{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6517766733784146,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.9869178533554077,
      "learning_rate": 9.99997860104108e-06,
      "loss": 1.4113,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7985073924064636,
      "learning_rate": 9.999914404347487e-06,
      "loss": 1.3054,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5533589720726013,
      "learning_rate": 9.999807410468713e-06,
      "loss": 1.2698,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.527658224105835,
      "learning_rate": 9.999657620320587e-06,
      "loss": 1.1817,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5116757750511169,
      "learning_rate": 9.999465035185248e-06,
      "loss": 1.1681,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.40034493803977966,
      "learning_rate": 9.999229656711143e-06,
      "loss": 1.1709,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.41553106904029846,
      "learning_rate": 9.998951486913015e-06,
      "loss": 1.133,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.38187628984451294,
      "learning_rate": 9.998630528171882e-06,
      "loss": 1.0915,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.39334940910339355,
      "learning_rate": 9.998266783235018e-06,
      "loss": 1.1154,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.38754093647003174,
      "learning_rate": 9.997860255215926e-06,
      "loss": 1.0616,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3934289216995239,
      "learning_rate": 9.997410947594317e-06,
      "loss": 1.1031,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.44751420617103577,
      "learning_rate": 9.996918864216081e-06,
      "loss": 1.0547,
      "step": 120
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.4995039701461792,
      "learning_rate": 9.99638400929324e-06,
      "loss": 1.0634,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.519430935382843,
      "learning_rate": 9.995806387403935e-06,
      "loss": 1.0366,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4063739478588104,
      "learning_rate": 9.995186003492366e-06,
      "loss": 1.0329,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.47834908962249756,
      "learning_rate": 9.994522862868763e-06,
      "loss": 1.0168,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4666057825088501,
      "learning_rate": 9.993816971209332e-06,
      "loss": 1.044,
      "step": 170
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5252623558044434,
      "learning_rate": 9.99306833455621e-06,
      "loss": 1.0311,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.41999587416648865,
      "learning_rate": 9.992276959317419e-06,
      "loss": 1.0273,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4322344660758972,
      "learning_rate": 9.9914428522668e-06,
      "loss": 1.0174,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.44230663776397705,
      "learning_rate": 9.99056602054396e-06,
      "loss": 0.9849,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.48284780979156494,
      "learning_rate": 9.989646471654216e-06,
      "loss": 1.0219,
      "step": 220
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.478462815284729,
      "learning_rate": 9.988684213468521e-06,
      "loss": 1.01,
      "step": 230
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4927475154399872,
      "learning_rate": 9.987679254223405e-06,
      "loss": 1.0199,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4281039237976074,
      "learning_rate": 9.986631602520904e-06,
      "loss": 1.0072,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5018512010574341,
      "learning_rate": 9.985541267328479e-06,
      "loss": 0.9704,
      "step": 260
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4972495138645172,
      "learning_rate": 9.98440825797894e-06,
      "loss": 0.9858,
      "step": 270
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5815178751945496,
      "learning_rate": 9.98323258417038e-06,
      "loss": 0.9752,
      "step": 280
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6393250823020935,
      "learning_rate": 9.982014255966078e-06,
      "loss": 1.0034,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5562496781349182,
      "learning_rate": 9.980753283794414e-06,
      "loss": 1.0004,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.48535385727882385,
      "learning_rate": 9.979449678448785e-06,
      "loss": 0.995,
      "step": 310
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6010926365852356,
      "learning_rate": 9.978103451087511e-06,
      "loss": 0.9669,
      "step": 320
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5188761949539185,
      "learning_rate": 9.976714613233736e-06,
      "loss": 0.978,
      "step": 330
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7140072584152222,
      "learning_rate": 9.975283176775332e-06,
      "loss": 0.9778,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5040972232818604,
      "learning_rate": 9.973809153964803e-06,
      "loss": 0.984,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.557113528251648,
      "learning_rate": 9.97229255741917e-06,
      "loss": 0.9871,
      "step": 360
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6509271860122681,
      "learning_rate": 9.970733400119865e-06,
      "loss": 0.9797,
      "step": 370
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6498264670372009,
      "learning_rate": 9.969131695412629e-06,
      "loss": 0.9665,
      "step": 380
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6095420718193054,
      "learning_rate": 9.967487457007382e-06,
      "loss": 0.9589,
      "step": 390
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5704444646835327,
      "learning_rate": 9.965800698978126e-06,
      "loss": 0.9666,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5877268314361572,
      "learning_rate": 9.964071435762802e-06,
      "loss": 0.9999,
      "step": 410
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5909533500671387,
      "learning_rate": 9.962299682163185e-06,
      "loss": 0.9381,
      "step": 420
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5712386965751648,
      "learning_rate": 9.96048545334475e-06,
      "loss": 0.972,
      "step": 430
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.737307608127594,
      "learning_rate": 9.95862876483654e-06,
      "loss": 0.9754,
      "step": 440
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7787168025970459,
      "learning_rate": 9.956729632531032e-06,
      "loss": 1.0197,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5988175868988037,
      "learning_rate": 9.954788072684011e-06,
      "loss": 0.9596,
      "step": 460
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.981139600276947,
      "learning_rate": 9.952804101914418e-06,
      "loss": 0.9876,
      "step": 470
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6938815712928772,
      "learning_rate": 9.95077773720422e-06,
      "loss": 0.9533,
      "step": 480
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.671062707901001,
      "learning_rate": 9.948708995898251e-06,
      "loss": 0.9812,
      "step": 490
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6615872979164124,
      "learning_rate": 9.94659789570408e-06,
      "loss": 0.9522,
      "step": 500
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5831172466278076,
      "learning_rate": 9.94444445469184e-06,
      "loss": 0.9454,
      "step": 510
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6584324836730957,
      "learning_rate": 9.942248691294092e-06,
      "loss": 0.951,
      "step": 520
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6748599410057068,
      "learning_rate": 9.940010624305658e-06,
      "loss": 0.927,
      "step": 530
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5425313115119934,
      "learning_rate": 9.937730272883458e-06,
      "loss": 0.9542,
      "step": 540
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.551030158996582,
      "learning_rate": 9.93540765654635e-06,
      "loss": 0.933,
      "step": 550
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.743243396282196,
      "learning_rate": 9.933042795174964e-06,
      "loss": 0.9573,
      "step": 560
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6725818514823914,
      "learning_rate": 9.930635709011524e-06,
      "loss": 0.9745,
      "step": 570
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.743486225605011,
      "learning_rate": 9.928186418659693e-06,
      "loss": 0.9705,
      "step": 580
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6531779170036316,
      "learning_rate": 9.925694945084369e-06,
      "loss": 0.9526,
      "step": 590
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7471463680267334,
      "learning_rate": 9.923161309611534e-06,
      "loss": 0.9718,
      "step": 600
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6551310420036316,
      "learning_rate": 9.920585533928049e-06,
      "loss": 0.9738,
      "step": 610
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.835099458694458,
      "learning_rate": 9.917967640081482e-06,
      "loss": 0.9559,
      "step": 620
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6788685321807861,
      "learning_rate": 9.915307650479915e-06,
      "loss": 0.9448,
      "step": 630
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.640802800655365,
      "learning_rate": 9.912605587891748e-06,
      "loss": 0.9632,
      "step": 640
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7521352767944336,
      "learning_rate": 9.909861475445517e-06,
      "loss": 0.9947,
      "step": 650
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6148037314414978,
      "learning_rate": 9.907075336629679e-06,
      "loss": 0.8992,
      "step": 660
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7000744938850403,
      "learning_rate": 9.90424719529242e-06,
      "loss": 0.9978,
      "step": 670
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6232642531394958,
      "learning_rate": 9.901377075641457e-06,
      "loss": 0.9916,
      "step": 680
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6425381898880005,
      "learning_rate": 9.898465002243813e-06,
      "loss": 0.9478,
      "step": 690
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6344175934791565,
      "learning_rate": 9.89551100002563e-06,
      "loss": 0.9749,
      "step": 700
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.677696704864502,
      "learning_rate": 9.892515094271931e-06,
      "loss": 0.9815,
      "step": 710
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6174888610839844,
      "learning_rate": 9.889477310626426e-06,
      "loss": 0.9444,
      "step": 720
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7845975756645203,
      "learning_rate": 9.886397675091275e-06,
      "loss": 0.9751,
      "step": 730
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.844460666179657,
      "learning_rate": 9.883276214026877e-06,
      "loss": 0.9412,
      "step": 740
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6752843260765076,
      "learning_rate": 9.880112954151639e-06,
      "loss": 0.95,
      "step": 750
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7257810831069946,
      "learning_rate": 9.876907922541748e-06,
      "loss": 0.9468,
      "step": 760
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6597475409507751,
      "learning_rate": 9.87366114663094e-06,
      "loss": 0.9219,
      "step": 770
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6881552934646606,
      "learning_rate": 9.870372654210265e-06,
      "loss": 0.9571,
      "step": 780
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8410021662712097,
      "learning_rate": 9.867042473427848e-06,
      "loss": 0.9429,
      "step": 790
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6908501982688904,
      "learning_rate": 9.863670632788652e-06,
      "loss": 0.9618,
      "step": 800
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6102294921875,
      "learning_rate": 9.860257161154224e-06,
      "loss": 0.9539,
      "step": 810
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8168625831604004,
      "learning_rate": 9.856802087742463e-06,
      "loss": 0.9384,
      "step": 820
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6453863978385925,
      "learning_rate": 9.85330544212736e-06,
      "loss": 0.9341,
      "step": 830
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6028039455413818,
      "learning_rate": 9.849767254238741e-06,
      "loss": 0.95,
      "step": 840
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6181658506393433,
      "learning_rate": 9.846187554362026e-06,
      "loss": 0.9184,
      "step": 850
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7346929907798767,
      "learning_rate": 9.842566373137949e-06,
      "loss": 0.9714,
      "step": 860
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6204361915588379,
      "learning_rate": 9.83890374156232e-06,
      "loss": 0.9161,
      "step": 870
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7579419016838074,
      "learning_rate": 9.835199690985737e-06,
      "loss": 0.943,
      "step": 880
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6947230100631714,
      "learning_rate": 9.831454253113328e-06,
      "loss": 0.9092,
      "step": 890
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.887786865234375,
      "learning_rate": 9.827667460004487e-06,
      "loss": 0.9149,
      "step": 900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7788590788841248,
      "learning_rate": 9.823839344072582e-06,
      "loss": 0.9187,
      "step": 910
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6010196208953857,
      "learning_rate": 9.81996993808469e-06,
      "loss": 0.9445,
      "step": 920
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6567085385322571,
      "learning_rate": 9.816059275161318e-06,
      "loss": 0.9492,
      "step": 930
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8992276191711426,
      "learning_rate": 9.812107388776113e-06,
      "loss": 0.8785,
      "step": 940
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6781167387962341,
      "learning_rate": 9.808114312755574e-06,
      "loss": 0.902,
      "step": 950
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6629174947738647,
      "learning_rate": 9.804080081278768e-06,
      "loss": 0.9416,
      "step": 960
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8017804026603699,
      "learning_rate": 9.800004728877042e-06,
      "loss": 0.9273,
      "step": 970
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6197686195373535,
      "learning_rate": 9.795888290433709e-06,
      "loss": 0.9335,
      "step": 980
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7496291995048523,
      "learning_rate": 9.791730801183772e-06,
      "loss": 0.9036,
      "step": 990
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6641804575920105,
      "learning_rate": 9.787532296713605e-06,
      "loss": 0.9768,
      "step": 1000
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7570294737815857,
      "learning_rate": 9.78329281296066e-06,
      "loss": 0.9162,
      "step": 1010
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7247979044914246,
      "learning_rate": 9.779012386213151e-06,
      "loss": 0.9261,
      "step": 1020
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7128037214279175,
      "learning_rate": 9.77469105310975e-06,
      "loss": 0.9403,
      "step": 1030
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7633296847343445,
      "learning_rate": 9.770328850639268e-06,
      "loss": 0.9184,
      "step": 1040
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7282195687294006,
      "learning_rate": 9.76592581614034e-06,
      "loss": 0.9144,
      "step": 1050
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7485966086387634,
      "learning_rate": 9.761481987301111e-06,
      "loss": 0.9307,
      "step": 1060
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7050462961196899,
      "learning_rate": 9.756997402158904e-06,
      "loss": 0.9269,
      "step": 1070
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6218549609184265,
      "learning_rate": 9.752472099099897e-06,
      "loss": 0.9497,
      "step": 1080
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.019995927810669,
      "learning_rate": 9.747906116858805e-06,
      "loss": 0.9218,
      "step": 1090
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6491205096244812,
      "learning_rate": 9.743299494518532e-06,
      "loss": 0.9447,
      "step": 1100
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6476668119430542,
      "learning_rate": 9.738652271509846e-06,
      "loss": 0.9189,
      "step": 1110
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6570378541946411,
      "learning_rate": 9.733964487611044e-06,
      "loss": 0.938,
      "step": 1120
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6387359499931335,
      "learning_rate": 9.729236182947597e-06,
      "loss": 0.933,
      "step": 1130
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5747265815734863,
      "learning_rate": 9.724467397991833e-06,
      "loss": 0.9155,
      "step": 1140
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6549525260925293,
      "learning_rate": 9.719658173562558e-06,
      "loss": 0.9293,
      "step": 1150
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6779614686965942,
      "learning_rate": 9.714808550824735e-06,
      "loss": 0.9377,
      "step": 1160
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6967174410820007,
      "learning_rate": 9.709918571289114e-06,
      "loss": 0.9253,
      "step": 1170
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7676512598991394,
      "learning_rate": 9.704988276811883e-06,
      "loss": 0.9454,
      "step": 1180
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.655295193195343,
      "learning_rate": 9.70001770959431e-06,
      "loss": 0.9125,
      "step": 1190
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7563107013702393,
      "learning_rate": 9.695006912182379e-06,
      "loss": 0.9405,
      "step": 1200
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.8399559259414673,
      "learning_rate": 9.68995592746643e-06,
      "loss": 0.9427,
      "step": 1210
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7510868906974792,
      "learning_rate": 9.684864798680789e-06,
      "loss": 0.9303,
      "step": 1220
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9126793146133423,
      "learning_rate": 9.679733569403398e-06,
      "loss": 0.9218,
      "step": 1230
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6488207578659058,
      "learning_rate": 9.674562283555445e-06,
      "loss": 0.943,
      "step": 1240
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9493321776390076,
      "learning_rate": 9.669350985400979e-06,
      "loss": 0.9429,
      "step": 1250
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7507259845733643,
      "learning_rate": 9.664099719546547e-06,
      "loss": 0.9041,
      "step": 1260
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7052815556526184,
      "learning_rate": 9.658808530940794e-06,
      "loss": 0.9214,
      "step": 1270
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7954320311546326,
      "learning_rate": 9.653477464874093e-06,
      "loss": 0.9178,
      "step": 1280
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0493558645248413,
      "learning_rate": 9.648106566978149e-06,
      "loss": 0.9019,
      "step": 1290
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.767572820186615,
      "learning_rate": 9.64269588322561e-06,
      "loss": 0.9551,
      "step": 1300
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7655388712882996,
      "learning_rate": 9.637245459929678e-06,
      "loss": 0.9281,
      "step": 1310
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2685116529464722,
      "learning_rate": 9.631755343743707e-06,
      "loss": 0.9443,
      "step": 1320
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6689302325248718,
      "learning_rate": 9.626225581660802e-06,
      "loss": 0.9128,
      "step": 1330
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.8155603408813477,
      "learning_rate": 9.620656221013428e-06,
      "loss": 0.9437,
      "step": 1340
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7464457154273987,
      "learning_rate": 9.61504730947299e-06,
      "loss": 0.891,
      "step": 1350
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.629400908946991,
      "learning_rate": 9.609398895049435e-06,
      "loss": 0.8955,
      "step": 1360
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7425062656402588,
      "learning_rate": 9.60371102609084e-06,
      "loss": 0.8861,
      "step": 1370
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6443877816200256,
      "learning_rate": 9.597983751282993e-06,
      "loss": 0.9051,
      "step": 1380
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7057684063911438,
      "learning_rate": 9.592217119648982e-06,
      "loss": 0.8885,
      "step": 1390
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.8697768449783325,
      "learning_rate": 9.586411180548771e-06,
      "loss": 0.914,
      "step": 1400
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7217336893081665,
      "learning_rate": 9.580565983678784e-06,
      "loss": 0.9314,
      "step": 1410
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.609015703201294,
      "learning_rate": 9.57468157907147e-06,
      "loss": 0.9134,
      "step": 1420
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7643072605133057,
      "learning_rate": 9.568758017094884e-06,
      "loss": 0.9209,
      "step": 1430
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6023324728012085,
      "learning_rate": 9.562795348452245e-06,
      "loss": 0.929,
      "step": 1440
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6700206398963928,
      "learning_rate": 9.55679362418152e-06,
      "loss": 0.9361,
      "step": 1450
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6590617895126343,
      "learning_rate": 9.550752895654963e-06,
      "loss": 0.8946,
      "step": 1460
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6463683843612671,
      "learning_rate": 9.544673214578699e-06,
      "loss": 0.9069,
      "step": 1470
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6586316823959351,
      "learning_rate": 9.538554632992265e-06,
      "loss": 0.9218,
      "step": 1480
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6628296375274658,
      "learning_rate": 9.532397203268172e-06,
      "loss": 0.8845,
      "step": 1490
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6902598142623901,
      "learning_rate": 9.526200978111452e-06,
      "loss": 0.9196,
      "step": 1500
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6551186442375183,
      "learning_rate": 9.519966010559217e-06,
      "loss": 0.9147,
      "step": 1510
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6625112295150757,
      "learning_rate": 9.513692353980186e-06,
      "loss": 0.912,
      "step": 1520
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7705367207527161,
      "learning_rate": 9.507380062074252e-06,
      "loss": 0.9298,
      "step": 1530
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6840114593505859,
      "learning_rate": 9.501029188872005e-06,
      "loss": 0.9199,
      "step": 1540
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6210880875587463,
      "learning_rate": 9.49463978873427e-06,
      "loss": 0.9077,
      "step": 1550
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7556228637695312,
      "learning_rate": 9.488211916351656e-06,
      "loss": 0.9089,
      "step": 1560
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.735604465007782,
      "learning_rate": 9.481745626744071e-06,
      "loss": 0.9305,
      "step": 1570
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7971885800361633,
      "learning_rate": 9.475240975260266e-06,
      "loss": 0.9319,
      "step": 1580
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.6014077067375183,
      "learning_rate": 9.468698017577344e-06,
      "loss": 0.9292,
      "step": 1590
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7462084889411926,
      "learning_rate": 9.4621168097003e-06,
      "loss": 0.8942,
      "step": 1600
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7277187705039978,
      "learning_rate": 9.455497407961533e-06,
      "loss": 0.8978,
      "step": 1610
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7430045008659363,
      "learning_rate": 9.448839869020364e-06,
      "loss": 0.9175,
      "step": 1620
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.693151593208313,
      "learning_rate": 9.442144249862555e-06,
      "loss": 0.9416,
      "step": 1630
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6412799954414368,
      "learning_rate": 9.435410607799818e-06,
      "loss": 0.8838,
      "step": 1640
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8170580863952637,
      "learning_rate": 9.428639000469326e-06,
      "loss": 0.924,
      "step": 1650
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.699052095413208,
      "learning_rate": 9.421829485833214e-06,
      "loss": 0.9344,
      "step": 1660
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7370940446853638,
      "learning_rate": 9.414982122178095e-06,
      "loss": 0.935,
      "step": 1670
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7777985334396362,
      "learning_rate": 9.40809696811455e-06,
      "loss": 0.933,
      "step": 1680
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7101726531982422,
      "learning_rate": 9.40117408257663e-06,
      "loss": 0.9002,
      "step": 1690
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6398364305496216,
      "learning_rate": 9.394213524821351e-06,
      "loss": 0.9449,
      "step": 1700
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8372628688812256,
      "learning_rate": 9.387215354428192e-06,
      "loss": 0.916,
      "step": 1710
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6209415793418884,
      "learning_rate": 9.380179631298573e-06,
      "loss": 0.9391,
      "step": 1720
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6791089177131653,
      "learning_rate": 9.373106415655359e-06,
      "loss": 0.8643,
      "step": 1730
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.681890070438385,
      "learning_rate": 9.365995768042327e-06,
      "loss": 0.9468,
      "step": 1740
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7921962738037109,
      "learning_rate": 9.35884774932366e-06,
      "loss": 0.9127,
      "step": 1750
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7955833077430725,
      "learning_rate": 9.351662420683421e-06,
      "loss": 0.9091,
      "step": 1760
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7748980522155762,
      "learning_rate": 9.344439843625034e-06,
      "loss": 0.897,
      "step": 1770
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8075698614120483,
      "learning_rate": 9.337180079970747e-06,
      "loss": 0.9111,
      "step": 1780
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6480504274368286,
      "learning_rate": 9.329883191861116e-06,
      "loss": 0.9003,
      "step": 1790
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6438207626342773,
      "learning_rate": 9.322549241754465e-06,
      "loss": 0.8922,
      "step": 1800
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7411364912986755,
      "learning_rate": 9.31517829242635e-06,
      "loss": 0.8988,
      "step": 1810
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.864782989025116,
      "learning_rate": 9.307770406969032e-06,
      "loss": 0.9395,
      "step": 1820
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7258287668228149,
      "learning_rate": 9.300325648790922e-06,
      "loss": 0.9025,
      "step": 1830
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7146714925765991,
      "learning_rate": 9.29284408161605e-06,
      "loss": 0.9223,
      "step": 1840
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6053746342658997,
      "learning_rate": 9.285325769483517e-06,
      "loss": 0.9332,
      "step": 1850
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8638799786567688,
      "learning_rate": 9.277770776746944e-06,
      "loss": 0.942,
      "step": 1860
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8022944927215576,
      "learning_rate": 9.270179168073923e-06,
      "loss": 0.909,
      "step": 1870
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7210090756416321,
      "learning_rate": 9.26255100844546e-06,
      "loss": 0.9314,
      "step": 1880
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.650475263595581,
      "learning_rate": 9.254886363155429e-06,
      "loss": 0.9355,
      "step": 1890
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8607627153396606,
      "learning_rate": 9.247185297809997e-06,
      "loss": 0.9138,
      "step": 1900
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9340272545814514,
      "learning_rate": 9.239447878327081e-06,
      "loss": 0.9417,
      "step": 1910
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7494912147521973,
      "learning_rate": 9.231674170935767e-06,
      "loss": 0.8857,
      "step": 1920
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7654171586036682,
      "learning_rate": 9.223864242175756e-06,
      "loss": 0.8908,
      "step": 1930
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7642711400985718,
      "learning_rate": 9.21601815889678e-06,
      "loss": 0.8765,
      "step": 1940
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.696348249912262,
      "learning_rate": 9.20813598825805e-06,
      "loss": 0.9323,
      "step": 1950
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.6663697957992554,
      "learning_rate": 9.200217797727663e-06,
      "loss": 0.8775,
      "step": 1960
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7656813859939575,
      "learning_rate": 9.192263655082033e-06,
      "loss": 0.8997,
      "step": 1970
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.6377949714660645,
      "learning_rate": 9.184273628405304e-06,
      "loss": 0.8842,
      "step": 1980
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7093709707260132,
      "learning_rate": 9.176247786088783e-06,
      "loss": 0.9002,
      "step": 1990
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.875128984451294,
      "learning_rate": 9.168186196830336e-06,
      "loss": 0.9209,
      "step": 2000
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7550728917121887,
      "learning_rate": 9.16008892963381e-06,
      "loss": 0.9035,
      "step": 2010
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0108016729354858,
      "learning_rate": 9.15195605380844e-06,
      "loss": 0.875,
      "step": 2020
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.7439371943473816,
      "learning_rate": 9.143787638968255e-06,
      "loss": 0.8722,
      "step": 2030
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6964495182037354,
      "learning_rate": 9.135583755031486e-06,
      "loss": 0.8726,
      "step": 2040
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9313227534294128,
      "learning_rate": 9.127344472219964e-06,
      "loss": 0.9107,
      "step": 2050
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4056764841079712,
      "learning_rate": 9.119069861058516e-06,
      "loss": 0.9061,
      "step": 2060
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.680668830871582,
      "learning_rate": 9.110759992374369e-06,
      "loss": 0.9099,
      "step": 2070
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.7828279733657837,
      "learning_rate": 9.102414937296542e-06,
      "loss": 0.8803,
      "step": 2080
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6008317470550537,
      "learning_rate": 9.094034767255225e-06,
      "loss": 0.8987,
      "step": 2090
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8024834990501404,
      "learning_rate": 9.085619553981186e-06,
      "loss": 0.9235,
      "step": 2100
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8021408915519714,
      "learning_rate": 9.07716936950515e-06,
      "loss": 0.8767,
      "step": 2110
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.7997848391532898,
      "learning_rate": 9.068684286157174e-06,
      "loss": 0.8941,
      "step": 2120
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8140115141868591,
      "learning_rate": 9.060164376566037e-06,
      "loss": 0.8886,
      "step": 2130
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7906537652015686,
      "learning_rate": 9.051609713658619e-06,
      "loss": 0.8951,
      "step": 2140
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6860573291778564,
      "learning_rate": 9.04302037065927e-06,
      "loss": 0.8922,
      "step": 2150
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7657009363174438,
      "learning_rate": 9.034396421089192e-06,
      "loss": 0.9203,
      "step": 2160
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9350467920303345,
      "learning_rate": 9.025737938765803e-06,
      "loss": 0.9592,
      "step": 2170
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.8163906931877136,
      "learning_rate": 9.0170449978021e-06,
      "loss": 0.9156,
      "step": 2180
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6962751150131226,
      "learning_rate": 9.008317672606044e-06,
      "loss": 0.8692,
      "step": 2190
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6299543976783752,
      "learning_rate": 8.9995560378799e-06,
      "loss": 0.9052,
      "step": 2200
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6928566694259644,
      "learning_rate": 8.990760168619616e-06,
      "loss": 0.9231,
      "step": 2210
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.8014935851097107,
      "learning_rate": 8.981930140114167e-06,
      "loss": 0.9165,
      "step": 2220
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7766875624656677,
      "learning_rate": 8.97306602794492e-06,
      "loss": 0.9003,
      "step": 2230
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7637115120887756,
      "learning_rate": 8.964167907984989e-06,
      "loss": 0.8775,
      "step": 2240
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8758956789970398,
      "learning_rate": 8.955235856398568e-06,
      "loss": 0.8945,
      "step": 2250
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8250870704650879,
      "learning_rate": 8.946269949640306e-06,
      "loss": 0.8865,
      "step": 2260
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6754716634750366,
      "learning_rate": 8.937270264454624e-06,
      "loss": 0.8994,
      "step": 2270
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6988954544067383,
      "learning_rate": 8.928236877875084e-06,
      "loss": 0.9221,
      "step": 2280
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7391093373298645,
      "learning_rate": 8.919169867223713e-06,
      "loss": 0.9124,
      "step": 2290
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7323642373085022,
      "learning_rate": 8.910069310110346e-06,
      "loss": 0.887,
      "step": 2300
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9379021525382996,
      "learning_rate": 8.900935284431962e-06,
      "loss": 0.8925,
      "step": 2310
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7654469609260559,
      "learning_rate": 8.891767868372016e-06,
      "loss": 0.9569,
      "step": 2320
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.6094922423362732,
      "learning_rate": 8.882567140399775e-06,
      "loss": 0.878,
      "step": 2330
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7786396741867065,
      "learning_rate": 8.873333179269635e-06,
      "loss": 0.8933,
      "step": 2340
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6561703681945801,
      "learning_rate": 8.864066064020462e-06,
      "loss": 0.8821,
      "step": 2350
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6152118444442749,
      "learning_rate": 8.854765873974898e-06,
      "loss": 0.8905,
      "step": 2360
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8001115322113037,
      "learning_rate": 8.845432688738703e-06,
      "loss": 0.9002,
      "step": 2370
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.709892988204956,
      "learning_rate": 8.836066588200052e-06,
      "loss": 0.887,
      "step": 2380
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.6732577085494995,
      "learning_rate": 8.826667652528866e-06,
      "loss": 0.8866,
      "step": 2390
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9450706839561462,
      "learning_rate": 8.817235962176121e-06,
      "loss": 0.8808,
      "step": 2400
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9471026659011841,
      "learning_rate": 8.807771597873159e-06,
      "loss": 0.8943,
      "step": 2410
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7772449851036072,
      "learning_rate": 8.798274640630997e-06,
      "loss": 0.8859,
      "step": 2420
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.8146265745162964,
      "learning_rate": 8.788745171739632e-06,
      "loss": 0.8869,
      "step": 2430
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7327932715415955,
      "learning_rate": 8.779183272767353e-06,
      "loss": 0.874,
      "step": 2440
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6848121881484985,
      "learning_rate": 8.76958902556003e-06,
      "loss": 0.8677,
      "step": 2450
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.679011881351471,
      "learning_rate": 8.759962512240426e-06,
      "loss": 0.9292,
      "step": 2460
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8341182470321655,
      "learning_rate": 8.750303815207487e-06,
      "loss": 0.9098,
      "step": 2470
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.874434769153595,
      "learning_rate": 8.740613017135634e-06,
      "loss": 0.9081,
      "step": 2480
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7139870524406433,
      "learning_rate": 8.730890200974065e-06,
      "loss": 0.896,
      "step": 2490
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7836432456970215,
      "learning_rate": 8.721135449946037e-06,
      "loss": 0.8939,
      "step": 2500
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6674822568893433,
      "learning_rate": 8.711348847548157e-06,
      "loss": 0.9039,
      "step": 2510
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6287670135498047,
      "learning_rate": 8.701530477549666e-06,
      "loss": 0.8786,
      "step": 2520
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7027100324630737,
      "learning_rate": 8.691680423991722e-06,
      "loss": 0.8887,
      "step": 2530
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.713030993938446,
      "learning_rate": 8.681798771186683e-06,
      "loss": 0.9048,
      "step": 2540
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7505326867103577,
      "learning_rate": 8.67188560371738e-06,
      "loss": 0.868,
      "step": 2550
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6765055060386658,
      "learning_rate": 8.661941006436401e-06,
      "loss": 0.8686,
      "step": 2560
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7267132997512817,
      "learning_rate": 8.651965064465355e-06,
      "loss": 0.9243,
      "step": 2570
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6484735608100891,
      "learning_rate": 8.641957863194153e-06,
      "loss": 0.8846,
      "step": 2580
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6731644868850708,
      "learning_rate": 8.631919488280267e-06,
      "loss": 0.914,
      "step": 2590
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6963881850242615,
      "learning_rate": 8.621850025648008e-06,
      "loss": 0.8604,
      "step": 2600
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9081583619117737,
      "learning_rate": 8.611749561487785e-06,
      "loss": 0.9331,
      "step": 2610
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7570102214813232,
      "learning_rate": 8.601618182255364e-06,
      "loss": 0.8996,
      "step": 2620
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.820863664150238,
      "learning_rate": 8.59145597467113e-06,
      "loss": 0.8707,
      "step": 2630
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8625741004943848,
      "learning_rate": 8.581263025719352e-06,
      "loss": 0.8806,
      "step": 2640
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.1321467161178589,
      "learning_rate": 8.571039422647423e-06,
      "loss": 0.9222,
      "step": 2650
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7913120985031128,
      "learning_rate": 8.560785252965131e-06,
      "loss": 0.9212,
      "step": 2660
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7801851034164429,
      "learning_rate": 8.5505006044439e-06,
      "loss": 0.9001,
      "step": 2670
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.689361035823822,
      "learning_rate": 8.540185565116034e-06,
      "loss": 0.9105,
      "step": 2680
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6706151366233826,
      "learning_rate": 8.52984022327398e-06,
      "loss": 0.8955,
      "step": 2690
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8290680050849915,
      "learning_rate": 8.51946466746955e-06,
      "loss": 0.9029,
      "step": 2700
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.74539715051651,
      "learning_rate": 8.509058986513185e-06,
      "loss": 0.9072,
      "step": 2710
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6989156603813171,
      "learning_rate": 8.498623269473178e-06,
      "loss": 0.9111,
      "step": 2720
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8335017561912537,
      "learning_rate": 8.488157605674924e-06,
      "loss": 0.9043,
      "step": 2730
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8350189924240112,
      "learning_rate": 8.477662084700145e-06,
      "loss": 0.9496,
      "step": 2740
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7352144122123718,
      "learning_rate": 8.46713679638613e-06,
      "loss": 0.8805,
      "step": 2750
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8180497884750366,
      "learning_rate": 8.456581830824966e-06,
      "loss": 0.8715,
      "step": 2760
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8565731644630432,
      "learning_rate": 8.445997278362759e-06,
      "loss": 0.881,
      "step": 2770
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7524616122245789,
      "learning_rate": 8.435383229598872e-06,
      "loss": 0.873,
      "step": 2780
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8238971829414368,
      "learning_rate": 8.424739775385145e-06,
      "loss": 0.8766,
      "step": 2790
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7271744608879089,
      "learning_rate": 8.414067006825108e-06,
      "loss": 0.8659,
      "step": 2800
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.6863881945610046,
      "learning_rate": 8.403365015273223e-06,
      "loss": 0.9122,
      "step": 2810
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.7746860384941101,
      "learning_rate": 8.392633892334074e-06,
      "loss": 0.8661,
      "step": 2820
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.7054237723350525,
      "learning_rate": 8.381873729861609e-06,
      "loss": 0.8884,
      "step": 2830
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6736589074134827,
      "learning_rate": 8.371084619958337e-06,
      "loss": 0.8776,
      "step": 2840
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.671820878982544,
      "learning_rate": 8.360266654974543e-06,
      "loss": 0.866,
      "step": 2850
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8050073981285095,
      "learning_rate": 8.349419927507505e-06,
      "loss": 0.8884,
      "step": 2860
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8177492618560791,
      "learning_rate": 8.338544530400693e-06,
      "loss": 0.8678,
      "step": 2870
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8012502193450928,
      "learning_rate": 8.327640556742978e-06,
      "loss": 0.8792,
      "step": 2880
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6939083337783813,
      "learning_rate": 8.316708099867834e-06,
      "loss": 0.8928,
      "step": 2890
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7548695206642151,
      "learning_rate": 8.305747253352534e-06,
      "loss": 0.8564,
      "step": 2900
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7304048538208008,
      "learning_rate": 8.294758111017367e-06,
      "loss": 0.8836,
      "step": 2910
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8671132922172546,
      "learning_rate": 8.28374076692481e-06,
      "loss": 0.8864,
      "step": 2920
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7430797219276428,
      "learning_rate": 8.272695315378744e-06,
      "loss": 0.874,
      "step": 2930
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6934476494789124,
      "learning_rate": 8.261621850923634e-06,
      "loss": 0.8583,
      "step": 2940
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7404267191886902,
      "learning_rate": 8.250520468343722e-06,
      "loss": 0.8734,
      "step": 2950
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6465054154396057,
      "learning_rate": 8.239391262662221e-06,
      "loss": 0.8734,
      "step": 2960
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6825598478317261,
      "learning_rate": 8.228234329140498e-06,
      "loss": 0.8666,
      "step": 2970
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6966167688369751,
      "learning_rate": 8.217049763277257e-06,
      "loss": 0.8823,
      "step": 2980
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8441101908683777,
      "learning_rate": 8.205837660807726e-06,
      "loss": 0.8876,
      "step": 2990
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.78465336561203,
      "learning_rate": 8.194598117702828e-06,
      "loss": 0.8991,
      "step": 3000
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7731233835220337,
      "learning_rate": 8.183331230168376e-06,
      "loss": 0.8763,
      "step": 3010
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7912877202033997,
      "learning_rate": 8.172037094644236e-06,
      "loss": 0.867,
      "step": 3020
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7123367190361023,
      "learning_rate": 8.160715807803503e-06,
      "loss": 0.9098,
      "step": 3030
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.689221203327179,
      "learning_rate": 8.149367466551678e-06,
      "loss": 0.8859,
      "step": 3040
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7518553137779236,
      "learning_rate": 8.137992168025835e-06,
      "loss": 0.9198,
      "step": 3050
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7416417598724365,
      "learning_rate": 8.126590009593794e-06,
      "loss": 0.9011,
      "step": 3060
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6670682430267334,
      "learning_rate": 8.115161088853283e-06,
      "loss": 0.8884,
      "step": 3070
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7208391427993774,
      "learning_rate": 8.103705503631103e-06,
      "loss": 0.8903,
      "step": 3080
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.635645866394043,
      "learning_rate": 8.092223351982296e-06,
      "loss": 0.8773,
      "step": 3090
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7762060761451721,
      "learning_rate": 8.080714732189295e-06,
      "loss": 0.9039,
      "step": 3100
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.6757723689079285,
      "learning_rate": 8.069179742761092e-06,
      "loss": 0.8745,
      "step": 3110
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7416348457336426,
      "learning_rate": 8.057618482432399e-06,
      "loss": 0.9077,
      "step": 3120
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8212820887565613,
      "learning_rate": 8.046031050162784e-06,
      "loss": 0.8753,
      "step": 3130
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.6839666962623596,
      "learning_rate": 8.034417545135846e-06,
      "loss": 0.8759,
      "step": 3140
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.767983078956604,
      "learning_rate": 8.022778066758348e-06,
      "loss": 0.8778,
      "step": 3150
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.6773496270179749,
      "learning_rate": 8.01111271465938e-06,
      "loss": 0.868,
      "step": 3160
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.793113648891449,
      "learning_rate": 7.999421588689498e-06,
      "loss": 0.9105,
      "step": 3170
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7550264000892639,
      "learning_rate": 7.987704788919872e-06,
      "loss": 0.9105,
      "step": 3180
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7124478220939636,
      "learning_rate": 7.975962415641428e-06,
      "loss": 0.8728,
      "step": 3190
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.805101215839386,
      "learning_rate": 7.96419456936399e-06,
      "loss": 0.9488,
      "step": 3200
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8259230852127075,
      "learning_rate": 7.952401350815427e-06,
      "loss": 0.8665,
      "step": 3210
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7200030088424683,
      "learning_rate": 7.940582860940771e-06,
      "loss": 0.9036,
      "step": 3220
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6903145909309387,
      "learning_rate": 7.928739200901381e-06,
      "loss": 0.8781,
      "step": 3230
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6554933786392212,
      "learning_rate": 7.91687047207405e-06,
      "loss": 0.8773,
      "step": 3240
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7486357688903809,
      "learning_rate": 7.904976776050156e-06,
      "loss": 0.8914,
      "step": 3250
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7948868870735168,
      "learning_rate": 7.893058214634785e-06,
      "loss": 0.8497,
      "step": 3260
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.6839557886123657,
      "learning_rate": 7.881114889845858e-06,
      "loss": 0.9194,
      "step": 3270
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8078474998474121,
      "learning_rate": 7.869146903913261e-06,
      "loss": 0.8988,
      "step": 3280
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7275850176811218,
      "learning_rate": 7.857154359277972e-06,
      "loss": 0.8843,
      "step": 3290
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.638208270072937,
      "learning_rate": 7.845137358591176e-06,
      "loss": 0.8942,
      "step": 3300
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7862017750740051,
      "learning_rate": 7.833096004713398e-06,
      "loss": 0.8712,
      "step": 3310
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7157966494560242,
      "learning_rate": 7.82103040071361e-06,
      "loss": 0.8902,
      "step": 3320
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7477262020111084,
      "learning_rate": 7.80894064986836e-06,
      "loss": 0.9196,
      "step": 3330
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9382122755050659,
      "learning_rate": 7.796826855660876e-06,
      "loss": 0.8779,
      "step": 3340
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6241675615310669,
      "learning_rate": 7.7846891217802e-06,
      "loss": 0.891,
      "step": 3350
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.7218158841133118,
      "learning_rate": 7.772527552120274e-06,
      "loss": 0.8948,
      "step": 3360
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9833104610443115,
      "learning_rate": 7.760342250779069e-06,
      "loss": 0.8914,
      "step": 3370
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8101766109466553,
      "learning_rate": 7.748133322057693e-06,
      "loss": 0.8875,
      "step": 3380
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8606131672859192,
      "learning_rate": 7.73590087045949e-06,
      "loss": 0.9178,
      "step": 3390
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6935533881187439,
      "learning_rate": 7.723645000689153e-06,
      "loss": 0.9003,
      "step": 3400
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8851600289344788,
      "learning_rate": 7.711365817651822e-06,
      "loss": 0.8694,
      "step": 3410
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7385109663009644,
      "learning_rate": 7.699063426452191e-06,
      "loss": 0.8655,
      "step": 3420
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7656209468841553,
      "learning_rate": 7.686737932393606e-06,
      "loss": 0.8576,
      "step": 3430
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.6485370397567749,
      "learning_rate": 7.67438944097716e-06,
      "loss": 0.8662,
      "step": 3440
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7411097884178162,
      "learning_rate": 7.662018057900805e-06,
      "loss": 0.8384,
      "step": 3450
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.6974339485168457,
      "learning_rate": 7.649623889058423e-06,
      "loss": 0.8955,
      "step": 3460
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7264565825462341,
      "learning_rate": 7.637207040538937e-06,
      "loss": 0.8896,
      "step": 3470
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7551647424697876,
      "learning_rate": 7.6247676186253995e-06,
      "loss": 0.8735,
      "step": 3480
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8099594116210938,
      "learning_rate": 7.612305729794085e-06,
      "loss": 0.8714,
      "step": 3490
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8042902946472168,
      "learning_rate": 7.599821480713571e-06,
      "loss": 0.8959,
      "step": 3500
    }
  ],
  "logging_steps": 10,
  "max_steps": 10738,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1.2581155985441161e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}