{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1949,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.5175645371884534,
      "learning_rate": 1.0256410256410257e-06,
      "loss": 2.4689,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.5414392036662189,
      "learning_rate": 5.128205128205128e-06,
      "loss": 2.296,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5538901364950987,
      "learning_rate": 1.0256410256410256e-05,
      "loss": 2.3771,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6356059806505323,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 2.4349,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9396860738842863,
      "learning_rate": 2.0512820512820512e-05,
      "loss": 2.2378,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.076742844293254,
      "learning_rate": 2.564102564102564e-05,
      "loss": 2.2165,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3800536964350727,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 2.0858,
      "step": 30
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6831923951115355,
      "learning_rate": 3.58974358974359e-05,
      "loss": 1.835,
      "step": 35
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6855160654030399,
      "learning_rate": 4.1025641025641023e-05,
      "loss": 1.5935,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9033517207922789,
      "learning_rate": 4.615384615384616e-05,
      "loss": 1.3204,
      "step": 45
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6789318044371851,
      "learning_rate": 5.128205128205128e-05,
      "loss": 1.1966,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6612528285257852,
      "learning_rate": 5.6410256410256414e-05,
      "loss": 1.1654,
      "step": 55
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6824673739330733,
      "learning_rate": 6.153846153846155e-05,
      "loss": 1.2263,
      "step": 60
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9537688149291184,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.1394,
      "step": 65
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.7959286790470161,
      "learning_rate": 7.17948717948718e-05,
      "loss": 1.081,
      "step": 70
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5527526649901106,
      "learning_rate": 7.692307692307693e-05,
      "loss": 1.1526,
      "step": 75
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4146640624894086,
      "learning_rate": 8.205128205128205e-05,
      "loss": 1.0054,
      "step": 80
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.6339977668776766,
      "learning_rate": 8.717948717948718e-05,
      "loss": 1.0504,
      "step": 85
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5116071525445501,
      "learning_rate": 9.230769230769232e-05,
      "loss": 1.0481,
      "step": 90
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6110345975050266,
      "learning_rate": 9.743589743589744e-05,
      "loss": 1.085,
      "step": 95
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5408855638356486,
      "learning_rate": 0.00010256410256410256,
      "loss": 1.0849,
      "step": 100
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7481217171922416,
      "learning_rate": 0.0001076923076923077,
      "loss": 1.0474,
      "step": 105
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5909177861205187,
      "learning_rate": 0.00011282051282051283,
      "loss": 1.0301,
      "step": 110
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.648775964865753,
      "learning_rate": 0.00011794871794871796,
      "loss": 0.9806,
      "step": 115
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5194742680857717,
      "learning_rate": 0.0001230769230769231,
      "loss": 1.0154,
      "step": 120
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5866960495913472,
      "learning_rate": 0.00012820512820512823,
      "loss": 0.9774,
      "step": 125
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.838042235131286,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.9929,
      "step": 130
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.676210521480851,
      "learning_rate": 0.00013846153846153847,
      "loss": 0.9398,
      "step": 135
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5991863742896065,
      "learning_rate": 0.0001435897435897436,
      "loss": 0.9318,
      "step": 140
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5783257417623179,
      "learning_rate": 0.00014871794871794872,
      "loss": 0.8787,
      "step": 145
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7852234940550689,
      "learning_rate": 0.00015384615384615385,
      "loss": 1.0355,
      "step": 150
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5673613793254546,
      "learning_rate": 0.00015897435897435896,
      "loss": 0.9257,
      "step": 155
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.716323259532379,
      "learning_rate": 0.0001641025641025641,
      "loss": 0.9157,
      "step": 160
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6491233215866394,
      "learning_rate": 0.00016923076923076923,
      "loss": 0.9785,
      "step": 165
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8188832814719226,
      "learning_rate": 0.00017435897435897436,
      "loss": 1.0265,
      "step": 170
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6027544914280824,
      "learning_rate": 0.0001794871794871795,
      "loss": 1.0485,
      "step": 175
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5509987996201512,
      "learning_rate": 0.00018461538461538463,
      "loss": 0.9023,
      "step": 180
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7807717227302549,
      "learning_rate": 0.00018974358974358974,
      "loss": 1.007,
      "step": 185
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5296490906086143,
      "learning_rate": 0.00019487179487179487,
      "loss": 0.8691,
      "step": 190
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5131439344827929,
      "learning_rate": 0.0002,
      "loss": 0.9987,
      "step": 195
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5529964688311281,
      "learning_rate": 0.00019999598996948235,
      "loss": 0.9328,
      "step": 200
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6788038592543275,
      "learning_rate": 0.00019998396019953624,
      "loss": 0.9547,
      "step": 205
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5345212419451733,
      "learning_rate": 0.0001999639116549566,
      "loss": 0.8549,
      "step": 210
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5721084016275837,
      "learning_rate": 0.00019993584594364894,
      "loss": 0.8922,
      "step": 215
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.585457461148956,
      "learning_rate": 0.0001998997653165004,
      "loss": 0.9316,
      "step": 220
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9259431028197681,
      "learning_rate": 0.00019985567266719934,
      "loss": 1.037,
      "step": 225
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6728955086395567,
      "learning_rate": 0.00019980357153200315,
      "loss": 0.972,
      "step": 230
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5152186492594036,
      "learning_rate": 0.00019974346608945466,
      "loss": 0.9563,
      "step": 235
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5113808888103211,
      "learning_rate": 0.00019967536116004698,
      "loss": 0.9597,
      "step": 240
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5746530110327146,
      "learning_rate": 0.00019959926220583713,
      "loss": 0.9332,
      "step": 245
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5634704125813298,
      "learning_rate": 0.00019951517533000764,
      "loss": 0.8902,
      "step": 250
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4214647378593852,
      "learning_rate": 0.00019942310727637724,
      "loss": 0.9396,
      "step": 255
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5261412444526037,
      "learning_rate": 0.00019932306542886009,
      "loss": 0.9587,
      "step": 260
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5138255404441824,
      "learning_rate": 0.00019921505781087334,
      "loss": 0.9934,
      "step": 265
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5970926325926473,
      "learning_rate": 0.00019909909308469398,
      "loss": 0.997,
      "step": 270
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5766575485105531,
      "learning_rate": 0.0001989751805507637,
      "loss": 0.9655,
      "step": 275
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5395097951114133,
      "learning_rate": 0.00019884333014694345,
      "loss": 0.9967,
      "step": 280
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5016351936492325,
      "learning_rate": 0.00019870355244771607,
      "loss": 0.9681,
      "step": 285
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5677320493762555,
      "learning_rate": 0.00019855585866333835,
      "loss": 0.9529,
      "step": 290
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5911468398046378,
      "learning_rate": 0.00019840026063894193,
      "loss": 0.9605,
      "step": 295
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6411041567501414,
      "learning_rate": 0.00019823677085358335,
      "loss": 0.9069,
      "step": 300
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4804628818142298,
      "learning_rate": 0.00019806540241924317,
      "loss": 0.9314,
      "step": 305
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4812821665193446,
      "learning_rate": 0.00019788616907977441,
      "loss": 0.958,
      "step": 310
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4970057545050132,
      "learning_rate": 0.00019769908520980034,
      "loss": 0.9267,
      "step": 315
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5167967545423097,
      "learning_rate": 0.00019750416581356146,
      "loss": 0.9361,
      "step": 320
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6254260263696805,
      "learning_rate": 0.00019730142652371236,
      "loss": 0.8087,
      "step": 325
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.513385773188758,
      "learning_rate": 0.0001970908836000678,
      "loss": 0.9305,
      "step": 330
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5415659067668064,
      "learning_rate": 0.00019687255392829877,
      "loss": 0.9361,
      "step": 335
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.510989151659704,
      "learning_rate": 0.0001966464550185782,
      "loss": 0.9008,
      "step": 340
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5133445959515295,
      "learning_rate": 0.0001964126050041767,
      "loss": 0.8804,
      "step": 345
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5373718187593515,
      "learning_rate": 0.0001961710226400081,
      "loss": 0.8954,
      "step": 350
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5284526685899744,
      "learning_rate": 0.00019592172730112544,
      "loss": 0.9318,
      "step": 355
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5461758806263474,
      "learning_rate": 0.00019566473898116713,
      "loss": 0.9521,
      "step": 360
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4581181882064247,
      "learning_rate": 0.0001954000782907532,
      "loss": 0.9303,
      "step": 365
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4760166258411054,
      "learning_rate": 0.00019512776645583263,
      "loss": 0.9426,
      "step": 370
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5513038240051542,
      "learning_rate": 0.00019484782531598073,
      "loss": 0.9641,
      "step": 375
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.521150098206276,
      "learning_rate": 0.00019456027732264784,
      "loss": 0.998,
      "step": 380
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5242215798378737,
      "learning_rate": 0.00019426514553735848,
      "loss": 0.9679,
      "step": 385
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6519293243594394,
      "learning_rate": 0.00019396245362986197,
      "loss": 0.9287,
      "step": 390
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5450172606812117,
      "learning_rate": 0.00019365222587623405,
      "loss": 0.9075,
      "step": 395
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5309867024495617,
      "learning_rate": 0.00019333448715692995,
      "loss": 0.9354,
      "step": 400
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5923011900956465,
      "learning_rate": 0.00019300926295478884,
      "loss": 0.9876,
      "step": 405
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4957516682866912,
      "learning_rate": 0.0001926765793529902,
      "loss": 0.92,
      "step": 410
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5537126414506671,
      "learning_rate": 0.00019233646303296205,
      "loss": 0.9695,
      "step": 415
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5078410502232953,
      "learning_rate": 0.00019198894127224074,
      "loss": 0.9538,
      "step": 420
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4733618421142402,
      "learning_rate": 0.0001916340419422837,
      "loss": 0.9658,
      "step": 425
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.49209866706455047,
      "learning_rate": 0.00019127179350623372,
      "loss": 0.8984,
      "step": 430
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.49622757816937,
      "learning_rate": 0.0001909022250166365,
      "loss": 0.9138,
      "step": 435
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5184441328755123,
      "learning_rate": 0.00019052536611311046,
      "loss": 0.9442,
      "step": 440
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.48350706635252955,
      "learning_rate": 0.00019014124701996973,
      "loss": 0.9644,
      "step": 445
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5617561220383316,
      "learning_rate": 0.00018974989854379996,
      "loss": 0.9671,
      "step": 450
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5847538823797435,
      "learning_rate": 0.00018935135207098785,
      "loss": 0.9236,
      "step": 455
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5490720274515225,
      "learning_rate": 0.00018894563956520374,
      "loss": 0.9147,
      "step": 460
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.590496607441963,
      "learning_rate": 0.00018853279356483826,
      "loss": 1.0025,
      "step": 465
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.49977088297154965,
      "learning_rate": 0.00018811284718039256,
      "loss": 0.8315,
      "step": 470
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5850512439578014,
      "learning_rate": 0.00018768583409182305,
      "loss": 0.928,
      "step": 475
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5612278839658844,
      "learning_rate": 0.00018725178854584007,
      "loss": 0.9584,
      "step": 480
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.38575791182621044,
      "learning_rate": 0.00018681074535316125,
      "loss": 0.9835,
      "step": 485
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.590124333549399,
      "learning_rate": 0.00018636273988571991,
      "loss": 0.9219,
      "step": 490
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4549791547923319,
      "learning_rate": 0.0001859078080738279,
      "loss": 0.8532,
      "step": 495
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.48478262233698005,
      "learning_rate": 0.00018544598640329432,
      "loss": 0.9546,
      "step": 500
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4311420270366156,
      "learning_rate": 0.00018497731191249894,
      "loss": 0.9918,
      "step": 505
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.43490592457635724,
      "learning_rate": 0.000184501822189422,
      "loss": 0.9557,
      "step": 510
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5752673229536645,
      "learning_rate": 0.00018401955536862948,
      "loss": 0.9633,
      "step": 515
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5947855476321496,
      "learning_rate": 0.0001835305501282148,
      "loss": 1.1131,
      "step": 520
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5271682390355651,
      "learning_rate": 0.00018303484568669667,
      "loss": 0.9367,
      "step": 525
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5563821620964788,
      "learning_rate": 0.00018253248179987388,
      "loss": 0.9222,
      "step": 530
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.583085759157006,
      "learning_rate": 0.0001820234987576368,
      "loss": 0.9531,
      "step": 535
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5844550052753958,
      "learning_rate": 0.00018150793738073602,
      "loss": 0.9207,
      "step": 540
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4675785903498203,
      "learning_rate": 0.00018098583901750867,
      "loss": 0.9309,
      "step": 545
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5246100443420615,
      "learning_rate": 0.00018045724554056214,
      "loss": 0.9167,
      "step": 550
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6280973827882916,
      "learning_rate": 0.0001799221993434159,
      "loss": 0.9903,
      "step": 555
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7087687805202509,
      "learning_rate": 0.00017938074333710157,
      "loss": 0.9453,
      "step": 560
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6297486749925568,
      "learning_rate": 0.00017883292094672128,
      "loss": 0.922,
      "step": 565
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7063313887043356,
      "learning_rate": 0.00017827877610796514,
      "loss": 0.9278,
      "step": 570
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9652058316967166,
      "learning_rate": 0.00017771835326358743,
      "loss": 0.9618,
      "step": 575
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5107454246547803,
      "learning_rate": 0.00017715169735984233,
      "loss": 0.894,
      "step": 580
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5337950114330782,
      "learning_rate": 0.0001765788538428792,
      "loss": 1.0184,
      "step": 585
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7582115364725789,
      "learning_rate": 0.00017599986865509767,
      "loss": 0.9577,
      "step": 590
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4915133885648252,
      "learning_rate": 0.00017541478823146327,
      "loss": 1.0303,
      "step": 595
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.544288667972696,
      "learning_rate": 0.00017482365949578302,
      "loss": 0.9686,
      "step": 600
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5333100547273532,
      "learning_rate": 0.00017422652985694237,
      "loss": 1.0086,
      "step": 605
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.46285430749278295,
      "learning_rate": 0.00017362344720510278,
      "loss": 0.9188,
      "step": 610
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.43835857167409115,
      "learning_rate": 0.00017301445990786102,
      "loss": 0.9918,
      "step": 615
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.45378563171120695,
      "learning_rate": 0.00017239961680637,
      "loss": 0.9322,
      "step": 620
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.488588862520746,
      "learning_rate": 0.0001717789672114218,
      "loss": 1.0627,
      "step": 625
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5664300216234298,
      "learning_rate": 0.0001711525608994927,
      "loss": 0.9394,
      "step": 630
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.5866018044168315,
      "learning_rate": 0.00017052044810875126,
      "loss": 1.0217,
      "step": 635
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6215672900017715,
      "learning_rate": 0.00016988267953502913,
      "loss": 0.9602,
      "step": 640
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.5967382209757992,
      "learning_rate": 0.00016923930632775516,
      "loss": 0.9541,
      "step": 645
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6732298808281975,
      "learning_rate": 0.00016859038008585326,
      "loss": 1.0256,
      "step": 650
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6211638290762932,
      "learning_rate": 0.0001679359528536041,
      "loss": 1.0479,
      "step": 655
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.838914996602064,
      "learning_rate": 0.00016727607711647114,
      "loss": 1.143,
      "step": 660
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0833526800195692,
      "learning_rate": 0.00016661080579689132,
      "loss": 1.0484,
      "step": 665
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6305711211973876,
      "learning_rate": 0.0001659401922500304,
      "loss": 1.0016,
      "step": 670
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6150973517547568,
      "learning_rate": 0.00016526429025950424,
      "loss": 0.963,
      "step": 675
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8249952190210409,
      "learning_rate": 0.00016458315403306502,
      "loss": 1.1581,
      "step": 680
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5898518906871459,
      "learning_rate": 0.0001638968381982538,
      "loss": 0.9366,
      "step": 685
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4474253579572677,
      "learning_rate": 0.0001632053977980194,
      "loss": 0.997,
      "step": 690
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.6754235732624945,
      "learning_rate": 0.000162508888286304,
      "loss": 0.9602,
      "step": 695
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.48044459235384895,
      "learning_rate": 0.00016180736552359553,
      "loss": 1.0053,
      "step": 700
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5766935444157818,
      "learning_rate": 0.00016110088577244773,
      "loss": 1.0136,
      "step": 705
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5571885498100683,
      "learning_rate": 0.00016038950569296785,
      "loss": 0.968,
      "step": 710
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5445736103981007,
      "learning_rate": 0.00015967328233827249,
      "loss": 1.0162,
      "step": 715
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.49130995534885447,
      "learning_rate": 0.00015895227314991178,
      "loss": 0.8883,
      "step": 720
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4915834535077405,
      "learning_rate": 0.00015822653595326275,
      "loss": 0.9338,
      "step": 725
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5300748714740343,
      "learning_rate": 0.00015749612895289152,
      "loss": 0.9305,
      "step": 730
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5372273737913402,
      "learning_rate": 0.00015676111072788527,
      "loss": 0.9349,
      "step": 735
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6245609086190844,
      "learning_rate": 0.00015602154022715435,
      "loss": 0.9948,
      "step": 740
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5790462357457173,
      "learning_rate": 0.0001552774767647043,
      "loss": 0.9573,
      "step": 745
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6447510056373511,
      "learning_rate": 0.0001545289800148789,
      "loss": 0.9978,
      "step": 750
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.605962104404941,
      "learning_rate": 0.0001537761100075744,
      "loss": 1.0899,
      "step": 755
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.5624797807098281,
      "learning_rate": 0.00015301892712342482,
      "loss": 1.0878,
      "step": 760
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.5705928920573031,
      "learning_rate": 0.00015225749208895968,
      "loss": 1.0121,
      "step": 765
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5438913601461941,
      "learning_rate": 0.0001514918659717335,
      "loss": 1.0091,
      "step": 770
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4945101274008925,
      "learning_rate": 0.00015072211017542813,
      "loss": 0.9938,
      "step": 775
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6603965507044199,
      "learning_rate": 0.00014994828643492827,
      "loss": 0.9679,
      "step": 780
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7851609218796439,
      "learning_rate": 0.00014917045681137026,
      "loss": 1.0093,
      "step": 785
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6765038114592608,
      "learning_rate": 0.0001483886836871646,
      "loss": 0.9677,
      "step": 790
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6596813866762753,
      "learning_rate": 0.00014760302976099304,
      "loss": 0.9667,
      "step": 795
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.8030701350894465,
      "learning_rate": 0.00014681355804278001,
      "loss": 0.9391,
      "step": 800
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6524743168094689,
      "learning_rate": 0.00014602033184863913,
      "loss": 0.9544,
      "step": 805
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6486909328335279,
      "learning_rate": 0.00014522341479579533,
      "loss": 1.0861,
      "step": 810
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.5875917335021977,
      "learning_rate": 0.00014442287079748263,
      "loss": 1.0407,
      "step": 815
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7233627090530691,
      "learning_rate": 0.00014361876405781832,
      "loss": 1.0423,
      "step": 820
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0652264291453544,
      "learning_rate": 0.00014281115906665374,
      "loss": 1.1462,
      "step": 825
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.1496976425386072,
      "learning_rate": 0.00014200012059440207,
      "loss": 1.4185,
      "step": 830
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.246140968343057,
      "learning_rate": 0.00014118571368684383,
      "loss": 1.5507,
      "step": 835
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5964924625105266,
      "learning_rate": 0.00014036800365991008,
      "loss": 0.9685,
      "step": 840
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7007609442996293,
      "learning_rate": 0.00013954705609444404,
      "loss": 1.0578,
      "step": 845
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.790069957991019,
      "learning_rate": 0.00013872293683094152,
      "loss": 1.0975,
      "step": 850
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.5667518297820778,
      "learning_rate": 0.00013789571196427055,
      "loss": 1.0221,
      "step": 855
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3519054125833088,
      "learning_rate": 0.00013706544783837022,
      "loss": 1.0336,
      "step": 860
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7653917292323358,
      "learning_rate": 0.00013623221104093025,
      "loss": 1.1741,
      "step": 865
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.6386175596358098,
      "learning_rate": 0.00013539606839805036,
      "loss": 1.046,
      "step": 870
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5919882177796002,
      "learning_rate": 0.00013455708696888085,
      "loss": 1.0329,
      "step": 875
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5016309091464144,
      "learning_rate": 0.00013371533404024438,
      "loss": 1.0209,
      "step": 880
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.6072914705011146,
      "learning_rate": 0.00013287087712123962,
      "loss": 1.0457,
      "step": 885
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5983687797100984,
      "learning_rate": 0.00013202378393782692,
      "loss": 1.0532,
      "step": 890
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9743847221993901,
      "learning_rate": 0.00013117412242739655,
      "loss": 0.9917,
      "step": 895
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.364596992687842,
      "learning_rate": 0.00013032196073332027,
      "loss": 1.2187,
      "step": 900
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6307333206330358,
      "learning_rate": 0.00012946736719948607,
      "loss": 1.0382,
      "step": 905
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.49659246347339786,
      "learning_rate": 0.000128610410364817,
      "loss": 0.971,
      "step": 910
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6693308860411511,
      "learning_rate": 0.00012775115895777417,
      "loss": 1.0628,
      "step": 915
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8040634419449191,
      "learning_rate": 0.00012688968189084493,
      "loss": 0.9951,
      "step": 920
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6757987735677503,
      "learning_rate": 0.00012602604825501587,
      "loss": 1.0313,
      "step": 925
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5473258526701311,
      "learning_rate": 0.00012516032731423165,
      "loss": 1.0723,
      "step": 930
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5590687589778759,
      "learning_rate": 0.00012429258849984014,
      "loss": 1.0267,
      "step": 935
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6243186898722551,
      "learning_rate": 0.00012342290140502388,
      "loss": 1.0273,
      "step": 940
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6468208292730475,
      "learning_rate": 0.00012255133577921868,
      "loss": 1.0404,
      "step": 945
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6163632161848674,
      "learning_rate": 0.0001216779615225197,
      "loss": 0.9746,
      "step": 950
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6530645620182826,
      "learning_rate": 0.00012080284868007541,
      "loss": 1.014,
      "step": 955
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6054201691352697,
      "learning_rate": 0.0001199260674364699,
      "loss": 1.0582,
      "step": 960
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6447850122660467,
      "learning_rate": 0.00011904768811009405,
      "loss": 1.0674,
      "step": 965
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.5198627185486991,
      "learning_rate": 0.00011816778114750593,
      "loss": 1.0467,
      "step": 970
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.5662285572860847,
      "learning_rate": 0.00011728641711778103,
      "loss": 1.0656,
      "step": 975
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6506545237321479,
      "learning_rate": 0.00011640366670685248,
      "loss": 0.9647,
      "step": 980
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8042796459925735,
      "learning_rate": 0.00011551960071184195,
      "loss": 1.1709,
      "step": 985
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.687108654869035,
      "learning_rate": 0.00011463429003538196,
      "loss": 1.0447,
      "step": 990
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6424656897816762,
      "learning_rate": 0.000113747805679929,
      "loss": 1.1066,
      "step": 995
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.895580663791374,
      "learning_rate": 0.00011286021874206952,
      "loss": 0.9909,
      "step": 1000
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7973376179813747,
      "learning_rate": 0.00011197160040681762,
      "loss": 1.2025,
      "step": 1005
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7168908878647755,
      "learning_rate": 0.0001110820219419062,
      "loss": 1.0947,
      "step": 1010
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.6538109383723899,
      "learning_rate": 0.0001101915546920711,
      "loss": 1.0962,
      "step": 1015
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7189531772805177,
      "learning_rate": 0.00010930027007332923,
      "loss": 1.0181,
      "step": 1020
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8326731896215877,
      "learning_rate": 0.00010840823956725103,
      "loss": 1.1334,
      "step": 1025
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8473033240689335,
      "learning_rate": 0.00010751553471522757,
      "loss": 1.1032,
      "step": 1030
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.7557996349026346,
      "learning_rate": 0.00010662222711273279,
      "loss": 1.0761,
      "step": 1035
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6513659950736571,
      "learning_rate": 0.00010572838840358168,
      "loss": 1.0373,
      "step": 1040
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7423562616150807,
      "learning_rate": 0.00010483409027418425,
      "loss": 1.0421,
      "step": 1045
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7158964464697141,
      "learning_rate": 0.00010393940444779635,
      "loss": 1.0541,
      "step": 1050
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7996387165453449,
      "learning_rate": 0.00010304440267876727,
      "loss": 1.2155,
      "step": 1055
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8216668434955695,
      "learning_rate": 0.00010214915674678523,
      "loss": 1.2363,
      "step": 1060
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7120501980824897,
      "learning_rate": 0.00010125373845112034,
      "loss": 1.1574,
      "step": 1065
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.913854792920572,
      "learning_rate": 0.00010035821960486643,
      "loss": 1.1805,
      "step": 1070
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8468128906491331,
      "learning_rate": 9.946267202918157e-05,
      "loss": 1.0998,
      "step": 1075
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.0456818160084107,
      "learning_rate": 9.856716754752796e-05,
      "loss": 1.1047,
      "step": 1080
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8350505078035804,
      "learning_rate": 9.767177797991155e-05,
      "loss": 1.1321,
      "step": 1085
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6783048186052755,
      "learning_rate": 9.677657513712221e-05,
      "loss": 1.061,
      "step": 1090
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7882949718690779,
      "learning_rate": 9.588163081497427e-05,
      "loss": 1.1711,
      "step": 1095
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6806692657836286,
      "learning_rate": 9.498701678854865e-05,
      "loss": 1.113,
      "step": 1100
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9030791057296585,
      "learning_rate": 9.409280480643628e-05,
      "loss": 1.1979,
      "step": 1105
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7521251870330353,
      "learning_rate": 9.319906658498389e-05,
      "loss": 1.1946,
      "step": 1110
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8157002643577485,
      "learning_rate": 9.230587380254237e-05,
      "loss": 1.1608,
      "step": 1115
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7465778977451432,
      "learning_rate": 9.141329809371803e-05,
      "loss": 1.1465,
      "step": 1120
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7624443580773561,
      "learning_rate": 9.052141104362748e-05,
      "loss": 1.1194,
      "step": 1125
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9071687976108164,
      "learning_rate": 8.963028418215653e-05,
      "loss": 1.282,
      "step": 1130
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9225211232736652,
      "learning_rate": 8.873998897822336e-05,
      "loss": 1.2844,
      "step": 1135
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.0150112698716645,
      "learning_rate": 8.785059683404672e-05,
      "loss": 1.5796,
      "step": 1140
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9261706177590988,
      "learning_rate": 8.696217907941941e-05,
      "loss": 1.1975,
      "step": 1145
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.938094505733298,
      "learning_rate": 8.607480696598762e-05,
      "loss": 1.0538,
      "step": 1150
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.0678420030431226,
      "learning_rate": 8.518855166153644e-05,
      "loss": 1.2,
      "step": 1155
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9566507672068523,
      "learning_rate": 8.43034842442822e-05,
      "loss": 1.1481,
      "step": 1160
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8514422421050399,
      "learning_rate": 8.341967569717202e-05,
      "loss": 1.0757,
      "step": 1165
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9002858843043067,
      "learning_rate": 8.253719690219079e-05,
      "loss": 1.1066,
      "step": 1170
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9185606255836889,
      "learning_rate": 8.165611863467644e-05,
      "loss": 1.1159,
      "step": 1175
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8498737163717386,
      "learning_rate": 8.077651155764387e-05,
      "loss": 1.0841,
      "step": 1180
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7818771458613086,
      "learning_rate": 7.98984462161175e-05,
      "loss": 1.1569,
      "step": 1185
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7994636695845656,
      "learning_rate": 7.902199303147363e-05,
      "loss": 1.1612,
      "step": 1190
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8725883660970624,
      "learning_rate": 7.814722229579264e-05,
      "loss": 1.142,
      "step": 1195
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0166557786134702,
      "learning_rate": 7.727420416622144e-05,
      "loss": 1.1821,
      "step": 1200
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.725978165251939,
      "learning_rate": 7.640300865934687e-05,
      "loss": 1.1213,
      "step": 1205
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0536229921804432,
      "learning_rate": 7.553370564558032e-05,
      "loss": 1.2072,
      "step": 1210
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8560456236796121,
      "learning_rate": 7.46663648435541e-05,
      "loss": 1.1977,
      "step": 1215
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.1384316732644513,
      "learning_rate": 7.380105581452987e-05,
      "loss": 1.1746,
      "step": 1220
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9337738140433541,
      "learning_rate": 7.293784795681994e-05,
      "loss": 1.2707,
      "step": 1225
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9525328413620964,
      "learning_rate": 7.207681050022132e-05,
      "loss": 1.1199,
      "step": 1230
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8288888174056264,
      "learning_rate": 7.121801250046363e-05,
      "loss": 1.2899,
      "step": 1235
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7121188159596955,
      "learning_rate": 7.036152283367056e-05,
      "loss": 1.1126,
      "step": 1240
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9439808750435214,
      "learning_rate": 6.950741019083617e-05,
      "loss": 1.2199,
      "step": 1245
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9081442463995935,
      "learning_rate": 6.865574307231575e-05,
      "loss": 1.1986,
      "step": 1250
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.2074522816329538,
      "learning_rate": 6.780658978233199e-05,
      "loss": 1.2642,
      "step": 1255
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.1591818963360307,
      "learning_rate": 6.696001842349702e-05,
      "loss": 1.2396,
      "step": 1260
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4378229246332925,
      "learning_rate": 6.611609689135056e-05,
      "loss": 1.3653,
      "step": 1265
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8248227152334574,
      "learning_rate": 6.527489286891459e-05,
      "loss": 1.0621,
      "step": 1270
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0010033999308974,
      "learning_rate": 6.443647382126509e-05,
      "loss": 1.2177,
      "step": 1275
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.0397952900383065,
      "learning_rate": 6.360090699012145e-05,
      "loss": 1.1591,
      "step": 1280
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8820591270051052,
      "learning_rate": 6.27682593884535e-05,
      "loss": 1.0823,
      "step": 1285
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8336491605072446,
      "learning_rate": 6.193859779510712e-05,
      "loss": 1.3229,
      "step": 1290
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8995355585429261,
      "learning_rate": 6.111198874944845e-05,
      "loss": 1.3138,
      "step": 1295
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9420446740151357,
      "learning_rate": 6.0288498546027536e-05,
      "loss": 1.2631,
      "step": 1300
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9859796536689945,
      "learning_rate": 5.946819322926127e-05,
      "loss": 1.1125,
      "step": 1305
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.1268253558322285,
      "learning_rate": 5.865113858813673e-05,
      "loss": 1.1855,
      "step": 1310
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.999493714287093,
      "learning_rate": 5.783740015093484e-05,
      "loss": 1.2303,
      "step": 1315
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.975237281525299,
      "learning_rate": 5.702704317997492e-05,
      "loss": 1.1431,
      "step": 1320
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9877406612582614,
      "learning_rate": 5.6220132666380635e-05,
      "loss": 1.0927,
      "step": 1325
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.0373246609079592,
      "learning_rate": 5.541673332486773e-05,
      "loss": 1.2512,
      "step": 1330
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.834253734571841,
      "learning_rate": 5.4616909588553674e-05,
      "loss": 1.3298,
      "step": 1335
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.045955594093587,
      "learning_rate": 5.3820725603790346e-05,
      "loss": 1.3265,
      "step": 1340
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9496400711428868,
      "learning_rate": 5.30282452250193e-05,
      "loss": 1.2492,
      "step": 1345
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9475902754002916,
      "learning_rate": 5.223953200965055e-05,
      "loss": 1.1695,
      "step": 1350
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.1065380759578969,
      "learning_rate": 5.145464921296537e-05,
      "loss": 1.3468,
      "step": 1355
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.0245170475484517,
      "learning_rate": 5.067365978304315e-05,
      "loss": 1.1842,
      "step": 1360
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9825535871117517,
      "learning_rate": 4.9896626355712805e-05,
      "loss": 1.1858,
      "step": 1365
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9745163728190328,
      "learning_rate": 4.912361124952948e-05,
      "loss": 1.2226,
      "step": 1370
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9499691689210246,
      "learning_rate": 4.835467646077656e-05,
      "loss": 1.1724,
      "step": 1375
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8591209639811482,
      "learning_rate": 4.7589883658493296e-05,
      "loss": 1.194,
      "step": 1380
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.0553349866626618,
      "learning_rate": 4.682929417952939e-05,
      "loss": 1.172,
      "step": 1385
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9861698028980803,
      "learning_rate": 4.6072969023625165e-05,
      "loss": 1.2485,
      "step": 1390
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.133626581213844,
      "learning_rate": 4.532096884851978e-05,
      "loss": 1.2671,
      "step": 1395
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9077262808519728,
      "learning_rate": 4.457335396508631e-05,
      "loss": 1.2497,
      "step": 1400
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9776537099776628,
      "learning_rate": 4.383018433249464e-05,
      "loss": 1.1841,
      "step": 1405
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9744593516723105,
      "learning_rate": 4.309151955340297e-05,
      "loss": 1.1423,
      "step": 1410
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9696343349780663,
      "learning_rate": 4.2357418869177354e-05,
      "loss": 1.2703,
      "step": 1415
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.1281479230206592,
      "learning_rate": 4.162794115514078e-05,
      "loss": 1.3003,
      "step": 1420
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.07249353342714,
      "learning_rate": 4.0903144915851174e-05,
      "loss": 1.099,
      "step": 1425
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.1909749243720829,
      "learning_rate": 4.018308828040924e-05,
      "loss": 1.1693,
      "step": 1430
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9745679726030496,
      "learning_rate": 3.946782899779667e-05,
      "loss": 1.2225,
      "step": 1435
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.3838207416992974,
      "learning_rate": 3.875742443224451e-05,
      "loss": 1.3124,
      "step": 1440
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.1624066489107725,
      "learning_rate": 3.805193155863247e-05,
      "loss": 1.3336,
      "step": 1445
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.1693228345737117,
      "learning_rate": 3.7351406957919636e-05,
      "loss": 1.888,
      "step": 1450
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.113237072027005,
      "learning_rate": 3.665590681260658e-05,
      "loss": 1.4847,
      "step": 1455
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.027432684993533,
      "learning_rate": 3.59654869022294e-05,
      "loss": 1.2917,
      "step": 1460
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9015272534258235,
      "learning_rate": 3.5280202598886324e-05,
      "loss": 1.2668,
      "step": 1465
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9989589902337217,
      "learning_rate": 3.4600108862796796e-05,
      "loss": 1.1778,
      "step": 1470
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9135433747040247,
      "learning_rate": 3.392526023789349e-05,
      "loss": 1.2165,
      "step": 1475
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9037720655106274,
      "learning_rate": 3.325571084744803e-05,
      "loss": 1.1215,
      "step": 1480
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.0089528495488433,
      "learning_rate": 3.259151438973024e-05,
      "loss": 1.2771,
      "step": 1485
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8578907836344709,
      "learning_rate": 3.1932724133701344e-05,
      "loss": 1.1198,
      "step": 1490
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9910666197470511,
      "learning_rate": 3.1279392914742046e-05,
      "loss": 1.2362,
      "step": 1495
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.906720175898158,
      "learning_rate": 3.06315731304148e-05,
      "loss": 1.2752,
      "step": 1500
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9727961797782265,
      "learning_rate": 2.998931673626175e-05,
      "loss": 1.1237,
      "step": 1505
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9952282526211972,
      "learning_rate": 2.935267524163774e-05,
      "loss": 1.3033,
      "step": 1510
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.0510019669013457,
      "learning_rate": 2.872169970557913e-05,
      "loss": 1.2973,
      "step": 1515
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.0691274991403132,
      "learning_rate": 2.8096440732709083e-05,
      "loss": 1.3736,
      "step": 1520
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.226603673595138,
      "learning_rate": 2.7476948469178887e-05,
      "loss": 1.2731,
      "step": 1525
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.1794742264910503,
      "learning_rate": 2.6863272598646106e-05,
      "loss": 1.3517,
      "step": 1530
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.525402988837181,
      "learning_rate": 2.625546233829016e-05,
      "loss": 1.6903,
      "step": 1535
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5714066559056612,
      "learning_rate": 2.5653566434864928e-05,
      "loss": 2.1873,
      "step": 1540
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.4716145750540761,
      "learning_rate": 2.5057633160789184e-05,
      "loss": 2.1076,
      "step": 1545
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.2754222415393928,
      "learning_rate": 2.446771031027527e-05,
      "loss": 2.6349,
      "step": 1550
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.791464425253221,
      "learning_rate": 2.3883845195495878e-05,
      "loss": 3.052,
      "step": 1555
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.7198825982139114,
      "learning_rate": 2.330608464278953e-05,
      "loss": 1.7635,
      "step": 1560
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.635207645118843,
      "learning_rate": 2.273447498890521e-05,
      "loss": 1.5937,
      "step": 1565
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.1772159776006166,
      "learning_rate": 2.2169062077286075e-05,
      "loss": 1.441,
      "step": 1570
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.201745038274024,
      "learning_rate": 2.1609891254392678e-05,
      "loss": 1.4796,
      "step": 1575
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.5021231016124097,
      "learning_rate": 2.1057007366066373e-05,
      "loss": 1.6289,
      "step": 1580
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.323524751892768,
      "learning_rate": 2.0510454753932395e-05,
      "loss": 1.5372,
      "step": 1585
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.1548514419384561,
      "learning_rate": 1.9970277251843862e-05,
      "loss": 1.6231,
      "step": 1590
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.181699286829025,
      "learning_rate": 1.9436518182366158e-05,
      "loss": 1.63,
      "step": 1595
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.1461725693025797,
      "learning_rate": 1.8909220353302392e-05,
      "loss": 1.5822,
      "step": 1600
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.341801520738588,
      "learning_rate": 1.838842605426031e-05,
      "loss": 1.4874,
      "step": 1605
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.0659249058996556,
      "learning_rate": 1.7874177053260598e-05,
      "loss": 1.4152,
      "step": 1610
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.4830623433840733,
      "learning_rate": 1.736651459338695e-05,
      "loss": 1.4297,
      "step": 1615
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.2632350719823107,
      "learning_rate": 1.6865479389478545e-05,
      "loss": 1.5458,
      "step": 1620
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.0804391841122525,
      "learning_rate": 1.6371111624864543e-05,
      "loss": 1.4844,
      "step": 1625
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.4354171503918984,
      "learning_rate": 1.5883450948141377e-05,
      "loss": 1.5549,
      "step": 1630
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.452642601949355,
      "learning_rate": 1.540253646999299e-05,
      "loss": 1.6123,
      "step": 1635
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.2836975505765336,
      "learning_rate": 1.4928406760054059e-05,
      "loss": 1.4825,
      "step": 1640
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3754719500363763,
      "learning_rate": 1.4461099843816684e-05,
      "loss": 1.496,
      "step": 1645
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.0399348065075504,
      "learning_rate": 1.4000653199580782e-05,
      "loss": 1.4858,
      "step": 1650
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.247304075740197,
      "learning_rate": 1.3547103755448287e-05,
      "loss": 1.4482,
      "step": 1655
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.0818653623296255,
      "learning_rate": 1.3100487886361379e-05,
      "loss": 1.4938,
      "step": 1660
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.027111694463334,
      "learning_rate": 1.266084141118542e-05,
      "loss": 1.3626,
      "step": 1665
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.062618725490322,
      "learning_rate": 1.2228199589835999e-05,
      "loss": 1.42,
      "step": 1670
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.231558384560607,
      "learning_rate": 1.1802597120451286e-05,
      "loss": 1.4311,
      "step": 1675
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3160612712927446,
      "learning_rate": 1.1384068136609105e-05,
      "loss": 1.4488,
      "step": 1680
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.1451058338292823,
      "learning_rate": 1.0972646204589377e-05,
      "loss": 1.3587,
      "step": 1685
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3622321749039823,
      "learning_rate": 1.0568364320682178e-05,
      "loss": 1.4835,
      "step": 1690
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3157150127080717,
      "learning_rate": 1.0171254908541372e-05,
      "loss": 1.3429,
      "step": 1695
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3168426340030872,
      "learning_rate": 9.781349816584162e-06,
      "loss": 1.3686,
      "step": 1700
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.212283535175554,
      "learning_rate": 9.398680315436903e-06,
      "loss": 1.3853,
      "step": 1705
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.3377304564373287,
      "learning_rate": 9.023277095427173e-06,
      "loss": 1.3042,
      "step": 1710
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.0133923360246824,
      "learning_rate": 8.655170264122303e-06,
      "loss": 1.4156,
      "step": 1715
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.0142755578646634,
      "learning_rate": 8.294389343914899e-06,
      "loss": 1.3661,
      "step": 1720
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.058349976360042,
      "learning_rate": 7.940963269654922e-06,
      "loss": 1.4371,
      "step": 1725
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.9781751990105597,
      "learning_rate": 7.594920386329252e-06,
      "loss": 1.3693,
      "step": 1730
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.0141996245377212,
      "learning_rate": 7.256288446788362e-06,
      "loss": 1.2786,
      "step": 1735
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.3054233686308656,
      "learning_rate": 6.925094609520455e-06,
      "loss": 1.413,
      "step": 1740
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.0692031855069113,
      "learning_rate": 6.601365436473439e-06,
      "loss": 1.3203,
      "step": 1745
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.231083792558415,
      "learning_rate": 6.2851268909245865e-06,
      "loss": 1.3064,
      "step": 1750
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.0994222164567315,
      "learning_rate": 5.976404335398256e-06,
      "loss": 1.3936,
      "step": 1755
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.9841529223340063,
      "learning_rate": 5.675222529631841e-06,
      "loss": 1.3629,
      "step": 1760
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.143186215443706,
      "learning_rate": 5.381605628590003e-06,
      "loss": 1.3115,
      "step": 1765
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.262725732625003,
      "learning_rate": 5.095577180527378e-06,
      "loss": 1.3718,
      "step": 1770
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.0907826292525469,
      "learning_rate": 4.817160125100106e-06,
      "loss": 1.4595,
      "step": 1775
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.2322750149837634,
      "learning_rate": 4.546376791525975e-06,
      "loss": 1.3603,
      "step": 1780
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.0939278850532899,
      "learning_rate": 4.2832488967935795e-06,
      "loss": 1.3568,
      "step": 1785
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.1448889400301656,
      "learning_rate": 4.02779754392072e-06,
      "loss": 1.3641,
      "step": 1790
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.266880119709807,
      "learning_rate": 3.780043220261764e-06,
      "loss": 1.4117,
      "step": 1795
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.3173572982116377,
      "learning_rate": 3.540005795864709e-06,
      "loss": 1.2435,
      "step": 1800
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.2174226024217492,
      "learning_rate": 3.3077045218775192e-06,
      "loss": 1.3869,
      "step": 1805
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.1558882314841081,
      "learning_rate": 3.0831580290041184e-06,
      "loss": 1.4034,
      "step": 1810
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.9757229002365276,
      "learning_rate": 2.8663843260103074e-06,
      "loss": 1.3497,
      "step": 1815
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.3067724280685595,
      "learning_rate": 2.6574007982793857e-06,
      "loss": 1.352,
      "step": 1820
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.135393851086464,
      "learning_rate": 2.456224206417812e-06,
      "loss": 1.2794,
      "step": 1825
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.078679103611306,
      "learning_rate": 2.262870684911045e-06,
      "loss": 1.3759,
      "step": 1830
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.1132669384116787,
      "learning_rate": 2.0773557408295343e-06,
      "loss": 1.3527,
      "step": 1835
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.0332287216058291,
      "learning_rate": 1.8996942525850047e-06,
      "loss": 1.3262,
      "step": 1840
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.1352071402113169,
      "learning_rate": 1.7299004687372665e-06,
      "loss": 1.3562,
      "step": 1845
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.2389706855715208,
      "learning_rate": 1.5679880068514174e-06,
      "loss": 1.2931,
      "step": 1850
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.3021234746737105,
      "learning_rate": 1.4139698524057165e-06,
      "loss": 1.4022,
      "step": 1855
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.2646020975420837,
      "learning_rate": 1.2678583577501624e-06,
      "loss": 1.3853,
      "step": 1860
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.240987442877643,
      "learning_rate": 1.1296652411158182e-06,
      "loss": 1.3658,
      "step": 1865
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.1726859315746156,
      "learning_rate": 9.994015856749527e-07,
      "loss": 1.3202,
      "step": 1870
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.1089411210805613,
      "learning_rate": 8.770778386522627e-07,
      "loss": 1.265,
      "step": 1875
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.2285374452780913,
      "learning_rate": 7.627038104869199e-07,
      "loss": 1.3509,
      "step": 1880
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.0602256257040228,
      "learning_rate": 6.562886740457797e-07,
      "loss": 1.3321,
      "step": 1885
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.1838941867679529,
      "learning_rate": 5.578409638877457e-07,
      "loss": 1.2253,
      "step": 1890
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.2088615096234103,
      "learning_rate": 4.6736857557925227e-07,
      "loss": 1.3172,
      "step": 1895
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.1429539879303683,
      "learning_rate": 3.8487876506106966e-07,
      "loss": 1.4324,
      "step": 1900
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.0563107580222184,
      "learning_rate": 3.1037814806634815e-07,
      "loss": 1.304,
      "step": 1905
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.0423096141514725,
      "learning_rate": 2.43872699590042e-07,
      "loss": 1.267,
      "step": 1910
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.1241174227772588,
      "learning_rate": 1.8536775340970425e-07,
      "loss": 1.437,
      "step": 1915
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.2829849821049102,
      "learning_rate": 1.348680016577397e-07,
      "loss": 1.4342,
      "step": 1920
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.2052610335123075,
      "learning_rate": 9.237749444505062e-08,
      "loss": 1.3375,
      "step": 1925
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.125000345854297,
      "learning_rate": 5.7899639536251883e-08,
      "loss": 1.3748,
      "step": 1930
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.2144039325069627,
      "learning_rate": 3.143720207635648e-08,
      "loss": 1.3634,
      "step": 1935
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3857093808309557,
      "learning_rate": 1.299230436898613e-08,
      "loss": 1.3961,
      "step": 1940
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.226862301795748,
      "learning_rate": 2.566425706218567e-09,
      "loss": 1.2739,
      "step": 1945
    },
    {
      "epoch": 1.0,
      "eval_loss": NaN,
      "eval_runtime": 1976.8002,
      "eval_samples_per_second": 3.507,
      "eval_steps_per_second": 0.877,
      "step": 1949
    },
    {
      "epoch": 1.0,
      "step": 1949,
      "total_flos": 1.2135540828143616e+16,
      "train_loss": 1.1757926806356431,
      "train_runtime": 18036.5775,
      "train_samples_per_second": 3.457,
      "train_steps_per_second": 0.108
    }
  ],
  "logging_steps": 5,
  "max_steps": 1949,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 1.2135540828143616e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}