{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010548523206751054,
      "grad_norm": 1.1911535263061523,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.514,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1261857748031616,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.971,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8440089821815491,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.3116,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 1.437301754951477,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.8497,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 1.4225043058395386,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.5173,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.9975273013114929,
      "learning_rate": 0.0009493670886075949,
      "loss": 5.1521,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.0305777788162231,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.7579,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.3440488576889038,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.4427,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 1.0169594287872314,
      "learning_rate": 0.0014240506329113926,
      "loss": 4.1901,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.8410079479217529,
      "learning_rate": 0.0015,
      "loss": 4.0012,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.6870416402816772,
      "learning_rate": 0.0015,
      "loss": 3.8088,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.6572973728179932,
      "learning_rate": 0.0015,
      "loss": 3.6725,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.535599410533905,
      "learning_rate": 0.0015,
      "loss": 3.5343,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.8405603766441345,
      "learning_rate": 0.0015,
      "loss": 3.4165,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.5135706663131714,
      "learning_rate": 0.0015,
      "loss": 3.3249,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.6049426794052124,
      "learning_rate": 0.0015,
      "loss": 3.2309,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.6043658256530762,
      "learning_rate": 0.0015,
      "loss": 3.1535,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.7351441383361816,
      "learning_rate": 0.0015,
      "loss": 3.1078,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.5547571182250977,
      "learning_rate": 0.0015,
      "loss": 3.0352,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.7150177955627441,
      "learning_rate": 0.0015,
      "loss": 2.9785,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.5605180263519287,
      "learning_rate": 0.0015,
      "loss": 2.9324,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.5501560568809509,
      "learning_rate": 0.0015,
      "loss": 2.892,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.5405913591384888,
      "learning_rate": 0.0015,
      "loss": 2.8553,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.44567427039146423,
      "learning_rate": 0.0015,
      "loss": 2.8098,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.5209677815437317,
      "learning_rate": 0.0015,
      "loss": 2.7686,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.5224133133888245,
      "learning_rate": 0.0015,
      "loss": 2.7305,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.561279296875,
      "learning_rate": 0.0015,
      "loss": 2.7045,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.48504865169525146,
      "learning_rate": 0.0015,
      "loss": 2.6616,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.5276456475257874,
      "learning_rate": 0.0015,
      "loss": 2.6373,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.7674102783203125,
      "learning_rate": 0.0015,
      "loss": 2.6083,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.5037146806716919,
      "learning_rate": 0.0015,
      "loss": 2.583,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.4920331835746765,
      "learning_rate": 0.0015,
      "loss": 2.5358,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.6064138412475586,
      "learning_rate": 0.0015,
      "loss": 2.5224,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.5508156418800354,
      "learning_rate": 0.0015,
      "loss": 2.4869,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.46913155913352966,
      "learning_rate": 0.0015,
      "loss": 2.4401,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.6472739577293396,
      "learning_rate": 0.0015,
      "loss": 2.4113,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.5063979029655457,
      "learning_rate": 0.0015,
      "loss": 2.3883,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.5765858888626099,
      "learning_rate": 0.0015,
      "loss": 2.3691,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.5593698620796204,
      "learning_rate": 0.0015,
      "loss": 2.3406,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.7147346138954163,
      "learning_rate": 0.0015,
      "loss": 2.3005,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.5318769812583923,
      "learning_rate": 0.0015,
      "loss": 2.2975,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.5068933963775635,
      "learning_rate": 0.0015,
      "loss": 2.2618,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.5029956102371216,
      "learning_rate": 0.0015,
      "loss": 2.2284,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.5769665241241455,
      "learning_rate": 0.0015,
      "loss": 2.2023,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.6128251552581787,
      "learning_rate": 0.0015,
      "loss": 2.2,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.530529260635376,
      "learning_rate": 0.0015,
      "loss": 2.1738,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.5000449419021606,
      "learning_rate": 0.0015,
      "loss": 2.1554,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.7125113606452942,
      "learning_rate": 0.0015,
      "loss": 2.1395,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.4939398467540741,
      "learning_rate": 0.0015,
      "loss": 2.105,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.5103901028633118,
      "learning_rate": 0.0015,
      "loss": 2.0902,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.4903666377067566,
      "learning_rate": 0.0015,
      "loss": 2.0743,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.6215901374816895,
      "learning_rate": 0.0015,
      "loss": 2.0572,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.5370311141014099,
      "learning_rate": 0.0015,
      "loss": 2.0422,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.583985447883606,
      "learning_rate": 0.0015,
      "loss": 2.0237,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.48627397418022156,
      "learning_rate": 0.0015,
      "loss": 2.0133,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.46253713965415955,
      "learning_rate": 0.0015,
      "loss": 1.9705,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.5268188118934631,
      "learning_rate": 0.0015,
      "loss": 1.9772,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 0.526195764541626,
      "learning_rate": 0.0015,
      "loss": 1.9743,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.501920759677887,
      "learning_rate": 0.0015,
      "loss": 1.9446,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.5201966166496277,
      "learning_rate": 0.0015,
      "loss": 1.9259,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.49704989790916443,
      "learning_rate": 0.0015,
      "loss": 1.9233,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.46477070450782776,
      "learning_rate": 0.0015,
      "loss": 1.9054,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.5311816334724426,
      "learning_rate": 0.0015,
      "loss": 1.892,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.6093870401382446,
      "learning_rate": 0.0015,
      "loss": 1.907,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.5302767157554626,
      "learning_rate": 0.0015,
      "loss": 1.8872,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.5349369049072266,
      "learning_rate": 0.0015,
      "loss": 1.8741,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.5523291230201721,
      "learning_rate": 0.0015,
      "loss": 1.8571,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.45728522539138794,
      "learning_rate": 0.0015,
      "loss": 1.8468,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.49975475668907166,
      "learning_rate": 0.0015,
      "loss": 1.8546,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.5351147651672363,
      "learning_rate": 0.0015,
      "loss": 1.8344,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.4953480362892151,
      "learning_rate": 0.0015,
      "loss": 1.8199,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.5598252415657043,
      "learning_rate": 0.0015,
      "loss": 1.806,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.636795699596405,
      "learning_rate": 0.0015,
      "loss": 1.795,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.5506213903427124,
      "learning_rate": 0.0015,
      "loss": 1.8043,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.5925042033195496,
      "learning_rate": 0.0015,
      "loss": 1.7829,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 0.7445921301841736,
      "learning_rate": 0.0015,
      "loss": 1.7803,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.48227658867836,
      "learning_rate": 0.0015,
      "loss": 1.7703,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.5093647837638855,
      "learning_rate": 0.0015,
      "loss": 1.768,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.4523041844367981,
      "learning_rate": 0.0015,
      "loss": 1.7647,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.48087066411972046,
      "learning_rate": 0.0015,
      "loss": 1.743,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.5096999406814575,
      "learning_rate": 0.0015,
      "loss": 1.7307,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.557677149772644,
      "learning_rate": 0.0015,
      "loss": 1.7368,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.5189062356948853,
      "learning_rate": 0.0015,
      "loss": 1.7393,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.48466941714286804,
      "learning_rate": 0.0015,
      "loss": 1.7153,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.508627712726593,
      "learning_rate": 0.0015,
      "loss": 1.713,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.44129225611686707,
      "learning_rate": 0.0015,
      "loss": 1.7096,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.5102670788764954,
      "learning_rate": 0.0015,
      "loss": 1.7063,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.5179425477981567,
      "learning_rate": 0.0015,
      "loss": 1.7001,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.46534618735313416,
      "learning_rate": 0.0015,
      "loss": 1.6878,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.4997912049293518,
      "learning_rate": 0.0015,
      "loss": 1.6861,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.508382260799408,
      "learning_rate": 0.0015,
      "loss": 1.6949,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.4784116744995117,
      "learning_rate": 0.0015,
      "loss": 1.6795,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.4688781797885895,
      "learning_rate": 0.0015,
      "loss": 1.6661,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.5456479787826538,
      "learning_rate": 0.0015,
      "loss": 1.6739,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.5274268984794617,
      "learning_rate": 0.0015,
      "loss": 1.6537,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.4864138960838318,
      "learning_rate": 0.0015,
      "loss": 1.6592,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 0.5758424997329712,
      "learning_rate": 0.0015,
      "loss": 1.6613,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.4512311816215515,
      "learning_rate": 0.0015,
      "loss": 1.6414,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.45962268114089966,
      "learning_rate": 0.0015,
      "loss": 1.6391,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.5761123895645142,
      "learning_rate": 0.0015,
      "loss": 1.6495,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.4733625650405884,
      "learning_rate": 0.0015,
      "loss": 1.6366,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.5327010154724121,
      "learning_rate": 0.0015,
      "loss": 1.6276,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.49683907628059387,
      "learning_rate": 0.0015,
      "loss": 1.6303,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.5400471687316895,
      "learning_rate": 0.0015,
      "loss": 1.6251,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.48789605498313904,
      "learning_rate": 0.0015,
      "loss": 1.6224,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.45513924956321716,
      "learning_rate": 0.0015,
      "loss": 1.6116,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.46881577372550964,
      "learning_rate": 0.0015,
      "loss": 1.602,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.4709598124027252,
      "learning_rate": 0.0015,
      "loss": 1.5996,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.43881508708000183,
      "learning_rate": 0.0015,
      "loss": 1.6087,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.4861796796321869,
      "learning_rate": 0.0015,
      "loss": 1.5978,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.4598830044269562,
      "learning_rate": 0.0015,
      "loss": 1.6007,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.46872007846832275,
      "learning_rate": 0.0015,
      "loss": 1.5916,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.5027231574058533,
      "learning_rate": 0.0015,
      "loss": 1.5836,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.4448629319667816,
      "learning_rate": 0.0015,
      "loss": 1.5928,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 0.5146149396896362,
      "learning_rate": 0.0015,
      "loss": 1.5788,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.4398585557937622,
      "learning_rate": 0.0015,
      "loss": 1.5838,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.43229490518569946,
      "learning_rate": 0.0015,
      "loss": 1.5731,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.44356974959373474,
      "learning_rate": 0.0015,
      "loss": 1.5594,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.4866366684436798,
      "learning_rate": 0.0015,
      "loss": 1.5628,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.4466060698032379,
      "learning_rate": 0.0015,
      "loss": 1.5753,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.5837816596031189,
      "learning_rate": 0.0015,
      "loss": 1.5668,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.4354594647884369,
      "learning_rate": 0.0015,
      "loss": 1.5527,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.43021681904792786,
      "learning_rate": 0.0015,
      "loss": 1.5443,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.5192949771881104,
      "learning_rate": 0.0015,
      "loss": 1.5602,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.41557225584983826,
      "learning_rate": 0.0015,
      "loss": 1.5582,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.4372217655181885,
      "learning_rate": 0.0015,
      "loss": 1.5435,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.4619849920272827,
      "learning_rate": 0.0015,
      "loss": 1.5426,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.5369654297828674,
      "learning_rate": 0.0015,
      "loss": 1.5554,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.43552911281585693,
      "learning_rate": 0.0015,
      "loss": 1.5421,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.4249727427959442,
      "learning_rate": 0.0015,
      "loss": 1.5318,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.4574583172798157,
      "learning_rate": 0.0015,
      "loss": 1.5332,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.4392021596431732,
      "learning_rate": 0.0015,
      "loss": 1.5371,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 0.6404235363006592,
      "learning_rate": 0.0015,
      "loss": 1.5272,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.4825304448604584,
      "learning_rate": 0.0015,
      "loss": 1.5276,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.4290938079357147,
      "learning_rate": 0.0015,
      "loss": 1.5289,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.42625662684440613,
      "learning_rate": 0.0015,
      "loss": 1.5239,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.5395568013191223,
      "learning_rate": 0.0015,
      "loss": 1.5224,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.4960879385471344,
      "learning_rate": 0.0015,
      "loss": 1.5122,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.4987901747226715,
      "learning_rate": 0.0015,
      "loss": 1.5219,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.5155091881752014,
      "learning_rate": 0.0015,
      "loss": 1.5035,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.4964143931865692,
      "learning_rate": 0.0015,
      "loss": 1.5091,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.5313318967819214,
      "learning_rate": 0.0015,
      "loss": 1.5126,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.5072162747383118,
      "learning_rate": 0.0015,
      "loss": 1.5043,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.4389128088951111,
      "learning_rate": 0.0015,
      "loss": 1.5022,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.44489845633506775,
      "learning_rate": 0.0015,
      "loss": 1.5024,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.4190062880516052,
      "learning_rate": 0.0015,
      "loss": 1.4957,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.5346754789352417,
      "learning_rate": 0.0015,
      "loss": 1.4955,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.472139447927475,
      "learning_rate": 0.0015,
      "loss": 1.5004,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.4273566007614136,
      "learning_rate": 0.0015,
      "loss": 1.4983,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.5061330199241638,
      "learning_rate": 0.0015,
      "loss": 1.4967,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.4574885964393616,
      "learning_rate": 0.0015,
      "loss": 1.4999,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.5211601257324219,
      "learning_rate": 0.0015,
      "loss": 1.4977,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.44409823417663574,
      "learning_rate": 0.0015,
      "loss": 1.4798,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.4143390357494354,
      "learning_rate": 0.0015,
      "loss": 1.4666,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.45285868644714355,
      "learning_rate": 0.0015,
      "loss": 1.4724,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.44844868779182434,
      "learning_rate": 0.0015,
      "loss": 1.4721,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.43528443574905396,
      "learning_rate": 0.0015,
      "loss": 1.4662,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.4710999131202698,
      "learning_rate": 0.0015,
      "loss": 1.4786,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.4777051508426666,
      "learning_rate": 0.0015,
      "loss": 1.4755,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.43375569581985474,
      "learning_rate": 0.0015,
      "loss": 1.4753,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.44435209035873413,
      "learning_rate": 0.0015,
      "loss": 1.4761,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.4641205370426178,
      "learning_rate": 0.0015,
      "loss": 1.4454,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.43654873967170715,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.43061891198158264,
      "learning_rate": 0.0015,
      "loss": 1.4605,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.4999637007713318,
      "learning_rate": 0.0015,
      "loss": 1.4506,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.4092038571834564,
      "learning_rate": 0.0015,
      "loss": 1.4538,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.4975314438343048,
      "learning_rate": 0.0015,
      "loss": 1.464,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.5646780133247375,
      "learning_rate": 0.0015,
      "loss": 1.4501,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.436535120010376,
      "learning_rate": 0.0015,
      "loss": 1.4464,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.43750905990600586,
      "learning_rate": 0.0015,
      "loss": 1.4623,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.4112912118434906,
      "learning_rate": 0.0015,
      "loss": 1.4389,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.39078304171562195,
      "learning_rate": 0.0015,
      "loss": 1.4371,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.4432763159275055,
      "learning_rate": 0.0015,
      "loss": 1.4397,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.42006704211235046,
      "learning_rate": 0.0015,
      "loss": 1.443,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.44937148690223694,
      "learning_rate": 0.0015,
      "loss": 1.4642,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.4456152617931366,
      "learning_rate": 0.0015,
      "loss": 1.4428,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.4034252464771271,
      "learning_rate": 0.0015,
      "loss": 1.4444,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.41375723481178284,
      "learning_rate": 0.0015,
      "loss": 1.4463,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.44860073924064636,
      "learning_rate": 0.0015,
      "loss": 1.4482,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.4023696482181549,
      "learning_rate": 0.0015,
      "loss": 1.4437,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.42736566066741943,
      "learning_rate": 0.0015,
      "loss": 1.4281,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.48598966002464294,
      "learning_rate": 0.0015,
      "loss": 1.4277,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.45729580521583557,
      "learning_rate": 0.0015,
      "loss": 1.4284,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.41488245129585266,
      "learning_rate": 0.0015,
      "loss": 1.4288,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.4101181626319885,
      "learning_rate": 0.0015,
      "loss": 1.4296,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.3972269296646118,
      "learning_rate": 0.0015,
      "loss": 1.4295,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.4474700391292572,
      "learning_rate": 0.0015,
      "loss": 1.4327,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.4976608157157898,
      "learning_rate": 0.0015,
      "loss": 1.421,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.47018617391586304,
      "learning_rate": 0.0015,
      "loss": 1.4293,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.43353503942489624,
      "learning_rate": 0.0015,
      "loss": 1.4208,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.4655958116054535,
      "learning_rate": 0.0015,
      "loss": 1.4235,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.40534839034080505,
      "learning_rate": 0.0015,
      "loss": 1.4226,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.46445077657699585,
      "learning_rate": 0.0015,
      "loss": 1.4226,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.3819921314716339,
      "learning_rate": 0.0015,
      "loss": 1.4135,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.3935282528400421,
      "learning_rate": 0.0015,
      "loss": 1.4088,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.42892390489578247,
      "learning_rate": 0.0015,
      "loss": 1.4121,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.40014711022377014,
      "learning_rate": 0.0015,
      "loss": 1.4235,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.48264893889427185,
      "learning_rate": 0.0015,
      "loss": 1.405,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.4335213303565979,
      "learning_rate": 0.0015,
      "loss": 1.4061,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.4485473036766052,
      "learning_rate": 0.0015,
      "loss": 1.4045,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.3801015317440033,
      "learning_rate": 0.0015,
      "loss": 1.4107,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.5835695862770081,
      "learning_rate": 0.0015,
      "loss": 1.4223,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.4383508861064911,
      "learning_rate": 0.0015,
      "loss": 1.4068,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.3987447917461395,
      "learning_rate": 0.0015,
      "loss": 1.4073,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.4258471727371216,
      "learning_rate": 0.0015,
      "loss": 1.4054,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.4014463424682617,
      "learning_rate": 0.0015,
      "loss": 1.3986,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.4350442588329315,
      "learning_rate": 0.0015,
      "loss": 1.4018,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.44266244769096375,
      "learning_rate": 0.0015,
      "loss": 1.4064,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.4371784031391144,
      "learning_rate": 0.0015,
      "loss": 1.3977,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.5309934020042419,
      "learning_rate": 0.0015,
      "loss": 1.4059,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.39169225096702576,
      "learning_rate": 0.0015,
      "loss": 1.3954,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.4109102189540863,
      "learning_rate": 0.0015,
      "loss": 1.3929,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.41619110107421875,
      "learning_rate": 0.0015,
      "loss": 1.395,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.38766422867774963,
      "learning_rate": 0.0015,
      "loss": 1.3884,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.4916270971298218,
      "learning_rate": 0.0015,
      "loss": 1.3973,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.5352882146835327,
      "learning_rate": 0.0015,
      "loss": 1.4078,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.4140731394290924,
      "learning_rate": 0.0015,
      "loss": 1.3921,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.41106846928596497,
      "learning_rate": 0.0015,
      "loss": 1.3766,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.5240281224250793,
      "learning_rate": 0.0015,
      "loss": 1.3942,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.48931968212127686,
      "learning_rate": 0.0015,
      "loss": 1.3958,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.40602126717567444,
      "learning_rate": 0.0015,
      "loss": 1.3791,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.36141782999038696,
      "learning_rate": 0.0015,
      "loss": 1.3915,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.550554096698761,
      "learning_rate": 0.0015,
      "loss": 1.3864,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.4383642077445984,
      "learning_rate": 0.0015,
      "loss": 1.3855,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.4155581295490265,
      "learning_rate": 0.0015,
      "loss": 1.3821,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.5224567651748657,
      "learning_rate": 0.0015,
      "loss": 1.3723,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.40350115299224854,
      "learning_rate": 0.0015,
      "loss": 1.4061,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.4014328420162201,
      "learning_rate": 0.0015,
      "loss": 1.3902,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.45540615916252136,
      "learning_rate": 0.0015,
      "loss": 1.3805,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.39433974027633667,
      "learning_rate": 0.0015,
      "loss": 1.3777,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.3825417459011078,
      "learning_rate": 0.0015,
      "loss": 1.3742,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.4290764331817627,
      "learning_rate": 0.0015,
      "loss": 1.3864,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.3883923590183258,
      "learning_rate": 0.0015,
      "loss": 1.3689,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.43312767148017883,
      "learning_rate": 0.0015,
      "loss": 1.3621,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.4548530876636505,
      "learning_rate": 0.0015,
      "loss": 1.3653,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.5046688914299011,
      "learning_rate": 0.0015,
      "loss": 1.3775,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4298846125602722,
      "learning_rate": 0.0015,
      "loss": 1.3738,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.4356420636177063,
      "learning_rate": 0.0015,
      "loss": 1.3683,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.4656219184398651,
      "learning_rate": 0.0015,
      "loss": 1.3709,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.3807707130908966,
      "learning_rate": 0.0015,
      "loss": 1.369,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.4016205370426178,
      "learning_rate": 0.0015,
      "loss": 1.3717,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.3906775116920471,
      "learning_rate": 0.0015,
      "loss": 1.3662,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.4661619961261749,
      "learning_rate": 0.0015,
      "loss": 1.3614,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.42744624614715576,
      "learning_rate": 0.0015,
      "loss": 1.3732,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.3593474328517914,
      "learning_rate": 0.0015,
      "loss": 1.3596,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.36683934926986694,
      "learning_rate": 0.0015,
      "loss": 1.37,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.47187283635139465,
      "learning_rate": 0.0015,
      "loss": 1.3627,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.36955729126930237,
      "learning_rate": 0.0015,
      "loss": 1.367,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.4172874689102173,
      "learning_rate": 0.0015,
      "loss": 1.3564,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.42587655782699585,
      "learning_rate": 0.0015,
      "loss": 1.3608,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.3867819905281067,
      "learning_rate": 0.0015,
      "loss": 1.3616,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.4664280414581299,
      "learning_rate": 0.0015,
      "loss": 1.3648,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.43253934383392334,
      "learning_rate": 0.0015,
      "loss": 1.3612,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.5740087628364563,
      "learning_rate": 0.0015,
      "loss": 1.3582,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.4363726079463959,
      "learning_rate": 0.0015,
      "loss": 1.3555,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.38443291187286377,
      "learning_rate": 0.0015,
      "loss": 1.3625,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.45384156703948975,
      "learning_rate": 0.0015,
      "loss": 1.3583,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.36326706409454346,
      "learning_rate": 0.0015,
      "loss": 1.3557,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.37362417578697205,
      "learning_rate": 0.0015,
      "loss": 1.3612,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.36792299151420593,
      "learning_rate": 0.0015,
      "loss": 1.3639,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.39010557532310486,
      "learning_rate": 0.0015,
      "loss": 1.3571,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.4335007071495056,
      "learning_rate": 0.0015,
      "loss": 1.3581,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.4066050350666046,
      "learning_rate": 0.0015,
      "loss": 1.344,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.39231425523757935,
      "learning_rate": 0.0015,
      "loss": 1.3571,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.47960364818573,
      "learning_rate": 0.0015,
      "loss": 1.3528,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.4317481219768524,
      "learning_rate": 0.0015,
      "loss": 1.3449,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.3781212568283081,
      "learning_rate": 0.0015,
      "loss": 1.3504,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.38963305950164795,
      "learning_rate": 0.0015,
      "loss": 1.3496,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.4155701696872711,
      "learning_rate": 0.0015,
      "loss": 1.3472,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.38998645544052124,
      "learning_rate": 0.0015,
      "loss": 1.342,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.40188562870025635,
      "learning_rate": 0.0015,
      "loss": 1.3501,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.38737237453460693,
      "learning_rate": 0.0015,
      "loss": 1.3491,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.40319275856018066,
      "learning_rate": 0.0015,
      "loss": 1.3301,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.39979711174964905,
      "learning_rate": 0.0015,
      "loss": 1.3415,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.4038510322570801,
      "learning_rate": 0.0015,
      "loss": 1.3488,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.5032958984375,
      "learning_rate": 0.0015,
      "loss": 1.3563,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.432725727558136,
      "learning_rate": 0.0015,
      "loss": 1.3524,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.4258490204811096,
      "learning_rate": 0.0015,
      "loss": 1.3465,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.370568186044693,
      "learning_rate": 0.0015,
      "loss": 1.3436,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.3846338093280792,
      "learning_rate": 0.0015,
      "loss": 1.3406,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.38603997230529785,
      "learning_rate": 0.0015,
      "loss": 1.3419,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.4352133870124817,
      "learning_rate": 0.0015,
      "loss": 1.3495,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.45890146493911743,
      "learning_rate": 0.0015,
      "loss": 1.3247,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.45596638321876526,
      "learning_rate": 0.0015,
      "loss": 1.3293,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.4347798526287079,
      "learning_rate": 0.0015,
      "loss": 1.3335,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.39894723892211914,
      "learning_rate": 0.0015,
      "loss": 1.3373,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.4256788194179535,
      "learning_rate": 0.0015,
      "loss": 1.3404,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.40725162625312805,
      "learning_rate": 0.0015,
      "loss": 1.326,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.3742321729660034,
      "learning_rate": 0.0015,
      "loss": 1.3286,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.4294082224369049,
      "learning_rate": 0.0015,
      "loss": 1.3371,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.38926073908805847,
      "learning_rate": 0.0015,
      "loss": 1.3369,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.3792055547237396,
      "learning_rate": 0.0015,
      "loss": 1.3318,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.3738500475883484,
      "learning_rate": 0.0015,
      "loss": 1.3283,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.3772472143173218,
      "learning_rate": 0.0015,
      "loss": 1.33,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.35470521450042725,
      "learning_rate": 0.0015,
      "loss": 1.3417,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.4564476013183594,
      "learning_rate": 0.0015,
      "loss": 1.3269,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.42535507678985596,
      "learning_rate": 0.0015,
      "loss": 1.3218,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.4032561779022217,
      "learning_rate": 0.0015,
      "loss": 1.3237,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.39573147892951965,
      "learning_rate": 0.0015,
      "loss": 1.3367,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.369488924741745,
      "learning_rate": 0.0015,
      "loss": 1.3198,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.3686826527118683,
      "learning_rate": 0.0015,
      "loss": 1.3266,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.4304807484149933,
      "learning_rate": 0.0015,
      "loss": 1.3289,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.4005890190601349,
      "learning_rate": 0.0015,
      "loss": 1.3159,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.4065329134464264,
      "learning_rate": 0.0015,
      "loss": 1.323,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.37223920226097107,
      "learning_rate": 0.0015,
      "loss": 1.3209,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.357913076877594,
      "learning_rate": 0.0015,
      "loss": 1.3307,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.4006459712982178,
      "learning_rate": 0.0015,
      "loss": 1.3211,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.37744468450546265,
      "learning_rate": 0.0015,
      "loss": 1.3126,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.3618320822715759,
      "learning_rate": 0.0015,
      "loss": 1.3311,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.38559070229530334,
      "learning_rate": 0.0015,
      "loss": 1.3167,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.4073399007320404,
      "learning_rate": 0.0015,
      "loss": 1.3192,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.3995159864425659,
      "learning_rate": 0.0015,
      "loss": 1.3211,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.40881553292274475,
      "learning_rate": 0.0015,
      "loss": 1.3339,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.4172428548336029,
      "learning_rate": 0.0015,
      "loss": 1.3081,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.38847237825393677,
      "learning_rate": 0.0015,
      "loss": 1.3214,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.3866778314113617,
      "learning_rate": 0.0015,
      "loss": 1.3199,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.3850679397583008,
      "learning_rate": 0.0015,
      "loss": 1.3249,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.415362149477005,
      "learning_rate": 0.0015,
      "loss": 1.318,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.41383853554725647,
      "learning_rate": 0.0015,
      "loss": 1.3173,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.39325055480003357,
      "learning_rate": 0.0015,
      "loss": 1.3133,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.4082551598548889,
      "learning_rate": 0.0015,
      "loss": 1.3242,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.3817073106765747,
      "learning_rate": 0.0015,
      "loss": 1.3168,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.3971477746963501,
      "learning_rate": 0.0015,
      "loss": 1.3096,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.3851396441459656,
      "learning_rate": 0.0015,
      "loss": 1.2907,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.36164918541908264,
      "learning_rate": 0.0015,
      "loss": 1.3179,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.3581262230873108,
      "learning_rate": 0.0015,
      "loss": 1.3071,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.4046621322631836,
      "learning_rate": 0.0015,
      "loss": 1.3231,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.45322754979133606,
      "learning_rate": 0.0015,
      "loss": 1.3222,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.47892409563064575,
      "learning_rate": 0.0015,
      "loss": 1.3055,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.3846244812011719,
      "learning_rate": 0.0015,
      "loss": 1.3126,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.36716458201408386,
      "learning_rate": 0.0015,
      "loss": 1.3135,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.37936967611312866,
      "learning_rate": 0.0015,
      "loss": 1.3194,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.36913326382637024,
      "learning_rate": 0.0015,
      "loss": 1.3094,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.40605461597442627,
      "learning_rate": 0.0015,
      "loss": 1.3109,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.3609521985054016,
      "learning_rate": 0.0015,
      "loss": 1.304,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.3889778256416321,
      "learning_rate": 0.0015,
      "loss": 1.3039,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.4098104238510132,
      "learning_rate": 0.0015,
      "loss": 1.3144,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.37357568740844727,
      "learning_rate": 0.0015,
      "loss": 1.3078,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.39619654417037964,
      "learning_rate": 0.0015,
      "loss": 1.3039,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.4084376394748688,
      "learning_rate": 0.0015,
      "loss": 1.3099,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.4111672043800354,
      "learning_rate": 0.0015,
      "loss": 1.3097,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.4097122550010681,
      "learning_rate": 0.0015,
      "loss": 1.2917,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.4278287887573242,
      "learning_rate": 0.0015,
      "loss": 1.3061,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.3902808725833893,
      "learning_rate": 0.0015,
      "loss": 1.3039,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.4041774868965149,
      "learning_rate": 0.0015,
      "loss": 1.3011,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.43103742599487305,
      "learning_rate": 0.0015,
      "loss": 1.2968,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.39118364453315735,
      "learning_rate": 0.0015,
      "loss": 1.3109,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.3834078013896942,
      "learning_rate": 0.0015,
      "loss": 1.3154,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.37766724824905396,
      "learning_rate": 0.0015,
      "loss": 1.3072,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.37089774012565613,
      "learning_rate": 0.0015,
      "loss": 1.2862,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.39285215735435486,
      "learning_rate": 0.0015,
      "loss": 1.3097,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.4295338988304138,
      "learning_rate": 0.0015,
      "loss": 1.3058,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.38859865069389343,
      "learning_rate": 0.0015,
      "loss": 1.3126,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.40371015667915344,
      "learning_rate": 0.0015,
      "loss": 1.3021,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.3651468753814697,
      "learning_rate": 0.0015,
      "loss": 1.2968,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.36767125129699707,
      "learning_rate": 0.0015,
      "loss": 1.2971,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.4188666045665741,
      "learning_rate": 0.0015,
      "loss": 1.2942,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.481037437915802,
      "learning_rate": 0.0015,
      "loss": 1.2977,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.4443666636943817,
      "learning_rate": 0.0015,
      "loss": 1.3077,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.36201488971710205,
      "learning_rate": 0.0015,
      "loss": 1.2909,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.3435400426387787,
      "learning_rate": 0.0015,
      "loss": 1.2924,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.38419613242149353,
      "learning_rate": 0.0015,
      "loss": 1.2967,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.4411642253398895,
      "learning_rate": 0.0015,
      "loss": 1.2929,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.3975316882133484,
      "learning_rate": 0.0015,
      "loss": 1.2966,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.3638233244419098,
      "learning_rate": 0.0015,
      "loss": 1.3035,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.3786154091358185,
      "learning_rate": 0.0015,
      "loss": 1.2938,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.3656991422176361,
      "learning_rate": 0.0015,
      "loss": 1.2899,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.5419597029685974,
      "learning_rate": 0.0015,
      "loss": 1.2922,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.4152900278568268,
      "learning_rate": 0.0015,
      "loss": 1.3027,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.3799019455909729,
      "learning_rate": 0.0015,
      "loss": 1.3026,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.3744471073150635,
      "learning_rate": 0.0015,
      "loss": 1.2933,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.3584083020687103,
      "learning_rate": 0.0015,
      "loss": 1.2744,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.375611811876297,
      "learning_rate": 0.0015,
      "loss": 1.2936,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.3821758031845093,
      "learning_rate": 0.0015,
      "loss": 1.2904,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.3981541097164154,
      "learning_rate": 0.0015,
      "loss": 1.2987,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.3971460461616516,
      "learning_rate": 0.0015,
      "loss": 1.2851,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.3641221225261688,
      "learning_rate": 0.0015,
      "loss": 1.3026,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.3688564896583557,
      "learning_rate": 0.0015,
      "loss": 1.2854,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.4649963080883026,
      "learning_rate": 0.0015,
      "loss": 1.293,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.43506956100463867,
      "learning_rate": 0.0015,
      "loss": 1.2983,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.3600989282131195,
      "learning_rate": 0.0015,
      "loss": 1.2861,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.35615450143814087,
      "learning_rate": 0.0015,
      "loss": 1.2795,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.3604353368282318,
      "learning_rate": 0.0015,
      "loss": 1.284,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.37614142894744873,
      "learning_rate": 0.0015,
      "loss": 1.2985,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.3707175850868225,
      "learning_rate": 0.0015,
      "loss": 1.2803,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.40577250719070435,
      "learning_rate": 0.0015,
      "loss": 1.2921,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.38110923767089844,
      "learning_rate": 0.0015,
      "loss": 1.2858,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.3657989203929901,
      "learning_rate": 0.0015,
      "loss": 1.2752,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.37295961380004883,
      "learning_rate": 0.0015,
      "loss": 1.2743,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.5426948070526123,
      "learning_rate": 0.0015,
      "loss": 1.2808,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.398045152425766,
      "learning_rate": 0.0015,
      "loss": 1.2877,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.37474942207336426,
      "learning_rate": 0.0015,
      "loss": 1.2795,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.38445577025413513,
      "learning_rate": 0.0015,
      "loss": 1.2844,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.4166354835033417,
      "learning_rate": 0.0015,
      "loss": 1.2782,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.3484591543674469,
      "learning_rate": 0.0015,
      "loss": 1.2835,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.3584078252315521,
      "learning_rate": 0.0015,
      "loss": 1.2837,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.4460738003253937,
      "learning_rate": 0.0015,
      "loss": 1.2736,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.37689462304115295,
      "learning_rate": 0.0015,
      "loss": 1.2851,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.3761577904224396,
      "learning_rate": 0.0015,
      "loss": 1.276,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.35418757796287537,
      "learning_rate": 0.0015,
      "loss": 1.2882,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.4295203387737274,
      "learning_rate": 0.0015,
      "loss": 1.2772,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.37253305315971375,
      "learning_rate": 0.0015,
      "loss": 1.2799,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.36368170380592346,
      "learning_rate": 0.0015,
      "loss": 1.2734,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.37631523609161377,
      "learning_rate": 0.0015,
      "loss": 1.2793,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.36368027329444885,
      "learning_rate": 0.0015,
      "loss": 1.282,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.4057527184486389,
      "learning_rate": 0.0015,
      "loss": 1.2864,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.34539783000946045,
      "learning_rate": 0.0015,
      "loss": 1.2814,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.5057688355445862,
      "learning_rate": 0.0015,
      "loss": 1.2873,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.430195689201355,
      "learning_rate": 0.0015,
      "loss": 1.2817,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.38058942556381226,
      "learning_rate": 0.0015,
      "loss": 1.2782,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.4111412763595581,
      "learning_rate": 0.0015,
      "loss": 1.27,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.38134413957595825,
      "learning_rate": 0.0015,
      "loss": 1.2816,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.37636449933052063,
      "learning_rate": 0.0015,
      "loss": 1.2669,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.3791978657245636,
      "learning_rate": 0.0015,
      "loss": 1.2718,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.37329211831092834,
      "learning_rate": 0.0015,
      "loss": 1.2758,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.37318155169487,
      "learning_rate": 0.0015,
      "loss": 1.2715,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.3652787506580353,
      "learning_rate": 0.0015,
      "loss": 1.2693,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.3359285295009613,
      "learning_rate": 0.0015,
      "loss": 1.2671,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.3559111952781677,
      "learning_rate": 0.0015,
      "loss": 1.2723,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.3659558892250061,
      "learning_rate": 0.0015,
      "loss": 1.2766,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.3589211404323578,
      "learning_rate": 0.0015,
      "loss": 1.2669,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.39184415340423584,
      "learning_rate": 0.0015,
      "loss": 1.2767,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.47252312302589417,
      "learning_rate": 0.0015,
      "loss": 1.2736,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.3448552191257477,
      "learning_rate": 0.0015,
      "loss": 1.2692,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.39452487230300903,
      "learning_rate": 0.0015,
      "loss": 1.2731,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.4135473072528839,
      "learning_rate": 0.0015,
      "loss": 1.2796,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.358565092086792,
      "learning_rate": 0.0015,
      "loss": 1.268,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.39048680663108826,
      "learning_rate": 0.0015,
      "loss": 1.2626,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.3444206416606903,
      "learning_rate": 0.0015,
      "loss": 1.2733,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.3811909556388855,
      "learning_rate": 0.0015,
      "loss": 1.2723,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.35479670763015747,
      "learning_rate": 0.0015,
      "loss": 1.2697,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.38825368881225586,
      "learning_rate": 0.0015,
      "loss": 1.2671,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.3594970405101776,
      "learning_rate": 0.0015,
      "loss": 1.2745,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.35297471284866333,
      "learning_rate": 0.0015,
      "loss": 1.2685,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.43033990263938904,
      "learning_rate": 0.0015,
      "loss": 1.2673,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.33862191438674927,
      "learning_rate": 0.0015,
      "loss": 1.2839,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.3441391587257385,
      "learning_rate": 0.0015,
      "loss": 1.2593,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.3779299855232239,
      "learning_rate": 0.0015,
      "loss": 1.2654,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.3549061119556427,
      "learning_rate": 0.0015,
      "loss": 1.2625,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.3566148281097412,
      "learning_rate": 0.0015,
      "loss": 1.2677,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.3833557367324829,
      "learning_rate": 0.0015,
      "loss": 1.2799,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.36007896065711975,
      "learning_rate": 0.0015,
      "loss": 1.2666,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.39347776770591736,
      "learning_rate": 0.0015,
      "loss": 1.2693,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.40295612812042236,
      "learning_rate": 0.0015,
      "loss": 1.2634,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.3890196979045868,
      "learning_rate": 0.0015,
      "loss": 1.2742,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.35265377163887024,
      "learning_rate": 0.0015,
      "loss": 1.258,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.373200923204422,
      "learning_rate": 0.0015,
      "loss": 1.2676,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.3729031980037689,
      "learning_rate": 0.0015,
      "loss": 1.2732,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.35276147723197937,
      "learning_rate": 0.0015,
      "loss": 1.2612,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.4667707085609436,
      "learning_rate": 0.0015,
      "loss": 1.2615,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.3366076648235321,
      "learning_rate": 0.0015,
      "loss": 1.2664,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.40432390570640564,
      "learning_rate": 0.0015,
      "loss": 1.2498,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.3696410655975342,
      "learning_rate": 0.0015,
      "loss": 1.2675,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.36451849341392517,
      "learning_rate": 0.0015,
      "loss": 1.268,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.4523102343082428,
      "learning_rate": 0.0015,
      "loss": 1.2688,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.3984118103981018,
      "learning_rate": 0.0015,
      "loss": 1.2572,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.3475066125392914,
      "learning_rate": 0.0015,
      "loss": 1.2389,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.4492590129375458,
      "learning_rate": 0.0015,
      "loss": 1.2604,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.398572713136673,
      "learning_rate": 0.0015,
      "loss": 1.2507,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.3685084581375122,
      "learning_rate": 0.0015,
      "loss": 1.269,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.4924905598163605,
      "learning_rate": 0.0015,
      "loss": 1.2708,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.38425061106681824,
      "learning_rate": 0.0015,
      "loss": 1.2492,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.3547668755054474,
      "learning_rate": 0.0015,
      "loss": 1.264,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.3492022156715393,
      "learning_rate": 0.0015,
      "loss": 1.2669,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.35016942024230957,
      "learning_rate": 0.0015,
      "loss": 1.2664,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.3903471827507019,
      "learning_rate": 0.0015,
      "loss": 1.2573,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.37766721844673157,
      "learning_rate": 0.0015,
      "loss": 1.2571,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.35377028584480286,
      "learning_rate": 0.0015,
      "loss": 1.2515,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.47878357768058777,
      "learning_rate": 0.0015,
      "loss": 1.2506,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.3595276176929474,
      "learning_rate": 0.0015,
      "loss": 1.2587,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.37340301275253296,
      "learning_rate": 0.0015,
      "loss": 1.2588,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.4304148554801941,
      "learning_rate": 0.0015,
      "loss": 1.2574,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.3435005843639374,
      "learning_rate": 0.0015,
      "loss": 1.253,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.36592572927474976,
      "learning_rate": 0.0015,
      "loss": 1.2501,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.371332585811615,
      "learning_rate": 0.0015,
      "loss": 1.2721,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.3589191734790802,
      "learning_rate": 0.0015,
      "loss": 1.2573,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.3282475769519806,
      "learning_rate": 0.0015,
      "loss": 1.2555,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.34423360228538513,
      "learning_rate": 0.0015,
      "loss": 1.2645,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.42229223251342773,
      "learning_rate": 0.0015,
      "loss": 1.2485,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.36169490218162537,
      "learning_rate": 0.0015,
      "loss": 1.2495,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.3373130261898041,
      "learning_rate": 0.0015,
      "loss": 1.2513,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.35851961374282837,
      "learning_rate": 0.0015,
      "loss": 1.2716,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.38497915863990784,
      "learning_rate": 0.0015,
      "loss": 1.2553,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.3687344491481781,
      "learning_rate": 0.0015,
      "loss": 1.2483,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.3644031286239624,
      "learning_rate": 0.0015,
      "loss": 1.2622,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.38985204696655273,
      "learning_rate": 0.0015,
      "loss": 1.2502,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.4006396234035492,
      "learning_rate": 0.0015,
      "loss": 1.2447,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.36055195331573486,
      "learning_rate": 0.0015,
      "loss": 1.263,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.3749941289424896,
      "learning_rate": 0.0015,
      "loss": 1.2526,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.3642890751361847,
      "learning_rate": 0.0015,
      "loss": 1.2617,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.37513530254364014,
      "learning_rate": 0.0015,
      "loss": 1.2548,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.41637057065963745,
      "learning_rate": 0.0015,
      "loss": 1.2468,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.4796326160430908,
      "learning_rate": 0.0015,
      "loss": 1.2583,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.35974326729774475,
      "learning_rate": 0.0015,
      "loss": 1.2535,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.37178534269332886,
      "learning_rate": 0.0015,
      "loss": 1.2599,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.349430650472641,
      "learning_rate": 0.0015,
      "loss": 1.2439,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.395673543214798,
      "learning_rate": 0.0015,
      "loss": 1.2508,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.40750932693481445,
      "learning_rate": 0.0015,
      "loss": 1.2449,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.3674182891845703,
      "learning_rate": 0.0015,
      "loss": 1.2476,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.3393429219722748,
      "learning_rate": 0.0015,
      "loss": 1.2561,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.33547094464302063,
      "learning_rate": 0.0015,
      "loss": 1.2501,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.36183106899261475,
      "learning_rate": 0.0015,
      "loss": 1.2452,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.3869287371635437,
      "learning_rate": 0.0015,
      "loss": 1.2315,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.36629679799079895,
      "learning_rate": 0.0015,
      "loss": 1.2587,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.349203884601593,
      "learning_rate": 0.0015,
      "loss": 1.2467,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.3906978666782379,
      "learning_rate": 0.0015,
      "loss": 1.2572,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.4080146253108978,
      "learning_rate": 0.0015,
      "loss": 1.2486,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.45458248257637024,
      "learning_rate": 0.0015,
      "loss": 1.2591,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.39543619751930237,
      "learning_rate": 0.0015,
      "loss": 1.2524,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.4424404799938202,
      "learning_rate": 0.0015,
      "loss": 1.2367,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.3873385488986969,
      "learning_rate": 0.0015,
      "loss": 1.2489,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.33514106273651123,
      "learning_rate": 0.0015,
      "loss": 1.2461,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.38259291648864746,
      "learning_rate": 0.0015,
      "loss": 1.2454,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.3605073094367981,
      "learning_rate": 0.0015,
      "loss": 1.2585,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.40182405710220337,
      "learning_rate": 0.0015,
      "loss": 1.2523,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.4059754014015198,
      "learning_rate": 0.0015,
      "loss": 1.2488,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.3528928756713867,
      "learning_rate": 0.0015,
      "loss": 1.2469,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.4284350574016571,
      "learning_rate": 0.0015,
      "loss": 1.2564,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.3745574951171875,
      "learning_rate": 0.0015,
      "loss": 1.2419,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.3614684045314789,
      "learning_rate": 0.0015,
      "loss": 1.2527,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.4084944427013397,
      "learning_rate": 0.0015,
      "loss": 1.2394,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.3417372405529022,
      "learning_rate": 0.0015,
      "loss": 1.248,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.35009732842445374,
      "learning_rate": 0.0015,
      "loss": 1.2431,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.37051039934158325,
      "learning_rate": 0.0015,
      "loss": 1.2475,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.44351595640182495,
      "learning_rate": 0.0015,
      "loss": 1.2384,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.3997798264026642,
      "learning_rate": 0.0015,
      "loss": 1.2459,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.4093949794769287,
      "learning_rate": 0.0015,
      "loss": 1.2505,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.457400381565094,
      "learning_rate": 0.0015,
      "loss": 1.2443,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.35332190990448,
      "learning_rate": 0.0015,
      "loss": 1.2468,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.3771302103996277,
      "learning_rate": 0.0015,
      "loss": 1.2499,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.47953101992607117,
      "learning_rate": 0.0015,
      "loss": 1.2372,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.3351249098777771,
      "learning_rate": 0.0015,
      "loss": 1.2373,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.35240134596824646,
      "learning_rate": 0.0015,
      "loss": 1.2459,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.35376760363578796,
      "learning_rate": 0.0015,
      "loss": 1.2396,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.36500561237335205,
      "learning_rate": 0.0015,
      "loss": 1.2361,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.45256397128105164,
      "learning_rate": 0.0015,
      "loss": 1.2449,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.3955960273742676,
      "learning_rate": 0.0015,
      "loss": 1.2435,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.40095025300979614,
      "learning_rate": 0.0015,
      "loss": 1.2355,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.3567625880241394,
      "learning_rate": 0.0015,
      "loss": 1.2386,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.36974769830703735,
      "learning_rate": 0.0015,
      "loss": 1.2374,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.3549600839614868,
      "learning_rate": 0.0015,
      "loss": 1.2426,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.34809744358062744,
      "learning_rate": 0.0015,
      "loss": 1.236,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.3890848755836487,
      "learning_rate": 0.0015,
      "loss": 1.2304,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.3576195538043976,
      "learning_rate": 0.0015,
      "loss": 1.2281,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.4001421630382538,
      "learning_rate": 0.0015,
      "loss": 1.2365,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.36096686124801636,
      "learning_rate": 0.0015,
      "loss": 1.24,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.42172369360923767,
      "learning_rate": 0.0015,
      "loss": 1.2386,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.3547919690608978,
      "learning_rate": 0.0015,
      "loss": 1.241,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.3332633674144745,
      "learning_rate": 0.0015,
      "loss": 1.2467,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.37135013937950134,
      "learning_rate": 0.0015,
      "loss": 1.2368,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.4235054850578308,
      "learning_rate": 0.0015,
      "loss": 1.2446,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.36324357986450195,
      "learning_rate": 0.0015,
      "loss": 1.2353,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.5144487023353577,
      "learning_rate": 0.0015,
      "loss": 1.2329,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.3375086784362793,
      "learning_rate": 0.0015,
      "loss": 1.2295,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.3670271337032318,
      "learning_rate": 0.0015,
      "loss": 1.2448,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.3790842890739441,
      "learning_rate": 0.0015,
      "loss": 1.2506,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.36140161752700806,
      "learning_rate": 0.0015,
      "loss": 1.2344,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.40632885694503784,
      "learning_rate": 0.0015,
      "loss": 1.2327,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.40224945545196533,
      "learning_rate": 0.0015,
      "loss": 1.2375,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.3521060347557068,
      "learning_rate": 0.0015,
      "loss": 1.2327,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.3723829686641693,
      "learning_rate": 0.0015,
      "loss": 1.2371,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.3640119731426239,
      "learning_rate": 0.0015,
      "loss": 1.2431,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.41383010149002075,
      "learning_rate": 0.0015,
      "loss": 1.2391,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.3435527980327606,
      "learning_rate": 0.0015,
      "loss": 1.2479,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.32942721247673035,
      "learning_rate": 0.0015,
      "loss": 1.2407,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.47213542461395264,
      "learning_rate": 0.0015,
      "loss": 1.239,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.34918421506881714,
      "learning_rate": 0.0015,
      "loss": 1.2427,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.34842443466186523,
      "learning_rate": 0.0015,
      "loss": 1.244,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.3522525727748871,
      "learning_rate": 0.0015,
      "loss": 1.2239,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.3676183223724365,
      "learning_rate": 0.0015,
      "loss": 1.2302,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.3557111322879791,
      "learning_rate": 0.0015,
      "loss": 1.2345,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.37497225403785706,
      "learning_rate": 0.0015,
      "loss": 1.2355,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.35429346561431885,
      "learning_rate": 0.0015,
      "loss": 1.2347,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.3666691780090332,
      "learning_rate": 0.0015,
      "loss": 1.2334,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.34536102414131165,
      "learning_rate": 0.0015,
      "loss": 1.2097,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.3289570212364197,
      "learning_rate": 0.0015,
      "loss": 1.2217,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.39452025294303894,
      "learning_rate": 0.0015,
      "loss": 1.2228,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.3545337915420532,
      "learning_rate": 0.0015,
      "loss": 1.2295,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.3574298024177551,
      "learning_rate": 0.0015,
      "loss": 1.2271,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.3354833126068115,
      "learning_rate": 0.0015,
      "loss": 1.2293,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.35068947076797485,
      "learning_rate": 0.0015,
      "loss": 1.2348,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.38195034861564636,
      "learning_rate": 0.0015,
      "loss": 1.2494,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.3787460923194885,
      "learning_rate": 0.0015,
      "loss": 1.2297,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.38402631878852844,
      "learning_rate": 0.0015,
      "loss": 1.243,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.40364325046539307,
      "learning_rate": 0.0015,
      "loss": 1.2472,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.36070314049720764,
      "learning_rate": 0.0015,
      "loss": 1.2321,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.4334326684474945,
      "learning_rate": 0.0015,
      "loss": 1.2288,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.3435908257961273,
      "learning_rate": 0.0015,
      "loss": 1.2362,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.3393274247646332,
      "learning_rate": 0.0015,
      "loss": 1.2285,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.3442111015319824,
      "learning_rate": 0.0015,
      "loss": 1.23,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.3644740581512451,
      "learning_rate": 0.0015,
      "loss": 1.2248,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.3478515148162842,
      "learning_rate": 0.0015,
      "loss": 1.2182,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.47216373682022095,
      "learning_rate": 0.0015,
      "loss": 1.2252,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.3889217674732208,
      "learning_rate": 0.0015,
      "loss": 1.2331,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.3746896982192993,
      "learning_rate": 0.0015,
      "loss": 1.2218,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.337329238653183,
      "learning_rate": 0.0015,
      "loss": 1.2327,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.45176035165786743,
      "learning_rate": 0.0015,
      "loss": 1.2321,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.34400928020477295,
      "learning_rate": 0.0015,
      "loss": 1.216,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.4225354790687561,
      "learning_rate": 0.0015,
      "loss": 1.2325,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.3400496244430542,
      "learning_rate": 0.0015,
      "loss": 1.2411,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.3938896358013153,
      "learning_rate": 0.0015,
      "loss": 1.2299,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.3652991056442261,
      "learning_rate": 0.0015,
      "loss": 1.2299,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.38011661171913147,
      "learning_rate": 0.0015,
      "loss": 1.2129,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.34210604429244995,
      "learning_rate": 0.0015,
      "loss": 1.2329,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.352566123008728,
      "learning_rate": 0.0015,
      "loss": 1.2176,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.34896865487098694,
      "learning_rate": 0.0015,
      "loss": 1.233,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.4462093710899353,
      "learning_rate": 0.0015,
      "loss": 1.2256,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.35512223839759827,
      "learning_rate": 0.0015,
      "loss": 1.2257,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.3445747196674347,
      "learning_rate": 0.0015,
      "loss": 1.2241,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.3473874628543854,
      "learning_rate": 0.0015,
      "loss": 1.2186,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.35191580653190613,
      "learning_rate": 0.0015,
      "loss": 1.2311,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.33353883028030396,
      "learning_rate": 0.0015,
      "loss": 1.2265,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.44887515902519226,
      "learning_rate": 0.0015,
      "loss": 1.2262,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.39968711137771606,
      "learning_rate": 0.0015,
      "loss": 1.2253,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.3577297627925873,
      "learning_rate": 0.0015,
      "loss": 1.2358,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.4742908179759979,
      "learning_rate": 0.0015,
      "loss": 1.23,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.4291151463985443,
      "learning_rate": 0.0014834368975312174,
      "loss": 1.2079,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.38173770904541016,
      "learning_rate": 0.0014629899726345957,
      "loss": 1.2299,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.37623050808906555,
      "learning_rate": 0.0014428248775471316,
      "loss": 1.2305,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.38231006264686584,
      "learning_rate": 0.00142293772767289,
      "loss": 1.2213,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.3671601116657257,
      "learning_rate": 0.001403324691959192,
      "loss": 1.2156,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.3565635085105896,
      "learning_rate": 0.0013839819921586025,
      "loss": 1.2229,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.3567785918712616,
      "learning_rate": 0.0013649059021010894,
      "loss": 1.2105,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.4190845489501953,
      "learning_rate": 0.0013460927469762154,
      "loss": 1.2122,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.3603142201900482,
      "learning_rate": 0.0013275389026252255,
      "loss": 1.2185,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.3586927652359009,
      "learning_rate": 0.0013092407948428887,
      "loss": 1.2087,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.3322678804397583,
      "learning_rate": 0.001291194898688966,
      "loss": 1.2146,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.3324128985404968,
      "learning_rate": 0.001273397737809166,
      "loss": 1.2128,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.34206292033195496,
      "learning_rate": 0.001255845883765463,
      "loss": 1.21,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.33811911940574646,
      "learning_rate": 0.001238535955375642,
      "loss": 1.2004,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.357466459274292,
      "learning_rate": 0.0012214646180619506,
      "loss": 1.2015,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.34239062666893005,
      "learning_rate": 0.001204628583208727,
      "loss": 1.1962,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.34912315011024475,
      "learning_rate": 0.0011880246075288827,
      "loss": 1.2038,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.3250739872455597,
      "learning_rate": 0.001171649492439115,
      "loss": 1.1965,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.335197776556015,
      "learning_rate": 0.0011555000834437364,
      "loss": 1.1989,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.3212112486362457,
      "learning_rate": 0.0011395732695269908,
      "loss": 1.1966,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.31406673789024353,
      "learning_rate": 0.0011238659825537505,
      "loss": 1.1773,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.37119054794311523,
      "learning_rate": 0.0011083751966784717,
      "loss": 1.18,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.4086954593658447,
      "learning_rate": 0.0010930979277622953,
      "loss": 1.1948,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.40815266966819763,
      "learning_rate": 0.0010780312327981854,
      "loss": 1.1939,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.35664618015289307,
      "learning_rate": 0.0010631722093439888,
      "loss": 1.1888,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.3347793519496918,
      "learning_rate": 0.00104851799496331,
      "loss": 1.1786,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.33713704347610474,
      "learning_rate": 0.0010340657666740914,
      "loss": 1.1852,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.40812811255455017,
      "learning_rate": 0.0010198127404047975,
      "loss": 1.1685,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.32857853174209595,
      "learning_rate": 0.0010057561704580897,
      "loss": 1.1727,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.36213618516921997,
      "learning_rate": 0.0009918933489818985,
      "loss": 1.1856,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.3340655267238617,
      "learning_rate": 0.0009782216054477827,
      "loss": 1.178,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.36028996109962463,
      "learning_rate": 0.0009647383061364801,
      "loss": 1.1836,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.32589006423950195,
      "learning_rate": 0.0009514408536305495,
      "loss": 1.1763,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.36935076117515564,
      "learning_rate": 0.0009383266863140042,
      "loss": 1.1913,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.35536494851112366,
      "learning_rate": 0.000925393277878844,
      "loss": 1.1876,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.3353438079357147,
      "learning_rate": 0.0009126381368383879,
      "loss": 1.1738,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.3449837267398834,
      "learning_rate": 0.0009000588060473156,
      "loss": 1.1637,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.32303208112716675,
      "learning_rate": 0.0008876528622283235,
      "loss": 1.1715,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.3335954248905182,
      "learning_rate": 0.0008754179155053053,
      "loss": 1.1685,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.3782043755054474,
      "learning_rate": 0.0008633516089429683,
      "loss": 1.1681,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.35206338763237,
      "learning_rate": 0.0008514516180927928,
      "loss": 1.1643,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.3493340313434601,
      "learning_rate": 0.0008397156505452524,
      "loss": 1.1596,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.34636741876602173,
      "learning_rate": 0.0008281414454882051,
      "loss": 1.1601,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.3398618996143341,
      "learning_rate": 0.0008167267732713704,
      "loss": 1.1666,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.36243852972984314,
      "learning_rate": 0.0008054694349768117,
      "loss": 1.1518,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.34618276357650757,
      "learning_rate": 0.0007943672619953359,
      "loss": 1.163,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.3229001462459564,
      "learning_rate": 0.0007834181156087356,
      "loss": 1.1538,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.3970085084438324,
      "learning_rate": 0.0007726198865777852,
      "loss": 1.1541,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.34208858013153076,
      "learning_rate": 0.0007619704947359191,
      "loss": 1.1508,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.31160768866539,
      "learning_rate": 0.0007514678885885087,
      "loss": 1.1521,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.3307741582393646,
      "learning_rate": 0.0007411100449176633,
      "loss": 1.1528,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.32905980944633484,
      "learning_rate": 0.0007308949683924791,
      "loss": 1.1547,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.3278239667415619,
      "learning_rate": 0.000720820691184658,
      "loss": 1.147,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.32193559408187866,
      "learning_rate": 0.0007108852725894269,
      "loss": 1.1469,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.34123092889785767,
      "learning_rate": 0.000701086798651681,
      "loss": 1.1485,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.3501998484134674,
      "learning_rate": 0.0006914233817972798,
      "loss": 1.1398,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.3438262939453125,
      "learning_rate": 0.0006818931604694261,
      "loss": 1.1436,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.3316311836242676,
      "learning_rate": 0.0006724942987700563,
      "loss": 1.1487,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.3469674289226532,
      "learning_rate": 0.0006632249861061732,
      "loss": 1.151,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.34641310572624207,
      "learning_rate": 0.0006540834368410549,
      "loss": 1.1452,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.33597779273986816,
      "learning_rate": 0.0006450678899502701,
      "loss": 1.1438,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.3230700194835663,
      "learning_rate": 0.0006361766086824345,
      "loss": 1.1433,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.32162338495254517,
      "learning_rate": 0.000627407880224645,
      "loss": 1.1462,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.32030555605888367,
      "learning_rate": 0.0006187600153725225,
      "loss": 1.1339,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.3984506130218506,
      "learning_rate": 0.0006102313482048055,
      "loss": 1.1385,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.34766748547554016,
      "learning_rate": 0.0006018202357624274,
      "loss": 1.1386,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.3280739486217499,
      "learning_rate": 0.0005935250577320168,
      "loss": 1.1369,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.35947754979133606,
      "learning_rate": 0.0005853442161337618,
      "loss": 1.1279,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.3208995759487152,
      "learning_rate": 0.0005772761350135759,
      "loss": 1.1348,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.34295305609703064,
      "learning_rate": 0.0005693192601395058,
      "loss": 1.1285,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.3107556402683258,
      "learning_rate": 0.000561472058702326,
      "loss": 1.1269,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.31452247500419617,
      "learning_rate": 0.000553733019020258,
      "loss": 1.1349,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.3466354012489319,
      "learning_rate": 0.0005461006502477612,
      "loss": 1.1244,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.3526834547519684,
      "learning_rate": 0.0005385734820883369,
      "loss": 1.1238,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.33901646733283997,
      "learning_rate": 0.0005311500645112907,
      "loss": 1.1459,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.3271549642086029,
      "learning_rate": 0.0005238289674723993,
      "loss": 1.1249,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.330404132604599,
      "learning_rate": 0.0005166087806384274,
      "loss": 1.1319,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.32270577549934387,
      "learning_rate": 0.0005094881131154418,
      "loss": 1.1339,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.34502968192100525,
      "learning_rate": 0.0005024655931808696,
      "loss": 1.1294,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.32282140851020813,
      "learning_rate": 0.0004955398680192508,
      "loss": 1.124,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.33826205134391785,
      "learning_rate": 0.000488709603461632,
      "loss": 1.1189,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.31764915585517883,
      "learning_rate": 0.000481973483728553,
      "loss": 1.1212,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.3235630989074707,
      "learning_rate": 0.0004753302111765748,
      "loss": 1.1206,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.3392946422100067,
      "learning_rate": 0.0004687785060483032,
      "loss": 1.1296,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.3291850984096527,
      "learning_rate": 0.0004623171062258558,
      "loss": 1.103,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.34219229221343994,
      "learning_rate": 0.0004559447669877288,
      "loss": 1.1152,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.32017433643341064,
      "learning_rate": 0.00044966026076901413,
      "loss": 1.1238,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.41779911518096924,
      "learning_rate": 0.00044346237692492177,
      "loss": 1.1272,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.37291568517684937,
      "learning_rate": 0.0004373499214975615,
      "loss": 1.111,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.4117240309715271,
      "learning_rate": 0.0004313217169859396,
      "loss": 1.1179,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.3707019090652466,
      "learning_rate": 0.0004253766021191256,
      "loss": 1.1227,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.3181743621826172,
      "learning_rate": 0.00041951343163254497,
      "loss": 1.1184,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.33088555932044983,
      "learning_rate": 0.00041373107604735626,
      "loss": 1.1144,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.3327450752258301,
      "learning_rate": 0.0004080284214528687,
      "loss": 1.1118,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3445628881454468,
      "learning_rate": 0.0004024043692919589,
      "loss": 1.124,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.36757397651672363,
      "learning_rate": 0.0003968578361494449,
      "loss": 1.1146,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.3530139923095703,
      "learning_rate": 0.000391387753543378,
      "loss": 1.1237,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.33693066239356995,
      "learning_rate": 0.00038599306771921023,
      "loss": 1.109,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.3698440492153168,
      "learning_rate": 0.0003806727394468004,
      "loss": 1.1047,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.3456558883190155,
      "learning_rate": 0.0003754257438202162,
      "loss": 1.1114,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.3305872082710266,
      "learning_rate": 0.0003702510700602974,
      "loss": 1.1189,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.33361759781837463,
      "learning_rate": 0.0003651477213199393,
      "loss": 1.0986,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.32957959175109863,
      "learning_rate": 0.000360114714492061,
      "loss": 1.1006,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.3591114580631256,
      "learning_rate": 0.0003551510800202195,
      "loss": 1.1035,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.39091235399246216,
      "learning_rate": 0.0003502558617118353,
      "loss": 1.1091,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.32055342197418213,
      "learning_rate": 0.0003454281165539914,
      "loss": 1.121,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.35132747888565063,
      "learning_rate": 0.00034066691453177176,
      "loss": 1.1166,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.31827104091644287,
      "learning_rate": 0.0003359713384491037,
      "loss": 1.1138,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.3282644748687744,
      "learning_rate": 0.00033134048375206944,
      "loss": 1.1073,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.32072553038597107,
      "learning_rate": 0.0003267734583546536,
      "loss": 1.1065,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.36000901460647583,
      "learning_rate": 0.00032226938246689157,
      "loss": 1.1056,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.33699488639831543,
      "learning_rate": 0.0003178273884253874,
      "loss": 1.1058,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.33975517749786377,
      "learning_rate": 0.0003134466205261674,
      "loss": 1.1156,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.3478587567806244,
      "learning_rate": 0.0003091262348598378,
      "loss": 1.1191,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.32292914390563965,
      "learning_rate": 0.0003048653991490141,
      "loss": 1.0999,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.32401716709136963,
      "learning_rate": 0.00030066329258799187,
      "loss": 1.0977,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.34817883372306824,
      "learning_rate": 0.0002965191056846266,
      "loss": 1.1029,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.3278037905693054,
      "learning_rate": 0.000292432040104394,
      "loss": 1.0939,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.3233949542045593,
      "learning_rate": 0.00028840130851659853,
      "loss": 1.0947,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.32060080766677856,
      "learning_rate": 0.0002844261344427028,
      "loss": 1.1021,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.33821195363998413,
      "learning_rate": 0.0002805057521067471,
      "loss": 1.1042,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.33443766832351685,
      "learning_rate": 0.00027663940628783017,
      "loss": 1.0892,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.34301623702049255,
      "learning_rate": 0.00027282635217462393,
      "loss": 1.1006,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.3341234028339386,
      "learning_rate": 0.0002690658552218937,
      "loss": 1.1055,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.31503307819366455,
      "learning_rate": 0.00026535719100899516,
      "loss": 1.0884,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.33604514598846436,
      "learning_rate": 0.00026169964510032245,
      "loss": 1.0909,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.3207562565803528,
      "learning_rate": 0.00025809251290767984,
      "loss": 1.0814,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.3265933394432068,
      "learning_rate": 0.00025453509955454957,
      "loss": 1.0837,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.3195817768573761,
      "learning_rate": 0.00025102671974223175,
      "loss": 1.0861,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.34826865792274475,
      "learning_rate": 0.00024756669761782815,
      "loss": 1.0951,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.3280653655529022,
      "learning_rate": 0.0002441543666440464,
      "loss": 1.0878,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.3230268657207489,
      "learning_rate": 0.00024078906947079878,
      "loss": 1.094,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.3275749385356903,
      "learning_rate": 0.00023747015780857005,
      "loss": 1.0986,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.3295617401599884,
      "learning_rate": 0.00023419699230353144,
      "loss": 1.0968,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.3250296115875244,
      "learning_rate": 0.00023096894241437586,
      "loss": 1.1005,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.3130958676338196,
      "learning_rate": 0.00022778538629085056,
      "loss": 1.0857,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.32801106572151184,
      "learning_rate": 0.00022464571065396427,
      "loss": 1.0861,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.3705176115036011,
      "learning_rate": 0.00022154931067784521,
      "loss": 1.085,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.34427809715270996,
      "learning_rate": 0.00021849558987322782,
      "loss": 1.082,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.3411599397659302,
      "learning_rate": 0.0002154839599725452,
      "loss": 1.0857,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.3404884934425354,
      "learning_rate": 0.00021251384081660544,
      "loss": 1.0935,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.31883886456489563,
      "learning_rate": 0.0002095846602428303,
      "loss": 1.0889,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.3307037949562073,
      "learning_rate": 0.00020669585397503358,
      "loss": 1.0802,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.3246137201786041,
      "learning_rate": 0.0002038468655147195,
      "loss": 1.0874,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.32273244857788086,
      "learning_rate": 0.00020103714603387894,
      "loss": 1.0986,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.31461453437805176,
      "learning_rate": 0.00019826615426926338,
      "loss": 1.0736,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.34340304136276245,
      "learning_rate": 0.00019553335641811625,
      "loss": 1.0947,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.3382534086704254,
      "learning_rate": 0.0001928382260353415,
      "loss": 1.0853,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.3437410891056061,
      "learning_rate": 0.00019018024393208902,
      "loss": 1.102,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.32129502296447754,
      "learning_rate": 0.00018755889807573872,
      "loss": 1.0771,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.32240381836891174,
      "learning_rate": 0.00018497368349126262,
      "loss": 1.0863,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.3478579819202423,
      "learning_rate": 0.00018242410216394648,
      "loss": 1.0964,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.3518509268760681,
      "learning_rate": 0.0001799096629434529,
      "loss": 1.0721,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.32270342111587524,
      "learning_rate": 0.00017742988144920578,
      "loss": 1.0828,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.323770135641098,
      "learning_rate": 0.00017498427997707976,
      "loss": 1.0796,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.33255451917648315,
      "learning_rate": 0.00017257238740737548,
      "loss": 1.0837,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.35608723759651184,
      "learning_rate": 0.00017019373911406307,
      "loss": 1.0903,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.31979310512542725,
      "learning_rate": 0.000167847876875277,
      "loss": 1.0904,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.35860174894332886,
      "learning_rate": 0.00016553434878504428,
      "loss": 1.078,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.3339579701423645,
      "learning_rate": 0.00016325270916622947,
      "loss": 1.074,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.33170056343078613,
      "learning_rate": 0.00016100251848467966,
      "loss": 1.0821,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.3272804021835327,
      "learning_rate": 0.0001587833432645528,
      "loss": 1.0775,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.3446336090564728,
      "learning_rate": 0.00015659475600481292,
      "loss": 1.0926,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.3152468502521515,
      "learning_rate": 0.00015443633509687688,
      "loss": 1.0805,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.3231417238712311,
      "learning_rate": 0.00015230766474339536,
      "loss": 1.0813,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.32221952080726624,
      "learning_rate": 0.00015020833487815416,
      "loss": 1.0871,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.3186984360218048,
      "learning_rate": 0.0001481379410870792,
      "loss": 1.079,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.3271889090538025,
      "learning_rate": 0.00014609608453033013,
      "loss": 1.0644,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.34038275480270386,
      "learning_rate": 0.00014408237186546807,
      "loss": 1.072,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.3195663094520569,
      "learning_rate": 0.00014209641517168273,
      "loss": 1.0609,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.3180614709854126,
      "learning_rate": 0.00014013783187506265,
      "loss": 1.0764,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.33012938499450684,
      "learning_rate": 0.00013820624467489697,
      "loss": 1.0924,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.3188948631286621,
      "learning_rate": 0.00013630128147099213,
      "loss": 1.0883,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.3357279896736145,
      "learning_rate": 0.00013442257529199068,
      "loss": 1.0711,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.32860061526298523,
      "learning_rate": 0.00013256976422467803,
      "loss": 1.0801,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.33959388732910156,
      "learning_rate": 0.00013074249134426366,
      "loss": 1.0766,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.323219895362854,
      "learning_rate": 0.0001289404046456233,
      "loss": 1.0863,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.32049325108528137,
      "learning_rate": 0.0001271631569754887,
      "loss": 1.0798,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.3425719141960144,
      "learning_rate": 0.0001254104059655723,
      "loss": 1.0843,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.31871697306632996,
      "learning_rate": 0.00012368181396661337,
      "loss": 1.0694,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.3283820152282715,
      "learning_rate": 0.00012197704798333364,
      "loss": 1.0705,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.31667330861091614,
      "learning_rate": 0.00012029577961028894,
      "loss": 1.0704,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.36018213629722595,
      "learning_rate": 0.00011863768496860542,
      "loss": 1.0803,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.3193598985671997,
      "learning_rate": 0.00011700244464358777,
      "loss": 1.0763,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.33096808195114136,
      "learning_rate": 0.00011538974362318715,
      "loss": 1.0784,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.3247699737548828,
      "learning_rate": 0.00011379927123731737,
      "loss": 1.0745,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.33818915486335754,
      "learning_rate": 0.0001122307210980077,
      "loss": 1.0732,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.3458991050720215,
      "learning_rate": 0.00011068379104038026,
      "loss": 1.0837,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.35434862971305847,
      "learning_rate": 0.00010915818306444116,
      "loss": 1.0648,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.3178817331790924,
      "learning_rate": 0.00010765360327767384,
      "loss": 1.0742,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.3386595547199249,
      "learning_rate": 0.00010616976183842376,
      "loss": 1.078,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.333050400018692,
      "learning_rate": 0.00010470637290006365,
      "loss": 1.0806,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.3198642432689667,
      "learning_rate": 0.00010326315455592764,
      "loss": 1.0683,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.3254936933517456,
      "learning_rate": 0.0001018398287850053,
      "loss": 1.0654,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.3687814176082611,
      "learning_rate": 0.00010043612139838357,
      "loss": 1.0796,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.34864723682403564,
      "learning_rate": 9.905176198642719e-05,
      "loss": 1.0716,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.3522048890590668,
      "learning_rate": 9.76864838666871e-05,
      "loss": 1.0719,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.3265816867351532,
      "learning_rate": 9.634002403252676e-05,
      "loss": 1.0686,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.3169366419315338,
      "learning_rate": 9.501212310245681e-05,
      "loss": 1.0684,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.33043038845062256,
      "learning_rate": 9.370252527016777e-05,
      "loss": 1.0774,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.3303966820240021,
      "learning_rate": 9.241097825525163e-05,
      "loss": 1.0683,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.33609864115715027,
      "learning_rate": 9.113723325460276e-05,
      "loss": 1.0735,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.3414512574672699,
      "learning_rate": 8.988104489448849e-05,
      "loss": 1.0685,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.31755390763282776,
      "learning_rate": 8.864217118328042e-05,
      "loss": 1.0803,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.3173295557498932,
      "learning_rate": 8.742037346483729e-05,
      "loss": 1.0733,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.3303743302822113,
      "learning_rate": 8.62154163725303e-05,
      "loss": 1.0794,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.3222305178642273,
      "learning_rate": 8.502706778390219e-05,
      "loss": 1.0764,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.3168421685695648,
      "learning_rate": 8.38550987759513e-05,
      "loss": 1.0713,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.3277452290058136,
      "learning_rate": 8.269928358103191e-05,
      "loss": 1.0841,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.31893619894981384,
      "learning_rate": 8.155939954336243e-05,
      "loss": 1.0776,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.3293255865573883,
      "learning_rate": 8.043522707613312e-05,
      "loss": 1.0716,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.3259051740169525,
      "learning_rate": 7.932654961920486e-05,
      "loss": 1.058,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.34142494201660156,
      "learning_rate": 7.823315359739135e-05,
      "loss": 1.0615,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.31675177812576294,
      "learning_rate": 7.715482837931577e-05,
      "loss": 1.0818,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.32128244638442993,
      "learning_rate": 7.6091366236835e-05,
      "loss": 1.0615,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.33591708540916443,
      "learning_rate": 7.504256230502289e-05,
      "loss": 1.0801,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.3197396397590637,
      "learning_rate": 7.400821454270524e-05,
      "loss": 1.0713,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.32733380794525146,
      "learning_rate": 7.29881236935386e-05,
      "loss": 1.0613,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.3151637017726898,
      "learning_rate": 7.198209324762562e-05,
      "loss": 1.0643,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.3231916129589081,
      "learning_rate": 7.098992940365946e-05,
      "loss": 1.0613,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.3338322937488556,
      "learning_rate": 7.001144103159e-05,
      "loss": 1.0692,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.3360345661640167,
      "learning_rate": 6.904643963580461e-05,
      "loss": 1.0743,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.3391522169113159,
      "learning_rate": 6.809473931881644e-05,
      "loss": 1.0661,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.32648056745529175,
      "learning_rate": 6.71561567454532e-05,
      "loss": 1.0661,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.3194596767425537,
      "learning_rate": 6.623051110753948e-05,
      "loss": 1.0827,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.31810277700424194,
      "learning_rate": 6.531762408906607e-05,
      "loss": 1.0726,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.3305910527706146,
      "learning_rate": 6.441731983183912e-05,
      "loss": 1.0709,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.32806795835494995,
      "learning_rate": 6.352942490160292e-05,
      "loss": 1.0623,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.333575040102005,
      "learning_rate": 6.265376825462966e-05,
      "loss": 1.0708,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.35335201025009155,
      "learning_rate": 6.179018120476945e-05,
      "loss": 1.0659,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.3264673948287964,
      "learning_rate": 6.0938497390954946e-05,
      "loss": 1.0669,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.3392159938812256,
      "learning_rate": 6.009855274515339e-05,
      "loss": 1.0623,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.3224976360797882,
      "learning_rate": 5.9270185460760735e-05,
      "loss": 1.0693,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.33346402645111084,
      "learning_rate": 5.8453235961431225e-05,
      "loss": 1.0586,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.32544925808906555,
      "learning_rate": 5.764754687033678e-05,
      "loss": 1.0688,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.33190402388572693,
      "learning_rate": 5.6852962979849836e-05,
      "loss": 1.0658,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.3238281011581421,
      "learning_rate": 5.6069331221644284e-05,
      "loss": 1.0782,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.32378917932510376,
      "learning_rate": 5.529650063720842e-05,
      "loss": 1.0722,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.3402920663356781,
      "learning_rate": 5.453432234876445e-05,
      "loss": 1.066,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.32455769181251526,
      "learning_rate": 5.37826495305886e-05,
      "loss": 1.0591,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.3707723319530487,
      "learning_rate": 5.304133738072674e-05,
      "loss": 1.0771,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.3264203369617462,
      "learning_rate": 5.2310243093099814e-05,
      "loss": 1.0685,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.343057245016098,
      "learning_rate": 5.158922582999368e-05,
      "loss": 1.0724,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.32586669921875,
      "learning_rate": 5.087814669492819e-05,
      "loss": 1.0629,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.33987531065940857,
      "learning_rate": 5.017686870590028e-05,
      "loss": 1.0668,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.32024064660072327,
      "learning_rate": 4.948525676899577e-05,
      "loss": 1.0642,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.33926844596862793,
      "learning_rate": 4.880317765236493e-05,
      "loss": 1.0714,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.36164024472236633,
      "learning_rate": 4.8130499960556755e-05,
      "loss": 1.0544,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.3245364725589752,
      "learning_rate": 4.746709410920699e-05,
      "loss": 1.0571,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.3439021408557892,
      "learning_rate": 4.681283230007507e-05,
      "loss": 1.0616,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.3390875458717346,
      "learning_rate": 4.616758849642509e-05,
      "loss": 1.0684,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.3170855641365051,
      "learning_rate": 4.553123839874615e-05,
      "loss": 1.076,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.324824720621109,
      "learning_rate": 4.490365942080736e-05,
      "loss": 1.0691,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.3236071467399597,
      "learning_rate": 4.428473066604285e-05,
      "loss": 1.0678,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.33034569025039673,
      "learning_rate": 4.367433290426233e-05,
      "loss": 1.0675,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.3257054388523102,
      "learning_rate": 4.3072348548682595e-05,
      "loss": 1.0606,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.31744372844696045,
      "learning_rate": 4.247866163327575e-05,
      "loss": 1.0726,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.33041128516197205,
      "learning_rate": 4.1893157790429404e-05,
      "loss": 1.0634,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.3342299461364746,
      "learning_rate": 4.1315724228915066e-05,
      "loss": 1.0609,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.32655617594718933,
      "learning_rate": 4.074624971216005e-05,
      "loss": 1.0516,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.32039201259613037,
      "learning_rate": 4.018462453681889e-05,
      "loss": 1.0596,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.33058446645736694,
      "learning_rate": 3.963074051164014e-05,
      "loss": 1.0648,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.32546499371528625,
      "learning_rate": 3.908449093662446e-05,
      "loss": 1.0598,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.3228848874568939,
      "learning_rate": 3.854577058246998e-05,
      "loss": 1.0557,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.33862391114234924,
      "learning_rate": 3.801447567030094e-05,
      "loss": 1.076,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.3227675259113312,
      "learning_rate": 3.7490503851675777e-05,
      "loss": 1.0605,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.3217218816280365,
      "learning_rate": 3.6973754188870806e-05,
      "loss": 1.0786,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.3212469220161438,
      "learning_rate": 3.6464127135435536e-05,
      "loss": 1.066,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.3378359377384186,
      "learning_rate": 3.596152451701616e-05,
      "loss": 1.0651,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.31813666224479675,
      "learning_rate": 3.5465849512443226e-05,
      "loss": 1.0633,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.32426491379737854,
      "learning_rate": 3.4977006635080086e-05,
      "loss": 1.064,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.34550321102142334,
      "learning_rate": 3.449490171442838e-05,
      "loss": 1.0705,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.3262031674385071,
      "learning_rate": 3.401944187798702e-05,
      "loss": 1.0669,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.31900161504745483,
      "learning_rate": 3.355053553336137e-05,
      "loss": 1.058,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.3401828706264496,
      "learning_rate": 3.308809235061882e-05,
      "loss": 1.0604,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.33225199580192566,
      "learning_rate": 3.263202324488772e-05,
      "loss": 1.0653,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.32263875007629395,
      "learning_rate": 3.218224035919609e-05,
      "loss": 1.0632,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.32119113206863403,
      "learning_rate": 3.173865704754688e-05,
      "loss": 1.0623,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.3211652338504791,
      "learning_rate": 3.130118785822657e-05,
      "loss": 1.0641,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.3391672372817993,
      "learning_rate": 3.08697485173437e-05,
      "loss": 1.0631,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.3661402761936188,
      "learning_rate": 3.0444255912594442e-05,
      "loss": 1.0639,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.35915690660476685,
      "learning_rate": 3.002462807725185e-05,
      "loss": 1.0599,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.32106488943099976,
      "learning_rate": 2.9610784174375868e-05,
      "loss": 1.0659,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.32112917304039,
      "learning_rate": 2.920264448124087e-05,
      "loss": 1.0639,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.32968273758888245,
      "learning_rate": 2.8800130373977936e-05,
      "loss": 1.0543,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.338516503572464,
      "learning_rate": 2.84031643124288e-05,
      "loss": 1.0592,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.330265074968338,
      "learning_rate": 2.8011669825208517e-05,
      "loss": 1.0776,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.3306685984134674,
      "learning_rate": 2.762557149497405e-05,
      "loss": 1.0539,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.3367357850074768,
      "learning_rate": 2.724479494389592e-05,
      "loss": 1.063,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.31615322828292847,
      "learning_rate": 2.6869266819330058e-05,
      "loss": 1.0667,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.3239525556564331,
      "learning_rate": 2.6498914779687228e-05,
      "loss": 1.0685,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.33848342299461365,
      "learning_rate": 2.6133667480497115e-05,
      "loss": 1.0673,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.32886168360710144,
      "learning_rate": 2.5773454560664597e-05,
      "loss": 1.0668,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.3179469704627991,
      "learning_rate": 2.541820662891541e-05,
      "loss": 1.0475,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.3251250982284546,
      "learning_rate": 2.5067855250428616e-05,
      "loss": 1.0579,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.3251715898513794,
      "learning_rate": 2.472233293365335e-05,
      "loss": 1.0719,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.33614784479141235,
      "learning_rate": 2.4381573117307307e-05,
      "loss": 1.0542,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.3413209021091461,
      "learning_rate": 2.4045510157554362e-05,
      "loss": 1.0709,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.3312832713127136,
      "learning_rate": 2.3714079315358985e-05,
      "loss": 1.0591,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.3249393701553345,
      "learning_rate": 2.338721674401494e-05,
      "loss": 1.0675,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.32825586199760437,
      "learning_rate": 2.30648594768459e-05,
      "loss": 1.0653,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.3269410729408264,
      "learning_rate": 2.2746945415075523e-05,
      "loss": 1.0617,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.3209409713745117,
      "learning_rate": 2.2433413315864803e-05,
      "loss": 1.0578,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.3257068395614624,
      "learning_rate": 2.2124202780514277e-05,
      "loss": 1.0625,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.3229939043521881,
      "learning_rate": 2.1819254242828815e-05,
      "loss": 1.0616,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.33057811856269836,
      "learning_rate": 2.151850895764285e-05,
      "loss": 1.0591,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.305545449256897,
      "learning_rate": 2.12219089895037e-05,
      "loss": 1.0513,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.32736852765083313,
      "learning_rate": 2.092939720151092e-05,
      "loss": 1.0479,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.3169935643672943,
      "learning_rate": 2.064091724430947e-05,
      "loss": 1.0569,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.3271348774433136,
      "learning_rate": 2.0356413545234603e-05,
      "loss": 1.0539,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.31698349118232727,
      "learning_rate": 2.0075831297606357e-05,
      "loss": 1.0645,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.3189639449119568,
      "learning_rate": 1.9799116450171627e-05,
      "loss": 1.0515,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.3284784257411957,
      "learning_rate": 1.952621569669175e-05,
      "loss": 1.0612,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.32952404022216797,
      "learning_rate": 1.9257076465673605e-05,
      "loss": 1.0593,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.32908713817596436,
      "learning_rate": 1.899164691024229e-05,
      "loss": 1.0613,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.3283434808254242,
      "learning_rate": 1.872987589815331e-05,
      "loss": 1.0548,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.33627134561538696,
      "learning_rate": 1.8471713001942538e-05,
      "loss": 1.0674,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.3174121677875519,
      "learning_rate": 1.8217108489211845e-05,
      "loss": 1.065,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.3188031315803528,
      "learning_rate": 1.7966013313048696e-05,
      "loss": 1.074,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.3233587145805359,
      "learning_rate": 1.7718379102577752e-05,
      "loss": 1.0646,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.32843634486198425,
      "learning_rate": 1.7474158153642745e-05,
      "loss": 1.0637,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.3318040668964386,
      "learning_rate": 1.7233303419616745e-05,
      "loss": 1.0532,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.343174546957016,
      "learning_rate": 1.699576850233916e-05,
      "loss": 1.0554,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.3253111243247986,
      "learning_rate": 1.6761507643177553e-05,
      "loss": 1.0696,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.3279666602611542,
      "learning_rate": 1.6530475714212752e-05,
      "loss": 1.046,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.32378527522087097,
      "learning_rate": 1.6302628209545423e-05,
      "loss": 1.0631,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.31868770718574524,
      "learning_rate": 1.6077921236722464e-05,
      "loss": 1.063,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.323849618434906,
      "learning_rate": 1.5856311508281594e-05,
      "loss": 1.0585,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.320055216550827,
      "learning_rate": 1.5637756333412454e-05,
      "loss": 1.0692,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.32591259479522705,
      "learning_rate": 1.542221360973268e-05,
      "loss": 1.0535,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.3283117413520813,
      "learning_rate": 1.5209641815177312e-05,
      "loss": 1.0675,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.9648054838180542,
      "learning_rate": 1.5e-05,
      "loss": 1.0556,
      "step": 9480
    }
  ],
  "logging_steps": 10,
  "max_steps": 9480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.037432118742016e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}