{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.10404203298132446,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00034680677660441485,
      "grad_norm": 8.286304473876953,
      "learning_rate": 2.6999999999999996e-06,
      "loss": 36.023321533203124,
      "step": 10
    },
    {
      "epoch": 0.0006936135532088297,
      "grad_norm": 5.303624153137207,
      "learning_rate": 5.7e-06,
      "loss": 35.81325988769531,
      "step": 20
    },
    {
      "epoch": 0.0010404203298132447,
      "grad_norm": 14.0803861618042,
      "learning_rate": 8.7e-06,
      "loss": 35.61415100097656,
      "step": 30
    },
    {
      "epoch": 0.0013872271064176594,
      "grad_norm": 42.10546875,
      "learning_rate": 1.17e-05,
      "loss": 35.36910400390625,
      "step": 40
    },
    {
      "epoch": 0.0017340338830220744,
      "grad_norm": 43.37276840209961,
      "learning_rate": 1.47e-05,
      "loss": 34.86077880859375,
      "step": 50
    },
    {
      "epoch": 0.0020808406596264893,
      "grad_norm": Infinity,
      "learning_rate": 1.7699999999999997e-05,
      "loss": 34.67452697753906,
      "step": 60
    },
    {
      "epoch": 0.002427647436230904,
      "grad_norm": 86.09446716308594,
      "learning_rate": 2.04e-05,
      "loss": 34.29078979492188,
      "step": 70
    },
    {
      "epoch": 0.002774454212835319,
      "grad_norm": 43.50123977661133,
      "learning_rate": 2.34e-05,
      "loss": 33.816766357421876,
      "step": 80
    },
    {
      "epoch": 0.0031212609894397335,
      "grad_norm": 104.7258529663086,
      "learning_rate": 2.6399999999999995e-05,
      "loss": 33.36159057617188,
      "step": 90
    },
    {
      "epoch": 0.0034680677660441487,
      "grad_norm": 66.46515655517578,
      "learning_rate": 2.91e-05,
      "loss": 33.181732177734375,
      "step": 100
    },
    {
      "epoch": 0.0038148745426485635,
      "grad_norm": 127.66901397705078,
      "learning_rate": 3.2099999999999994e-05,
      "loss": 32.83320007324219,
      "step": 110
    },
    {
      "epoch": 0.004161681319252979,
      "grad_norm": 71.67156982421875,
      "learning_rate": 3.51e-05,
      "loss": 32.48929138183594,
      "step": 120
    },
    {
      "epoch": 0.004508488095857393,
      "grad_norm": 76.15655517578125,
      "learning_rate": 3.81e-05,
      "loss": 32.10602722167969,
      "step": 130
    },
    {
      "epoch": 0.004855294872461808,
      "grad_norm": 38.65349578857422,
      "learning_rate": 4.11e-05,
      "loss": 31.449526977539062,
      "step": 140
    },
    {
      "epoch": 0.005202101649066222,
      "grad_norm": 59.55622863769531,
      "learning_rate": 4.4099999999999995e-05,
      "loss": 31.01339111328125,
      "step": 150
    },
    {
      "epoch": 0.005548908425670638,
      "grad_norm": 130.26136779785156,
      "learning_rate": 4.709999999999999e-05,
      "loss": 30.92586669921875,
      "step": 160
    },
    {
      "epoch": 0.005895715202275053,
      "grad_norm": 77.2778549194336,
      "learning_rate": 5.01e-05,
      "loss": 30.492510986328124,
      "step": 170
    },
    {
      "epoch": 0.006242521978879467,
      "grad_norm": 29.711563110351562,
      "learning_rate": 5.309999999999999e-05,
      "loss": 29.836746215820312,
      "step": 180
    },
    {
      "epoch": 0.006589328755483882,
      "grad_norm": 29.891956329345703,
      "learning_rate": 5.6099999999999995e-05,
      "loss": 29.291275024414062,
      "step": 190
    },
    {
      "epoch": 0.006936135532088297,
      "grad_norm": 30.56305503845215,
      "learning_rate": 5.91e-05,
      "loss": 29.14805908203125,
      "step": 200
    },
    {
      "epoch": 0.007282942308692712,
      "grad_norm": 29.489208221435547,
      "learning_rate": 6.209999999999999e-05,
      "loss": 28.80553283691406,
      "step": 210
    },
    {
      "epoch": 0.007629749085297127,
      "grad_norm": 37.147377014160156,
      "learning_rate": 6.51e-05,
      "loss": 28.694085693359376,
      "step": 220
    },
    {
      "epoch": 0.007976555861901541,
      "grad_norm": 28.496992111206055,
      "learning_rate": 6.81e-05,
      "loss": 28.166943359375,
      "step": 230
    },
    {
      "epoch": 0.008323362638505957,
      "grad_norm": 29.42328453063965,
      "learning_rate": 7.11e-05,
      "loss": 27.878482055664062,
      "step": 240
    },
    {
      "epoch": 0.008670169415110372,
      "grad_norm": 27.867637634277344,
      "learning_rate": 7.41e-05,
      "loss": 26.99749755859375,
      "step": 250
    },
    {
      "epoch": 0.009016976191714786,
      "grad_norm": 29.695758819580078,
      "learning_rate": 7.709999999999999e-05,
      "loss": 27.208102416992187,
      "step": 260
    },
    {
      "epoch": 0.0093637829683192,
      "grad_norm": 34.605934143066406,
      "learning_rate": 8.01e-05,
      "loss": 26.798681640625,
      "step": 270
    },
    {
      "epoch": 0.009710589744923616,
      "grad_norm": 35.05394744873047,
      "learning_rate": 8.31e-05,
      "loss": 27.192486572265626,
      "step": 280
    },
    {
      "epoch": 0.01005739652152803,
      "grad_norm": 19.137067794799805,
      "learning_rate": 8.609999999999999e-05,
      "loss": 26.89140930175781,
      "step": 290
    },
    {
      "epoch": 0.010404203298132445,
      "grad_norm": 24.543006896972656,
      "learning_rate": 8.879999999999999e-05,
      "loss": 26.167477416992188,
      "step": 300
    },
    {
      "epoch": 0.010751010074736861,
      "grad_norm": 18.849651336669922,
      "learning_rate": 9.18e-05,
      "loss": 26.41388854980469,
      "step": 310
    },
    {
      "epoch": 0.011097816851341275,
      "grad_norm": 21.179908752441406,
      "learning_rate": 9.479999999999999e-05,
      "loss": 26.003546142578124,
      "step": 320
    },
    {
      "epoch": 0.01144462362794569,
      "grad_norm": 19.344528198242188,
      "learning_rate": 9.779999999999999e-05,
      "loss": 26.271051025390626,
      "step": 330
    },
    {
      "epoch": 0.011791430404550106,
      "grad_norm": 19.74913215637207,
      "learning_rate": 0.0001008,
      "loss": 26.064590454101562,
      "step": 340
    },
    {
      "epoch": 0.01213823718115452,
      "grad_norm": 17.638103485107422,
      "learning_rate": 0.00010379999999999999,
      "loss": 25.863119506835936,
      "step": 350
    },
    {
      "epoch": 0.012485043957758934,
      "grad_norm": 24.873981475830078,
      "learning_rate": 0.00010679999999999998,
      "loss": 26.28948059082031,
      "step": 360
    },
    {
      "epoch": 0.01283185073436335,
      "grad_norm": 17.048828125,
      "learning_rate": 0.00010979999999999999,
      "loss": 26.215179443359375,
      "step": 370
    },
    {
      "epoch": 0.013178657510967764,
      "grad_norm": 21.703975677490234,
      "learning_rate": 0.00011279999999999999,
      "loss": 26.152392578125,
      "step": 380
    },
    {
      "epoch": 0.013525464287572179,
      "grad_norm": 22.66498565673828,
      "learning_rate": 0.0001158,
      "loss": 26.0640625,
      "step": 390
    },
    {
      "epoch": 0.013872271064176595,
      "grad_norm": 16.945016860961914,
      "learning_rate": 0.0001188,
      "loss": 26.023715209960937,
      "step": 400
    },
    {
      "epoch": 0.01421907784078101,
      "grad_norm": 27.866268157958984,
      "learning_rate": 0.00012179999999999999,
      "loss": 25.43433380126953,
      "step": 410
    },
    {
      "epoch": 0.014565884617385423,
      "grad_norm": 22.481279373168945,
      "learning_rate": 0.00012479999999999997,
      "loss": 25.464639282226564,
      "step": 420
    },
    {
      "epoch": 0.014912691393989838,
      "grad_norm": 14.993969917297363,
      "learning_rate": 0.0001278,
      "loss": 25.218385314941408,
      "step": 430
    },
    {
      "epoch": 0.015259498170594254,
      "grad_norm": 14.426427841186523,
      "learning_rate": 0.00013079999999999998,
      "loss": 24.94810333251953,
      "step": 440
    },
    {
      "epoch": 0.015606304947198668,
      "grad_norm": 15.592818260192871,
      "learning_rate": 0.0001338,
      "loss": 25.38585662841797,
      "step": 450
    },
    {
      "epoch": 0.015953111723803082,
      "grad_norm": 14.709174156188965,
      "learning_rate": 0.0001368,
      "loss": 24.31098175048828,
      "step": 460
    },
    {
      "epoch": 0.016299918500407497,
      "grad_norm": 13.651994705200195,
      "learning_rate": 0.00013979999999999998,
      "loss": 24.82652893066406,
      "step": 470
    },
    {
      "epoch": 0.016646725277011915,
      "grad_norm": 17.521072387695312,
      "learning_rate": 0.00014279999999999997,
      "loss": 25.121923828125,
      "step": 480
    },
    {
      "epoch": 0.01699353205361633,
      "grad_norm": 13.458234786987305,
      "learning_rate": 0.0001458,
      "loss": 24.447775268554686,
      "step": 490
    },
    {
      "epoch": 0.017340338830220743,
      "grad_norm": 13.57589340209961,
      "learning_rate": 0.00014879999999999998,
      "loss": 24.76665802001953,
      "step": 500
    },
    {
      "epoch": 0.017687145606825157,
      "grad_norm": 12.934654235839844,
      "learning_rate": 0.00015179999999999998,
      "loss": 24.170372009277344,
      "step": 510
    },
    {
      "epoch": 0.018033952383429572,
      "grad_norm": 13.884750366210938,
      "learning_rate": 0.0001548,
      "loss": 24.240016174316406,
      "step": 520
    },
    {
      "epoch": 0.018380759160033986,
      "grad_norm": 13.660422325134277,
      "learning_rate": 0.0001578,
      "loss": 23.897187805175783,
      "step": 530
    },
    {
      "epoch": 0.0187275659366384,
      "grad_norm": 11.849883079528809,
      "learning_rate": 0.0001608,
      "loss": 23.574449157714845,
      "step": 540
    },
    {
      "epoch": 0.019074372713242818,
      "grad_norm": 10.595759391784668,
      "learning_rate": 0.0001638,
      "loss": 23.569731140136717,
      "step": 550
    },
    {
      "epoch": 0.019421179489847232,
      "grad_norm": 12.363926887512207,
      "learning_rate": 0.0001668,
      "loss": 23.617283630371094,
      "step": 560
    },
    {
      "epoch": 0.019767986266451647,
      "grad_norm": 10.968243598937988,
      "learning_rate": 0.00016979999999999998,
      "loss": 23.990586853027345,
      "step": 570
    },
    {
      "epoch": 0.02011479304305606,
      "grad_norm": 12.141168594360352,
      "learning_rate": 0.00017279999999999997,
      "loss": 23.807048034667968,
      "step": 580
    },
    {
      "epoch": 0.020461599819660475,
      "grad_norm": 14.928491592407227,
      "learning_rate": 0.00017579999999999996,
      "loss": 22.980938720703126,
      "step": 590
    },
    {
      "epoch": 0.02080840659626489,
      "grad_norm": 11.39072036743164,
      "learning_rate": 0.00017879999999999998,
      "loss": 23.521900939941407,
      "step": 600
    },
    {
      "epoch": 0.021155213372869307,
      "grad_norm": 10.157258033752441,
      "learning_rate": 0.00018179999999999997,
      "loss": 23.075851440429688,
      "step": 610
    },
    {
      "epoch": 0.021502020149473722,
      "grad_norm": 12.673665046691895,
      "learning_rate": 0.0001848,
      "loss": 23.70465850830078,
      "step": 620
    },
    {
      "epoch": 0.021848826926078136,
      "grad_norm": 11.242607116699219,
      "learning_rate": 0.00018779999999999998,
      "loss": 22.982296752929688,
      "step": 630
    },
    {
      "epoch": 0.02219563370268255,
      "grad_norm": 25.496898651123047,
      "learning_rate": 0.00019079999999999998,
      "loss": 22.516278076171876,
      "step": 640
    },
    {
      "epoch": 0.022542440479286965,
      "grad_norm": 11.215885162353516,
      "learning_rate": 0.0001938,
      "loss": 22.14655456542969,
      "step": 650
    },
    {
      "epoch": 0.02288924725589138,
      "grad_norm": 9.149953842163086,
      "learning_rate": 0.00019679999999999999,
      "loss": 22.114730834960938,
      "step": 660
    },
    {
      "epoch": 0.023236054032495793,
      "grad_norm": 14.604310035705566,
      "learning_rate": 0.0001995,
      "loss": 22.670735168457032,
      "step": 670
    },
    {
      "epoch": 0.02358286080910021,
      "grad_norm": 11.181646347045898,
      "learning_rate": 0.0002025,
      "loss": 22.34198455810547,
      "step": 680
    },
    {
      "epoch": 0.023929667585704625,
      "grad_norm": 10.59487533569336,
      "learning_rate": 0.0002055,
      "loss": 22.268771362304687,
      "step": 690
    },
    {
      "epoch": 0.02427647436230904,
      "grad_norm": 11.758952140808105,
      "learning_rate": 0.00020849999999999997,
      "loss": 21.350621032714844,
      "step": 700
    },
    {
      "epoch": 0.024623281138913454,
      "grad_norm": 18.59368896484375,
      "learning_rate": 0.00021149999999999996,
      "loss": 21.507183837890626,
      "step": 710
    },
    {
      "epoch": 0.02497008791551787,
      "grad_norm": 9.710185050964355,
      "learning_rate": 0.00021449999999999998,
      "loss": 21.502236938476564,
      "step": 720
    },
    {
      "epoch": 0.025316894692122283,
      "grad_norm": 20.68129539489746,
      "learning_rate": 0.00021749999999999997,
      "loss": 21.232615661621093,
      "step": 730
    },
    {
      "epoch": 0.0256637014687267,
      "grad_norm": 9.399124145507812,
      "learning_rate": 0.00022049999999999997,
      "loss": 21.146585083007814,
      "step": 740
    },
    {
      "epoch": 0.026010508245331115,
      "grad_norm": 9.145174026489258,
      "learning_rate": 0.00022349999999999998,
      "loss": 21.06390838623047,
      "step": 750
    },
    {
      "epoch": 0.02635731502193553,
      "grad_norm": 11.161269187927246,
      "learning_rate": 0.00022649999999999998,
      "loss": 20.896176147460938,
      "step": 760
    },
    {
      "epoch": 0.026704121798539943,
      "grad_norm": 8.362421035766602,
      "learning_rate": 0.0002295,
      "loss": 20.884358215332032,
      "step": 770
    },
    {
      "epoch": 0.027050928575144358,
      "grad_norm": 9.504199981689453,
      "learning_rate": 0.00023249999999999999,
      "loss": 20.086947631835937,
      "step": 780
    },
    {
      "epoch": 0.027397735351748772,
      "grad_norm": 8.848526954650879,
      "learning_rate": 0.00023549999999999998,
      "loss": 20.411599731445314,
      "step": 790
    },
    {
      "epoch": 0.02774454212835319,
      "grad_norm": 9.612874984741211,
      "learning_rate": 0.0002385,
      "loss": 20.284182739257812,
      "step": 800
    },
    {
      "epoch": 0.028091348904957604,
      "grad_norm": 8.688453674316406,
      "learning_rate": 0.0002415,
      "loss": 20.024198913574217,
      "step": 810
    },
    {
      "epoch": 0.02843815568156202,
      "grad_norm": 9.779630661010742,
      "learning_rate": 0.0002445,
      "loss": 19.91345977783203,
      "step": 820
    },
    {
      "epoch": 0.028784962458166433,
      "grad_norm": 12.37082576751709,
      "learning_rate": 0.0002472,
      "loss": 19.54875030517578,
      "step": 830
    },
    {
      "epoch": 0.029131769234770847,
      "grad_norm": 8.736369132995605,
      "learning_rate": 0.00025019999999999996,
      "loss": 19.499098205566405,
      "step": 840
    },
    {
      "epoch": 0.02947857601137526,
      "grad_norm": 10.752023696899414,
      "learning_rate": 0.0002532,
      "loss": 19.327198791503907,
      "step": 850
    },
    {
      "epoch": 0.029825382787979676,
      "grad_norm": 9.755398750305176,
      "learning_rate": 0.0002562,
      "loss": 19.291722106933594,
      "step": 860
    },
    {
      "epoch": 0.030172189564584093,
      "grad_norm": 8.066308975219727,
      "learning_rate": 0.00025919999999999996,
      "loss": 18.516694641113283,
      "step": 870
    },
    {
      "epoch": 0.030518996341188508,
      "grad_norm": 8.660676002502441,
      "learning_rate": 0.0002622,
      "loss": 18.70814666748047,
      "step": 880
    },
    {
      "epoch": 0.030865803117792922,
      "grad_norm": 8.439098358154297,
      "learning_rate": 0.0002652,
      "loss": 18.787985229492186,
      "step": 890
    },
    {
      "epoch": 0.031212609894397336,
      "grad_norm": 8.525120735168457,
      "learning_rate": 0.00026819999999999996,
      "loss": 18.356887817382812,
      "step": 900
    },
    {
      "epoch": 0.031559416671001754,
      "grad_norm": 9.237743377685547,
      "learning_rate": 0.0002712,
      "loss": 18.07917022705078,
      "step": 910
    },
    {
      "epoch": 0.031906223447606165,
      "grad_norm": 8.137825965881348,
      "learning_rate": 0.0002742,
      "loss": 18.066172790527343,
      "step": 920
    },
    {
      "epoch": 0.03225303022421058,
      "grad_norm": 20.67136001586914,
      "learning_rate": 0.0002772,
      "loss": 18.19664306640625,
      "step": 930
    },
    {
      "epoch": 0.032599837000814993,
      "grad_norm": 8.68689250946045,
      "learning_rate": 0.0002802,
      "loss": 17.834249877929686,
      "step": 940
    },
    {
      "epoch": 0.03294664377741941,
      "grad_norm": 8.076814651489258,
      "learning_rate": 0.00028319999999999994,
      "loss": 17.645645141601562,
      "step": 950
    },
    {
      "epoch": 0.03329345055402383,
      "grad_norm": 8.278287887573242,
      "learning_rate": 0.00028619999999999996,
      "loss": 17.20880889892578,
      "step": 960
    },
    {
      "epoch": 0.03364025733062824,
      "grad_norm": 8.195657730102539,
      "learning_rate": 0.0002892,
      "loss": 17.02202453613281,
      "step": 970
    },
    {
      "epoch": 0.03398706410723266,
      "grad_norm": 35.29956817626953,
      "learning_rate": 0.00029219999999999995,
      "loss": 17.46351013183594,
      "step": 980
    },
    {
      "epoch": 0.03433387088383707,
      "grad_norm": 14.36285400390625,
      "learning_rate": 0.00029519999999999997,
      "loss": 17.128692626953125,
      "step": 990
    },
    {
      "epoch": 0.034680677660441486,
      "grad_norm": 117.96296691894531,
      "learning_rate": 0.0002979,
      "loss": 16.577227783203124,
      "step": 1000
    },
    {
      "epoch": 0.0350274844370459,
      "grad_norm": 33.18951416015625,
      "learning_rate": 0.00029999999932027514,
      "loss": 16.93028106689453,
      "step": 1010
    },
    {
      "epoch": 0.035374291213650315,
      "grad_norm": 16.420063018798828,
      "learning_rate": 0.00029999998723627847,
      "loss": 17.68489990234375,
      "step": 1020
    },
    {
      "epoch": 0.03572109799025473,
      "grad_norm": 37.7625732421875,
      "learning_rate": 0.0002999999600472873,
      "loss": 17.295022583007814,
      "step": 1030
    },
    {
      "epoch": 0.036067904766859143,
      "grad_norm": 32.17815399169922,
      "learning_rate": 0.00029999992266242694,
      "loss": 17.40197296142578,
      "step": 1040
    },
    {
      "epoch": 0.03641471154346356,
      "grad_norm": 10.255291938781738,
      "learning_rate": 0.00029999986677395476,
      "loss": 17.136293029785158,
      "step": 1050
    },
    {
      "epoch": 0.03676151832006797,
      "grad_norm": 7.522293567657471,
      "learning_rate": 0.0002999997957805001,
      "loss": 16.333123779296876,
      "step": 1060
    },
    {
      "epoch": 0.03710832509667239,
      "grad_norm": 7.725010395050049,
      "learning_rate": 0.00029999970968207007,
      "loss": 15.974671936035156,
      "step": 1070
    },
    {
      "epoch": 0.0374551318732768,
      "grad_norm": 7.349689483642578,
      "learning_rate": 0.00029999960847867344,
      "loss": 15.520753479003906,
      "step": 1080
    },
    {
      "epoch": 0.03780193864988122,
      "grad_norm": 6.910863399505615,
      "learning_rate": 0.00029999949217032024,
      "loss": 15.53778533935547,
      "step": 1090
    },
    {
      "epoch": 0.038148745426485636,
      "grad_norm": 7.105229377746582,
      "learning_rate": 0.00029999936075702236,
      "loss": 15.192759704589843,
      "step": 1100
    },
    {
      "epoch": 0.03849555220309005,
      "grad_norm": 6.765523910522461,
      "learning_rate": 0.0002999992142387929,
      "loss": 15.362620544433593,
      "step": 1110
    },
    {
      "epoch": 0.038842358979694465,
      "grad_norm": 7.596284866333008,
      "learning_rate": 0.0002999990526156467,
      "loss": 15.27081298828125,
      "step": 1120
    },
    {
      "epoch": 0.039189165756298876,
      "grad_norm": 7.164301872253418,
      "learning_rate": 0.0002999988758876,
      "loss": 14.56300048828125,
      "step": 1130
    },
    {
      "epoch": 0.039535972532903293,
      "grad_norm": 6.729367256164551,
      "learning_rate": 0.0002999986840546706,
      "loss": 14.51199951171875,
      "step": 1140
    },
    {
      "epoch": 0.039882779309507704,
      "grad_norm": 6.131908893585205,
      "learning_rate": 0.00029999847711687785,
      "loss": 14.647183227539063,
      "step": 1150
    },
    {
      "epoch": 0.04022958608611212,
      "grad_norm": 7.322798728942871,
      "learning_rate": 0.00029999825507424254,
      "loss": 14.303775024414062,
      "step": 1160
    },
    {
      "epoch": 0.04057639286271654,
      "grad_norm": 7.279247760772705,
      "learning_rate": 0.00029999801792678706,
      "loss": 14.378622436523438,
      "step": 1170
    },
    {
      "epoch": 0.04092319963932095,
      "grad_norm": 6.741374492645264,
      "learning_rate": 0.00029999776567453526,
      "loss": 13.964057922363281,
      "step": 1180
    },
    {
      "epoch": 0.04127000641592537,
      "grad_norm": 6.461069583892822,
      "learning_rate": 0.00029999749831751256,
      "loss": 14.041261291503906,
      "step": 1190
    },
    {
      "epoch": 0.04161681319252978,
      "grad_norm": 6.411841869354248,
      "learning_rate": 0.0002999972158557459,
      "loss": 13.568597412109375,
      "step": 1200
    },
    {
      "epoch": 0.0419636199691342,
      "grad_norm": 6.214684963226318,
      "learning_rate": 0.00029999691828926366,
      "loss": 13.064254760742188,
      "step": 1210
    },
    {
      "epoch": 0.042310426745738615,
      "grad_norm": 6.115879058837891,
      "learning_rate": 0.0002999966056180959,
      "loss": 14.200587463378906,
      "step": 1220
    },
    {
      "epoch": 0.042657233522343026,
      "grad_norm": 6.505321979522705,
      "learning_rate": 0.00029999627784227406,
      "loss": 13.873959350585938,
      "step": 1230
    },
    {
      "epoch": 0.043004040298947444,
      "grad_norm": 6.690847873687744,
      "learning_rate": 0.0002999959349618311,
      "loss": 13.29089813232422,
      "step": 1240
    },
    {
      "epoch": 0.043350847075551854,
      "grad_norm": 6.375129699707031,
      "learning_rate": 0.0002999955769768017,
      "loss": 13.220611572265625,
      "step": 1250
    },
    {
      "epoch": 0.04369765385215627,
      "grad_norm": 5.927398204803467,
      "learning_rate": 0.00029999520388722167,
      "loss": 13.184442138671875,
      "step": 1260
    },
    {
      "epoch": 0.04404446062876068,
      "grad_norm": 6.170562267303467,
      "learning_rate": 0.00029999481569312875,
      "loss": 12.594881439208985,
      "step": 1270
    },
    {
      "epoch": 0.0443912674053651,
      "grad_norm": 7.415892601013184,
      "learning_rate": 0.000299994412394562,
      "loss": 12.665945434570313,
      "step": 1280
    },
    {
      "epoch": 0.04473807418196952,
      "grad_norm": 6.139370441436768,
      "learning_rate": 0.0002999939939915621,
      "loss": 12.897845458984374,
      "step": 1290
    },
    {
      "epoch": 0.04508488095857393,
      "grad_norm": 5.625839710235596,
      "learning_rate": 0.0002999935604841711,
      "loss": 13.06021728515625,
      "step": 1300
    },
    {
      "epoch": 0.04543168773517835,
      "grad_norm": 7.4519877433776855,
      "learning_rate": 0.00029999311187243257,
      "loss": 12.539090728759765,
      "step": 1310
    },
    {
      "epoch": 0.04577849451178276,
      "grad_norm": 5.8146071434021,
      "learning_rate": 0.0002999926481563918,
      "loss": 12.719709014892578,
      "step": 1320
    },
    {
      "epoch": 0.046125301288387176,
      "grad_norm": 6.683107376098633,
      "learning_rate": 0.0002999921693360955,
      "loss": 13.211663818359375,
      "step": 1330
    },
    {
      "epoch": 0.04647210806499159,
      "grad_norm": 5.467896938323975,
      "learning_rate": 0.00029999167541159183,
      "loss": 11.408037567138672,
      "step": 1340
    },
    {
      "epoch": 0.046818914841596004,
      "grad_norm": 9.282605171203613,
      "learning_rate": 0.00029999116638293054,
      "loss": 12.204759979248047,
      "step": 1350
    },
    {
      "epoch": 0.04716572161820042,
      "grad_norm": 6.287093162536621,
      "learning_rate": 0.00029999064225016297,
      "loss": 12.018901062011718,
      "step": 1360
    },
    {
      "epoch": 0.04751252839480483,
      "grad_norm": 5.694602012634277,
      "learning_rate": 0.00029999010301334176,
      "loss": 11.694412231445312,
      "step": 1370
    },
    {
      "epoch": 0.04785933517140925,
      "grad_norm": 5.267282962799072,
      "learning_rate": 0.00029998954867252133,
      "loss": 11.589154052734376,
      "step": 1380
    },
    {
      "epoch": 0.04820614194801366,
      "grad_norm": 6.381001949310303,
      "learning_rate": 0.00029998897922775737,
      "loss": 11.301937103271484,
      "step": 1390
    },
    {
      "epoch": 0.04855294872461808,
      "grad_norm": 4.900434970855713,
      "learning_rate": 0.0002999883946791074,
      "loss": 11.892779541015624,
      "step": 1400
    },
    {
      "epoch": 0.0488997555012225,
      "grad_norm": 6.233642578125,
      "learning_rate": 0.00029998779502663013,
      "loss": 11.379587554931641,
      "step": 1410
    },
    {
      "epoch": 0.04924656227782691,
      "grad_norm": 4.936466693878174,
      "learning_rate": 0.000299987180270386,
      "loss": 11.712295532226562,
      "step": 1420
    },
    {
      "epoch": 0.049593369054431326,
      "grad_norm": 4.68892240524292,
      "learning_rate": 0.000299986550410437,
      "loss": 11.267239379882813,
      "step": 1430
    },
    {
      "epoch": 0.04994017583103574,
      "grad_norm": 4.828918933868408,
      "learning_rate": 0.00029998590544684636,
      "loss": 11.923808288574218,
      "step": 1440
    },
    {
      "epoch": 0.050286982607640154,
      "grad_norm": 5.1969380378723145,
      "learning_rate": 0.0002999852453796792,
      "loss": 11.340673828125,
      "step": 1450
    },
    {
      "epoch": 0.050633789384244565,
      "grad_norm": 5.5531325340271,
      "learning_rate": 0.00029998457020900197,
      "loss": 11.629721069335938,
      "step": 1460
    },
    {
      "epoch": 0.05098059616084898,
      "grad_norm": 4.863478183746338,
      "learning_rate": 0.0002999838799348826,
      "loss": 10.981313323974609,
      "step": 1470
    },
    {
      "epoch": 0.0513274029374534,
      "grad_norm": 5.0770368576049805,
      "learning_rate": 0.0002999831745573906,
      "loss": 10.801561737060547,
      "step": 1480
    },
    {
      "epoch": 0.05167420971405781,
      "grad_norm": 4.6786088943481445,
      "learning_rate": 0.00029998245407659703,
      "loss": 11.578643798828125,
      "step": 1490
    },
    {
      "epoch": 0.05202101649066223,
      "grad_norm": 5.357450485229492,
      "learning_rate": 0.00029998171849257446,
      "loss": 11.369083404541016,
      "step": 1500
    },
    {
      "epoch": 0.05236782326726664,
      "grad_norm": 5.213536739349365,
      "learning_rate": 0.00029998096780539695,
      "loss": 11.568294525146484,
      "step": 1510
    },
    {
      "epoch": 0.05271463004387106,
      "grad_norm": 5.254374027252197,
      "learning_rate": 0.00029998020201514006,
      "loss": 10.894474029541016,
      "step": 1520
    },
    {
      "epoch": 0.05306143682047547,
      "grad_norm": 4.989107131958008,
      "learning_rate": 0.00029997942112188093,
      "loss": 11.209683990478515,
      "step": 1530
    },
    {
      "epoch": 0.05340824359707989,
      "grad_norm": 4.966851234436035,
      "learning_rate": 0.00029997862512569823,
      "loss": 10.634807586669922,
      "step": 1540
    },
    {
      "epoch": 0.053755050373684304,
      "grad_norm": 4.828151702880859,
      "learning_rate": 0.000299977814026672,
      "loss": 11.01966552734375,
      "step": 1550
    },
    {
      "epoch": 0.054101857150288715,
      "grad_norm": 4.8093953132629395,
      "learning_rate": 0.0002999769878248841,
      "loss": 10.188558959960938,
      "step": 1560
    },
    {
      "epoch": 0.05444866392689313,
      "grad_norm": 4.421850204467773,
      "learning_rate": 0.00029997614652041763,
      "loss": 10.327834320068359,
      "step": 1570
    },
    {
      "epoch": 0.054795470703497544,
      "grad_norm": 5.034534931182861,
      "learning_rate": 0.00029997529011335723,
      "loss": 10.527660369873047,
      "step": 1580
    },
    {
      "epoch": 0.05514227748010196,
      "grad_norm": 5.502425670623779,
      "learning_rate": 0.00029997441860378927,
      "loss": 10.757402038574218,
      "step": 1590
    },
    {
      "epoch": 0.05548908425670638,
      "grad_norm": 8.620652198791504,
      "learning_rate": 0.00029997353199180145,
      "loss": 10.318006896972657,
      "step": 1600
    },
    {
      "epoch": 0.05583589103331079,
      "grad_norm": 4.712747573852539,
      "learning_rate": 0.00029997263027748306,
      "loss": 10.284696197509765,
      "step": 1610
    },
    {
      "epoch": 0.05618269780991521,
      "grad_norm": 4.792940139770508,
      "learning_rate": 0.00029997171346092495,
      "loss": 9.87569808959961,
      "step": 1620
    },
    {
      "epoch": 0.05652950458651962,
      "grad_norm": 4.812749862670898,
      "learning_rate": 0.00029997078154221933,
      "loss": 10.548392486572265,
      "step": 1630
    },
    {
      "epoch": 0.05687631136312404,
      "grad_norm": 5.6575422286987305,
      "learning_rate": 0.0002999698345214601,
      "loss": 10.364289855957031,
      "step": 1640
    },
    {
      "epoch": 0.05722311813972845,
      "grad_norm": 4.242788314819336,
      "learning_rate": 0.00029996887239874267,
      "loss": 10.221109008789062,
      "step": 1650
    },
    {
      "epoch": 0.057569924916332865,
      "grad_norm": 4.478567600250244,
      "learning_rate": 0.0002999678951741639,
      "loss": 9.714788818359375,
      "step": 1660
    },
    {
      "epoch": 0.05791673169293728,
      "grad_norm": 5.814032077789307,
      "learning_rate": 0.0002999669028478222,
      "loss": 10.347315216064453,
      "step": 1670
    },
    {
      "epoch": 0.058263538469541694,
      "grad_norm": 4.448582649230957,
      "learning_rate": 0.00029996589541981747,
      "loss": 10.337659454345703,
      "step": 1680
    },
    {
      "epoch": 0.05861034524614611,
      "grad_norm": 4.868472576141357,
      "learning_rate": 0.00029996487289025113,
      "loss": 10.064467620849609,
      "step": 1690
    },
    {
      "epoch": 0.05895715202275052,
      "grad_norm": 4.490790843963623,
      "learning_rate": 0.0002999638352592263,
      "loss": 9.652999877929688,
      "step": 1700
    },
    {
      "epoch": 0.05930395879935494,
      "grad_norm": 4.921995639801025,
      "learning_rate": 0.0002999627825268473,
      "loss": 9.934603881835937,
      "step": 1710
    },
    {
      "epoch": 0.05965076557595935,
      "grad_norm": 4.262001037597656,
      "learning_rate": 0.0002999617146932202,
      "loss": 9.353959655761718,
      "step": 1720
    },
    {
      "epoch": 0.05999757235256377,
      "grad_norm": 4.506113052368164,
      "learning_rate": 0.0002999606317584525,
      "loss": 9.367766571044921,
      "step": 1730
    },
    {
      "epoch": 0.06034437912916819,
      "grad_norm": 4.591375827789307,
      "learning_rate": 0.00029995953372265335,
      "loss": 9.564098358154297,
      "step": 1740
    },
    {
      "epoch": 0.0606911859057726,
      "grad_norm": 4.821197509765625,
      "learning_rate": 0.00029995842058593323,
      "loss": 10.153470611572265,
      "step": 1750
    },
    {
      "epoch": 0.061037992682377015,
      "grad_norm": 4.2808308601379395,
      "learning_rate": 0.0002999572923484042,
      "loss": 9.919622802734375,
      "step": 1760
    },
    {
      "epoch": 0.061384799458981426,
      "grad_norm": 4.291468143463135,
      "learning_rate": 0.0002999561490101801,
      "loss": 9.118783569335937,
      "step": 1770
    },
    {
      "epoch": 0.061731606235585844,
      "grad_norm": 4.747788906097412,
      "learning_rate": 0.00029995499057137575,
      "loss": 9.513807678222657,
      "step": 1780
    },
    {
      "epoch": 0.06207841301219026,
      "grad_norm": 4.285642147064209,
      "learning_rate": 0.000299953817032108,
      "loss": 9.9878173828125,
      "step": 1790
    },
    {
      "epoch": 0.06242521978879467,
      "grad_norm": 4.473206520080566,
      "learning_rate": 0.00029995262839249497,
      "loss": 9.633001708984375,
      "step": 1800
    },
    {
      "epoch": 0.06277202656539908,
      "grad_norm": 4.4688005447387695,
      "learning_rate": 0.00029995142465265637,
      "loss": 9.59437255859375,
      "step": 1810
    },
    {
      "epoch": 0.06311883334200351,
      "grad_norm": 4.453664779663086,
      "learning_rate": 0.0002999502058127134,
      "loss": 8.898344421386719,
      "step": 1820
    },
    {
      "epoch": 0.06346564011860792,
      "grad_norm": 4.2538275718688965,
      "learning_rate": 0.0002999489718727888,
      "loss": 9.250914764404296,
      "step": 1830
    },
    {
      "epoch": 0.06381244689521233,
      "grad_norm": 4.033658981323242,
      "learning_rate": 0.0002999477228330069,
      "loss": 9.316190338134765,
      "step": 1840
    },
    {
      "epoch": 0.06415925367181674,
      "grad_norm": 4.747572898864746,
      "learning_rate": 0.00029994645869349334,
      "loss": 8.979618072509766,
      "step": 1850
    },
    {
      "epoch": 0.06450606044842117,
      "grad_norm": 3.9432504177093506,
      "learning_rate": 0.0002999451794543755,
      "loss": 9.21793441772461,
      "step": 1860
    },
    {
      "epoch": 0.06485286722502558,
      "grad_norm": 4.528207302093506,
      "learning_rate": 0.0002999438851157822,
      "loss": 9.081039428710938,
      "step": 1870
    },
    {
      "epoch": 0.06519967400162999,
      "grad_norm": 4.3286542892456055,
      "learning_rate": 0.0002999425756778438,
      "loss": 9.034320831298828,
      "step": 1880
    },
    {
      "epoch": 0.06554648077823441,
      "grad_norm": 5.157027721405029,
      "learning_rate": 0.0002999412511406921,
      "loss": 8.857444763183594,
      "step": 1890
    },
    {
      "epoch": 0.06589328755483882,
      "grad_norm": 4.572315692901611,
      "learning_rate": 0.00029993991150446055,
      "loss": 9.055838012695313,
      "step": 1900
    },
    {
      "epoch": 0.06624009433144323,
      "grad_norm": 4.572640419006348,
      "learning_rate": 0.00029993855676928397,
      "loss": 9.332460021972656,
      "step": 1910
    },
    {
      "epoch": 0.06658690110804766,
      "grad_norm": 5.196224212646484,
      "learning_rate": 0.00029993718693529885,
      "loss": 8.875639343261719,
      "step": 1920
    },
    {
      "epoch": 0.06693370788465207,
      "grad_norm": 4.320059299468994,
      "learning_rate": 0.0002999358020026431,
      "loss": 8.493698120117188,
      "step": 1930
    },
    {
      "epoch": 0.06728051466125648,
      "grad_norm": 5.197443008422852,
      "learning_rate": 0.0002999344019714562,
      "loss": 8.634622192382812,
      "step": 1940
    },
    {
      "epoch": 0.06762732143786089,
      "grad_norm": 5.580541133880615,
      "learning_rate": 0.0002999329868418791,
      "loss": 9.131027984619141,
      "step": 1950
    },
    {
      "epoch": 0.06797412821446532,
      "grad_norm": 4.569746017456055,
      "learning_rate": 0.00029993155661405435,
      "loss": 8.806238555908203,
      "step": 1960
    },
    {
      "epoch": 0.06832093499106973,
      "grad_norm": 5.034822940826416,
      "learning_rate": 0.00029993011128812597,
      "loss": 8.735447692871094,
      "step": 1970
    },
    {
      "epoch": 0.06866774176767414,
      "grad_norm": 5.717297077178955,
      "learning_rate": 0.00029992865086423944,
      "loss": 8.358311462402344,
      "step": 1980
    },
    {
      "epoch": 0.06901454854427856,
      "grad_norm": 5.722506046295166,
      "learning_rate": 0.00029992717534254194,
      "loss": 8.968340301513672,
      "step": 1990
    },
    {
      "epoch": 0.06936135532088297,
      "grad_norm": 4.75577974319458,
      "learning_rate": 0.0002999256847231819,
      "loss": 8.24468765258789,
      "step": 2000
    },
    {
      "epoch": 0.06970816209748738,
      "grad_norm": 4.7462029457092285,
      "learning_rate": 0.00029992417900630953,
      "loss": 9.001568603515626,
      "step": 2010
    },
    {
      "epoch": 0.0700549688740918,
      "grad_norm": 4.48958158493042,
      "learning_rate": 0.00029992265819207646,
      "loss": 9.029542541503906,
      "step": 2020
    },
    {
      "epoch": 0.07040177565069622,
      "grad_norm": 4.929230213165283,
      "learning_rate": 0.0002999211222806358,
      "loss": 8.544095611572265,
      "step": 2030
    },
    {
      "epoch": 0.07074858242730063,
      "grad_norm": 5.174428939819336,
      "learning_rate": 0.0002999195712721422,
      "loss": 9.152230834960937,
      "step": 2040
    },
    {
      "epoch": 0.07109538920390504,
      "grad_norm": 5.165338516235352,
      "learning_rate": 0.00029991800516675194,
      "loss": 8.47047119140625,
      "step": 2050
    },
    {
      "epoch": 0.07144219598050947,
      "grad_norm": 4.739821910858154,
      "learning_rate": 0.00029991642396462264,
      "loss": 8.763507080078124,
      "step": 2060
    },
    {
      "epoch": 0.07178900275711388,
      "grad_norm": 5.1035637855529785,
      "learning_rate": 0.00029991482766591354,
      "loss": 8.246512603759765,
      "step": 2070
    },
    {
      "epoch": 0.07213580953371829,
      "grad_norm": 5.471373558044434,
      "learning_rate": 0.0002999132162707854,
      "loss": 8.864594268798829,
      "step": 2080
    },
    {
      "epoch": 0.0724826163103227,
      "grad_norm": 5.771513938903809,
      "learning_rate": 0.0002999115897794005,
      "loss": 8.889707946777344,
      "step": 2090
    },
    {
      "epoch": 0.07282942308692712,
      "grad_norm": 5.27165412902832,
      "learning_rate": 0.00029990994819192255,
      "loss": 8.567472839355469,
      "step": 2100
    },
    {
      "epoch": 0.07317622986353153,
      "grad_norm": 4.975783824920654,
      "learning_rate": 0.00029990829150851697,
      "loss": 8.457414245605468,
      "step": 2110
    },
    {
      "epoch": 0.07352303664013594,
      "grad_norm": 6.3124680519104,
      "learning_rate": 0.00029990661972935053,
      "loss": 8.194977569580079,
      "step": 2120
    },
    {
      "epoch": 0.07386984341674037,
      "grad_norm": 4.776012897491455,
      "learning_rate": 0.00029990493285459157,
      "loss": 8.464956665039063,
      "step": 2130
    },
    {
      "epoch": 0.07421665019334478,
      "grad_norm": 4.533716678619385,
      "learning_rate": 0.00029990323088440997,
      "loss": 8.268666839599609,
      "step": 2140
    },
    {
      "epoch": 0.07456345696994919,
      "grad_norm": 5.62864875793457,
      "learning_rate": 0.00029990151381897715,
      "loss": 7.639604949951172,
      "step": 2150
    },
    {
      "epoch": 0.0749102637465536,
      "grad_norm": 4.756488800048828,
      "learning_rate": 0.000299899781658466,
      "loss": 8.60247573852539,
      "step": 2160
    },
    {
      "epoch": 0.07525707052315803,
      "grad_norm": 5.337765693664551,
      "learning_rate": 0.00029989803440305093,
      "loss": 8.40919189453125,
      "step": 2170
    },
    {
      "epoch": 0.07560387729976244,
      "grad_norm": 4.989221572875977,
      "learning_rate": 0.0002998962720529079,
      "loss": 8.433936309814452,
      "step": 2180
    },
    {
      "epoch": 0.07595068407636685,
      "grad_norm": 5.895046710968018,
      "learning_rate": 0.00029989449460821434,
      "loss": 8.606028747558593,
      "step": 2190
    },
    {
      "epoch": 0.07629749085297127,
      "grad_norm": 4.604063034057617,
      "learning_rate": 0.0002998927020691493,
      "loss": 8.863967132568359,
      "step": 2200
    },
    {
      "epoch": 0.07664429762957568,
      "grad_norm": 4.635347366333008,
      "learning_rate": 0.0002998908944358932,
      "loss": 8.632648468017578,
      "step": 2210
    },
    {
      "epoch": 0.0769911044061801,
      "grad_norm": 5.013646125793457,
      "learning_rate": 0.0002998890717086282,
      "loss": 8.241293334960938,
      "step": 2220
    },
    {
      "epoch": 0.0773379111827845,
      "grad_norm": 5.209301948547363,
      "learning_rate": 0.0002998872338875378,
      "loss": 8.100372314453125,
      "step": 2230
    },
    {
      "epoch": 0.07768471795938893,
      "grad_norm": 5.827475547790527,
      "learning_rate": 0.00029988538097280696,
      "loss": 7.576439666748047,
      "step": 2240
    },
    {
      "epoch": 0.07803152473599334,
      "grad_norm": 5.064399242401123,
      "learning_rate": 0.00029988351296462247,
      "loss": 8.33707275390625,
      "step": 2250
    },
    {
      "epoch": 0.07837833151259775,
      "grad_norm": 5.7265706062316895,
      "learning_rate": 0.0002998816298631723,
      "loss": 8.343417358398437,
      "step": 2260
    },
    {
      "epoch": 0.07872513828920218,
      "grad_norm": 5.61683988571167,
      "learning_rate": 0.00029987973166864605,
      "loss": 7.654642486572266,
      "step": 2270
    },
    {
      "epoch": 0.07907194506580659,
      "grad_norm": 6.139464855194092,
      "learning_rate": 0.0002998778183812349,
      "loss": 8.174369812011719,
      "step": 2280
    },
    {
      "epoch": 0.079418751842411,
      "grad_norm": 5.3133225440979,
      "learning_rate": 0.0002998758900011316,
      "loss": 8.249425506591797,
      "step": 2290
    },
    {
      "epoch": 0.07976555861901541,
      "grad_norm": 5.132144927978516,
      "learning_rate": 0.0002998739465285303,
      "loss": 8.233519744873046,
      "step": 2300
    },
    {
      "epoch": 0.08011236539561983,
      "grad_norm": 6.415826320648193,
      "learning_rate": 0.00029987198796362666,
      "loss": 8.221486663818359,
      "step": 2310
    },
    {
      "epoch": 0.08045917217222424,
      "grad_norm": 5.217477798461914,
      "learning_rate": 0.0002998700143066179,
      "loss": 8.524445343017579,
      "step": 2320
    },
    {
      "epoch": 0.08080597894882866,
      "grad_norm": 5.741121292114258,
      "learning_rate": 0.0002998680255577028,
      "loss": 8.453343200683594,
      "step": 2330
    },
    {
      "epoch": 0.08115278572543308,
      "grad_norm": 6.204251289367676,
      "learning_rate": 0.0002998660217170816,
      "loss": 8.372715759277344,
      "step": 2340
    },
    {
      "epoch": 0.08149959250203749,
      "grad_norm": 7.696920394897461,
      "learning_rate": 0.00029986400278495624,
      "loss": 8.396324157714844,
      "step": 2350
    },
    {
      "epoch": 0.0818463992786419,
      "grad_norm": 6.234743118286133,
      "learning_rate": 0.00029986196876152977,
      "loss": 8.096724700927734,
      "step": 2360
    },
    {
      "epoch": 0.08219320605524633,
      "grad_norm": 4.549152374267578,
      "learning_rate": 0.0002998599196470072,
      "loss": 7.865915679931641,
      "step": 2370
    },
    {
      "epoch": 0.08254001283185074,
      "grad_norm": 5.453365325927734,
      "learning_rate": 0.00029985785544159484,
      "loss": 7.964163970947266,
      "step": 2380
    },
    {
      "epoch": 0.08288681960845515,
      "grad_norm": 4.967399597167969,
      "learning_rate": 0.0002998557761455005,
      "loss": 7.947079467773437,
      "step": 2390
    },
    {
      "epoch": 0.08323362638505956,
      "grad_norm": 6.34585428237915,
      "learning_rate": 0.00029985368175893364,
      "loss": 8.035052490234374,
      "step": 2400
    },
    {
      "epoch": 0.08358043316166398,
      "grad_norm": 5.457056999206543,
      "learning_rate": 0.0002998515722821051,
      "loss": 7.74532470703125,
      "step": 2410
    },
    {
      "epoch": 0.0839272399382684,
      "grad_norm": 4.875389099121094,
      "learning_rate": 0.00029984944771522734,
      "loss": 7.901406860351562,
      "step": 2420
    },
    {
      "epoch": 0.0842740467148728,
      "grad_norm": 5.598247051239014,
      "learning_rate": 0.00029984730805851425,
      "loss": 8.011869812011719,
      "step": 2430
    },
    {
      "epoch": 0.08462085349147723,
      "grad_norm": 5.847703456878662,
      "learning_rate": 0.0002998451533121814,
      "loss": 8.203024291992188,
      "step": 2440
    },
    {
      "epoch": 0.08496766026808164,
      "grad_norm": 7.487412452697754,
      "learning_rate": 0.00029984298347644566,
      "loss": 7.6920112609863285,
      "step": 2450
    },
    {
      "epoch": 0.08531446704468605,
      "grad_norm": 6.7232255935668945,
      "learning_rate": 0.0002998407985515257,
      "loss": 7.558551788330078,
      "step": 2460
    },
    {
      "epoch": 0.08566127382129046,
      "grad_norm": 7.134231090545654,
      "learning_rate": 0.0002998385985376413,
      "loss": 7.462080383300782,
      "step": 2470
    },
    {
      "epoch": 0.08600808059789489,
      "grad_norm": 5.4149909019470215,
      "learning_rate": 0.00029983638343501415,
      "loss": 8.575581359863282,
      "step": 2480
    },
    {
      "epoch": 0.0863548873744993,
      "grad_norm": 6.803240776062012,
      "learning_rate": 0.0002998341532438674,
      "loss": 7.64642333984375,
      "step": 2490
    },
    {
      "epoch": 0.08670169415110371,
      "grad_norm": 5.699243068695068,
      "learning_rate": 0.0002998319079644254,
      "loss": 7.960725402832031,
      "step": 2500
    },
    {
      "epoch": 0.08704850092770813,
      "grad_norm": 6.425755500793457,
      "learning_rate": 0.00029982964759691435,
      "loss": 7.6616157531738285,
      "step": 2510
    },
    {
      "epoch": 0.08739530770431254,
      "grad_norm": 5.376076698303223,
      "learning_rate": 0.00029982737214156194,
      "loss": 7.7346656799316404,
      "step": 2520
    },
    {
      "epoch": 0.08774211448091696,
      "grad_norm": 6.2727484703063965,
      "learning_rate": 0.0002998250815985972,
      "loss": 7.836188507080078,
      "step": 2530
    },
    {
      "epoch": 0.08808892125752137,
      "grad_norm": 5.631091117858887,
      "learning_rate": 0.0002998227759682509,
      "loss": 7.705145263671875,
      "step": 2540
    },
    {
      "epoch": 0.08843572803412579,
      "grad_norm": 6.217864036560059,
      "learning_rate": 0.00029982045525075513,
      "loss": 8.142205047607423,
      "step": 2550
    },
    {
      "epoch": 0.0887825348107302,
      "grad_norm": 5.994324684143066,
      "learning_rate": 0.00029981811944634364,
      "loss": 7.400997924804687,
      "step": 2560
    },
    {
      "epoch": 0.08912934158733461,
      "grad_norm": 4.685929298400879,
      "learning_rate": 0.00029981576855525165,
      "loss": 8.070906829833984,
      "step": 2570
    },
    {
      "epoch": 0.08947614836393904,
      "grad_norm": 5.380947589874268,
      "learning_rate": 0.00029981340257771586,
      "loss": 7.440650177001953,
      "step": 2580
    },
    {
      "epoch": 0.08982295514054345,
      "grad_norm": 5.548213005065918,
      "learning_rate": 0.00029981102151397445,
      "loss": 7.631075286865235,
      "step": 2590
    },
    {
      "epoch": 0.09016976191714786,
      "grad_norm": 7.113351821899414,
      "learning_rate": 0.00029980862536426736,
      "loss": 7.411952972412109,
      "step": 2600
    },
    {
      "epoch": 0.09051656869375227,
      "grad_norm": 4.830175399780273,
      "learning_rate": 0.0002998062141288357,
      "loss": 8.161653900146485,
      "step": 2610
    },
    {
      "epoch": 0.0908633754703567,
      "grad_norm": 5.5586018562316895,
      "learning_rate": 0.00029980378780792243,
      "loss": 7.329508209228516,
      "step": 2620
    },
    {
      "epoch": 0.0912101822469611,
      "grad_norm": 6.112547397613525,
      "learning_rate": 0.0002998013464017719,
      "loss": 7.714527893066406,
      "step": 2630
    },
    {
      "epoch": 0.09155698902356552,
      "grad_norm": 4.658982276916504,
      "learning_rate": 0.00029979888991062977,
      "loss": 7.437145233154297,
      "step": 2640
    },
    {
      "epoch": 0.09190379580016994,
      "grad_norm": 6.236598014831543,
      "learning_rate": 0.0002997964183347436,
      "loss": 7.697496795654297,
      "step": 2650
    },
    {
      "epoch": 0.09225060257677435,
      "grad_norm": 4.987429618835449,
      "learning_rate": 0.00029979393167436214,
      "loss": 7.137907409667969,
      "step": 2660
    },
    {
      "epoch": 0.09259740935337876,
      "grad_norm": 7.033418655395508,
      "learning_rate": 0.0002997914299297359,
      "loss": 7.323359680175781,
      "step": 2670
    },
    {
      "epoch": 0.09294421612998317,
      "grad_norm": 5.532907009124756,
      "learning_rate": 0.00029978891310111674,
      "loss": 7.454933166503906,
      "step": 2680
    },
    {
      "epoch": 0.0932910229065876,
      "grad_norm": 6.004257678985596,
      "learning_rate": 0.00029978638118875815,
      "loss": 7.570569610595703,
      "step": 2690
    },
    {
      "epoch": 0.09363782968319201,
      "grad_norm": 6.463276386260986,
      "learning_rate": 0.0002997838341929151,
      "loss": 7.730841827392578,
      "step": 2700
    },
    {
      "epoch": 0.09398463645979642,
      "grad_norm": 6.122608184814453,
      "learning_rate": 0.000299781272113844,
      "loss": 7.185985565185547,
      "step": 2710
    },
    {
      "epoch": 0.09433144323640084,
      "grad_norm": 5.712291717529297,
      "learning_rate": 0.00029977869495180287,
      "loss": 8.223866271972657,
      "step": 2720
    },
    {
      "epoch": 0.09467825001300526,
      "grad_norm": 6.126311302185059,
      "learning_rate": 0.00029977610270705133,
      "loss": 7.804103088378906,
      "step": 2730
    },
    {
      "epoch": 0.09502505678960967,
      "grad_norm": 4.903704643249512,
      "learning_rate": 0.0002997734953798503,
      "loss": 7.276256561279297,
      "step": 2740
    },
    {
      "epoch": 0.09537186356621409,
      "grad_norm": 6.626465797424316,
      "learning_rate": 0.0002997708729704624,
      "loss": 7.943316650390625,
      "step": 2750
    },
    {
      "epoch": 0.0957186703428185,
      "grad_norm": 6.401085376739502,
      "learning_rate": 0.0002997682354791517,
      "loss": 7.571746826171875,
      "step": 2760
    },
    {
      "epoch": 0.09606547711942291,
      "grad_norm": 6.731732368469238,
      "learning_rate": 0.00029976558290618373,
      "loss": 8.095571136474609,
      "step": 2770
    },
    {
      "epoch": 0.09641228389602732,
      "grad_norm": 6.18290901184082,
      "learning_rate": 0.00029976291525182574,
      "loss": 7.459332275390625,
      "step": 2780
    },
    {
      "epoch": 0.09675909067263175,
      "grad_norm": 4.505118370056152,
      "learning_rate": 0.00029976023251634623,
      "loss": 8.029454040527344,
      "step": 2790
    },
    {
      "epoch": 0.09710589744923616,
      "grad_norm": 7.047544479370117,
      "learning_rate": 0.00029975753470001543,
      "loss": 7.250439453125,
      "step": 2800
    },
    {
      "epoch": 0.09745270422584057,
      "grad_norm": 7.906260967254639,
      "learning_rate": 0.00029975482180310495,
      "loss": 7.352965545654297,
      "step": 2810
    },
    {
      "epoch": 0.097799511002445,
      "grad_norm": 5.261754989624023,
      "learning_rate": 0.00029975209382588806,
      "loss": 7.319126892089844,
      "step": 2820
    },
    {
      "epoch": 0.0981463177790494,
      "grad_norm": 5.247056007385254,
      "learning_rate": 0.0002997493507686394,
      "loss": 7.372547912597656,
      "step": 2830
    },
    {
      "epoch": 0.09849312455565382,
      "grad_norm": 5.888732433319092,
      "learning_rate": 0.00029974659263163516,
      "loss": 7.207266235351563,
      "step": 2840
    },
    {
      "epoch": 0.09883993133225823,
      "grad_norm": 6.898653030395508,
      "learning_rate": 0.0002997438194151532,
      "loss": 7.474554443359375,
      "step": 2850
    },
    {
      "epoch": 0.09918673810886265,
      "grad_norm": 5.344205379486084,
      "learning_rate": 0.0002997410311194727,
      "loss": 7.531422424316406,
      "step": 2860
    },
    {
      "epoch": 0.09953354488546706,
      "grad_norm": 8.75732421875,
      "learning_rate": 0.00029973822774487447,
      "loss": 7.705049896240235,
      "step": 2870
    },
    {
      "epoch": 0.09988035166207147,
      "grad_norm": 5.712015151977539,
      "learning_rate": 0.0002997354092916408,
      "loss": 7.514460754394531,
      "step": 2880
    },
    {
      "epoch": 0.1002271584386759,
      "grad_norm": 6.071193218231201,
      "learning_rate": 0.0002997325757600555,
      "loss": 8.2220947265625,
      "step": 2890
    },
    {
      "epoch": 0.10057396521528031,
      "grad_norm": 4.527872085571289,
      "learning_rate": 0.00029972972715040395,
      "loss": 7.812737274169922,
      "step": 2900
    },
    {
      "epoch": 0.10092077199188472,
      "grad_norm": 5.524644374847412,
      "learning_rate": 0.00029972686346297294,
      "loss": 7.216932678222657,
      "step": 2910
    },
    {
      "epoch": 0.10126757876848913,
      "grad_norm": 5.723361015319824,
      "learning_rate": 0.0002997239846980509,
      "loss": 7.164757537841797,
      "step": 2920
    },
    {
      "epoch": 0.10161438554509356,
      "grad_norm": 6.528432369232178,
      "learning_rate": 0.0002997210908559277,
      "loss": 7.451429748535157,
      "step": 2930
    },
    {
      "epoch": 0.10196119232169797,
      "grad_norm": 6.940042018890381,
      "learning_rate": 0.00029971818193689473,
      "loss": 7.0673583984375,
      "step": 2940
    },
    {
      "epoch": 0.10230799909830238,
      "grad_norm": 5.519460678100586,
      "learning_rate": 0.00029971525794124495,
      "loss": 7.4651947021484375,
      "step": 2950
    },
    {
      "epoch": 0.1026548058749068,
      "grad_norm": 5.205141544342041,
      "learning_rate": 0.0002997123188692728,
      "loss": 7.1986549377441404,
      "step": 2960
    },
    {
      "epoch": 0.10300161265151121,
      "grad_norm": 4.912453651428223,
      "learning_rate": 0.0002997093647212742,
      "loss": 7.043345642089844,
      "step": 2970
    },
    {
      "epoch": 0.10334841942811562,
      "grad_norm": 5.264575481414795,
      "learning_rate": 0.00029970639549754667,
      "loss": 7.150616455078125,
      "step": 2980
    },
    {
      "epoch": 0.10369522620472003,
      "grad_norm": 8.516182899475098,
      "learning_rate": 0.0002997034111983892,
      "loss": 6.536961364746094,
      "step": 2990
    },
    {
      "epoch": 0.10404203298132446,
      "grad_norm": 5.379635810852051,
      "learning_rate": 0.00029970041182410235,
      "loss": 7.469226837158203,
      "step": 3000
    }
  ],
  "logging_steps": 10,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.8659740483584e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}