{
  "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": 1.3075357675552368,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.5213,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.180863380432129,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.8987,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8489615321159363,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.2384,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.9192500114440918,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.747,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.9643535614013672,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.2708,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 1.082902193069458,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.7644,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.1981812715530396,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3723,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.9766468405723572,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.1225,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 1.1699119806289673,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.9235,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 1.0279523134231567,
      "learning_rate": 0.0015,
      "loss": 3.7655,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.9181833863258362,
      "learning_rate": 0.0015,
      "loss": 3.6204,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.8613824248313904,
      "learning_rate": 0.0015,
      "loss": 3.5071,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 1.1336807012557983,
      "learning_rate": 0.0015,
      "loss": 3.395,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.9157042503356934,
      "learning_rate": 0.0015,
      "loss": 3.2994,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.9386192560195923,
      "learning_rate": 0.0015,
      "loss": 3.2244,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.9380351901054382,
      "learning_rate": 0.0015,
      "loss": 3.1574,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.6966649889945984,
      "learning_rate": 0.0015,
      "loss": 3.0937,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.7090355157852173,
      "learning_rate": 0.0015,
      "loss": 3.0498,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.7136836647987366,
      "learning_rate": 0.0015,
      "loss": 2.9903,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.8697147369384766,
      "learning_rate": 0.0015,
      "loss": 2.9377,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.7215303182601929,
      "learning_rate": 0.0015,
      "loss": 2.9076,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.8969533443450928,
      "learning_rate": 0.0015,
      "loss": 2.8731,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.6747064590454102,
      "learning_rate": 0.0015,
      "loss": 2.8266,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.9318302273750305,
      "learning_rate": 0.0015,
      "loss": 2.7889,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.71550452709198,
      "learning_rate": 0.0015,
      "loss": 2.7549,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.7597529292106628,
      "learning_rate": 0.0015,
      "loss": 2.7197,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.7744061946868896,
      "learning_rate": 0.0015,
      "loss": 2.6953,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.8805778622627258,
      "learning_rate": 0.0015,
      "loss": 2.6567,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.8690669536590576,
      "learning_rate": 0.0015,
      "loss": 2.6336,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 1.0534539222717285,
      "learning_rate": 0.0015,
      "loss": 2.6088,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.9530712366104126,
      "learning_rate": 0.0015,
      "loss": 2.5892,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 1.4720412492752075,
      "learning_rate": 0.0015,
      "loss": 2.5548,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.9948314428329468,
      "learning_rate": 0.0015,
      "loss": 2.5482,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.9229395985603333,
      "learning_rate": 0.0015,
      "loss": 2.5241,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 1.1244056224822998,
      "learning_rate": 0.0015,
      "loss": 2.4848,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.7936338186264038,
      "learning_rate": 0.0015,
      "loss": 2.4668,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.9502090215682983,
      "learning_rate": 0.0015,
      "loss": 2.4538,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.7836471796035767,
      "learning_rate": 0.0015,
      "loss": 2.445,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.8590785264968872,
      "learning_rate": 0.0015,
      "loss": 2.4195,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 1.0463407039642334,
      "learning_rate": 0.0015,
      "loss": 2.3997,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 1.039176344871521,
      "learning_rate": 0.0015,
      "loss": 2.3899,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.8505009412765503,
      "learning_rate": 0.0015,
      "loss": 2.3748,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.8631711006164551,
      "learning_rate": 0.0015,
      "loss": 2.3399,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.9587793946266174,
      "learning_rate": 0.0015,
      "loss": 2.326,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.7890628576278687,
      "learning_rate": 0.0015,
      "loss": 2.3267,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.763205349445343,
      "learning_rate": 0.0015,
      "loss": 2.3002,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.7925857901573181,
      "learning_rate": 0.0015,
      "loss": 2.2826,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.7465478181838989,
      "learning_rate": 0.0015,
      "loss": 2.2844,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.9607719779014587,
      "learning_rate": 0.0015,
      "loss": 2.2541,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.7803492546081543,
      "learning_rate": 0.0015,
      "loss": 2.2435,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.9835359454154968,
      "learning_rate": 0.0015,
      "loss": 2.2327,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 1.1225404739379883,
      "learning_rate": 0.0015,
      "loss": 2.2231,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.8247981667518616,
      "learning_rate": 0.0015,
      "loss": 2.2064,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.8044012188911438,
      "learning_rate": 0.0015,
      "loss": 2.1905,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.8903504610061646,
      "learning_rate": 0.0015,
      "loss": 2.1921,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.8271088600158691,
      "learning_rate": 0.0015,
      "loss": 2.1587,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.6447353363037109,
      "learning_rate": 0.0015,
      "loss": 2.1608,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 1.0156145095825195,
      "learning_rate": 0.0015,
      "loss": 2.162,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.8628018498420715,
      "learning_rate": 0.0015,
      "loss": 2.1352,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.7424734830856323,
      "learning_rate": 0.0015,
      "loss": 2.1174,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.8701589107513428,
      "learning_rate": 0.0015,
      "loss": 2.1179,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.76248699426651,
      "learning_rate": 0.0015,
      "loss": 2.1089,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.8957002758979797,
      "learning_rate": 0.0015,
      "loss": 2.0911,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.9295898675918579,
      "learning_rate": 0.0015,
      "loss": 2.1067,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.792677104473114,
      "learning_rate": 0.0015,
      "loss": 2.0931,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.877686083316803,
      "learning_rate": 0.0015,
      "loss": 2.083,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.8630133271217346,
      "learning_rate": 0.0015,
      "loss": 2.0581,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.8127005696296692,
      "learning_rate": 0.0015,
      "loss": 2.0547,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7369611263275146,
      "learning_rate": 0.0015,
      "loss": 2.0627,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.9251187443733215,
      "learning_rate": 0.0015,
      "loss": 2.0457,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.752349853515625,
      "learning_rate": 0.0015,
      "loss": 2.0381,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.7315734624862671,
      "learning_rate": 0.0015,
      "loss": 2.0194,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.926827609539032,
      "learning_rate": 0.0015,
      "loss": 2.0147,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.8080822229385376,
      "learning_rate": 0.0015,
      "loss": 2.0244,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.8982332348823547,
      "learning_rate": 0.0015,
      "loss": 2.001,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.2816462516784668,
      "learning_rate": 0.0015,
      "loss": 2.0012,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.6519502997398376,
      "learning_rate": 0.0015,
      "loss": 1.9916,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.7490759491920471,
      "learning_rate": 0.0015,
      "loss": 1.9894,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.8788653612136841,
      "learning_rate": 0.0015,
      "loss": 1.9916,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.837259829044342,
      "learning_rate": 0.0015,
      "loss": 1.9636,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.9153476357460022,
      "learning_rate": 0.0015,
      "loss": 1.9544,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.7409590482711792,
      "learning_rate": 0.0015,
      "loss": 1.9623,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.8565800786018372,
      "learning_rate": 0.0015,
      "loss": 1.9658,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.9489105939865112,
      "learning_rate": 0.0015,
      "loss": 1.9448,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.6747156381607056,
      "learning_rate": 0.0015,
      "loss": 1.9454,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.73032146692276,
      "learning_rate": 0.0015,
      "loss": 1.9453,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 1.097557544708252,
      "learning_rate": 0.0015,
      "loss": 1.9388,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.7871012091636658,
      "learning_rate": 0.0015,
      "loss": 1.9311,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.9786339998245239,
      "learning_rate": 0.0015,
      "loss": 1.9229,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 1.0217005014419556,
      "learning_rate": 0.0015,
      "loss": 1.9212,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.6946851015090942,
      "learning_rate": 0.0015,
      "loss": 1.9254,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.6087393760681152,
      "learning_rate": 0.0015,
      "loss": 1.914,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.9960092902183533,
      "learning_rate": 0.0015,
      "loss": 1.897,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.7217803001403809,
      "learning_rate": 0.0015,
      "loss": 1.9084,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.8259333372116089,
      "learning_rate": 0.0015,
      "loss": 1.8858,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.7509357333183289,
      "learning_rate": 0.0015,
      "loss": 1.8917,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 0.8337006568908691,
      "learning_rate": 0.0015,
      "loss": 1.8975,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.7414672374725342,
      "learning_rate": 0.0015,
      "loss": 1.876,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 1.2368232011795044,
      "learning_rate": 0.0015,
      "loss": 1.8823,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.7442910075187683,
      "learning_rate": 0.0015,
      "loss": 1.879,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.7759189605712891,
      "learning_rate": 0.0015,
      "loss": 1.8702,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.9012283682823181,
      "learning_rate": 0.0015,
      "loss": 1.8629,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.8135190606117249,
      "learning_rate": 0.0015,
      "loss": 1.8652,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.6859939694404602,
      "learning_rate": 0.0015,
      "loss": 1.858,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 1.2869263887405396,
      "learning_rate": 0.0015,
      "loss": 1.8613,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.7284495830535889,
      "learning_rate": 0.0015,
      "loss": 1.8531,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.8010578751564026,
      "learning_rate": 0.0015,
      "loss": 1.8417,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 1.0110236406326294,
      "learning_rate": 0.0015,
      "loss": 1.841,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.6968798637390137,
      "learning_rate": 0.0015,
      "loss": 1.8465,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.6804802417755127,
      "learning_rate": 0.0015,
      "loss": 1.8316,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.6878808736801147,
      "learning_rate": 0.0015,
      "loss": 1.8405,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.7354303002357483,
      "learning_rate": 0.0015,
      "loss": 1.8306,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.9895655512809753,
      "learning_rate": 0.0015,
      "loss": 1.8165,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.8318318724632263,
      "learning_rate": 0.0015,
      "loss": 1.8319,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.169092059135437,
      "learning_rate": 0.0015,
      "loss": 1.8156,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.9082102179527283,
      "learning_rate": 0.0015,
      "loss": 1.8278,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.8206037878990173,
      "learning_rate": 0.0015,
      "loss": 1.8203,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.6738736033439636,
      "learning_rate": 0.0015,
      "loss": 1.7983,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.6992834210395813,
      "learning_rate": 0.0015,
      "loss": 1.8003,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.9246848821640015,
      "learning_rate": 0.0015,
      "loss": 1.8136,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.7774592041969299,
      "learning_rate": 0.0015,
      "loss": 1.8062,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.6720554232597351,
      "learning_rate": 0.0015,
      "loss": 1.7889,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.6804599761962891,
      "learning_rate": 0.0015,
      "loss": 1.78,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.6456968188285828,
      "learning_rate": 0.0015,
      "loss": 1.7974,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.6627588272094727,
      "learning_rate": 0.0015,
      "loss": 1.7982,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.8834444284439087,
      "learning_rate": 0.0015,
      "loss": 1.781,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.7322362065315247,
      "learning_rate": 0.0015,
      "loss": 1.7838,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.9749155640602112,
      "learning_rate": 0.0015,
      "loss": 1.8028,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.7019423842430115,
      "learning_rate": 0.0015,
      "loss": 1.7743,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.686527669429779,
      "learning_rate": 0.0015,
      "loss": 1.7653,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.6187251806259155,
      "learning_rate": 0.0015,
      "loss": 1.7691,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.9021650552749634,
      "learning_rate": 0.0015,
      "loss": 1.7711,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 0.9765862822532654,
      "learning_rate": 0.0015,
      "loss": 1.7638,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.7598349452018738,
      "learning_rate": 0.0015,
      "loss": 1.7652,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.7436577677726746,
      "learning_rate": 0.0015,
      "loss": 1.77,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.7047667503356934,
      "learning_rate": 0.0015,
      "loss": 1.76,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 1.006347417831421,
      "learning_rate": 0.0015,
      "loss": 1.7616,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.7401061058044434,
      "learning_rate": 0.0015,
      "loss": 1.7485,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 1.2228999137878418,
      "learning_rate": 0.0015,
      "loss": 1.7608,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.6520518660545349,
      "learning_rate": 0.0015,
      "loss": 1.7426,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.8179234266281128,
      "learning_rate": 0.0015,
      "loss": 1.7431,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.6742578148841858,
      "learning_rate": 0.0015,
      "loss": 1.7486,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.6870294809341431,
      "learning_rate": 0.0015,
      "loss": 1.7452,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.7341146469116211,
      "learning_rate": 0.0015,
      "loss": 1.7434,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.6733317375183105,
      "learning_rate": 0.0015,
      "loss": 1.7385,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.6803165078163147,
      "learning_rate": 0.0015,
      "loss": 1.7352,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 1.4749171733856201,
      "learning_rate": 0.0015,
      "loss": 1.7365,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 1.347564935684204,
      "learning_rate": 0.0015,
      "loss": 1.7489,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.7148910164833069,
      "learning_rate": 0.0015,
      "loss": 1.7339,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.8039283752441406,
      "learning_rate": 0.0015,
      "loss": 1.7286,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 1.3167616128921509,
      "learning_rate": 0.0015,
      "loss": 1.7352,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 1.0010050535202026,
      "learning_rate": 0.0015,
      "loss": 1.7346,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.6768319606781006,
      "learning_rate": 0.0015,
      "loss": 1.7145,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6522865891456604,
      "learning_rate": 0.0015,
      "loss": 1.7039,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.9729596376419067,
      "learning_rate": 0.0015,
      "loss": 1.7063,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.7230920791625977,
      "learning_rate": 0.0015,
      "loss": 1.7116,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.7281738519668579,
      "learning_rate": 0.0015,
      "loss": 1.7023,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.8933695554733276,
      "learning_rate": 0.0015,
      "loss": 1.7135,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.8795140981674194,
      "learning_rate": 0.0015,
      "loss": 1.7093,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.702780544757843,
      "learning_rate": 0.0015,
      "loss": 1.7153,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.6340116858482361,
      "learning_rate": 0.0015,
      "loss": 1.7115,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6516766548156738,
      "learning_rate": 0.0015,
      "loss": 1.6822,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.7751317620277405,
      "learning_rate": 0.0015,
      "loss": 1.6931,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.8222350478172302,
      "learning_rate": 0.0015,
      "loss": 1.7007,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.9241092205047607,
      "learning_rate": 0.0015,
      "loss": 1.6929,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.758102536201477,
      "learning_rate": 0.0015,
      "loss": 1.6892,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.742407500743866,
      "learning_rate": 0.0015,
      "loss": 1.7018,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.2279940843582153,
      "learning_rate": 0.0015,
      "loss": 1.6859,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.8028030395507812,
      "learning_rate": 0.0015,
      "loss": 1.6828,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.9893059730529785,
      "learning_rate": 0.0015,
      "loss": 1.6995,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.623496949672699,
      "learning_rate": 0.0015,
      "loss": 1.6756,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.654557466506958,
      "learning_rate": 0.0015,
      "loss": 1.6716,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.6914863586425781,
      "learning_rate": 0.0015,
      "loss": 1.6772,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.68093341588974,
      "learning_rate": 0.0015,
      "loss": 1.6788,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.6601630449295044,
      "learning_rate": 0.0015,
      "loss": 1.6985,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.677162766456604,
      "learning_rate": 0.0015,
      "loss": 1.6771,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6156236529350281,
      "learning_rate": 0.0015,
      "loss": 1.6796,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.7677043676376343,
      "learning_rate": 0.0015,
      "loss": 1.6823,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.7379671335220337,
      "learning_rate": 0.0015,
      "loss": 1.6842,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.6247692108154297,
      "learning_rate": 0.0015,
      "loss": 1.6766,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.9126911759376526,
      "learning_rate": 0.0015,
      "loss": 1.6644,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.6720184087753296,
      "learning_rate": 0.0015,
      "loss": 1.6601,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.7707266211509705,
      "learning_rate": 0.0015,
      "loss": 1.6621,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.5960696935653687,
      "learning_rate": 0.0015,
      "loss": 1.6582,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.9515610933303833,
      "learning_rate": 0.0015,
      "loss": 1.6641,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.6111634969711304,
      "learning_rate": 0.0015,
      "loss": 1.6679,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.7194884419441223,
      "learning_rate": 0.0015,
      "loss": 1.6666,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6791244149208069,
      "learning_rate": 0.0015,
      "loss": 1.6587,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.8795470595359802,
      "learning_rate": 0.0015,
      "loss": 1.662,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.6828414797782898,
      "learning_rate": 0.0015,
      "loss": 1.6587,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.8992465138435364,
      "learning_rate": 0.0015,
      "loss": 1.6628,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.6946663856506348,
      "learning_rate": 0.0015,
      "loss": 1.6554,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.765633761882782,
      "learning_rate": 0.0015,
      "loss": 1.651,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.6072712540626526,
      "learning_rate": 0.0015,
      "loss": 1.6436,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.6243547201156616,
      "learning_rate": 0.0015,
      "loss": 1.6451,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.602359414100647,
      "learning_rate": 0.0015,
      "loss": 1.6494,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.6976801753044128,
      "learning_rate": 0.0015,
      "loss": 1.6617,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.7268016934394836,
      "learning_rate": 0.0015,
      "loss": 1.642,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.6199648380279541,
      "learning_rate": 0.0015,
      "loss": 1.6393,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.8665949702262878,
      "learning_rate": 0.0015,
      "loss": 1.6367,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.6769253611564636,
      "learning_rate": 0.0015,
      "loss": 1.6474,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.9666388630867004,
      "learning_rate": 0.0015,
      "loss": 1.657,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.6327275633811951,
      "learning_rate": 0.0015,
      "loss": 1.6425,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6882089376449585,
      "learning_rate": 0.0015,
      "loss": 1.639,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.7690213918685913,
      "learning_rate": 0.0015,
      "loss": 1.64,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.6751629710197449,
      "learning_rate": 0.0015,
      "loss": 1.6321,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 1.079460859298706,
      "learning_rate": 0.0015,
      "loss": 1.633,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 1.1796042919158936,
      "learning_rate": 0.0015,
      "loss": 1.6458,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.6678057312965393,
      "learning_rate": 0.0015,
      "loss": 1.6244,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.7857685089111328,
      "learning_rate": 0.0015,
      "loss": 1.6326,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.7463678121566772,
      "learning_rate": 0.0015,
      "loss": 1.6266,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6068580746650696,
      "learning_rate": 0.0015,
      "loss": 1.6228,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.6354256272315979,
      "learning_rate": 0.0015,
      "loss": 1.6263,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6060561537742615,
      "learning_rate": 0.0015,
      "loss": 1.6196,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.9400723576545715,
      "learning_rate": 0.0015,
      "loss": 1.6286,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.2904577255249023,
      "learning_rate": 0.0015,
      "loss": 1.6442,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 1.4418649673461914,
      "learning_rate": 0.0015,
      "loss": 1.6288,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 1.1155215501785278,
      "learning_rate": 0.0015,
      "loss": 1.6123,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.7112396359443665,
      "learning_rate": 0.0015,
      "loss": 1.6179,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.6585453152656555,
      "learning_rate": 0.0015,
      "loss": 1.621,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.6442099213600159,
      "learning_rate": 0.0015,
      "loss": 1.6067,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.5695006251335144,
      "learning_rate": 0.0015,
      "loss": 1.6236,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.6747961044311523,
      "learning_rate": 0.0015,
      "loss": 1.6154,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.809987485408783,
      "learning_rate": 0.0015,
      "loss": 1.6164,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.7946650385856628,
      "learning_rate": 0.0015,
      "loss": 1.6143,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.8359740376472473,
      "learning_rate": 0.0015,
      "loss": 1.6009,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.756756603717804,
      "learning_rate": 0.0015,
      "loss": 1.6346,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.9783498048782349,
      "learning_rate": 0.0015,
      "loss": 1.6222,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.6283764243125916,
      "learning_rate": 0.0015,
      "loss": 1.608,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.6477097272872925,
      "learning_rate": 0.0015,
      "loss": 1.6049,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.5955271124839783,
      "learning_rate": 0.0015,
      "loss": 1.6075,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.7311010360717773,
      "learning_rate": 0.0015,
      "loss": 1.6151,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.8355544209480286,
      "learning_rate": 0.0015,
      "loss": 1.5992,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.8351632952690125,
      "learning_rate": 0.0015,
      "loss": 1.5954,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.6420930027961731,
      "learning_rate": 0.0015,
      "loss": 1.5986,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.6042001247406006,
      "learning_rate": 0.0015,
      "loss": 1.6036,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7402383089065552,
      "learning_rate": 0.0015,
      "loss": 1.6031,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.6115971207618713,
      "learning_rate": 0.0015,
      "loss": 1.6009,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.6082727909088135,
      "learning_rate": 0.0015,
      "loss": 1.6004,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6307869553565979,
      "learning_rate": 0.0015,
      "loss": 1.6019,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.6242306232452393,
      "learning_rate": 0.0015,
      "loss": 1.6032,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.7853202223777771,
      "learning_rate": 0.0015,
      "loss": 1.5959,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.6395176649093628,
      "learning_rate": 0.0015,
      "loss": 1.5949,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.6518899202346802,
      "learning_rate": 0.0015,
      "loss": 1.6045,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.575395941734314,
      "learning_rate": 0.0015,
      "loss": 1.5894,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.6950863599777222,
      "learning_rate": 0.0015,
      "loss": 1.6033,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.952235996723175,
      "learning_rate": 0.0015,
      "loss": 1.5943,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.6923096179962158,
      "learning_rate": 0.0015,
      "loss": 1.5995,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.6995301246643066,
      "learning_rate": 0.0015,
      "loss": 1.5874,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.6522321701049805,
      "learning_rate": 0.0015,
      "loss": 1.5907,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.5644729137420654,
      "learning_rate": 0.0015,
      "loss": 1.5911,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.7484664916992188,
      "learning_rate": 0.0015,
      "loss": 1.5943,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.6378340125083923,
      "learning_rate": 0.0015,
      "loss": 1.5892,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.9534720778465271,
      "learning_rate": 0.0015,
      "loss": 1.5907,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.6464568972587585,
      "learning_rate": 0.0015,
      "loss": 1.5843,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.8279476761817932,
      "learning_rate": 0.0015,
      "loss": 1.5937,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.7225745320320129,
      "learning_rate": 0.0015,
      "loss": 1.5865,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.6332727670669556,
      "learning_rate": 0.0015,
      "loss": 1.5872,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.5766628384590149,
      "learning_rate": 0.0015,
      "loss": 1.5933,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.6092755794525146,
      "learning_rate": 0.0015,
      "loss": 1.5898,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.7969714999198914,
      "learning_rate": 0.0015,
      "loss": 1.588,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6936610341072083,
      "learning_rate": 0.0015,
      "loss": 1.5855,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.6494637727737427,
      "learning_rate": 0.0015,
      "loss": 1.5805,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 1.3193278312683105,
      "learning_rate": 0.0015,
      "loss": 1.5948,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.6159821152687073,
      "learning_rate": 0.0015,
      "loss": 1.5799,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.5923860669136047,
      "learning_rate": 0.0015,
      "loss": 1.5702,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.626798152923584,
      "learning_rate": 0.0015,
      "loss": 1.5768,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.5723259449005127,
      "learning_rate": 0.0015,
      "loss": 1.5735,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.6969932913780212,
      "learning_rate": 0.0015,
      "loss": 1.5726,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6743830442428589,
      "learning_rate": 0.0015,
      "loss": 1.569,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.6520766615867615,
      "learning_rate": 0.0015,
      "loss": 1.5784,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.6736236214637756,
      "learning_rate": 0.0015,
      "loss": 1.5783,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.5693079233169556,
      "learning_rate": 0.0015,
      "loss": 1.5557,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 1.0078039169311523,
      "learning_rate": 0.0015,
      "loss": 1.5693,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.5904132127761841,
      "learning_rate": 0.0015,
      "loss": 1.5736,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.6193687319755554,
      "learning_rate": 0.0015,
      "loss": 1.5841,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6845226287841797,
      "learning_rate": 0.0015,
      "loss": 1.5785,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.5668240785598755,
      "learning_rate": 0.0015,
      "loss": 1.5748,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.6082599759101868,
      "learning_rate": 0.0015,
      "loss": 1.5727,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.9084484577178955,
      "learning_rate": 0.0015,
      "loss": 1.5677,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.6633672714233398,
      "learning_rate": 0.0015,
      "loss": 1.5699,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6207756400108337,
      "learning_rate": 0.0015,
      "loss": 1.577,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.9366081357002258,
      "learning_rate": 0.0015,
      "loss": 1.5507,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 1.137453317642212,
      "learning_rate": 0.0015,
      "loss": 1.5553,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.5851289629936218,
      "learning_rate": 0.0015,
      "loss": 1.5616,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.5535052418708801,
      "learning_rate": 0.0015,
      "loss": 1.5642,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.968014657497406,
      "learning_rate": 0.0015,
      "loss": 1.5655,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.7300151586532593,
      "learning_rate": 0.0015,
      "loss": 1.5521,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.629642903804779,
      "learning_rate": 0.0015,
      "loss": 1.5599,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.8954982161521912,
      "learning_rate": 0.0015,
      "loss": 1.5642,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.7289522886276245,
      "learning_rate": 0.0015,
      "loss": 1.5622,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.6284463405609131,
      "learning_rate": 0.0015,
      "loss": 1.5567,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.679797887802124,
      "learning_rate": 0.0015,
      "loss": 1.5544,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.5774010419845581,
      "learning_rate": 0.0015,
      "loss": 1.5625,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.8260715007781982,
      "learning_rate": 0.0015,
      "loss": 1.5688,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.7063205242156982,
      "learning_rate": 0.0015,
      "loss": 1.5584,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.9290918111801147,
      "learning_rate": 0.0015,
      "loss": 1.539,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.5621309876441956,
      "learning_rate": 0.0015,
      "loss": 1.5439,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.7395849823951721,
      "learning_rate": 0.0015,
      "loss": 1.5643,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.6708174347877502,
      "learning_rate": 0.0015,
      "loss": 1.5452,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.6967434883117676,
      "learning_rate": 0.0015,
      "loss": 1.553,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.7203395366668701,
      "learning_rate": 0.0015,
      "loss": 1.5571,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6107951402664185,
      "learning_rate": 0.0015,
      "loss": 1.5397,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.6455133557319641,
      "learning_rate": 0.0015,
      "loss": 1.5494,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.559616208076477,
      "learning_rate": 0.0015,
      "loss": 1.5457,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.6274845600128174,
      "learning_rate": 0.0015,
      "loss": 1.5574,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.5791829228401184,
      "learning_rate": 0.0015,
      "loss": 1.547,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.7367576360702515,
      "learning_rate": 0.0015,
      "loss": 1.5377,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.5722604393959045,
      "learning_rate": 0.0015,
      "loss": 1.5517,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.6680841445922852,
      "learning_rate": 0.0015,
      "loss": 1.5399,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.7069068551063538,
      "learning_rate": 0.0015,
      "loss": 1.5465,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.5908547639846802,
      "learning_rate": 0.0015,
      "loss": 1.5464,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.5545928478240967,
      "learning_rate": 0.0015,
      "loss": 1.558,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.6596634984016418,
      "learning_rate": 0.0015,
      "loss": 1.5356,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.6853034496307373,
      "learning_rate": 0.0015,
      "loss": 1.5465,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5891876220703125,
      "learning_rate": 0.0015,
      "loss": 1.5466,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.6577898263931274,
      "learning_rate": 0.0015,
      "loss": 1.5521,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.9291559457778931,
      "learning_rate": 0.0015,
      "loss": 1.5427,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.7408413290977478,
      "learning_rate": 0.0015,
      "loss": 1.5436,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.6851930022239685,
      "learning_rate": 0.0015,
      "loss": 1.537,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 1.133357048034668,
      "learning_rate": 0.0015,
      "loss": 1.5526,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.6708566546440125,
      "learning_rate": 0.0015,
      "loss": 1.5415,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.7159783244132996,
      "learning_rate": 0.0015,
      "loss": 1.532,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.5234776139259338,
      "learning_rate": 0.0015,
      "loss": 1.516,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.5658848285675049,
      "learning_rate": 0.0015,
      "loss": 1.5417,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.5536794066429138,
      "learning_rate": 0.0015,
      "loss": 1.5261,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.5448280572891235,
      "learning_rate": 0.0015,
      "loss": 1.546,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.6722742319107056,
      "learning_rate": 0.0015,
      "loss": 1.5487,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.8077312707901001,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.5687286853790283,
      "learning_rate": 0.0015,
      "loss": 1.5384,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.5500969886779785,
      "learning_rate": 0.0015,
      "loss": 1.5391,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.5969871282577515,
      "learning_rate": 0.0015,
      "loss": 1.5442,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.6553987264633179,
      "learning_rate": 0.0015,
      "loss": 1.5301,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.6266419291496277,
      "learning_rate": 0.0015,
      "loss": 1.5368,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6171289682388306,
      "learning_rate": 0.0015,
      "loss": 1.5289,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.6962433457374573,
      "learning_rate": 0.0015,
      "loss": 1.5242,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.8192803859710693,
      "learning_rate": 0.0015,
      "loss": 1.5352,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.6736462712287903,
      "learning_rate": 0.0015,
      "loss": 1.532,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.5962381958961487,
      "learning_rate": 0.0015,
      "loss": 1.5257,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.6057121753692627,
      "learning_rate": 0.0015,
      "loss": 1.5357,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.564690113067627,
      "learning_rate": 0.0015,
      "loss": 1.5301,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.8714531660079956,
      "learning_rate": 0.0015,
      "loss": 1.512,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.5465991497039795,
      "learning_rate": 0.0015,
      "loss": 1.5288,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.6148397922515869,
      "learning_rate": 0.0015,
      "loss": 1.5248,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.6495022177696228,
      "learning_rate": 0.0015,
      "loss": 1.5226,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 1.464675784111023,
      "learning_rate": 0.0015,
      "loss": 1.5173,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.7403339743614197,
      "learning_rate": 0.0015,
      "loss": 1.54,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.6511922478675842,
      "learning_rate": 0.0015,
      "loss": 1.5349,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.5631682872772217,
      "learning_rate": 0.0015,
      "loss": 1.5281,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.6367287039756775,
      "learning_rate": 0.0015,
      "loss": 1.5106,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.6354228854179382,
      "learning_rate": 0.0015,
      "loss": 1.5302,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.6151289343833923,
      "learning_rate": 0.0015,
      "loss": 1.5257,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.7224065661430359,
      "learning_rate": 0.0015,
      "loss": 1.536,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 1.1419929265975952,
      "learning_rate": 0.0015,
      "loss": 1.5298,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.7123492956161499,
      "learning_rate": 0.0015,
      "loss": 1.5215,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.5686929225921631,
      "learning_rate": 0.0015,
      "loss": 1.5196,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 1.0632009506225586,
      "learning_rate": 0.0015,
      "loss": 1.5181,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.9212638139724731,
      "learning_rate": 0.0015,
      "loss": 1.5188,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.560971200466156,
      "learning_rate": 0.0015,
      "loss": 1.5254,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.5467254519462585,
      "learning_rate": 0.0015,
      "loss": 1.5096,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.599979817867279,
      "learning_rate": 0.0015,
      "loss": 1.5132,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.6306496858596802,
      "learning_rate": 0.0015,
      "loss": 1.5147,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.879368007183075,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.8159598112106323,
      "learning_rate": 0.0015,
      "loss": 1.518,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.6509981155395508,
      "learning_rate": 0.0015,
      "loss": 1.5275,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.6290169358253479,
      "learning_rate": 0.0015,
      "loss": 1.5173,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.6807018518447876,
      "learning_rate": 0.0015,
      "loss": 1.5122,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.7448897361755371,
      "learning_rate": 0.0015,
      "loss": 1.5149,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.5696559548377991,
      "learning_rate": 0.0015,
      "loss": 1.5254,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.6409304738044739,
      "learning_rate": 0.0015,
      "loss": 1.527,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.6279841065406799,
      "learning_rate": 0.0015,
      "loss": 1.5162,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.6621096730232239,
      "learning_rate": 0.0015,
      "loss": 1.4963,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.6872855424880981,
      "learning_rate": 0.0015,
      "loss": 1.5135,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.5671707391738892,
      "learning_rate": 0.0015,
      "loss": 1.5123,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.8312143087387085,
      "learning_rate": 0.0015,
      "loss": 1.5174,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.995048463344574,
      "learning_rate": 0.0015,
      "loss": 1.5071,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.6123533248901367,
      "learning_rate": 0.0015,
      "loss": 1.5221,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.5263500809669495,
      "learning_rate": 0.0015,
      "loss": 1.507,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.7213681936264038,
      "learning_rate": 0.0015,
      "loss": 1.5124,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.6691168546676636,
      "learning_rate": 0.0015,
      "loss": 1.5206,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.5371190309524536,
      "learning_rate": 0.0015,
      "loss": 1.5025,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.5635185837745667,
      "learning_rate": 0.0015,
      "loss": 1.4976,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.535293698310852,
      "learning_rate": 0.0015,
      "loss": 1.4994,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.48217204213142395,
      "learning_rate": 0.0015,
      "loss": 1.5166,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.5847092270851135,
      "learning_rate": 0.0015,
      "loss": 1.499,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.5289793014526367,
      "learning_rate": 0.0015,
      "loss": 1.5136,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 1.0073657035827637,
      "learning_rate": 0.0015,
      "loss": 1.5029,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.9138542413711548,
      "learning_rate": 0.0015,
      "loss": 1.4957,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.5173036456108093,
      "learning_rate": 0.0015,
      "loss": 1.4944,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 1.1745268106460571,
      "learning_rate": 0.0015,
      "loss": 1.4964,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.5849131941795349,
      "learning_rate": 0.0015,
      "loss": 1.5001,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.5363203287124634,
      "learning_rate": 0.0015,
      "loss": 1.4939,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.6402507424354553,
      "learning_rate": 0.0015,
      "loss": 1.5021,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.688439130783081,
      "learning_rate": 0.0015,
      "loss": 1.4966,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.8214023113250732,
      "learning_rate": 0.0015,
      "loss": 1.5042,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.6306218504905701,
      "learning_rate": 0.0015,
      "loss": 1.5045,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.6060662269592285,
      "learning_rate": 0.0015,
      "loss": 1.492,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.6276333928108215,
      "learning_rate": 0.0015,
      "loss": 1.5038,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.5927751064300537,
      "learning_rate": 0.0015,
      "loss": 1.4971,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.5483892560005188,
      "learning_rate": 0.0015,
      "loss": 1.5104,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.7140218019485474,
      "learning_rate": 0.0015,
      "loss": 1.4954,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.5532450079917908,
      "learning_rate": 0.0015,
      "loss": 1.4969,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.5548282265663147,
      "learning_rate": 0.0015,
      "loss": 1.4868,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5832592844963074,
      "learning_rate": 0.0015,
      "loss": 1.4982,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5230129957199097,
      "learning_rate": 0.0015,
      "loss": 1.4993,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.8107106685638428,
      "learning_rate": 0.0015,
      "loss": 1.5049,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.7404172420501709,
      "learning_rate": 0.0015,
      "loss": 1.4998,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6545730233192444,
      "learning_rate": 0.0015,
      "loss": 1.506,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.8511345386505127,
      "learning_rate": 0.0015,
      "loss": 1.5044,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.7161096334457397,
      "learning_rate": 0.0015,
      "loss": 1.4939,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.7610675692558289,
      "learning_rate": 0.0015,
      "loss": 1.489,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.7774899005889893,
      "learning_rate": 0.0015,
      "loss": 1.4954,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.7286903858184814,
      "learning_rate": 0.0015,
      "loss": 1.4849,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5786338448524475,
      "learning_rate": 0.0015,
      "loss": 1.4891,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.6982702016830444,
      "learning_rate": 0.0015,
      "loss": 1.4962,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.6745831370353699,
      "learning_rate": 0.0015,
      "loss": 1.4873,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.5979346632957458,
      "learning_rate": 0.0015,
      "loss": 1.4905,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.522210419178009,
      "learning_rate": 0.0015,
      "loss": 1.4871,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.6977109313011169,
      "learning_rate": 0.0015,
      "loss": 1.4918,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.7419961094856262,
      "learning_rate": 0.0015,
      "loss": 1.4962,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.5456410050392151,
      "learning_rate": 0.0015,
      "loss": 1.4819,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.5922154188156128,
      "learning_rate": 0.0015,
      "loss": 1.4933,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.8679487705230713,
      "learning_rate": 0.0015,
      "loss": 1.4899,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.574184238910675,
      "learning_rate": 0.0015,
      "loss": 1.4861,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.6575837135314941,
      "learning_rate": 0.0015,
      "loss": 1.4904,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.8064940571784973,
      "learning_rate": 0.0015,
      "loss": 1.4969,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5215798020362854,
      "learning_rate": 0.0015,
      "loss": 1.4875,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.7631582617759705,
      "learning_rate": 0.0015,
      "loss": 1.4789,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.720287561416626,
      "learning_rate": 0.0015,
      "loss": 1.4913,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.9794846177101135,
      "learning_rate": 0.0015,
      "loss": 1.4893,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.6552988886833191,
      "learning_rate": 0.0015,
      "loss": 1.4904,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.8640376329421997,
      "learning_rate": 0.0015,
      "loss": 1.4821,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.5411359667778015,
      "learning_rate": 0.0015,
      "loss": 1.4917,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.7179622650146484,
      "learning_rate": 0.0015,
      "loss": 1.4809,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 1.0754817724227905,
      "learning_rate": 0.0015,
      "loss": 1.4849,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.6185563206672668,
      "learning_rate": 0.0015,
      "loss": 1.4987,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.5712226033210754,
      "learning_rate": 0.0015,
      "loss": 1.4738,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.7067052125930786,
      "learning_rate": 0.0015,
      "loss": 1.4837,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.5491397380828857,
      "learning_rate": 0.0015,
      "loss": 1.4791,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.558575451374054,
      "learning_rate": 0.0015,
      "loss": 1.4818,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.6383283138275146,
      "learning_rate": 0.0015,
      "loss": 1.5,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.6881840825080872,
      "learning_rate": 0.0015,
      "loss": 1.4799,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.7439998984336853,
      "learning_rate": 0.0015,
      "loss": 1.4844,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.755620539188385,
      "learning_rate": 0.0015,
      "loss": 1.4788,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.5531437397003174,
      "learning_rate": 0.0015,
      "loss": 1.4896,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5616311430931091,
      "learning_rate": 0.0015,
      "loss": 1.4752,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.6229474544525146,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.7021309733390808,
      "learning_rate": 0.0015,
      "loss": 1.4911,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.7292089462280273,
      "learning_rate": 0.0015,
      "loss": 1.4769,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.6191797256469727,
      "learning_rate": 0.0015,
      "loss": 1.4763,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5511068105697632,
      "learning_rate": 0.0015,
      "loss": 1.4833,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.6581600904464722,
      "learning_rate": 0.0015,
      "loss": 1.464,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.6747670769691467,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.6507686376571655,
      "learning_rate": 0.0015,
      "loss": 1.4845,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.607556164264679,
      "learning_rate": 0.0015,
      "loss": 1.4824,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.6154383420944214,
      "learning_rate": 0.0015,
      "loss": 1.4718,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.6015588641166687,
      "learning_rate": 0.0015,
      "loss": 1.4541,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.6513546705245972,
      "learning_rate": 0.0015,
      "loss": 1.4762,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.6503914594650269,
      "learning_rate": 0.0015,
      "loss": 1.4657,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.5230239629745483,
      "learning_rate": 0.0015,
      "loss": 1.4827,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 1.0571868419647217,
      "learning_rate": 0.0015,
      "loss": 1.4849,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.7897891998291016,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.569659948348999,
      "learning_rate": 0.0015,
      "loss": 1.4813,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.6289198994636536,
      "learning_rate": 0.0015,
      "loss": 1.4831,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.6410478353500366,
      "learning_rate": 0.0015,
      "loss": 1.4835,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.5901730060577393,
      "learning_rate": 0.0015,
      "loss": 1.4699,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.8757472634315491,
      "learning_rate": 0.0015,
      "loss": 1.4705,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.6517742276191711,
      "learning_rate": 0.0015,
      "loss": 1.4671,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.973685622215271,
      "learning_rate": 0.0015,
      "loss": 1.4637,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.6177852749824524,
      "learning_rate": 0.0015,
      "loss": 1.4731,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.501401960849762,
      "learning_rate": 0.0015,
      "loss": 1.4743,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.8149515986442566,
      "learning_rate": 0.0015,
      "loss": 1.4726,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.6071144938468933,
      "learning_rate": 0.0015,
      "loss": 1.4676,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7226423025131226,
      "learning_rate": 0.0015,
      "loss": 1.4671,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.6534337401390076,
      "learning_rate": 0.0015,
      "loss": 1.4874,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5539942383766174,
      "learning_rate": 0.0015,
      "loss": 1.4672,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.6125468611717224,
      "learning_rate": 0.0015,
      "loss": 1.4676,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.6089324355125427,
      "learning_rate": 0.0015,
      "loss": 1.4772,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.6028252840042114,
      "learning_rate": 0.0015,
      "loss": 1.4603,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.6291878819465637,
      "learning_rate": 0.0015,
      "loss": 1.4662,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.5844051241874695,
      "learning_rate": 0.0015,
      "loss": 1.4694,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.6730778813362122,
      "learning_rate": 0.0015,
      "loss": 1.4857,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.7433723211288452,
      "learning_rate": 0.0015,
      "loss": 1.4687,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.6143871545791626,
      "learning_rate": 0.0015,
      "loss": 1.463,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.514320433139801,
      "learning_rate": 0.0015,
      "loss": 1.4734,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.5280232429504395,
      "learning_rate": 0.0015,
      "loss": 1.4622,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.9742357134819031,
      "learning_rate": 0.0015,
      "loss": 1.4572,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.6315768361091614,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.6527271270751953,
      "learning_rate": 0.0015,
      "loss": 1.4685,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.6024441719055176,
      "learning_rate": 0.0015,
      "loss": 1.4738,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.68658447265625,
      "learning_rate": 0.0015,
      "loss": 1.4693,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.5601794719696045,
      "learning_rate": 0.0015,
      "loss": 1.4581,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.5640733242034912,
      "learning_rate": 0.0015,
      "loss": 1.4681,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.5403666496276855,
      "learning_rate": 0.0015,
      "loss": 1.4651,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.6531521677970886,
      "learning_rate": 0.0015,
      "loss": 1.474,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.6115059852600098,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.8114550709724426,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.6079219579696655,
      "learning_rate": 0.0015,
      "loss": 1.4546,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.578087568283081,
      "learning_rate": 0.0015,
      "loss": 1.459,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.672719419002533,
      "learning_rate": 0.0015,
      "loss": 1.4712,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.5922591686248779,
      "learning_rate": 0.0015,
      "loss": 1.4642,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.5584644079208374,
      "learning_rate": 0.0015,
      "loss": 1.4551,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.5503262281417847,
      "learning_rate": 0.0015,
      "loss": 1.4439,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.5341739654541016,
      "learning_rate": 0.0015,
      "loss": 1.4708,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.5860607624053955,
      "learning_rate": 0.0015,
      "loss": 1.4585,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.7299894690513611,
      "learning_rate": 0.0015,
      "loss": 1.4713,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.8018727898597717,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.5662050843238831,
      "learning_rate": 0.0015,
      "loss": 1.4733,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.5539529919624329,
      "learning_rate": 0.0015,
      "loss": 1.4636,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.6735826134681702,
      "learning_rate": 0.0015,
      "loss": 1.4452,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.5980088710784912,
      "learning_rate": 0.0015,
      "loss": 1.4619,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6005414724349976,
      "learning_rate": 0.0015,
      "loss": 1.4603,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.5914118885993958,
      "learning_rate": 0.0015,
      "loss": 1.4553,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 1.06390380859375,
      "learning_rate": 0.0015,
      "loss": 1.4698,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.6869582533836365,
      "learning_rate": 0.0015,
      "loss": 1.4669,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.559586226940155,
      "learning_rate": 0.0015,
      "loss": 1.4606,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5051333904266357,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.7952111959457397,
      "learning_rate": 0.0015,
      "loss": 1.4712,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.8145313262939453,
      "learning_rate": 0.0015,
      "loss": 1.4487,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.5636731386184692,
      "learning_rate": 0.0015,
      "loss": 1.464,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.6466506719589233,
      "learning_rate": 0.0015,
      "loss": 1.4512,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.6597478985786438,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.6992036700248718,
      "learning_rate": 0.0015,
      "loss": 1.4528,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.5877173542976379,
      "learning_rate": 0.0015,
      "loss": 1.4613,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.8450697064399719,
      "learning_rate": 0.0015,
      "loss": 1.4529,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.5091837048530579,
      "learning_rate": 0.0015,
      "loss": 1.4591,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.7349790334701538,
      "learning_rate": 0.0015,
      "loss": 1.4653,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 1.0718812942504883,
      "learning_rate": 0.0015,
      "loss": 1.4561,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.5732921957969666,
      "learning_rate": 0.0015,
      "loss": 1.4563,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.5573410987854004,
      "learning_rate": 0.0015,
      "loss": 1.4617,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.6892499923706055,
      "learning_rate": 0.0015,
      "loss": 1.4465,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.5653085112571716,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.8573117256164551,
      "learning_rate": 0.0015,
      "loss": 1.4575,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.5884286165237427,
      "learning_rate": 0.0015,
      "loss": 1.4494,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.5683361887931824,
      "learning_rate": 0.0015,
      "loss": 1.4454,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.8135273456573486,
      "learning_rate": 0.0015,
      "loss": 1.4528,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.5589624643325806,
      "learning_rate": 0.0015,
      "loss": 1.4534,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.5421573519706726,
      "learning_rate": 0.0015,
      "loss": 1.4472,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.6375049352645874,
      "learning_rate": 0.0015,
      "loss": 1.4455,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5435382723808289,
      "learning_rate": 0.0015,
      "loss": 1.4458,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5885555148124695,
      "learning_rate": 0.0015,
      "loss": 1.451,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.7123895287513733,
      "learning_rate": 0.0015,
      "loss": 1.4447,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.7274113297462463,
      "learning_rate": 0.0015,
      "loss": 1.4421,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.8285456299781799,
      "learning_rate": 0.0015,
      "loss": 1.441,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.7984604239463806,
      "learning_rate": 0.0015,
      "loss": 1.4515,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.6042565107345581,
      "learning_rate": 0.0015,
      "loss": 1.4512,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.6467163562774658,
      "learning_rate": 0.0015,
      "loss": 1.4507,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.7072213292121887,
      "learning_rate": 0.0015,
      "loss": 1.4507,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.5475664734840393,
      "learning_rate": 0.0015,
      "loss": 1.4571,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.6362132430076599,
      "learning_rate": 0.0015,
      "loss": 1.4446,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 1.106300711631775,
      "learning_rate": 0.0015,
      "loss": 1.4537,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.5137573480606079,
      "learning_rate": 0.0015,
      "loss": 1.4442,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.6358821988105774,
      "learning_rate": 0.0015,
      "loss": 1.4389,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.5367628335952759,
      "learning_rate": 0.0015,
      "loss": 1.4381,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.6821813583374023,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.5292353630065918,
      "learning_rate": 0.0015,
      "loss": 1.4606,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.4966440796852112,
      "learning_rate": 0.0015,
      "loss": 1.4436,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.5308849215507507,
      "learning_rate": 0.0015,
      "loss": 1.4429,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.6342505216598511,
      "learning_rate": 0.0015,
      "loss": 1.4468,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5645837783813477,
      "learning_rate": 0.0015,
      "loss": 1.4409,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.637165904045105,
      "learning_rate": 0.0015,
      "loss": 1.4491,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.6109039783477783,
      "learning_rate": 0.0015,
      "loss": 1.4531,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.6971760392189026,
      "learning_rate": 0.0015,
      "loss": 1.447,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.6155142188072205,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.7192047238349915,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 1.393180251121521,
      "learning_rate": 0.0015,
      "loss": 1.4544,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 1.2101657390594482,
      "learning_rate": 0.0015,
      "loss": 1.4603,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.9806029200553894,
      "learning_rate": 0.0015,
      "loss": 1.4542,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 1.0299819707870483,
      "learning_rate": 0.0015,
      "loss": 1.4311,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.563425600528717,
      "learning_rate": 0.0015,
      "loss": 1.4381,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.5351040959358215,
      "learning_rate": 0.0015,
      "loss": 1.4414,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5208850502967834,
      "learning_rate": 0.0015,
      "loss": 1.4461,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.5404925346374512,
      "learning_rate": 0.0015,
      "loss": 1.4455,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.5373325943946838,
      "learning_rate": 0.0015,
      "loss": 1.4394,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.6910935044288635,
      "learning_rate": 0.0015,
      "loss": 1.4179,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5569452047348022,
      "learning_rate": 0.0015,
      "loss": 1.4294,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.7494074702262878,
      "learning_rate": 0.0015,
      "loss": 1.4305,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.5173893570899963,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.6150003671646118,
      "learning_rate": 0.0015,
      "loss": 1.4346,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.541309118270874,
      "learning_rate": 0.0015,
      "loss": 1.4406,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.523231565952301,
      "learning_rate": 0.0015,
      "loss": 1.4462,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5977320671081543,
      "learning_rate": 0.0015,
      "loss": 1.4574,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.5601499080657959,
      "learning_rate": 0.0015,
      "loss": 1.4371,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.731503963470459,
      "learning_rate": 0.0015,
      "loss": 1.4526,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.599065363407135,
      "learning_rate": 0.0015,
      "loss": 1.4543,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.5344615578651428,
      "learning_rate": 0.0015,
      "loss": 1.4414,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.5567651987075806,
      "learning_rate": 0.0015,
      "loss": 1.4368,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.681366503238678,
      "learning_rate": 0.0015,
      "loss": 1.4456,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5168014168739319,
      "learning_rate": 0.0015,
      "loss": 1.437,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.6140822768211365,
      "learning_rate": 0.0015,
      "loss": 1.4383,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.6157610416412354,
      "learning_rate": 0.0015,
      "loss": 1.4337,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.5748449563980103,
      "learning_rate": 0.0015,
      "loss": 1.4274,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 1.1245657205581665,
      "learning_rate": 0.0015,
      "loss": 1.4369,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.6548138856887817,
      "learning_rate": 0.0015,
      "loss": 1.4441,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.5102262496948242,
      "learning_rate": 0.0015,
      "loss": 1.4316,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.5147098898887634,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5202301740646362,
      "learning_rate": 0.0015,
      "loss": 1.4394,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.536027729511261,
      "learning_rate": 0.0015,
      "loss": 1.4223,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 1.0913490056991577,
      "learning_rate": 0.0015,
      "loss": 1.4456,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.5582550168037415,
      "learning_rate": 0.0015,
      "loss": 1.4539,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.5896278023719788,
      "learning_rate": 0.0015,
      "loss": 1.4339,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.9601753354072571,
      "learning_rate": 0.0015,
      "loss": 1.4385,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.634880006313324,
      "learning_rate": 0.0015,
      "loss": 1.4206,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.5148939490318298,
      "learning_rate": 0.0015,
      "loss": 1.4398,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.5344108939170837,
      "learning_rate": 0.0015,
      "loss": 1.4243,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.6601157188415527,
      "learning_rate": 0.0015,
      "loss": 1.4424,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.7422851324081421,
      "learning_rate": 0.0015,
      "loss": 1.4334,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.5776787400245667,
      "learning_rate": 0.0015,
      "loss": 1.4334,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5289221405982971,
      "learning_rate": 0.0015,
      "loss": 1.4328,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.8894216418266296,
      "learning_rate": 0.0015,
      "loss": 1.4276,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.5304521918296814,
      "learning_rate": 0.0015,
      "loss": 1.4398,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.6131283044815063,
      "learning_rate": 0.0015,
      "loss": 1.4323,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.7603117823600769,
      "learning_rate": 0.0015,
      "loss": 1.4328,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": Infinity,
      "learning_rate": 0.0015,
      "loss": 1.4344,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.5572797060012817,
      "learning_rate": 0.0015,
      "loss": 1.4415,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.6372722387313843,
      "learning_rate": 0.0015,
      "loss": 1.4372,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5281897187232971,
      "learning_rate": 0.001487560447745699,
      "loss": 1.4152,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.6531746983528137,
      "learning_rate": 0.0014670566859713624,
      "loss": 1.4363,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5641706585884094,
      "learning_rate": 0.0014468355374162303,
      "loss": 1.4428,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.6546826362609863,
      "learning_rate": 0.0014268931066862504,
      "loss": 1.4312,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5389937162399292,
      "learning_rate": 0.0014072255520794614,
      "loss": 1.4232,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.5590718388557434,
      "learning_rate": 0.0013878290848459301,
      "loss": 1.431,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.7184352874755859,
      "learning_rate": 0.0013686999684578874,
      "loss": 1.4193,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5333797335624695,
      "learning_rate": 0.001349834517889925,
      "loss": 1.4206,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.48487091064453125,
      "learning_rate": 0.001331229098909114,
      "loss": 1.4279,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.5883477926254272,
      "learning_rate": 0.0013128801273749075,
      "loss": 1.4172,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.5650054216384888,
      "learning_rate": 0.0012947840685486932,
      "loss": 1.4268,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.6160814166069031,
      "learning_rate": 0.0012769374364128628,
      "loss": 1.4233,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.5125714540481567,
      "learning_rate": 0.0012593367929992667,
      "loss": 1.417,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.6019501090049744,
      "learning_rate": 0.0012419787477269257,
      "loss": 1.4113,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.5925447344779968,
      "learning_rate": 0.0012248599567488698,
      "loss": 1.4096,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.47715866565704346,
      "learning_rate": 0.0012079771223079822,
      "loss": 1.4019,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.5875175595283508,
      "learning_rate": 0.0011913269921017202,
      "loss": 1.4112,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.7462964057922363,
      "learning_rate": 0.0011749063586555919,
      "loss": 1.4049,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.6075565218925476,
      "learning_rate": 0.001158712058705271,
      "loss": 1.41,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.8765844106674194,
      "learning_rate": 0.0011427409725872262,
      "loss": 1.4071,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.7356165051460266,
      "learning_rate": 0.00112699002363775,
      "loss": 1.387,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.47462791204452515,
      "learning_rate": 0.0011114561776002726,
      "loss": 1.389,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.9150704741477966,
      "learning_rate": 0.001096136442040843,
      "loss": 1.4043,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.7336153388023376,
      "learning_rate": 0.001081027865771668,
      "loss": 1.4043,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.5432242155075073,
      "learning_rate": 0.0010661275382825958,
      "loss": 1.3978,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.5393098592758179,
      "learning_rate": 0.0010514325891804379,
      "loss": 1.3866,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.4978395700454712,
      "learning_rate": 0.0010369401876360166,
      "loss": 1.3949,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.47768574953079224,
      "learning_rate": 0.001022647541838836,
      "loss": 1.38,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.6149685978889465,
      "learning_rate": 0.0010085518984592678,
      "loss": 1.3817,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.5218510031700134,
      "learning_rate": 0.0009946505421181513,
      "loss": 1.3933,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5168776512145996,
      "learning_rate": 0.0009809407948637044,
      "loss": 1.3879,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.6201766729354858,
      "learning_rate": 0.0009674200156556436,
      "loss": 1.3931,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.5909963846206665,
      "learning_rate": 0.0009540855998564147,
      "loss": 1.389,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.5744777917861938,
      "learning_rate": 0.000940934978729437,
      "loss": 1.4006,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.4588823914527893,
      "learning_rate": 0.0009279656189442628,
      "loss": 1.4011,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.4933320879936218,
      "learning_rate": 0.0009151750220885573,
      "loss": 1.3857,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.5646079778671265,
      "learning_rate": 0.0009025607241868057,
      "loss": 1.3734,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.6859690546989441,
      "learning_rate": 0.0008901202952256545,
      "loss": 1.3858,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.5678068399429321,
      "learning_rate": 0.0008778513386857928,
      "loss": 1.3776,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.5483217835426331,
      "learning_rate": 0.0008657514910802905,
      "loss": 1.3772,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.4671037495136261,
      "learning_rate": 0.0008538184214992943,
      "loss": 1.3731,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.5548893809318542,
      "learning_rate": 0.0008420498311610049,
      "loss": 1.3706,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.5803453922271729,
      "learning_rate": 0.0008304434529688382,
      "loss": 1.3719,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5223375558853149,
      "learning_rate": 0.0008189970510746938,
      "loss": 1.3799,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.5233261585235596,
      "learning_rate": 0.0008077084204482425,
      "loss": 1.3666,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.530835747718811,
      "learning_rate": 0.0007965753864521494,
      "loss": 1.3752,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.4769565463066101,
      "learning_rate": 0.0007855958044231527,
      "loss": 1.3673,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.6102154850959778,
      "learning_rate": 0.000774767559258917,
      "loss": 1.3712,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.5593146085739136,
      "learning_rate": 0.0007640885650105806,
      "loss": 1.3645,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.594312846660614,
      "learning_rate": 0.0007535567644809191,
      "loss": 1.3636,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5139537453651428,
      "learning_rate": 0.0007431701288280478,
      "loss": 1.3638,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.46587347984313965,
      "learning_rate": 0.0007329266571745864,
      "loss": 1.3649,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.5560813546180725,
      "learning_rate": 0.0007228243762222109,
      "loss": 1.3593,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.4940660297870636,
      "learning_rate": 0.0007128613398715179,
      "loss": 1.3623,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.49038827419281006,
      "learning_rate": 0.0007030356288471288,
      "loss": 1.3619,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.6032112240791321,
      "learning_rate": 0.0006933453503279619,
      "loss": 1.3517,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.5079308152198792,
      "learning_rate": 0.0006837886375825994,
      "loss": 1.3612,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.5179148316383362,
      "learning_rate": 0.0006743636496096813,
      "loss": 1.3622,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.4907872974872589,
      "learning_rate": 0.0006650685707832559,
      "loss": 1.3621,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.6771067976951599,
      "learning_rate": 0.0006559016105030176,
      "loss": 1.3606,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.6246157288551331,
      "learning_rate": 0.000646861002849367,
      "loss": 1.3595,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.4946833550930023,
      "learning_rate": 0.0006379450062432248,
      "loss": 1.3566,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.503964900970459,
      "learning_rate": 0.0006291519031105347,
      "loss": 1.3611,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.48558634519577026,
      "learning_rate": 0.00062047999955139,
      "loss": 1.3469,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.6461641788482666,
      "learning_rate": 0.000611927625013722,
      "loss": 1.3491,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.5477721095085144,
      "learning_rate": 0.0006034931319714858,
      "loss": 1.353,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.6451165080070496,
      "learning_rate": 0.0005951748956072806,
      "loss": 1.3514,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.7312910556793213,
      "learning_rate": 0.0005869713134993463,
      "loss": 1.3396,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.4609370529651642,
      "learning_rate": 0.0005788808053128734,
      "loss": 1.3474,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5799684524536133,
      "learning_rate": 0.0005709018124955674,
      "loss": 1.342,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.5226260423660278,
      "learning_rate": 0.0005630327979774111,
      "loss": 1.3405,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.6010406017303467,
      "learning_rate": 0.0005552722458745627,
      "loss": 1.348,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.5826901197433472,
      "learning_rate": 0.0005476186611973374,
      "loss": 1.3383,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.5549695491790771,
      "learning_rate": 0.000540070569562213,
      "loss": 1.3397,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5249995589256287,
      "learning_rate": 0.0005326265169078048,
      "loss": 1.3595,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.565153181552887,
      "learning_rate": 0.0005252850692147567,
      "loss": 1.3401,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.48969659209251404,
      "learning_rate": 0.0005180448122294913,
      "loss": 1.3475,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.5739888548851013,
      "learning_rate": 0.0005109043511917693,
      "loss": 1.3471,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.5284473896026611,
      "learning_rate": 0.0005038623105660032,
      "loss": 1.3465,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.4692670404911041,
      "learning_rate": 0.0004969173337762747,
      "loss": 1.3364,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.4796403646469116,
      "learning_rate": 0.0004900680829450042,
      "loss": 1.33,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.48727184534072876,
      "learning_rate": 0.0004833132386352233,
      "loss": 1.336,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.4695288836956024,
      "learning_rate": 0.00047665149959639813,
      "loss": 1.334,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.5014991164207458,
      "learning_rate": 0.0004700815825137577,
      "loss": 1.3422,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.44504332542419434,
      "learning_rate": 0.00046360222176107584,
      "loss": 1.3159,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.6361677646636963,
      "learning_rate": 0.0004572121691568625,
      "loss": 1.331,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.5153830647468567,
      "learning_rate": 0.00045091019372391354,
      "loss": 1.3341,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.5851036906242371,
      "learning_rate": 0.0004446950814521764,
      "loss": 1.3415,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5900114178657532,
      "learning_rate": 0.0004385656350648835,
      "loss": 1.328,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.5923007726669312,
      "learning_rate": 0.00043252067378790946,
      "loss": 1.3333,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.4834819734096527,
      "learning_rate": 0.00042655903312230673,
      "loss": 1.3377,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.499101847410202,
      "learning_rate": 0.0004206795646199778,
      "loss": 1.3337,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.5846497416496277,
      "learning_rate": 0.0004148811356624379,
      "loss": 1.3341,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.471291720867157,
      "learning_rate": 0.0004091626292426282,
      "loss": 1.3285,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4688068628311157,
      "learning_rate": 0.0004035229437497357,
      "loss": 1.3382,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.5641908645629883,
      "learning_rate": 0.00039796099275697986,
      "loss": 1.3303,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.4740554392337799,
      "learning_rate": 0.0003924757048123232,
      "loss": 1.3412,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.4787723422050476,
      "learning_rate": 0.0003870660232320675,
      "loss": 1.3223,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5571549534797668,
      "learning_rate": 0.000381730905897295,
      "loss": 1.321,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.4965110421180725,
      "learning_rate": 0.0003764693250531141,
      "loss": 1.3276,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.7557862401008606,
      "learning_rate": 0.0003712802671106742,
      "loss": 1.3378,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.5894241333007812,
      "learning_rate": 0.0003661627324519073,
      "loss": 1.3146,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.49034640192985535,
      "learning_rate": 0.0003611157352369628,
      "loss": 1.3182,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.48033463954925537,
      "learning_rate": 0.00035613830321429534,
      "loss": 1.322,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.5243462324142456,
      "learning_rate": 0.00035122947753337037,
      "loss": 1.3255,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.5140429139137268,
      "learning_rate": 0.0003463883125599521,
      "loss": 1.3377,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.5377463102340698,
      "learning_rate": 0.00034161387569393647,
      "loss": 1.3328,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.5283532738685608,
      "learning_rate": 0.00033690524718969593,
      "loss": 1.3296,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.4901208281517029,
      "learning_rate": 0.0003322615199788993,
      "loss": 1.3233,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.4646488130092621,
      "learning_rate": 0.00032768179949577516,
      "loss": 1.321,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.49830129742622375,
      "learning_rate": 0.0003231652035047826,
      "loss": 1.3192,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.6454711556434631,
      "learning_rate": 0.000318710861930658,
      "loss": 1.3227,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.4911431074142456,
      "learning_rate": 0.0003143179166908038,
      "loss": 1.3326,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.5197914242744446,
      "learning_rate": 0.00030998552152998834,
      "loss": 1.3362,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.4570811688899994,
      "learning_rate": 0.00030571284185732276,
      "loss": 1.3172,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.5702477693557739,
      "learning_rate": 0.0003014990545854864,
      "loss": 1.3117,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5021108984947205,
      "learning_rate": 0.0002973433479721675,
      "loss": 1.3169,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.5911805629730225,
      "learning_rate": 0.00029324492146368906,
      "loss": 1.3139,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.4756425619125366,
      "learning_rate": 0.00028920298554079113,
      "loss": 1.3101,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.4732058048248291,
      "learning_rate": 0.00028521676156653756,
      "loss": 1.3195,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.5115228891372681,
      "learning_rate": 0.00028128548163632006,
      "loss": 1.3191,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.5534390807151794,
      "learning_rate": 0.0002774083884299292,
      "loss": 1.3064,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.46912050247192383,
      "learning_rate": 0.0002735847350656645,
      "loss": 1.3158,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.5076785683631897,
      "learning_rate": 0.0002698137849564556,
      "loss": 1.3241,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.5120282173156738,
      "learning_rate": 0.0002660948116679665,
      "loss": 1.3036,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.5615637898445129,
      "learning_rate": 0.00026242709877865493,
      "loss": 1.309,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.4717916250228882,
      "learning_rate": 0.00025880993974176204,
      "loss": 1.2957,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.47996142506599426,
      "learning_rate": 0.0002552426377492028,
      "loss": 1.2992,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.45766481757164,
      "learning_rate": 0.0002517245055973337,
      "loss": 1.303,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.482280433177948,
      "learning_rate": 0.00024825486555456975,
      "loss": 1.3137,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.5143918395042419,
      "learning_rate": 0.00024483304923082663,
      "loss": 1.3052,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.513719916343689,
      "learning_rate": 0.0002414583974487624,
      "loss": 1.3109,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.45668476819992065,
      "learning_rate": 0.00023813026011679372,
      "loss": 1.3182,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.49208977818489075,
      "learning_rate": 0.0002348479961038625,
      "loss": 1.313,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.46777087450027466,
      "learning_rate": 0.00023161097311592867,
      "loss": 1.3178,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.4836950898170471,
      "learning_rate": 0.00022841856757416538,
      "loss": 1.303,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.46816742420196533,
      "learning_rate": 0.0002252701644948328,
      "loss": 1.3028,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.5724418759346008,
      "learning_rate": 0.00022216515737080817,
      "loss": 1.3023,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.4637795388698578,
      "learning_rate": 0.00021910294805474833,
      "loss": 1.2987,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.5904604196548462,
      "learning_rate": 0.0002160829466438629,
      "loss": 1.3005,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.4693807065486908,
      "learning_rate": 0.00021310457136627562,
      "loss": 1.3132,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.47196266055107117,
      "learning_rate": 0.00021016724846895213,
      "loss": 1.3096,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.4710361957550049,
      "learning_rate": 0.00020727041210717235,
      "loss": 1.3005,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.4702761471271515,
      "learning_rate": 0.00020441350423552624,
      "loss": 1.3018,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.5067238211631775,
      "learning_rate": 0.00020159597450041257,
      "loss": 1.3151,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.4618867039680481,
      "learning_rate": 0.00019881728013401842,
      "loss": 1.2896,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.534237265586853,
      "learning_rate": 0.00019607688584976116,
      "loss": 1.3143,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.5107449293136597,
      "learning_rate": 0.00019337426373917076,
      "loss": 1.3021,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.47250908613204956,
      "learning_rate": 0.00019070889317019375,
      "loss": 1.3203,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.463882178068161,
      "learning_rate": 0.00018808026068689883,
      "loss": 1.2975,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.4829825460910797,
      "learning_rate": 0.00018548785991056508,
      "loss": 1.3031,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5524773001670837,
      "learning_rate": 0.00018293119144213328,
      "loss": 1.3172,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.6024357676506042,
      "learning_rate": 0.00018040976276600176,
      "loss": 1.2894,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.5167730450630188,
      "learning_rate": 0.00017792308815514854,
      "loss": 1.2996,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.537597119808197,
      "learning_rate": 0.00017547068857756104,
      "loss": 1.2992,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.47882741689682007,
      "learning_rate": 0.00017305209160395547,
      "loss": 1.3024,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.5075975060462952,
      "learning_rate": 0.00017066683131676825,
      "loss": 1.3086,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.45599648356437683,
      "learning_rate": 0.00016831444822040207,
      "loss": 1.3073,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.4692286252975464,
      "learning_rate": 0.00016599448915270845,
      "loss": 1.2932,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5363489985466003,
      "learning_rate": 0.000163706507197691,
      "loss": 1.294,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.46856990456581116,
      "learning_rate": 0.0001614500615994117,
      "loss": 1.302,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.529060959815979,
      "learning_rate": 0.00015922471767708377,
      "loss": 1.2952,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.6456242799758911,
      "learning_rate": 0.00015703004674133498,
      "loss": 1.3112,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4637324810028076,
      "learning_rate": 0.00015486562601162512,
      "loss": 1.2997,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.523307204246521,
      "learning_rate": 0.0001527310385348017,
      "loss": 1.2974,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.45750248432159424,
      "learning_rate": 0.00015062587310477813,
      "loss": 1.304,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.463380366563797,
      "learning_rate": 0.00014854972418331948,
      "loss": 1.2948,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.48942434787750244,
      "learning_rate": 0.00014650219182191934,
      "loss": 1.2825,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.525461733341217,
      "learning_rate": 0.00014448288158475423,
      "loss": 1.2923,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.45308318734169006,
      "learning_rate": 0.0001424914044726995,
      "loss": 1.2776,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.6268229484558105,
      "learning_rate": 0.0001405273768483926,
      "loss": 1.291,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.496448278427124,
      "learning_rate": 0.0001385904203623296,
      "loss": 1.3135,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.47262054681777954,
      "learning_rate": 0.00013668016187997964,
      "loss": 1.3109,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.4877476692199707,
      "learning_rate": 0.0001347962334099052,
      "loss": 1.2887,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5657548904418945,
      "learning_rate": 0.00013293827203287141,
      "loss": 1.2999,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5200432538986206,
      "learning_rate": 0.00013110591983193424,
      "loss": 1.296,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.4625248908996582,
      "learning_rate": 0.00012929882382349103,
      "loss": 1.3053,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.5191882252693176,
      "learning_rate": 0.0001275166358892821,
      "loss": 1.3006,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.49986517429351807,
      "learning_rate": 0.00012575901270932944,
      "loss": 1.3017,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.4724440276622772,
      "learning_rate": 0.00012402561569579935,
      "loss": 1.287,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.4518606662750244,
      "learning_rate": 0.00012231611092777743,
      "loss": 1.2891,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.46727433800697327,
      "learning_rate": 0.00012063016908694192,
      "loss": 1.2874,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.5409188866615295,
      "learning_rate": 0.00011896746539412405,
      "loss": 1.3011,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.5100900530815125,
      "learning_rate": 0.00011732767954674264,
      "loss": 1.2931,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.49695006012916565,
      "learning_rate": 0.00011571049565710122,
      "loss": 1.3001,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.5343964695930481,
      "learning_rate": 0.00011411560219153552,
      "loss": 1.2943,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.4976387023925781,
      "learning_rate": 0.0001125426919103997,
      "loss": 1.2914,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.47220513224601746,
      "learning_rate": 0.00011099146180887992,
      "loss": 1.3017,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.465561181306839,
      "learning_rate": 0.0001094616130586235,
      "loss": 1.284,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.49773499369621277,
      "learning_rate": 0.00010795285095017282,
      "loss": 1.2936,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.4696478545665741,
      "learning_rate": 0.00010646488483619263,
      "loss": 1.2975,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4973534643650055,
      "learning_rate": 0.00010499742807547978,
      "loss": 1.2995,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.5054043531417847,
      "learning_rate": 0.0001035501979777448,
      "loss": 1.2895,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.5566696524620056,
      "learning_rate": 0.00010212291574915464,
      "loss": 1.2813,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.5192328691482544,
      "learning_rate": 0.00010071530643862575,
      "loss": 1.299,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.4706285893917084,
      "learning_rate": 9.932709888485788e-05,
      "loss": 1.2908,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.5187116861343384,
      "learning_rate": 9.79580256640974e-05,
      "loss": 1.2899,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.46268996596336365,
      "learning_rate": 9.660782303862107e-05,
      "loss": 1.2889,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.4860462546348572,
      "learning_rate": 9.527623090592962e-05,
      "loss": 1.282,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.49108442664146423,
      "learning_rate": 9.396299274864176e-05,
      "loss": 1.296,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.5382658839225769,
      "learning_rate": 9.266785558507876e-05,
      "loss": 1.2887,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.5346512198448181,
      "learning_rate": 9.139056992053017e-05,
      "loss": 1.2945,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.5191641449928284,
      "learning_rate": 9.01308896991912e-05,
      "loss": 1.286,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.46642035245895386,
      "learning_rate": 8.88885722567627e-05,
      "loss": 1.2987,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.4673319160938263,
      "learning_rate": 8.766337827370438e-05,
      "loss": 1.2916,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.4642171263694763,
      "learning_rate": 8.645507172913238e-05,
      "loss": 1.2993,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.47717443108558655,
      "learning_rate": 8.52634198553523e-05,
      "loss": 1.2958,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.4526664912700653,
      "learning_rate": 8.408819309301891e-05,
      "loss": 1.2922,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.4961718022823334,
      "learning_rate": 8.292916504691398e-05,
      "loss": 1.3044,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.4467519521713257,
      "learning_rate": 8.178611244233354e-05,
      "loss": 1.2958,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.4841761887073517,
      "learning_rate": 8.065881508207636e-05,
      "loss": 1.293,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.4580075442790985,
      "learning_rate": 7.954705580402525e-05,
      "loss": 1.2756,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.5612879395484924,
      "learning_rate": 7.845062043931299e-05,
      "loss": 1.2776,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.4572225511074066,
      "learning_rate": 7.736929777106499e-05,
      "loss": 1.3018,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.46267932653427124,
      "learning_rate": 7.630287949371051e-05,
      "loss": 1.2814,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.44602736830711365,
      "learning_rate": 7.525116017285479e-05,
      "loss": 1.3011,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.4606595039367676,
      "learning_rate": 7.421393720570416e-05,
      "loss": 1.2942,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.45960938930511475,
      "learning_rate": 7.319101078203692e-05,
      "loss": 1.2803,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.49523669481277466,
      "learning_rate": 7.218218384571176e-05,
      "loss": 1.283,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.45749959349632263,
      "learning_rate": 7.118726205670702e-05,
      "loss": 1.2813,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.4498620629310608,
      "learning_rate": 7.020605375368314e-05,
      "loss": 1.2887,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.4596899449825287,
      "learning_rate": 6.923836991706108e-05,
      "loss": 1.2957,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.4731309115886688,
      "learning_rate": 6.828402413260965e-05,
      "loss": 1.2858,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.50437992811203,
      "learning_rate": 6.73428325555347e-05,
      "loss": 1.2861,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.5009710788726807,
      "learning_rate": 6.641461387506347e-05,
      "loss": 1.3037,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4667570888996124,
      "learning_rate": 6.549918927951679e-05,
      "loss": 1.2925,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.46593964099884033,
      "learning_rate": 6.459638242186298e-05,
      "loss": 1.2886,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.5209925770759583,
      "learning_rate": 6.370601938574637e-05,
      "loss": 1.2809,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5734021067619324,
      "learning_rate": 6.282792865198421e-05,
      "loss": 1.2907,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.5093693733215332,
      "learning_rate": 6.196194106552512e-05,
      "loss": 1.2863,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.46313154697418213,
      "learning_rate": 6.110788980286329e-05,
      "loss": 1.2886,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.5093756318092346,
      "learning_rate": 6.026561033990159e-05,
      "loss": 1.2821,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.49674439430236816,
      "learning_rate": 5.943494042025771e-05,
      "loss": 1.2866,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.49339917302131653,
      "learning_rate": 5.8615720024007174e-05,
      "loss": 1.2775,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.46326667070388794,
      "learning_rate": 5.780779133685717e-05,
      "loss": 1.2904,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.4716956913471222,
      "learning_rate": 5.701099871974525e-05,
      "loss": 1.2855,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.46486932039260864,
      "learning_rate": 5.6225188678857095e-05,
      "loss": 1.2982,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.45782825350761414,
      "learning_rate": 5.545020983605749e-05,
      "loss": 1.2956,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.46856826543807983,
      "learning_rate": 5.4685912899728965e-05,
      "loss": 1.2851,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.4586803913116455,
      "learning_rate": 5.39321506360123e-05,
      "loss": 1.2774,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5621423125267029,
      "learning_rate": 5.318877784044342e-05,
      "loss": 1.2984,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.5149664878845215,
      "learning_rate": 5.245565130998124e-05,
      "loss": 1.2885,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5729714632034302,
      "learning_rate": 5.173262981542119e-05,
      "loss": 1.2936,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.4599650204181671,
      "learning_rate": 5.101957407418877e-05,
      "loss": 1.2832,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.5617707371711731,
      "learning_rate": 5.0316346723508287e-05,
      "loss": 1.2872,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.5002422332763672,
      "learning_rate": 4.962281229394129e-05,
      "loss": 1.2828,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.478383868932724,
      "learning_rate": 4.893883718328984e-05,
      "loss": 1.2912,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5419098734855652,
      "learning_rate": 4.8264289630859386e-05,
      "loss": 1.2732,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.4829871952533722,
      "learning_rate": 4.759903969207646e-05,
      "loss": 1.2783,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.4976823627948761,
      "learning_rate": 4.694295921345623e-05,
      "loss": 1.2804,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.48330676555633545,
      "learning_rate": 4.629592180791501e-05,
      "loss": 1.2887,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.43191611766815186,
      "learning_rate": 4.565780283042316e-05,
      "loss": 1.2961,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4623429775238037,
      "learning_rate": 4.502847935399348e-05,
      "loss": 1.2887,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.49294161796569824,
      "learning_rate": 4.440783014600059e-05,
      "loss": 1.2865,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.4549841582775116,
      "learning_rate": 4.3795735644826776e-05,
      "loss": 1.2851,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.4817006587982178,
      "learning_rate": 4.319207793682965e-05,
      "loss": 1.2789,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.4689021408557892,
      "learning_rate": 4.259674073362732e-05,
      "loss": 1.2934,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.49034053087234497,
      "learning_rate": 4.200960934969664e-05,
      "loss": 1.2804,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.4780392646789551,
      "learning_rate": 4.143057068028024e-05,
      "loss": 1.2799,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.5309856534004211,
      "learning_rate": 4.0859513179598096e-05,
      "loss": 1.2662,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.45172807574272156,
      "learning_rate": 4.02963268393593e-05,
      "loss": 1.2796,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.4595884084701538,
      "learning_rate": 3.974090316757029e-05,
      "loss": 1.2822,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.4415097236633301,
      "learning_rate": 3.919313516763478e-05,
      "loss": 1.2811,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.4719064235687256,
      "learning_rate": 3.8652917317742106e-05,
      "loss": 1.2778,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.44753560423851013,
      "learning_rate": 3.812014555053955e-05,
      "loss": 1.2965,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.45936092734336853,
      "learning_rate": 3.759471723308477e-05,
      "loss": 1.2829,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.4726845622062683,
      "learning_rate": 3.707653114707471e-05,
      "loss": 1.2988,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.44957664608955383,
      "learning_rate": 3.6565487469346904e-05,
      "loss": 1.284,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.5235864520072937,
      "learning_rate": 3.606148775264958e-05,
      "loss": 1.2835,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5054678320884705,
      "learning_rate": 3.5564434906676834e-05,
      "loss": 1.2825,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.4912639260292053,
      "learning_rate": 3.507423317936521e-05,
      "loss": 1.2856,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.4780506193637848,
      "learning_rate": 3.4590788138448004e-05,
      "loss": 1.292,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.4532028138637543,
      "learning_rate": 3.411400665326393e-05,
      "loss": 1.2881,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.46156346797943115,
      "learning_rate": 3.364379687681642e-05,
      "loss": 1.2791,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.5188421010971069,
      "learning_rate": 3.31800682280803e-05,
      "loss": 1.2781,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.457661896944046,
      "learning_rate": 3.272273137455225e-05,
      "loss": 1.2867,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.46978896856307983,
      "learning_rate": 3.227169821504187e-05,
      "loss": 1.2823,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.4533792734146118,
      "learning_rate": 3.182688186269985e-05,
      "loss": 1.2842,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.4478567838668823,
      "learning_rate": 3.138819662828018e-05,
      "loss": 1.2831,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.48361706733703613,
      "learning_rate": 3.095555800363297e-05,
      "loss": 1.2866,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5194976925849915,
      "learning_rate": 3.052888264542484e-05,
      "loss": 1.285,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.593368411064148,
      "learning_rate": 3.0108088359083675e-05,
      "loss": 1.2823,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.470674067735672,
      "learning_rate": 2.9693094082964775e-05,
      "loss": 1.2863,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.45366424322128296,
      "learning_rate": 2.928381987273507e-05,
      "loss": 1.2808,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.5190941095352173,
      "learning_rate": 2.8880186885972716e-05,
      "loss": 1.2803,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.4770827293395996,
      "learning_rate": 2.8482117366978935e-05,
      "loss": 1.2788,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.5441578030586243,
      "learning_rate": 2.808953463179918e-05,
      "loss": 1.3002,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.4609360694885254,
      "learning_rate": 2.770236305345076e-05,
      "loss": 1.2742,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.47006136178970337,
      "learning_rate": 2.732052804735409e-05,
      "loss": 1.2855,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.4748915135860443,
      "learning_rate": 2.6943956056964773e-05,
      "loss": 1.2858,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.46649369597435,
      "learning_rate": 2.6572574539603643e-05,
      "loss": 1.2875,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.44580161571502686,
      "learning_rate": 2.6206311952482224e-05,
      "loss": 1.2886,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.4662837088108063,
      "learning_rate": 2.584509773892073e-05,
      "loss": 1.2885,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.4643775522708893,
      "learning_rate": 2.5488862314756066e-05,
      "loss": 1.2682,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.44580066204071045,
      "learning_rate": 2.513753705493713e-05,
      "loss": 1.2811,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.47444525361061096,
      "learning_rate": 2.4791054280304972e-05,
      "loss": 1.2887,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.48020312190055847,
      "learning_rate": 2.4449347244555043e-05,
      "loss": 1.2739,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.531441330909729,
      "learning_rate": 2.4112350121379255e-05,
      "loss": 1.2936,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.5161461234092712,
      "learning_rate": 2.3779997991785207e-05,
      "loss": 1.2806,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.4328412115573883,
      "learning_rate": 2.3452226831590232e-05,
      "loss": 1.2885,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.45244327187538147,
      "learning_rate": 2.3128973499087785e-05,
      "loss": 1.2857,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.46364492177963257,
      "learning_rate": 2.2810175722883866e-05,
      "loss": 1.2832,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.45162323117256165,
      "learning_rate": 2.2495772089901067e-05,
      "loss": 1.2791,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.4856041967868805,
      "learning_rate": 2.218570203354799e-05,
      "loss": 1.2836,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.4696802496910095,
      "learning_rate": 2.187990582205175e-05,
      "loss": 1.2824,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.4935973882675171,
      "learning_rate": 2.157832454695122e-05,
      "loss": 1.2809,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.4482496380805969,
      "learning_rate": 2.1280900111748943e-05,
      "loss": 1.2713,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.4511680006980896,
      "learning_rate": 2.0987575220719476e-05,
      "loss": 1.265,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.45410269498825073,
      "learning_rate": 2.069829336787193e-05,
      "loss": 1.2768,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.455393522977829,
      "learning_rate": 2.0412998826064695e-05,
      "loss": 1.2747,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.48198139667510986,
      "learning_rate": 2.0131636636270178e-05,
      "loss": 1.2863,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.4690043330192566,
      "learning_rate": 1.9854152596987523e-05,
      "loss": 1.2736,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.46250012516975403,
      "learning_rate": 1.9580493253801253e-05,
      "loss": 1.2801,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.45912715792655945,
      "learning_rate": 1.9310605889083842e-05,
      "loss": 1.2798,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.5150806307792664,
      "learning_rate": 1.904443851184018e-05,
      "loss": 1.28,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.4904801845550537,
      "learning_rate": 1.87819398476921e-05,
      "loss": 1.2755,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4826321601867676,
      "learning_rate": 1.8523059329000848e-05,
      "loss": 1.2878,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.47633230686187744,
      "learning_rate": 1.826774708512579e-05,
      "loss": 1.2845,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.4657934010028839,
      "learning_rate": 1.8015953932817347e-05,
      "loss": 1.2959,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.4584238827228546,
      "learning_rate": 1.7767631366742332e-05,
      "loss": 1.2879,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.48173826932907104,
      "learning_rate": 1.7522731550139926e-05,
      "loss": 1.2852,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.46752113103866577,
      "learning_rate": 1.728120730560641e-05,
      "loss": 1.2714,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.48856937885284424,
      "learning_rate": 1.704301210600693e-05,
      "loss": 1.2767,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.49191349744796753,
      "learning_rate": 1.6808100065512536e-05,
      "loss": 1.2902,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.49473485350608826,
      "learning_rate": 1.657642593076074e-05,
      "loss": 1.2649,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.4490512013435364,
      "learning_rate": 1.634794507213793e-05,
      "loss": 1.2829,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.45531249046325684,
      "learning_rate": 1.6122613475181976e-05,
      "loss": 1.2841,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.4522455334663391,
      "learning_rate": 1.590038773210323e-05,
      "loss": 1.2796,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.4630568027496338,
      "learning_rate": 1.568122503342252e-05,
      "loss": 1.2891,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.4376571774482727,
      "learning_rate": 1.5465083159724344e-05,
      "loss": 1.2732,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.4659065902233124,
      "learning_rate": 1.5251920473523708e-05,
      "loss": 1.2874,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.316353678703308,
      "learning_rate": 1.5041695911245136e-05,
      "loss": 1.276,
      "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": 1.832308198648013e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}