{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2422,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041288191577208916,
      "grad_norm": 24.94086805103786,
      "learning_rate": 4.1152263374485605e-08,
      "loss": 1.4345,
      "step": 1
    },
    {
      "epoch": 0.002064409578860446,
      "grad_norm": 24.898038168161193,
      "learning_rate": 2.05761316872428e-07,
      "loss": 1.4101,
      "step": 5
    },
    {
      "epoch": 0.004128819157720892,
      "grad_norm": 15.938300194174698,
      "learning_rate": 4.11522633744856e-07,
      "loss": 1.3943,
      "step": 10
    },
    {
      "epoch": 0.006193228736581338,
      "grad_norm": 8.846886408597587,
      "learning_rate": 6.17283950617284e-07,
      "loss": 1.2771,
      "step": 15
    },
    {
      "epoch": 0.008257638315441783,
      "grad_norm": 9.948610282503696,
      "learning_rate": 8.23045267489712e-07,
      "loss": 1.1572,
      "step": 20
    },
    {
      "epoch": 0.01032204789430223,
      "grad_norm": 4.744946112205854,
      "learning_rate": 1.02880658436214e-06,
      "loss": 1.0575,
      "step": 25
    },
    {
      "epoch": 0.012386457473162676,
      "grad_norm": 3.3793530185545926,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.9708,
      "step": 30
    },
    {
      "epoch": 0.014450867052023121,
      "grad_norm": 3.4379702794812124,
      "learning_rate": 1.440329218106996e-06,
      "loss": 0.9668,
      "step": 35
    },
    {
      "epoch": 0.016515276630883566,
      "grad_norm": 3.110466178670075,
      "learning_rate": 1.646090534979424e-06,
      "loss": 0.9508,
      "step": 40
    },
    {
      "epoch": 0.018579686209744012,
      "grad_norm": 3.0934122499344614,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.9496,
      "step": 45
    },
    {
      "epoch": 0.02064409578860446,
      "grad_norm": 3.2046299727793572,
      "learning_rate": 2.05761316872428e-06,
      "loss": 0.9256,
      "step": 50
    },
    {
      "epoch": 0.022708505367464906,
      "grad_norm": 2.9857050496874358,
      "learning_rate": 2.263374485596708e-06,
      "loss": 0.9146,
      "step": 55
    },
    {
      "epoch": 0.02477291494632535,
      "grad_norm": 3.010697976532333,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.9185,
      "step": 60
    },
    {
      "epoch": 0.026837324525185797,
      "grad_norm": 2.9986835294430274,
      "learning_rate": 2.674897119341564e-06,
      "loss": 0.9008,
      "step": 65
    },
    {
      "epoch": 0.028901734104046242,
      "grad_norm": 3.2527007862069186,
      "learning_rate": 2.880658436213992e-06,
      "loss": 0.8855,
      "step": 70
    },
    {
      "epoch": 0.030966143682906688,
      "grad_norm": 3.2007644630686833,
      "learning_rate": 3.08641975308642e-06,
      "loss": 0.8783,
      "step": 75
    },
    {
      "epoch": 0.03303055326176713,
      "grad_norm": 3.196196143650837,
      "learning_rate": 3.292181069958848e-06,
      "loss": 0.8775,
      "step": 80
    },
    {
      "epoch": 0.03509496284062758,
      "grad_norm": 3.004102067065696,
      "learning_rate": 3.4979423868312762e-06,
      "loss": 0.8808,
      "step": 85
    },
    {
      "epoch": 0.037159372419488024,
      "grad_norm": 3.118135767316795,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.8775,
      "step": 90
    },
    {
      "epoch": 0.03922378199834847,
      "grad_norm": 3.0856528450327354,
      "learning_rate": 3.909465020576132e-06,
      "loss": 0.8634,
      "step": 95
    },
    {
      "epoch": 0.04128819157720892,
      "grad_norm": 2.9700335979493167,
      "learning_rate": 4.11522633744856e-06,
      "loss": 0.872,
      "step": 100
    },
    {
      "epoch": 0.04335260115606936,
      "grad_norm": 3.1350056589977533,
      "learning_rate": 4.3209876543209875e-06,
      "loss": 0.8706,
      "step": 105
    },
    {
      "epoch": 0.04541701073492981,
      "grad_norm": 3.2497160658033852,
      "learning_rate": 4.526748971193416e-06,
      "loss": 0.8539,
      "step": 110
    },
    {
      "epoch": 0.047481420313790254,
      "grad_norm": 3.382546616433861,
      "learning_rate": 4.732510288065844e-06,
      "loss": 0.8573,
      "step": 115
    },
    {
      "epoch": 0.0495458298926507,
      "grad_norm": 3.1741955682207124,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.8433,
      "step": 120
    },
    {
      "epoch": 0.051610239471511145,
      "grad_norm": 3.13507489217465,
      "learning_rate": 5.1440329218107e-06,
      "loss": 0.8546,
      "step": 125
    },
    {
      "epoch": 0.053674649050371594,
      "grad_norm": 3.182938519722128,
      "learning_rate": 5.349794238683128e-06,
      "loss": 0.8505,
      "step": 130
    },
    {
      "epoch": 0.05573905862923204,
      "grad_norm": 3.3061978280140916,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8596,
      "step": 135
    },
    {
      "epoch": 0.057803468208092484,
      "grad_norm": 3.172304485577032,
      "learning_rate": 5.761316872427984e-06,
      "loss": 0.8454,
      "step": 140
    },
    {
      "epoch": 0.05986787778695293,
      "grad_norm": 3.222279073856588,
      "learning_rate": 5.967078189300412e-06,
      "loss": 0.8552,
      "step": 145
    },
    {
      "epoch": 0.061932287365813375,
      "grad_norm": 3.026540545603275,
      "learning_rate": 6.17283950617284e-06,
      "loss": 0.8365,
      "step": 150
    },
    {
      "epoch": 0.06399669694467382,
      "grad_norm": 3.1724013332685304,
      "learning_rate": 6.3786008230452675e-06,
      "loss": 0.8397,
      "step": 155
    },
    {
      "epoch": 0.06606110652353427,
      "grad_norm": 3.4516704362207222,
      "learning_rate": 6.584362139917696e-06,
      "loss": 0.8346,
      "step": 160
    },
    {
      "epoch": 0.06812551610239471,
      "grad_norm": 3.1199583778413285,
      "learning_rate": 6.790123456790124e-06,
      "loss": 0.8296,
      "step": 165
    },
    {
      "epoch": 0.07018992568125516,
      "grad_norm": 3.448087429701013,
      "learning_rate": 6.9958847736625525e-06,
      "loss": 0.8384,
      "step": 170
    },
    {
      "epoch": 0.07225433526011561,
      "grad_norm": 3.072968814322973,
      "learning_rate": 7.201646090534981e-06,
      "loss": 0.846,
      "step": 175
    },
    {
      "epoch": 0.07431874483897605,
      "grad_norm": 3.1647169459938076,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.8256,
      "step": 180
    },
    {
      "epoch": 0.0763831544178365,
      "grad_norm": 3.105263189462399,
      "learning_rate": 7.613168724279836e-06,
      "loss": 0.8011,
      "step": 185
    },
    {
      "epoch": 0.07844756399669695,
      "grad_norm": 2.9946543968073294,
      "learning_rate": 7.818930041152263e-06,
      "loss": 0.8214,
      "step": 190
    },
    {
      "epoch": 0.0805119735755574,
      "grad_norm": 3.25501723509638,
      "learning_rate": 8.024691358024692e-06,
      "loss": 0.8201,
      "step": 195
    },
    {
      "epoch": 0.08257638315441784,
      "grad_norm": 3.0087568189695237,
      "learning_rate": 8.23045267489712e-06,
      "loss": 0.8139,
      "step": 200
    },
    {
      "epoch": 0.08464079273327828,
      "grad_norm": 3.0911976211351817,
      "learning_rate": 8.43621399176955e-06,
      "loss": 0.8242,
      "step": 205
    },
    {
      "epoch": 0.08670520231213873,
      "grad_norm": 3.310286578361364,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.8364,
      "step": 210
    },
    {
      "epoch": 0.08876961189099918,
      "grad_norm": 2.986473336622259,
      "learning_rate": 8.847736625514404e-06,
      "loss": 0.8316,
      "step": 215
    },
    {
      "epoch": 0.09083402146985962,
      "grad_norm": 3.102015428130825,
      "learning_rate": 9.053497942386832e-06,
      "loss": 0.8128,
      "step": 220
    },
    {
      "epoch": 0.09289843104872006,
      "grad_norm": 2.906374211555821,
      "learning_rate": 9.25925925925926e-06,
      "loss": 0.7945,
      "step": 225
    },
    {
      "epoch": 0.09496284062758051,
      "grad_norm": 3.0506200713500826,
      "learning_rate": 9.465020576131688e-06,
      "loss": 0.8158,
      "step": 230
    },
    {
      "epoch": 0.09702725020644096,
      "grad_norm": 3.0384361806454017,
      "learning_rate": 9.670781893004116e-06,
      "loss": 0.8228,
      "step": 235
    },
    {
      "epoch": 0.0990916597853014,
      "grad_norm": 3.100843008960617,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.8099,
      "step": 240
    },
    {
      "epoch": 0.10115606936416185,
      "grad_norm": 2.935521796714598,
      "learning_rate": 9.999979213327006e-06,
      "loss": 0.8108,
      "step": 245
    },
    {
      "epoch": 0.10322047894302229,
      "grad_norm": 3.361172390200279,
      "learning_rate": 9.999745365240722e-06,
      "loss": 0.7923,
      "step": 250
    },
    {
      "epoch": 0.10528488852188274,
      "grad_norm": 2.915556545903911,
      "learning_rate": 9.999251697919687e-06,
      "loss": 0.8045,
      "step": 255
    },
    {
      "epoch": 0.10734929810074319,
      "grad_norm": 3.212286307862521,
      "learning_rate": 9.998498237018056e-06,
      "loss": 0.8112,
      "step": 260
    },
    {
      "epoch": 0.10941370767960364,
      "grad_norm": 3.191428090938715,
      "learning_rate": 9.997485021690553e-06,
      "loss": 0.8001,
      "step": 265
    },
    {
      "epoch": 0.11147811725846409,
      "grad_norm": 2.838764527991799,
      "learning_rate": 9.996212104590426e-06,
      "loss": 0.8074,
      "step": 270
    },
    {
      "epoch": 0.11354252683732452,
      "grad_norm": 2.9644134550569943,
      "learning_rate": 9.994679551866712e-06,
      "loss": 0.8033,
      "step": 275
    },
    {
      "epoch": 0.11560693641618497,
      "grad_norm": 3.118728479888768,
      "learning_rate": 9.992887443160804e-06,
      "loss": 0.7905,
      "step": 280
    },
    {
      "epoch": 0.11767134599504542,
      "grad_norm": 2.8425828854985737,
      "learning_rate": 9.990835871602305e-06,
      "loss": 0.7987,
      "step": 285
    },
    {
      "epoch": 0.11973575557390587,
      "grad_norm": 3.091935940986862,
      "learning_rate": 9.988524943804196e-06,
      "loss": 0.7919,
      "step": 290
    },
    {
      "epoch": 0.12180016515276632,
      "grad_norm": 2.9471270363090976,
      "learning_rate": 9.985954779857292e-06,
      "loss": 0.7885,
      "step": 295
    },
    {
      "epoch": 0.12386457473162675,
      "grad_norm": 2.761224162387467,
      "learning_rate": 9.983125513324e-06,
      "loss": 0.784,
      "step": 300
    },
    {
      "epoch": 0.1259289843104872,
      "grad_norm": 2.779726805954528,
      "learning_rate": 9.98003729123138e-06,
      "loss": 0.7762,
      "step": 305
    },
    {
      "epoch": 0.12799339388934763,
      "grad_norm": 2.8806127663541625,
      "learning_rate": 9.976690274063507e-06,
      "loss": 0.7995,
      "step": 310
    },
    {
      "epoch": 0.13005780346820808,
      "grad_norm": 3.0861309652893163,
      "learning_rate": 9.973084635753125e-06,
      "loss": 0.8002,
      "step": 315
    },
    {
      "epoch": 0.13212221304706853,
      "grad_norm": 2.886159344081087,
      "learning_rate": 9.969220563672614e-06,
      "loss": 0.7742,
      "step": 320
    },
    {
      "epoch": 0.13418662262592898,
      "grad_norm": 2.9262415184999835,
      "learning_rate": 9.965098258624253e-06,
      "loss": 0.7807,
      "step": 325
    },
    {
      "epoch": 0.13625103220478943,
      "grad_norm": 2.9922747460068324,
      "learning_rate": 9.960717934829777e-06,
      "loss": 0.7582,
      "step": 330
    },
    {
      "epoch": 0.13831544178364988,
      "grad_norm": 2.9066701618612596,
      "learning_rate": 9.956079819919254e-06,
      "loss": 0.7849,
      "step": 335
    },
    {
      "epoch": 0.14037985136251033,
      "grad_norm": 2.9577891790066766,
      "learning_rate": 9.951184154919253e-06,
      "loss": 0.7676,
      "step": 340
    },
    {
      "epoch": 0.14244426094137078,
      "grad_norm": 2.7242674149455923,
      "learning_rate": 9.946031194240317e-06,
      "loss": 0.7613,
      "step": 345
    },
    {
      "epoch": 0.14450867052023122,
      "grad_norm": 2.738187421683147,
      "learning_rate": 9.940621205663745e-06,
      "loss": 0.7476,
      "step": 350
    },
    {
      "epoch": 0.14657308009909165,
      "grad_norm": 2.959447667783807,
      "learning_rate": 9.934954470327671e-06,
      "loss": 0.757,
      "step": 355
    },
    {
      "epoch": 0.1486374896779521,
      "grad_norm": 2.9344152274329693,
      "learning_rate": 9.92903128271246e-06,
      "loss": 0.7559,
      "step": 360
    },
    {
      "epoch": 0.15070189925681254,
      "grad_norm": 3.0358566235662554,
      "learning_rate": 9.9228519506254e-06,
      "loss": 0.7671,
      "step": 365
    },
    {
      "epoch": 0.152766308835673,
      "grad_norm": 2.917481842444454,
      "learning_rate": 9.916416795184716e-06,
      "loss": 0.7423,
      "step": 370
    },
    {
      "epoch": 0.15483071841453344,
      "grad_norm": 2.858632389269523,
      "learning_rate": 9.90972615080287e-06,
      "loss": 0.7625,
      "step": 375
    },
    {
      "epoch": 0.1568951279933939,
      "grad_norm": 2.7940389389869997,
      "learning_rate": 9.902780365169186e-06,
      "loss": 0.7344,
      "step": 380
    },
    {
      "epoch": 0.15895953757225434,
      "grad_norm": 2.966296190863466,
      "learning_rate": 9.895579799231791e-06,
      "loss": 0.7431,
      "step": 385
    },
    {
      "epoch": 0.1610239471511148,
      "grad_norm": 2.734933032890328,
      "learning_rate": 9.888124827178847e-06,
      "loss": 0.7543,
      "step": 390
    },
    {
      "epoch": 0.16308835672997524,
      "grad_norm": 3.0106930594709183,
      "learning_rate": 9.880415836419108e-06,
      "loss": 0.7322,
      "step": 395
    },
    {
      "epoch": 0.16515276630883569,
      "grad_norm": 3.1160711879023277,
      "learning_rate": 9.872453227561795e-06,
      "loss": 0.7217,
      "step": 400
    },
    {
      "epoch": 0.1672171758876961,
      "grad_norm": 3.025574483100628,
      "learning_rate": 9.864237414395768e-06,
      "loss": 0.7234,
      "step": 405
    },
    {
      "epoch": 0.16928158546655656,
      "grad_norm": 3.0800133917199024,
      "learning_rate": 9.855768823868026e-06,
      "loss": 0.7451,
      "step": 410
    },
    {
      "epoch": 0.171345995045417,
      "grad_norm": 2.878566581474389,
      "learning_rate": 9.847047896061526e-06,
      "loss": 0.7479,
      "step": 415
    },
    {
      "epoch": 0.17341040462427745,
      "grad_norm": 3.168491396422218,
      "learning_rate": 9.838075084172303e-06,
      "loss": 0.7258,
      "step": 420
    },
    {
      "epoch": 0.1754748142031379,
      "grad_norm": 2.815061772988165,
      "learning_rate": 9.828850854485929e-06,
      "loss": 0.7055,
      "step": 425
    },
    {
      "epoch": 0.17753922378199835,
      "grad_norm": 2.6873037207867356,
      "learning_rate": 9.819375686353277e-06,
      "loss": 0.7253,
      "step": 430
    },
    {
      "epoch": 0.1796036333608588,
      "grad_norm": 2.8648079803750077,
      "learning_rate": 9.809650072165606e-06,
      "loss": 0.7218,
      "step": 435
    },
    {
      "epoch": 0.18166804293971925,
      "grad_norm": 2.6485478231791446,
      "learning_rate": 9.799674517328985e-06,
      "loss": 0.7211,
      "step": 440
    },
    {
      "epoch": 0.1837324525185797,
      "grad_norm": 2.9086971891718365,
      "learning_rate": 9.78944954023802e-06,
      "loss": 0.7113,
      "step": 445
    },
    {
      "epoch": 0.18579686209744012,
      "grad_norm": 2.777331603392391,
      "learning_rate": 9.778975672248909e-06,
      "loss": 0.7236,
      "step": 450
    },
    {
      "epoch": 0.18786127167630057,
      "grad_norm": 2.896862579927988,
      "learning_rate": 9.768253457651853e-06,
      "loss": 0.7239,
      "step": 455
    },
    {
      "epoch": 0.18992568125516102,
      "grad_norm": 2.788383349290103,
      "learning_rate": 9.757283453642741e-06,
      "loss": 0.7083,
      "step": 460
    },
    {
      "epoch": 0.19199009083402147,
      "grad_norm": 2.7799217861819816,
      "learning_rate": 9.746066230294216e-06,
      "loss": 0.7008,
      "step": 465
    },
    {
      "epoch": 0.19405450041288191,
      "grad_norm": 3.0828748770941603,
      "learning_rate": 9.734602370526042e-06,
      "loss": 0.6949,
      "step": 470
    },
    {
      "epoch": 0.19611890999174236,
      "grad_norm": 2.8583233928158145,
      "learning_rate": 9.722892470074811e-06,
      "loss": 0.6923,
      "step": 475
    },
    {
      "epoch": 0.1981833195706028,
      "grad_norm": 2.850888921108626,
      "learning_rate": 9.710937137462992e-06,
      "loss": 0.7056,
      "step": 480
    },
    {
      "epoch": 0.20024772914946326,
      "grad_norm": 2.8804381008997697,
      "learning_rate": 9.698736993967295e-06,
      "loss": 0.7134,
      "step": 485
    },
    {
      "epoch": 0.2023121387283237,
      "grad_norm": 2.5675849510127473,
      "learning_rate": 9.686292673586397e-06,
      "loss": 0.6854,
      "step": 490
    },
    {
      "epoch": 0.20437654830718416,
      "grad_norm": 2.752726728070582,
      "learning_rate": 9.673604823007994e-06,
      "loss": 0.6944,
      "step": 495
    },
    {
      "epoch": 0.20644095788604458,
      "grad_norm": 2.8340105200706507,
      "learning_rate": 9.660674101575188e-06,
      "loss": 0.6959,
      "step": 500
    },
    {
      "epoch": 0.20850536746490503,
      "grad_norm": 2.6749836979567325,
      "learning_rate": 9.647501181252228e-06,
      "loss": 0.7181,
      "step": 505
    },
    {
      "epoch": 0.21056977704376548,
      "grad_norm": 2.871887495235821,
      "learning_rate": 9.634086746589597e-06,
      "loss": 0.6855,
      "step": 510
    },
    {
      "epoch": 0.21263418662262593,
      "grad_norm": 3.022510499989922,
      "learning_rate": 9.62043149468842e-06,
      "loss": 0.6786,
      "step": 515
    },
    {
      "epoch": 0.21469859620148637,
      "grad_norm": 2.798631786402729,
      "learning_rate": 9.60653613516426e-06,
      "loss": 0.6878,
      "step": 520
    },
    {
      "epoch": 0.21676300578034682,
      "grad_norm": 2.7665584034808557,
      "learning_rate": 9.592401390110224e-06,
      "loss": 0.6689,
      "step": 525
    },
    {
      "epoch": 0.21882741535920727,
      "grad_norm": 2.6972597849127995,
      "learning_rate": 9.578027994059448e-06,
      "loss": 0.6938,
      "step": 530
    },
    {
      "epoch": 0.22089182493806772,
      "grad_norm": 2.876624388604322,
      "learning_rate": 9.563416693946927e-06,
      "loss": 0.6831,
      "step": 535
    },
    {
      "epoch": 0.22295623451692817,
      "grad_norm": 2.760614779991045,
      "learning_rate": 9.548568249070688e-06,
      "loss": 0.6721,
      "step": 540
    },
    {
      "epoch": 0.2250206440957886,
      "grad_norm": 2.711332679097759,
      "learning_rate": 9.533483431052347e-06,
      "loss": 0.6734,
      "step": 545
    },
    {
      "epoch": 0.22708505367464904,
      "grad_norm": 2.9419038213760573,
      "learning_rate": 9.518163023797001e-06,
      "loss": 0.6826,
      "step": 550
    },
    {
      "epoch": 0.2291494632535095,
      "grad_norm": 2.708449916455307,
      "learning_rate": 9.502607823452492e-06,
      "loss": 0.6909,
      "step": 555
    },
    {
      "epoch": 0.23121387283236994,
      "grad_norm": 2.6744723555279712,
      "learning_rate": 9.486818638368037e-06,
      "loss": 0.6727,
      "step": 560
    },
    {
      "epoch": 0.2332782824112304,
      "grad_norm": 2.740519766192663,
      "learning_rate": 9.47079628905222e-06,
      "loss": 0.6643,
      "step": 565
    },
    {
      "epoch": 0.23534269199009084,
      "grad_norm": 2.6127044855477544,
      "learning_rate": 9.45454160813035e-06,
      "loss": 0.641,
      "step": 570
    },
    {
      "epoch": 0.23740710156895128,
      "grad_norm": 2.5898360039465547,
      "learning_rate": 9.438055440301198e-06,
      "loss": 0.654,
      "step": 575
    },
    {
      "epoch": 0.23947151114781173,
      "grad_norm": 2.594315428105211,
      "learning_rate": 9.421338642293096e-06,
      "loss": 0.6612,
      "step": 580
    },
    {
      "epoch": 0.24153592072667218,
      "grad_norm": 2.675279309210325,
      "learning_rate": 9.404392082819418e-06,
      "loss": 0.6835,
      "step": 585
    },
    {
      "epoch": 0.24360033030553263,
      "grad_norm": 2.720481221371884,
      "learning_rate": 9.387216642533436e-06,
      "loss": 0.638,
      "step": 590
    },
    {
      "epoch": 0.24566473988439305,
      "grad_norm": 2.7274539164197567,
      "learning_rate": 9.369813213982557e-06,
      "loss": 0.6742,
      "step": 595
    },
    {
      "epoch": 0.2477291494632535,
      "grad_norm": 2.881159963630694,
      "learning_rate": 9.352182701561932e-06,
      "loss": 0.643,
      "step": 600
    },
    {
      "epoch": 0.24979355904211395,
      "grad_norm": 2.672785271513205,
      "learning_rate": 9.334326021467473e-06,
      "loss": 0.6518,
      "step": 605
    },
    {
      "epoch": 0.2518579686209744,
      "grad_norm": 2.6806455710314365,
      "learning_rate": 9.316244101648229e-06,
      "loss": 0.644,
      "step": 610
    },
    {
      "epoch": 0.2539223781998348,
      "grad_norm": 2.5531072290720296,
      "learning_rate": 9.297937881758165e-06,
      "loss": 0.6738,
      "step": 615
    },
    {
      "epoch": 0.25598678777869527,
      "grad_norm": 2.8562519058356335,
      "learning_rate": 9.279408313107342e-06,
      "loss": 0.6497,
      "step": 620
    },
    {
      "epoch": 0.2580511973575557,
      "grad_norm": 2.9118129566241273,
      "learning_rate": 9.260656358612461e-06,
      "loss": 0.6544,
      "step": 625
    },
    {
      "epoch": 0.26011560693641617,
      "grad_norm": 2.6582669873812117,
      "learning_rate": 9.241682992746847e-06,
      "loss": 0.6492,
      "step": 630
    },
    {
      "epoch": 0.2621800165152766,
      "grad_norm": 3.135088950824475,
      "learning_rate": 9.222489201489792e-06,
      "loss": 0.6324,
      "step": 635
    },
    {
      "epoch": 0.26424442609413706,
      "grad_norm": 2.5674033607777282,
      "learning_rate": 9.203075982275323e-06,
      "loss": 0.6461,
      "step": 640
    },
    {
      "epoch": 0.2663088356729975,
      "grad_norm": 2.615329482313224,
      "learning_rate": 9.18344434394037e-06,
      "loss": 0.6322,
      "step": 645
    },
    {
      "epoch": 0.26837324525185796,
      "grad_norm": 2.6662637432037144,
      "learning_rate": 9.163595306672337e-06,
      "loss": 0.6245,
      "step": 650
    },
    {
      "epoch": 0.2704376548307184,
      "grad_norm": 2.788831071229617,
      "learning_rate": 9.143529901956089e-06,
      "loss": 0.6395,
      "step": 655
    },
    {
      "epoch": 0.27250206440957886,
      "grad_norm": 2.7354638041171344,
      "learning_rate": 9.12324917252035e-06,
      "loss": 0.6239,
      "step": 660
    },
    {
      "epoch": 0.2745664739884393,
      "grad_norm": 2.6074213157412682,
      "learning_rate": 9.102754172283513e-06,
      "loss": 0.6146,
      "step": 665
    },
    {
      "epoch": 0.27663088356729976,
      "grad_norm": 2.6642140176865485,
      "learning_rate": 9.082045966298871e-06,
      "loss": 0.6207,
      "step": 670
    },
    {
      "epoch": 0.2786952931461602,
      "grad_norm": 2.7423993871208547,
      "learning_rate": 9.061125630699276e-06,
      "loss": 0.6244,
      "step": 675
    },
    {
      "epoch": 0.28075970272502065,
      "grad_norm": 2.639661943202539,
      "learning_rate": 9.039994252641214e-06,
      "loss": 0.6212,
      "step": 680
    },
    {
      "epoch": 0.2828241123038811,
      "grad_norm": 2.7917609886441443,
      "learning_rate": 9.018652930248299e-06,
      "loss": 0.6279,
      "step": 685
    },
    {
      "epoch": 0.28488852188274155,
      "grad_norm": 2.629871903545716,
      "learning_rate": 8.997102772554226e-06,
      "loss": 0.6229,
      "step": 690
    },
    {
      "epoch": 0.286952931461602,
      "grad_norm": 2.625664182045412,
      "learning_rate": 8.975344899445119e-06,
      "loss": 0.6177,
      "step": 695
    },
    {
      "epoch": 0.28901734104046245,
      "grad_norm": 2.80144553747866,
      "learning_rate": 8.953380441601352e-06,
      "loss": 0.6225,
      "step": 700
    },
    {
      "epoch": 0.2910817506193229,
      "grad_norm": 2.672640679384628,
      "learning_rate": 8.931210540438778e-06,
      "loss": 0.6169,
      "step": 705
    },
    {
      "epoch": 0.2931461601981833,
      "grad_norm": 2.709692648262891,
      "learning_rate": 8.908836348049421e-06,
      "loss": 0.6198,
      "step": 710
    },
    {
      "epoch": 0.29521056977704374,
      "grad_norm": 2.681528317767069,
      "learning_rate": 8.886259027141603e-06,
      "loss": 0.6149,
      "step": 715
    },
    {
      "epoch": 0.2972749793559042,
      "grad_norm": 3.03293644975451,
      "learning_rate": 8.863479750979523e-06,
      "loss": 0.6048,
      "step": 720
    },
    {
      "epoch": 0.29933938893476464,
      "grad_norm": 2.670868153191048,
      "learning_rate": 8.840499703322286e-06,
      "loss": 0.6244,
      "step": 725
    },
    {
      "epoch": 0.3014037985136251,
      "grad_norm": 2.7608395704880646,
      "learning_rate": 8.817320078362388e-06,
      "loss": 0.6035,
      "step": 730
    },
    {
      "epoch": 0.30346820809248554,
      "grad_norm": 2.62588384982977,
      "learning_rate": 8.793942080663658e-06,
      "loss": 0.6172,
      "step": 735
    },
    {
      "epoch": 0.305532617671346,
      "grad_norm": 2.5889068256927006,
      "learning_rate": 8.77036692509866e-06,
      "loss": 0.6128,
      "step": 740
    },
    {
      "epoch": 0.30759702725020643,
      "grad_norm": 2.8092080143766216,
      "learning_rate": 8.746595836785561e-06,
      "loss": 0.6218,
      "step": 745
    },
    {
      "epoch": 0.3096614368290669,
      "grad_norm": 2.6419072436310205,
      "learning_rate": 8.722630051024467e-06,
      "loss": 0.5877,
      "step": 750
    },
    {
      "epoch": 0.31172584640792733,
      "grad_norm": 2.4997349328580363,
      "learning_rate": 8.698470813233223e-06,
      "loss": 0.6091,
      "step": 755
    },
    {
      "epoch": 0.3137902559867878,
      "grad_norm": 2.6466710278909074,
      "learning_rate": 8.674119378882707e-06,
      "loss": 0.6087,
      "step": 760
    },
    {
      "epoch": 0.31585466556564823,
      "grad_norm": 2.9297154437714363,
      "learning_rate": 8.649577013431569e-06,
      "loss": 0.592,
      "step": 765
    },
    {
      "epoch": 0.3179190751445087,
      "grad_norm": 2.549997409403058,
      "learning_rate": 8.624844992260483e-06,
      "loss": 0.6023,
      "step": 770
    },
    {
      "epoch": 0.3199834847233691,
      "grad_norm": 2.7080008761741827,
      "learning_rate": 8.599924600605865e-06,
      "loss": 0.6087,
      "step": 775
    },
    {
      "epoch": 0.3220478943022296,
      "grad_norm": 2.567143331934158,
      "learning_rate": 8.574817133493085e-06,
      "loss": 0.5847,
      "step": 780
    },
    {
      "epoch": 0.32411230388109,
      "grad_norm": 3.0113192932757418,
      "learning_rate": 8.549523895669167e-06,
      "loss": 0.5908,
      "step": 785
    },
    {
      "epoch": 0.3261767134599505,
      "grad_norm": 2.7819521209578206,
      "learning_rate": 8.52404620153499e-06,
      "loss": 0.5953,
      "step": 790
    },
    {
      "epoch": 0.3282411230388109,
      "grad_norm": 2.747216748021648,
      "learning_rate": 8.498385375076979e-06,
      "loss": 0.6114,
      "step": 795
    },
    {
      "epoch": 0.33030553261767137,
      "grad_norm": 2.8476886670194523,
      "learning_rate": 8.472542749798302e-06,
      "loss": 0.5923,
      "step": 800
    },
    {
      "epoch": 0.33236994219653176,
      "grad_norm": 3.078148326221217,
      "learning_rate": 8.44651966864958e-06,
      "loss": 0.5854,
      "step": 805
    },
    {
      "epoch": 0.3344343517753922,
      "grad_norm": 2.7069355571781553,
      "learning_rate": 8.420317483959086e-06,
      "loss": 0.5963,
      "step": 810
    },
    {
      "epoch": 0.33649876135425266,
      "grad_norm": 2.651998023029158,
      "learning_rate": 8.39393755736248e-06,
      "loss": 0.578,
      "step": 815
    },
    {
      "epoch": 0.3385631709331131,
      "grad_norm": 2.6251615353567117,
      "learning_rate": 8.367381259732042e-06,
      "loss": 0.587,
      "step": 820
    },
    {
      "epoch": 0.34062758051197356,
      "grad_norm": 2.5308341360997595,
      "learning_rate": 8.340649971105444e-06,
      "loss": 0.5732,
      "step": 825
    },
    {
      "epoch": 0.342691990090834,
      "grad_norm": 2.823077213728183,
      "learning_rate": 8.313745080614017e-06,
      "loss": 0.5672,
      "step": 830
    },
    {
      "epoch": 0.34475639966969446,
      "grad_norm": 2.659469137057744,
      "learning_rate": 8.286667986410578e-06,
      "loss": 0.5858,
      "step": 835
    },
    {
      "epoch": 0.3468208092485549,
      "grad_norm": 2.845018948853141,
      "learning_rate": 8.259420095596766e-06,
      "loss": 0.5662,
      "step": 840
    },
    {
      "epoch": 0.34888521882741536,
      "grad_norm": 2.6028478895400116,
      "learning_rate": 8.232002824149917e-06,
      "loss": 0.5606,
      "step": 845
    },
    {
      "epoch": 0.3509496284062758,
      "grad_norm": 2.664709732227379,
      "learning_rate": 8.204417596849492e-06,
      "loss": 0.5807,
      "step": 850
    },
    {
      "epoch": 0.35301403798513625,
      "grad_norm": 2.6109659834823478,
      "learning_rate": 8.176665847203023e-06,
      "loss": 0.5678,
      "step": 855
    },
    {
      "epoch": 0.3550784475639967,
      "grad_norm": 2.576771155175357,
      "learning_rate": 8.148749017371627e-06,
      "loss": 0.5656,
      "step": 860
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 2.6688801593883817,
      "learning_rate": 8.120668558095056e-06,
      "loss": 0.5749,
      "step": 865
    },
    {
      "epoch": 0.3592072667217176,
      "grad_norm": 2.801881409304358,
      "learning_rate": 8.092425928616317e-06,
      "loss": 0.5603,
      "step": 870
    },
    {
      "epoch": 0.36127167630057805,
      "grad_norm": 2.7158675577145512,
      "learning_rate": 8.064022596605825e-06,
      "loss": 0.5878,
      "step": 875
    },
    {
      "epoch": 0.3633360858794385,
      "grad_norm": 2.7112475864205208,
      "learning_rate": 8.03546003808515e-06,
      "loss": 0.5697,
      "step": 880
    },
    {
      "epoch": 0.36540049545829895,
      "grad_norm": 2.757081302105068,
      "learning_rate": 8.006739737350303e-06,
      "loss": 0.5549,
      "step": 885
    },
    {
      "epoch": 0.3674649050371594,
      "grad_norm": 2.6798695533176,
      "learning_rate": 7.977863186894598e-06,
      "loss": 0.5621,
      "step": 890
    },
    {
      "epoch": 0.36952931461601984,
      "grad_norm": 2.901794116790705,
      "learning_rate": 7.948831887331113e-06,
      "loss": 0.5531,
      "step": 895
    },
    {
      "epoch": 0.37159372419488024,
      "grad_norm": 2.6209628611019333,
      "learning_rate": 7.91964734731468e-06,
      "loss": 0.5529,
      "step": 900
    },
    {
      "epoch": 0.3736581337737407,
      "grad_norm": 2.631776999408232,
      "learning_rate": 7.890311083463511e-06,
      "loss": 0.5341,
      "step": 905
    },
    {
      "epoch": 0.37572254335260113,
      "grad_norm": 2.6460839052940424,
      "learning_rate": 7.860824620280367e-06,
      "loss": 0.5589,
      "step": 910
    },
    {
      "epoch": 0.3777869529314616,
      "grad_norm": 2.7027858769480604,
      "learning_rate": 7.831189490073343e-06,
      "loss": 0.5662,
      "step": 915
    },
    {
      "epoch": 0.37985136251032203,
      "grad_norm": 2.686204866691535,
      "learning_rate": 7.80140723287624e-06,
      "loss": 0.5557,
      "step": 920
    },
    {
      "epoch": 0.3819157720891825,
      "grad_norm": 2.537669294604416,
      "learning_rate": 7.771479396368533e-06,
      "loss": 0.5392,
      "step": 925
    },
    {
      "epoch": 0.38398018166804293,
      "grad_norm": 2.634245727933428,
      "learning_rate": 7.741407535794939e-06,
      "loss": 0.5476,
      "step": 930
    },
    {
      "epoch": 0.3860445912469034,
      "grad_norm": 2.63998365984595,
      "learning_rate": 7.711193213884602e-06,
      "loss": 0.5411,
      "step": 935
    },
    {
      "epoch": 0.38810900082576383,
      "grad_norm": 2.5066641406139163,
      "learning_rate": 7.680838000769889e-06,
      "loss": 0.5436,
      "step": 940
    },
    {
      "epoch": 0.3901734104046243,
      "grad_norm": 2.4698510563626384,
      "learning_rate": 7.650343473904776e-06,
      "loss": 0.536,
      "step": 945
    },
    {
      "epoch": 0.3922378199834847,
      "grad_norm": 2.5102808206533753,
      "learning_rate": 7.619711217982899e-06,
      "loss": 0.5391,
      "step": 950
    },
    {
      "epoch": 0.3943022295623452,
      "grad_norm": 2.648139467398325,
      "learning_rate": 7.5889428248551854e-06,
      "loss": 0.5368,
      "step": 955
    },
    {
      "epoch": 0.3963666391412056,
      "grad_norm": 2.6648738415553765,
      "learning_rate": 7.558039893447131e-06,
      "loss": 0.5496,
      "step": 960
    },
    {
      "epoch": 0.39843104872006607,
      "grad_norm": 2.6157509498551526,
      "learning_rate": 7.527004029675722e-06,
      "loss": 0.5318,
      "step": 965
    },
    {
      "epoch": 0.4004954582989265,
      "grad_norm": 2.5975829607560574,
      "learning_rate": 7.495836846365965e-06,
      "loss": 0.5352,
      "step": 970
    },
    {
      "epoch": 0.40255986787778697,
      "grad_norm": 2.92450937390704,
      "learning_rate": 7.464539963167091e-06,
      "loss": 0.5177,
      "step": 975
    },
    {
      "epoch": 0.4046242774566474,
      "grad_norm": 2.7355952122295264,
      "learning_rate": 7.433115006468373e-06,
      "loss": 0.5322,
      "step": 980
    },
    {
      "epoch": 0.40668868703550787,
      "grad_norm": 2.5829736114469184,
      "learning_rate": 7.40156360931462e-06,
      "loss": 0.5298,
      "step": 985
    },
    {
      "epoch": 0.4087530966143683,
      "grad_norm": 2.705958846647064,
      "learning_rate": 7.369887411321305e-06,
      "loss": 0.5477,
      "step": 990
    },
    {
      "epoch": 0.4108175061932287,
      "grad_norm": 2.4846959730474145,
      "learning_rate": 7.338088058589362e-06,
      "loss": 0.5324,
      "step": 995
    },
    {
      "epoch": 0.41288191577208916,
      "grad_norm": 2.5252160694956762,
      "learning_rate": 7.306167203619647e-06,
      "loss": 0.5335,
      "step": 1000
    },
    {
      "epoch": 0.4149463253509496,
      "grad_norm": 2.509240890701763,
      "learning_rate": 7.27412650522706e-06,
      "loss": 0.5219,
      "step": 1005
    },
    {
      "epoch": 0.41701073492981006,
      "grad_norm": 2.567378202198639,
      "learning_rate": 7.241967628454345e-06,
      "loss": 0.5142,
      "step": 1010
    },
    {
      "epoch": 0.4190751445086705,
      "grad_norm": 2.479116521394592,
      "learning_rate": 7.20969224448556e-06,
      "loss": 0.5198,
      "step": 1015
    },
    {
      "epoch": 0.42113955408753095,
      "grad_norm": 2.456783739959803,
      "learning_rate": 7.1773020305592355e-06,
      "loss": 0.5141,
      "step": 1020
    },
    {
      "epoch": 0.4232039636663914,
      "grad_norm": 2.6855797799019356,
      "learning_rate": 7.1447986698812115e-06,
      "loss": 0.5307,
      "step": 1025
    },
    {
      "epoch": 0.42526837324525185,
      "grad_norm": 2.5323199215752203,
      "learning_rate": 7.112183851537166e-06,
      "loss": 0.5256,
      "step": 1030
    },
    {
      "epoch": 0.4273327828241123,
      "grad_norm": 2.735418738176186,
      "learning_rate": 7.079459270404842e-06,
      "loss": 0.5112,
      "step": 1035
    },
    {
      "epoch": 0.42939719240297275,
      "grad_norm": 2.6703449875287735,
      "learning_rate": 7.046626627065974e-06,
      "loss": 0.5099,
      "step": 1040
    },
    {
      "epoch": 0.4314616019818332,
      "grad_norm": 2.4631394786044747,
      "learning_rate": 7.013687627717905e-06,
      "loss": 0.4978,
      "step": 1045
    },
    {
      "epoch": 0.43352601156069365,
      "grad_norm": 2.8328555379260196,
      "learning_rate": 6.980643984084927e-06,
      "loss": 0.5179,
      "step": 1050
    },
    {
      "epoch": 0.4355904211395541,
      "grad_norm": 2.504430665948721,
      "learning_rate": 6.947497413329333e-06,
      "loss": 0.5199,
      "step": 1055
    },
    {
      "epoch": 0.43765483071841454,
      "grad_norm": 2.5711442943268663,
      "learning_rate": 6.914249637962174e-06,
      "loss": 0.5133,
      "step": 1060
    },
    {
      "epoch": 0.439719240297275,
      "grad_norm": 2.5576807334846747,
      "learning_rate": 6.880902385753749e-06,
      "loss": 0.51,
      "step": 1065
    },
    {
      "epoch": 0.44178364987613544,
      "grad_norm": 2.4378114750971833,
      "learning_rate": 6.847457389643823e-06,
      "loss": 0.4966,
      "step": 1070
    },
    {
      "epoch": 0.4438480594549959,
      "grad_norm": 2.57922332477375,
      "learning_rate": 6.8139163876515694e-06,
      "loss": 0.5018,
      "step": 1075
    },
    {
      "epoch": 0.44591246903385634,
      "grad_norm": 2.4923563722799145,
      "learning_rate": 6.780281122785243e-06,
      "loss": 0.5089,
      "step": 1080
    },
    {
      "epoch": 0.4479768786127168,
      "grad_norm": 2.380488675820629,
      "learning_rate": 6.74655334295162e-06,
      "loss": 0.4969,
      "step": 1085
    },
    {
      "epoch": 0.4500412881915772,
      "grad_norm": 2.622744707083321,
      "learning_rate": 6.712734800865146e-06,
      "loss": 0.5006,
      "step": 1090
    },
    {
      "epoch": 0.45210569777043763,
      "grad_norm": 2.5279298037448514,
      "learning_rate": 6.67882725395687e-06,
      "loss": 0.4919,
      "step": 1095
    },
    {
      "epoch": 0.4541701073492981,
      "grad_norm": 2.5696353167432076,
      "learning_rate": 6.644832464283105e-06,
      "loss": 0.5006,
      "step": 1100
    },
    {
      "epoch": 0.45623451692815853,
      "grad_norm": 2.6472120693562675,
      "learning_rate": 6.6107521984338654e-06,
      "loss": 0.5222,
      "step": 1105
    },
    {
      "epoch": 0.458298926507019,
      "grad_norm": 2.5641259833492915,
      "learning_rate": 6.576588227441063e-06,
      "loss": 0.5021,
      "step": 1110
    },
    {
      "epoch": 0.4603633360858794,
      "grad_norm": 2.508674684822535,
      "learning_rate": 6.5423423266864715e-06,
      "loss": 0.5019,
      "step": 1115
    },
    {
      "epoch": 0.4624277456647399,
      "grad_norm": 2.5803067400324546,
      "learning_rate": 6.508016275809472e-06,
      "loss": 0.5116,
      "step": 1120
    },
    {
      "epoch": 0.4644921552436003,
      "grad_norm": 2.4783246238450296,
      "learning_rate": 6.473611858614557e-06,
      "loss": 0.4939,
      "step": 1125
    },
    {
      "epoch": 0.4665565648224608,
      "grad_norm": 2.3961470840043986,
      "learning_rate": 6.439130862978653e-06,
      "loss": 0.4964,
      "step": 1130
    },
    {
      "epoch": 0.4686209744013212,
      "grad_norm": 2.523268688947089,
      "learning_rate": 6.4045750807581875e-06,
      "loss": 0.4878,
      "step": 1135
    },
    {
      "epoch": 0.47068538398018167,
      "grad_norm": 2.464908065215536,
      "learning_rate": 6.369946307695994e-06,
      "loss": 0.5004,
      "step": 1140
    },
    {
      "epoch": 0.4727497935590421,
      "grad_norm": 2.4978854529114556,
      "learning_rate": 6.335246343327983e-06,
      "loss": 0.4928,
      "step": 1145
    },
    {
      "epoch": 0.47481420313790257,
      "grad_norm": 2.5910220265362405,
      "learning_rate": 6.300476990889622e-06,
      "loss": 0.4859,
      "step": 1150
    },
    {
      "epoch": 0.476878612716763,
      "grad_norm": 2.6474515998209314,
      "learning_rate": 6.2656400572222374e-06,
      "loss": 0.4968,
      "step": 1155
    },
    {
      "epoch": 0.47894302229562347,
      "grad_norm": 2.5111508875593627,
      "learning_rate": 6.230737352679115e-06,
      "loss": 0.4825,
      "step": 1160
    },
    {
      "epoch": 0.4810074318744839,
      "grad_norm": 2.6001140755401067,
      "learning_rate": 6.1957706910314195e-06,
      "loss": 0.4903,
      "step": 1165
    },
    {
      "epoch": 0.48307184145334436,
      "grad_norm": 2.4405093084108853,
      "learning_rate": 6.160741889373948e-06,
      "loss": 0.4738,
      "step": 1170
    },
    {
      "epoch": 0.4851362510322048,
      "grad_norm": 2.6122999506914817,
      "learning_rate": 6.12565276803069e-06,
      "loss": 0.4851,
      "step": 1175
    },
    {
      "epoch": 0.48720066061106526,
      "grad_norm": 2.5491172182391284,
      "learning_rate": 6.090505150460239e-06,
      "loss": 0.4969,
      "step": 1180
    },
    {
      "epoch": 0.48926507018992565,
      "grad_norm": 2.5615374385855256,
      "learning_rate": 6.055300863161035e-06,
      "loss": 0.4778,
      "step": 1185
    },
    {
      "epoch": 0.4913294797687861,
      "grad_norm": 2.4469903993231514,
      "learning_rate": 6.020041735576444e-06,
      "loss": 0.4787,
      "step": 1190
    },
    {
      "epoch": 0.49339388934764655,
      "grad_norm": 2.4860974782023346,
      "learning_rate": 5.9847295999996835e-06,
      "loss": 0.4914,
      "step": 1195
    },
    {
      "epoch": 0.495458298926507,
      "grad_norm": 2.4900103589842395,
      "learning_rate": 5.949366291478621e-06,
      "loss": 0.4906,
      "step": 1200
    },
    {
      "epoch": 0.49752270850536745,
      "grad_norm": 2.5283147963068964,
      "learning_rate": 5.9139536477203945e-06,
      "loss": 0.4759,
      "step": 1205
    },
    {
      "epoch": 0.4995871180842279,
      "grad_norm": 2.573137486004109,
      "learning_rate": 5.878493508995924e-06,
      "loss": 0.4768,
      "step": 1210
    },
    {
      "epoch": 0.5016515276630884,
      "grad_norm": 2.579476813837634,
      "learning_rate": 5.842987718044275e-06,
      "loss": 0.4679,
      "step": 1215
    },
    {
      "epoch": 0.5037159372419489,
      "grad_norm": 2.497492462831134,
      "learning_rate": 5.807438119976904e-06,
      "loss": 0.4698,
      "step": 1220
    },
    {
      "epoch": 0.5057803468208093,
      "grad_norm": 2.590655156263669,
      "learning_rate": 5.771846562181761e-06,
      "loss": 0.4728,
      "step": 1225
    },
    {
      "epoch": 0.5078447563996696,
      "grad_norm": 2.5177297962486493,
      "learning_rate": 5.736214894227304e-06,
      "loss": 0.473,
      "step": 1230
    },
    {
      "epoch": 0.5099091659785301,
      "grad_norm": 2.6021365054723775,
      "learning_rate": 5.700544967766373e-06,
      "loss": 0.4753,
      "step": 1235
    },
    {
      "epoch": 0.5119735755573905,
      "grad_norm": 2.604951347166713,
      "learning_rate": 5.6648386364399645e-06,
      "loss": 0.47,
      "step": 1240
    },
    {
      "epoch": 0.514037985136251,
      "grad_norm": 2.4157020679627794,
      "learning_rate": 5.629097755780913e-06,
      "loss": 0.4711,
      "step": 1245
    },
    {
      "epoch": 0.5161023947151114,
      "grad_norm": 2.3659395254457167,
      "learning_rate": 5.59332418311746e-06,
      "loss": 0.4773,
      "step": 1250
    },
    {
      "epoch": 0.5181668042939719,
      "grad_norm": 2.499361991412955,
      "learning_rate": 5.557519777476733e-06,
      "loss": 0.4527,
      "step": 1255
    },
    {
      "epoch": 0.5202312138728323,
      "grad_norm": 2.5585105365426513,
      "learning_rate": 5.521686399488145e-06,
      "loss": 0.4573,
      "step": 1260
    },
    {
      "epoch": 0.5222956234516928,
      "grad_norm": 2.4814508055210567,
      "learning_rate": 5.485825911286697e-06,
      "loss": 0.4691,
      "step": 1265
    },
    {
      "epoch": 0.5243600330305532,
      "grad_norm": 2.6081291748684716,
      "learning_rate": 5.449940176416213e-06,
      "loss": 0.4526,
      "step": 1270
    },
    {
      "epoch": 0.5264244426094137,
      "grad_norm": 2.587126119981003,
      "learning_rate": 5.414031059732502e-06,
      "loss": 0.4605,
      "step": 1275
    },
    {
      "epoch": 0.5284888521882741,
      "grad_norm": 2.5458768564010765,
      "learning_rate": 5.378100427306439e-06,
      "loss": 0.4467,
      "step": 1280
    },
    {
      "epoch": 0.5305532617671346,
      "grad_norm": 2.422344978356131,
      "learning_rate": 5.3421501463269976e-06,
      "loss": 0.4563,
      "step": 1285
    },
    {
      "epoch": 0.532617671345995,
      "grad_norm": 2.481246363328885,
      "learning_rate": 5.30618208500422e-06,
      "loss": 0.4496,
      "step": 1290
    },
    {
      "epoch": 0.5346820809248555,
      "grad_norm": 2.5964265890691274,
      "learning_rate": 5.270198112472128e-06,
      "loss": 0.4487,
      "step": 1295
    },
    {
      "epoch": 0.5367464905037159,
      "grad_norm": 2.375375823310664,
      "learning_rate": 5.234200098691597e-06,
      "loss": 0.4492,
      "step": 1300
    },
    {
      "epoch": 0.5388109000825764,
      "grad_norm": 2.482312531712164,
      "learning_rate": 5.19818991435317e-06,
      "loss": 0.4524,
      "step": 1305
    },
    {
      "epoch": 0.5408753096614368,
      "grad_norm": 2.6268917122173154,
      "learning_rate": 5.162169430779861e-06,
      "loss": 0.4343,
      "step": 1310
    },
    {
      "epoch": 0.5429397192402973,
      "grad_norm": 2.543083790336233,
      "learning_rate": 5.126140519829888e-06,
      "loss": 0.4738,
      "step": 1315
    },
    {
      "epoch": 0.5450041288191577,
      "grad_norm": 2.4547121862030865,
      "learning_rate": 5.090105053799418e-06,
      "loss": 0.4604,
      "step": 1320
    },
    {
      "epoch": 0.5470685383980182,
      "grad_norm": 2.5806078709462295,
      "learning_rate": 5.054064905325262e-06,
      "loss": 0.4571,
      "step": 1325
    },
    {
      "epoch": 0.5491329479768786,
      "grad_norm": 2.538503365518751,
      "learning_rate": 5.018021947287556e-06,
      "loss": 0.4432,
      "step": 1330
    },
    {
      "epoch": 0.5511973575557391,
      "grad_norm": 2.4205037183193157,
      "learning_rate": 4.981978052712446e-06,
      "loss": 0.4545,
      "step": 1335
    },
    {
      "epoch": 0.5532617671345995,
      "grad_norm": 2.5910957054221817,
      "learning_rate": 4.9459350946747405e-06,
      "loss": 0.4584,
      "step": 1340
    },
    {
      "epoch": 0.55532617671346,
      "grad_norm": 2.423593643872453,
      "learning_rate": 4.909894946200583e-06,
      "loss": 0.454,
      "step": 1345
    },
    {
      "epoch": 0.5573905862923204,
      "grad_norm": 2.491439793223575,
      "learning_rate": 4.873859480170113e-06,
      "loss": 0.4535,
      "step": 1350
    },
    {
      "epoch": 0.5594549958711809,
      "grad_norm": 2.6251023268801106,
      "learning_rate": 4.83783056922014e-06,
      "loss": 0.4475,
      "step": 1355
    },
    {
      "epoch": 0.5615194054500413,
      "grad_norm": 2.379012672081149,
      "learning_rate": 4.801810085646831e-06,
      "loss": 0.4427,
      "step": 1360
    },
    {
      "epoch": 0.5635838150289018,
      "grad_norm": 2.438307211574731,
      "learning_rate": 4.765799901308405e-06,
      "loss": 0.4392,
      "step": 1365
    },
    {
      "epoch": 0.5656482246077622,
      "grad_norm": 2.4743966890129534,
      "learning_rate": 4.729801887527872e-06,
      "loss": 0.4376,
      "step": 1370
    },
    {
      "epoch": 0.5677126341866227,
      "grad_norm": 2.4438402043185636,
      "learning_rate": 4.6938179149957815e-06,
      "loss": 0.4448,
      "step": 1375
    },
    {
      "epoch": 0.5697770437654831,
      "grad_norm": 2.3233363056641587,
      "learning_rate": 4.657849853673004e-06,
      "loss": 0.4332,
      "step": 1380
    },
    {
      "epoch": 0.5718414533443436,
      "grad_norm": 2.3513159156694536,
      "learning_rate": 4.621899572693564e-06,
      "loss": 0.4286,
      "step": 1385
    },
    {
      "epoch": 0.573905862923204,
      "grad_norm": 2.452790907750448,
      "learning_rate": 4.585968940267499e-06,
      "loss": 0.4329,
      "step": 1390
    },
    {
      "epoch": 0.5759702725020645,
      "grad_norm": 2.4876165011978544,
      "learning_rate": 4.550059823583788e-06,
      "loss": 0.4368,
      "step": 1395
    },
    {
      "epoch": 0.5780346820809249,
      "grad_norm": 2.4110190508295926,
      "learning_rate": 4.514174088713305e-06,
      "loss": 0.4294,
      "step": 1400
    },
    {
      "epoch": 0.5800990916597853,
      "grad_norm": 2.4790636104654724,
      "learning_rate": 4.478313600511856e-06,
      "loss": 0.438,
      "step": 1405
    },
    {
      "epoch": 0.5821635012386458,
      "grad_norm": 2.5304633815274062,
      "learning_rate": 4.442480222523269e-06,
      "loss": 0.4383,
      "step": 1410
    },
    {
      "epoch": 0.5842279108175062,
      "grad_norm": 2.365309103378621,
      "learning_rate": 4.4066758168825414e-06,
      "loss": 0.4383,
      "step": 1415
    },
    {
      "epoch": 0.5862923203963666,
      "grad_norm": 2.4670833952729687,
      "learning_rate": 4.370902244219088e-06,
      "loss": 0.4389,
      "step": 1420
    },
    {
      "epoch": 0.588356729975227,
      "grad_norm": 2.3390938676909734,
      "learning_rate": 4.335161363560037e-06,
      "loss": 0.4291,
      "step": 1425
    },
    {
      "epoch": 0.5904211395540875,
      "grad_norm": 2.362644766829775,
      "learning_rate": 4.299455032233631e-06,
      "loss": 0.4288,
      "step": 1430
    },
    {
      "epoch": 0.5924855491329479,
      "grad_norm": 2.445730007800526,
      "learning_rate": 4.263785105772696e-06,
      "loss": 0.4335,
      "step": 1435
    },
    {
      "epoch": 0.5945499587118084,
      "grad_norm": 2.303973634970439,
      "learning_rate": 4.22815343781824e-06,
      "loss": 0.4366,
      "step": 1440
    },
    {
      "epoch": 0.5966143682906688,
      "grad_norm": 2.391887494831502,
      "learning_rate": 4.192561880023099e-06,
      "loss": 0.4261,
      "step": 1445
    },
    {
      "epoch": 0.5986787778695293,
      "grad_norm": 2.293302291662098,
      "learning_rate": 4.157012281955726e-06,
      "loss": 0.424,
      "step": 1450
    },
    {
      "epoch": 0.6007431874483897,
      "grad_norm": 2.3712447663578655,
      "learning_rate": 4.1215064910040795e-06,
      "loss": 0.4314,
      "step": 1455
    },
    {
      "epoch": 0.6028075970272502,
      "grad_norm": 2.381812200033987,
      "learning_rate": 4.086046352279606e-06,
      "loss": 0.4227,
      "step": 1460
    },
    {
      "epoch": 0.6048720066061106,
      "grad_norm": 2.4743735981326114,
      "learning_rate": 4.050633708521381e-06,
      "loss": 0.4158,
      "step": 1465
    },
    {
      "epoch": 0.6069364161849711,
      "grad_norm": 2.396322395044398,
      "learning_rate": 4.015270400000317e-06,
      "loss": 0.42,
      "step": 1470
    },
    {
      "epoch": 0.6090008257638315,
      "grad_norm": 2.565114051952279,
      "learning_rate": 3.979958264423558e-06,
      "loss": 0.4155,
      "step": 1475
    },
    {
      "epoch": 0.611065235342692,
      "grad_norm": 2.4934554445972954,
      "learning_rate": 3.944699136838965e-06,
      "loss": 0.4126,
      "step": 1480
    },
    {
      "epoch": 0.6131296449215524,
      "grad_norm": 2.4162398419293187,
      "learning_rate": 3.909494849539761e-06,
      "loss": 0.4213,
      "step": 1485
    },
    {
      "epoch": 0.6151940545004129,
      "grad_norm": 2.464180248251729,
      "learning_rate": 3.874347231969312e-06,
      "loss": 0.4235,
      "step": 1490
    },
    {
      "epoch": 0.6172584640792733,
      "grad_norm": 2.4941582294474434,
      "learning_rate": 3.839258110626053e-06,
      "loss": 0.4307,
      "step": 1495
    },
    {
      "epoch": 0.6193228736581338,
      "grad_norm": 2.276754213405921,
      "learning_rate": 3.8042293089685813e-06,
      "loss": 0.4065,
      "step": 1500
    },
    {
      "epoch": 0.6213872832369942,
      "grad_norm": 2.324659425092549,
      "learning_rate": 3.769262647320886e-06,
      "loss": 0.4179,
      "step": 1505
    },
    {
      "epoch": 0.6234516928158547,
      "grad_norm": 2.3013999943076815,
      "learning_rate": 3.7343599427777634e-06,
      "loss": 0.4157,
      "step": 1510
    },
    {
      "epoch": 0.6255161023947151,
      "grad_norm": 2.3853366654087633,
      "learning_rate": 3.69952300911038e-06,
      "loss": 0.4087,
      "step": 1515
    },
    {
      "epoch": 0.6275805119735756,
      "grad_norm": 2.3274220274703206,
      "learning_rate": 3.6647536566720186e-06,
      "loss": 0.4203,
      "step": 1520
    },
    {
      "epoch": 0.629644921552436,
      "grad_norm": 2.4392433402643947,
      "learning_rate": 3.630053692304005e-06,
      "loss": 0.4091,
      "step": 1525
    },
    {
      "epoch": 0.6317093311312965,
      "grad_norm": 2.4428540121878894,
      "learning_rate": 3.5954249192418133e-06,
      "loss": 0.4255,
      "step": 1530
    },
    {
      "epoch": 0.6337737407101569,
      "grad_norm": 2.3810770043048692,
      "learning_rate": 3.560869137021349e-06,
      "loss": 0.4103,
      "step": 1535
    },
    {
      "epoch": 0.6358381502890174,
      "grad_norm": 2.301863999851205,
      "learning_rate": 3.526388141385444e-06,
      "loss": 0.4155,
      "step": 1540
    },
    {
      "epoch": 0.6379025598678778,
      "grad_norm": 2.400363519907276,
      "learning_rate": 3.4919837241905308e-06,
      "loss": 0.4224,
      "step": 1545
    },
    {
      "epoch": 0.6399669694467383,
      "grad_norm": 2.477498090820749,
      "learning_rate": 3.4576576733135284e-06,
      "loss": 0.3938,
      "step": 1550
    },
    {
      "epoch": 0.6420313790255987,
      "grad_norm": 2.707387721012179,
      "learning_rate": 3.4234117725589385e-06,
      "loss": 0.4111,
      "step": 1555
    },
    {
      "epoch": 0.6440957886044592,
      "grad_norm": 2.341459106750255,
      "learning_rate": 3.3892478015661362e-06,
      "loss": 0.4139,
      "step": 1560
    },
    {
      "epoch": 0.6461601981833196,
      "grad_norm": 2.4202740830209137,
      "learning_rate": 3.355167535716897e-06,
      "loss": 0.4127,
      "step": 1565
    },
    {
      "epoch": 0.64822460776218,
      "grad_norm": 2.3816575196973577,
      "learning_rate": 3.321172746043131e-06,
      "loss": 0.4002,
      "step": 1570
    },
    {
      "epoch": 0.6502890173410405,
      "grad_norm": 2.454435864594131,
      "learning_rate": 3.2872651991348548e-06,
      "loss": 0.403,
      "step": 1575
    },
    {
      "epoch": 0.652353426919901,
      "grad_norm": 2.591163202406022,
      "learning_rate": 3.253446657048382e-06,
      "loss": 0.4072,
      "step": 1580
    },
    {
      "epoch": 0.6544178364987614,
      "grad_norm": 2.4282921305894676,
      "learning_rate": 3.2197188772147593e-06,
      "loss": 0.3869,
      "step": 1585
    },
    {
      "epoch": 0.6564822460776218,
      "grad_norm": 2.4682745998960827,
      "learning_rate": 3.186083612348434e-06,
      "loss": 0.4028,
      "step": 1590
    },
    {
      "epoch": 0.6585466556564823,
      "grad_norm": 2.376999029803135,
      "learning_rate": 3.1525426103561774e-06,
      "loss": 0.4064,
      "step": 1595
    },
    {
      "epoch": 0.6606110652353427,
      "grad_norm": 2.370838413822066,
      "learning_rate": 3.119097614246252e-06,
      "loss": 0.3978,
      "step": 1600
    },
    {
      "epoch": 0.6626754748142032,
      "grad_norm": 2.362260369144417,
      "learning_rate": 3.0857503620378284e-06,
      "loss": 0.3905,
      "step": 1605
    },
    {
      "epoch": 0.6647398843930635,
      "grad_norm": 2.421231306606393,
      "learning_rate": 3.052502586670669e-06,
      "loss": 0.4053,
      "step": 1610
    },
    {
      "epoch": 0.666804293971924,
      "grad_norm": 2.395970410200359,
      "learning_rate": 3.0193560159150747e-06,
      "loss": 0.3943,
      "step": 1615
    },
    {
      "epoch": 0.6688687035507844,
      "grad_norm": 2.5057121725436917,
      "learning_rate": 2.986312372282097e-06,
      "loss": 0.4182,
      "step": 1620
    },
    {
      "epoch": 0.6709331131296449,
      "grad_norm": 2.3776161687385207,
      "learning_rate": 2.9533733729340274e-06,
      "loss": 0.4077,
      "step": 1625
    },
    {
      "epoch": 0.6729975227085053,
      "grad_norm": 2.5915693629163217,
      "learning_rate": 2.920540729595159e-06,
      "loss": 0.4085,
      "step": 1630
    },
    {
      "epoch": 0.6750619322873658,
      "grad_norm": 2.443160283840937,
      "learning_rate": 2.8878161484628364e-06,
      "loss": 0.4028,
      "step": 1635
    },
    {
      "epoch": 0.6771263418662262,
      "grad_norm": 2.420398270818125,
      "learning_rate": 2.8552013301187898e-06,
      "loss": 0.3972,
      "step": 1640
    },
    {
      "epoch": 0.6791907514450867,
      "grad_norm": 2.4196367974051265,
      "learning_rate": 2.8226979694407657e-06,
      "loss": 0.4067,
      "step": 1645
    },
    {
      "epoch": 0.6812551610239471,
      "grad_norm": 2.438274119472032,
      "learning_rate": 2.790307755514442e-06,
      "loss": 0.4032,
      "step": 1650
    },
    {
      "epoch": 0.6833195706028076,
      "grad_norm": 3.0023110501423944,
      "learning_rate": 2.7580323715456564e-06,
      "loss": 0.4121,
      "step": 1655
    },
    {
      "epoch": 0.685383980181668,
      "grad_norm": 2.3597852522402474,
      "learning_rate": 2.7258734947729428e-06,
      "loss": 0.3938,
      "step": 1660
    },
    {
      "epoch": 0.6874483897605285,
      "grad_norm": 2.4413156249499854,
      "learning_rate": 2.6938327963803545e-06,
      "loss": 0.3925,
      "step": 1665
    },
    {
      "epoch": 0.6895127993393889,
      "grad_norm": 2.441149846551036,
      "learning_rate": 2.6619119414106385e-06,
      "loss": 0.3902,
      "step": 1670
    },
    {
      "epoch": 0.6915772089182494,
      "grad_norm": 2.4274483132890685,
      "learning_rate": 2.6301125886786965e-06,
      "loss": 0.3933,
      "step": 1675
    },
    {
      "epoch": 0.6936416184971098,
      "grad_norm": 2.514301355730002,
      "learning_rate": 2.5984363906853814e-06,
      "loss": 0.3847,
      "step": 1680
    },
    {
      "epoch": 0.6957060280759703,
      "grad_norm": 2.23025942549343,
      "learning_rate": 2.566884993531628e-06,
      "loss": 0.3868,
      "step": 1685
    },
    {
      "epoch": 0.6977704376548307,
      "grad_norm": 2.3910909958649764,
      "learning_rate": 2.5354600368329108e-06,
      "loss": 0.379,
      "step": 1690
    },
    {
      "epoch": 0.6998348472336912,
      "grad_norm": 2.5413951969823554,
      "learning_rate": 2.5041631536340374e-06,
      "loss": 0.3909,
      "step": 1695
    },
    {
      "epoch": 0.7018992568125516,
      "grad_norm": 2.360003867942285,
      "learning_rate": 2.472995970324282e-06,
      "loss": 0.3888,
      "step": 1700
    },
    {
      "epoch": 0.7039636663914121,
      "grad_norm": 2.4621248065886516,
      "learning_rate": 2.4419601065528703e-06,
      "loss": 0.3803,
      "step": 1705
    },
    {
      "epoch": 0.7060280759702725,
      "grad_norm": 2.4221633580536865,
      "learning_rate": 2.4110571751448154e-06,
      "loss": 0.3804,
      "step": 1710
    },
    {
      "epoch": 0.708092485549133,
      "grad_norm": 2.3044799343595685,
      "learning_rate": 2.3802887820171012e-06,
      "loss": 0.3709,
      "step": 1715
    },
    {
      "epoch": 0.7101568951279934,
      "grad_norm": 2.3101555840305585,
      "learning_rate": 2.349656526095224e-06,
      "loss": 0.382,
      "step": 1720
    },
    {
      "epoch": 0.7122213047068539,
      "grad_norm": 2.439524803446898,
      "learning_rate": 2.3191619992301144e-06,
      "loss": 0.3969,
      "step": 1725
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 2.3771584605894174,
      "learning_rate": 2.2888067861153983e-06,
      "loss": 0.3792,
      "step": 1730
    },
    {
      "epoch": 0.7163501238645748,
      "grad_norm": 2.5666012183468587,
      "learning_rate": 2.2585924642050638e-06,
      "loss": 0.3802,
      "step": 1735
    },
    {
      "epoch": 0.7184145334434352,
      "grad_norm": 2.3145141740784307,
      "learning_rate": 2.2285206036314684e-06,
      "loss": 0.3835,
      "step": 1740
    },
    {
      "epoch": 0.7204789430222956,
      "grad_norm": 2.520653602273766,
      "learning_rate": 2.1985927671237605e-06,
      "loss": 0.3766,
      "step": 1745
    },
    {
      "epoch": 0.7225433526011561,
      "grad_norm": 2.470460484865602,
      "learning_rate": 2.168810509926659e-06,
      "loss": 0.3897,
      "step": 1750
    },
    {
      "epoch": 0.7246077621800165,
      "grad_norm": 2.408249418201136,
      "learning_rate": 2.1391753797196328e-06,
      "loss": 0.3952,
      "step": 1755
    },
    {
      "epoch": 0.726672171758877,
      "grad_norm": 2.4402970653227247,
      "learning_rate": 2.1096889165364894e-06,
      "loss": 0.375,
      "step": 1760
    },
    {
      "epoch": 0.7287365813377374,
      "grad_norm": 2.4710285995197188,
      "learning_rate": 2.0803526526853206e-06,
      "loss": 0.3748,
      "step": 1765
    },
    {
      "epoch": 0.7308009909165979,
      "grad_norm": 2.2841746369286633,
      "learning_rate": 2.0511681126688883e-06,
      "loss": 0.3799,
      "step": 1770
    },
    {
      "epoch": 0.7328654004954583,
      "grad_norm": 2.46102171882109,
      "learning_rate": 2.022136813105401e-06,
      "loss": 0.3729,
      "step": 1775
    },
    {
      "epoch": 0.7349298100743188,
      "grad_norm": 2.2829340264643156,
      "learning_rate": 1.9932602626496994e-06,
      "loss": 0.3722,
      "step": 1780
    },
    {
      "epoch": 0.7369942196531792,
      "grad_norm": 2.4439441720908,
      "learning_rate": 1.9645399619148515e-06,
      "loss": 0.3774,
      "step": 1785
    },
    {
      "epoch": 0.7390586292320397,
      "grad_norm": 2.363486720634483,
      "learning_rate": 1.9359774033941757e-06,
      "loss": 0.3775,
      "step": 1790
    },
    {
      "epoch": 0.7411230388109001,
      "grad_norm": 2.2893262909558065,
      "learning_rate": 1.9075740713836854e-06,
      "loss": 0.3851,
      "step": 1795
    },
    {
      "epoch": 0.7431874483897605,
      "grad_norm": 2.381579584567696,
      "learning_rate": 1.8793314419049446e-06,
      "loss": 0.3736,
      "step": 1800
    },
    {
      "epoch": 0.7452518579686209,
      "grad_norm": 2.3588892144435714,
      "learning_rate": 1.8512509826283752e-06,
      "loss": 0.3647,
      "step": 1805
    },
    {
      "epoch": 0.7473162675474814,
      "grad_norm": 2.468444814623241,
      "learning_rate": 1.8233341527969777e-06,
      "loss": 0.3731,
      "step": 1810
    },
    {
      "epoch": 0.7493806771263418,
      "grad_norm": 2.4216200751888217,
      "learning_rate": 1.7955824031505097e-06,
      "loss": 0.3741,
      "step": 1815
    },
    {
      "epoch": 0.7514450867052023,
      "grad_norm": 2.3925423466955884,
      "learning_rate": 1.7679971758500836e-06,
      "loss": 0.3652,
      "step": 1820
    },
    {
      "epoch": 0.7535094962840627,
      "grad_norm": 2.386508421187067,
      "learning_rate": 1.7405799044032356e-06,
      "loss": 0.3685,
      "step": 1825
    },
    {
      "epoch": 0.7555739058629232,
      "grad_norm": 2.439989821363405,
      "learning_rate": 1.7133320135894233e-06,
      "loss": 0.3653,
      "step": 1830
    },
    {
      "epoch": 0.7576383154417836,
      "grad_norm": 2.3039350711469186,
      "learning_rate": 1.686254919385985e-06,
      "loss": 0.3783,
      "step": 1835
    },
    {
      "epoch": 0.7597027250206441,
      "grad_norm": 2.216200410510777,
      "learning_rate": 1.6593500288945574e-06,
      "loss": 0.3646,
      "step": 1840
    },
    {
      "epoch": 0.7617671345995045,
      "grad_norm": 2.4989495842816405,
      "learning_rate": 1.6326187402679577e-06,
      "loss": 0.3741,
      "step": 1845
    },
    {
      "epoch": 0.763831544178365,
      "grad_norm": 2.4154932431289673,
      "learning_rate": 1.6060624426375222e-06,
      "loss": 0.3698,
      "step": 1850
    },
    {
      "epoch": 0.7658959537572254,
      "grad_norm": 2.473649546182009,
      "learning_rate": 1.5796825160409168e-06,
      "loss": 0.3672,
      "step": 1855
    },
    {
      "epoch": 0.7679603633360859,
      "grad_norm": 2.355038901564694,
      "learning_rate": 1.5534803313504215e-06,
      "loss": 0.3566,
      "step": 1860
    },
    {
      "epoch": 0.7700247729149463,
      "grad_norm": 2.3588413583977625,
      "learning_rate": 1.5274572502016982e-06,
      "loss": 0.3711,
      "step": 1865
    },
    {
      "epoch": 0.7720891824938068,
      "grad_norm": 2.401065534914905,
      "learning_rate": 1.5016146249230229e-06,
      "loss": 0.3736,
      "step": 1870
    },
    {
      "epoch": 0.7741535920726672,
      "grad_norm": 2.342195370497388,
      "learning_rate": 1.4759537984650114e-06,
      "loss": 0.3626,
      "step": 1875
    },
    {
      "epoch": 0.7762180016515277,
      "grad_norm": 2.3601036223952634,
      "learning_rate": 1.4504761043308346e-06,
      "loss": 0.3534,
      "step": 1880
    },
    {
      "epoch": 0.7782824112303881,
      "grad_norm": 2.3058771526048405,
      "learning_rate": 1.425182866506918e-06,
      "loss": 0.3633,
      "step": 1885
    },
    {
      "epoch": 0.7803468208092486,
      "grad_norm": 2.308152081660379,
      "learning_rate": 1.4000753993941369e-06,
      "loss": 0.3723,
      "step": 1890
    },
    {
      "epoch": 0.782411230388109,
      "grad_norm": 2.4604302877257664,
      "learning_rate": 1.3751550077395181e-06,
      "loss": 0.3523,
      "step": 1895
    },
    {
      "epoch": 0.7844756399669695,
      "grad_norm": 2.551739458655361,
      "learning_rate": 1.3504229865684326e-06,
      "loss": 0.3706,
      "step": 1900
    },
    {
      "epoch": 0.7865400495458299,
      "grad_norm": 2.4143979686568104,
      "learning_rate": 1.3258806211172954e-06,
      "loss": 0.3588,
      "step": 1905
    },
    {
      "epoch": 0.7886044591246903,
      "grad_norm": 2.364426767604513,
      "learning_rate": 1.3015291867667757e-06,
      "loss": 0.3508,
      "step": 1910
    },
    {
      "epoch": 0.7906688687035508,
      "grad_norm": 2.3233651206895147,
      "learning_rate": 1.2773699489755343e-06,
      "loss": 0.3571,
      "step": 1915
    },
    {
      "epoch": 0.7927332782824112,
      "grad_norm": 2.3196733264987675,
      "learning_rate": 1.2534041632144407e-06,
      "loss": 0.3405,
      "step": 1920
    },
    {
      "epoch": 0.7947976878612717,
      "grad_norm": 2.347467115991222,
      "learning_rate": 1.2296330749013408e-06,
      "loss": 0.3502,
      "step": 1925
    },
    {
      "epoch": 0.7968620974401321,
      "grad_norm": 2.3004029534314174,
      "learning_rate": 1.2060579193363431e-06,
      "loss": 0.3564,
      "step": 1930
    },
    {
      "epoch": 0.7989265070189926,
      "grad_norm": 2.5374233309055083,
      "learning_rate": 1.1826799216376128e-06,
      "loss": 0.3778,
      "step": 1935
    },
    {
      "epoch": 0.800990916597853,
      "grad_norm": 2.4431483706896007,
      "learning_rate": 1.1595002966777152e-06,
      "loss": 0.356,
      "step": 1940
    },
    {
      "epoch": 0.8030553261767135,
      "grad_norm": 2.35423434746266,
      "learning_rate": 1.136520249020478e-06,
      "loss": 0.3491,
      "step": 1945
    },
    {
      "epoch": 0.8051197357555739,
      "grad_norm": 2.3767268596250672,
      "learning_rate": 1.113740972858398e-06,
      "loss": 0.348,
      "step": 1950
    },
    {
      "epoch": 0.8071841453344344,
      "grad_norm": 2.4631890684934965,
      "learning_rate": 1.0911636519505791e-06,
      "loss": 0.3671,
      "step": 1955
    },
    {
      "epoch": 0.8092485549132948,
      "grad_norm": 2.201423417502954,
      "learning_rate": 1.068789459561223e-06,
      "loss": 0.3459,
      "step": 1960
    },
    {
      "epoch": 0.8113129644921553,
      "grad_norm": 2.36077824921438,
      "learning_rate": 1.0466195583986487e-06,
      "loss": 0.3639,
      "step": 1965
    },
    {
      "epoch": 0.8133773740710157,
      "grad_norm": 2.354003763601073,
      "learning_rate": 1.024655100554882e-06,
      "loss": 0.3643,
      "step": 1970
    },
    {
      "epoch": 0.8154417836498762,
      "grad_norm": 2.4101822702917004,
      "learning_rate": 1.002897227445777e-06,
      "loss": 0.3617,
      "step": 1975
    },
    {
      "epoch": 0.8175061932287366,
      "grad_norm": 2.4833584786174163,
      "learning_rate": 9.813470697517009e-07,
      "loss": 0.3501,
      "step": 1980
    },
    {
      "epoch": 0.8195706028075971,
      "grad_norm": 2.3627937000377393,
      "learning_rate": 9.600057473587876e-07,
      "loss": 0.3671,
      "step": 1985
    },
    {
      "epoch": 0.8216350123864574,
      "grad_norm": 2.415763997145714,
      "learning_rate": 9.388743693007247e-07,
      "loss": 0.3541,
      "step": 1990
    },
    {
      "epoch": 0.8236994219653179,
      "grad_norm": 2.4671751077694406,
      "learning_rate": 9.1795403370113e-07,
      "loss": 0.3535,
      "step": 1995
    },
    {
      "epoch": 0.8257638315441783,
      "grad_norm": 2.3775799288896287,
      "learning_rate": 8.972458277164886e-07,
      "loss": 0.3516,
      "step": 2000
    },
    {
      "epoch": 0.8278282411230388,
      "grad_norm": 2.2938470233646604,
      "learning_rate": 8.76750827479651e-07,
      "loss": 0.3398,
      "step": 2005
    },
    {
      "epoch": 0.8298926507018992,
      "grad_norm": 2.2933382256010497,
      "learning_rate": 8.564700980439122e-07,
      "loss": 0.3608,
      "step": 2010
    },
    {
      "epoch": 0.8319570602807597,
      "grad_norm": 2.3382609787220785,
      "learning_rate": 8.364046933276642e-07,
      "loss": 0.3666,
      "step": 2015
    },
    {
      "epoch": 0.8340214698596201,
      "grad_norm": 2.44152050789561,
      "learning_rate": 8.165556560596316e-07,
      "loss": 0.3485,
      "step": 2020
    },
    {
      "epoch": 0.8360858794384806,
      "grad_norm": 2.3858823443159034,
      "learning_rate": 7.969240177246779e-07,
      "loss": 0.3547,
      "step": 2025
    },
    {
      "epoch": 0.838150289017341,
      "grad_norm": 2.453983334202833,
      "learning_rate": 7.775107985102087e-07,
      "loss": 0.3651,
      "step": 2030
    },
    {
      "epoch": 0.8402146985962015,
      "grad_norm": 2.5029504947887875,
      "learning_rate": 7.583170072531543e-07,
      "loss": 0.3509,
      "step": 2035
    },
    {
      "epoch": 0.8422791081750619,
      "grad_norm": 2.482629951124892,
      "learning_rate": 7.393436413875411e-07,
      "loss": 0.3476,
      "step": 2040
    },
    {
      "epoch": 0.8443435177539224,
      "grad_norm": 2.283216056409345,
      "learning_rate": 7.205916868926604e-07,
      "loss": 0.3526,
      "step": 2045
    },
    {
      "epoch": 0.8464079273327828,
      "grad_norm": 2.3590631737094845,
      "learning_rate": 7.020621182418347e-07,
      "loss": 0.34,
      "step": 2050
    },
    {
      "epoch": 0.8484723369116433,
      "grad_norm": 2.6201296500024553,
      "learning_rate": 6.837558983517723e-07,
      "loss": 0.3535,
      "step": 2055
    },
    {
      "epoch": 0.8505367464905037,
      "grad_norm": 2.4227114258373095,
      "learning_rate": 6.656739785325278e-07,
      "loss": 0.3507,
      "step": 2060
    },
    {
      "epoch": 0.8526011560693642,
      "grad_norm": 2.390954329942295,
      "learning_rate": 6.478172984380687e-07,
      "loss": 0.3532,
      "step": 2065
    },
    {
      "epoch": 0.8546655656482246,
      "grad_norm": 2.556923085001048,
      "learning_rate": 6.301867860174443e-07,
      "loss": 0.3434,
      "step": 2070
    },
    {
      "epoch": 0.856729975227085,
      "grad_norm": 2.3169295549044335,
      "learning_rate": 6.127833574665642e-07,
      "loss": 0.3459,
      "step": 2075
    },
    {
      "epoch": 0.8587943848059455,
      "grad_norm": 2.3497867089138644,
      "learning_rate": 5.956079171805818e-07,
      "loss": 0.3415,
      "step": 2080
    },
    {
      "epoch": 0.860858794384806,
      "grad_norm": 2.1735667040509137,
      "learning_rate": 5.78661357706905e-07,
      "loss": 0.3379,
      "step": 2085
    },
    {
      "epoch": 0.8629232039636664,
      "grad_norm": 2.5779842919745355,
      "learning_rate": 5.619445596988026e-07,
      "loss": 0.3425,
      "step": 2090
    },
    {
      "epoch": 0.8649876135425268,
      "grad_norm": 2.401992025146321,
      "learning_rate": 5.45458391869651e-07,
      "loss": 0.3489,
      "step": 2095
    },
    {
      "epoch": 0.8670520231213873,
      "grad_norm": 2.3105243423015587,
      "learning_rate": 5.29203710947781e-07,
      "loss": 0.335,
      "step": 2100
    },
    {
      "epoch": 0.8691164327002477,
      "grad_norm": 2.5027833045418943,
      "learning_rate": 5.131813616319641e-07,
      "loss": 0.35,
      "step": 2105
    },
    {
      "epoch": 0.8711808422791082,
      "grad_norm": 2.492666189587263,
      "learning_rate": 4.973921765475093e-07,
      "loss": 0.3419,
      "step": 2110
    },
    {
      "epoch": 0.8732452518579686,
      "grad_norm": 2.476326337718451,
      "learning_rate": 4.818369762030001e-07,
      "loss": 0.3437,
      "step": 2115
    },
    {
      "epoch": 0.8753096614368291,
      "grad_norm": 2.4895934072647923,
      "learning_rate": 4.6651656894765307e-07,
      "loss": 0.345,
      "step": 2120
    },
    {
      "epoch": 0.8773740710156895,
      "grad_norm": 2.432351776306209,
      "learning_rate": 4.5143175092931257e-07,
      "loss": 0.3457,
      "step": 2125
    },
    {
      "epoch": 0.87943848059455,
      "grad_norm": 2.5446256871769894,
      "learning_rate": 4.3658330605307496e-07,
      "loss": 0.3501,
      "step": 2130
    },
    {
      "epoch": 0.8815028901734104,
      "grad_norm": 2.400237240106358,
      "learning_rate": 4.2197200594055135e-07,
      "loss": 0.3417,
      "step": 2135
    },
    {
      "epoch": 0.8835672997522709,
      "grad_norm": 2.420863056343347,
      "learning_rate": 4.0759860988977715e-07,
      "loss": 0.3336,
      "step": 2140
    },
    {
      "epoch": 0.8856317093311313,
      "grad_norm": 2.5855831775345317,
      "learning_rate": 3.934638648357414e-07,
      "loss": 0.3473,
      "step": 2145
    },
    {
      "epoch": 0.8876961189099918,
      "grad_norm": 2.4571094742538633,
      "learning_rate": 3.795685053115805e-07,
      "loss": 0.3425,
      "step": 2150
    },
    {
      "epoch": 0.8897605284888522,
      "grad_norm": 2.318969751255526,
      "learning_rate": 3.659132534104054e-07,
      "loss": 0.339,
      "step": 2155
    },
    {
      "epoch": 0.8918249380677127,
      "grad_norm": 2.464561015161846,
      "learning_rate": 3.5249881874777235e-07,
      "loss": 0.3415,
      "step": 2160
    },
    {
      "epoch": 0.8938893476465731,
      "grad_norm": 2.5739185699856653,
      "learning_rate": 3.3932589842481434e-07,
      "loss": 0.3549,
      "step": 2165
    },
    {
      "epoch": 0.8959537572254336,
      "grad_norm": 2.3924860465104554,
      "learning_rate": 3.2639517699200765e-07,
      "loss": 0.3501,
      "step": 2170
    },
    {
      "epoch": 0.898018166804294,
      "grad_norm": 2.4561049545962983,
      "learning_rate": 3.1370732641360424e-07,
      "loss": 0.3382,
      "step": 2175
    },
    {
      "epoch": 0.9000825763831544,
      "grad_norm": 2.4531360622529985,
      "learning_rate": 3.0126300603270763e-07,
      "loss": 0.3386,
      "step": 2180
    },
    {
      "epoch": 0.9021469859620148,
      "grad_norm": 2.3726671856862698,
      "learning_rate": 2.89062862537009e-07,
      "loss": 0.3382,
      "step": 2185
    },
    {
      "epoch": 0.9042113955408753,
      "grad_norm": 2.4508531032953704,
      "learning_rate": 2.7710752992518906e-07,
      "loss": 0.3475,
      "step": 2190
    },
    {
      "epoch": 0.9062758051197357,
      "grad_norm": 2.344783384097571,
      "learning_rate": 2.653976294739596e-07,
      "loss": 0.3435,
      "step": 2195
    },
    {
      "epoch": 0.9083402146985962,
      "grad_norm": 2.3832943118751495,
      "learning_rate": 2.539337697057853e-07,
      "loss": 0.3386,
      "step": 2200
    },
    {
      "epoch": 0.9104046242774566,
      "grad_norm": 2.2656029102160917,
      "learning_rate": 2.4271654635726003e-07,
      "loss": 0.326,
      "step": 2205
    },
    {
      "epoch": 0.9124690338563171,
      "grad_norm": 2.531439971555802,
      "learning_rate": 2.3174654234814854e-07,
      "loss": 0.3336,
      "step": 2210
    },
    {
      "epoch": 0.9145334434351775,
      "grad_norm": 2.526736652879497,
      "learning_rate": 2.2102432775109151e-07,
      "loss": 0.3431,
      "step": 2215
    },
    {
      "epoch": 0.916597853014038,
      "grad_norm": 2.499234877151574,
      "learning_rate": 2.1055045976198286e-07,
      "loss": 0.3274,
      "step": 2220
    },
    {
      "epoch": 0.9186622625928984,
      "grad_norm": 2.347953826058282,
      "learning_rate": 2.0032548267101547e-07,
      "loss": 0.3415,
      "step": 2225
    },
    {
      "epoch": 0.9207266721717589,
      "grad_norm": 2.2781319110873923,
      "learning_rate": 1.9034992783439398e-07,
      "loss": 0.3313,
      "step": 2230
    },
    {
      "epoch": 0.9227910817506193,
      "grad_norm": 2.4985215803253777,
      "learning_rate": 1.8062431364672394e-07,
      "loss": 0.347,
      "step": 2235
    },
    {
      "epoch": 0.9248554913294798,
      "grad_norm": 2.3523550820055807,
      "learning_rate": 1.7114914551407125e-07,
      "loss": 0.3405,
      "step": 2240
    },
    {
      "epoch": 0.9269199009083402,
      "grad_norm": 2.35149253079431,
      "learning_rate": 1.619249158276981e-07,
      "loss": 0.3396,
      "step": 2245
    },
    {
      "epoch": 0.9289843104872006,
      "grad_norm": 2.5391046088636724,
      "learning_rate": 1.529521039384757e-07,
      "loss": 0.3382,
      "step": 2250
    },
    {
      "epoch": 0.9310487200660611,
      "grad_norm": 2.436611958199382,
      "learning_rate": 1.4423117613197535e-07,
      "loss": 0.3507,
      "step": 2255
    },
    {
      "epoch": 0.9331131296449215,
      "grad_norm": 2.368900058293189,
      "learning_rate": 1.357625856042344e-07,
      "loss": 0.3436,
      "step": 2260
    },
    {
      "epoch": 0.935177539223782,
      "grad_norm": 2.483713260867796,
      "learning_rate": 1.275467724382068e-07,
      "loss": 0.3493,
      "step": 2265
    },
    {
      "epoch": 0.9372419488026424,
      "grad_norm": 2.3378032035177556,
      "learning_rate": 1.1958416358089308e-07,
      "loss": 0.3366,
      "step": 2270
    },
    {
      "epoch": 0.9393063583815029,
      "grad_norm": 2.427445932737846,
      "learning_rate": 1.1187517282115478e-07,
      "loss": 0.3357,
      "step": 2275
    },
    {
      "epoch": 0.9413707679603633,
      "grad_norm": 2.45979849427157,
      "learning_rate": 1.0442020076820992e-07,
      "loss": 0.3432,
      "step": 2280
    },
    {
      "epoch": 0.9434351775392238,
      "grad_norm": 2.4102972284673485,
      "learning_rate": 9.72196348308152e-08,
      "loss": 0.3421,
      "step": 2285
    },
    {
      "epoch": 0.9454995871180842,
      "grad_norm": 2.4147926771751127,
      "learning_rate": 9.027384919713167e-08,
      "loss": 0.3449,
      "step": 2290
    },
    {
      "epoch": 0.9475639966969447,
      "grad_norm": 2.492908131419299,
      "learning_rate": 8.358320481528404e-08,
      "loss": 0.3349,
      "step": 2295
    },
    {
      "epoch": 0.9496284062758051,
      "grad_norm": 2.445696827488344,
      "learning_rate": 7.714804937459964e-08,
      "loss": 0.3327,
      "step": 2300
    },
    {
      "epoch": 0.9516928158546656,
      "grad_norm": 2.4859237981346434,
      "learning_rate": 7.096871728754173e-08,
      "loss": 0.3369,
      "step": 2305
    },
    {
      "epoch": 0.953757225433526,
      "grad_norm": 2.4755163895408483,
      "learning_rate": 6.504552967233014e-08,
      "loss": 0.3341,
      "step": 2310
    },
    {
      "epoch": 0.9558216350123865,
      "grad_norm": 2.4903780675355947,
      "learning_rate": 5.9378794336255595e-08,
      "loss": 0.3259,
      "step": 2315
    },
    {
      "epoch": 0.9578860445912469,
      "grad_norm": 2.3608897794698986,
      "learning_rate": 5.396880575968266e-08,
      "loss": 0.34,
      "step": 2320
    },
    {
      "epoch": 0.9599504541701074,
      "grad_norm": 2.349066629113589,
      "learning_rate": 4.881584508074688e-08,
      "loss": 0.3285,
      "step": 2325
    },
    {
      "epoch": 0.9620148637489678,
      "grad_norm": 2.475193042134275,
      "learning_rate": 4.39201800807465e-08,
      "loss": 0.339,
      "step": 2330
    },
    {
      "epoch": 0.9640792733278283,
      "grad_norm": 2.3656425365762384,
      "learning_rate": 3.92820651702247e-08,
      "loss": 0.3358,
      "step": 2335
    },
    {
      "epoch": 0.9661436829066887,
      "grad_norm": 2.441572724635035,
      "learning_rate": 3.4901741375747975e-08,
      "loss": 0.3471,
      "step": 2340
    },
    {
      "epoch": 0.9682080924855492,
      "grad_norm": 2.3678044079850413,
      "learning_rate": 3.077943632738556e-08,
      "loss": 0.3499,
      "step": 2345
    },
    {
      "epoch": 0.9702725020644096,
      "grad_norm": 2.630292946781589,
      "learning_rate": 2.6915364246875574e-08,
      "loss": 0.341,
      "step": 2350
    },
    {
      "epoch": 0.9723369116432701,
      "grad_norm": 2.388897601971072,
      "learning_rate": 2.3309725936493387e-08,
      "loss": 0.3437,
      "step": 2355
    },
    {
      "epoch": 0.9744013212221305,
      "grad_norm": 2.4665523794028785,
      "learning_rate": 1.9962708768620498e-08,
      "loss": 0.3453,
      "step": 2360
    },
    {
      "epoch": 0.976465730800991,
      "grad_norm": 2.448813469296534,
      "learning_rate": 1.687448667600067e-08,
      "loss": 0.3408,
      "step": 2365
    },
    {
      "epoch": 0.9785301403798513,
      "grad_norm": 2.3718128749961576,
      "learning_rate": 1.4045220142708816e-08,
      "loss": 0.35,
      "step": 2370
    },
    {
      "epoch": 0.9805945499587118,
      "grad_norm": 2.4841177332012325,
      "learning_rate": 1.1475056195804358e-08,
      "loss": 0.3518,
      "step": 2375
    },
    {
      "epoch": 0.9826589595375722,
      "grad_norm": 2.493744247830282,
      "learning_rate": 9.164128397695648e-09,
      "loss": 0.332,
      "step": 2380
    },
    {
      "epoch": 0.9847233691164327,
      "grad_norm": 2.2819847476640107,
      "learning_rate": 7.112556839196649e-09,
      "loss": 0.3301,
      "step": 2385
    },
    {
      "epoch": 0.9867877786952931,
      "grad_norm": 2.4247247484873173,
      "learning_rate": 5.3204481332880255e-09,
      "loss": 0.3433,
      "step": 2390
    },
    {
      "epoch": 0.9888521882741536,
      "grad_norm": 2.4373319136694285,
      "learning_rate": 3.787895409574915e-09,
      "loss": 0.3388,
      "step": 2395
    },
    {
      "epoch": 0.990916597853014,
      "grad_norm": 2.3711173222964037,
      "learning_rate": 2.514978309447469e-09,
      "loss": 0.3433,
      "step": 2400
    },
    {
      "epoch": 0.9929810074318745,
      "grad_norm": 2.3426000581900652,
      "learning_rate": 1.501762981945265e-09,
      "loss": 0.3368,
      "step": 2405
    },
    {
      "epoch": 0.9950454170107349,
      "grad_norm": 2.35822764782016,
      "learning_rate": 7.483020803145114e-10,
      "loss": 0.3396,
      "step": 2410
    },
    {
      "epoch": 0.9971098265895953,
      "grad_norm": 2.4725534953927233,
      "learning_rate": 2.5463475927745007e-10,
      "loss": 0.3401,
      "step": 2415
    },
    {
      "epoch": 0.9991742361684558,
      "grad_norm": 2.407515233988016,
      "learning_rate": 2.078667299343362e-11,
      "loss": 0.3298,
      "step": 2420
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.3946,
      "eval_samples_per_second": 2.946,
      "eval_steps_per_second": 0.884,
      "step": 2422
    },
    {
      "epoch": 1.0,
      "step": 2422,
      "total_flos": 253558763028480.0,
      "train_loss": 0.530042272082643,
      "train_runtime": 21564.4437,
      "train_samples_per_second": 1.797,
      "train_steps_per_second": 0.112
    }
  ],
  "logging_steps": 5,
  "max_steps": 2422,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 253558763028480.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}