{
  "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": 1.150813341140747,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.6386,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 0.9222255349159241,
      "learning_rate": 0.00031645569620253165,
      "loss": 7.2705,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8442890644073486,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.5303,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.6285983324050903,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.8289,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.3649296462535858,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.3788,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.243421733379364,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.9696,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 0.21124279499053955,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.6049,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.3619547486305237,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.3648,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.3652024269104004,
      "learning_rate": 0.0014240506329113926,
      "loss": 4.1957,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.2144092321395874,
      "learning_rate": 0.0015,
      "loss": 4.0701,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.3157788813114166,
      "learning_rate": 0.0015,
      "loss": 3.9488,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.3224923610687256,
      "learning_rate": 0.0015,
      "loss": 3.8612,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.26904231309890747,
      "learning_rate": 0.0015,
      "loss": 3.7804,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.2192022055387497,
      "learning_rate": 0.0015,
      "loss": 3.6944,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.5886008739471436,
      "learning_rate": 0.0015,
      "loss": 3.6382,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.2867976725101471,
      "learning_rate": 0.0015,
      "loss": 3.579,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.2692849636077881,
      "learning_rate": 0.0015,
      "loss": 3.5003,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.31116002798080444,
      "learning_rate": 0.0015,
      "loss": 3.4577,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.617526650428772,
      "learning_rate": 0.0015,
      "loss": 3.4048,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.31083545088768005,
      "learning_rate": 0.0015,
      "loss": 3.3565,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.33985719084739685,
      "learning_rate": 0.0015,
      "loss": 3.3076,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.42189735174179077,
      "learning_rate": 0.0015,
      "loss": 3.2698,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.3632897138595581,
      "learning_rate": 0.0015,
      "loss": 3.2174,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.5148869752883911,
      "learning_rate": 0.0015,
      "loss": 3.1702,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.558694064617157,
      "learning_rate": 0.0015,
      "loss": 3.1347,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.3564925789833069,
      "learning_rate": 0.0015,
      "loss": 3.0917,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.5106367468833923,
      "learning_rate": 0.0015,
      "loss": 3.0564,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.5460816025733948,
      "learning_rate": 0.0015,
      "loss": 3.0089,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.4177979826927185,
      "learning_rate": 0.0015,
      "loss": 2.9935,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.4817129075527191,
      "learning_rate": 0.0015,
      "loss": 2.9533,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.45403867959976196,
      "learning_rate": 0.0015,
      "loss": 2.9139,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.5642548203468323,
      "learning_rate": 0.0015,
      "loss": 2.8754,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.3790704905986786,
      "learning_rate": 0.0015,
      "loss": 2.8677,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.6472330689430237,
      "learning_rate": 0.0015,
      "loss": 2.8357,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.6580318212509155,
      "learning_rate": 0.0015,
      "loss": 2.7878,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.6148264408111572,
      "learning_rate": 0.0015,
      "loss": 2.7579,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.5459489226341248,
      "learning_rate": 0.0015,
      "loss": 2.7384,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.38705775141716003,
      "learning_rate": 0.0015,
      "loss": 2.7189,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.7032163143157959,
      "learning_rate": 0.0015,
      "loss": 2.6934,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.5982773900032043,
      "learning_rate": 0.0015,
      "loss": 2.6665,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.568385899066925,
      "learning_rate": 0.0015,
      "loss": 2.6469,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.602777898311615,
      "learning_rate": 0.0015,
      "loss": 2.6266,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.5950765609741211,
      "learning_rate": 0.0015,
      "loss": 2.5922,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.5538167953491211,
      "learning_rate": 0.0015,
      "loss": 2.5684,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.5699306130409241,
      "learning_rate": 0.0015,
      "loss": 2.5589,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 1.100414514541626,
      "learning_rate": 0.0015,
      "loss": 2.5376,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.555873453617096,
      "learning_rate": 0.0015,
      "loss": 2.5088,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.44411933422088623,
      "learning_rate": 0.0015,
      "loss": 2.5043,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.6461852788925171,
      "learning_rate": 0.0015,
      "loss": 2.4712,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.4979498088359833,
      "learning_rate": 0.0015,
      "loss": 2.463,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.6500505208969116,
      "learning_rate": 0.0015,
      "loss": 2.4463,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.5891791582107544,
      "learning_rate": 0.0015,
      "loss": 2.4269,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.5033569931983948,
      "learning_rate": 0.0015,
      "loss": 2.4104,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 1.0817065238952637,
      "learning_rate": 0.0015,
      "loss": 2.3933,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.8942646384239197,
      "learning_rate": 0.0015,
      "loss": 2.3918,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.4584704041481018,
      "learning_rate": 0.0015,
      "loss": 2.3453,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.4871450960636139,
      "learning_rate": 0.0015,
      "loss": 2.3442,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 0.8289301991462708,
      "learning_rate": 0.0015,
      "loss": 2.3445,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.72162926197052,
      "learning_rate": 0.0015,
      "loss": 2.3166,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.40924233198165894,
      "learning_rate": 0.0015,
      "loss": 2.3003,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.4829574525356293,
      "learning_rate": 0.0015,
      "loss": 2.2965,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.49946269392967224,
      "learning_rate": 0.0015,
      "loss": 2.2814,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.5377524495124817,
      "learning_rate": 0.0015,
      "loss": 2.2663,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.39217859506607056,
      "learning_rate": 0.0015,
      "loss": 2.2744,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.4565280079841614,
      "learning_rate": 0.0015,
      "loss": 2.2613,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.639655590057373,
      "learning_rate": 0.0015,
      "loss": 2.2461,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.44999587535858154,
      "learning_rate": 0.0015,
      "loss": 2.2282,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.48550304770469666,
      "learning_rate": 0.0015,
      "loss": 2.2224,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.47652631998062134,
      "learning_rate": 0.0015,
      "loss": 2.2308,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.4794254004955292,
      "learning_rate": 0.0015,
      "loss": 2.2026,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.5449780225753784,
      "learning_rate": 0.0015,
      "loss": 2.1991,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.7790537476539612,
      "learning_rate": 0.0015,
      "loss": 2.1739,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.9798956513404846,
      "learning_rate": 0.0015,
      "loss": 2.1706,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.49852877855300903,
      "learning_rate": 0.0015,
      "loss": 2.1797,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.6039758324623108,
      "learning_rate": 0.0015,
      "loss": 2.1567,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 0.7530763745307922,
      "learning_rate": 0.0015,
      "loss": 2.1565,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.9843342304229736,
      "learning_rate": 0.0015,
      "loss": 2.1471,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.5783156752586365,
      "learning_rate": 0.0015,
      "loss": 2.1435,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.44985154271125793,
      "learning_rate": 0.0015,
      "loss": 2.144,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.539827287197113,
      "learning_rate": 0.0015,
      "loss": 2.1162,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.4987821578979492,
      "learning_rate": 0.0015,
      "loss": 2.1021,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.4859192371368408,
      "learning_rate": 0.0015,
      "loss": 2.1101,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.41457563638687134,
      "learning_rate": 0.0015,
      "loss": 2.1076,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 1.0149922370910645,
      "learning_rate": 0.0015,
      "loss": 2.0877,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.49760764837265015,
      "learning_rate": 0.0015,
      "loss": 2.0949,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.43229302763938904,
      "learning_rate": 0.0015,
      "loss": 2.0843,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.6546627283096313,
      "learning_rate": 0.0015,
      "loss": 2.0812,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.575627863407135,
      "learning_rate": 0.0015,
      "loss": 2.0758,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.47545164823532104,
      "learning_rate": 0.0015,
      "loss": 2.0634,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.4198952913284302,
      "learning_rate": 0.0015,
      "loss": 2.0588,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.521074652671814,
      "learning_rate": 0.0015,
      "loss": 2.0657,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.6115285158157349,
      "learning_rate": 0.0015,
      "loss": 2.0513,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.6493558287620544,
      "learning_rate": 0.0015,
      "loss": 2.0448,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.8226097226142883,
      "learning_rate": 0.0015,
      "loss": 2.0504,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.6946693658828735,
      "learning_rate": 0.0015,
      "loss": 2.0276,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.5198275446891785,
      "learning_rate": 0.0015,
      "loss": 2.0359,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 0.7822059988975525,
      "learning_rate": 0.0015,
      "loss": 2.0298,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.6713688373565674,
      "learning_rate": 0.0015,
      "loss": 2.0114,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.5232925415039062,
      "learning_rate": 0.0015,
      "loss": 2.0095,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.7252864837646484,
      "learning_rate": 0.0015,
      "loss": 2.0188,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.6016257405281067,
      "learning_rate": 0.0015,
      "loss": 2.0124,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.6161438822746277,
      "learning_rate": 0.0015,
      "loss": 1.9974,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.6386392116546631,
      "learning_rate": 0.0015,
      "loss": 1.9978,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.4686180651187897,
      "learning_rate": 0.0015,
      "loss": 1.9909,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.8074453473091125,
      "learning_rate": 0.0015,
      "loss": 1.9951,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.6009097099304199,
      "learning_rate": 0.0015,
      "loss": 1.9872,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.5426019430160522,
      "learning_rate": 0.0015,
      "loss": 1.9725,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.6576230525970459,
      "learning_rate": 0.0015,
      "loss": 1.9745,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.4691566824913025,
      "learning_rate": 0.0015,
      "loss": 1.9826,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.46580421924591064,
      "learning_rate": 0.0015,
      "loss": 1.9651,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.47598549723625183,
      "learning_rate": 0.0015,
      "loss": 1.9696,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.4900604784488678,
      "learning_rate": 0.0015,
      "loss": 1.962,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 1.0558894872665405,
      "learning_rate": 0.0015,
      "loss": 1.9495,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.4379860460758209,
      "learning_rate": 0.0015,
      "loss": 1.9671,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 0.86305832862854,
      "learning_rate": 0.0015,
      "loss": 1.9446,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.5544790029525757,
      "learning_rate": 0.0015,
      "loss": 1.9562,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.6956931948661804,
      "learning_rate": 0.0015,
      "loss": 1.9462,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.472929447889328,
      "learning_rate": 0.0015,
      "loss": 1.9301,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.4986661970615387,
      "learning_rate": 0.0015,
      "loss": 1.9298,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.6288496851921082,
      "learning_rate": 0.0015,
      "loss": 1.9431,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.4661789536476135,
      "learning_rate": 0.0015,
      "loss": 1.9342,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.4968653619289398,
      "learning_rate": 0.0015,
      "loss": 1.9187,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.7806333303451538,
      "learning_rate": 0.0015,
      "loss": 1.9089,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.4545213282108307,
      "learning_rate": 0.0015,
      "loss": 1.9265,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.4266795516014099,
      "learning_rate": 0.0015,
      "loss": 1.9248,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.9585714936256409,
      "learning_rate": 0.0015,
      "loss": 1.9094,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.8499062061309814,
      "learning_rate": 0.0015,
      "loss": 1.9159,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.48624470829963684,
      "learning_rate": 0.0015,
      "loss": 1.9249,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.43314653635025024,
      "learning_rate": 0.0015,
      "loss": 1.9068,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.42100489139556885,
      "learning_rate": 0.0015,
      "loss": 1.8916,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.44472599029541016,
      "learning_rate": 0.0015,
      "loss": 1.8974,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.8936122059822083,
      "learning_rate": 0.0015,
      "loss": 1.8974,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 0.44905006885528564,
      "learning_rate": 0.0015,
      "loss": 1.8947,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.49804428219795227,
      "learning_rate": 0.0015,
      "loss": 1.89,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.5527216196060181,
      "learning_rate": 0.0015,
      "loss": 1.8954,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.5694833993911743,
      "learning_rate": 0.0015,
      "loss": 1.8879,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.5084961652755737,
      "learning_rate": 0.0015,
      "loss": 1.8835,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.541684091091156,
      "learning_rate": 0.0015,
      "loss": 1.8736,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.4399120509624481,
      "learning_rate": 0.0015,
      "loss": 1.8841,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.49186035990715027,
      "learning_rate": 0.0015,
      "loss": 1.8662,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.47912323474884033,
      "learning_rate": 0.0015,
      "loss": 1.8666,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.6098805069923401,
      "learning_rate": 0.0015,
      "loss": 1.8768,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.6260626316070557,
      "learning_rate": 0.0015,
      "loss": 1.8675,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.8695244789123535,
      "learning_rate": 0.0015,
      "loss": 1.8666,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.5537676215171814,
      "learning_rate": 0.0015,
      "loss": 1.8635,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.5429517030715942,
      "learning_rate": 0.0015,
      "loss": 1.8584,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.4902762770652771,
      "learning_rate": 0.0015,
      "loss": 1.8526,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.5063686966896057,
      "learning_rate": 0.0015,
      "loss": 1.8556,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.4521937966346741,
      "learning_rate": 0.0015,
      "loss": 1.8542,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.7287399172782898,
      "learning_rate": 0.0015,
      "loss": 1.8565,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.5698118805885315,
      "learning_rate": 0.0015,
      "loss": 1.854,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.5222446322441101,
      "learning_rate": 0.0015,
      "loss": 1.8559,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.5126030445098877,
      "learning_rate": 0.0015,
      "loss": 1.8343,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6289701461791992,
      "learning_rate": 0.0015,
      "loss": 1.828,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.5062997341156006,
      "learning_rate": 0.0015,
      "loss": 1.8307,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.4616747498512268,
      "learning_rate": 0.0015,
      "loss": 1.8323,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.6381959915161133,
      "learning_rate": 0.0015,
      "loss": 1.8252,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.8569875955581665,
      "learning_rate": 0.0015,
      "loss": 1.8402,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.4836009740829468,
      "learning_rate": 0.0015,
      "loss": 1.8276,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.5724925398826599,
      "learning_rate": 0.0015,
      "loss": 1.8315,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.5199964642524719,
      "learning_rate": 0.0015,
      "loss": 1.8291,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6030388474464417,
      "learning_rate": 0.0015,
      "loss": 1.8022,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.5144745707511902,
      "learning_rate": 0.0015,
      "loss": 1.8165,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.5104593634605408,
      "learning_rate": 0.0015,
      "loss": 1.8167,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.686623752117157,
      "learning_rate": 0.0015,
      "loss": 1.808,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.4637073278427124,
      "learning_rate": 0.0015,
      "loss": 1.8073,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.5601900815963745,
      "learning_rate": 0.0015,
      "loss": 1.8147,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.8115816116333008,
      "learning_rate": 0.0015,
      "loss": 1.8063,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.5719322562217712,
      "learning_rate": 0.0015,
      "loss": 1.7955,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.4701850712299347,
      "learning_rate": 0.0015,
      "loss": 1.8115,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.535297155380249,
      "learning_rate": 0.0015,
      "loss": 1.7919,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.5622710585594177,
      "learning_rate": 0.0015,
      "loss": 1.7956,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.4849448502063751,
      "learning_rate": 0.0015,
      "loss": 1.7942,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.6260956525802612,
      "learning_rate": 0.0015,
      "loss": 1.7974,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.5183302164077759,
      "learning_rate": 0.0015,
      "loss": 1.8183,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.49603813886642456,
      "learning_rate": 0.0015,
      "loss": 1.7976,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.5160091519355774,
      "learning_rate": 0.0015,
      "loss": 1.7964,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.6302099227905273,
      "learning_rate": 0.0015,
      "loss": 1.7947,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.6066163778305054,
      "learning_rate": 0.0015,
      "loss": 1.7939,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.8006702065467834,
      "learning_rate": 0.0015,
      "loss": 1.7888,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.44408902525901794,
      "learning_rate": 0.0015,
      "loss": 1.7812,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.6326849460601807,
      "learning_rate": 0.0015,
      "loss": 1.7835,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.5474056005477905,
      "learning_rate": 0.0015,
      "loss": 1.7788,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.4583626687526703,
      "learning_rate": 0.0015,
      "loss": 1.7756,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.5443903207778931,
      "learning_rate": 0.0015,
      "loss": 1.776,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.6028438806533813,
      "learning_rate": 0.0015,
      "loss": 1.7816,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.8050011992454529,
      "learning_rate": 0.0015,
      "loss": 1.7832,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.5138239860534668,
      "learning_rate": 0.0015,
      "loss": 1.7663,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.5934531688690186,
      "learning_rate": 0.0015,
      "loss": 1.7715,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.5334857702255249,
      "learning_rate": 0.0015,
      "loss": 1.7711,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.6304555535316467,
      "learning_rate": 0.0015,
      "loss": 1.7701,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.5386019945144653,
      "learning_rate": 0.0015,
      "loss": 1.7676,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.8461247086524963,
      "learning_rate": 0.0015,
      "loss": 1.7696,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.5659627318382263,
      "learning_rate": 0.0015,
      "loss": 1.7576,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.45056530833244324,
      "learning_rate": 0.0015,
      "loss": 1.7592,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.5300774574279785,
      "learning_rate": 0.0015,
      "loss": 1.7616,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.4294365644454956,
      "learning_rate": 0.0015,
      "loss": 1.7708,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.7624651193618774,
      "learning_rate": 0.0015,
      "loss": 1.7542,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.8281710743904114,
      "learning_rate": 0.0015,
      "loss": 1.751,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.49018529057502747,
      "learning_rate": 0.0015,
      "loss": 1.752,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.6264277696609497,
      "learning_rate": 0.0015,
      "loss": 1.7565,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.720066249370575,
      "learning_rate": 0.0015,
      "loss": 1.7686,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.48484906554222107,
      "learning_rate": 0.0015,
      "loss": 1.7502,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.4189296364784241,
      "learning_rate": 0.0015,
      "loss": 1.7496,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.43970799446105957,
      "learning_rate": 0.0015,
      "loss": 1.751,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.9812093377113342,
      "learning_rate": 0.0015,
      "loss": 1.7439,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.8057104349136353,
      "learning_rate": 0.0015,
      "loss": 1.7427,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.5278838872909546,
      "learning_rate": 0.0015,
      "loss": 1.7466,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.5162642598152161,
      "learning_rate": 0.0015,
      "loss": 1.7364,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.4243667721748352,
      "learning_rate": 0.0015,
      "loss": 1.7465,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.5826160907745361,
      "learning_rate": 0.0015,
      "loss": 1.7353,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6756936311721802,
      "learning_rate": 0.0015,
      "loss": 1.7338,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.43055227398872375,
      "learning_rate": 0.0015,
      "loss": 1.7379,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6738662719726562,
      "learning_rate": 0.0015,
      "loss": 1.7307,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.5482190251350403,
      "learning_rate": 0.0015,
      "loss": 1.7379,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.8880805373191833,
      "learning_rate": 0.0015,
      "loss": 1.752,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.44593849778175354,
      "learning_rate": 0.0015,
      "loss": 1.7368,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.5845939517021179,
      "learning_rate": 0.0015,
      "loss": 1.7188,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.4610976278781891,
      "learning_rate": 0.0015,
      "loss": 1.736,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.7984962463378906,
      "learning_rate": 0.0015,
      "loss": 1.735,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.479261189699173,
      "learning_rate": 0.0015,
      "loss": 1.7216,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.5528878569602966,
      "learning_rate": 0.0015,
      "loss": 1.7316,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.47382694482803345,
      "learning_rate": 0.0015,
      "loss": 1.7248,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.4474259912967682,
      "learning_rate": 0.0015,
      "loss": 1.7175,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.4632035791873932,
      "learning_rate": 0.0015,
      "loss": 1.7171,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.7560294270515442,
      "learning_rate": 0.0015,
      "loss": 1.7133,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.6277971863746643,
      "learning_rate": 0.0015,
      "loss": 1.7402,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.5490138530731201,
      "learning_rate": 0.0015,
      "loss": 1.7278,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.9364339113235474,
      "learning_rate": 0.0015,
      "loss": 1.7159,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.5979941487312317,
      "learning_rate": 0.0015,
      "loss": 1.7124,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.8417498469352722,
      "learning_rate": 0.0015,
      "loss": 1.717,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.4808991551399231,
      "learning_rate": 0.0015,
      "loss": 1.7226,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.4588165879249573,
      "learning_rate": 0.0015,
      "loss": 1.7037,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.6004635095596313,
      "learning_rate": 0.0015,
      "loss": 1.7037,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.7591840028762817,
      "learning_rate": 0.0015,
      "loss": 1.6986,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.6121309995651245,
      "learning_rate": 0.0015,
      "loss": 1.7124,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5905430912971497,
      "learning_rate": 0.0015,
      "loss": 1.7085,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.8464624285697937,
      "learning_rate": 0.0015,
      "loss": 1.708,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.8337712287902832,
      "learning_rate": 0.0015,
      "loss": 1.7081,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.4713364839553833,
      "learning_rate": 0.0015,
      "loss": 1.7023,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.6810732483863831,
      "learning_rate": 0.0015,
      "loss": 1.7072,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.5745030045509338,
      "learning_rate": 0.0015,
      "loss": 1.7042,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.5170700550079346,
      "learning_rate": 0.0015,
      "loss": 1.7009,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.7049311399459839,
      "learning_rate": 0.0015,
      "loss": 1.7082,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.5195064544677734,
      "learning_rate": 0.0015,
      "loss": 1.6935,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.46363475918769836,
      "learning_rate": 0.0015,
      "loss": 1.7066,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.4979170858860016,
      "learning_rate": 0.0015,
      "loss": 1.697,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.6032502055168152,
      "learning_rate": 0.0015,
      "loss": 1.7018,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 1.4362196922302246,
      "learning_rate": 0.0015,
      "loss": 1.6966,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.49109920859336853,
      "learning_rate": 0.0015,
      "loss": 1.6975,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.46409809589385986,
      "learning_rate": 0.0015,
      "loss": 1.6911,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.5044241547584534,
      "learning_rate": 0.0015,
      "loss": 1.6978,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.6585612297058105,
      "learning_rate": 0.0015,
      "loss": 1.688,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.5030466318130493,
      "learning_rate": 0.0015,
      "loss": 1.6959,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.44553372263908386,
      "learning_rate": 0.0015,
      "loss": 1.6851,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.4382231831550598,
      "learning_rate": 0.0015,
      "loss": 1.6926,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.572163462638855,
      "learning_rate": 0.0015,
      "loss": 1.6882,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.4450538754463196,
      "learning_rate": 0.0015,
      "loss": 1.6888,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.46374884247779846,
      "learning_rate": 0.0015,
      "loss": 1.6928,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.5283554792404175,
      "learning_rate": 0.0015,
      "loss": 1.6939,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6039606332778931,
      "learning_rate": 0.0015,
      "loss": 1.6931,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.8294399380683899,
      "learning_rate": 0.0015,
      "loss": 1.6863,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.8275244832038879,
      "learning_rate": 0.0015,
      "loss": 1.6808,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.5573651194572449,
      "learning_rate": 0.0015,
      "loss": 1.69,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.512090265750885,
      "learning_rate": 0.0015,
      "loss": 1.682,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.5057612657546997,
      "learning_rate": 0.0015,
      "loss": 1.6739,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.4737897217273712,
      "learning_rate": 0.0015,
      "loss": 1.6771,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.5719987154006958,
      "learning_rate": 0.0015,
      "loss": 1.6803,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.8795638680458069,
      "learning_rate": 0.0015,
      "loss": 1.6787,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.4607473313808441,
      "learning_rate": 0.0015,
      "loss": 1.6759,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.7981528639793396,
      "learning_rate": 0.0015,
      "loss": 1.6809,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.4333970546722412,
      "learning_rate": 0.0015,
      "loss": 1.6762,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.50799560546875,
      "learning_rate": 0.0015,
      "loss": 1.6593,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.9140480160713196,
      "learning_rate": 0.0015,
      "loss": 1.6704,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.4954560697078705,
      "learning_rate": 0.0015,
      "loss": 1.6766,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.5669369697570801,
      "learning_rate": 0.0015,
      "loss": 1.6824,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.5241600275039673,
      "learning_rate": 0.0015,
      "loss": 1.6805,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.5225156545639038,
      "learning_rate": 0.0015,
      "loss": 1.6742,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.8474009037017822,
      "learning_rate": 0.0015,
      "loss": 1.6732,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.9528505802154541,
      "learning_rate": 0.0015,
      "loss": 1.6725,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.6440209746360779,
      "learning_rate": 0.0015,
      "loss": 1.6707,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.4494965374469757,
      "learning_rate": 0.0015,
      "loss": 1.6792,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.6342407464981079,
      "learning_rate": 0.0015,
      "loss": 1.6491,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.5306560397148132,
      "learning_rate": 0.0015,
      "loss": 1.6538,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.887423038482666,
      "learning_rate": 0.0015,
      "loss": 1.66,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.5342133045196533,
      "learning_rate": 0.0015,
      "loss": 1.6641,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.5690568089485168,
      "learning_rate": 0.0015,
      "loss": 1.665,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.9180155396461487,
      "learning_rate": 0.0015,
      "loss": 1.6522,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.49240967631340027,
      "learning_rate": 0.0015,
      "loss": 1.6577,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.5052701830863953,
      "learning_rate": 0.0015,
      "loss": 1.6636,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.5250066518783569,
      "learning_rate": 0.0015,
      "loss": 1.6677,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.6492303013801575,
      "learning_rate": 0.0015,
      "loss": 1.6564,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.739266574382782,
      "learning_rate": 0.0015,
      "loss": 1.6542,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.4702005386352539,
      "learning_rate": 0.0015,
      "loss": 1.6595,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.6040682196617126,
      "learning_rate": 0.0015,
      "loss": 1.6675,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.6438997983932495,
      "learning_rate": 0.0015,
      "loss": 1.6472,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.5325720906257629,
      "learning_rate": 0.0015,
      "loss": 1.6406,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.6444641947746277,
      "learning_rate": 0.0015,
      "loss": 1.6481,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.4740985333919525,
      "learning_rate": 0.0015,
      "loss": 1.6659,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.4967401325702667,
      "learning_rate": 0.0015,
      "loss": 1.6432,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 1.3561769723892212,
      "learning_rate": 0.0015,
      "loss": 1.6535,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.9485900402069092,
      "learning_rate": 0.0015,
      "loss": 1.6623,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.46477529406547546,
      "learning_rate": 0.0015,
      "loss": 1.6381,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.5185746550559998,
      "learning_rate": 0.0015,
      "loss": 1.645,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.44782137870788574,
      "learning_rate": 0.0015,
      "loss": 1.6417,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.5505303740501404,
      "learning_rate": 0.0015,
      "loss": 1.6533,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.5244767069816589,
      "learning_rate": 0.0015,
      "loss": 1.6446,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.684838056564331,
      "learning_rate": 0.0015,
      "loss": 1.6319,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.6142449378967285,
      "learning_rate": 0.0015,
      "loss": 1.6519,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.8353508710861206,
      "learning_rate": 0.0015,
      "loss": 1.6386,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.44266948103904724,
      "learning_rate": 0.0015,
      "loss": 1.6413,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.5067423582077026,
      "learning_rate": 0.0015,
      "loss": 1.644,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.5164125561714172,
      "learning_rate": 0.0015,
      "loss": 1.656,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 1.0247501134872437,
      "learning_rate": 0.0015,
      "loss": 1.6339,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5098726153373718,
      "learning_rate": 0.0015,
      "loss": 1.6439,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5366957187652588,
      "learning_rate": 0.0015,
      "loss": 1.6428,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.5414708852767944,
      "learning_rate": 0.0015,
      "loss": 1.6438,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.5522865653038025,
      "learning_rate": 0.0015,
      "loss": 1.6341,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.7166458964347839,
      "learning_rate": 0.0015,
      "loss": 1.6359,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.4831211268901825,
      "learning_rate": 0.0015,
      "loss": 1.6298,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.454694539308548,
      "learning_rate": 0.0015,
      "loss": 1.6484,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.5077452659606934,
      "learning_rate": 0.0015,
      "loss": 1.6366,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.6335849165916443,
      "learning_rate": 0.0015,
      "loss": 1.6281,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.5139283537864685,
      "learning_rate": 0.0015,
      "loss": 1.6125,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.680316686630249,
      "learning_rate": 0.0015,
      "loss": 1.6342,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.5644792318344116,
      "learning_rate": 0.0015,
      "loss": 1.6269,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.4449726641178131,
      "learning_rate": 0.0015,
      "loss": 1.641,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.6795740723609924,
      "learning_rate": 0.0015,
      "loss": 1.6442,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 1.046998143196106,
      "learning_rate": 0.0015,
      "loss": 1.6308,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.5865319967269897,
      "learning_rate": 0.0015,
      "loss": 1.6334,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.4926215708255768,
      "learning_rate": 0.0015,
      "loss": 1.6301,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.47046005725860596,
      "learning_rate": 0.0015,
      "loss": 1.6353,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.47757425904273987,
      "learning_rate": 0.0015,
      "loss": 1.6253,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.5514276623725891,
      "learning_rate": 0.0015,
      "loss": 1.627,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.7312896847724915,
      "learning_rate": 0.0015,
      "loss": 1.6243,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.4865380823612213,
      "learning_rate": 0.0015,
      "loss": 1.6234,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.47738945484161377,
      "learning_rate": 0.0015,
      "loss": 1.6317,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.456064373254776,
      "learning_rate": 0.0015,
      "loss": 1.6228,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.9585252404212952,
      "learning_rate": 0.0015,
      "loss": 1.6196,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.8172290921211243,
      "learning_rate": 0.0015,
      "loss": 1.6293,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.6002830862998962,
      "learning_rate": 0.0015,
      "loss": 1.6266,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.5298130512237549,
      "learning_rate": 0.0015,
      "loss": 1.6045,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.4257142245769501,
      "learning_rate": 0.0015,
      "loss": 1.6215,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.8374847769737244,
      "learning_rate": 0.0015,
      "loss": 1.619,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.4771282970905304,
      "learning_rate": 0.0015,
      "loss": 1.6136,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.6299778819084167,
      "learning_rate": 0.0015,
      "loss": 1.6117,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.5360429286956787,
      "learning_rate": 0.0015,
      "loss": 1.6264,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.49939456582069397,
      "learning_rate": 0.0015,
      "loss": 1.6315,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 1.1296309232711792,
      "learning_rate": 0.0015,
      "loss": 1.6277,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.5995752215385437,
      "learning_rate": 0.0015,
      "loss": 1.606,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.5539703369140625,
      "learning_rate": 0.0015,
      "loss": 1.6215,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.4397030472755432,
      "learning_rate": 0.0015,
      "loss": 1.6184,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.7149756550788879,
      "learning_rate": 0.0015,
      "loss": 1.6267,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.5444366335868835,
      "learning_rate": 0.0015,
      "loss": 1.6232,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.5677608251571655,
      "learning_rate": 0.0015,
      "loss": 1.6142,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.5791630744934082,
      "learning_rate": 0.0015,
      "loss": 1.6148,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.5979585647583008,
      "learning_rate": 0.0015,
      "loss": 1.6111,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.4636039435863495,
      "learning_rate": 0.0015,
      "loss": 1.6126,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.4689747989177704,
      "learning_rate": 0.0015,
      "loss": 1.6212,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 1.0875957012176514,
      "learning_rate": 0.0015,
      "loss": 1.6053,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.8120474219322205,
      "learning_rate": 0.0015,
      "loss": 1.6089,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.5953245162963867,
      "learning_rate": 0.0015,
      "loss": 1.6054,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.8343967795372009,
      "learning_rate": 0.0015,
      "loss": 1.6097,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.50998854637146,
      "learning_rate": 0.0015,
      "loss": 1.6093,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.6544301509857178,
      "learning_rate": 0.0015,
      "loss": 1.6149,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.5464091300964355,
      "learning_rate": 0.0015,
      "loss": 1.6105,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.6660382151603699,
      "learning_rate": 0.0015,
      "loss": 1.6008,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.7359334826469421,
      "learning_rate": 0.0015,
      "loss": 1.6077,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.5046261548995972,
      "learning_rate": 0.0015,
      "loss": 1.6176,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.5216406583786011,
      "learning_rate": 0.0015,
      "loss": 1.62,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.6172910332679749,
      "learning_rate": 0.0015,
      "loss": 1.6043,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.849410355091095,
      "learning_rate": 0.0015,
      "loss": 1.5866,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.6826720833778381,
      "learning_rate": 0.0015,
      "loss": 1.6075,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.49372223019599915,
      "learning_rate": 0.0015,
      "loss": 1.6071,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.6839715838432312,
      "learning_rate": 0.0015,
      "loss": 1.6076,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.46212440729141235,
      "learning_rate": 0.0015,
      "loss": 1.5991,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.7888845205307007,
      "learning_rate": 0.0015,
      "loss": 1.615,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.48591920733451843,
      "learning_rate": 0.0015,
      "loss": 1.6026,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.46042948961257935,
      "learning_rate": 0.0015,
      "loss": 1.6032,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.5190149545669556,
      "learning_rate": 0.0015,
      "loss": 1.6085,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.7054757475852966,
      "learning_rate": 0.0015,
      "loss": 1.6004,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.5426971316337585,
      "learning_rate": 0.0015,
      "loss": 1.5902,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.49252286553382874,
      "learning_rate": 0.0015,
      "loss": 1.5953,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.4810856580734253,
      "learning_rate": 0.0015,
      "loss": 1.6094,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.5096778869628906,
      "learning_rate": 0.0015,
      "loss": 1.5915,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.4748266637325287,
      "learning_rate": 0.0015,
      "loss": 1.6084,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 1.0327509641647339,
      "learning_rate": 0.0015,
      "loss": 1.5977,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.6517208814620972,
      "learning_rate": 0.0015,
      "loss": 1.5859,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 1.1022732257843018,
      "learning_rate": 0.0015,
      "loss": 1.5864,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.5637365579605103,
      "learning_rate": 0.0015,
      "loss": 1.5921,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.6358108520507812,
      "learning_rate": 0.0015,
      "loss": 1.5922,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.6844465136528015,
      "learning_rate": 0.0015,
      "loss": 1.5887,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.5282321572303772,
      "learning_rate": 0.0015,
      "loss": 1.5947,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.5245356559753418,
      "learning_rate": 0.0015,
      "loss": 1.5885,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.5230286121368408,
      "learning_rate": 0.0015,
      "loss": 1.5932,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.6528234481811523,
      "learning_rate": 0.0015,
      "loss": 1.5927,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.5511032938957214,
      "learning_rate": 0.0015,
      "loss": 1.5808,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.5456432104110718,
      "learning_rate": 0.0015,
      "loss": 1.5926,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.5152608156204224,
      "learning_rate": 0.0015,
      "loss": 1.5834,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.4645211696624756,
      "learning_rate": 0.0015,
      "loss": 1.5966,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.509320855140686,
      "learning_rate": 0.0015,
      "loss": 1.5838,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.48627740144729614,
      "learning_rate": 0.0015,
      "loss": 1.5858,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.48657694458961487,
      "learning_rate": 0.0015,
      "loss": 1.5801,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5566064715385437,
      "learning_rate": 0.0015,
      "loss": 1.5885,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.7756157517433167,
      "learning_rate": 0.0015,
      "loss": 1.5908,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.8503503203392029,
      "learning_rate": 0.0015,
      "loss": 1.5986,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.6750513911247253,
      "learning_rate": 0.0015,
      "loss": 1.5913,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.4914165735244751,
      "learning_rate": 0.0015,
      "loss": 1.5943,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.48406654596328735,
      "learning_rate": 0.0015,
      "loss": 1.5843,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5713526010513306,
      "learning_rate": 0.0015,
      "loss": 1.5858,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.5185654163360596,
      "learning_rate": 0.0015,
      "loss": 1.5794,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.6192831993103027,
      "learning_rate": 0.0015,
      "loss": 1.5816,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.5061731934547424,
      "learning_rate": 0.0015,
      "loss": 1.5766,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.9296662211418152,
      "learning_rate": 0.0015,
      "loss": 1.5838,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.8087186217308044,
      "learning_rate": 0.0015,
      "loss": 1.5884,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.5286755561828613,
      "learning_rate": 0.0015,
      "loss": 1.5772,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.4991167187690735,
      "learning_rate": 0.0015,
      "loss": 1.5772,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.6365343928337097,
      "learning_rate": 0.0015,
      "loss": 1.5744,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.5686373114585876,
      "learning_rate": 0.0015,
      "loss": 1.5819,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.46001219749450684,
      "learning_rate": 0.0015,
      "loss": 1.5793,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.9407059550285339,
      "learning_rate": 0.0015,
      "loss": 1.5722,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.5759108066558838,
      "learning_rate": 0.0015,
      "loss": 1.5795,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.5303200483322144,
      "learning_rate": 0.0015,
      "loss": 1.5784,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.8242844939231873,
      "learning_rate": 0.0015,
      "loss": 1.5766,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.6142957210540771,
      "learning_rate": 0.0015,
      "loss": 1.5806,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 1.1863654851913452,
      "learning_rate": 0.0015,
      "loss": 1.5896,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 1.060115098953247,
      "learning_rate": 0.0015,
      "loss": 1.5747,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 1.1346009969711304,
      "learning_rate": 0.0015,
      "loss": 1.5703,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.9194135665893555,
      "learning_rate": 0.0015,
      "loss": 1.581,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.6078530550003052,
      "learning_rate": 0.0015,
      "loss": 1.5789,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.47440075874328613,
      "learning_rate": 0.0015,
      "loss": 1.5752,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.4443087875843048,
      "learning_rate": 0.0015,
      "loss": 1.5714,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.5200234055519104,
      "learning_rate": 0.0015,
      "loss": 1.5812,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.5711065530776978,
      "learning_rate": 0.0015,
      "loss": 1.5692,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.674248218536377,
      "learning_rate": 0.0015,
      "loss": 1.5711,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.7093497514724731,
      "learning_rate": 0.0015,
      "loss": 1.5866,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.7658103704452515,
      "learning_rate": 0.0015,
      "loss": 1.5653,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5601438283920288,
      "learning_rate": 0.0015,
      "loss": 1.5691,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.479261577129364,
      "learning_rate": 0.0015,
      "loss": 1.5663,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.4836485981941223,
      "learning_rate": 0.0015,
      "loss": 1.5732,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.6983446478843689,
      "learning_rate": 0.0015,
      "loss": 1.5847,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.49608248472213745,
      "learning_rate": 0.0015,
      "loss": 1.5705,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.8320929408073425,
      "learning_rate": 0.0015,
      "loss": 1.5751,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.6829028725624084,
      "learning_rate": 0.0015,
      "loss": 1.5726,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.6526251435279846,
      "learning_rate": 0.0015,
      "loss": 1.5789,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5063625574111938,
      "learning_rate": 0.0015,
      "loss": 1.564,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.5352622866630554,
      "learning_rate": 0.0015,
      "loss": 1.5717,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.7437189817428589,
      "learning_rate": 0.0015,
      "loss": 1.5761,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.7176907658576965,
      "learning_rate": 0.0015,
      "loss": 1.5656,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.4548720419406891,
      "learning_rate": 0.0015,
      "loss": 1.5674,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.44671133160591125,
      "learning_rate": 0.0015,
      "loss": 1.5713,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.6885534524917603,
      "learning_rate": 0.0015,
      "loss": 1.553,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.7057598829269409,
      "learning_rate": 0.0015,
      "loss": 1.572,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.5296266078948975,
      "learning_rate": 0.0015,
      "loss": 1.5731,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.4908561110496521,
      "learning_rate": 0.0015,
      "loss": 1.5682,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.5058237314224243,
      "learning_rate": 0.0015,
      "loss": 1.5601,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.5028045177459717,
      "learning_rate": 0.0015,
      "loss": 1.5414,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.5612614154815674,
      "learning_rate": 0.0015,
      "loss": 1.5652,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.5249614715576172,
      "learning_rate": 0.0015,
      "loss": 1.5504,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.8108306527137756,
      "learning_rate": 0.0015,
      "loss": 1.5664,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.8428634405136108,
      "learning_rate": 0.0015,
      "loss": 1.5711,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.47969523072242737,
      "learning_rate": 0.0015,
      "loss": 1.5482,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.5133702754974365,
      "learning_rate": 0.0015,
      "loss": 1.5648,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.6160013675689697,
      "learning_rate": 0.0015,
      "loss": 1.5682,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5954915881156921,
      "learning_rate": 0.0015,
      "loss": 1.5704,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.5409538149833679,
      "learning_rate": 0.0015,
      "loss": 1.5567,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.5427380204200745,
      "learning_rate": 0.0015,
      "loss": 1.5589,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.5295004844665527,
      "learning_rate": 0.0015,
      "loss": 1.5561,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.7764521241188049,
      "learning_rate": 0.0015,
      "loss": 1.5526,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.7982751727104187,
      "learning_rate": 0.0015,
      "loss": 1.5621,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.47664445638656616,
      "learning_rate": 0.0015,
      "loss": 1.563,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 1.288316249847412,
      "learning_rate": 0.0015,
      "loss": 1.5592,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.724997341632843,
      "learning_rate": 0.0015,
      "loss": 1.5585,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.5051091313362122,
      "learning_rate": 0.0015,
      "loss": 1.5522,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.67183518409729,
      "learning_rate": 0.0015,
      "loss": 1.5719,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.6073653101921082,
      "learning_rate": 0.0015,
      "loss": 1.5577,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.7339580655097961,
      "learning_rate": 0.0015,
      "loss": 1.5544,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.493580162525177,
      "learning_rate": 0.0015,
      "loss": 1.5672,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.7570888996124268,
      "learning_rate": 0.0015,
      "loss": 1.5516,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.4718903601169586,
      "learning_rate": 0.0015,
      "loss": 1.5537,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.5226156115531921,
      "learning_rate": 0.0015,
      "loss": 1.5511,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.47913801670074463,
      "learning_rate": 0.0015,
      "loss": 1.5711,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.8636896014213562,
      "learning_rate": 0.0015,
      "loss": 1.5537,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.6331192255020142,
      "learning_rate": 0.0015,
      "loss": 1.5467,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.4399349093437195,
      "learning_rate": 0.0015,
      "loss": 1.5637,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.7091642022132874,
      "learning_rate": 0.0015,
      "loss": 1.5481,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.5377468466758728,
      "learning_rate": 0.0015,
      "loss": 1.5447,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.5458089113235474,
      "learning_rate": 0.0015,
      "loss": 1.5617,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.5292337536811829,
      "learning_rate": 0.0015,
      "loss": 1.5541,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.5375266671180725,
      "learning_rate": 0.0015,
      "loss": 1.5641,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.6999505758285522,
      "learning_rate": 0.0015,
      "loss": 1.5532,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.5864401459693909,
      "learning_rate": 0.0015,
      "loss": 1.5475,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.5812068581581116,
      "learning_rate": 0.0015,
      "loss": 1.5536,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.8708783984184265,
      "learning_rate": 0.0015,
      "loss": 1.5544,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.7917665839195251,
      "learning_rate": 0.0015,
      "loss": 1.5648,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.6949357390403748,
      "learning_rate": 0.0015,
      "loss": 1.5444,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.5002421140670776,
      "learning_rate": 0.0015,
      "loss": 1.549,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.6171126961708069,
      "learning_rate": 0.0015,
      "loss": 1.5442,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.5782960057258606,
      "learning_rate": 0.0015,
      "loss": 1.5465,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.6581275463104248,
      "learning_rate": 0.0015,
      "loss": 1.5573,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.5274462699890137,
      "learning_rate": 0.0015,
      "loss": 1.5497,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.5273053646087646,
      "learning_rate": 0.0015,
      "loss": 1.5455,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.6722688674926758,
      "learning_rate": 0.0015,
      "loss": 1.5301,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.6710203289985657,
      "learning_rate": 0.0015,
      "loss": 1.5565,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.6958735585212708,
      "learning_rate": 0.0015,
      "loss": 1.5458,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.6558091640472412,
      "learning_rate": 0.0015,
      "loss": 1.5572,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.679487943649292,
      "learning_rate": 0.0015,
      "loss": 1.5454,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.5793918371200562,
      "learning_rate": 0.0015,
      "loss": 1.5607,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.6777279376983643,
      "learning_rate": 0.0015,
      "loss": 1.5491,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.8515413999557495,
      "learning_rate": 0.0015,
      "loss": 1.5308,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.4885188937187195,
      "learning_rate": 0.0015,
      "loss": 1.5454,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.9769313335418701,
      "learning_rate": 0.0015,
      "loss": 1.5474,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.5005014538764954,
      "learning_rate": 0.0015,
      "loss": 1.543,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.4980844259262085,
      "learning_rate": 0.0015,
      "loss": 1.5585,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.5155827403068542,
      "learning_rate": 0.0015,
      "loss": 1.5529,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.4385599195957184,
      "learning_rate": 0.0015,
      "loss": 1.5474,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5264326333999634,
      "learning_rate": 0.0015,
      "loss": 1.5497,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.620452344417572,
      "learning_rate": 0.0015,
      "loss": 1.5556,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.7450008392333984,
      "learning_rate": 0.0015,
      "loss": 1.54,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.7327592372894287,
      "learning_rate": 0.0015,
      "loss": 1.5486,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.8912865519523621,
      "learning_rate": 0.0015,
      "loss": 1.5328,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.5738736987113953,
      "learning_rate": 0.0015,
      "loss": 1.5458,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.4828597605228424,
      "learning_rate": 0.0015,
      "loss": 1.537,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.6224974393844604,
      "learning_rate": 0.0015,
      "loss": 1.5478,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 1.001340627670288,
      "learning_rate": 0.0015,
      "loss": 1.5381,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.5912613272666931,
      "learning_rate": 0.0015,
      "loss": 1.5467,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.5086495876312256,
      "learning_rate": 0.0015,
      "loss": 1.5539,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.5097530484199524,
      "learning_rate": 0.0015,
      "loss": 1.545,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.747732400894165,
      "learning_rate": 0.0015,
      "loss": 1.5443,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.595024585723877,
      "learning_rate": 0.0015,
      "loss": 1.5479,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.4828443229198456,
      "learning_rate": 0.0015,
      "loss": 1.5348,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.6312142610549927,
      "learning_rate": 0.0015,
      "loss": 1.534,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.7723662257194519,
      "learning_rate": 0.0015,
      "loss": 1.5471,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.8001915812492371,
      "learning_rate": 0.0015,
      "loss": 1.5378,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.48704344034194946,
      "learning_rate": 0.0015,
      "loss": 1.5275,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.7280847430229187,
      "learning_rate": 0.0015,
      "loss": 1.5398,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.5214621424674988,
      "learning_rate": 0.0015,
      "loss": 1.5429,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.5062591433525085,
      "learning_rate": 0.0015,
      "loss": 1.5317,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.5541534423828125,
      "learning_rate": 0.0015,
      "loss": 1.5338,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.48535457253456116,
      "learning_rate": 0.0015,
      "loss": 1.5381,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.6548792123794556,
      "learning_rate": 0.0015,
      "loss": 1.5399,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.4913672208786011,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.6058489680290222,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.8538274168968201,
      "learning_rate": 0.0015,
      "loss": 1.5249,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.6668264269828796,
      "learning_rate": 0.0015,
      "loss": 1.5341,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.47492551803588867,
      "learning_rate": 0.0015,
      "loss": 1.5349,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.5480576157569885,
      "learning_rate": 0.0015,
      "loss": 1.534,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.7709435820579529,
      "learning_rate": 0.0015,
      "loss": 1.5384,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.5125923156738281,
      "learning_rate": 0.0015,
      "loss": 1.5391,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.5231842994689941,
      "learning_rate": 0.0015,
      "loss": 1.5292,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.4863499402999878,
      "learning_rate": 0.0015,
      "loss": 1.5369,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.47522854804992676,
      "learning_rate": 0.0015,
      "loss": 1.532,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.5916144847869873,
      "learning_rate": 0.0015,
      "loss": 1.5247,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.6398744583129883,
      "learning_rate": 0.0015,
      "loss": 1.5251,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.6109040975570679,
      "learning_rate": 0.0015,
      "loss": 1.541,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.5315248966217041,
      "learning_rate": 0.0015,
      "loss": 1.5444,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.7278485298156738,
      "learning_rate": 0.0015,
      "loss": 1.5283,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.4984261989593506,
      "learning_rate": 0.0015,
      "loss": 1.5282,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.6012659668922424,
      "learning_rate": 0.0015,
      "loss": 1.5306,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.612429141998291,
      "learning_rate": 0.0015,
      "loss": 1.5246,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.6125677824020386,
      "learning_rate": 0.0015,
      "loss": 1.5305,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.5053579807281494,
      "learning_rate": 0.0015,
      "loss": 1.5426,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.632543683052063,
      "learning_rate": 0.0015,
      "loss": 1.5334,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.6583037972450256,
      "learning_rate": 0.0015,
      "loss": 1.5465,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.5003215670585632,
      "learning_rate": 0.0015,
      "loss": 1.5416,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.8263731598854065,
      "learning_rate": 0.0015,
      "loss": 1.5368,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.6069696545600891,
      "learning_rate": 0.0015,
      "loss": 1.5423,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5096368789672852,
      "learning_rate": 0.0015,
      "loss": 1.5429,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.5679385662078857,
      "learning_rate": 0.0015,
      "loss": 1.516,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 1.0392451286315918,
      "learning_rate": 0.0015,
      "loss": 1.525,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.5319361686706543,
      "learning_rate": 0.0015,
      "loss": 1.5272,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.45275771617889404,
      "learning_rate": 0.0015,
      "loss": 1.5289,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 1.0679144859313965,
      "learning_rate": 0.0015,
      "loss": 1.5261,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.514147937297821,
      "learning_rate": 0.0015,
      "loss": 1.5268,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.6027361750602722,
      "learning_rate": 0.0015,
      "loss": 1.5008,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5092279314994812,
      "learning_rate": 0.0015,
      "loss": 1.5128,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.5995198488235474,
      "learning_rate": 0.0015,
      "loss": 1.517,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.7631956934928894,
      "learning_rate": 0.0015,
      "loss": 1.5262,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.5148959755897522,
      "learning_rate": 0.0015,
      "loss": 1.5172,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.5230889320373535,
      "learning_rate": 0.0015,
      "loss": 1.5247,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5816003084182739,
      "learning_rate": 0.0015,
      "loss": 1.5316,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.9210548400878906,
      "learning_rate": 0.0015,
      "loss": 1.543,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.5702227354049683,
      "learning_rate": 0.0015,
      "loss": 1.5244,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.4419424533843994,
      "learning_rate": 0.0015,
      "loss": 1.5363,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.4318900406360626,
      "learning_rate": 0.0015,
      "loss": 1.5411,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.478669673204422,
      "learning_rate": 0.0015,
      "loss": 1.5248,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.4842545688152313,
      "learning_rate": 0.0015,
      "loss": 1.5194,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.6825987100601196,
      "learning_rate": 0.0015,
      "loss": 1.5313,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.7300772070884705,
      "learning_rate": 0.0015,
      "loss": 1.5223,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.47134003043174744,
      "learning_rate": 0.0015,
      "loss": 1.526,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.5139616131782532,
      "learning_rate": 0.0015,
      "loss": 1.5195,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.5042129755020142,
      "learning_rate": 0.0015,
      "loss": 1.5151,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 1.1311372518539429,
      "learning_rate": 0.0015,
      "loss": 1.5218,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.7219382524490356,
      "learning_rate": 0.0015,
      "loss": 1.5309,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.565650224685669,
      "learning_rate": 0.0015,
      "loss": 1.5184,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.5066986680030823,
      "learning_rate": 0.0015,
      "loss": 1.5221,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5538550615310669,
      "learning_rate": 0.0015,
      "loss": 1.5211,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.5575450658798218,
      "learning_rate": 0.0015,
      "loss": 1.505,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.5579926371574402,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.5696278214454651,
      "learning_rate": 0.0015,
      "loss": 1.5369,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.7768396139144897,
      "learning_rate": 0.0015,
      "loss": 1.521,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.9120811820030212,
      "learning_rate": 0.0015,
      "loss": 1.5236,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.4908900260925293,
      "learning_rate": 0.0015,
      "loss": 1.51,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.7604925036430359,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.49551641941070557,
      "learning_rate": 0.0015,
      "loss": 1.5081,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.6577268242835999,
      "learning_rate": 0.0015,
      "loss": 1.526,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.5374493598937988,
      "learning_rate": 0.0015,
      "loss": 1.5183,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.5176852345466614,
      "learning_rate": 0.0015,
      "loss": 1.5138,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5398962497711182,
      "learning_rate": 0.0015,
      "loss": 1.5131,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.6885660886764526,
      "learning_rate": 0.0015,
      "loss": 1.5078,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.6670694351196289,
      "learning_rate": 0.0015,
      "loss": 1.5249,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.49489378929138184,
      "learning_rate": 0.0015,
      "loss": 1.5181,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.4911707639694214,
      "learning_rate": 0.0015,
      "loss": 1.5163,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.524797260761261,
      "learning_rate": 0.0015,
      "loss": 1.5178,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.5137598514556885,
      "learning_rate": 0.0015,
      "loss": 1.5273,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.4799788296222687,
      "learning_rate": 0.0015,
      "loss": 1.522,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.4757637679576874,
      "learning_rate": 0.0014834368975312174,
      "loss": 1.498,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.680515468120575,
      "learning_rate": 0.0014629899726345957,
      "loss": 1.5224,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.6326233148574829,
      "learning_rate": 0.0014428248775471316,
      "loss": 1.5227,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.5564861297607422,
      "learning_rate": 0.00142293772767289,
      "loss": 1.5154,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5272548198699951,
      "learning_rate": 0.001403324691959192,
      "loss": 1.5069,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.6546555161476135,
      "learning_rate": 0.0013839819921586025,
      "loss": 1.5155,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.4702341854572296,
      "learning_rate": 0.0013649059021010894,
      "loss": 1.5034,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5010920763015747,
      "learning_rate": 0.0013460927469762154,
      "loss": 1.5067,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.564780592918396,
      "learning_rate": 0.0013275389026252255,
      "loss": 1.5091,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.482165664434433,
      "learning_rate": 0.0013092407948428887,
      "loss": 1.5,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.44305357336997986,
      "learning_rate": 0.001291194898688966,
      "loss": 1.5095,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5847101211547852,
      "learning_rate": 0.001273397737809166,
      "loss": 1.5062,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.6253090500831604,
      "learning_rate": 0.001255845883765463,
      "loss": 1.5005,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.5310482978820801,
      "learning_rate": 0.001238535955375642,
      "loss": 1.4954,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.4950437545776367,
      "learning_rate": 0.0012214646180619506,
      "loss": 1.4917,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5149294137954712,
      "learning_rate": 0.001204628583208727,
      "loss": 1.4887,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.6041187644004822,
      "learning_rate": 0.0011880246075288827,
      "loss": 1.4972,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.4914081394672394,
      "learning_rate": 0.001171649492439115,
      "loss": 1.4875,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.6181724071502686,
      "learning_rate": 0.0011555000834437364,
      "loss": 1.4907,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.8684509992599487,
      "learning_rate": 0.0011395732695269908,
      "loss": 1.4908,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.7119426727294922,
      "learning_rate": 0.0011238659825537505,
      "loss": 1.4732,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.5087348222732544,
      "learning_rate": 0.0011083751966784717,
      "loss": 1.4737,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.8575738668441772,
      "learning_rate": 0.0010930979277622953,
      "loss": 1.4907,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.4962826073169708,
      "learning_rate": 0.0010780312327981854,
      "loss": 1.4884,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.6311264038085938,
      "learning_rate": 0.0010631722093439888,
      "loss": 1.482,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.6116746664047241,
      "learning_rate": 0.00104851799496331,
      "loss": 1.4725,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.5390072464942932,
      "learning_rate": 0.0010340657666740914,
      "loss": 1.4821,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.4817228615283966,
      "learning_rate": 0.0010198127404047975,
      "loss": 1.4635,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.46289730072021484,
      "learning_rate": 0.0010057561704580897,
      "loss": 1.4678,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.4937571883201599,
      "learning_rate": 0.0009918933489818985,
      "loss": 1.4778,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5377094149589539,
      "learning_rate": 0.0009782216054477827,
      "loss": 1.4709,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.49664416909217834,
      "learning_rate": 0.0009647383061364801,
      "loss": 1.4785,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.590204119682312,
      "learning_rate": 0.0009514408536305495,
      "loss": 1.4728,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.4623895585536957,
      "learning_rate": 0.0009383266863140042,
      "loss": 1.485,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.4693305194377899,
      "learning_rate": 0.000925393277878844,
      "loss": 1.4844,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.4652727544307709,
      "learning_rate": 0.0009126381368383879,
      "loss": 1.4676,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.46819135546684265,
      "learning_rate": 0.0009000588060473156,
      "loss": 1.46,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.4959070086479187,
      "learning_rate": 0.0008876528622283235,
      "loss": 1.4679,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.4959053695201874,
      "learning_rate": 0.0008754179155053053,
      "loss": 1.4645,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.5161389708518982,
      "learning_rate": 0.0008633516089429683,
      "loss": 1.4646,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.5014788508415222,
      "learning_rate": 0.0008514516180927928,
      "loss": 1.4604,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.523698091506958,
      "learning_rate": 0.0008397156505452524,
      "loss": 1.4589,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.4866982698440552,
      "learning_rate": 0.0008281414454882051,
      "loss": 1.4617,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5741299986839294,
      "learning_rate": 0.0008167267732713704,
      "loss": 1.4661,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.45538657903671265,
      "learning_rate": 0.0008054694349768117,
      "loss": 1.4533,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.4680151641368866,
      "learning_rate": 0.0007943672619953359,
      "loss": 1.4581,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.49672621488571167,
      "learning_rate": 0.0007834181156087356,
      "loss": 1.4481,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.5573347806930542,
      "learning_rate": 0.0007726198865777852,
      "loss": 1.4574,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.4367733895778656,
      "learning_rate": 0.0007619704947359191,
      "loss": 1.4494,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.4344213306903839,
      "learning_rate": 0.0007514678885885087,
      "loss": 1.4513,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5328627824783325,
      "learning_rate": 0.0007411100449176633,
      "loss": 1.4513,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.5584678649902344,
      "learning_rate": 0.0007308949683924791,
      "loss": 1.4539,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.5603947639465332,
      "learning_rate": 0.000720820691184658,
      "loss": 1.4487,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.4115470349788666,
      "learning_rate": 0.0007108852725894269,
      "loss": 1.4479,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.46082454919815063,
      "learning_rate": 0.000701086798651681,
      "loss": 1.447,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.44467219710350037,
      "learning_rate": 0.0006914233817972798,
      "loss": 1.4367,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.5105217695236206,
      "learning_rate": 0.0006818931604694261,
      "loss": 1.4451,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.4288281500339508,
      "learning_rate": 0.0006724942987700563,
      "loss": 1.4483,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.46505606174468994,
      "learning_rate": 0.0006632249861061732,
      "loss": 1.4471,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.45374882221221924,
      "learning_rate": 0.0006540834368410549,
      "loss": 1.4445,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.4485350549221039,
      "learning_rate": 0.0006450678899502701,
      "loss": 1.4483,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.7305462956428528,
      "learning_rate": 0.0006361766086824345,
      "loss": 1.4414,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.7582378387451172,
      "learning_rate": 0.000627407880224645,
      "loss": 1.4491,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.51595538854599,
      "learning_rate": 0.0006187600153725225,
      "loss": 1.4358,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.7243607044219971,
      "learning_rate": 0.0006102313482048055,
      "loss": 1.4363,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.5584043860435486,
      "learning_rate": 0.0006018202357624274,
      "loss": 1.4404,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.8587073683738708,
      "learning_rate": 0.0005935250577320168,
      "loss": 1.4365,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.46687424182891846,
      "learning_rate": 0.0005853442161337618,
      "loss": 1.4277,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.4609874188899994,
      "learning_rate": 0.0005772761350135759,
      "loss": 1.4365,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.42891716957092285,
      "learning_rate": 0.0005693192601395058,
      "loss": 1.4314,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.5215572118759155,
      "learning_rate": 0.000561472058702326,
      "loss": 1.4267,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.4409372806549072,
      "learning_rate": 0.000553733019020258,
      "loss": 1.4355,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.45947423577308655,
      "learning_rate": 0.0005461006502477612,
      "loss": 1.4243,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.48462724685668945,
      "learning_rate": 0.0005385734820883369,
      "loss": 1.4232,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5507255792617798,
      "learning_rate": 0.0005311500645112907,
      "loss": 1.4494,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.44307002425193787,
      "learning_rate": 0.0005238289674723993,
      "loss": 1.4274,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.4315953552722931,
      "learning_rate": 0.0005166087806384274,
      "loss": 1.4332,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.47853460907936096,
      "learning_rate": 0.0005094881131154418,
      "loss": 1.4365,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.44437679648399353,
      "learning_rate": 0.0005024655931808696,
      "loss": 1.4315,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.48510003089904785,
      "learning_rate": 0.0004955398680192508,
      "loss": 1.4223,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.4431687891483307,
      "learning_rate": 0.000488709603461632,
      "loss": 1.4171,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.4383626878261566,
      "learning_rate": 0.000481973483728553,
      "loss": 1.4189,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.4964883327484131,
      "learning_rate": 0.0004753302111765748,
      "loss": 1.4192,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.4469358026981354,
      "learning_rate": 0.0004687785060483032,
      "loss": 1.4332,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.4128322899341583,
      "learning_rate": 0.0004623171062258558,
      "loss": 1.4017,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.5924244523048401,
      "learning_rate": 0.0004559447669877288,
      "loss": 1.4194,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.4384585916996002,
      "learning_rate": 0.00044966026076901413,
      "loss": 1.4236,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.47069668769836426,
      "learning_rate": 0.00044346237692492177,
      "loss": 1.4272,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.45984041690826416,
      "learning_rate": 0.0004373499214975615,
      "loss": 1.4135,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.4670739471912384,
      "learning_rate": 0.0004313217169859396,
      "loss": 1.4184,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.5275494456291199,
      "learning_rate": 0.0004253766021191256,
      "loss": 1.425,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.459009051322937,
      "learning_rate": 0.00041951343163254497,
      "loss": 1.4232,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.5222692489624023,
      "learning_rate": 0.00041373107604735626,
      "loss": 1.4203,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.44252029061317444,
      "learning_rate": 0.0004080284214528687,
      "loss": 1.4156,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.595230221748352,
      "learning_rate": 0.0004024043692919589,
      "loss": 1.4238,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.4670724868774414,
      "learning_rate": 0.0003968578361494449,
      "loss": 1.4218,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.44240885972976685,
      "learning_rate": 0.000391387753543378,
      "loss": 1.4288,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.48701536655426025,
      "learning_rate": 0.00038599306771921023,
      "loss": 1.4099,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.6511821746826172,
      "learning_rate": 0.0003806727394468004,
      "loss": 1.4067,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.6622909307479858,
      "learning_rate": 0.0003754257438202162,
      "loss": 1.4164,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.5599359273910522,
      "learning_rate": 0.0003702510700602974,
      "loss": 1.4218,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.6296157836914062,
      "learning_rate": 0.0003651477213199393,
      "loss": 1.4009,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.48194336891174316,
      "learning_rate": 0.000360114714492061,
      "loss": 1.4026,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.4437825083732605,
      "learning_rate": 0.0003551510800202195,
      "loss": 1.4078,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.4649067223072052,
      "learning_rate": 0.0003502558617118353,
      "loss": 1.4127,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.4865807890892029,
      "learning_rate": 0.0003454281165539914,
      "loss": 1.4245,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.4343038499355316,
      "learning_rate": 0.00034066691453177176,
      "loss": 1.4223,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.4302985668182373,
      "learning_rate": 0.0003359713384491037,
      "loss": 1.4188,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.547858476638794,
      "learning_rate": 0.00033134048375206944,
      "loss": 1.4084,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.4173900783061981,
      "learning_rate": 0.0003267734583546536,
      "loss": 1.409,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.4922495186328888,
      "learning_rate": 0.00032226938246689157,
      "loss": 1.4061,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.48684826493263245,
      "learning_rate": 0.0003178273884253874,
      "loss": 1.4051,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.5038044452667236,
      "learning_rate": 0.0003134466205261674,
      "loss": 1.4208,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.4553186893463135,
      "learning_rate": 0.0003091262348598378,
      "loss": 1.4231,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.42974764108657837,
      "learning_rate": 0.0003048653991490141,
      "loss": 1.4037,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.4593367278575897,
      "learning_rate": 0.00030066329258799187,
      "loss": 1.3998,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5465492606163025,
      "learning_rate": 0.0002965191056846266,
      "loss": 1.4055,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.7009845972061157,
      "learning_rate": 0.000292432040104394,
      "loss": 1.4044,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.5206081867218018,
      "learning_rate": 0.00028840130851659853,
      "loss": 1.3981,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.43879786133766174,
      "learning_rate": 0.0002844261344427028,
      "loss": 1.4068,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.43642282485961914,
      "learning_rate": 0.0002805057521067471,
      "loss": 1.4047,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.48048076033592224,
      "learning_rate": 0.00027663940628783017,
      "loss": 1.3936,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.4274750053882599,
      "learning_rate": 0.00027282635217462393,
      "loss": 1.405,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.4558883607387543,
      "learning_rate": 0.0002690658552218937,
      "loss": 1.4144,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.4264092743396759,
      "learning_rate": 0.00026535719100899516,
      "loss": 1.3914,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.5099205374717712,
      "learning_rate": 0.00026169964510032245,
      "loss": 1.3956,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.44471004605293274,
      "learning_rate": 0.00025809251290767984,
      "loss": 1.3827,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.4197140634059906,
      "learning_rate": 0.00025453509955454957,
      "loss": 1.3892,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.4443294107913971,
      "learning_rate": 0.00025102671974223175,
      "loss": 1.392,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.41616514325141907,
      "learning_rate": 0.00024756669761782815,
      "loss": 1.4034,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.4350857436656952,
      "learning_rate": 0.0002441543666440464,
      "loss": 1.3926,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.6285386085510254,
      "learning_rate": 0.00024078906947079878,
      "loss": 1.3982,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.3976116478443146,
      "learning_rate": 0.00023747015780857005,
      "loss": 1.4051,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.43629202246665955,
      "learning_rate": 0.00023419699230353144,
      "loss": 1.4026,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.4500335454940796,
      "learning_rate": 0.00023096894241437586,
      "loss": 1.4079,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.44755616784095764,
      "learning_rate": 0.00022778538629085056,
      "loss": 1.3917,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.4716295301914215,
      "learning_rate": 0.00022464571065396427,
      "loss": 1.3913,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.40633416175842285,
      "learning_rate": 0.00022154931067784521,
      "loss": 1.3893,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.5139057040214539,
      "learning_rate": 0.00021849558987322782,
      "loss": 1.3872,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.48982763290405273,
      "learning_rate": 0.0002154839599725452,
      "loss": 1.3852,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.4648490250110626,
      "learning_rate": 0.00021251384081660544,
      "loss": 1.4016,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.46889105439186096,
      "learning_rate": 0.0002095846602428303,
      "loss": 1.3948,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.4210684299468994,
      "learning_rate": 0.00020669585397503358,
      "loss": 1.389,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.4468730092048645,
      "learning_rate": 0.0002038468655147195,
      "loss": 1.392,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.5256270170211792,
      "learning_rate": 0.00020103714603387894,
      "loss": 1.4042,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.40575408935546875,
      "learning_rate": 0.00019826615426926338,
      "loss": 1.3776,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.4396148920059204,
      "learning_rate": 0.00019553335641811625,
      "loss": 1.4004,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.40318024158477783,
      "learning_rate": 0.0001928382260353415,
      "loss": 1.3912,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.4133901000022888,
      "learning_rate": 0.00019018024393208902,
      "loss": 1.4067,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.4269525408744812,
      "learning_rate": 0.00018755889807573872,
      "loss": 1.3841,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.3969095051288605,
      "learning_rate": 0.00018497368349126262,
      "loss": 1.3892,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.4360499978065491,
      "learning_rate": 0.00018242410216394648,
      "loss": 1.4075,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.4163929522037506,
      "learning_rate": 0.0001799096629434529,
      "loss": 1.3792,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.40780094265937805,
      "learning_rate": 0.00017742988144920578,
      "loss": 1.3891,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.42543014883995056,
      "learning_rate": 0.00017498427997707976,
      "loss": 1.3851,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.5237059593200684,
      "learning_rate": 0.00017257238740737548,
      "loss": 1.3897,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.46695801615715027,
      "learning_rate": 0.00017019373911406307,
      "loss": 1.3961,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.4350501596927643,
      "learning_rate": 0.000167847876875277,
      "loss": 1.3964,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.4048670828342438,
      "learning_rate": 0.00016553434878504428,
      "loss": 1.3802,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.4566376209259033,
      "learning_rate": 0.00016325270916622947,
      "loss": 1.3824,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.42859938740730286,
      "learning_rate": 0.00016100251848467966,
      "loss": 1.3904,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.5645683407783508,
      "learning_rate": 0.0001587833432645528,
      "loss": 1.3821,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5286628603935242,
      "learning_rate": 0.00015659475600481292,
      "loss": 1.4003,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4128339886665344,
      "learning_rate": 0.00015443633509687688,
      "loss": 1.3883,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.3946399986743927,
      "learning_rate": 0.00015230766474339536,
      "loss": 1.3856,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.4302919805049896,
      "learning_rate": 0.00015020833487815416,
      "loss": 1.3938,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.43755945563316345,
      "learning_rate": 0.0001481379410870792,
      "loss": 1.3839,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.4731377959251404,
      "learning_rate": 0.00014609608453033013,
      "loss": 1.3694,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.3955100476741791,
      "learning_rate": 0.00014408237186546807,
      "loss": 1.379,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.5021982789039612,
      "learning_rate": 0.00014209641517168273,
      "loss": 1.3686,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.43806201219558716,
      "learning_rate": 0.00014013783187506265,
      "loss": 1.3822,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.4424991011619568,
      "learning_rate": 0.00013820624467489697,
      "loss": 1.4,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.4218648672103882,
      "learning_rate": 0.00013630128147099213,
      "loss": 1.3986,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.4457917809486389,
      "learning_rate": 0.00013442257529199068,
      "loss": 1.3781,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.4868776798248291,
      "learning_rate": 0.00013256976422467803,
      "loss": 1.3866,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.39558809995651245,
      "learning_rate": 0.00013074249134426366,
      "loss": 1.3851,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.4246436655521393,
      "learning_rate": 0.0001289404046456233,
      "loss": 1.395,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.4155665338039398,
      "learning_rate": 0.0001271631569754887,
      "loss": 1.3873,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.41908028721809387,
      "learning_rate": 0.0001254104059655723,
      "loss": 1.3876,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.39711275696754456,
      "learning_rate": 0.00012368181396661337,
      "loss": 1.3755,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.4529458284378052,
      "learning_rate": 0.00012197704798333364,
      "loss": 1.3784,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.4224666357040405,
      "learning_rate": 0.00012029577961028894,
      "loss": 1.3757,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.45166996121406555,
      "learning_rate": 0.00011863768496860542,
      "loss": 1.3883,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.4248290956020355,
      "learning_rate": 0.00011700244464358777,
      "loss": 1.3829,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.40860509872436523,
      "learning_rate": 0.00011538974362318715,
      "loss": 1.3879,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.42114412784576416,
      "learning_rate": 0.00011379927123731737,
      "loss": 1.3823,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.40649139881134033,
      "learning_rate": 0.0001122307210980077,
      "loss": 1.3805,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.42806771397590637,
      "learning_rate": 0.00011068379104038026,
      "loss": 1.3943,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.4196814000606537,
      "learning_rate": 0.00010915818306444116,
      "loss": 1.3726,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.4241830110549927,
      "learning_rate": 0.00010765360327767384,
      "loss": 1.3823,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.43294644355773926,
      "learning_rate": 0.00010616976183842376,
      "loss": 1.3876,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4089681804180145,
      "learning_rate": 0.00010470637290006365,
      "loss": 1.3886,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.42662033438682556,
      "learning_rate": 0.00010326315455592764,
      "loss": 1.3767,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.45557141304016113,
      "learning_rate": 0.0001018398287850053,
      "loss": 1.3683,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.4296424388885498,
      "learning_rate": 0.00010043612139838357,
      "loss": 1.3888,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.4354464113712311,
      "learning_rate": 9.905176198642719e-05,
      "loss": 1.3785,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.41208481788635254,
      "learning_rate": 9.76864838666871e-05,
      "loss": 1.3798,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.40287283062934875,
      "learning_rate": 9.634002403252676e-05,
      "loss": 1.3746,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.41079938411712646,
      "learning_rate": 9.501212310245681e-05,
      "loss": 1.3717,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.4065066874027252,
      "learning_rate": 9.370252527016777e-05,
      "loss": 1.3829,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.4285900592803955,
      "learning_rate": 9.241097825525163e-05,
      "loss": 1.3764,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.4585988521575928,
      "learning_rate": 9.113723325460276e-05,
      "loss": 1.3844,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.4628828465938568,
      "learning_rate": 8.988104489448849e-05,
      "loss": 1.3768,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.43181532621383667,
      "learning_rate": 8.864217118328042e-05,
      "loss": 1.3857,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.43050840497016907,
      "learning_rate": 8.742037346483729e-05,
      "loss": 1.3775,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.45058807730674744,
      "learning_rate": 8.62154163725303e-05,
      "loss": 1.3881,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.4476744532585144,
      "learning_rate": 8.502706778390219e-05,
      "loss": 1.3836,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.38526806235313416,
      "learning_rate": 8.38550987759513e-05,
      "loss": 1.3793,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.4117099642753601,
      "learning_rate": 8.269928358103191e-05,
      "loss": 1.3951,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.4176345765590668,
      "learning_rate": 8.155939954336243e-05,
      "loss": 1.3857,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.4403482675552368,
      "learning_rate": 8.043522707613312e-05,
      "loss": 1.3821,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.3984360694885254,
      "learning_rate": 7.932654961920486e-05,
      "loss": 1.3666,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.41802310943603516,
      "learning_rate": 7.823315359739135e-05,
      "loss": 1.3669,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.4352799952030182,
      "learning_rate": 7.715482837931577e-05,
      "loss": 1.3935,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.4808656871318817,
      "learning_rate": 7.6091366236835e-05,
      "loss": 1.3694,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.6241698265075684,
      "learning_rate": 7.504256230502289e-05,
      "loss": 1.3899,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.4258475601673126,
      "learning_rate": 7.400821454270524e-05,
      "loss": 1.383,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.391531378030777,
      "learning_rate": 7.29881236935386e-05,
      "loss": 1.3693,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.40577179193496704,
      "learning_rate": 7.198209324762562e-05,
      "loss": 1.3727,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.4294425845146179,
      "learning_rate": 7.098992940365946e-05,
      "loss": 1.3679,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.41527122259140015,
      "learning_rate": 7.001144103159e-05,
      "loss": 1.3798,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.4196541905403137,
      "learning_rate": 6.904643963580461e-05,
      "loss": 1.3858,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.39948469400405884,
      "learning_rate": 6.809473931881644e-05,
      "loss": 1.3744,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.3883092403411865,
      "learning_rate": 6.71561567454532e-05,
      "loss": 1.3719,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.5183804035186768,
      "learning_rate": 6.623051110753948e-05,
      "loss": 1.3942,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4039570689201355,
      "learning_rate": 6.531762408906607e-05,
      "loss": 1.3797,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.3927137553691864,
      "learning_rate": 6.441731983183912e-05,
      "loss": 1.3762,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.4067143499851227,
      "learning_rate": 6.352942490160292e-05,
      "loss": 1.3733,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.4075525999069214,
      "learning_rate": 6.265376825462966e-05,
      "loss": 1.3816,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.44645389914512634,
      "learning_rate": 6.179018120476945e-05,
      "loss": 1.3761,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.4125687777996063,
      "learning_rate": 6.0938497390954946e-05,
      "loss": 1.376,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.4031289517879486,
      "learning_rate": 6.009855274515339e-05,
      "loss": 1.3706,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.4718487560749054,
      "learning_rate": 5.9270185460760735e-05,
      "loss": 1.3763,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.40616875886917114,
      "learning_rate": 5.8453235961431225e-05,
      "loss": 1.3667,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.4215424954891205,
      "learning_rate": 5.764754687033678e-05,
      "loss": 1.3789,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.41651642322540283,
      "learning_rate": 5.6852962979849836e-05,
      "loss": 1.3731,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.4115312099456787,
      "learning_rate": 5.6069331221644284e-05,
      "loss": 1.3873,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.4101486802101135,
      "learning_rate": 5.529650063720842e-05,
      "loss": 1.3839,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.40822911262512207,
      "learning_rate": 5.453432234876445e-05,
      "loss": 1.3754,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.4204903244972229,
      "learning_rate": 5.37826495305886e-05,
      "loss": 1.3666,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.4433811902999878,
      "learning_rate": 5.304133738072674e-05,
      "loss": 1.3862,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.49682924151420593,
      "learning_rate": 5.2310243093099814e-05,
      "loss": 1.3777,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.45827022194862366,
      "learning_rate": 5.158922582999368e-05,
      "loss": 1.3803,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.46369123458862305,
      "learning_rate": 5.087814669492819e-05,
      "loss": 1.373,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.4033257067203522,
      "learning_rate": 5.017686870590028e-05,
      "loss": 1.3736,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.405099093914032,
      "learning_rate": 4.948525676899577e-05,
      "loss": 1.3702,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.41203173995018005,
      "learning_rate": 4.880317765236493e-05,
      "loss": 1.3816,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.4495888650417328,
      "learning_rate": 4.8130499960556755e-05,
      "loss": 1.3636,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.427905797958374,
      "learning_rate": 4.746709410920699e-05,
      "loss": 1.3674,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.49238380789756775,
      "learning_rate": 4.681283230007507e-05,
      "loss": 1.3683,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.3994861841201782,
      "learning_rate": 4.616758849642509e-05,
      "loss": 1.3769,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.41125601530075073,
      "learning_rate": 4.553123839874615e-05,
      "loss": 1.3825,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4175708591938019,
      "learning_rate": 4.490365942080736e-05,
      "loss": 1.3785,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.3910943269729614,
      "learning_rate": 4.428473066604285e-05,
      "loss": 1.3742,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.4074828624725342,
      "learning_rate": 4.367433290426233e-05,
      "loss": 1.3754,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.42127725481987,
      "learning_rate": 4.3072348548682595e-05,
      "loss": 1.3683,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.4069540202617645,
      "learning_rate": 4.247866163327575e-05,
      "loss": 1.382,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.41554608941078186,
      "learning_rate": 4.1893157790429404e-05,
      "loss": 1.3706,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.38450121879577637,
      "learning_rate": 4.1315724228915066e-05,
      "loss": 1.3681,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.47636377811431885,
      "learning_rate": 4.074624971216005e-05,
      "loss": 1.3567,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4643012285232544,
      "learning_rate": 4.018462453681889e-05,
      "loss": 1.3678,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.39813393354415894,
      "learning_rate": 3.963074051164014e-05,
      "loss": 1.376,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.39325088262557983,
      "learning_rate": 3.908449093662446e-05,
      "loss": 1.3689,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.39915817975997925,
      "learning_rate": 3.854577058246998e-05,
      "loss": 1.3668,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.42343389987945557,
      "learning_rate": 3.801447567030094e-05,
      "loss": 1.3843,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.39450734853744507,
      "learning_rate": 3.7490503851675777e-05,
      "loss": 1.369,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.401849627494812,
      "learning_rate": 3.6973754188870806e-05,
      "loss": 1.3879,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.39773768186569214,
      "learning_rate": 3.6464127135435536e-05,
      "loss": 1.3732,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.4499163031578064,
      "learning_rate": 3.596152451701616e-05,
      "loss": 1.3719,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.4330551028251648,
      "learning_rate": 3.5465849512443226e-05,
      "loss": 1.3714,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.4155457317829132,
      "learning_rate": 3.4977006635080086e-05,
      "loss": 1.3739,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.38983967900276184,
      "learning_rate": 3.449490171442838e-05,
      "loss": 1.3767,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.4134129583835602,
      "learning_rate": 3.401944187798702e-05,
      "loss": 1.3765,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.3873719274997711,
      "learning_rate": 3.355053553336137e-05,
      "loss": 1.3664,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.4332026243209839,
      "learning_rate": 3.308809235061882e-05,
      "loss": 1.3679,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.4026283025741577,
      "learning_rate": 3.263202324488772e-05,
      "loss": 1.3738,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.39853334426879883,
      "learning_rate": 3.218224035919609e-05,
      "loss": 1.3718,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.40445834398269653,
      "learning_rate": 3.173865704754688e-05,
      "loss": 1.3713,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.39231348037719727,
      "learning_rate": 3.130118785822657e-05,
      "loss": 1.3728,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.41678184270858765,
      "learning_rate": 3.08697485173437e-05,
      "loss": 1.3771,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.4322419762611389,
      "learning_rate": 3.0444255912594442e-05,
      "loss": 1.373,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.4527577757835388,
      "learning_rate": 3.002462807725185e-05,
      "loss": 1.3696,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.4193917512893677,
      "learning_rate": 2.9610784174375868e-05,
      "loss": 1.3765,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.3905143141746521,
      "learning_rate": 2.920264448124087e-05,
      "loss": 1.371,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.48122960329055786,
      "learning_rate": 2.8800130373977936e-05,
      "loss": 1.3686,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.40320464968681335,
      "learning_rate": 2.84031643124288e-05,
      "loss": 1.3676,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.43494367599487305,
      "learning_rate": 2.8011669825208517e-05,
      "loss": 1.3892,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.38821348547935486,
      "learning_rate": 2.762557149497405e-05,
      "loss": 1.3635,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.3894213140010834,
      "learning_rate": 2.724479494389592e-05,
      "loss": 1.3751,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.3882591724395752,
      "learning_rate": 2.6869266819330058e-05,
      "loss": 1.3764,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.402079313993454,
      "learning_rate": 2.6498914779687228e-05,
      "loss": 1.3762,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.3952219784259796,
      "learning_rate": 2.6133667480497115e-05,
      "loss": 1.3765,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.4140000641345978,
      "learning_rate": 2.5773454560664597e-05,
      "loss": 1.3779,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.3977586030960083,
      "learning_rate": 2.541820662891541e-05,
      "loss": 1.3577,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.3956061601638794,
      "learning_rate": 2.5067855250428616e-05,
      "loss": 1.3683,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.41328442096710205,
      "learning_rate": 2.472233293365335e-05,
      "loss": 1.3782,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.42529335618019104,
      "learning_rate": 2.4381573117307307e-05,
      "loss": 1.3662,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.413034588098526,
      "learning_rate": 2.4045510157554362e-05,
      "loss": 1.3834,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.4039822816848755,
      "learning_rate": 2.3714079315358985e-05,
      "loss": 1.3703,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.4148334562778473,
      "learning_rate": 2.338721674401494e-05,
      "loss": 1.3787,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.4227692484855652,
      "learning_rate": 2.30648594768459e-05,
      "loss": 1.3752,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.40048426389694214,
      "learning_rate": 2.2746945415075523e-05,
      "loss": 1.3696,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.40551966428756714,
      "learning_rate": 2.2433413315864803e-05,
      "loss": 1.368,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.40273284912109375,
      "learning_rate": 2.2124202780514277e-05,
      "loss": 1.371,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.396890789270401,
      "learning_rate": 2.1819254242828815e-05,
      "loss": 1.3713,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.41630643606185913,
      "learning_rate": 2.151850895764285e-05,
      "loss": 1.3697,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.38820475339889526,
      "learning_rate": 2.12219089895037e-05,
      "loss": 1.3586,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.3982078731060028,
      "learning_rate": 2.092939720151092e-05,
      "loss": 1.3547,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.4186069667339325,
      "learning_rate": 2.064091724430947e-05,
      "loss": 1.3669,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.4142046570777893,
      "learning_rate": 2.0356413545234603e-05,
      "loss": 1.3631,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.38553792238235474,
      "learning_rate": 2.0075831297606357e-05,
      "loss": 1.3746,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.39433085918426514,
      "learning_rate": 1.9799116450171627e-05,
      "loss": 1.3629,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.39306923747062683,
      "learning_rate": 1.952621569669175e-05,
      "loss": 1.3693,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.41047433018684387,
      "learning_rate": 1.9257076465673605e-05,
      "loss": 1.3689,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.44673049449920654,
      "learning_rate": 1.899164691024229e-05,
      "loss": 1.3686,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.4185176193714142,
      "learning_rate": 1.872987589815331e-05,
      "loss": 1.3641,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.39824265241622925,
      "learning_rate": 1.8471713001942538e-05,
      "loss": 1.3758,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.3979812562465668,
      "learning_rate": 1.8217108489211845e-05,
      "loss": 1.3734,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.38724735379219055,
      "learning_rate": 1.7966013313048696e-05,
      "loss": 1.3855,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.3822114169597626,
      "learning_rate": 1.7718379102577752e-05,
      "loss": 1.377,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.43143442273139954,
      "learning_rate": 1.7474158153642745e-05,
      "loss": 1.3738,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.42795488238334656,
      "learning_rate": 1.7233303419616745e-05,
      "loss": 1.3594,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.41713598370552063,
      "learning_rate": 1.699576850233916e-05,
      "loss": 1.3648,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.4310470521450043,
      "learning_rate": 1.6761507643177553e-05,
      "loss": 1.3777,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.39297470450401306,
      "learning_rate": 1.6530475714212752e-05,
      "loss": 1.3532,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.409742534160614,
      "learning_rate": 1.6302628209545423e-05,
      "loss": 1.3728,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.40209704637527466,
      "learning_rate": 1.6077921236722464e-05,
      "loss": 1.3732,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.399766743183136,
      "learning_rate": 1.5856311508281594e-05,
      "loss": 1.3689,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.3950158655643463,
      "learning_rate": 1.5637756333412454e-05,
      "loss": 1.3781,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.41801828145980835,
      "learning_rate": 1.542221360973268e-05,
      "loss": 1.3632,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.3901999890804291,
      "learning_rate": 1.5209641815177312e-05,
      "loss": 1.3744,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.1799888610839844,
      "learning_rate": 1.5e-05,
      "loss": 1.364,
      "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": 1.839757817279693e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}