{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "eval_steps": 500,
  "global_step": 3075,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04878048780487805,
      "grad_norm": 0.4036892056465149,
      "learning_rate": 0.00019999478113897612,
      "loss": 1.0282,
      "step": 10
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 0.3629762828350067,
      "learning_rate": 0.0001999791251006346,
      "loss": 0.7875,
      "step": 20
    },
    {
      "epoch": 0.14634146341463414,
      "grad_norm": 0.4877622425556183,
      "learning_rate": 0.0001999530335191093,
      "loss": 0.5942,
      "step": 30
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 0.4466260075569153,
      "learning_rate": 0.00019991650911776695,
      "loss": 0.3866,
      "step": 40
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.649118959903717,
      "learning_rate": 0.000199869555708923,
      "loss": 0.3928,
      "step": 50
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 0.8762800097465515,
      "learning_rate": 0.0001998121781934438,
      "loss": 0.3258,
      "step": 60
    },
    {
      "epoch": 0.34146341463414637,
      "grad_norm": 0.9195622801780701,
      "learning_rate": 0.0001997443825602349,
      "loss": 0.2885,
      "step": 70
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 0.5856262445449829,
      "learning_rate": 0.00019966617588561609,
      "loss": 0.2888,
      "step": 80
    },
    {
      "epoch": 0.43902439024390244,
      "grad_norm": 0.5520443320274353,
      "learning_rate": 0.00019957756633258265,
      "loss": 0.2242,
      "step": 90
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.9435800909996033,
      "learning_rate": 0.00019947856314995349,
      "loss": 0.1629,
      "step": 100
    },
    {
      "epoch": 0.5365853658536586,
      "grad_norm": 0.9416623115539551,
      "learning_rate": 0.00019936917667140555,
      "loss": 0.1555,
      "step": 110
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 0.802065372467041,
      "learning_rate": 0.0001992494183143955,
      "loss": 0.1339,
      "step": 120
    },
    {
      "epoch": 0.6341463414634146,
      "grad_norm": 0.7007794380187988,
      "learning_rate": 0.00019911930057896774,
      "loss": 0.1191,
      "step": 130
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 0.6755990386009216,
      "learning_rate": 0.00019897883704644983,
      "loss": 0.1571,
      "step": 140
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 1.6951078176498413,
      "learning_rate": 0.00019882804237803488,
      "loss": 0.1309,
      "step": 150
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 0.567158579826355,
      "learning_rate": 0.0001986669323132512,
      "loss": 0.0766,
      "step": 160
    },
    {
      "epoch": 0.8292682926829268,
      "grad_norm": 0.8820038437843323,
      "learning_rate": 0.0001984955236683196,
      "loss": 0.0839,
      "step": 170
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 0.6520794034004211,
      "learning_rate": 0.00019831383433439797,
      "loss": 0.0863,
      "step": 180
    },
    {
      "epoch": 0.926829268292683,
      "grad_norm": 0.45519864559173584,
      "learning_rate": 0.00019812188327571399,
      "loss": 0.0889,
      "step": 190
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.614235520362854,
      "learning_rate": 0.00019791969052758562,
      "loss": 0.0725,
      "step": 200
    },
    {
      "epoch": 1.024390243902439,
      "grad_norm": 0.2764686644077301,
      "learning_rate": 0.00019770727719432994,
      "loss": 0.0407,
      "step": 210
    },
    {
      "epoch": 1.0731707317073171,
      "grad_norm": 0.6082726716995239,
      "learning_rate": 0.00019748466544706022,
      "loss": 0.044,
      "step": 220
    },
    {
      "epoch": 1.1219512195121952,
      "grad_norm": 0.9295619130134583,
      "learning_rate": 0.00019725187852137195,
      "loss": 0.0675,
      "step": 230
    },
    {
      "epoch": 1.170731707317073,
      "grad_norm": 0.3758924603462219,
      "learning_rate": 0.00019700894071491732,
      "loss": 0.0439,
      "step": 240
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 0.46514585614204407,
      "learning_rate": 0.00019675587738486936,
      "loss": 0.0398,
      "step": 250
    },
    {
      "epoch": 1.2682926829268293,
      "grad_norm": 0.5870018005371094,
      "learning_rate": 0.0001964927149452751,
      "loss": 0.0406,
      "step": 260
    },
    {
      "epoch": 1.3170731707317074,
      "grad_norm": 0.30292996764183044,
      "learning_rate": 0.00019621948086429844,
      "loss": 0.028,
      "step": 270
    },
    {
      "epoch": 1.3658536585365852,
      "grad_norm": 0.47037121653556824,
      "learning_rate": 0.00019593620366135337,
      "loss": 0.0239,
      "step": 280
    },
    {
      "epoch": 1.4146341463414633,
      "grad_norm": 0.4176475405693054,
      "learning_rate": 0.00019564291290412688,
      "loss": 0.0281,
      "step": 290
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 0.3179157078266144,
      "learning_rate": 0.00019533963920549306,
      "loss": 0.0281,
      "step": 300
    },
    {
      "epoch": 1.5121951219512195,
      "grad_norm": 0.5817562937736511,
      "learning_rate": 0.00019502641422031763,
      "loss": 0.0296,
      "step": 310
    },
    {
      "epoch": 1.5609756097560976,
      "grad_norm": 0.7409655451774597,
      "learning_rate": 0.00019470327064215383,
      "loss": 0.029,
      "step": 320
    },
    {
      "epoch": 1.6097560975609757,
      "grad_norm": 0.4418310225009918,
      "learning_rate": 0.00019437024219983028,
      "loss": 0.0583,
      "step": 330
    },
    {
      "epoch": 1.6585365853658538,
      "grad_norm": 0.31637728214263916,
      "learning_rate": 0.0001940273636539301,
      "loss": 0.0354,
      "step": 340
    },
    {
      "epoch": 1.7073170731707317,
      "grad_norm": 0.22175493836402893,
      "learning_rate": 0.00019367467079316279,
      "loss": 0.0514,
      "step": 350
    },
    {
      "epoch": 1.7560975609756098,
      "grad_norm": 0.6636152267456055,
      "learning_rate": 0.00019331220043062894,
      "loss": 0.034,
      "step": 360
    },
    {
      "epoch": 1.8048780487804879,
      "grad_norm": 0.8424332141876221,
      "learning_rate": 0.00019293999039997746,
      "loss": 0.0299,
      "step": 370
    },
    {
      "epoch": 1.8536585365853657,
      "grad_norm": 0.6435155272483826,
      "learning_rate": 0.00019255807955145677,
      "loss": 0.0508,
      "step": 380
    },
    {
      "epoch": 1.9024390243902438,
      "grad_norm": 0.7734220027923584,
      "learning_rate": 0.00019216650774785972,
      "loss": 0.035,
      "step": 390
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 0.2854250967502594,
      "learning_rate": 0.0001917653158603628,
      "loss": 0.0339,
      "step": 400
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6165639758110046,
      "learning_rate": 0.0001913545457642601,
      "loss": 0.0323,
      "step": 410
    },
    {
      "epoch": 2.048780487804878,
      "grad_norm": 0.2167205959558487,
      "learning_rate": 0.00019093424033459248,
      "loss": 0.026,
      "step": 420
    },
    {
      "epoch": 2.097560975609756,
      "grad_norm": 0.2723434269428253,
      "learning_rate": 0.0001905044434416725,
      "loss": 0.0176,
      "step": 430
    },
    {
      "epoch": 2.1463414634146343,
      "grad_norm": 0.4085879325866699,
      "learning_rate": 0.00019006519994650513,
      "loss": 0.0138,
      "step": 440
    },
    {
      "epoch": 2.1951219512195124,
      "grad_norm": 0.17931397259235382,
      "learning_rate": 0.00018961655569610557,
      "loss": 0.0358,
      "step": 450
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 0.3886450231075287,
      "learning_rate": 0.00018915855751871363,
      "loss": 0.0209,
      "step": 460
    },
    {
      "epoch": 2.292682926829268,
      "grad_norm": 0.11022531986236572,
      "learning_rate": 0.0001886912532189061,
      "loss": 0.0101,
      "step": 470
    },
    {
      "epoch": 2.341463414634146,
      "grad_norm": 0.14626094698905945,
      "learning_rate": 0.00018821469157260685,
      "loss": 0.0156,
      "step": 480
    },
    {
      "epoch": 2.3902439024390243,
      "grad_norm": 0.18015721440315247,
      "learning_rate": 0.00018772892232199592,
      "loss": 0.0156,
      "step": 490
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 0.3262254595756531,
      "learning_rate": 0.00018723399617031751,
      "loss": 0.0295,
      "step": 500
    },
    {
      "epoch": 2.4878048780487805,
      "grad_norm": 0.10063759982585907,
      "learning_rate": 0.00018672996477658767,
      "loss": 0.0164,
      "step": 510
    },
    {
      "epoch": 2.5365853658536586,
      "grad_norm": 0.11382050812244415,
      "learning_rate": 0.00018621688075020227,
      "loss": 0.0207,
      "step": 520
    },
    {
      "epoch": 2.5853658536585367,
      "grad_norm": 0.12307539582252502,
      "learning_rate": 0.0001856947976454459,
      "loss": 0.0326,
      "step": 530
    },
    {
      "epoch": 2.6341463414634148,
      "grad_norm": 0.18902955949306488,
      "learning_rate": 0.00018516376995590187,
      "loss": 0.0144,
      "step": 540
    },
    {
      "epoch": 2.682926829268293,
      "grad_norm": 0.18087974190711975,
      "learning_rate": 0.00018462385310876443,
      "loss": 0.0111,
      "step": 550
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 0.2894444465637207,
      "learning_rate": 0.00018407510345905332,
      "loss": 0.0081,
      "step": 560
    },
    {
      "epoch": 2.7804878048780486,
      "grad_norm": 0.24273361265659332,
      "learning_rate": 0.0001835175782837318,
      "loss": 0.0301,
      "step": 570
    },
    {
      "epoch": 2.8292682926829267,
      "grad_norm": 0.069428451359272,
      "learning_rate": 0.00018295133577572799,
      "loss": 0.0234,
      "step": 580
    },
    {
      "epoch": 2.8780487804878048,
      "grad_norm": 0.1845165342092514,
      "learning_rate": 0.00018237643503786095,
      "loss": 0.0112,
      "step": 590
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 0.06373828649520874,
      "learning_rate": 0.00018179293607667178,
      "loss": 0.0241,
      "step": 600
    },
    {
      "epoch": 2.975609756097561,
      "grad_norm": 0.09466666728258133,
      "learning_rate": 0.0001812008997961602,
      "loss": 0.0151,
      "step": 610
    },
    {
      "epoch": 3.024390243902439,
      "grad_norm": 0.5068451762199402,
      "learning_rate": 0.00018060038799142759,
      "loss": 0.02,
      "step": 620
    },
    {
      "epoch": 3.073170731707317,
      "grad_norm": 0.17547158896923065,
      "learning_rate": 0.00017999146334222695,
      "loss": 0.0111,
      "step": 630
    },
    {
      "epoch": 3.1219512195121952,
      "grad_norm": 0.24108292162418365,
      "learning_rate": 0.00017937418940642074,
      "loss": 0.0095,
      "step": 640
    },
    {
      "epoch": 3.1707317073170733,
      "grad_norm": 0.24457822740077972,
      "learning_rate": 0.00017874863061334657,
      "loss": 0.0134,
      "step": 650
    },
    {
      "epoch": 3.2195121951219514,
      "grad_norm": 0.2185467779636383,
      "learning_rate": 0.00017811485225709256,
      "loss": 0.0135,
      "step": 660
    },
    {
      "epoch": 3.2682926829268295,
      "grad_norm": 0.12849357724189758,
      "learning_rate": 0.00017747292048968187,
      "loss": 0.0154,
      "step": 670
    },
    {
      "epoch": 3.317073170731707,
      "grad_norm": 0.09158976376056671,
      "learning_rate": 0.0001768229023141682,
      "loss": 0.0137,
      "step": 680
    },
    {
      "epoch": 3.3658536585365852,
      "grad_norm": 0.09520118683576584,
      "learning_rate": 0.00017616486557764187,
      "loss": 0.0147,
      "step": 690
    },
    {
      "epoch": 3.4146341463414633,
      "grad_norm": 0.11151307821273804,
      "learning_rate": 0.00017549887896414851,
      "loss": 0.0168,
      "step": 700
    },
    {
      "epoch": 3.4634146341463414,
      "grad_norm": 0.1175757572054863,
      "learning_rate": 0.00017482501198751965,
      "loss": 0.015,
      "step": 710
    },
    {
      "epoch": 3.5121951219512195,
      "grad_norm": 0.12314116209745407,
      "learning_rate": 0.00017414333498411733,
      "loss": 0.0179,
      "step": 720
    },
    {
      "epoch": 3.5609756097560976,
      "grad_norm": 0.08803991228342056,
      "learning_rate": 0.00017345391910549238,
      "loss": 0.0105,
      "step": 730
    },
    {
      "epoch": 3.6097560975609757,
      "grad_norm": 0.06382381916046143,
      "learning_rate": 0.000172756836310958,
      "loss": 0.0106,
      "step": 740
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 0.47523975372314453,
      "learning_rate": 0.0001720521593600787,
      "loss": 0.0085,
      "step": 750
    },
    {
      "epoch": 3.7073170731707314,
      "grad_norm": 0.1268441379070282,
      "learning_rate": 0.000171339961805076,
      "loss": 0.0155,
      "step": 760
    },
    {
      "epoch": 3.7560975609756095,
      "grad_norm": 0.23719309270381927,
      "learning_rate": 0.000170620317983151,
      "loss": 0.015,
      "step": 770
    },
    {
      "epoch": 3.8048780487804876,
      "grad_norm": 0.0664207935333252,
      "learning_rate": 0.00016989330300872576,
      "loss": 0.0179,
      "step": 780
    },
    {
      "epoch": 3.8536585365853657,
      "grad_norm": 0.3121427893638611,
      "learning_rate": 0.00016915899276560237,
      "loss": 0.0138,
      "step": 790
    },
    {
      "epoch": 3.902439024390244,
      "grad_norm": 0.0784049853682518,
      "learning_rate": 0.00016841746389904304,
      "loss": 0.0114,
      "step": 800
    },
    {
      "epoch": 3.951219512195122,
      "grad_norm": 0.7239044904708862,
      "learning_rate": 0.0001676687938077698,
      "loss": 0.0251,
      "step": 810
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2333860546350479,
      "learning_rate": 0.00016691306063588583,
      "loss": 0.0135,
      "step": 820
    },
    {
      "epoch": 4.048780487804878,
      "grad_norm": 0.7087911367416382,
      "learning_rate": 0.00016615034326471898,
      "loss": 0.0195,
      "step": 830
    },
    {
      "epoch": 4.097560975609756,
      "grad_norm": 0.07190815359354019,
      "learning_rate": 0.00016538072130458853,
      "loss": 0.0095,
      "step": 840
    },
    {
      "epoch": 4.146341463414634,
      "grad_norm": 0.33951500058174133,
      "learning_rate": 0.00016460427508649546,
      "loss": 0.0131,
      "step": 850
    },
    {
      "epoch": 4.195121951219512,
      "grad_norm": 0.06381627917289734,
      "learning_rate": 0.00016382108565373785,
      "loss": 0.0119,
      "step": 860
    },
    {
      "epoch": 4.2439024390243905,
      "grad_norm": 0.44717633724212646,
      "learning_rate": 0.00016303123475345182,
      "loss": 0.0127,
      "step": 870
    },
    {
      "epoch": 4.2926829268292686,
      "grad_norm": 0.6674973368644714,
      "learning_rate": 0.00016223480482807894,
      "loss": 0.0111,
      "step": 880
    },
    {
      "epoch": 4.341463414634147,
      "grad_norm": 0.3303108215332031,
      "learning_rate": 0.00016143187900676112,
      "loss": 0.0159,
      "step": 890
    },
    {
      "epoch": 4.390243902439025,
      "grad_norm": 0.2972947359085083,
      "learning_rate": 0.0001606225410966638,
      "loss": 0.0086,
      "step": 900
    },
    {
      "epoch": 4.439024390243903,
      "grad_norm": 0.059205561876297,
      "learning_rate": 0.00015980687557422854,
      "loss": 0.0104,
      "step": 910
    },
    {
      "epoch": 4.487804878048781,
      "grad_norm": 0.08515360206365585,
      "learning_rate": 0.00015898496757635536,
      "loss": 0.0079,
      "step": 920
    },
    {
      "epoch": 4.536585365853659,
      "grad_norm": 0.14600469172000885,
      "learning_rate": 0.00015815690289151658,
      "loss": 0.0101,
      "step": 930
    },
    {
      "epoch": 4.585365853658536,
      "grad_norm": 0.05901546776294708,
      "learning_rate": 0.0001573227679508024,
      "loss": 0.0075,
      "step": 940
    },
    {
      "epoch": 4.634146341463414,
      "grad_norm": 0.3404502868652344,
      "learning_rate": 0.00015648264981889934,
      "loss": 0.0158,
      "step": 950
    },
    {
      "epoch": 4.682926829268292,
      "grad_norm": 0.0968211218714714,
      "learning_rate": 0.00015563663618500302,
      "loss": 0.0092,
      "step": 960
    },
    {
      "epoch": 4.7317073170731705,
      "grad_norm": 0.08040373027324677,
      "learning_rate": 0.00015478481535366494,
      "loss": 0.0088,
      "step": 970
    },
    {
      "epoch": 4.780487804878049,
      "grad_norm": 0.7208348512649536,
      "learning_rate": 0.00015392727623557585,
      "loss": 0.0132,
      "step": 980
    },
    {
      "epoch": 4.829268292682927,
      "grad_norm": 0.26821044087409973,
      "learning_rate": 0.00015306410833828535,
      "loss": 0.0113,
      "step": 990
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 0.09240720421075821,
      "learning_rate": 0.00015219540175685938,
      "loss": 0.0096,
      "step": 1000
    },
    {
      "epoch": 4.926829268292683,
      "grad_norm": 0.3397311270236969,
      "learning_rate": 0.00015132124716447627,
      "loss": 0.0072,
      "step": 1010
    },
    {
      "epoch": 4.975609756097561,
      "grad_norm": 0.07910922169685364,
      "learning_rate": 0.00015044173580296266,
      "loss": 0.0115,
      "step": 1020
    },
    {
      "epoch": 5.024390243902439,
      "grad_norm": 0.05629584938287735,
      "learning_rate": 0.00014955695947326987,
      "loss": 0.0066,
      "step": 1030
    },
    {
      "epoch": 5.073170731707317,
      "grad_norm": 0.7522541880607605,
      "learning_rate": 0.00014866701052589193,
      "loss": 0.0159,
      "step": 1040
    },
    {
      "epoch": 5.121951219512195,
      "grad_norm": 0.09192178398370743,
      "learning_rate": 0.0001477719818512263,
      "loss": 0.0085,
      "step": 1050
    },
    {
      "epoch": 5.170731707317073,
      "grad_norm": 0.3984428346157074,
      "learning_rate": 0.00014687196686987832,
      "loss": 0.0091,
      "step": 1060
    },
    {
      "epoch": 5.219512195121951,
      "grad_norm": 0.4188997149467468,
      "learning_rate": 0.00014596705952291017,
      "loss": 0.0083,
      "step": 1070
    },
    {
      "epoch": 5.2682926829268295,
      "grad_norm": 0.07686212658882141,
      "learning_rate": 0.00014505735426203543,
      "loss": 0.0064,
      "step": 1080
    },
    {
      "epoch": 5.317073170731708,
      "grad_norm": 0.38933801651000977,
      "learning_rate": 0.00014414294603976076,
      "loss": 0.0086,
      "step": 1090
    },
    {
      "epoch": 5.365853658536586,
      "grad_norm": 0.0843789279460907,
      "learning_rate": 0.00014322393029947468,
      "loss": 0.0074,
      "step": 1100
    },
    {
      "epoch": 5.414634146341464,
      "grad_norm": 0.19899946451187134,
      "learning_rate": 0.0001423004029654859,
      "loss": 0.0102,
      "step": 1110
    },
    {
      "epoch": 5.463414634146342,
      "grad_norm": 0.06790963560342789,
      "learning_rate": 0.00014137246043301042,
      "loss": 0.0102,
      "step": 1120
    },
    {
      "epoch": 5.512195121951219,
      "grad_norm": 0.08267515152692795,
      "learning_rate": 0.00014044019955811065,
      "loss": 0.009,
      "step": 1130
    },
    {
      "epoch": 5.560975609756097,
      "grad_norm": 0.3616844713687897,
      "learning_rate": 0.00013950371764758542,
      "loss": 0.0119,
      "step": 1140
    },
    {
      "epoch": 5.609756097560975,
      "grad_norm": 0.052261725068092346,
      "learning_rate": 0.0001385631124488136,
      "loss": 0.0078,
      "step": 1150
    },
    {
      "epoch": 5.658536585365853,
      "grad_norm": 0.07086987048387527,
      "learning_rate": 0.00013761848213955134,
      "loss": 0.0112,
      "step": 1160
    },
    {
      "epoch": 5.7073170731707314,
      "grad_norm": 0.05841855704784393,
      "learning_rate": 0.00013666992531768482,
      "loss": 0.0095,
      "step": 1170
    },
    {
      "epoch": 5.7560975609756095,
      "grad_norm": 0.2936477065086365,
      "learning_rate": 0.00013571754099093848,
      "loss": 0.0082,
      "step": 1180
    },
    {
      "epoch": 5.804878048780488,
      "grad_norm": 0.7195134162902832,
      "learning_rate": 0.0001347614285665412,
      "loss": 0.0093,
      "step": 1190
    },
    {
      "epoch": 5.853658536585366,
      "grad_norm": 0.09665755927562714,
      "learning_rate": 0.00013380168784085027,
      "loss": 0.0125,
      "step": 1200
    },
    {
      "epoch": 5.902439024390244,
      "grad_norm": 0.08146821707487106,
      "learning_rate": 0.0001328384189889351,
      "loss": 0.0094,
      "step": 1210
    },
    {
      "epoch": 5.951219512195122,
      "grad_norm": 1.6456434726715088,
      "learning_rate": 0.00013187172255412096,
      "loss": 0.0104,
      "step": 1220
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.46437689661979675,
      "learning_rate": 0.00013090169943749476,
      "loss": 0.0103,
      "step": 1230
    },
    {
      "epoch": 6.048780487804878,
      "grad_norm": 0.40744251012802124,
      "learning_rate": 0.00012992845088737322,
      "loss": 0.0105,
      "step": 1240
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 0.09202691912651062,
      "learning_rate": 0.00012895207848873487,
      "loss": 0.0088,
      "step": 1250
    },
    {
      "epoch": 6.146341463414634,
      "grad_norm": 0.41562455892562866,
      "learning_rate": 0.00012797268415261682,
      "loss": 0.0105,
      "step": 1260
    },
    {
      "epoch": 6.195121951219512,
      "grad_norm": 0.061398524791002274,
      "learning_rate": 0.0001269903701054776,
      "loss": 0.0084,
      "step": 1270
    },
    {
      "epoch": 6.2439024390243905,
      "grad_norm": 0.0851488932967186,
      "learning_rate": 0.00012600523887852706,
      "loss": 0.0081,
      "step": 1280
    },
    {
      "epoch": 6.2926829268292686,
      "grad_norm": 0.4565141201019287,
      "learning_rate": 0.00012501739329702453,
      "loss": 0.0122,
      "step": 1290
    },
    {
      "epoch": 6.341463414634147,
      "grad_norm": 0.06614430993795395,
      "learning_rate": 0.00012402693646954607,
      "loss": 0.0079,
      "step": 1300
    },
    {
      "epoch": 6.390243902439025,
      "grad_norm": 0.3719758987426758,
      "learning_rate": 0.00012303397177722234,
      "loss": 0.0096,
      "step": 1310
    },
    {
      "epoch": 6.439024390243903,
      "grad_norm": 0.06896215677261353,
      "learning_rate": 0.0001220386028629479,
      "loss": 0.0059,
      "step": 1320
    },
    {
      "epoch": 6.487804878048781,
      "grad_norm": 0.05981654301285744,
      "learning_rate": 0.00012104093362056341,
      "loss": 0.007,
      "step": 1330
    },
    {
      "epoch": 6.536585365853659,
      "grad_norm": 0.0625968649983406,
      "learning_rate": 0.00012004106818401135,
      "loss": 0.0104,
      "step": 1340
    },
    {
      "epoch": 6.585365853658536,
      "grad_norm": 0.04070024937391281,
      "learning_rate": 0.00011903911091646684,
      "loss": 0.0082,
      "step": 1350
    },
    {
      "epoch": 6.634146341463414,
      "grad_norm": 0.05095053091645241,
      "learning_rate": 0.00011803516639944452,
      "loss": 0.008,
      "step": 1360
    },
    {
      "epoch": 6.682926829268292,
      "grad_norm": 0.058545567095279694,
      "learning_rate": 0.00011702933942188252,
      "loss": 0.0079,
      "step": 1370
    },
    {
      "epoch": 6.7317073170731705,
      "grad_norm": 0.06065903976559639,
      "learning_rate": 0.0001160217349692051,
      "loss": 0.0075,
      "step": 1380
    },
    {
      "epoch": 6.780487804878049,
      "grad_norm": 0.13544753193855286,
      "learning_rate": 0.00011501245821236439,
      "loss": 0.0075,
      "step": 1390
    },
    {
      "epoch": 6.829268292682927,
      "grad_norm": 0.06476018577814102,
      "learning_rate": 0.00011400161449686293,
      "loss": 0.0086,
      "step": 1400
    },
    {
      "epoch": 6.878048780487805,
      "grad_norm": 0.0454210601747036,
      "learning_rate": 0.00011298930933175804,
      "loss": 0.0074,
      "step": 1410
    },
    {
      "epoch": 6.926829268292683,
      "grad_norm": 0.05906907469034195,
      "learning_rate": 0.00011197564837864922,
      "loss": 0.007,
      "step": 1420
    },
    {
      "epoch": 6.975609756097561,
      "grad_norm": 0.2888021171092987,
      "learning_rate": 0.00011096073744064919,
      "loss": 0.0076,
      "step": 1430
    },
    {
      "epoch": 7.024390243902439,
      "grad_norm": 0.05438758432865143,
      "learning_rate": 0.00010994468245134071,
      "loss": 0.008,
      "step": 1440
    },
    {
      "epoch": 7.073170731707317,
      "grad_norm": 0.04347096383571625,
      "learning_rate": 0.00010892758946371944,
      "loss": 0.006,
      "step": 1450
    },
    {
      "epoch": 7.121951219512195,
      "grad_norm": 0.06834473460912704,
      "learning_rate": 0.00010790956463912438,
      "loss": 0.0079,
      "step": 1460
    },
    {
      "epoch": 7.170731707317073,
      "grad_norm": 0.04391489177942276,
      "learning_rate": 0.00010689071423615709,
      "loss": 0.0061,
      "step": 1470
    },
    {
      "epoch": 7.219512195121951,
      "grad_norm": 0.0990622267127037,
      "learning_rate": 0.00010587114459959071,
      "loss": 0.0071,
      "step": 1480
    },
    {
      "epoch": 7.2682926829268295,
      "grad_norm": 0.048575662076473236,
      "learning_rate": 0.00010485096214927002,
      "loss": 0.0066,
      "step": 1490
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 0.05991552770137787,
      "learning_rate": 0.00010383027336900355,
      "loss": 0.0075,
      "step": 1500
    },
    {
      "epoch": 7.365853658536586,
      "grad_norm": 0.08045367151498795,
      "learning_rate": 0.00010280918479544914,
      "loss": 0.0062,
      "step": 1510
    },
    {
      "epoch": 7.414634146341464,
      "grad_norm": 0.044166866689920425,
      "learning_rate": 0.00010178780300699395,
      "loss": 0.008,
      "step": 1520
    },
    {
      "epoch": 7.463414634146342,
      "grad_norm": 0.06291426718235016,
      "learning_rate": 0.00010076623461263018,
      "loss": 0.0066,
      "step": 1530
    },
    {
      "epoch": 7.512195121951219,
      "grad_norm": 0.056730519980192184,
      "learning_rate": 9.974458624082727e-05,
      "loss": 0.0063,
      "step": 1540
    },
    {
      "epoch": 7.560975609756097,
      "grad_norm": 0.18145255744457245,
      "learning_rate": 9.872296452840264e-05,
      "loss": 0.007,
      "step": 1550
    },
    {
      "epoch": 7.609756097560975,
      "grad_norm": 0.038042183965444565,
      "learning_rate": 9.770147610939097e-05,
      "loss": 0.0071,
      "step": 1560
    },
    {
      "epoch": 7.658536585365853,
      "grad_norm": 0.09980273991823196,
      "learning_rate": 9.668022760391418e-05,
      "loss": 0.0083,
      "step": 1570
    },
    {
      "epoch": 7.7073170731707314,
      "grad_norm": 0.04334135353565216,
      "learning_rate": 9.565932560705277e-05,
      "loss": 0.0073,
      "step": 1580
    },
    {
      "epoch": 7.7560975609756095,
      "grad_norm": 0.08293572813272476,
      "learning_rate": 9.463887667771946e-05,
      "loss": 0.007,
      "step": 1590
    },
    {
      "epoch": 7.804878048780488,
      "grad_norm": 0.06575772166252136,
      "learning_rate": 9.361898732753716e-05,
      "loss": 0.007,
      "step": 1600
    },
    {
      "epoch": 7.853658536585366,
      "grad_norm": 0.13714659214019775,
      "learning_rate": 9.259976400972147e-05,
      "loss": 0.0071,
      "step": 1610
    },
    {
      "epoch": 7.902439024390244,
      "grad_norm": 0.05362860858440399,
      "learning_rate": 9.15813131079693e-05,
      "loss": 0.0069,
      "step": 1620
    },
    {
      "epoch": 7.951219512195122,
      "grad_norm": 0.036402106285095215,
      "learning_rate": 9.056374092535504e-05,
      "loss": 0.0068,
      "step": 1630
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.06819606572389603,
      "learning_rate": 8.954715367323468e-05,
      "loss": 0.0074,
      "step": 1640
    },
    {
      "epoch": 8.048780487804878,
      "grad_norm": 0.051237184554338455,
      "learning_rate": 8.853165746015997e-05,
      "loss": 0.0061,
      "step": 1650
    },
    {
      "epoch": 8.097560975609756,
      "grad_norm": 0.04173099994659424,
      "learning_rate": 8.751735828080308e-05,
      "loss": 0.0071,
      "step": 1660
    },
    {
      "epoch": 8.146341463414634,
      "grad_norm": 0.04740411415696144,
      "learning_rate": 8.650436200489303e-05,
      "loss": 0.0062,
      "step": 1670
    },
    {
      "epoch": 8.195121951219512,
      "grad_norm": 0.07791474461555481,
      "learning_rate": 8.549277436616551e-05,
      "loss": 0.0063,
      "step": 1680
    },
    {
      "epoch": 8.24390243902439,
      "grad_norm": 0.10498108714818954,
      "learning_rate": 8.448270095132652e-05,
      "loss": 0.0061,
      "step": 1690
    },
    {
      "epoch": 8.292682926829269,
      "grad_norm": 0.07452121376991272,
      "learning_rate": 8.347424718903151e-05,
      "loss": 0.0063,
      "step": 1700
    },
    {
      "epoch": 8.341463414634147,
      "grad_norm": 0.07632943242788315,
      "learning_rate": 8.246751833888122e-05,
      "loss": 0.0067,
      "step": 1710
    },
    {
      "epoch": 8.390243902439025,
      "grad_norm": 0.040240757167339325,
      "learning_rate": 8.146261948043485e-05,
      "loss": 0.0063,
      "step": 1720
    },
    {
      "epoch": 8.439024390243903,
      "grad_norm": 0.06094524264335632,
      "learning_rate": 8.0459655502242e-05,
      "loss": 0.0061,
      "step": 1730
    },
    {
      "epoch": 8.487804878048781,
      "grad_norm": 0.09173769503831863,
      "learning_rate": 7.945873109089503e-05,
      "loss": 0.0066,
      "step": 1740
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 0.07450529932975769,
      "learning_rate": 7.845995072010188e-05,
      "loss": 0.0065,
      "step": 1750
    },
    {
      "epoch": 8.585365853658537,
      "grad_norm": 0.07175474613904953,
      "learning_rate": 7.74634186397815e-05,
      "loss": 0.0063,
      "step": 1760
    },
    {
      "epoch": 8.634146341463415,
      "grad_norm": 0.1444873958826065,
      "learning_rate": 7.646923886518264e-05,
      "loss": 0.0066,
      "step": 1770
    },
    {
      "epoch": 8.682926829268293,
      "grad_norm": 0.06203252822160721,
      "learning_rate": 7.54775151660267e-05,
      "loss": 0.0061,
      "step": 1780
    },
    {
      "epoch": 8.731707317073171,
      "grad_norm": 0.07840217649936676,
      "learning_rate": 7.448835105567696e-05,
      "loss": 0.0066,
      "step": 1790
    },
    {
      "epoch": 8.78048780487805,
      "grad_norm": 0.05611024424433708,
      "learning_rate": 7.350184978033386e-05,
      "loss": 0.0063,
      "step": 1800
    },
    {
      "epoch": 8.829268292682928,
      "grad_norm": 0.05446132645010948,
      "learning_rate": 7.251811430825846e-05,
      "loss": 0.0072,
      "step": 1810
    },
    {
      "epoch": 8.878048780487806,
      "grad_norm": 0.06147909164428711,
      "learning_rate": 7.153724731902506e-05,
      "loss": 0.0063,
      "step": 1820
    },
    {
      "epoch": 8.926829268292684,
      "grad_norm": 0.137003555893898,
      "learning_rate": 7.055935119280369e-05,
      "loss": 0.0066,
      "step": 1830
    },
    {
      "epoch": 8.975609756097562,
      "grad_norm": 0.15653973817825317,
      "learning_rate": 6.958452799967385e-05,
      "loss": 0.007,
      "step": 1840
    },
    {
      "epoch": 9.024390243902438,
      "grad_norm": 0.06786433607339859,
      "learning_rate": 6.861287948897091e-05,
      "loss": 0.0069,
      "step": 1850
    },
    {
      "epoch": 9.073170731707316,
      "grad_norm": 0.07620502263307571,
      "learning_rate": 6.764450707866576e-05,
      "loss": 0.0062,
      "step": 1860
    },
    {
      "epoch": 9.121951219512194,
      "grad_norm": 0.05350011587142944,
      "learning_rate": 6.667951184477893e-05,
      "loss": 0.0059,
      "step": 1870
    },
    {
      "epoch": 9.170731707317072,
      "grad_norm": 0.05874161049723625,
      "learning_rate": 6.57179945108307e-05,
      "loss": 0.0054,
      "step": 1880
    },
    {
      "epoch": 9.21951219512195,
      "grad_norm": 0.07297080755233765,
      "learning_rate": 6.476005543732782e-05,
      "loss": 0.0067,
      "step": 1890
    },
    {
      "epoch": 9.268292682926829,
      "grad_norm": 0.08384808897972107,
      "learning_rate": 6.380579461128819e-05,
      "loss": 0.0065,
      "step": 1900
    },
    {
      "epoch": 9.317073170731707,
      "grad_norm": 0.06501758098602295,
      "learning_rate": 6.285531163580449e-05,
      "loss": 0.0058,
      "step": 1910
    },
    {
      "epoch": 9.365853658536585,
      "grad_norm": 0.051126834005117416,
      "learning_rate": 6.190870571964774e-05,
      "loss": 0.0061,
      "step": 1920
    },
    {
      "epoch": 9.414634146341463,
      "grad_norm": 0.07436596602201462,
      "learning_rate": 6.096607566691235e-05,
      "loss": 0.0064,
      "step": 1930
    },
    {
      "epoch": 9.463414634146341,
      "grad_norm": 0.0661911889910698,
      "learning_rate": 6.002751986670323e-05,
      "loss": 0.0061,
      "step": 1940
    },
    {
      "epoch": 9.512195121951219,
      "grad_norm": 0.058736588805913925,
      "learning_rate": 5.909313628286601e-05,
      "loss": 0.0063,
      "step": 1950
    },
    {
      "epoch": 9.560975609756097,
      "grad_norm": 0.051081206649541855,
      "learning_rate": 5.816302244376206e-05,
      "loss": 0.0074,
      "step": 1960
    },
    {
      "epoch": 9.609756097560975,
      "grad_norm": 0.06269817799329758,
      "learning_rate": 5.7237275432088664e-05,
      "loss": 0.0057,
      "step": 1970
    },
    {
      "epoch": 9.658536585365853,
      "grad_norm": 0.0546838752925396,
      "learning_rate": 5.63159918747457e-05,
      "loss": 0.0063,
      "step": 1980
    },
    {
      "epoch": 9.707317073170731,
      "grad_norm": 0.09113246202468872,
      "learning_rate": 5.539926793275021e-05,
      "loss": 0.0067,
      "step": 1990
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 0.06070900708436966,
      "learning_rate": 5.448719929119915e-05,
      "loss": 0.0065,
      "step": 2000
    },
    {
      "epoch": 9.804878048780488,
      "grad_norm": 0.06381196528673172,
      "learning_rate": 5.3579881149282204e-05,
      "loss": 0.0063,
      "step": 2010
    },
    {
      "epoch": 9.853658536585366,
      "grad_norm": 0.05914902687072754,
      "learning_rate": 5.267740821034524e-05,
      "loss": 0.0059,
      "step": 2020
    },
    {
      "epoch": 9.902439024390244,
      "grad_norm": 0.0414993092417717,
      "learning_rate": 5.177987467200504e-05,
      "loss": 0.0065,
      "step": 2030
    },
    {
      "epoch": 9.951219512195122,
      "grad_norm": 0.051615435630083084,
      "learning_rate": 5.088737421631767e-05,
      "loss": 0.0061,
      "step": 2040
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.08131363987922668,
      "learning_rate": 5.000000000000002e-05,
      "loss": 0.0058,
      "step": 2050
    },
    {
      "epoch": 10.048780487804878,
      "grad_norm": 0.06244876608252525,
      "learning_rate": 4.911784464470623e-05,
      "loss": 0.0057,
      "step": 2060
    },
    {
      "epoch": 10.097560975609756,
      "grad_norm": 0.060385096818208694,
      "learning_rate": 4.8241000227360354e-05,
      "loss": 0.0062,
      "step": 2070
    },
    {
      "epoch": 10.146341463414634,
      "grad_norm": 0.052863430231809616,
      "learning_rate": 4.73695582705455e-05,
      "loss": 0.0062,
      "step": 2080
    },
    {
      "epoch": 10.195121951219512,
      "grad_norm": 0.07917828112840652,
      "learning_rate": 4.650360973295086e-05,
      "loss": 0.0061,
      "step": 2090
    },
    {
      "epoch": 10.24390243902439,
      "grad_norm": 0.054362453520298004,
      "learning_rate": 4.56432449998779e-05,
      "loss": 0.0069,
      "step": 2100
    },
    {
      "epoch": 10.292682926829269,
      "grad_norm": 0.06768287718296051,
      "learning_rate": 4.478855387380605e-05,
      "loss": 0.0062,
      "step": 2110
    },
    {
      "epoch": 10.341463414634147,
      "grad_norm": 0.06945596635341644,
      "learning_rate": 4.3939625565019416e-05,
      "loss": 0.0062,
      "step": 2120
    },
    {
      "epoch": 10.390243902439025,
      "grad_norm": 0.07058496028184891,
      "learning_rate": 4.30965486822953e-05,
      "loss": 0.0064,
      "step": 2130
    },
    {
      "epoch": 10.439024390243903,
      "grad_norm": 0.06080787256360054,
      "learning_rate": 4.225941122365529e-05,
      "loss": 0.0058,
      "step": 2140
    },
    {
      "epoch": 10.487804878048781,
      "grad_norm": 0.059627074748277664,
      "learning_rate": 4.142830056718052e-05,
      "loss": 0.0062,
      "step": 2150
    },
    {
      "epoch": 10.536585365853659,
      "grad_norm": 0.05086175352334976,
      "learning_rate": 4.0603303461891253e-05,
      "loss": 0.0056,
      "step": 2160
    },
    {
      "epoch": 10.585365853658537,
      "grad_norm": 0.1199130192399025,
      "learning_rate": 3.978450601869217e-05,
      "loss": 0.0062,
      "step": 2170
    },
    {
      "epoch": 10.634146341463415,
      "grad_norm": 0.1347583681344986,
      "learning_rate": 3.8971993701384554e-05,
      "loss": 0.0064,
      "step": 2180
    },
    {
      "epoch": 10.682926829268293,
      "grad_norm": 0.050621725618839264,
      "learning_rate": 3.816585131774571e-05,
      "loss": 0.0057,
      "step": 2190
    },
    {
      "epoch": 10.731707317073171,
      "grad_norm": 0.04940760135650635,
      "learning_rate": 3.736616301067694e-05,
      "loss": 0.0062,
      "step": 2200
    },
    {
      "epoch": 10.78048780487805,
      "grad_norm": 0.052099861204624176,
      "learning_rate": 3.657301224942098e-05,
      "loss": 0.0058,
      "step": 2210
    },
    {
      "epoch": 10.829268292682928,
      "grad_norm": 0.08436452597379684,
      "learning_rate": 3.578648182084975e-05,
      "loss": 0.0058,
      "step": 2220
    },
    {
      "epoch": 10.878048780487806,
      "grad_norm": 0.0537634901702404,
      "learning_rate": 3.5006653820823135e-05,
      "loss": 0.0056,
      "step": 2230
    },
    {
      "epoch": 10.926829268292684,
      "grad_norm": 0.06204523518681526,
      "learning_rate": 3.423360964562026e-05,
      "loss": 0.0058,
      "step": 2240
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 0.1142859160900116,
      "learning_rate": 3.346742998344348e-05,
      "loss": 0.0062,
      "step": 2250
    },
    {
      "epoch": 11.024390243902438,
      "grad_norm": 0.06821559369564056,
      "learning_rate": 3.270819480599625e-05,
      "loss": 0.006,
      "step": 2260
    },
    {
      "epoch": 11.073170731707316,
      "grad_norm": 0.05169494450092316,
      "learning_rate": 3.195598336013615e-05,
      "loss": 0.0053,
      "step": 2270
    },
    {
      "epoch": 11.121951219512194,
      "grad_norm": 0.05910908058285713,
      "learning_rate": 3.121087415960304e-05,
      "loss": 0.0059,
      "step": 2280
    },
    {
      "epoch": 11.170731707317072,
      "grad_norm": 0.07483071833848953,
      "learning_rate": 3.0472944976824224e-05,
      "loss": 0.0058,
      "step": 2290
    },
    {
      "epoch": 11.21951219512195,
      "grad_norm": 0.0738058090209961,
      "learning_rate": 2.9742272834796813e-05,
      "loss": 0.0055,
      "step": 2300
    },
    {
      "epoch": 11.268292682926829,
      "grad_norm": 0.09469424933195114,
      "learning_rate": 2.901893399904797e-05,
      "loss": 0.0062,
      "step": 2310
    },
    {
      "epoch": 11.317073170731707,
      "grad_norm": 0.05931226164102554,
      "learning_rate": 2.830300396967487e-05,
      "loss": 0.0059,
      "step": 2320
    },
    {
      "epoch": 11.365853658536585,
      "grad_norm": 0.07385493069887161,
      "learning_rate": 2.7594557473464066e-05,
      "loss": 0.0059,
      "step": 2330
    },
    {
      "epoch": 11.414634146341463,
      "grad_norm": 0.08171911537647247,
      "learning_rate": 2.6893668456091625e-05,
      "loss": 0.0057,
      "step": 2340
    },
    {
      "epoch": 11.463414634146341,
      "grad_norm": 0.05290233716368675,
      "learning_rate": 2.620041007440508e-05,
      "loss": 0.0064,
      "step": 2350
    },
    {
      "epoch": 11.512195121951219,
      "grad_norm": 0.06285463273525238,
      "learning_rate": 2.5514854688787405e-05,
      "loss": 0.0062,
      "step": 2360
    },
    {
      "epoch": 11.560975609756097,
      "grad_norm": 0.06230178475379944,
      "learning_rate": 2.4837073855604188e-05,
      "loss": 0.0061,
      "step": 2370
    },
    {
      "epoch": 11.609756097560975,
      "grad_norm": 0.05689137428998947,
      "learning_rate": 2.4167138319734905e-05,
      "loss": 0.006,
      "step": 2380
    },
    {
      "epoch": 11.658536585365853,
      "grad_norm": 0.07083682715892792,
      "learning_rate": 2.350511800718871e-05,
      "loss": 0.0057,
      "step": 2390
    },
    {
      "epoch": 11.707317073170731,
      "grad_norm": 0.056767258793115616,
      "learning_rate": 2.2851082017805703e-05,
      "loss": 0.0056,
      "step": 2400
    },
    {
      "epoch": 11.75609756097561,
      "grad_norm": 0.06561973690986633,
      "learning_rate": 2.2205098618044583e-05,
      "loss": 0.0058,
      "step": 2410
    },
    {
      "epoch": 11.804878048780488,
      "grad_norm": 0.05094975233078003,
      "learning_rate": 2.1567235233857054e-05,
      "loss": 0.0057,
      "step": 2420
    },
    {
      "epoch": 11.853658536585366,
      "grad_norm": 0.0570388026535511,
      "learning_rate": 2.0937558443650206e-05,
      "loss": 0.006,
      "step": 2430
    },
    {
      "epoch": 11.902439024390244,
      "grad_norm": 0.06019320338964462,
      "learning_rate": 2.031613397133724e-05,
      "loss": 0.006,
      "step": 2440
    },
    {
      "epoch": 11.951219512195122,
      "grad_norm": 0.09152466058731079,
      "learning_rate": 1.9703026679477256e-05,
      "loss": 0.0059,
      "step": 2450
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.06767109036445618,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 0.0058,
      "step": 2460
    },
    {
      "epoch": 12.048780487804878,
      "grad_norm": 0.05434449389576912,
      "learning_rate": 1.8502018740052496e-05,
      "loss": 0.0057,
      "step": 2470
    },
    {
      "epoch": 12.097560975609756,
      "grad_norm": 0.061113059520721436,
      "learning_rate": 1.7914243450358216e-05,
      "loss": 0.0054,
      "step": 2480
    },
    {
      "epoch": 12.146341463414634,
      "grad_norm": 0.05813341587781906,
      "learning_rate": 1.73350360437734e-05,
      "loss": 0.0057,
      "step": 2490
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 0.07934972643852234,
      "learning_rate": 1.676445697635728e-05,
      "loss": 0.0057,
      "step": 2500
    },
    {
      "epoch": 12.24390243902439,
      "grad_norm": 0.08221197873353958,
      "learning_rate": 1.6202565803566917e-05,
      "loss": 0.0056,
      "step": 2510
    },
    {
      "epoch": 12.292682926829269,
      "grad_norm": 0.07954658567905426,
      "learning_rate": 1.564942117404119e-05,
      "loss": 0.0056,
      "step": 2520
    },
    {
      "epoch": 12.341463414634147,
      "grad_norm": 0.07448497414588928,
      "learning_rate": 1.510508082347899e-05,
      "loss": 0.0058,
      "step": 2530
    },
    {
      "epoch": 12.390243902439025,
      "grad_norm": 0.06170118972659111,
      "learning_rate": 1.4569601568613144e-05,
      "loss": 0.0058,
      "step": 2540
    },
    {
      "epoch": 12.439024390243903,
      "grad_norm": 0.05958204343914986,
      "learning_rate": 1.4043039301279903e-05,
      "loss": 0.0058,
      "step": 2550
    },
    {
      "epoch": 12.487804878048781,
      "grad_norm": 0.05719635635614395,
      "learning_rate": 1.352544898258511e-05,
      "loss": 0.0056,
      "step": 2560
    },
    {
      "epoch": 12.536585365853659,
      "grad_norm": 0.056574221700429916,
      "learning_rate": 1.301688463716757e-05,
      "loss": 0.0063,
      "step": 2570
    },
    {
      "epoch": 12.585365853658537,
      "grad_norm": 0.14904461801052094,
      "learning_rate": 1.2517399347560167e-05,
      "loss": 0.0059,
      "step": 2580
    },
    {
      "epoch": 12.634146341463415,
      "grad_norm": 0.03999248892068863,
      "learning_rate": 1.2027045248649015e-05,
      "loss": 0.0057,
      "step": 2590
    },
    {
      "epoch": 12.682926829268293,
      "grad_norm": 0.05714278668165207,
      "learning_rate": 1.1545873522232053e-05,
      "loss": 0.0056,
      "step": 2600
    },
    {
      "epoch": 12.731707317073171,
      "grad_norm": 0.06596580147743225,
      "learning_rate": 1.1073934391676667e-05,
      "loss": 0.0058,
      "step": 2610
    },
    {
      "epoch": 12.78048780487805,
      "grad_norm": 0.06003756448626518,
      "learning_rate": 1.0611277116677487e-05,
      "loss": 0.0059,
      "step": 2620
    },
    {
      "epoch": 12.829268292682928,
      "grad_norm": 0.05459544062614441,
      "learning_rate": 1.0157949988114967e-05,
      "loss": 0.0059,
      "step": 2630
    },
    {
      "epoch": 12.878048780487806,
      "grad_norm": 0.061511751264333725,
      "learning_rate": 9.714000323014705e-06,
      "loss": 0.0058,
      "step": 2640
    },
    {
      "epoch": 12.926829268292684,
      "grad_norm": 0.048676956444978714,
      "learning_rate": 9.279474459608805e-06,
      "loss": 0.0059,
      "step": 2650
    },
    {
      "epoch": 12.975609756097562,
      "grad_norm": 0.05746001750230789,
      "learning_rate": 8.854417752499111e-06,
      "loss": 0.0057,
      "step": 2660
    },
    {
      "epoch": 13.024390243902438,
      "grad_norm": 0.05339556187391281,
      "learning_rate": 8.43887456792326e-06,
      "loss": 0.0057,
      "step": 2670
    },
    {
      "epoch": 13.073170731707316,
      "grad_norm": 0.06085890159010887,
      "learning_rate": 8.032888279123829e-06,
      "loss": 0.0056,
      "step": 2680
    },
    {
      "epoch": 13.121951219512194,
      "grad_norm": 0.06632421165704727,
      "learning_rate": 7.636501261821238e-06,
      "loss": 0.0057,
      "step": 2690
    },
    {
      "epoch": 13.170731707317072,
      "grad_norm": 0.07371345162391663,
      "learning_rate": 7.249754889790539e-06,
      "loss": 0.0058,
      "step": 2700
    },
    {
      "epoch": 13.21951219512195,
      "grad_norm": 0.06796526908874512,
      "learning_rate": 6.872689530543086e-06,
      "loss": 0.0059,
      "step": 2710
    },
    {
      "epoch": 13.268292682926829,
      "grad_norm": 0.05836968868970871,
      "learning_rate": 6.505344541113046e-06,
      "loss": 0.0058,
      "step": 2720
    },
    {
      "epoch": 13.317073170731707,
      "grad_norm": 0.05631383880972862,
      "learning_rate": 6.147758263949321e-06,
      "loss": 0.0055,
      "step": 2730
    },
    {
      "epoch": 13.365853658536585,
      "grad_norm": 0.057296037673950195,
      "learning_rate": 5.7999680229136375e-06,
      "loss": 0.0057,
      "step": 2740
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 0.059217844158411026,
      "learning_rate": 5.462010119384664e-06,
      "loss": 0.0056,
      "step": 2750
    },
    {
      "epoch": 13.463414634146341,
      "grad_norm": 0.052348434925079346,
      "learning_rate": 5.1339198284689916e-06,
      "loss": 0.0057,
      "step": 2760
    },
    {
      "epoch": 13.512195121951219,
      "grad_norm": 0.04569809511303902,
      "learning_rate": 4.815731395319278e-06,
      "loss": 0.0055,
      "step": 2770
    },
    {
      "epoch": 13.560975609756097,
      "grad_norm": 0.05811962112784386,
      "learning_rate": 4.507478031559731e-06,
      "loss": 0.0057,
      "step": 2780
    },
    {
      "epoch": 13.609756097560975,
      "grad_norm": 0.04801028221845627,
      "learning_rate": 4.209191911819688e-06,
      "loss": 0.0055,
      "step": 2790
    },
    {
      "epoch": 13.658536585365853,
      "grad_norm": 0.04857472702860832,
      "learning_rate": 3.9209041703752395e-06,
      "loss": 0.0056,
      "step": 2800
    },
    {
      "epoch": 13.707317073170731,
      "grad_norm": 0.09129537642002106,
      "learning_rate": 3.6426448978995056e-06,
      "loss": 0.006,
      "step": 2810
    },
    {
      "epoch": 13.75609756097561,
      "grad_norm": 0.06310546398162842,
      "learning_rate": 3.3744431383219367e-06,
      "loss": 0.0055,
      "step": 2820
    },
    {
      "epoch": 13.804878048780488,
      "grad_norm": 0.055802859365940094,
      "learning_rate": 3.116326885796728e-06,
      "loss": 0.0056,
      "step": 2830
    },
    {
      "epoch": 13.853658536585366,
      "grad_norm": 0.0634494498372078,
      "learning_rate": 2.868323081780877e-06,
      "loss": 0.0055,
      "step": 2840
    },
    {
      "epoch": 13.902439024390244,
      "grad_norm": 0.05695090815424919,
      "learning_rate": 2.6304576122221035e-06,
      "loss": 0.0055,
      "step": 2850
    },
    {
      "epoch": 13.951219512195122,
      "grad_norm": 0.06994400918483734,
      "learning_rate": 2.402755304856974e-06,
      "loss": 0.0057,
      "step": 2860
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.0930507555603981,
      "learning_rate": 2.1852399266194314e-06,
      "loss": 0.006,
      "step": 2870
    },
    {
      "epoch": 14.048780487804878,
      "grad_norm": 0.056339725852012634,
      "learning_rate": 1.9779341811600794e-06,
      "loss": 0.0056,
      "step": 2880
    },
    {
      "epoch": 14.097560975609756,
      "grad_norm": 0.05481722205877304,
      "learning_rate": 1.7808597064764009e-06,
      "loss": 0.0056,
      "step": 2890
    },
    {
      "epoch": 14.146341463414634,
      "grad_norm": 0.07605459541082382,
      "learning_rate": 1.5940370726542863e-06,
      "loss": 0.0055,
      "step": 2900
    },
    {
      "epoch": 14.195121951219512,
      "grad_norm": 0.07780090719461441,
      "learning_rate": 1.417485779720995e-06,
      "loss": 0.0055,
      "step": 2910
    },
    {
      "epoch": 14.24390243902439,
      "grad_norm": 0.06754057854413986,
      "learning_rate": 1.2512242556097398e-06,
      "loss": 0.0055,
      "step": 2920
    },
    {
      "epoch": 14.292682926829269,
      "grad_norm": 0.055912312120199203,
      "learning_rate": 1.0952698542362805e-06,
      "loss": 0.0056,
      "step": 2930
    },
    {
      "epoch": 14.341463414634147,
      "grad_norm": 0.06355050206184387,
      "learning_rate": 9.496388536875623e-07,
      "loss": 0.0058,
      "step": 2940
    },
    {
      "epoch": 14.390243902439025,
      "grad_norm": 0.05702829360961914,
      "learning_rate": 8.143464545226298e-07,
      "loss": 0.0057,
      "step": 2950
    },
    {
      "epoch": 14.439024390243903,
      "grad_norm": 0.05529710277915001,
      "learning_rate": 6.894067781860635e-07,
      "loss": 0.0057,
      "step": 2960
    },
    {
      "epoch": 14.487804878048781,
      "grad_norm": 0.073829285800457,
      "learning_rate": 5.74832865534014e-07,
      "loss": 0.0055,
      "step": 2970
    },
    {
      "epoch": 14.536585365853659,
      "grad_norm": 0.054862912744283676,
      "learning_rate": 4.7063667547301294e-07,
      "loss": 0.0058,
      "step": 2980
    },
    {
      "epoch": 14.585365853658537,
      "grad_norm": 0.07511288672685623,
      "learning_rate": 3.768290837117605e-07,
      "loss": 0.0056,
      "step": 2990
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 0.05845462158322334,
      "learning_rate": 2.934198816259559e-07,
      "loss": 0.0055,
      "step": 3000
    },
    {
      "epoch": 14.682926829268293,
      "grad_norm": 0.0624593086540699,
      "learning_rate": 2.2041777523627017e-07,
      "loss": 0.0057,
      "step": 3010
    },
    {
      "epoch": 14.731707317073171,
      "grad_norm": 0.056128259748220444,
      "learning_rate": 1.5783038429965092e-07,
      "loss": 0.0055,
      "step": 3020
    },
    {
      "epoch": 14.78048780487805,
      "grad_norm": 0.0808698758482933,
      "learning_rate": 1.0566424151401411e-07,
      "loss": 0.0057,
      "step": 3030
    },
    {
      "epoch": 14.829268292682928,
      "grad_norm": 0.0584842748939991,
      "learning_rate": 6.392479183633394e-08,
      "loss": 0.0055,
      "step": 3040
    },
    {
      "epoch": 14.878048780487806,
      "grad_norm": 0.07366840541362762,
      "learning_rate": 3.2616391914364054e-08,
      "loss": 0.0055,
      "step": 3050
    },
    {
      "epoch": 14.926829268292684,
      "grad_norm": 0.06805972754955292,
      "learning_rate": 1.1742309631845861e-08,
      "loss": 0.0055,
      "step": 3060
    },
    {
      "epoch": 14.975609756097562,
      "grad_norm": 0.057060301303863525,
      "learning_rate": 1.304723767492355e-09,
      "loss": 0.0056,
      "step": 3070
    }
  ],
  "logging_steps": 10,
  "max_steps": 3075,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.23238405675008e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}