{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1656,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006038647342995169,
      "grad_norm": 11.647146166653753,
      "learning_rate": 6.02409638554217e-08,
      "loss": 1.4319,
      "step": 1
    },
    {
      "epoch": 0.0030193236714975845,
      "grad_norm": 10.76596246435424,
      "learning_rate": 3.0120481927710845e-07,
      "loss": 1.4423,
      "step": 5
    },
    {
      "epoch": 0.006038647342995169,
      "grad_norm": 8.345396159148025,
      "learning_rate": 6.024096385542169e-07,
      "loss": 1.4123,
      "step": 10
    },
    {
      "epoch": 0.009057971014492754,
      "grad_norm": 5.728658238969367,
      "learning_rate": 9.036144578313254e-07,
      "loss": 1.3779,
      "step": 15
    },
    {
      "epoch": 0.012077294685990338,
      "grad_norm": 8.858123210879867,
      "learning_rate": 1.2048192771084338e-06,
      "loss": 1.307,
      "step": 20
    },
    {
      "epoch": 0.015096618357487922,
      "grad_norm": 5.956632109318297,
      "learning_rate": 1.5060240963855425e-06,
      "loss": 1.2575,
      "step": 25
    },
    {
      "epoch": 0.018115942028985508,
      "grad_norm": 2.691768583064308,
      "learning_rate": 1.8072289156626508e-06,
      "loss": 1.1412,
      "step": 30
    },
    {
      "epoch": 0.021135265700483092,
      "grad_norm": 5.521729994682871,
      "learning_rate": 2.1084337349397595e-06,
      "loss": 1.0901,
      "step": 35
    },
    {
      "epoch": 0.024154589371980676,
      "grad_norm": 1.950978681738152,
      "learning_rate": 2.4096385542168676e-06,
      "loss": 1.0489,
      "step": 40
    },
    {
      "epoch": 0.02717391304347826,
      "grad_norm": 1.6004592652392855,
      "learning_rate": 2.710843373493976e-06,
      "loss": 1.0294,
      "step": 45
    },
    {
      "epoch": 0.030193236714975844,
      "grad_norm": 1.8752688704592362,
      "learning_rate": 3.012048192771085e-06,
      "loss": 1.0009,
      "step": 50
    },
    {
      "epoch": 0.03321256038647343,
      "grad_norm": 1.730463603619638,
      "learning_rate": 3.313253012048193e-06,
      "loss": 0.9798,
      "step": 55
    },
    {
      "epoch": 0.036231884057971016,
      "grad_norm": 1.5964460177303024,
      "learning_rate": 3.6144578313253016e-06,
      "loss": 0.9684,
      "step": 60
    },
    {
      "epoch": 0.0392512077294686,
      "grad_norm": 1.6451870176416807,
      "learning_rate": 3.91566265060241e-06,
      "loss": 0.9598,
      "step": 65
    },
    {
      "epoch": 0.042270531400966184,
      "grad_norm": 1.7104520603402644,
      "learning_rate": 4.216867469879519e-06,
      "loss": 0.9603,
      "step": 70
    },
    {
      "epoch": 0.04528985507246377,
      "grad_norm": 1.5767703064322158,
      "learning_rate": 4.518072289156627e-06,
      "loss": 0.9292,
      "step": 75
    },
    {
      "epoch": 0.04830917874396135,
      "grad_norm": 1.451795891571848,
      "learning_rate": 4.819277108433735e-06,
      "loss": 0.9191,
      "step": 80
    },
    {
      "epoch": 0.051328502415458936,
      "grad_norm": 1.5778301532020658,
      "learning_rate": 5.120481927710844e-06,
      "loss": 0.9131,
      "step": 85
    },
    {
      "epoch": 0.05434782608695652,
      "grad_norm": 1.5490469939143814,
      "learning_rate": 5.421686746987952e-06,
      "loss": 0.9147,
      "step": 90
    },
    {
      "epoch": 0.057367149758454104,
      "grad_norm": 1.6958988602467524,
      "learning_rate": 5.722891566265061e-06,
      "loss": 0.9031,
      "step": 95
    },
    {
      "epoch": 0.06038647342995169,
      "grad_norm": 1.695663023457705,
      "learning_rate": 6.02409638554217e-06,
      "loss": 0.8903,
      "step": 100
    },
    {
      "epoch": 0.06340579710144928,
      "grad_norm": 1.696466007675861,
      "learning_rate": 6.325301204819277e-06,
      "loss": 0.8852,
      "step": 105
    },
    {
      "epoch": 0.06642512077294686,
      "grad_norm": 1.607366527826758,
      "learning_rate": 6.626506024096386e-06,
      "loss": 0.8729,
      "step": 110
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 1.592578389550239,
      "learning_rate": 6.927710843373494e-06,
      "loss": 0.8685,
      "step": 115
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 1.5599550038753036,
      "learning_rate": 7.228915662650603e-06,
      "loss": 0.8643,
      "step": 120
    },
    {
      "epoch": 0.07548309178743962,
      "grad_norm": 1.7526452577051699,
      "learning_rate": 7.530120481927712e-06,
      "loss": 0.8472,
      "step": 125
    },
    {
      "epoch": 0.0785024154589372,
      "grad_norm": 1.570734587958465,
      "learning_rate": 7.83132530120482e-06,
      "loss": 0.8457,
      "step": 130
    },
    {
      "epoch": 0.08152173913043478,
      "grad_norm": 1.3578392369216397,
      "learning_rate": 8.132530120481928e-06,
      "loss": 0.8419,
      "step": 135
    },
    {
      "epoch": 0.08454106280193237,
      "grad_norm": 1.5345711918538654,
      "learning_rate": 8.433734939759038e-06,
      "loss": 0.8293,
      "step": 140
    },
    {
      "epoch": 0.08756038647342995,
      "grad_norm": 1.5199887652875657,
      "learning_rate": 8.734939759036145e-06,
      "loss": 0.8116,
      "step": 145
    },
    {
      "epoch": 0.09057971014492754,
      "grad_norm": 1.4329178750495808,
      "learning_rate": 9.036144578313254e-06,
      "loss": 0.8131,
      "step": 150
    },
    {
      "epoch": 0.09359903381642512,
      "grad_norm": 1.5357940171741657,
      "learning_rate": 9.337349397590362e-06,
      "loss": 0.8109,
      "step": 155
    },
    {
      "epoch": 0.0966183574879227,
      "grad_norm": 1.4127066136547346,
      "learning_rate": 9.63855421686747e-06,
      "loss": 0.8116,
      "step": 160
    },
    {
      "epoch": 0.09963768115942029,
      "grad_norm": 1.5153723323216768,
      "learning_rate": 9.93975903614458e-06,
      "loss": 0.8067,
      "step": 165
    },
    {
      "epoch": 0.10265700483091787,
      "grad_norm": 1.3541934052887161,
      "learning_rate": 9.999822178354131e-06,
      "loss": 0.7957,
      "step": 170
    },
    {
      "epoch": 0.10567632850241546,
      "grad_norm": 1.3822332075059867,
      "learning_rate": 9.999099799595088e-06,
      "loss": 0.7915,
      "step": 175
    },
    {
      "epoch": 0.10869565217391304,
      "grad_norm": 1.4372310169992935,
      "learning_rate": 9.997821830092095e-06,
      "loss": 0.7999,
      "step": 180
    },
    {
      "epoch": 0.11171497584541062,
      "grad_norm": 1.6716440887834658,
      "learning_rate": 9.995988411876328e-06,
      "loss": 0.7992,
      "step": 185
    },
    {
      "epoch": 0.11473429951690821,
      "grad_norm": 1.5723808182751033,
      "learning_rate": 9.993599748710505e-06,
      "loss": 0.783,
      "step": 190
    },
    {
      "epoch": 0.11775362318840579,
      "grad_norm": 1.4610043673248276,
      "learning_rate": 9.990656106066257e-06,
      "loss": 0.7827,
      "step": 195
    },
    {
      "epoch": 0.12077294685990338,
      "grad_norm": 1.4846567788944018,
      "learning_rate": 9.9871578110946e-06,
      "loss": 0.7773,
      "step": 200
    },
    {
      "epoch": 0.12379227053140096,
      "grad_norm": 1.4132792684184563,
      "learning_rate": 9.983105252589599e-06,
      "loss": 0.7662,
      "step": 205
    },
    {
      "epoch": 0.12681159420289856,
      "grad_norm": 1.3392248870506875,
      "learning_rate": 9.978498880945138e-06,
      "loss": 0.7643,
      "step": 210
    },
    {
      "epoch": 0.12983091787439613,
      "grad_norm": 1.4340057279445353,
      "learning_rate": 9.97333920810488e-06,
      "loss": 0.7791,
      "step": 215
    },
    {
      "epoch": 0.13285024154589373,
      "grad_norm": 1.2837053197488815,
      "learning_rate": 9.967626807505359e-06,
      "loss": 0.7647,
      "step": 220
    },
    {
      "epoch": 0.1358695652173913,
      "grad_norm": 1.4133054167156096,
      "learning_rate": 9.961362314012258e-06,
      "loss": 0.7851,
      "step": 225
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.4609903990792592,
      "learning_rate": 9.954546423849842e-06,
      "loss": 0.7689,
      "step": 230
    },
    {
      "epoch": 0.14190821256038647,
      "grad_norm": 1.4430590677898403,
      "learning_rate": 9.947179894523594e-06,
      "loss": 0.7483,
      "step": 235
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 1.1927964292128603,
      "learning_rate": 9.93926354473601e-06,
      "loss": 0.7549,
      "step": 240
    },
    {
      "epoch": 0.14794685990338163,
      "grad_norm": 1.267141348725353,
      "learning_rate": 9.930798254295628e-06,
      "loss": 0.7528,
      "step": 245
    },
    {
      "epoch": 0.15096618357487923,
      "grad_norm": 1.2738788294787875,
      "learning_rate": 9.921784964019234e-06,
      "loss": 0.7588,
      "step": 250
    },
    {
      "epoch": 0.1539855072463768,
      "grad_norm": 1.3665251025923502,
      "learning_rate": 9.91222467562731e-06,
      "loss": 0.752,
      "step": 255
    },
    {
      "epoch": 0.1570048309178744,
      "grad_norm": 1.2895638701049197,
      "learning_rate": 9.902118451632694e-06,
      "loss": 0.7443,
      "step": 260
    },
    {
      "epoch": 0.16002415458937197,
      "grad_norm": 1.2792248819171141,
      "learning_rate": 9.891467415222511e-06,
      "loss": 0.7389,
      "step": 265
    },
    {
      "epoch": 0.16304347826086957,
      "grad_norm": 1.2577426141477723,
      "learning_rate": 9.880272750133328e-06,
      "loss": 0.7495,
      "step": 270
    },
    {
      "epoch": 0.16606280193236714,
      "grad_norm": 1.2581092756100691,
      "learning_rate": 9.868535700519605e-06,
      "loss": 0.7574,
      "step": 275
    },
    {
      "epoch": 0.16908212560386474,
      "grad_norm": 1.3915721437755781,
      "learning_rate": 9.856257570815415e-06,
      "loss": 0.7451,
      "step": 280
    },
    {
      "epoch": 0.1721014492753623,
      "grad_norm": 1.4182529730794977,
      "learning_rate": 9.843439725589481e-06,
      "loss": 0.7392,
      "step": 285
    },
    {
      "epoch": 0.1751207729468599,
      "grad_norm": 1.4007115307336426,
      "learning_rate": 9.83008358939351e-06,
      "loss": 0.7488,
      "step": 290
    },
    {
      "epoch": 0.17814009661835747,
      "grad_norm": 1.2602901130922803,
      "learning_rate": 9.81619064660388e-06,
      "loss": 0.7481,
      "step": 295
    },
    {
      "epoch": 0.18115942028985507,
      "grad_norm": 1.230299206050831,
      "learning_rate": 9.801762441256663e-06,
      "loss": 0.7277,
      "step": 300
    },
    {
      "epoch": 0.18417874396135267,
      "grad_norm": 1.1834363014701523,
      "learning_rate": 9.786800576876026e-06,
      "loss": 0.7458,
      "step": 305
    },
    {
      "epoch": 0.18719806763285024,
      "grad_norm": 1.257138541703525,
      "learning_rate": 9.77130671629602e-06,
      "loss": 0.7338,
      "step": 310
    },
    {
      "epoch": 0.19021739130434784,
      "grad_norm": 1.316511247899504,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.7382,
      "step": 315
    },
    {
      "epoch": 0.1932367149758454,
      "grad_norm": 1.2036218489959571,
      "learning_rate": 9.738729953308104e-06,
      "loss": 0.734,
      "step": 320
    },
    {
      "epoch": 0.196256038647343,
      "grad_norm": 1.3147689783592325,
      "learning_rate": 9.72165067142163e-06,
      "loss": 0.7133,
      "step": 325
    },
    {
      "epoch": 0.19927536231884058,
      "grad_norm": 1.2043088743921089,
      "learning_rate": 9.70404663397628e-06,
      "loss": 0.724,
      "step": 330
    },
    {
      "epoch": 0.20229468599033817,
      "grad_norm": 1.184487132016919,
      "learning_rate": 9.68591979745235e-06,
      "loss": 0.7227,
      "step": 335
    },
    {
      "epoch": 0.20531400966183574,
      "grad_norm": 1.1868803234306302,
      "learning_rate": 9.667272176433063e-06,
      "loss": 0.7211,
      "step": 340
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 1.354442450720458,
      "learning_rate": 9.648105843380674e-06,
      "loss": 0.7246,
      "step": 345
    },
    {
      "epoch": 0.2113526570048309,
      "grad_norm": 1.3221102565499465,
      "learning_rate": 9.628422928406133e-06,
      "loss": 0.723,
      "step": 350
    },
    {
      "epoch": 0.2143719806763285,
      "grad_norm": 1.298626503464365,
      "learning_rate": 9.608225619032361e-06,
      "loss": 0.7204,
      "step": 355
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 1.2235482139383798,
      "learning_rate": 9.587516159951118e-06,
      "loss": 0.7157,
      "step": 360
    },
    {
      "epoch": 0.22041062801932368,
      "grad_norm": 1.4044097057246043,
      "learning_rate": 9.566296852773541e-06,
      "loss": 0.7444,
      "step": 365
    },
    {
      "epoch": 0.22342995169082125,
      "grad_norm": 1.204069094083874,
      "learning_rate": 9.544570055774348e-06,
      "loss": 0.7068,
      "step": 370
    },
    {
      "epoch": 0.22644927536231885,
      "grad_norm": 1.433962682853488,
      "learning_rate": 9.522338183629737e-06,
      "loss": 0.7136,
      "step": 375
    },
    {
      "epoch": 0.22946859903381642,
      "grad_norm": 1.3644548503760257,
      "learning_rate": 9.499603707149035e-06,
      "loss": 0.7121,
      "step": 380
    },
    {
      "epoch": 0.23248792270531402,
      "grad_norm": 1.3126699782875395,
      "learning_rate": 9.476369153000076e-06,
      "loss": 0.7184,
      "step": 385
    },
    {
      "epoch": 0.23550724637681159,
      "grad_norm": 1.1868983270075697,
      "learning_rate": 9.45263710342842e-06,
      "loss": 0.6992,
      "step": 390
    },
    {
      "epoch": 0.23852657004830918,
      "grad_norm": 1.4139162160182768,
      "learning_rate": 9.428410195970337e-06,
      "loss": 0.6926,
      "step": 395
    },
    {
      "epoch": 0.24154589371980675,
      "grad_norm": 1.1181104001140345,
      "learning_rate": 9.403691123159707e-06,
      "loss": 0.7041,
      "step": 400
    },
    {
      "epoch": 0.24456521739130435,
      "grad_norm": 1.4069308087516696,
      "learning_rate": 9.378482632228745e-06,
      "loss": 0.6956,
      "step": 405
    },
    {
      "epoch": 0.24758454106280192,
      "grad_norm": 1.230108931004278,
      "learning_rate": 9.352787524802707e-06,
      "loss": 0.7121,
      "step": 410
    },
    {
      "epoch": 0.2506038647342995,
      "grad_norm": 1.2509235012582307,
      "learning_rate": 9.326608656588502e-06,
      "loss": 0.7013,
      "step": 415
    },
    {
      "epoch": 0.2536231884057971,
      "grad_norm": 1.36604045335286,
      "learning_rate": 9.299948937057325e-06,
      "loss": 0.6953,
      "step": 420
    },
    {
      "epoch": 0.25664251207729466,
      "grad_norm": 1.1730575669226555,
      "learning_rate": 9.272811329121305e-06,
      "loss": 0.7144,
      "step": 425
    },
    {
      "epoch": 0.25966183574879226,
      "grad_norm": 1.1622429028280914,
      "learning_rate": 9.245198848804197e-06,
      "loss": 0.6991,
      "step": 430
    },
    {
      "epoch": 0.26268115942028986,
      "grad_norm": 1.264307120570429,
      "learning_rate": 9.217114564906208e-06,
      "loss": 0.6867,
      "step": 435
    },
    {
      "epoch": 0.26570048309178745,
      "grad_norm": 1.1848039169987097,
      "learning_rate": 9.188561598662921e-06,
      "loss": 0.691,
      "step": 440
    },
    {
      "epoch": 0.26871980676328505,
      "grad_norm": 1.2376999426889066,
      "learning_rate": 9.159543123398416e-06,
      "loss": 0.7033,
      "step": 445
    },
    {
      "epoch": 0.2717391304347826,
      "grad_norm": 1.2317420665829562,
      "learning_rate": 9.130062364172582e-06,
      "loss": 0.6907,
      "step": 450
    },
    {
      "epoch": 0.2747584541062802,
      "grad_norm": 1.2724956183147942,
      "learning_rate": 9.1001225974227e-06,
      "loss": 0.6805,
      "step": 455
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 1.1495012873362647,
      "learning_rate": 9.0697271505993e-06,
      "loss": 0.6813,
      "step": 460
    },
    {
      "epoch": 0.2807971014492754,
      "grad_norm": 1.267845133752999,
      "learning_rate": 9.038879401796358e-06,
      "loss": 0.6947,
      "step": 465
    },
    {
      "epoch": 0.28381642512077293,
      "grad_norm": 1.2127192185018594,
      "learning_rate": 9.00758277937586e-06,
      "loss": 0.6966,
      "step": 470
    },
    {
      "epoch": 0.28683574879227053,
      "grad_norm": 1.2447551629635505,
      "learning_rate": 8.975840761586772e-06,
      "loss": 0.7047,
      "step": 475
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 1.1928908039430874,
      "learning_rate": 8.94365687617849e-06,
      "loss": 0.6935,
      "step": 480
    },
    {
      "epoch": 0.2928743961352657,
      "grad_norm": 1.2672585178249127,
      "learning_rate": 8.911034700008757e-06,
      "loss": 0.6739,
      "step": 485
    },
    {
      "epoch": 0.29589371980676327,
      "grad_norm": 1.3798934886196015,
      "learning_rate": 8.87797785864615e-06,
      "loss": 0.6877,
      "step": 490
    },
    {
      "epoch": 0.29891304347826086,
      "grad_norm": 1.2248493689194777,
      "learning_rate": 8.844490025967126e-06,
      "loss": 0.6758,
      "step": 495
    },
    {
      "epoch": 0.30193236714975846,
      "grad_norm": 1.1804949407776295,
      "learning_rate": 8.810574923747729e-06,
      "loss": 0.6798,
      "step": 500
    },
    {
      "epoch": 0.30495169082125606,
      "grad_norm": 1.2545058603527033,
      "learning_rate": 8.776236321249955e-06,
      "loss": 0.6891,
      "step": 505
    },
    {
      "epoch": 0.3079710144927536,
      "grad_norm": 1.2224106907939065,
      "learning_rate": 8.741478034802835e-06,
      "loss": 0.6866,
      "step": 510
    },
    {
      "epoch": 0.3109903381642512,
      "grad_norm": 1.2220324268766434,
      "learning_rate": 8.706303927378306e-06,
      "loss": 0.6537,
      "step": 515
    },
    {
      "epoch": 0.3140096618357488,
      "grad_norm": 1.2102281315861285,
      "learning_rate": 8.670717908161878e-06,
      "loss": 0.6656,
      "step": 520
    },
    {
      "epoch": 0.3170289855072464,
      "grad_norm": 1.3602250350216378,
      "learning_rate": 8.634723932118184e-06,
      "loss": 0.6738,
      "step": 525
    },
    {
      "epoch": 0.32004830917874394,
      "grad_norm": 1.289145913095343,
      "learning_rate": 8.598325999551425e-06,
      "loss": 0.6717,
      "step": 530
    },
    {
      "epoch": 0.32306763285024154,
      "grad_norm": 1.1804864914238673,
      "learning_rate": 8.56152815566078e-06,
      "loss": 0.6579,
      "step": 535
    },
    {
      "epoch": 0.32608695652173914,
      "grad_norm": 1.2188640624977634,
      "learning_rate": 8.524334490090848e-06,
      "loss": 0.6802,
      "step": 540
    },
    {
      "epoch": 0.32910628019323673,
      "grad_norm": 1.277685319237274,
      "learning_rate": 8.486749136477112e-06,
      "loss": 0.6693,
      "step": 545
    },
    {
      "epoch": 0.3321256038647343,
      "grad_norm": 1.300215594029609,
      "learning_rate": 8.448776271986542e-06,
      "loss": 0.6681,
      "step": 550
    },
    {
      "epoch": 0.3351449275362319,
      "grad_norm": 1.1966744788724837,
      "learning_rate": 8.41042011685336e-06,
      "loss": 0.6698,
      "step": 555
    },
    {
      "epoch": 0.33816425120772947,
      "grad_norm": 1.354369397725386,
      "learning_rate": 8.371684933909996e-06,
      "loss": 0.6662,
      "step": 560
    },
    {
      "epoch": 0.34118357487922707,
      "grad_norm": 1.2209167981844395,
      "learning_rate": 8.33257502811334e-06,
      "loss": 0.6714,
      "step": 565
    },
    {
      "epoch": 0.3442028985507246,
      "grad_norm": 1.3116820491493628,
      "learning_rate": 8.293094746066283e-06,
      "loss": 0.6686,
      "step": 570
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 1.1920221462379554,
      "learning_rate": 8.253248475534656e-06,
      "loss": 0.666,
      "step": 575
    },
    {
      "epoch": 0.3502415458937198,
      "grad_norm": 1.2619018537421245,
      "learning_rate": 8.213040644959572e-06,
      "loss": 0.6376,
      "step": 580
    },
    {
      "epoch": 0.3532608695652174,
      "grad_norm": 1.3255977529395933,
      "learning_rate": 8.172475722965263e-06,
      "loss": 0.6485,
      "step": 585
    },
    {
      "epoch": 0.35628019323671495,
      "grad_norm": 1.1572256141420745,
      "learning_rate": 8.131558217862444e-06,
      "loss": 0.6265,
      "step": 590
    },
    {
      "epoch": 0.35929951690821255,
      "grad_norm": 1.2055573426722468,
      "learning_rate": 8.090292677147268e-06,
      "loss": 0.6488,
      "step": 595
    },
    {
      "epoch": 0.36231884057971014,
      "grad_norm": 1.2861724281724545,
      "learning_rate": 8.048683686995921e-06,
      "loss": 0.6577,
      "step": 600
    },
    {
      "epoch": 0.36533816425120774,
      "grad_norm": 1.1614784125218114,
      "learning_rate": 8.006735871754932e-06,
      "loss": 0.6639,
      "step": 605
    },
    {
      "epoch": 0.36835748792270534,
      "grad_norm": 1.2651442426402355,
      "learning_rate": 7.96445389342722e-06,
      "loss": 0.6429,
      "step": 610
    },
    {
      "epoch": 0.3713768115942029,
      "grad_norm": 1.1763415581175696,
      "learning_rate": 7.921842451153982e-06,
      "loss": 0.642,
      "step": 615
    },
    {
      "epoch": 0.3743961352657005,
      "grad_norm": 1.2194809852155362,
      "learning_rate": 7.878906280692424e-06,
      "loss": 0.658,
      "step": 620
    },
    {
      "epoch": 0.3774154589371981,
      "grad_norm": 1.234621233448178,
      "learning_rate": 7.835650153889449e-06,
      "loss": 0.6455,
      "step": 625
    },
    {
      "epoch": 0.3804347826086957,
      "grad_norm": 1.2755378740652312,
      "learning_rate": 7.792078878151318e-06,
      "loss": 0.6475,
      "step": 630
    },
    {
      "epoch": 0.3834541062801932,
      "grad_norm": 1.2958591755788766,
      "learning_rate": 7.748197295909359e-06,
      "loss": 0.6387,
      "step": 635
    },
    {
      "epoch": 0.3864734299516908,
      "grad_norm": 1.2667202622939782,
      "learning_rate": 7.704010284081801e-06,
      "loss": 0.6403,
      "step": 640
    },
    {
      "epoch": 0.3894927536231884,
      "grad_norm": 1.5195261481113882,
      "learning_rate": 7.65952275353175e-06,
      "loss": 0.6303,
      "step": 645
    },
    {
      "epoch": 0.392512077294686,
      "grad_norm": 1.2745676764475649,
      "learning_rate": 7.614739648521412e-06,
      "loss": 0.6427,
      "step": 650
    },
    {
      "epoch": 0.39553140096618356,
      "grad_norm": 1.249145405044073,
      "learning_rate": 7.56966594616259e-06,
      "loss": 0.6324,
      "step": 655
    },
    {
      "epoch": 0.39855072463768115,
      "grad_norm": 1.3163147586943797,
      "learning_rate": 7.524306655863544e-06,
      "loss": 0.6467,
      "step": 660
    },
    {
      "epoch": 0.40157004830917875,
      "grad_norm": 1.2102246889732293,
      "learning_rate": 7.478666818772252e-06,
      "loss": 0.637,
      "step": 665
    },
    {
      "epoch": 0.40458937198067635,
      "grad_norm": 1.270031191494333,
      "learning_rate": 7.432751507216146e-06,
      "loss": 0.6375,
      "step": 670
    },
    {
      "epoch": 0.4076086956521739,
      "grad_norm": 1.2566934217617964,
      "learning_rate": 7.386565824138378e-06,
      "loss": 0.6315,
      "step": 675
    },
    {
      "epoch": 0.4106280193236715,
      "grad_norm": 1.2617759238489816,
      "learning_rate": 7.3401149025306995e-06,
      "loss": 0.6243,
      "step": 680
    },
    {
      "epoch": 0.4136473429951691,
      "grad_norm": 1.2024909559860102,
      "learning_rate": 7.293403904862981e-06,
      "loss": 0.6369,
      "step": 685
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 1.3149707857289934,
      "learning_rate": 7.246438022509465e-06,
      "loss": 0.6272,
      "step": 690
    },
    {
      "epoch": 0.4196859903381642,
      "grad_norm": 1.3641187440427756,
      "learning_rate": 7.199222475171812e-06,
      "loss": 0.6286,
      "step": 695
    },
    {
      "epoch": 0.4227053140096618,
      "grad_norm": 1.2521464928430228,
      "learning_rate": 7.151762510298985e-06,
      "loss": 0.6292,
      "step": 700
    },
    {
      "epoch": 0.4257246376811594,
      "grad_norm": 1.4667183264262524,
      "learning_rate": 7.104063402504065e-06,
      "loss": 0.6159,
      "step": 705
    },
    {
      "epoch": 0.428743961352657,
      "grad_norm": 1.2306489207096507,
      "learning_rate": 7.056130452978039e-06,
      "loss": 0.6184,
      "step": 710
    },
    {
      "epoch": 0.43176328502415456,
      "grad_norm": 1.2724217204325428,
      "learning_rate": 7.0079689889006275e-06,
      "loss": 0.6166,
      "step": 715
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 1.363470409728175,
      "learning_rate": 6.959584362848239e-06,
      "loss": 0.6115,
      "step": 720
    },
    {
      "epoch": 0.43780193236714976,
      "grad_norm": 1.2907389058316032,
      "learning_rate": 6.910981952199097e-06,
      "loss": 0.6184,
      "step": 725
    },
    {
      "epoch": 0.44082125603864736,
      "grad_norm": 1.2785918958945928,
      "learning_rate": 6.862167158535599e-06,
      "loss": 0.6075,
      "step": 730
    },
    {
      "epoch": 0.4438405797101449,
      "grad_norm": 1.3363100906946352,
      "learning_rate": 6.813145407044003e-06,
      "loss": 0.6062,
      "step": 735
    },
    {
      "epoch": 0.4468599033816425,
      "grad_norm": 1.2872953040326178,
      "learning_rate": 6.763922145911474e-06,
      "loss": 0.6133,
      "step": 740
    },
    {
      "epoch": 0.4498792270531401,
      "grad_norm": 1.2195163567464784,
      "learning_rate": 6.714502845720595e-06,
      "loss": 0.6081,
      "step": 745
    },
    {
      "epoch": 0.4528985507246377,
      "grad_norm": 1.2421351910574,
      "learning_rate": 6.664892998841361e-06,
      "loss": 0.6114,
      "step": 750
    },
    {
      "epoch": 0.4559178743961353,
      "grad_norm": 1.199322165560994,
      "learning_rate": 6.61509811882078e-06,
      "loss": 0.5977,
      "step": 755
    },
    {
      "epoch": 0.45893719806763283,
      "grad_norm": 1.3225200571161657,
      "learning_rate": 6.565123739770102e-06,
      "loss": 0.6035,
      "step": 760
    },
    {
      "epoch": 0.46195652173913043,
      "grad_norm": 1.2787776232265113,
      "learning_rate": 6.5149754157497645e-06,
      "loss": 0.5973,
      "step": 765
    },
    {
      "epoch": 0.46497584541062803,
      "grad_norm": 1.254549030567658,
      "learning_rate": 6.464658720152135e-06,
      "loss": 0.5986,
      "step": 770
    },
    {
      "epoch": 0.46799516908212563,
      "grad_norm": 1.2314288543177414,
      "learning_rate": 6.41417924508208e-06,
      "loss": 0.5894,
      "step": 775
    },
    {
      "epoch": 0.47101449275362317,
      "grad_norm": 1.2493709608575305,
      "learning_rate": 6.363542600735486e-06,
      "loss": 0.602,
      "step": 780
    },
    {
      "epoch": 0.47403381642512077,
      "grad_norm": 1.2411573705254966,
      "learning_rate": 6.312754414775737e-06,
      "loss": 0.5984,
      "step": 785
    },
    {
      "epoch": 0.47705314009661837,
      "grad_norm": 1.5223815896896797,
      "learning_rate": 6.261820331708275e-06,
      "loss": 0.5917,
      "step": 790
    },
    {
      "epoch": 0.48007246376811596,
      "grad_norm": 1.3324419027195697,
      "learning_rate": 6.210746012253277e-06,
      "loss": 0.5894,
      "step": 795
    },
    {
      "epoch": 0.4830917874396135,
      "grad_norm": 1.44284253649467,
      "learning_rate": 6.159537132716532e-06,
      "loss": 0.608,
      "step": 800
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 1.2847962576743663,
      "learning_rate": 6.108199384358595e-06,
      "loss": 0.5699,
      "step": 805
    },
    {
      "epoch": 0.4891304347826087,
      "grad_norm": 1.2281985675073568,
      "learning_rate": 6.0567384727622566e-06,
      "loss": 0.5978,
      "step": 810
    },
    {
      "epoch": 0.4921497584541063,
      "grad_norm": 1.2429721815217631,
      "learning_rate": 6.005160117198448e-06,
      "loss": 0.5655,
      "step": 815
    },
    {
      "epoch": 0.49516908212560384,
      "grad_norm": 1.2553916316159304,
      "learning_rate": 5.953470049990605e-06,
      "loss": 0.5824,
      "step": 820
    },
    {
      "epoch": 0.49818840579710144,
      "grad_norm": 1.2732162189081664,
      "learning_rate": 5.90167401587759e-06,
      "loss": 0.582,
      "step": 825
    },
    {
      "epoch": 0.501207729468599,
      "grad_norm": 1.2868094593799089,
      "learning_rate": 5.84977777137523e-06,
      "loss": 0.5755,
      "step": 830
    },
    {
      "epoch": 0.5042270531400966,
      "grad_norm": 1.2418575696704444,
      "learning_rate": 5.797787084136556e-06,
      "loss": 0.5887,
      "step": 835
    },
    {
      "epoch": 0.5072463768115942,
      "grad_norm": 1.346918393906291,
      "learning_rate": 5.745707732310781e-06,
      "loss": 0.5955,
      "step": 840
    },
    {
      "epoch": 0.5102657004830918,
      "grad_norm": 1.34120555394378,
      "learning_rate": 5.693545503901149e-06,
      "loss": 0.5855,
      "step": 845
    },
    {
      "epoch": 0.5132850241545893,
      "grad_norm": 1.2970124804625551,
      "learning_rate": 5.641306196121643e-06,
      "loss": 0.5703,
      "step": 850
    },
    {
      "epoch": 0.5163043478260869,
      "grad_norm": 1.3089147536533883,
      "learning_rate": 5.5889956147527156e-06,
      "loss": 0.581,
      "step": 855
    },
    {
      "epoch": 0.5193236714975845,
      "grad_norm": 1.317588086682751,
      "learning_rate": 5.536619573496027e-06,
      "loss": 0.5827,
      "step": 860
    },
    {
      "epoch": 0.5223429951690821,
      "grad_norm": 1.3414854828076408,
      "learning_rate": 5.484183893328332e-06,
      "loss": 0.5832,
      "step": 865
    },
    {
      "epoch": 0.5253623188405797,
      "grad_norm": 1.3014560301267408,
      "learning_rate": 5.431694401854545e-06,
      "loss": 0.576,
      "step": 870
    },
    {
      "epoch": 0.5283816425120773,
      "grad_norm": 1.3095457716443242,
      "learning_rate": 5.379156932660067e-06,
      "loss": 0.575,
      "step": 875
    },
    {
      "epoch": 0.5314009661835749,
      "grad_norm": 1.2198501596042746,
      "learning_rate": 5.326577324662459e-06,
      "loss": 0.5731,
      "step": 880
    },
    {
      "epoch": 0.5344202898550725,
      "grad_norm": 1.6231967828184084,
      "learning_rate": 5.273961421462505e-06,
      "loss": 0.5641,
      "step": 885
    },
    {
      "epoch": 0.5374396135265701,
      "grad_norm": 1.308363933960681,
      "learning_rate": 5.221315070694775e-06,
      "loss": 0.5597,
      "step": 890
    },
    {
      "epoch": 0.5404589371980676,
      "grad_norm": 1.2349082122280453,
      "learning_rate": 5.168644123377725e-06,
      "loss": 0.5665,
      "step": 895
    },
    {
      "epoch": 0.5434782608695652,
      "grad_norm": 1.360538833482079,
      "learning_rate": 5.1159544332634256e-06,
      "loss": 0.5584,
      "step": 900
    },
    {
      "epoch": 0.5464975845410628,
      "grad_norm": 1.3836112606752502,
      "learning_rate": 5.063251856186991e-06,
      "loss": 0.5613,
      "step": 905
    },
    {
      "epoch": 0.5495169082125604,
      "grad_norm": 1.3886831951922487,
      "learning_rate": 5.010542249415761e-06,
      "loss": 0.5709,
      "step": 910
    },
    {
      "epoch": 0.552536231884058,
      "grad_norm": 1.4704931160983408,
      "learning_rate": 4.95783147099835e-06,
      "loss": 0.5594,
      "step": 915
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 1.2850129485673656,
      "learning_rate": 4.90512537911358e-06,
      "loss": 0.5483,
      "step": 920
    },
    {
      "epoch": 0.5585748792270532,
      "grad_norm": 1.3661715593612251,
      "learning_rate": 4.852429831419428e-06,
      "loss": 0.5619,
      "step": 925
    },
    {
      "epoch": 0.5615942028985508,
      "grad_norm": 1.3522573537703417,
      "learning_rate": 4.799750684402006e-06,
      "loss": 0.5415,
      "step": 930
    },
    {
      "epoch": 0.5646135265700483,
      "grad_norm": 1.2608087316548307,
      "learning_rate": 4.747093792724679e-06,
      "loss": 0.552,
      "step": 935
    },
    {
      "epoch": 0.5676328502415459,
      "grad_norm": 1.2839946310245753,
      "learning_rate": 4.6944650085774095e-06,
      "loss": 0.5632,
      "step": 940
    },
    {
      "epoch": 0.5706521739130435,
      "grad_norm": 1.3007085793096578,
      "learning_rate": 4.641870181026322e-06,
      "loss": 0.5625,
      "step": 945
    },
    {
      "epoch": 0.5736714975845411,
      "grad_norm": 1.3004895083183792,
      "learning_rate": 4.589315155363683e-06,
      "loss": 0.5507,
      "step": 950
    },
    {
      "epoch": 0.5766908212560387,
      "grad_norm": 1.325648774927767,
      "learning_rate": 4.53680577245824e-06,
      "loss": 0.5596,
      "step": 955
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 1.3082395762281465,
      "learning_rate": 4.484347868106097e-06,
      "loss": 0.5478,
      "step": 960
    },
    {
      "epoch": 0.5827294685990339,
      "grad_norm": 1.3393641276484884,
      "learning_rate": 4.431947272382118e-06,
      "loss": 0.5441,
      "step": 965
    },
    {
      "epoch": 0.5857487922705314,
      "grad_norm": 1.312719486472206,
      "learning_rate": 4.379609808992e-06,
      "loss": 0.5497,
      "step": 970
    },
    {
      "epoch": 0.5887681159420289,
      "grad_norm": 1.3204903978926055,
      "learning_rate": 4.327341294625019e-06,
      "loss": 0.5402,
      "step": 975
    },
    {
      "epoch": 0.5917874396135265,
      "grad_norm": 1.3284223155848323,
      "learning_rate": 4.275147538307594e-06,
      "loss": 0.5454,
      "step": 980
    },
    {
      "epoch": 0.5948067632850241,
      "grad_norm": 1.3534667187989002,
      "learning_rate": 4.223034340757666e-06,
      "loss": 0.5449,
      "step": 985
    },
    {
      "epoch": 0.5978260869565217,
      "grad_norm": 1.4006362214343153,
      "learning_rate": 4.171007493740023e-06,
      "loss": 0.5488,
      "step": 990
    },
    {
      "epoch": 0.6008454106280193,
      "grad_norm": 1.344509170501366,
      "learning_rate": 4.1190727794226175e-06,
      "loss": 0.5548,
      "step": 995
    },
    {
      "epoch": 0.6038647342995169,
      "grad_norm": 1.3164908246322924,
      "learning_rate": 4.067235969733937e-06,
      "loss": 0.5457,
      "step": 1000
    },
    {
      "epoch": 0.6068840579710145,
      "grad_norm": 1.303498686637293,
      "learning_rate": 4.015502825721537e-06,
      "loss": 0.5409,
      "step": 1005
    },
    {
      "epoch": 0.6099033816425121,
      "grad_norm": 1.2724683031487374,
      "learning_rate": 3.963879096911751e-06,
      "loss": 0.5312,
      "step": 1010
    },
    {
      "epoch": 0.6129227053140096,
      "grad_norm": 1.2726486865020912,
      "learning_rate": 3.91237052067072e-06,
      "loss": 0.5401,
      "step": 1015
    },
    {
      "epoch": 0.6159420289855072,
      "grad_norm": 1.343440037009517,
      "learning_rate": 3.860982821566729e-06,
      "loss": 0.5301,
      "step": 1020
    },
    {
      "epoch": 0.6189613526570048,
      "grad_norm": 1.353619163212925,
      "learning_rate": 3.8097217107340107e-06,
      "loss": 0.5365,
      "step": 1025
    },
    {
      "epoch": 0.6219806763285024,
      "grad_norm": 1.3477813018813276,
      "learning_rate": 3.7585928852380025e-06,
      "loss": 0.5393,
      "step": 1030
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.3731933102965312,
      "learning_rate": 3.7076020274421996e-06,
      "loss": 0.5464,
      "step": 1035
    },
    {
      "epoch": 0.6280193236714976,
      "grad_norm": 1.3390473233403632,
      "learning_rate": 3.6567548043766157e-06,
      "loss": 0.5444,
      "step": 1040
    },
    {
      "epoch": 0.6310386473429952,
      "grad_norm": 1.3046493242835504,
      "learning_rate": 3.6060568671079658e-06,
      "loss": 0.5353,
      "step": 1045
    },
    {
      "epoch": 0.6340579710144928,
      "grad_norm": 1.4419256695572908,
      "learning_rate": 3.5555138501116247e-06,
      "loss": 0.5244,
      "step": 1050
    },
    {
      "epoch": 0.6370772946859904,
      "grad_norm": 1.2976760979714546,
      "learning_rate": 3.5051313706453995e-06,
      "loss": 0.5404,
      "step": 1055
    },
    {
      "epoch": 0.6400966183574879,
      "grad_norm": 1.2919651906622562,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.5183,
      "step": 1060
    },
    {
      "epoch": 0.6431159420289855,
      "grad_norm": 1.3937729306554194,
      "learning_rate": 3.40487040350303e-06,
      "loss": 0.5279,
      "step": 1065
    },
    {
      "epoch": 0.6461352657004831,
      "grad_norm": 1.4754894489154289,
      "learning_rate": 3.355003058646105e-06,
      "loss": 0.5262,
      "step": 1070
    },
    {
      "epoch": 0.6491545893719807,
      "grad_norm": 1.39657778630638,
      "learning_rate": 3.305318535719343e-06,
      "loss": 0.5375,
      "step": 1075
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 1.350395780922992,
      "learning_rate": 3.2558223565691104e-06,
      "loss": 0.5238,
      "step": 1080
    },
    {
      "epoch": 0.6551932367149759,
      "grad_norm": 1.3685872873162936,
      "learning_rate": 3.2065200221095905e-06,
      "loss": 0.5298,
      "step": 1085
    },
    {
      "epoch": 0.6582125603864735,
      "grad_norm": 1.3233219201374211,
      "learning_rate": 3.1574170117114293e-06,
      "loss": 0.5184,
      "step": 1090
    },
    {
      "epoch": 0.6612318840579711,
      "grad_norm": 1.3288039422625464,
      "learning_rate": 3.1085187825927555e-06,
      "loss": 0.5278,
      "step": 1095
    },
    {
      "epoch": 0.6642512077294686,
      "grad_norm": 1.3601324138928483,
      "learning_rate": 3.0598307692126904e-06,
      "loss": 0.5216,
      "step": 1100
    },
    {
      "epoch": 0.6672705314009661,
      "grad_norm": 1.2974323886442631,
      "learning_rate": 3.0113583826673655e-06,
      "loss": 0.5281,
      "step": 1105
    },
    {
      "epoch": 0.6702898550724637,
      "grad_norm": 1.3817579947889416,
      "learning_rate": 2.9631070100885373e-06,
      "loss": 0.5344,
      "step": 1110
    },
    {
      "epoch": 0.6733091787439613,
      "grad_norm": 1.373246896186375,
      "learning_rate": 2.915082014044883e-06,
      "loss": 0.5118,
      "step": 1115
    },
    {
      "epoch": 0.6763285024154589,
      "grad_norm": 1.3914109961429577,
      "learning_rate": 2.867288731946004e-06,
      "loss": 0.5219,
      "step": 1120
    },
    {
      "epoch": 0.6793478260869565,
      "grad_norm": 1.3456129482520864,
      "learning_rate": 2.8197324754492456e-06,
      "loss": 0.5123,
      "step": 1125
    },
    {
      "epoch": 0.6823671497584541,
      "grad_norm": 1.3516053089371904,
      "learning_rate": 2.7724185298693596e-06,
      "loss": 0.5211,
      "step": 1130
    },
    {
      "epoch": 0.6853864734299517,
      "grad_norm": 1.6512789595808264,
      "learning_rate": 2.7253521535911144e-06,
      "loss": 0.5069,
      "step": 1135
    },
    {
      "epoch": 0.6884057971014492,
      "grad_norm": 1.3963436494331172,
      "learning_rate": 2.678538577484871e-06,
      "loss": 0.5163,
      "step": 1140
    },
    {
      "epoch": 0.6914251207729468,
      "grad_norm": 1.327652930753795,
      "learning_rate": 2.6319830043252616e-06,
      "loss": 0.5062,
      "step": 1145
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 1.3883407663503802,
      "learning_rate": 2.5856906082129313e-06,
      "loss": 0.5099,
      "step": 1150
    },
    {
      "epoch": 0.697463768115942,
      "grad_norm": 1.3541895548105385,
      "learning_rate": 2.53966653399952e-06,
      "loss": 0.5151,
      "step": 1155
    },
    {
      "epoch": 0.7004830917874396,
      "grad_norm": 1.6671674084618924,
      "learning_rate": 2.4939158967158657e-06,
      "loss": 0.5234,
      "step": 1160
    },
    {
      "epoch": 0.7035024154589372,
      "grad_norm": 1.4194726508864435,
      "learning_rate": 2.448443781003527e-06,
      "loss": 0.5009,
      "step": 1165
    },
    {
      "epoch": 0.7065217391304348,
      "grad_norm": 1.4328105129938522,
      "learning_rate": 2.403255240549693e-06,
      "loss": 0.5055,
      "step": 1170
    },
    {
      "epoch": 0.7095410628019324,
      "grad_norm": 1.4936167581096147,
      "learning_rate": 2.3583552975255108e-06,
      "loss": 0.5163,
      "step": 1175
    },
    {
      "epoch": 0.7125603864734299,
      "grad_norm": 1.366182685953503,
      "learning_rate": 2.313748942027956e-06,
      "loss": 0.5014,
      "step": 1180
    },
    {
      "epoch": 0.7155797101449275,
      "grad_norm": 1.4219678396595035,
      "learning_rate": 2.269441131525213e-06,
      "loss": 0.5043,
      "step": 1185
    },
    {
      "epoch": 0.7185990338164251,
      "grad_norm": 1.3494187919079106,
      "learning_rate": 2.225436790305733e-06,
      "loss": 0.5017,
      "step": 1190
    },
    {
      "epoch": 0.7216183574879227,
      "grad_norm": 1.3118718901779507,
      "learning_rate": 2.181740808930947e-06,
      "loss": 0.4854,
      "step": 1195
    },
    {
      "epoch": 0.7246376811594203,
      "grad_norm": 1.410018165296287,
      "learning_rate": 2.1383580436917452e-06,
      "loss": 0.4977,
      "step": 1200
    },
    {
      "epoch": 0.7276570048309179,
      "grad_norm": 1.4006801036455434,
      "learning_rate": 2.0952933160687456e-06,
      "loss": 0.4985,
      "step": 1205
    },
    {
      "epoch": 0.7306763285024155,
      "grad_norm": 1.4335252946653403,
      "learning_rate": 2.052551412196456e-06,
      "loss": 0.4939,
      "step": 1210
    },
    {
      "epoch": 0.7336956521739131,
      "grad_norm": 1.3979757625658358,
      "learning_rate": 2.010137082331354e-06,
      "loss": 0.5048,
      "step": 1215
    },
    {
      "epoch": 0.7367149758454107,
      "grad_norm": 1.2854697344016206,
      "learning_rate": 1.96805504032393e-06,
      "loss": 0.5047,
      "step": 1220
    },
    {
      "epoch": 0.7397342995169082,
      "grad_norm": 1.3683323463695072,
      "learning_rate": 1.9263099630948274e-06,
      "loss": 0.5042,
      "step": 1225
    },
    {
      "epoch": 0.7427536231884058,
      "grad_norm": 1.398514848998387,
      "learning_rate": 1.8849064901150372e-06,
      "loss": 0.473,
      "step": 1230
    },
    {
      "epoch": 0.7457729468599034,
      "grad_norm": 1.4022273688712774,
      "learning_rate": 1.8438492228902893e-06,
      "loss": 0.4995,
      "step": 1235
    },
    {
      "epoch": 0.748792270531401,
      "grad_norm": 1.293867549241767,
      "learning_rate": 1.8031427244496357e-06,
      "loss": 0.5009,
      "step": 1240
    },
    {
      "epoch": 0.7518115942028986,
      "grad_norm": 1.2978900181373894,
      "learning_rate": 1.7627915188383382e-06,
      "loss": 0.4901,
      "step": 1245
    },
    {
      "epoch": 0.7548309178743962,
      "grad_norm": 1.4448114494500728,
      "learning_rate": 1.7228000906150672e-06,
      "loss": 0.5034,
      "step": 1250
    },
    {
      "epoch": 0.7578502415458938,
      "grad_norm": 1.3817066267485774,
      "learning_rate": 1.6831728843534962e-06,
      "loss": 0.4928,
      "step": 1255
    },
    {
      "epoch": 0.7608695652173914,
      "grad_norm": 1.5474646288550162,
      "learning_rate": 1.6439143041483352e-06,
      "loss": 0.4753,
      "step": 1260
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 1.3836879466510739,
      "learning_rate": 1.6050287131258862e-06,
      "loss": 0.4929,
      "step": 1265
    },
    {
      "epoch": 0.7669082125603864,
      "grad_norm": 1.352247909907531,
      "learning_rate": 1.5665204329591066e-06,
      "loss": 0.4915,
      "step": 1270
    },
    {
      "epoch": 0.769927536231884,
      "grad_norm": 1.3795806073347745,
      "learning_rate": 1.528393743387328e-06,
      "loss": 0.486,
      "step": 1275
    },
    {
      "epoch": 0.7729468599033816,
      "grad_norm": 1.3399181762266046,
      "learning_rate": 1.4906528817406052e-06,
      "loss": 0.4765,
      "step": 1280
    },
    {
      "epoch": 0.7759661835748792,
      "grad_norm": 1.432793399552329,
      "learning_rate": 1.453302042468786e-06,
      "loss": 0.4793,
      "step": 1285
    },
    {
      "epoch": 0.7789855072463768,
      "grad_norm": 1.440038491030672,
      "learning_rate": 1.4163453766753537e-06,
      "loss": 0.489,
      "step": 1290
    },
    {
      "epoch": 0.7820048309178744,
      "grad_norm": 1.3868984310668784,
      "learning_rate": 1.3797869916560692e-06,
      "loss": 0.4933,
      "step": 1295
    },
    {
      "epoch": 0.785024154589372,
      "grad_norm": 1.382680178173945,
      "learning_rate": 1.3436309504425137e-06,
      "loss": 0.4816,
      "step": 1300
    },
    {
      "epoch": 0.7880434782608695,
      "grad_norm": 1.3699907284486668,
      "learning_rate": 1.3078812713505079e-06,
      "loss": 0.4836,
      "step": 1305
    },
    {
      "epoch": 0.7910628019323671,
      "grad_norm": 1.3748734244080174,
      "learning_rate": 1.2725419275335404e-06,
      "loss": 0.4787,
      "step": 1310
    },
    {
      "epoch": 0.7940821256038647,
      "grad_norm": 1.4416418877555286,
      "learning_rate": 1.237616846541192e-06,
      "loss": 0.4861,
      "step": 1315
    },
    {
      "epoch": 0.7971014492753623,
      "grad_norm": 1.4582733607311282,
      "learning_rate": 1.2031099098826376e-06,
      "loss": 0.4902,
      "step": 1320
    },
    {
      "epoch": 0.8001207729468599,
      "grad_norm": 1.3951099319990903,
      "learning_rate": 1.1690249525952569e-06,
      "loss": 0.4698,
      "step": 1325
    },
    {
      "epoch": 0.8031400966183575,
      "grad_norm": 1.3211083162574102,
      "learning_rate": 1.1353657628184217e-06,
      "loss": 0.4654,
      "step": 1330
    },
    {
      "epoch": 0.8061594202898551,
      "grad_norm": 1.39861267444744,
      "learning_rate": 1.1021360813724924e-06,
      "loss": 0.4814,
      "step": 1335
    },
    {
      "epoch": 0.8091787439613527,
      "grad_norm": 1.3855723067060435,
      "learning_rate": 1.0693396013430552e-06,
      "loss": 0.4744,
      "step": 1340
    },
    {
      "epoch": 0.8121980676328503,
      "grad_norm": 1.4445315702507702,
      "learning_rate": 1.036979967670494e-06,
      "loss": 0.4724,
      "step": 1345
    },
    {
      "epoch": 0.8152173913043478,
      "grad_norm": 1.4005662044580711,
      "learning_rate": 1.0050607767448928e-06,
      "loss": 0.4692,
      "step": 1350
    },
    {
      "epoch": 0.8182367149758454,
      "grad_norm": 1.3423573516670364,
      "learning_rate": 9.735855760063412e-07,
      "loss": 0.4769,
      "step": 1355
    },
    {
      "epoch": 0.821256038647343,
      "grad_norm": 1.3388950541234161,
      "learning_rate": 9.425578635506721e-07,
      "loss": 0.4801,
      "step": 1360
    },
    {
      "epoch": 0.8242753623188406,
      "grad_norm": 1.3520391195043011,
      "learning_rate": 9.119810877406998e-07,
      "loss": 0.4812,
      "step": 1365
    },
    {
      "epoch": 0.8272946859903382,
      "grad_norm": 1.364798918561803,
      "learning_rate": 8.818586468229695e-07,
      "loss": 0.4694,
      "step": 1370
    },
    {
      "epoch": 0.8303140096618358,
      "grad_norm": 1.3718140961001366,
      "learning_rate": 8.521938885500825e-07,
      "loss": 0.4695,
      "step": 1375
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 1.3996474606181322,
      "learning_rate": 8.229901098086335e-07,
      "loss": 0.4873,
      "step": 1380
    },
    {
      "epoch": 0.836352657004831,
      "grad_norm": 1.3941499294551503,
      "learning_rate": 7.942505562528024e-07,
      "loss": 0.4812,
      "step": 1385
    },
    {
      "epoch": 0.8393719806763285,
      "grad_norm": 1.4317232059058598,
      "learning_rate": 7.659784219436373e-07,
      "loss": 0.4775,
      "step": 1390
    },
    {
      "epoch": 0.842391304347826,
      "grad_norm": 1.3368338252764984,
      "learning_rate": 7.381768489940678e-07,
      "loss": 0.4834,
      "step": 1395
    },
    {
      "epoch": 0.8454106280193237,
      "grad_norm": 1.376767316724929,
      "learning_rate": 7.108489272197089e-07,
      "loss": 0.4777,
      "step": 1400
    },
    {
      "epoch": 0.8484299516908212,
      "grad_norm": 1.4416146692158918,
      "learning_rate": 6.839976937954479e-07,
      "loss": 0.4805,
      "step": 1405
    },
    {
      "epoch": 0.8514492753623188,
      "grad_norm": 1.3884346338383682,
      "learning_rate": 6.576261329179123e-07,
      "loss": 0.4659,
      "step": 1410
    },
    {
      "epoch": 0.8544685990338164,
      "grad_norm": 1.461661357035372,
      "learning_rate": 6.317371754738044e-07,
      "loss": 0.4835,
      "step": 1415
    },
    {
      "epoch": 0.857487922705314,
      "grad_norm": 1.3155537677854425,
      "learning_rate": 6.06333698714171e-07,
      "loss": 0.4612,
      "step": 1420
    },
    {
      "epoch": 0.8605072463768116,
      "grad_norm": 1.430549233517389,
      "learning_rate": 5.814185259346267e-07,
      "loss": 0.4652,
      "step": 1425
    },
    {
      "epoch": 0.8635265700483091,
      "grad_norm": 1.5438182616910798,
      "learning_rate": 5.56994426161584e-07,
      "loss": 0.4816,
      "step": 1430
    },
    {
      "epoch": 0.8665458937198067,
      "grad_norm": 1.3944382700425997,
      "learning_rate": 5.330641138445064e-07,
      "loss": 0.4668,
      "step": 1435
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 1.3524996972490821,
      "learning_rate": 5.096302485542265e-07,
      "loss": 0.4626,
      "step": 1440
    },
    {
      "epoch": 0.8725845410628019,
      "grad_norm": 1.3677343406038267,
      "learning_rate": 4.866954346873715e-07,
      "loss": 0.4658,
      "step": 1445
    },
    {
      "epoch": 0.8756038647342995,
      "grad_norm": 1.3464641098418646,
      "learning_rate": 4.642622211769099e-07,
      "loss": 0.4685,
      "step": 1450
    },
    {
      "epoch": 0.8786231884057971,
      "grad_norm": 1.390326156036712,
      "learning_rate": 4.4233310120887387e-07,
      "loss": 0.4591,
      "step": 1455
    },
    {
      "epoch": 0.8816425120772947,
      "grad_norm": 1.4052855328615665,
      "learning_rate": 4.209105119452628e-07,
      "loss": 0.4581,
      "step": 1460
    },
    {
      "epoch": 0.8846618357487923,
      "grad_norm": 1.4105848251200876,
      "learning_rate": 3.999968342531918e-07,
      "loss": 0.4707,
      "step": 1465
    },
    {
      "epoch": 0.8876811594202898,
      "grad_norm": 1.333713097837263,
      "learning_rate": 3.7959439244027727e-07,
      "loss": 0.4647,
      "step": 1470
    },
    {
      "epoch": 0.8907004830917874,
      "grad_norm": 1.356947580984905,
      "learning_rate": 3.5970545399632574e-07,
      "loss": 0.4572,
      "step": 1475
    },
    {
      "epoch": 0.893719806763285,
      "grad_norm": 1.3949650199499017,
      "learning_rate": 3.4033222934131914e-07,
      "loss": 0.466,
      "step": 1480
    },
    {
      "epoch": 0.8967391304347826,
      "grad_norm": 1.485033631723171,
      "learning_rate": 3.214768715797656e-07,
      "loss": 0.4735,
      "step": 1485
    },
    {
      "epoch": 0.8997584541062802,
      "grad_norm": 1.3806955768204767,
      "learning_rate": 3.0314147626139543e-07,
      "loss": 0.4671,
      "step": 1490
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 1.4448739661661896,
      "learning_rate": 2.853280811482734e-07,
      "loss": 0.4746,
      "step": 1495
    },
    {
      "epoch": 0.9057971014492754,
      "grad_norm": 1.4190366783020854,
      "learning_rate": 2.6803866598832216e-07,
      "loss": 0.4544,
      "step": 1500
    },
    {
      "epoch": 0.908816425120773,
      "grad_norm": 1.3988884827045063,
      "learning_rate": 2.5127515229529665e-07,
      "loss": 0.4659,
      "step": 1505
    },
    {
      "epoch": 0.9118357487922706,
      "grad_norm": 1.3539792184833292,
      "learning_rate": 2.350394031352343e-07,
      "loss": 0.4683,
      "step": 1510
    },
    {
      "epoch": 0.9148550724637681,
      "grad_norm": 1.3390472231534103,
      "learning_rate": 2.1933322291938897e-07,
      "loss": 0.4587,
      "step": 1515
    },
    {
      "epoch": 0.9178743961352657,
      "grad_norm": 1.2982303956915597,
      "learning_rate": 2.041583572037037e-07,
      "loss": 0.4705,
      "step": 1520
    },
    {
      "epoch": 0.9208937198067633,
      "grad_norm": 1.398922259288625,
      "learning_rate": 1.8951649249480287e-07,
      "loss": 0.4674,
      "step": 1525
    },
    {
      "epoch": 0.9239130434782609,
      "grad_norm": 1.3706043106328762,
      "learning_rate": 1.7540925606256088e-07,
      "loss": 0.4738,
      "step": 1530
    },
    {
      "epoch": 0.9269323671497585,
      "grad_norm": 1.4051091915658231,
      "learning_rate": 1.6183821575925186e-07,
      "loss": 0.4678,
      "step": 1535
    },
    {
      "epoch": 0.9299516908212561,
      "grad_norm": 1.3953334052871644,
      "learning_rate": 1.4880487984529846e-07,
      "loss": 0.4789,
      "step": 1540
    },
    {
      "epoch": 0.9329710144927537,
      "grad_norm": 1.4473194738633308,
      "learning_rate": 1.363106968216482e-07,
      "loss": 0.474,
      "step": 1545
    },
    {
      "epoch": 0.9359903381642513,
      "grad_norm": 1.3696328617667175,
      "learning_rate": 1.243570552687895e-07,
      "loss": 0.4549,
      "step": 1550
    },
    {
      "epoch": 0.9390096618357487,
      "grad_norm": 1.4005643653184559,
      "learning_rate": 1.1294528369242663e-07,
      "loss": 0.4545,
      "step": 1555
    },
    {
      "epoch": 0.9420289855072463,
      "grad_norm": 1.3929824476134263,
      "learning_rate": 1.020766503758347e-07,
      "loss": 0.4793,
      "step": 1560
    },
    {
      "epoch": 0.9450483091787439,
      "grad_norm": 1.3824037261570252,
      "learning_rate": 9.175236323890058e-08,
      "loss": 0.4773,
      "step": 1565
    },
    {
      "epoch": 0.9480676328502415,
      "grad_norm": 1.4006805147153165,
      "learning_rate": 8.197356970388148e-08,
      "loss": 0.4672,
      "step": 1570
    },
    {
      "epoch": 0.9510869565217391,
      "grad_norm": 1.3886975460206281,
      "learning_rate": 7.274135656787917e-08,
      "loss": 0.4526,
      "step": 1575
    },
    {
      "epoch": 0.9541062801932367,
      "grad_norm": 1.4838070116928936,
      "learning_rate": 6.405674988205602e-08,
      "loss": 0.4572,
      "step": 1580
    },
    {
      "epoch": 0.9571256038647343,
      "grad_norm": 1.4235233786161414,
      "learning_rate": 5.592071483760397e-08,
      "loss": 0.4599,
      "step": 1585
    },
    {
      "epoch": 0.9601449275362319,
      "grad_norm": 1.402808009230018,
      "learning_rate": 4.833415565847155e-08,
      "loss": 0.4634,
      "step": 1590
    },
    {
      "epoch": 0.9631642512077294,
      "grad_norm": 1.397718470953625,
      "learning_rate": 4.1297915500873034e-08,
      "loss": 0.4605,
      "step": 1595
    },
    {
      "epoch": 0.966183574879227,
      "grad_norm": 1.3622090829571152,
      "learning_rate": 3.481277635957903e-08,
      "loss": 0.4644,
      "step": 1600
    },
    {
      "epoch": 0.9692028985507246,
      "grad_norm": 1.4192128536229283,
      "learning_rate": 2.88794589810093e-08,
      "loss": 0.4513,
      "step": 1605
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 1.4607543198292594,
      "learning_rate": 2.3498622783128533e-08,
      "loss": 0.4643,
      "step": 1610
    },
    {
      "epoch": 0.9752415458937198,
      "grad_norm": 1.4519572105646263,
      "learning_rate": 1.8670865782161042e-08,
      "loss": 0.4737,
      "step": 1615
    },
    {
      "epoch": 0.9782608695652174,
      "grad_norm": 1.5587138432284067,
      "learning_rate": 1.4396724526127282e-08,
      "loss": 0.4719,
      "step": 1620
    },
    {
      "epoch": 0.981280193236715,
      "grad_norm": 1.3504712276503201,
      "learning_rate": 1.067667403521433e-08,
      "loss": 0.4597,
      "step": 1625
    },
    {
      "epoch": 0.9842995169082126,
      "grad_norm": 1.3482863134408452,
      "learning_rate": 7.51112774898144e-09,
      "loss": 0.4678,
      "step": 1630
    },
    {
      "epoch": 0.9873188405797102,
      "grad_norm": 1.3970409533706654,
      "learning_rate": 4.900437480413467e-09,
      "loss": 0.4718,
      "step": 1635
    },
    {
      "epoch": 0.9903381642512077,
      "grad_norm": 1.4616532167767295,
      "learning_rate": 2.844893376816593e-09,
      "loss": 0.4653,
      "step": 1640
    },
    {
      "epoch": 0.9933574879227053,
      "grad_norm": 1.3667351204068807,
      "learning_rate": 1.3447238875774482e-09,
      "loss": 0.468,
      "step": 1645
    },
    {
      "epoch": 0.9963768115942029,
      "grad_norm": 1.3873085164877705,
      "learning_rate": 4.000957387700899e-10,
      "loss": 0.4735,
      "step": 1650
    },
    {
      "epoch": 0.9993961352657005,
      "grad_norm": 1.369426340332071,
      "learning_rate": 1.1113914626381672e-11,
      "loss": 0.4693,
      "step": 1655
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.6076,
      "eval_samples_per_second": 2.772,
      "eval_steps_per_second": 0.832,
      "step": 1656
    },
    {
      "epoch": 1.0,
      "step": 1656,
      "total_flos": 173366354903040.0,
      "train_loss": 0.6223726071021407,
      "train_runtime": 17321.1955,
      "train_samples_per_second": 1.529,
      "train_steps_per_second": 0.096
    }
  ],
  "logging_steps": 5,
  "max_steps": 1656,
  "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": 173366354903040.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}