587443946188341,
      "grad_norm": 1.5686573872779637,
      "learning_rate": 3.587443946188341e-06,
      "loss": 0.8581,
      "step": 80
    },
    {
      "epoch": 0.04035874439461883,
      "grad_norm": 1.8398749741247653,
      "learning_rate": 4.0358744394618836e-06,
      "loss": 0.8272,
      "step": 90
    },
    {
      "epoch": 0.04484304932735426,
      "grad_norm": 1.431522929426351,
      "learning_rate": 4.484304932735426e-06,
      "loss": 0.837,
      "step": 100
    },
    {
      "epoch": 0.04932735426008968,
      "grad_norm": 1.5183209831163855,
      "learning_rate": 4.932735426008969e-06,
      "loss": 0.7993,
      "step": 110
    },
    {
      "epoch": 0.053811659192825115,
      "grad_norm": 1.440741479973038,
      "learning_rate": 5.381165919282512e-06,
      "loss": 0.7768,
      "step": 120
    },
    {
      "epoch": 0.05829596412556054,
      "grad_norm": 1.6957957755628803,
      "learning_rate": 5.8295964125560544e-06,
      "loss": 0.793,
      "step": 130
    },
    {
      "epoch": 0.06278026905829596,
      "grad_norm": 1.5847929433417824,
      "learning_rate": 6.278026905829597e-06,
      "loss": 0.8145,
      "step": 140
    },
    {
      "epoch": 0.06726457399103139,
      "grad_norm": 1.721638040499256,
      "learning_rate": 6.72645739910314e-06,
      "loss": 0.7677,
      "step": 150
    },
    {
      "epoch": 0.07174887892376682,
      "grad_norm": 1.8900581111883776,
      "learning_rate": 7.174887892376682e-06,
      "loss": 0.784,
      "step": 160
    },
    {
      "epoch": 0.07623318385650224,
      "grad_norm": 1.554784427401762,
      "learning_rate": 7.6233183856502244e-06,
      "loss": 0.7595,
      "step": 170
    },
    {
      "epoch": 0.08071748878923767,
      "grad_norm": 1.8326262271816938,
      "learning_rate": 8.071748878923767e-06,
      "loss": 0.7722,
      "step": 180
    },
    {
      "epoch": 0.08520179372197309,
      "grad_norm": 1.4707798772479566,
      "learning_rate": 8.52017937219731e-06,
      "loss": 0.764,
      "step": 190
    },
    {
      "epoch": 0.08968609865470852,
      "grad_norm": 1.449476026227771,
      "learning_rate": 8.968609865470853e-06,
      "loss": 0.773,
      "step": 200
    },
    {
      "epoch": 0.09417040358744394,
      "grad_norm": 1.5945882538155545,
      "learning_rate": 9.417040358744395e-06,
      "loss": 0.7549,
      "step": 210
    },
    {
      "epoch": 0.09865470852017937,
      "grad_norm": 1.4765341188830319,
      "learning_rate": 9.865470852017938e-06,
      "loss": 0.784,
      "step": 220
    },
    {
      "epoch": 0.1031390134529148,
      "grad_norm": 1.452955640004767,
      "learning_rate": 9.999699851108367e-06,
      "loss": 0.7909,
      "step": 230
    },
    {
      "epoch": 0.10762331838565023,
      "grad_norm": 1.5997501860541272,
      "learning_rate": 9.9982298208374e-06,
      "loss": 0.7651,
      "step": 240
    },
    {
      "epoch": 0.11210762331838565,
      "grad_norm": 1.554511883278329,
      "learning_rate": 9.995535139530904e-06,
      "loss": 0.7621,
      "step": 250
    },
    {
      "epoch": 0.11659192825112108,
      "grad_norm": 1.6319742757477633,
      "learning_rate": 9.991616467431486e-06,
      "loss": 0.7906,
      "step": 260
    },
    {
      "epoch": 0.1210762331838565,
      "grad_norm": 1.6863475759473823,
      "learning_rate": 9.986474764680236e-06,
      "loss": 0.7684,
      "step": 270
    },
    {
      "epoch": 0.12556053811659193,
      "grad_norm": 1.5563949560805244,
      "learning_rate": 9.98011129108149e-06,
      "loss": 0.793,
      "step": 280
    },
    {
      "epoch": 0.13004484304932734,
      "grad_norm": 1.6203440473254576,
      "learning_rate": 9.972527605794151e-06,
      "loss": 0.771,
      "step": 290
    },
    {
      "epoch": 0.13452914798206278,
      "grad_norm": 1.5211967329248808,
      "learning_rate": 9.963725566949674e-06,
      "loss": 0.7865,
      "step": 300
    },
    {
      "epoch": 0.13901345291479822,
      "grad_norm": 1.3996609263865165,
      "learning_rate": 9.953707331196787e-06,
      "loss": 0.7397,
      "step": 310
    },
    {
      "epoch": 0.14349775784753363,
      "grad_norm": 1.7467903395183983,
      "learning_rate": 9.94247535317308e-06,
      "loss": 0.802,
      "step": 320
    },
    {
      "epoch": 0.14798206278026907,
      "grad_norm": 1.4561862746225176,
      "learning_rate": 9.930032384903566e-06,
      "loss": 0.78,
      "step": 330
    },
    {
      "epoch": 0.15246636771300448,
      "grad_norm": 1.261592243705757,
      "learning_rate": 9.916381475126406e-06,
      "loss": 0.7678,
      "step": 340
    },
    {
      "epoch": 0.15695067264573992,
      "grad_norm": 1.3845441861346746,
      "learning_rate": 9.901525968545907e-06,
      "loss": 0.7462,
      "step": 350
    },
    {
      "epoch": 0.16143497757847533,
      "grad_norm": 1.330783975393604,
      "learning_rate": 9.885469505013006e-06,
      "loss": 0.7516,
      "step": 360
    },
    {
      "epoch": 0.16591928251121077,
      "grad_norm": 1.5411411223309597,
      "learning_rate": 9.868216018633456e-06,
      "loss": 0.7617,
      "step": 370
    },
    {
      "epoch": 0.17040358744394618,
      "grad_norm": 1.4997661320107978,
      "learning_rate": 9.8497697368039e-06,
      "loss": 0.7433,
      "step": 380
    },
    {
      "epoch": 0.17488789237668162,
      "grad_norm": 1.4613487182945122,
      "learning_rate": 9.830135179176086e-06,
      "loss": 0.7977,
      "step": 390
    },
    {
      "epoch": 0.17937219730941703,
      "grad_norm": 1.3823002584421413,
      "learning_rate": 9.809317156549476e-06,
      "loss": 0.7668,
      "step": 400
    },
    {
      "epoch": 0.18385650224215247,
      "grad_norm": 1.3405610913825137,
      "learning_rate": 9.787320769692517e-06,
      "loss": 0.755,
      "step": 410
    },
    {
      "epoch": 0.18834080717488788,
      "grad_norm": 1.321174566440371,
      "learning_rate": 9.76415140809287e-06,
      "loss": 0.7712,
      "step": 420
    },
    {
      "epoch": 0.19282511210762332,
      "grad_norm": 1.5855307049280556,
      "learning_rate": 9.739814748636892e-06,
      "loss": 0.7876,
      "step": 430
    },
    {
      "epoch": 0.19730941704035873,
      "grad_norm": 1.3277684374580685,
      "learning_rate": 9.7143167542187e-06,
      "loss": 0.7497,
      "step": 440
    },
    {
      "epoch": 0.20179372197309417,
      "grad_norm": 1.3316052843966186,
      "learning_rate": 9.687663672279167e-06,
      "loss": 0.7742,
      "step": 450
    },
    {
      "epoch": 0.2062780269058296,
      "grad_norm": 1.3422686862026139,
      "learning_rate": 9.659862033275188e-06,
      "loss": 0.7443,
      "step": 460
    },
    {
      "epoch": 0.21076233183856502,
      "grad_norm": 1.3794382112115433,
      "learning_rate": 9.630918649079606e-06,
      "loss": 0.7423,
      "step": 470
    },
    {
      "epoch": 0.21524663677130046,
      "grad_norm": 1.3473781761704757,
      "learning_rate": 9.600840611312198e-06,
      "loss": 0.756,
      "step": 480
    },
    {
      "epoch": 0.21973094170403587,
      "grad_norm": 1.477159074283593,
      "learning_rate": 9.569635289602098e-06,
      "loss": 0.758,
      "step": 490
    },
    {
      "epoch": 0.2242152466367713,
      "grad_norm": 1.4406496372056654,
      "learning_rate": 9.537310329782109e-06,
      "loss": 0.7373,
      "step": 500
    },
    {
      "epoch": 0.22869955156950672,
      "grad_norm": 1.3236307975388621,
      "learning_rate": 9.503873652015358e-06,
      "loss": 0.7485,
      "step": 510
    },
    {
      "epoch": 0.23318385650224216,
      "grad_norm": 1.2168213830447414,
      "learning_rate": 9.469333448854713e-06,
      "loss": 0.7518,
      "step": 520
    },
    {
      "epoch": 0.23766816143497757,
      "grad_norm": 1.4695219464522695,
      "learning_rate": 9.433698183235468e-06,
      "loss": 0.7389,
      "step": 530
    },
    {
      "epoch": 0.242152466367713,
      "grad_norm": 1.5460371795366352,
      "learning_rate": 9.39697658640179e-06,
      "loss": 0.7606,
      "step": 540
    },
    {
      "epoch": 0.24663677130044842,
      "grad_norm": 1.36384817307445,
      "learning_rate": 9.359177655767398e-06,
      "loss": 0.7573,
      "step": 550
    },
    {
      "epoch": 0.25112107623318386,
      "grad_norm": 1.2974873879986306,
      "learning_rate": 9.320310652711062e-06,
      "loss": 0.7447,
      "step": 560
    },
    {
      "epoch": 0.2556053811659193,
      "grad_norm": 1.3924395179060305,
      "learning_rate": 9.2803851003074e-06,
      "loss": 0.7346,
      "step": 570
    },
    {
      "epoch": 0.2600896860986547,
      "grad_norm": 1.3646463661968233,
      "learning_rate": 9.239410780993565e-06,
      "loss": 0.7637,
      "step": 580
    },
    {
      "epoch": 0.2645739910313901,
      "grad_norm": 1.5500883315093192,
      "learning_rate": 9.197397734172381e-06,
      "loss": 0.7352,
      "step": 590
    },
    {
      "epoch": 0.26905829596412556,
      "grad_norm": 1.2384578247121611,
      "learning_rate": 9.154356253752519e-06,
      "loss": 0.7467,
      "step": 600
    },
    {
      "epoch": 0.273542600896861,
      "grad_norm": 1.4632407956897133,
      "learning_rate": 9.110296885626315e-06,
      "loss": 0.7592,
      "step": 610
    },
    {
      "epoch": 0.27802690582959644,
      "grad_norm": 1.4743516068749583,
      "learning_rate": 9.065230425085849e-06,
      "loss": 0.7471,
      "step": 620
    },
    {
      "epoch": 0.2825112107623318,
      "grad_norm": 1.609459180353317,
      "learning_rate": 9.01916791417792e-06,
      "loss": 0.7411,
      "step": 630
    },
    {
      "epoch": 0.28699551569506726,
      "grad_norm": 1.4163468213726333,
      "learning_rate": 8.97212063899854e-06,
      "loss": 0.7583,
      "step": 640
    },
    {
      "epoch": 0.2914798206278027,
      "grad_norm": 1.2789206960042645,
      "learning_rate": 8.924100126927672e-06,
      "loss": 0.7637,
      "step": 650
    },
    {
      "epoch": 0.29596412556053814,
      "grad_norm": 1.4348891847742615,
      "learning_rate": 8.87511814380481e-06,
      "loss": 0.7376,
      "step": 660
    },
    {
      "epoch": 0.3004484304932735,
      "grad_norm": 1.4055551714674843,
      "learning_rate": 8.825186691046156e-06,
      "loss": 0.7544,
      "step": 670
    },
    {
      "epoch": 0.30493273542600896,
      "grad_norm": 1.3076750983715024,
      "learning_rate": 8.774318002704072e-06,
      "loss": 0.7388,
      "step": 680
    },
    {
      "epoch": 0.3094170403587444,
      "grad_norm": 1.2832634236951583,
      "learning_rate": 8.722524542469517e-06,
      "loss": 0.7386,
      "step": 690
    },
    {
      "epoch": 0.31390134529147984,
      "grad_norm": 1.3612825976793976,
      "learning_rate": 8.669819000618248e-06,
      "loss": 0.768,
      "step": 700
    },
    {
      "epoch": 0.3183856502242152,
      "grad_norm": 1.3343477240590562,
      "learning_rate": 8.616214290901474e-06,
      "loss": 0.7244,
      "step": 710
    },
    {
      "epoch": 0.32286995515695066,
      "grad_norm": 1.3061898085537815,
      "learning_rate": 8.56172354738178e-06,
      "loss": 0.7368,
      "step": 720
    },
    {
      "epoch": 0.3273542600896861,
      "grad_norm": 1.423008056010291,
      "learning_rate": 8.506360121215046e-06,
      "loss": 0.7297,
      "step": 730
    },
    {
      "epoch": 0.33183856502242154,
      "grad_norm": 1.258418518673196,
      "learning_rate": 8.4501375773792e-06,
      "loss": 0.7322,
      "step": 740
    },
    {
      "epoch": 0.336322869955157,
      "grad_norm": 1.4805131801943718,
      "learning_rate": 8.39306969135056e-06,
      "loss": 0.7284,
      "step": 750
    },
    {
      "epoch": 0.34080717488789236,
      "grad_norm": 1.3260411374117855,
      "learning_rate": 8.335170445728609e-06,
      "loss": 0.7618,
      "step": 760
    },
    {
      "epoch": 0.3452914798206278,
      "grad_norm": 1.3351013471897553,
      "learning_rate": 8.276454026810026e-06,
      "loss": 0.7454,
      "step": 770
    },
    {
      "epoch": 0.34977578475336324,
      "grad_norm": 1.267536863507402,
      "learning_rate": 8.216934821112803e-06,
      "loss": 0.742,
      "step": 780
    },
    {
      "epoch": 0.3542600896860987,
      "grad_norm": 1.4791659570865663,
      "learning_rate": 8.156627411851295e-06,
      "loss": 0.7483,
      "step": 790
    },
    {
      "epoch": 0.35874439461883406,
      "grad_norm": 1.4042375363677306,
      "learning_rate": 8.095546575363098e-06,
      "loss": 0.7134,
      "step": 800
    },
    {
      "epoch": 0.3632286995515695,
      "grad_norm": 1.2552675050466975,
      "learning_rate": 8.033707277488585e-06,
      "loss": 0.7186,
      "step": 810
    },
    {
      "epoch": 0.36771300448430494,
      "grad_norm": 1.281776208560821,
      "learning_rate": 7.97112466990403e-06,
      "loss": 0.7367,
      "step": 820
    },
    {
      "epoch": 0.3721973094170404,
      "grad_norm": 1.417675486477273,
      "learning_rate": 7.907814086409183e-06,
      "loss": 0.7399,
      "step": 830
    },
    {
      "epoch": 0.37668161434977576,
      "grad_norm": 1.4911782746859528,
      "learning_rate": 7.843791039170232e-06,
      "loss": 0.738,
      "step": 840
    },
    {
      "epoch": 0.3811659192825112,
      "grad_norm": 1.3628167592658191,
      "learning_rate": 7.779071214919068e-06,
      "loss": 0.7404,
      "step": 850
    },
    {
      "epoch": 0.38565022421524664,
      "grad_norm": 1.256534318600846,
      "learning_rate": 7.713670471109749e-06,
      "loss": 0.7364,
      "step": 860
    },
    {
      "epoch": 0.3901345291479821,
      "grad_norm": 1.3061985247991272,
      "learning_rate": 7.647604832033178e-06,
      "loss": 0.7535,
      "step": 870
    },
    {
      "epoch": 0.39461883408071746,
      "grad_norm": 1.3626743194809932,
      "learning_rate": 7.580890484890864e-06,
      "loss": 0.7212,
      "step": 880
    },
    {
      "epoch": 0.3991031390134529,
      "grad_norm": 1.283405821881289,
      "learning_rate": 7.513543775828791e-06,
      "loss": 0.7336,
      "step": 890
    },
    {
      "epoch": 0.40358744394618834,
      "grad_norm": 1.4217858183255085,
      "learning_rate": 7.445581205932335e-06,
      "loss": 0.7349,
      "step": 900
    },
    {
      "epoch": 0.4080717488789238,
      "grad_norm": 1.3304359323377464,
      "learning_rate": 7.377019427183213e-06,
      "loss": 0.7265,
      "step": 910
    },
    {
      "epoch": 0.4125560538116592,
      "grad_norm": 1.3324230635898773,
      "learning_rate": 7.30787523837947e-06,
      "loss": 0.7451,
      "step": 920
    },
    {
      "epoch": 0.4170403587443946,
      "grad_norm": 1.2976336030674929,
      "learning_rate": 7.238165581019488e-06,
      "loss": 0.7415,
      "step": 930
    },
    {
      "epoch": 0.42152466367713004,
      "grad_norm": 1.3082251443805497,
      "learning_rate": 7.167907535151027e-06,
      "loss": 0.7405,
      "step": 940
    },
    {
      "epoch": 0.4260089686098655,
      "grad_norm": 1.368310990039969,
      "learning_rate": 7.097118315186335e-06,
      "loss": 0.7141,
      "step": 950
    },
    {
      "epoch": 0.4304932735426009,
      "grad_norm": 1.3507885090012453,
      "learning_rate": 7.025815265684315e-06,
      "loss": 0.744,
      "step": 960
    },
    {
      "epoch": 0.4349775784753363,
      "grad_norm": 1.3870418574466943,
      "learning_rate": 6.9540158571008105e-06,
      "loss": 0.7344,
      "step": 970
    },
    {
      "epoch": 0.43946188340807174,
      "grad_norm": 1.2702369270052376,
      "learning_rate": 6.881737681508065e-06,
      "loss": 0.7131,
      "step": 980
    },
    {
      "epoch": 0.4439461883408072,
      "grad_norm": 1.3694518474478212,
      "learning_rate": 6.808998448284347e-06,
      "loss": 0.7516,
      "step": 990
    },
    {
      "epoch": 0.4484304932735426,
      "grad_norm": 1.3967080566658139,
      "learning_rate": 6.735815979774865e-06,
      "loss": 0.7208,
      "step": 1000
    },
    {
      "epoch": 0.452914798206278,
      "grad_norm": 1.4517646369323314,
      "learning_rate": 6.662208206924986e-06,
      "loss": 0.7455,
      "step": 1010
    },
    {
      "epoch": 0.45739910313901344,
      "grad_norm": 1.2752229454885209,
      "learning_rate": 6.588193164886847e-06,
      "loss": 0.7555,
      "step": 1020
    },
    {
      "epoch": 0.4618834080717489,
      "grad_norm": 1.3528261238688069,
      "learning_rate": 6.513788988600441e-06,
      "loss": 0.7428,
      "step": 1030
    },
    {
      "epoch": 0.4663677130044843,
      "grad_norm": 1.340831615562883,
      "learning_rate": 6.439013908350249e-06,
      "loss": 0.7446,
      "step": 1040
    },
    {
      "epoch": 0.47085201793721976,
      "grad_norm": 1.3233981064583105,
      "learning_rate": 6.363886245298514e-06,
      "loss": 0.6945,
      "step": 1050
    },
    {
      "epoch": 0.47533632286995514,
      "grad_norm": 1.29683857481815,
      "learning_rate": 6.288424406996237e-06,
      "loss": 0.7085,
      "step": 1060
    },
    {
      "epoch": 0.4798206278026906,
      "grad_norm": 1.3009424922765027,
      "learning_rate": 6.2126468828730225e-06,
      "loss": 0.7294,
      "step": 1070
    },
    {
      "epoch": 0.484304932735426,
      "grad_norm": 1.244204104767217,
      "learning_rate": 6.136572239706854e-06,
      "loss": 0.7091,
      "step": 1080
    },
    {
      "epoch": 0.48878923766816146,
      "grad_norm": 1.4080424781447183,
      "learning_rate": 6.060219117074914e-06,
      "loss": 0.724,
      "step": 1090
    },
    {
      "epoch": 0.49327354260089684,
      "grad_norm": 1.4328593518662633,
      "learning_rate": 5.983606222786577e-06,
      "loss": 0.7106,
      "step": 1100
    },
    {
      "epoch": 0.4977578475336323,
      "grad_norm": 1.3515007322722326,
      "learning_rate": 5.9067523282996775e-06,
      "loss": 0.7111,
      "step": 1110
    },
    {
      "epoch": 0.5022421524663677,
      "grad_norm": 1.3343622572750706,
      "learning_rate": 5.829676264121184e-06,
      "loss": 0.7323,
      "step": 1120
    },
    {
      "epoch": 0.5067264573991032,
      "grad_norm": 1.3346769678858248,
      "learning_rate": 5.752396915193403e-06,
      "loss": 0.744,
      "step": 1130
    },
    {
      "epoch": 0.5112107623318386,
      "grad_norm": 1.4071259742434898,
      "learning_rate": 5.6749332162668525e-06,
      "loss": 0.7181,
      "step": 1140
    },
    {
      "epoch": 0.515695067264574,
      "grad_norm": 1.314030857289351,
      "learning_rate": 5.5973041472609265e-06,
      "loss": 0.7278,
      "step": 1150
    },
    {
      "epoch": 0.5201793721973094,
      "grad_norm": 1.2893634754111014,
      "learning_rate": 5.519528728613491e-06,
      "loss": 0.722,
      "step": 1160
    },
    {
      "epoch": 0.5246636771300448,
      "grad_norm": 1.2894378069172643,
      "learning_rate": 5.4416260166205525e-06,
      "loss": 0.7282,
      "step": 1170
    },
    {
      "epoch": 0.5291479820627802,
      "grad_norm": 1.2691657277439665,
      "learning_rate": 5.363615098767149e-06,
      "loss": 0.7439,
      "step": 1180
    },
    {
      "epoch": 0.5336322869955157,
      "grad_norm": 1.3478124675299437,
      "learning_rate": 5.285515089050587e-06,
      "loss": 0.7164,
      "step": 1190
    },
    {
      "epoch": 0.5381165919282511,
      "grad_norm": 1.3052899854504807,
      "learning_rate": 5.207345123297187e-06,
      "loss": 0.7171,
      "step": 1200
    },
    {
      "epoch": 0.5426008968609866,
      "grad_norm": 1.27131464716386,
      "learning_rate": 5.129124354473688e-06,
      "loss": 0.7235,
      "step": 1210
    },
    {
      "epoch": 0.547085201793722,
      "grad_norm": 1.257329353514232,
      "learning_rate": 5.050871947994443e-06,
      "loss": 0.6999,
      "step": 1220
    },
    {
      "epoch": 0.5515695067264574,
      "grad_norm": 1.3863466652299603,
      "learning_rate": 4.972607077025563e-06,
      "loss": 0.7251,
      "step": 1230
    },
    {
      "epoch": 0.5560538116591929,
      "grad_norm": 1.394706696869187,
      "learning_rate": 4.894348917787174e-06,
      "loss": 0.6963,
      "step": 1240
    },
    {
      "epoch": 0.5605381165919282,
      "grad_norm": 1.288516832053435,
      "learning_rate": 4.816116644854912e-06,
      "loss": 0.7207,
      "step": 1250
    },
    {
      "epoch": 0.5650224215246636,
      "grad_norm": 1.291568839795731,
      "learning_rate": 4.73792942646183e-06,
      "loss": 0.7168,
      "step": 1260
    },
    {
      "epoch": 0.5695067264573991,
      "grad_norm": 1.3692501176044114,
      "learning_rate": 4.659806419801855e-06,
      "loss": 0.7311,
      "step": 1270
    },
    {
      "epoch": 0.5739910313901345,
      "grad_norm": 1.2214111714932894,
      "learning_rate": 4.581766766335953e-06,
      "loss": 0.7175,
      "step": 1280
    },
    {
      "epoch": 0.57847533632287,
      "grad_norm": 1.2261021201524966,
      "learning_rate": 4.503829587102138e-06,
      "loss": 0.722,
      "step": 1290
    },
    {
      "epoch": 0.5829596412556054,
      "grad_norm": 1.279552435182188,
      "learning_rate": 4.426013978030508e-06,
      "loss": 0.7407,
      "step": 1300
    },
    {
      "epoch": 0.5874439461883408,
      "grad_norm": 1.2639914975822624,
      "learning_rate": 4.348339005264406e-06,
      "loss": 0.7174,
      "step": 1310
    },
    {
      "epoch": 0.5919282511210763,
      "grad_norm": 1.2931987303236723,
      "learning_rate": 4.270823700488896e-06,
      "loss": 0.7236,
      "step": 1320
    },
    {
      "epoch": 0.5964125560538116,
      "grad_norm": 1.3083460389811294,
      "learning_rate": 4.19348705626768e-06,
      "loss": 0.7247,
      "step": 1330
    },
    {
      "epoch": 0.600896860986547,
      "grad_norm": 1.359946232054244,
      "learning_rate": 4.116348021389595e-06,
      "loss": 0.7289,
      "step": 1340
    },
    {
      "epoch": 0.6053811659192825,
      "grad_norm": 1.3038409489179155,
      "learning_rate": 4.039425496225834e-06,
      "loss": 0.723,
      "step": 1350
    },
    {
      "epoch": 0.6098654708520179,
      "grad_norm": 1.337031671252276,
      "learning_rate": 3.962738328099047e-06,
      "loss": 0.718,
      "step": 1360
    },
    {
      "epoch": 0.6143497757847534,
      "grad_norm": 1.3006506595251124,
      "learning_rate": 3.88630530666542e-06,
      "loss": 0.7372,
      "step": 1370
    },
    {
      "epoch": 0.6188340807174888,
      "grad_norm": 1.3038120667886732,
      "learning_rate": 3.8101451593108816e-06,
      "loss": 0.732,
      "step": 1380
    },
    {
      "epoch": 0.6233183856502242,
      "grad_norm": 1.2544712968929104,
      "learning_rate": 3.7342765465625953e-06,
      "loss": 0.7347,
      "step": 1390
    },
    {
      "epoch": 0.6278026905829597,
      "grad_norm": 1.3352755014667614,
      "learning_rate": 3.658718057516803e-06,
      "loss": 0.7332,
      "step": 1400
    },
    {
      "epoch": 0.6322869955156951,
      "grad_norm": 1.3389617347187606,
      "learning_rate": 3.5834882052841744e-06,
      "loss": 0.7154,
      "step": 1410
    },
    {
      "epoch": 0.6367713004484304,
      "grad_norm": 1.2654799213890686,
      "learning_rate": 3.508605422453799e-06,
      "loss": 0.7002,
      "step": 1420
    },
    {
      "epoch": 0.6412556053811659,
      "grad_norm": 1.311398422880929,
      "learning_rate": 3.4340880565768707e-06,
      "loss": 0.7098,
      "step": 1430
    },
    {
      "epoch": 0.6457399103139013,
      "grad_norm": 1.3577211660369808,
      "learning_rate": 3.359954365671241e-06,
      "loss": 0.7024,
      "step": 1440
    },
    {
      "epoch": 0.6502242152466368,
      "grad_norm": 1.145375472595952,
      "learning_rate": 3.2862225137478897e-06,
      "loss": 0.7097,
      "step": 1450
    },
    {
      "epoch": 0.6547085201793722,
      "grad_norm": 1.3358950336855993,
      "learning_rate": 3.2129105663604275e-06,
      "loss": 0.7148,
      "step": 1460
    },
    {
      "epoch": 0.6591928251121076,
      "grad_norm": 1.2495395838860759,
      "learning_rate": 3.1400364861787434e-06,
      "loss": 0.7483,
      "step": 1470
    },
    {
      "epoch": 0.6636771300448431,
      "grad_norm": 1.2742823576583961,
      "learning_rate": 3.0676181285878343e-06,
      "loss": 0.7063,
      "step": 1480
    },
    {
      "epoch": 0.6681614349775785,
      "grad_norm": 1.2596513250823083,
      "learning_rate": 2.9956732373129378e-06,
      "loss": 0.7201,
      "step": 1490
    },
    {
      "epoch": 0.672645739910314,
      "grad_norm": 1.4183600229677984,
      "learning_rate": 2.9242194400720157e-06,
      "loss": 0.7202,
      "step": 1500
    },
    {
      "epoch": 0.6771300448430493,
      "grad_norm": 1.224500332009707,
      "learning_rate": 2.8532742442566735e-06,
      "loss": 0.7228,
      "step": 1510
    },
    {
      "epoch": 0.6816143497757847,
      "grad_norm": 1.20210393613667,
      "learning_rate": 2.782855032642535e-06,
      "loss": 0.7386,
      "step": 1520
    },
    {
      "epoch": 0.6860986547085202,
      "grad_norm": 1.2835056973370584,
      "learning_rate": 2.712979059130187e-06,
      "loss": 0.7207,
      "step": 1530
    },
    {
      "epoch": 0.6905829596412556,
      "grad_norm": 1.180714987729606,
      "learning_rate": 2.643663444517671e-06,
      "loss": 0.6981,
      "step": 1540
    },
    {
      "epoch": 0.695067264573991,
      "grad_norm": 1.2871858226590431,
      "learning_rate": 2.5749251723055933e-06,
      "loss": 0.6853,
      "step": 1550
    },
    {
      "epoch": 0.6995515695067265,
      "grad_norm": 1.3219720717807693,
      "learning_rate": 2.5067810845358926e-06,
      "loss": 0.7192,
      "step": 1560
    },
    {
      "epoch": 0.7040358744394619,
      "grad_norm": 1.391893981214182,
      "learning_rate": 2.439247877665244e-06,
      "loss": 0.7103,
      "step": 1570
    },
    {
      "epoch": 0.7085201793721974,
      "grad_norm": 1.2636799158865641,
      "learning_rate": 2.3723420984741417e-06,
      "loss": 0.684,
      "step": 1580
    },
    {
      "epoch": 0.7130044843049327,
      "grad_norm": 1.3557464635552046,
      "learning_rate": 2.3060801400126693e-06,
      "loss": 0.7207,
      "step": 1590
    },
    {
      "epoch": 0.7174887892376681,
      "grad_norm": 1.3569256088684083,
      "learning_rate": 2.240478237583915e-06,
      "loss": 0.7077,
      "step": 1600
    },
    {
      "epoch": 0.7219730941704036,
      "grad_norm": 1.37192661939199,
      "learning_rate": 2.1755524647660514e-06,
      "loss": 0.693,
      "step": 1610
    },
    {
      "epoch": 0.726457399103139,
      "grad_norm": 1.248391921620642,
      "learning_rate": 2.1113187294740294e-06,
      "loss": 0.6911,
      "step": 1620
    },
    {
      "epoch": 0.7309417040358744,
      "grad_norm": 1.357856212054366,
      "learning_rate": 2.047792770061881e-06,
      "loss": 0.6838,
      "step": 1630
    },
    {
      "epoch": 0.7354260089686099,
      "grad_norm": 1.4147463764446673,
      "learning_rate": 1.9849901514665458e-06,
      "loss": 0.7122,
      "step": 1640
    },
    {
      "epoch": 0.7399103139013453,
      "grad_norm": 1.3587967144630926,
      "learning_rate": 1.922926261394206e-06,
      "loss": 0.6927,
      "step": 1650
    },
    {
      "epoch": 0.7443946188340808,
      "grad_norm": 1.326112035320906,
      "learning_rate": 1.8616163065500231e-06,
      "loss": 0.6931,
      "step": 1660
    },
    {
      "epoch": 0.7488789237668162,
      "grad_norm": 1.2337199420172695,
      "learning_rate": 1.8010753089122572e-06,
      "loss": 0.6934,
      "step": 1670
    },
    {
      "epoch": 0.7533632286995515,
      "grad_norm": 1.2176765529306792,
      "learning_rate": 1.7413181020516146e-06,
      "loss": 0.7164,
      "step": 1680
    },
    {
      "epoch": 0.757847533632287,
      "grad_norm": 1.2724919849788396,
      "learning_rate": 1.6823593274967703e-06,
      "loss": 0.7267,
      "step": 1690
    },
    {
      "epoch": 0.7623318385650224,
      "grad_norm": 1.3274428154634692,
      "learning_rate": 1.6242134311469538e-06,
      "loss": 0.6824,
      "step": 1700
    },
    {
      "epoch": 0.7668161434977578,
      "grad_norm": 1.3661211722387332,
      "learning_rate": 1.5668946597324558e-06,
      "loss": 0.7182,
      "step": 1710
    },
    {
      "epoch": 0.7713004484304933,
      "grad_norm": 1.2959174525555186,
      "learning_rate": 1.51041705732393e-06,
      "loss": 0.7118,
      "step": 1720
    },
    {
      "epoch": 0.7757847533632287,
      "grad_norm": 1.2773849053623296,
      "learning_rate": 1.4547944618913706e-06,
      "loss": 0.6929,
      "step": 1730
    },
    {
      "epoch": 0.7802690582959642,
      "grad_norm": 1.297150269984566,
      "learning_rate": 1.4000405019135676e-06,
      "loss": 0.6883,
      "step": 1740
    },
    {
      "epoch": 0.7847533632286996,
      "grad_norm": 1.400539444832659,
      "learning_rate": 1.3461685930388958e-06,
      "loss": 0.6911,
      "step": 1750
    },
    {
      "epoch": 0.7892376681614349,
      "grad_norm": 1.2337657683341194,
      "learning_rate": 1.2931919347982607e-06,
      "loss": 0.6921,
      "step": 1760
    },
    {
      "epoch": 0.7937219730941704,
      "grad_norm": 1.2072415924732212,
      "learning_rate": 1.2411235073709883e-06,
      "loss": 0.7102,
      "step": 1770
    },
    {
      "epoch": 0.7982062780269058,
      "grad_norm": 1.381367292756694,
      "learning_rate": 1.1899760684044515e-06,
      "loss": 0.6838,
      "step": 1780
    },
    {
      "epoch": 0.8026905829596412,
      "grad_norm": 1.3102643665842388,
      "learning_rate": 1.1397621498882471e-06,
      "loss": 0.6945,
      "step": 1790
    },
    {
      "epoch": 0.8071748878923767,
      "grad_norm": 1.3214717866253802,
      "learning_rate": 1.0904940550836285e-06,
      "loss": 0.7016,
      "step": 1800
    },
    {
      "epoch": 0.8116591928251121,
      "grad_norm": 1.1240219719921938,
      "learning_rate": 1.0421838555090119e-06,
      "loss": 0.7018,
      "step": 1810
    },
    {
      "epoch": 0.8161434977578476,
      "grad_norm": 1.2844428659053804,
      "learning_rate": 9.948433879822428e-07,
      "loss": 0.7361,
      "step": 1820
    },
    {
      "epoch": 0.820627802690583,
      "grad_norm": 1.249492840397432,
      "learning_rate": 9.484842517203735e-07,
      "loss": 0.707,
      "step": 1830
    },
    {
      "epoch": 0.8251121076233184,
      "grad_norm": 1.2793612843154911,
      "learning_rate": 9.031178054976636e-07,
      "loss": 0.7226,
      "step": 1840
    },
    {
      "epoch": 0.8295964125560538,
      "grad_norm": 1.3447642426455195,
      "learning_rate": 8.587551648624859e-07,
      "loss": 0.6906,
      "step": 1850
    },
    {
      "epoch": 0.8340807174887892,
      "grad_norm": 1.282168137173779,
      "learning_rate": 8.154071994138241e-07,
      "loss": 0.698,
      "step": 1860
    },
    {
      "epoch": 0.8385650224215246,
      "grad_norm": 1.4629920435128856,
      "learning_rate": 7.730845301380441e-07,
      "loss": 0.7212,
      "step": 1870
    },
    {
      "epoch": 0.8430493273542601,
      "grad_norm": 1.3143459341111923,
      "learning_rate": 7.317975268065685e-07,
      "loss": 0.6942,
      "step": 1880
    },
    {
      "epoch": 0.8475336322869955,
      "grad_norm": 1.3274198546425116,
      "learning_rate": 6.915563054351037e-07,
      "loss": 0.6944,
      "step": 1890
    },
    {
      "epoch": 0.852017937219731,
      "grad_norm": 1.2958298618538142,
      "learning_rate": 6.523707258050516e-07,
      "loss": 0.6692,
      "step": 1900
    },
    {
      "epoch": 0.8565022421524664,
      "grad_norm": 1.3403229426415875,
      "learning_rate": 6.14250389047692e-07,
      "loss": 0.7034,
      "step": 1910
    },
    {
      "epoch": 0.8609865470852018,
      "grad_norm": 1.258601487476543,
      "learning_rate": 5.772046352917399e-07,
      "loss": 0.7144,
      "step": 1920
    },
    {
      "epoch": 0.8654708520179372,
      "grad_norm": 1.1222145256140716,
      "learning_rate": 5.412425413748623e-07,
      "loss": 0.6988,
      "step": 1930
    },
    {
      "epoch": 0.8699551569506726,
      "grad_norm": 1.2674996138499859,
      "learning_rate": 5.063729186196948e-07,
      "loss": 0.7089,
      "step": 1940
    },
    {
      "epoch": 0.874439461883408,
      "grad_norm": 1.3633828734617575,
      "learning_rate": 4.7260431067491617e-07,
      "loss": 0.733,
      "step": 1950
    },
    {
      "epoch": 0.8789237668161435,
      "grad_norm": 1.2652212276863493,
      "learning_rate": 4.399449914219167e-07,
      "loss": 0.7209,
      "step": 1960
    },
    {
      "epoch": 0.8834080717488789,
      "grad_norm": 1.2975811004623845,
      "learning_rate": 4.084029629475478e-07,
      "loss": 0.7252,
      "step": 1970
    },
    {
      "epoch": 0.8878923766816144,
      "grad_norm": 1.2744740809644324,
      "learning_rate": 3.7798595358348457e-07,
      "loss": 0.7083,
      "step": 1980
    },
    {
      "epoch": 0.8923766816143498,
      "grad_norm": 1.2924446335095698,
      "learning_rate": 3.487014160126467e-07,
      "loss": 0.7077,
      "step": 1990
    },
    {
      "epoch": 0.8968609865470852,
      "grad_norm": 1.283938673421291,
      "learning_rate": 3.2055652544316695e-07,
      "loss": 0.7038,
      "step": 2000
    },
    {
      "epoch": 0.9013452914798207,
      "grad_norm": 1.3428158271865258,
      "learning_rate": 2.9355817785034325e-07,
      "loss": 0.7177,
      "step": 2010
    },
    {
      "epoch": 0.905829596412556,
      "grad_norm": 1.3237588390932393,
      "learning_rate": 2.6771298828700885e-07,
      "loss": 0.7079,
      "step": 2020
    },
    {
      "epoch": 0.9103139013452914,
      "grad_norm": 1.379188659318735,
      "learning_rate": 2.4302728926273224e-07,
      "loss": 0.7159,
      "step": 2030
    },
    {
      "epoch": 0.9147982062780269,
      "grad_norm": 1.2853687562313414,
      "learning_rate": 2.195071291922435e-07,
      "loss": 0.6842,
      "step": 2040
    },
    {
      "epoch": 0.9192825112107623,
      "grad_norm": 1.3212853096546313,
      "learning_rate": 1.9715827091347005e-07,
      "loss": 0.6994,
      "step": 2050
    },
    {
      "epoch": 0.9237668161434978,
      "grad_norm": 1.4231610496002076,
      "learning_rate": 1.7598619027554553e-07,
      "loss": 0.7032,
      "step": 2060
    },
    {
      "epoch": 0.9282511210762332,
      "grad_norm": 1.288660209861224,
      "learning_rate": 1.5599607479713396e-07,
      "loss": 0.6856,
      "step": 2070
    },
    {
      "epoch": 0.9327354260089686,
      "grad_norm": 1.291080758227027,
      "learning_rate": 1.3719282239539722e-07,
      "loss": 0.7183,
      "step": 2080
    },
    {
      "epoch": 0.9372197309417041,
      "grad_norm": 1.3527938026636528,
      "learning_rate": 1.1958104018592376e-07,
      "loss": 0.7022,
      "step": 2090
    },
    {
      "epoch": 0.9417040358744395,
      "grad_norm": 1.306930563286188,
      "learning_rate": 1.0316504335390775e-07,
      "loss": 0.7202,
      "step": 2100
    },
    {
      "epoch": 0.9461883408071748,
      "grad_norm": 1.34728936572921,
      "learning_rate": 8.79488540968565e-08,
      "loss": 0.7128,
      "step": 2110
    },
    {
      "epoch": 0.9506726457399103,
      "grad_norm": 1.5138997567977701,
      "learning_rate": 7.39362006390798e-08,
      "loss": 0.6841,
      "step": 2120
    },
    {
      "epoch": 0.9551569506726457,
      "grad_norm": 1.2518587224976194,
      "learning_rate": 6.113051631821631e-08,
      "loss": 0.71,
      "step": 2130
    },
    {
      "epoch": 0.9596412556053812,
      "grad_norm": 1.4419766059722212,
      "learning_rate": 4.9534938744004723e-08,
      "loss": 0.6944,
      "step": 2140
    },
    {
      "epoch": 0.9641255605381166,
      "grad_norm": 1.3591089992663685,
      "learning_rate": 3.915230902951761e-08,
      "loss": 0.717,
      "step": 2150
    },
    {
      "epoch": 0.968609865470852,
      "grad_norm": 1.281084457332566,
      "learning_rate": 2.9985171095041066e-08,
      "loss": 0.703,
      "step": 2160
    },
    {
      "epoch": 0.9730941704035875,
      "grad_norm": 1.1922556783094496,
      "learning_rate": 2.203577104476773e-08,
      "loss": 0.7085,
      "step": 2170
    },
    {
      "epoch": 0.9775784753363229,
      "grad_norm": 1.3917918614257676,
      "learning_rate": 1.5306056616468666e-08,
      "loss": 0.709,
      "step": 2180
    },
    {
      "epoch": 0.9820627802690582,
      "grad_norm": 1.178733370314234,
      "learning_rate": 9.797676704259574e-09,
      "loss": 0.7009,
      "step": 2190
    },
    {
      "epoch": 0.9865470852017937,
      "grad_norm": 1.2352452900540438,
      "learning_rate": 5.511980954596152e-09,
      "loss": 0.7085,
      "step": 2200
    },
    {
      "epoch": 0.9910313901345291,
      "grad_norm": 1.519249312059187,
      "learning_rate": 2.4500194355880913e-09,
      "loss": 0.6865,
      "step": 2210
    },
    {
      "epoch": 0.9955156950672646,
      "grad_norm": 1.2975150315542752,
      "learning_rate": 6.125423797137541e-10,
      "loss": 0.6935,
      "step": 2220
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.2570272992562246,
      "learning_rate": 0.0,
      "loss": 0.6946,
      "step": 2230
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7042345404624939,
      "eval_runtime": 381.2635,
      "eval_samples_per_second": 131.345,
      "eval_steps_per_second": 2.054,
      "step": 2230
    },
    {
      "epoch": 1.0,
      "step": 2230,
      "total_flos": 244955314192384.0,
      "train_loss": 0.7376568270371099,
      "train_runtime": 8720.6202,
      "train_samples_per_second": 32.731,
      "train_steps_per_second": 0.256
    }
  ],
  "logging_steps": 10,
  "max_steps": 2230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 244955314192384.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}