{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 0.9864585399627686,
      "learning_rate": 3.6e-06,
      "loss": 4.570732879638672,
      "step": 10
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.9653582572937012,
      "learning_rate": 7.6e-06,
      "loss": 4.4472816467285154,
      "step": 20
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.865614116191864,
      "learning_rate": 1.16e-05,
      "loss": 4.528660583496094,
      "step": 30
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.8203653693199158,
      "learning_rate": 1.56e-05,
      "loss": 4.469075012207031,
      "step": 40
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.8913492560386658,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 4.3623191833496096,
      "step": 50
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.7883138060569763,
      "learning_rate": 2.36e-05,
      "loss": 4.489534759521485,
      "step": 60
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.8199988007545471,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 4.3617607116699215,
      "step": 70
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.5262730121612549,
      "learning_rate": 3.16e-05,
      "loss": 4.352099609375,
      "step": 80
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.7354528903961182,
      "learning_rate": 3.56e-05,
      "loss": 4.413776016235351,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9181965589523315,
      "learning_rate": 3.960000000000001e-05,
      "loss": 4.542580413818359,
      "step": 100
    },
    {
      "epoch": 0.011,
      "grad_norm": 1.0277456045150757,
      "learning_rate": 4.36e-05,
      "loss": 4.520787048339844,
      "step": 110
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.9300506711006165,
      "learning_rate": 4.72e-05,
      "loss": 4.408308410644532,
      "step": 120
    },
    {
      "epoch": 0.013,
      "grad_norm": 1.7911187410354614,
      "learning_rate": 5.1200000000000004e-05,
      "loss": 4.257415771484375,
      "step": 130
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.6802721619606018,
      "learning_rate": 5.520000000000001e-05,
      "loss": 4.411191177368164,
      "step": 140
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.747800886631012,
      "learning_rate": 5.92e-05,
      "loss": 4.295530700683594,
      "step": 150
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.8360177278518677,
      "learning_rate": 6.280000000000001e-05,
      "loss": 4.434665679931641,
      "step": 160
    },
    {
      "epoch": 0.017,
      "grad_norm": 1.1511788368225098,
      "learning_rate": 6.680000000000001e-05,
      "loss": 4.305290603637696,
      "step": 170
    },
    {
      "epoch": 0.018,
      "grad_norm": 1.0238330364227295,
      "learning_rate": 7.08e-05,
      "loss": 4.361919021606445,
      "step": 180
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.8761144280433655,
      "learning_rate": 7.48e-05,
      "loss": 4.483421325683594,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7632453441619873,
      "learning_rate": 7.88e-05,
      "loss": 4.434427642822266,
      "step": 200
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.7385545969009399,
      "learning_rate": 8.28e-05,
      "loss": 4.31959228515625,
      "step": 210
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.9063302874565125,
      "learning_rate": 8.680000000000001e-05,
      "loss": 4.432417297363282,
      "step": 220
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.872031569480896,
      "learning_rate": 9.080000000000001e-05,
      "loss": 4.314726257324219,
      "step": 230
    },
    {
      "epoch": 0.024,
      "grad_norm": 1.190935730934143,
      "learning_rate": 9.48e-05,
      "loss": 4.348759460449219,
      "step": 240
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.71515291929245,
      "learning_rate": 9.88e-05,
      "loss": 4.343405914306641,
      "step": 250
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.9696555733680725,
      "learning_rate": 0.0001028,
      "loss": 4.372357940673828,
      "step": 260
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.7644748091697693,
      "learning_rate": 0.00010680000000000001,
      "loss": 4.257550430297852,
      "step": 270
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.7904483079910278,
      "learning_rate": 0.00011080000000000001,
      "loss": 4.393317413330078,
      "step": 280
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.7877936363220215,
      "learning_rate": 0.0001148,
      "loss": 4.269060134887695,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7665325403213501,
      "learning_rate": 0.0001188,
      "loss": 4.257502746582031,
      "step": 300
    },
    {
      "epoch": 0.031,
      "grad_norm": 1.269228458404541,
      "learning_rate": 0.0001228,
      "loss": 4.269769668579102,
      "step": 310
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.8273214101791382,
      "learning_rate": 0.00012680000000000002,
      "loss": 4.264021682739258,
      "step": 320
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.8829242587089539,
      "learning_rate": 0.0001308,
      "loss": 4.224617767333984,
      "step": 330
    },
    {
      "epoch": 0.034,
      "grad_norm": 1.170655369758606,
      "learning_rate": 0.00013480000000000002,
      "loss": 4.237782287597656,
      "step": 340
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.9054440259933472,
      "learning_rate": 0.00013879999999999999,
      "loss": 4.303476715087891,
      "step": 350
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.908544659614563,
      "learning_rate": 0.0001428,
      "loss": 4.118827819824219,
      "step": 360
    },
    {
      "epoch": 0.037,
      "grad_norm": 1.1054596900939941,
      "learning_rate": 0.00014680000000000002,
      "loss": 4.252466201782227,
      "step": 370
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.8284249901771545,
      "learning_rate": 0.0001508,
      "loss": 4.308754730224609,
      "step": 380
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.9838734865188599,
      "learning_rate": 0.0001544,
      "loss": 4.270053863525391,
      "step": 390
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1506909132003784,
      "learning_rate": 0.00015840000000000003,
      "loss": 4.273390579223633,
      "step": 400
    },
    {
      "epoch": 0.041,
      "grad_norm": 1.1765968799591064,
      "learning_rate": 0.00016240000000000002,
      "loss": 4.30668830871582,
      "step": 410
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.832719624042511,
      "learning_rate": 0.0001664,
      "loss": 4.189241409301758,
      "step": 420
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.7932992577552795,
      "learning_rate": 0.0001704,
      "loss": 4.213620376586914,
      "step": 430
    },
    {
      "epoch": 0.044,
      "grad_norm": 1.056486964225769,
      "learning_rate": 0.0001744,
      "loss": 4.136186218261718,
      "step": 440
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.8109835982322693,
      "learning_rate": 0.0001784,
      "loss": 4.234612655639649,
      "step": 450
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.7274177074432373,
      "learning_rate": 0.00018240000000000002,
      "loss": 4.074836730957031,
      "step": 460
    },
    {
      "epoch": 0.047,
      "grad_norm": 1.00718092918396,
      "learning_rate": 0.00018640000000000003,
      "loss": 4.227546691894531,
      "step": 470
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.7283305525779724,
      "learning_rate": 0.0001904,
      "loss": 4.220362854003906,
      "step": 480
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.8631565570831299,
      "learning_rate": 0.0001944,
      "loss": 4.183969116210937,
      "step": 490
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9393625259399414,
      "learning_rate": 0.0001984,
      "loss": 4.123226547241211,
      "step": 500
    },
    {
      "epoch": 0.051,
      "grad_norm": 1.0558165311813354,
      "learning_rate": 0.00019999980315476906,
      "loss": 4.196321868896485,
      "step": 510
    },
    {
      "epoch": 0.052,
      "grad_norm": 1.348644733428955,
      "learning_rate": 0.00019999860021449744,
      "loss": 4.480307769775391,
      "step": 520
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.7213611602783203,
      "learning_rate": 0.0001999963037055551,
      "loss": 4.237163543701172,
      "step": 530
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.8646374344825745,
      "learning_rate": 0.00019999291365305623,
      "loss": 4.309125137329102,
      "step": 540
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.9939764738082886,
      "learning_rate": 0.00019998843009407392,
      "loss": 4.144802856445312,
      "step": 550
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.7328234314918518,
      "learning_rate": 0.0001999828530776397,
      "loss": 4.328020477294922,
      "step": 560
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.9008225202560425,
      "learning_rate": 0.00019997618266474285,
      "loss": 4.094429397583008,
      "step": 570
    },
    {
      "epoch": 0.058,
      "grad_norm": 1.0807420015335083,
      "learning_rate": 0.00019996841892833,
      "loss": 4.087076568603516,
      "step": 580
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.8722774982452393,
      "learning_rate": 0.0001999595619533041,
      "loss": 4.256121826171875,
      "step": 590
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8614450693130493,
      "learning_rate": 0.00019994961183652366,
      "loss": 4.161228179931641,
      "step": 600
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.7335379123687744,
      "learning_rate": 0.00019993856868680146,
      "loss": 4.139397430419922,
      "step": 610
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.7370278239250183,
      "learning_rate": 0.00019992643262490368,
      "loss": 4.107318878173828,
      "step": 620
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.7478401064872742,
      "learning_rate": 0.00019991320378354838,
      "loss": 4.03961067199707,
      "step": 630
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.7328233122825623,
      "learning_rate": 0.00019989888230740398,
      "loss": 4.134631729125976,
      "step": 640
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.7537006735801697,
      "learning_rate": 0.0001998834683530879,
      "loss": 4.0941509246826175,
      "step": 650
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.7562767863273621,
      "learning_rate": 0.00019986696208916467,
      "loss": 4.0329750061035154,
      "step": 660
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.9703223705291748,
      "learning_rate": 0.00019984936369614413,
      "loss": 4.171240615844726,
      "step": 670
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.9317129254341125,
      "learning_rate": 0.00019983067336647948,
      "loss": 4.03155288696289,
      "step": 680
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.7133350968360901,
      "learning_rate": 0.00019981089130456517,
      "loss": 4.033761215209961,
      "step": 690
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8120133876800537,
      "learning_rate": 0.00019979001772673472,
      "loss": 4.046437835693359,
      "step": 700
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.899259626865387,
      "learning_rate": 0.00019976805286125814,
      "loss": 4.054334259033203,
      "step": 710
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.8513051867485046,
      "learning_rate": 0.00019974499694833974,
      "loss": 4.01842155456543,
      "step": 720
    },
    {
      "epoch": 0.073,
      "grad_norm": 1.0023386478424072,
      "learning_rate": 0.00019972085024011526,
      "loss": 4.078408813476562,
      "step": 730
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.7001808881759644,
      "learning_rate": 0.00019969561300064927,
      "loss": 4.034831237792969,
      "step": 740
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.6555860042572021,
      "learning_rate": 0.0001996692855059321,
      "loss": 3.9754600524902344,
      "step": 750
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.7931729555130005,
      "learning_rate": 0.00019964186804387714,
      "loss": 4.084543609619141,
      "step": 760
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.7387161254882812,
      "learning_rate": 0.00019961336091431727,
      "loss": 3.9783470153808596,
      "step": 770
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.7579309940338135,
      "learning_rate": 0.00019958376442900187,
      "loss": 4.114356231689453,
      "step": 780
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.6719383001327515,
      "learning_rate": 0.00019955307891159345,
      "loss": 3.928887939453125,
      "step": 790
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8183976411819458,
      "learning_rate": 0.00019952130469766386,
      "loss": 4.034050369262696,
      "step": 800
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.8277682662010193,
      "learning_rate": 0.0001994884421346908,
      "loss": 3.9189529418945312,
      "step": 810
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.6997612118721008,
      "learning_rate": 0.00019945449158205404,
      "loss": 3.9775638580322266,
      "step": 820
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.759364128112793,
      "learning_rate": 0.00019941945341103132,
      "loss": 3.9857391357421874,
      "step": 830
    },
    {
      "epoch": 0.084,
      "grad_norm": 1.3590269088745117,
      "learning_rate": 0.00019938332800479452,
      "loss": 4.000398254394531,
      "step": 840
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.8666356801986694,
      "learning_rate": 0.0001993461157584053,
      "loss": 4.033834457397461,
      "step": 850
    },
    {
      "epoch": 0.086,
      "grad_norm": 1.0900291204452515,
      "learning_rate": 0.0001993078170788108,
      "loss": 4.094419860839844,
      "step": 860
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.8870823383331299,
      "learning_rate": 0.0001992684323848392,
      "loss": 3.952312469482422,
      "step": 870
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.7740331292152405,
      "learning_rate": 0.00019922796210719527,
      "loss": 4.108131408691406,
      "step": 880
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.8540732264518738,
      "learning_rate": 0.00019918640668845544,
      "loss": 4.127989196777344,
      "step": 890
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7788583636283875,
      "learning_rate": 0.0001991437665830632,
      "loss": 3.9711688995361327,
      "step": 900
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.6887449622154236,
      "learning_rate": 0.0001991000422573238,
      "loss": 3.9237648010253907,
      "step": 910
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.6372981071472168,
      "learning_rate": 0.0001990552341893995,
      "loss": 4.003054809570313,
      "step": 920
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.7574194073677063,
      "learning_rate": 0.00019900934286930415,
      "loss": 3.9570274353027344,
      "step": 930
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.8783987760543823,
      "learning_rate": 0.0001989623687988979,
      "loss": 3.9508705139160156,
      "step": 940
    },
    {
      "epoch": 0.095,
      "grad_norm": 1.1051958799362183,
      "learning_rate": 0.00019891431249188161,
      "loss": 4.038083267211914,
      "step": 950
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.6932600140571594,
      "learning_rate": 0.0001988651744737914,
      "loss": 3.9600696563720703,
      "step": 960
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.7513547539710999,
      "learning_rate": 0.00019881495528199276,
      "loss": 3.945067596435547,
      "step": 970
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.7377809882164001,
      "learning_rate": 0.0001987636554656747,
      "loss": 3.959481430053711,
      "step": 980
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.6857086420059204,
      "learning_rate": 0.0001987112755858438,
      "loss": 3.981589508056641,
      "step": 990
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.708402693271637,
      "learning_rate": 0.00019865781621531805,
      "loss": 3.954851531982422,
      "step": 1000
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.9832116961479187,
      "learning_rate": 0.00019860327793872054,
      "loss": 3.9482818603515626,
      "step": 1010
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.8365011215209961,
      "learning_rate": 0.00019854766135247315,
      "loss": 3.8467781066894533,
      "step": 1020
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.6943666338920593,
      "learning_rate": 0.0001984909670647899,
      "loss": 3.9436370849609377,
      "step": 1030
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.6381665468215942,
      "learning_rate": 0.00019843319569567053,
      "loss": 3.854142761230469,
      "step": 1040
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.7425286769866943,
      "learning_rate": 0.00019837434787689338,
      "loss": 3.8861854553222654,
      "step": 1050
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.896165668964386,
      "learning_rate": 0.00019831442425200882,
      "loss": 3.964174270629883,
      "step": 1060
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.9097633957862854,
      "learning_rate": 0.00019825342547633202,
      "loss": 3.9671138763427733,
      "step": 1070
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.8225612640380859,
      "learning_rate": 0.00019819135221693575,
      "loss": 3.8377418518066406,
      "step": 1080
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.918106734752655,
      "learning_rate": 0.00019812820515264335,
      "loss": 3.985578918457031,
      "step": 1090
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8639251589775085,
      "learning_rate": 0.0001980639849740209,
      "loss": 3.9658134460449217,
      "step": 1100
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.6783847808837891,
      "learning_rate": 0.00019799869238337002,
      "loss": 3.9555068969726563,
      "step": 1110
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.0506054162979126,
      "learning_rate": 0.00019793232809472001,
      "loss": 3.9431129455566407,
      "step": 1120
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.7837322354316711,
      "learning_rate": 0.00019786489283382013,
      "loss": 3.9178836822509764,
      "step": 1130
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.674963116645813,
      "learning_rate": 0.00019779638733813154,
      "loss": 3.789346694946289,
      "step": 1140
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.7334838509559631,
      "learning_rate": 0.00019772681235681936,
      "loss": 3.9166748046875,
      "step": 1150
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.8206011056900024,
      "learning_rate": 0.00019765616865074449,
      "loss": 3.902676010131836,
      "step": 1160
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.6592251658439636,
      "learning_rate": 0.00019758445699245516,
      "loss": 3.8085838317871095,
      "step": 1170
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.7758939862251282,
      "learning_rate": 0.00019751167816617856,
      "loss": 3.845978927612305,
      "step": 1180
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.7027841806411743,
      "learning_rate": 0.00019743783296781228,
      "loss": 3.8182716369628906,
      "step": 1190
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7206117510795593,
      "learning_rate": 0.0001973629222049156,
      "loss": 3.7027477264404296,
      "step": 1200
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.7593057751655579,
      "learning_rate": 0.0001972869466967006,
      "loss": 3.7599334716796875,
      "step": 1210
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.7293952107429504,
      "learning_rate": 0.00019720990727402327,
      "loss": 3.9083782196044923,
      "step": 1220
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.695371150970459,
      "learning_rate": 0.00019713180477937443,
      "loss": 3.732463073730469,
      "step": 1230
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.589763343334198,
      "learning_rate": 0.00019705264006687035,
      "loss": 3.8740089416503904,
      "step": 1240
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.7158348560333252,
      "learning_rate": 0.00019697241400224377,
      "loss": 3.847222900390625,
      "step": 1250
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.6988404393196106,
      "learning_rate": 0.00019689112746283397,
      "loss": 3.735087585449219,
      "step": 1260
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.6927365660667419,
      "learning_rate": 0.0001968087813375776,
      "loss": 3.708781051635742,
      "step": 1270
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.6658136248588562,
      "learning_rate": 0.00019672537652699865,
      "loss": 3.8875518798828126,
      "step": 1280
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.859686553478241,
      "learning_rate": 0.00019664091394319875,
      "loss": 3.7763099670410156,
      "step": 1290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.61698317527771,
      "learning_rate": 0.00019655539450984724,
      "loss": 3.796756362915039,
      "step": 1300
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.926706075668335,
      "learning_rate": 0.0001964688191621709,
      "loss": 3.71102294921875,
      "step": 1310
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.793457567691803,
      "learning_rate": 0.00019638118884694386,
      "loss": 3.858362579345703,
      "step": 1320
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.847754180431366,
      "learning_rate": 0.00019629250452247725,
      "loss": 3.7236656188964843,
      "step": 1330
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.6211801767349243,
      "learning_rate": 0.0001962027671586086,
      "loss": 3.6885189056396483,
      "step": 1340
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.6702644228935242,
      "learning_rate": 0.00019611197773669134,
      "loss": 3.711442565917969,
      "step": 1350
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.8904920816421509,
      "learning_rate": 0.00019602013724958405,
      "loss": 3.637826156616211,
      "step": 1360
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.8204942345619202,
      "learning_rate": 0.0001959272467016396,
      "loss": 3.7548507690429687,
      "step": 1370
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.7706054449081421,
      "learning_rate": 0.0001958333071086941,
      "loss": 3.721475601196289,
      "step": 1380
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.824061393737793,
      "learning_rate": 0.00019573831949805596,
      "loss": 3.753992462158203,
      "step": 1390
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6780307292938232,
      "learning_rate": 0.00019564228490849436,
      "loss": 3.709253692626953,
      "step": 1400
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.678337574005127,
      "learning_rate": 0.0001955452043902283,
      "loss": 3.716749572753906,
      "step": 1410
    },
    {
      "epoch": 0.142,
      "grad_norm": 1.3171943426132202,
      "learning_rate": 0.00019544707900491469,
      "loss": 3.793317413330078,
      "step": 1420
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.6768912672996521,
      "learning_rate": 0.0001953479098256371,
      "loss": 3.7399524688720702,
      "step": 1430
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.6930646300315857,
      "learning_rate": 0.00019524769793689367,
      "loss": 3.7181812286376954,
      "step": 1440
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.7317301630973816,
      "learning_rate": 0.00019514644443458567,
      "loss": 3.713946533203125,
      "step": 1450
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.7954129576683044,
      "learning_rate": 0.00019504415042600516,
      "loss": 3.6904296875,
      "step": 1460
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.6594303250312805,
      "learning_rate": 0.00019494081702982297,
      "loss": 3.682525634765625,
      "step": 1470
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.7191218137741089,
      "learning_rate": 0.00019483644537607667,
      "loss": 3.7779449462890624,
      "step": 1480
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.6406489014625549,
      "learning_rate": 0.00019473103660615785,
      "loss": 3.791914367675781,
      "step": 1490
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7773824334144592,
      "learning_rate": 0.00019462459187279994,
      "loss": 3.72423095703125,
      "step": 1500
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.7769690752029419,
      "learning_rate": 0.0001945171123400655,
      "loss": 3.76646728515625,
      "step": 1510
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.7133404016494751,
      "learning_rate": 0.0001944085991833335,
      "loss": 3.732075881958008,
      "step": 1520
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.7259360551834106,
      "learning_rate": 0.00019429905358928646,
      "loss": 3.7340469360351562,
      "step": 1530
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.8133737444877625,
      "learning_rate": 0.00019418847675589738,
      "loss": 3.681981658935547,
      "step": 1540
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.8349258303642273,
      "learning_rate": 0.0001940768698924168,
      "loss": 3.6743228912353514,
      "step": 1550
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.6476967334747314,
      "learning_rate": 0.0001939642342193596,
      "loss": 3.833597183227539,
      "step": 1560
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.6458984017372131,
      "learning_rate": 0.00019385057096849134,
      "loss": 3.671551513671875,
      "step": 1570
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.7774168252944946,
      "learning_rate": 0.00019373588138281512,
      "loss": 3.673431396484375,
      "step": 1580
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.729218065738678,
      "learning_rate": 0.00019362016671655796,
      "loss": 3.7072010040283203,
      "step": 1590
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6772584915161133,
      "learning_rate": 0.00019350342823515684,
      "loss": 3.777060699462891,
      "step": 1600
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.6383146047592163,
      "learning_rate": 0.00019338566721524515,
      "loss": 3.6807861328125,
      "step": 1610
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.9132865071296692,
      "learning_rate": 0.00019326688494463845,
      "loss": 3.7850833892822267,
      "step": 1620
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.8380615711212158,
      "learning_rate": 0.0001931470827223207,
      "loss": 3.8162258148193358,
      "step": 1630
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.7553097009658813,
      "learning_rate": 0.0001930262618584298,
      "loss": 3.8104087829589846,
      "step": 1640
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.6871218085289001,
      "learning_rate": 0.0001929044236742433,
      "loss": 3.7746299743652343,
      "step": 1650
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.6442969441413879,
      "learning_rate": 0.00019278156950216412,
      "loss": 3.7820144653320313,
      "step": 1660
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.8019017577171326,
      "learning_rate": 0.00019265770068570582,
      "loss": 3.6772369384765624,
      "step": 1670
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.8048709034919739,
      "learning_rate": 0.00019253281857947783,
      "loss": 3.736543655395508,
      "step": 1680
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.6705429553985596,
      "learning_rate": 0.00019240692454917095,
      "loss": 3.769464874267578,
      "step": 1690
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5765128135681152,
      "learning_rate": 0.00019228001997154213,
      "loss": 3.7056846618652344,
      "step": 1700
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.604391872882843,
      "learning_rate": 0.00019215210623439944,
      "loss": 3.6604618072509765,
      "step": 1710
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.689091145992279,
      "learning_rate": 0.00019202318473658705,
      "loss": 3.6473930358886717,
      "step": 1720
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.7933512330055237,
      "learning_rate": 0.0001918932568879698,
      "loss": 3.6436813354492186,
      "step": 1730
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.7541588544845581,
      "learning_rate": 0.00019176232410941782,
      "loss": 3.6962326049804686,
      "step": 1740
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.7435059547424316,
      "learning_rate": 0.00019163038783279098,
      "loss": 3.7371444702148438,
      "step": 1750
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.6405462622642517,
      "learning_rate": 0.00019149744950092333,
      "loss": 3.6468109130859374,
      "step": 1760
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.8383358120918274,
      "learning_rate": 0.00019136351056760708,
      "loss": 3.7069969177246094,
      "step": 1770
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.7338976263999939,
      "learning_rate": 0.00019122857249757703,
      "loss": 3.6671722412109373,
      "step": 1780
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.8859654664993286,
      "learning_rate": 0.00019109263676649426,
      "loss": 3.777562713623047,
      "step": 1790
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.633763313293457,
      "learning_rate": 0.0001909557048609301,
      "loss": 3.687790298461914,
      "step": 1800
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.7197203636169434,
      "learning_rate": 0.00019081777827835002,
      "loss": 3.5835197448730467,
      "step": 1810
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.6443367600440979,
      "learning_rate": 0.00019067885852709694,
      "loss": 3.6920188903808593,
      "step": 1820
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.62919682264328,
      "learning_rate": 0.00019053894712637507,
      "loss": 3.72003059387207,
      "step": 1830
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.5908259749412537,
      "learning_rate": 0.00019039804560623305,
      "loss": 3.741707611083984,
      "step": 1840
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.6945422291755676,
      "learning_rate": 0.00019025615550754733,
      "loss": 3.687003326416016,
      "step": 1850
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.8403092622756958,
      "learning_rate": 0.00019011327838200527,
      "loss": 3.7636627197265624,
      "step": 1860
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.7481213212013245,
      "learning_rate": 0.00018996941579208827,
      "loss": 3.7506473541259764,
      "step": 1870
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.718259871006012,
      "learning_rate": 0.00018982456931105452,
      "loss": 3.7587440490722654,
      "step": 1880
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.7435288429260254,
      "learning_rate": 0.000189678740522922,
      "loss": 3.7968624114990233,
      "step": 1890
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7345373034477234,
      "learning_rate": 0.00018953193102245086,
      "loss": 3.735291290283203,
      "step": 1900
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.7988681793212891,
      "learning_rate": 0.0001893841424151264,
      "loss": 3.7723670959472657,
      "step": 1910
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.5639992952346802,
      "learning_rate": 0.00018923537631714105,
      "loss": 3.669097900390625,
      "step": 1920
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.6562656164169312,
      "learning_rate": 0.0001890856343553771,
      "loss": 3.7212371826171875,
      "step": 1930
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.5875832438468933,
      "learning_rate": 0.00018893491816738859,
      "loss": 3.6742832183837892,
      "step": 1940
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.8865353465080261,
      "learning_rate": 0.00018878322940138367,
      "loss": 3.9563770294189453,
      "step": 1950
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.8625441789627075,
      "learning_rate": 0.00018863056971620633,
      "loss": 3.6709442138671875,
      "step": 1960
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.5986853241920471,
      "learning_rate": 0.00018847694078131844,
      "loss": 3.6971904754638674,
      "step": 1970
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.8145214915275574,
      "learning_rate": 0.00018832234427678142,
      "loss": 3.6733352661132814,
      "step": 1980
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.7707452774047852,
      "learning_rate": 0.00018818238154774588,
      "loss": 3.642641067504883,
      "step": 1990
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.695024847984314,
      "learning_rate": 0.00018802595132732635,
      "loss": 3.7839202880859375,
      "step": 2000
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.6134204864501953,
      "learning_rate": 0.00018786855846920615,
      "loss": 3.731365966796875,
      "step": 2010
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.6342029571533203,
      "learning_rate": 0.0001877102046946083,
      "loss": 3.7211891174316407,
      "step": 2020
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.6614246368408203,
      "learning_rate": 0.0001875508917352643,
      "loss": 3.692976379394531,
      "step": 2030
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.8162015676498413,
      "learning_rate": 0.00018739062133339517,
      "loss": 3.6312129974365233,
      "step": 2040
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.7029600143432617,
      "learning_rate": 0.0001872293952416924,
      "loss": 3.6751358032226564,
      "step": 2050
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.8595512509346008,
      "learning_rate": 0.00018706721522329862,
      "loss": 3.7104766845703123,
      "step": 2060
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.7532731890678406,
      "learning_rate": 0.00018690408305178856,
      "loss": 3.881061553955078,
      "step": 2070
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.820124626159668,
      "learning_rate": 0.00018674000051114952,
      "loss": 3.7093475341796873,
      "step": 2080
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.7748057246208191,
      "learning_rate": 0.00018657496939576186,
      "loss": 3.6785037994384764,
      "step": 2090
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8756980895996094,
      "learning_rate": 0.00018640899151037944,
      "loss": 3.8046417236328125,
      "step": 2100
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.7910929918289185,
      "learning_rate": 0.0001862420686701098,
      "loss": 3.7161251068115235,
      "step": 2110
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.8564070463180542,
      "learning_rate": 0.0001860742027003944,
      "loss": 3.717102813720703,
      "step": 2120
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.654425859451294,
      "learning_rate": 0.00018590539543698854,
      "loss": 3.6956653594970703,
      "step": 2130
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.6396397352218628,
      "learning_rate": 0.00018573564872594145,
      "loss": 3.6549213409423826,
      "step": 2140
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.7105404734611511,
      "learning_rate": 0.00018556496442357595,
      "loss": 3.6959983825683596,
      "step": 2150
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.6370300650596619,
      "learning_rate": 0.00018539334439646824,
      "loss": 3.646870803833008,
      "step": 2160
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.7575487494468689,
      "learning_rate": 0.00018522079052142747,
      "loss": 3.638621520996094,
      "step": 2170
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.6455351710319519,
      "learning_rate": 0.0001850473046854751,
      "loss": 3.8218017578125,
      "step": 2180
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.6081588864326477,
      "learning_rate": 0.00018487288878582447,
      "loss": 3.680555725097656,
      "step": 2190
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.617534875869751,
      "learning_rate": 0.00018469754472985993,
      "loss": 3.683754730224609,
      "step": 2200
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.6604626178741455,
      "learning_rate": 0.00018452127443511598,
      "loss": 3.6632713317871093,
      "step": 2210
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.6190357804298401,
      "learning_rate": 0.0001843440798292563,
      "loss": 3.7039043426513674,
      "step": 2220
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.7622014284133911,
      "learning_rate": 0.00018416596285005272,
      "loss": 3.6635696411132814,
      "step": 2230
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.6593033075332642,
      "learning_rate": 0.00018398692544536397,
      "loss": 3.654493713378906,
      "step": 2240
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.6298798322677612,
      "learning_rate": 0.00018380696957311449,
      "loss": 3.6862766265869142,
      "step": 2250
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.7487024664878845,
      "learning_rate": 0.0001836260972012728,
      "loss": 3.6696929931640625,
      "step": 2260
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.703930139541626,
      "learning_rate": 0.0001834443103078302,
      "loss": 3.7097728729248045,
      "step": 2270
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.7945974469184875,
      "learning_rate": 0.00018326161088077903,
      "loss": 3.684752655029297,
      "step": 2280
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.9177495837211609,
      "learning_rate": 0.00018307800091809097,
      "loss": 3.6755802154541017,
      "step": 2290
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.8014151453971863,
      "learning_rate": 0.00018289348242769515,
      "loss": 3.4207176208496093,
      "step": 2300
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.6547634601593018,
      "learning_rate": 0.00018270805742745617,
      "loss": 3.5766365051269533,
      "step": 2310
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.6312336921691895,
      "learning_rate": 0.00018252172794515223,
      "loss": 3.6193904876708984,
      "step": 2320
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.5844429135322571,
      "learning_rate": 0.00018233449601845258,
      "loss": 3.5989307403564452,
      "step": 2330
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.7731897830963135,
      "learning_rate": 0.00018214636369489563,
      "loss": 3.5829071044921874,
      "step": 2340
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.9310140609741211,
      "learning_rate": 0.00018195733303186633,
      "loss": 3.579779052734375,
      "step": 2350
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.6296517252922058,
      "learning_rate": 0.0001817674060965737,
      "loss": 3.6821571350097657,
      "step": 2360
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.6074180603027344,
      "learning_rate": 0.00018157658496602833,
      "loss": 3.443158721923828,
      "step": 2370
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.6092249155044556,
      "learning_rate": 0.0001813848717270195,
      "loss": 3.6482177734375,
      "step": 2380
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.8526555895805359,
      "learning_rate": 0.00018119226847609245,
      "loss": 3.603512191772461,
      "step": 2390
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.732833206653595,
      "learning_rate": 0.0001809987773195255,
      "loss": 3.630801773071289,
      "step": 2400
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.7713726162910461,
      "learning_rate": 0.00018080440037330695,
      "loss": 3.6026840209960938,
      "step": 2410
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.8152686953544617,
      "learning_rate": 0.00018060913976311192,
      "loss": 3.6279972076416014,
      "step": 2420
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.7530510425567627,
      "learning_rate": 0.00018041299762427916,
      "loss": 3.5415550231933595,
      "step": 2430
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.6138831377029419,
      "learning_rate": 0.00018021597610178768,
      "loss": 3.5664520263671875,
      "step": 2440
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.6716843247413635,
      "learning_rate": 0.0001800180773502333,
      "loss": 3.600559616088867,
      "step": 2450
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.5920769572257996,
      "learning_rate": 0.00017981930353380503,
      "loss": 3.5558128356933594,
      "step": 2460
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.6642026901245117,
      "learning_rate": 0.0001796196568262615,
      "loss": 3.627911376953125,
      "step": 2470
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.7580263018608093,
      "learning_rate": 0.0001794191394109071,
      "loss": 3.6858451843261717,
      "step": 2480
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.9087399840354919,
      "learning_rate": 0.0001792177534805682,
      "loss": 3.5743579864501953,
      "step": 2490
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7173874974250793,
      "learning_rate": 0.00017901550123756906,
      "loss": 3.6358150482177733,
      "step": 2500
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.7219273447990417,
      "learning_rate": 0.00017881238489370776,
      "loss": 3.5945159912109377,
      "step": 2510
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.7742475867271423,
      "learning_rate": 0.00017860840667023212,
      "loss": 3.7099586486816407,
      "step": 2520
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.5952266454696655,
      "learning_rate": 0.0001784035687978153,
      "loss": 3.6410453796386717,
      "step": 2530
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.889819324016571,
      "learning_rate": 0.0001781978735165315,
      "loss": 3.6311595916748045,
      "step": 2540
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.6258213520050049,
      "learning_rate": 0.00017799132307583132,
      "loss": 3.6003307342529296,
      "step": 2550
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.6678746938705444,
      "learning_rate": 0.0001777839197345173,
      "loss": 3.6641471862792967,
      "step": 2560
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.6566543579101562,
      "learning_rate": 0.00017757566576071914,
      "loss": 3.5644149780273438,
      "step": 2570
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.7452837824821472,
      "learning_rate": 0.00017736656343186896,
      "loss": 3.5835647583007812,
      "step": 2580
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.5735037922859192,
      "learning_rate": 0.0001771566150346763,
      "loss": 3.5213489532470703,
      "step": 2590
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6488205194473267,
      "learning_rate": 0.0001769458228651032,
      "loss": 3.6145263671875,
      "step": 2600
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.8526797294616699,
      "learning_rate": 0.00017673418922833903,
      "loss": 3.648969268798828,
      "step": 2610
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.661071240901947,
      "learning_rate": 0.0001765217164387754,
      "loss": 3.6022281646728516,
      "step": 2620
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.6927957534790039,
      "learning_rate": 0.00017630840681998066,
      "loss": 3.5860767364501953,
      "step": 2630
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.8917861580848694,
      "learning_rate": 0.00017609426270467462,
      "loss": 3.531083679199219,
      "step": 2640
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.6344000697135925,
      "learning_rate": 0.000175879286434703,
      "loss": 3.5141754150390625,
      "step": 2650
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.5514779686927795,
      "learning_rate": 0.00017566348036101187,
      "loss": 3.5212158203125,
      "step": 2660
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.7369913458824158,
      "learning_rate": 0.0001754468468436219,
      "loss": 3.5689849853515625,
      "step": 2670
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.6635401248931885,
      "learning_rate": 0.0001752293882516025,
      "loss": 3.5931190490722655,
      "step": 2680
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.6905002593994141,
      "learning_rate": 0.00017501110696304596,
      "loss": 3.6259151458740235,
      "step": 2690
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7066922783851624,
      "learning_rate": 0.00017479200536504156,
      "loss": 3.5688690185546874,
      "step": 2700
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.762281060218811,
      "learning_rate": 0.00017457208585364918,
      "loss": 3.472735595703125,
      "step": 2710
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.7107834815979004,
      "learning_rate": 0.00017435135083387345,
      "loss": 3.5385597229003904,
      "step": 2720
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.7819568514823914,
      "learning_rate": 0.0001741298027196371,
      "loss": 3.558522033691406,
      "step": 2730
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.7242825031280518,
      "learning_rate": 0.00017390744393375486,
      "loss": 3.494080352783203,
      "step": 2740
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.7194530367851257,
      "learning_rate": 0.00017368427690790677,
      "loss": 3.5231224060058595,
      "step": 2750
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.6645846366882324,
      "learning_rate": 0.00017346030408261172,
      "loss": 3.61165771484375,
      "step": 2760
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.6156540513038635,
      "learning_rate": 0.00017323552790720058,
      "loss": 3.5236202239990235,
      "step": 2770
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.6109545826911926,
      "learning_rate": 0.00017300995083978965,
      "loss": 3.4879127502441407,
      "step": 2780
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.5747758746147156,
      "learning_rate": 0.0001727835753472535,
      "loss": 3.4643646240234376,
      "step": 2790
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6963766813278198,
      "learning_rate": 0.00017255640390519836,
      "loss": 3.469551849365234,
      "step": 2800
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.6529024839401245,
      "learning_rate": 0.00017232843899793468,
      "loss": 3.504813385009766,
      "step": 2810
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.8313033580780029,
      "learning_rate": 0.00017209968311845012,
      "loss": 3.522869110107422,
      "step": 2820
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.7342233061790466,
      "learning_rate": 0.0001718701387683824,
      "loss": 3.456130599975586,
      "step": 2830
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.6717584133148193,
      "learning_rate": 0.0001716398084579917,
      "loss": 3.386435699462891,
      "step": 2840
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.6578766107559204,
      "learning_rate": 0.00017140869470613342,
      "loss": 3.313336181640625,
      "step": 2850
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.5547276735305786,
      "learning_rate": 0.00017117680004023056,
      "loss": 3.3712467193603515,
      "step": 2860
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.5347000956535339,
      "learning_rate": 0.00017094412699624595,
      "loss": 3.441112518310547,
      "step": 2870
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.5903313159942627,
      "learning_rate": 0.00017071067811865476,
      "loss": 3.3869243621826173,
      "step": 2880
    },
    {
      "epoch": 0.289,
      "grad_norm": 0.6711658835411072,
      "learning_rate": 0.00017047645596041653,
      "loss": 3.687914276123047,
      "step": 2890
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6462357044219971,
      "learning_rate": 0.00017024146308294724,
      "loss": 3.393331527709961,
      "step": 2900
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.770625114440918,
      "learning_rate": 0.00017000570205609136,
      "loss": 3.404887390136719,
      "step": 2910
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.7213258147239685,
      "learning_rate": 0.00016976917545809367,
      "loss": 3.4106369018554688,
      "step": 2920
    },
    {
      "epoch": 0.293,
      "grad_norm": 1.1367720365524292,
      "learning_rate": 0.00016953188587557122,
      "loss": 3.3777629852294924,
      "step": 2930
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.5897477269172668,
      "learning_rate": 0.0001692938359034848,
      "loss": 3.56163330078125,
      "step": 2940
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.7523307204246521,
      "learning_rate": 0.00016905502814511082,
      "loss": 3.446942138671875,
      "step": 2950
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.6670454144477844,
      "learning_rate": 0.0001688154652120126,
      "loss": 3.6020389556884767,
      "step": 2960
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.7542824149131775,
      "learning_rate": 0.00016857514972401207,
      "loss": 3.5000465393066404,
      "step": 2970
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.6708978414535522,
      "learning_rate": 0.00016833408430916085,
      "loss": 3.554585266113281,
      "step": 2980
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.555879533290863,
      "learning_rate": 0.0001680922716037117,
      "loss": 3.5675426483154298,
      "step": 2990
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.6604810357093811,
      "learning_rate": 0.00016784971425208965,
      "loss": 3.420892333984375,
      "step": 3000
    },
    {
      "epoch": 0.301,
      "grad_norm": 0.6359909772872925,
      "learning_rate": 0.00016760641490686307,
      "loss": 3.4853782653808594,
      "step": 3010
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.5913683772087097,
      "learning_rate": 0.00016736237622871452,
      "loss": 3.5785594940185548,
      "step": 3020
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.5765400528907776,
      "learning_rate": 0.00016711760088641196,
      "loss": 3.4214027404785154,
      "step": 3030
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.7350262403488159,
      "learning_rate": 0.00016687209155677929,
      "loss": 3.5618011474609377,
      "step": 3040
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.9404264092445374,
      "learning_rate": 0.00016662585092466723,
      "loss": 3.5618026733398436,
      "step": 3050
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.6636123061180115,
      "learning_rate": 0.00016637888168292384,
      "loss": 3.466056060791016,
      "step": 3060
    },
    {
      "epoch": 0.307,
      "grad_norm": 0.6642017960548401,
      "learning_rate": 0.00016613118653236518,
      "loss": 3.4905704498291015,
      "step": 3070
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.6573847532272339,
      "learning_rate": 0.0001658827681817458,
      "loss": 3.463816833496094,
      "step": 3080
    },
    {
      "epoch": 0.309,
      "grad_norm": 0.643352210521698,
      "learning_rate": 0.00016563362934772892,
      "loss": 3.6454715728759766,
      "step": 3090
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6746664643287659,
      "learning_rate": 0.00016538377275485691,
      "loss": 3.5315715789794924,
      "step": 3100
    },
    {
      "epoch": 0.311,
      "grad_norm": 0.7330307364463806,
      "learning_rate": 0.00016513320113552152,
      "loss": 3.5648765563964844,
      "step": 3110
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.6295796632766724,
      "learning_rate": 0.0001648819172299337,
      "loss": 3.453710174560547,
      "step": 3120
    },
    {
      "epoch": 0.313,
      "grad_norm": 0.6997371315956116,
      "learning_rate": 0.00016462992378609407,
      "loss": 3.6060836791992186,
      "step": 3130
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.5787609815597534,
      "learning_rate": 0.00016437722355976258,
      "loss": 3.5101036071777343,
      "step": 3140
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.6718562245368958,
      "learning_rate": 0.00016412381931442838,
      "loss": 3.493669128417969,
      "step": 3150
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.5359857082366943,
      "learning_rate": 0.0001638697138212797,
      "loss": 3.450001525878906,
      "step": 3160
    },
    {
      "epoch": 0.317,
      "grad_norm": 0.650127112865448,
      "learning_rate": 0.0001636149098591735,
      "loss": 3.5727283477783205,
      "step": 3170
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.5813530683517456,
      "learning_rate": 0.00016335941021460506,
      "loss": 3.5950172424316404,
      "step": 3180
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.6389915943145752,
      "learning_rate": 0.00016310321768167762,
      "loss": 3.4239906311035155,
      "step": 3190
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6675479412078857,
      "learning_rate": 0.0001628463350620716,
      "loss": 3.5842430114746096,
      "step": 3200
    },
    {
      "epoch": 0.321,
      "grad_norm": 0.6422111988067627,
      "learning_rate": 0.00016258876516501424,
      "loss": 3.507898712158203,
      "step": 3210
    },
    {
      "epoch": 0.322,
      "grad_norm": 1.193140983581543,
      "learning_rate": 0.00016233051080724868,
      "loss": 3.678624725341797,
      "step": 3220
    },
    {
      "epoch": 0.323,
      "grad_norm": 0.8654516339302063,
      "learning_rate": 0.00016207157481300312,
      "loss": 3.5951072692871096,
      "step": 3230
    },
    {
      "epoch": 0.324,
      "grad_norm": 1.0018895864486694,
      "learning_rate": 0.00016181196001396019,
      "loss": 3.636985015869141,
      "step": 3240
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.6855871677398682,
      "learning_rate": 0.00016155166924922566,
      "loss": 3.487339401245117,
      "step": 3250
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.6390538811683655,
      "learning_rate": 0.00016129070536529766,
      "loss": 3.4002151489257812,
      "step": 3260
    },
    {
      "epoch": 0.327,
      "grad_norm": 0.6536672711372375,
      "learning_rate": 0.00016102907121603543,
      "loss": 3.448774719238281,
      "step": 3270
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.6245121955871582,
      "learning_rate": 0.00016076676966262813,
      "loss": 3.5433879852294923,
      "step": 3280
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.6213312149047852,
      "learning_rate": 0.0001605038035735635,
      "loss": 3.53321533203125,
      "step": 3290
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6818357110023499,
      "learning_rate": 0.00016024017582459652,
      "loss": 3.498116683959961,
      "step": 3300
    },
    {
      "epoch": 0.331,
      "grad_norm": 0.8187089562416077,
      "learning_rate": 0.00015997588929871808,
      "loss": 3.52230224609375,
      "step": 3310
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.5973637104034424,
      "learning_rate": 0.00015971094688612322,
      "loss": 3.686695861816406,
      "step": 3320
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.6574364304542542,
      "learning_rate": 0.00015944535148417982,
      "loss": 3.490721893310547,
      "step": 3330
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.7805285453796387,
      "learning_rate": 0.0001591791059973965,
      "loss": 3.6205764770507813,
      "step": 3340
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.6246685981750488,
      "learning_rate": 0.00015891221333739136,
      "loss": 3.548096466064453,
      "step": 3350
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.6953656077384949,
      "learning_rate": 0.00015864467642285975,
      "loss": 3.588268280029297,
      "step": 3360
    },
    {
      "epoch": 0.337,
      "grad_norm": 0.6035398244857788,
      "learning_rate": 0.00015837649817954246,
      "loss": 3.5697952270507813,
      "step": 3370
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.6679559350013733,
      "learning_rate": 0.00015810768154019385,
      "loss": 3.608984375,
      "step": 3380
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.688679575920105,
      "learning_rate": 0.00015783822944454956,
      "loss": 3.5157821655273436,
      "step": 3390
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6748696565628052,
      "learning_rate": 0.00015756814483929458,
      "loss": 3.360149383544922,
      "step": 3400
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.61775141954422,
      "learning_rate": 0.00015729743067803087,
      "loss": 3.497563934326172,
      "step": 3410
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.6262500286102295,
      "learning_rate": 0.0001570260899212451,
      "loss": 3.523365020751953,
      "step": 3420
    },
    {
      "epoch": 0.343,
      "grad_norm": 0.6370222568511963,
      "learning_rate": 0.00015675412553627639,
      "loss": 3.554864501953125,
      "step": 3430
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.5437333583831787,
      "learning_rate": 0.00015648154049728358,
      "loss": 3.5245609283447266,
      "step": 3440
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.5970487594604492,
      "learning_rate": 0.00015620833778521307,
      "loss": 3.5683326721191406,
      "step": 3450
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.7150463461875916,
      "learning_rate": 0.0001559345203877659,
      "loss": 3.490159606933594,
      "step": 3460
    },
    {
      "epoch": 0.347,
      "grad_norm": 0.7041430473327637,
      "learning_rate": 0.00015566009129936527,
      "loss": 3.5158573150634767,
      "step": 3470
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.6444110870361328,
      "learning_rate": 0.00015538505352112375,
      "loss": 3.469894027709961,
      "step": 3480
    },
    {
      "epoch": 0.349,
      "grad_norm": 0.5570189356803894,
      "learning_rate": 0.00015510941006081032,
      "loss": 3.470653533935547,
      "step": 3490
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6065769791603088,
      "learning_rate": 0.00015483316393281784,
      "loss": 3.467085266113281,
      "step": 3500
    },
    {
      "epoch": 0.351,
      "grad_norm": 0.669251561164856,
      "learning_rate": 0.00015455631815812962,
      "loss": 3.5677433013916016,
      "step": 3510
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.5979546904563904,
      "learning_rate": 0.00015427887576428674,
      "loss": 3.484996032714844,
      "step": 3520
    },
    {
      "epoch": 0.353,
      "grad_norm": 0.6913435459136963,
      "learning_rate": 0.00015400083978535473,
      "loss": 3.5166908264160157,
      "step": 3530
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.5269363522529602,
      "learning_rate": 0.0001537222132618905,
      "loss": 3.486831283569336,
      "step": 3540
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.6342379450798035,
      "learning_rate": 0.00015344299924090915,
      "loss": 3.524557113647461,
      "step": 3550
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.7631462812423706,
      "learning_rate": 0.00015316320077585037,
      "loss": 3.4986946105957033,
      "step": 3560
    },
    {
      "epoch": 0.357,
      "grad_norm": 0.6603431701660156,
      "learning_rate": 0.00015288282092654543,
      "loss": 3.559967803955078,
      "step": 3570
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.6937748789787292,
      "learning_rate": 0.00015260186275918342,
      "loss": 3.5320461273193358,
      "step": 3580
    },
    {
      "epoch": 0.359,
      "grad_norm": 0.5924033522605896,
      "learning_rate": 0.00015232032934627782,
      "loss": 3.3522659301757813,
      "step": 3590
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.602088212966919,
      "learning_rate": 0.000152038223766633,
      "loss": 3.4664031982421877,
      "step": 3600
    },
    {
      "epoch": 0.361,
      "grad_norm": 0.6828586459159851,
      "learning_rate": 0.0001517555491053103,
      "loss": 3.3773681640625,
      "step": 3610
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.6344897747039795,
      "learning_rate": 0.00015147230845359452,
      "loss": 3.4670860290527346,
      "step": 3620
    },
    {
      "epoch": 0.363,
      "grad_norm": 0.6472758650779724,
      "learning_rate": 0.00015118850490896012,
      "loss": 3.4802452087402345,
      "step": 3630
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.6187395453453064,
      "learning_rate": 0.00015090414157503714,
      "loss": 3.4608078002929688,
      "step": 3640
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.6757095456123352,
      "learning_rate": 0.00015061922156157746,
      "loss": 3.52484130859375,
      "step": 3650
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.6977061033248901,
      "learning_rate": 0.00015033374798442074,
      "loss": 3.4902244567871095,
      "step": 3660
    },
    {
      "epoch": 0.367,
      "grad_norm": 0.7214674949645996,
      "learning_rate": 0.0001500477239654602,
      "loss": 3.46387939453125,
      "step": 3670
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.6358008980751038,
      "learning_rate": 0.00014976115263260875,
      "loss": 3.5002384185791016,
      "step": 3680
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.5897611975669861,
      "learning_rate": 0.00014947403711976454,
      "loss": 3.4682540893554688,
      "step": 3690
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8321526050567627,
      "learning_rate": 0.0001491863805667768,
      "loss": 3.4693061828613283,
      "step": 3700
    },
    {
      "epoch": 0.371,
      "grad_norm": 0.6433870196342468,
      "learning_rate": 0.00014889818611941156,
      "loss": 3.461981201171875,
      "step": 3710
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.7365756034851074,
      "learning_rate": 0.00014860945692931712,
      "loss": 3.5233833312988283,
      "step": 3720
    },
    {
      "epoch": 0.373,
      "grad_norm": 0.6342639327049255,
      "learning_rate": 0.00014832019615398963,
      "loss": 3.4597244262695312,
      "step": 3730
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.6796419620513916,
      "learning_rate": 0.00014803040695673852,
      "loss": 3.5028820037841797,
      "step": 3740
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.608890950679779,
      "learning_rate": 0.00014774009250665215,
      "loss": 3.504289245605469,
      "step": 3750
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.5925471782684326,
      "learning_rate": 0.00014744925597856275,
      "loss": 3.600018310546875,
      "step": 3760
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.6019928455352783,
      "learning_rate": 0.00014715790055301196,
      "loss": 3.5456558227539063,
      "step": 3770
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.7246829867362976,
      "learning_rate": 0.00014686602941621615,
      "loss": 3.5188270568847657,
      "step": 3780
    },
    {
      "epoch": 0.379,
      "grad_norm": 0.8092122077941895,
      "learning_rate": 0.00014657364576003122,
      "loss": 3.408944320678711,
      "step": 3790
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6834470629692078,
      "learning_rate": 0.000146280752781918,
      "loss": 3.528093719482422,
      "step": 3800
    },
    {
      "epoch": 0.381,
      "grad_norm": 0.6125589609146118,
      "learning_rate": 0.00014598735368490725,
      "loss": 3.5244441986083985,
      "step": 3810
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.7575392127037048,
      "learning_rate": 0.00014569345167756437,
      "loss": 3.4380584716796876,
      "step": 3820
    },
    {
      "epoch": 0.383,
      "grad_norm": 0.6488502621650696,
      "learning_rate": 0.00014539904997395468,
      "loss": 3.4256725311279297,
      "step": 3830
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.5788398385047913,
      "learning_rate": 0.00014510415179360805,
      "loss": 3.5277938842773438,
      "step": 3840
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.5839400887489319,
      "learning_rate": 0.00014480876036148378,
      "loss": 3.4080963134765625,
      "step": 3850
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.6609331965446472,
      "learning_rate": 0.00014451287890793513,
      "loss": 3.4297019958496096,
      "step": 3860
    },
    {
      "epoch": 0.387,
      "grad_norm": 0.6030558943748474,
      "learning_rate": 0.00014421651066867438,
      "loss": 3.4403148651123048,
      "step": 3870
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.593529224395752,
      "learning_rate": 0.00014391965888473703,
      "loss": 3.4834789276123046,
      "step": 3880
    },
    {
      "epoch": 0.389,
      "grad_norm": 0.7416579723358154,
      "learning_rate": 0.00014362232680244672,
      "loss": 3.4777900695800783,
      "step": 3890
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.5268731713294983,
      "learning_rate": 0.00014332451767337938,
      "loss": 3.508534240722656,
      "step": 3900
    },
    {
      "epoch": 0.391,
      "grad_norm": 0.5780554413795471,
      "learning_rate": 0.00014302623475432794,
      "loss": 3.53543701171875,
      "step": 3910
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.5260859727859497,
      "learning_rate": 0.00014272748130726668,
      "loss": 3.4163360595703125,
      "step": 3920
    },
    {
      "epoch": 0.393,
      "grad_norm": 0.6335863471031189,
      "learning_rate": 0.00014242826059931537,
      "loss": 3.514918899536133,
      "step": 3930
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.6461051106452942,
      "learning_rate": 0.0001421285759027038,
      "loss": 3.403517150878906,
      "step": 3940
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.6143651008605957,
      "learning_rate": 0.00014182843049473568,
      "loss": 3.4832229614257812,
      "step": 3950
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.594697117805481,
      "learning_rate": 0.00014152782765775322,
      "loss": 3.4563465118408203,
      "step": 3960
    },
    {
      "epoch": 0.397,
      "grad_norm": 0.5396756529808044,
      "learning_rate": 0.00014122677067910085,
      "loss": 3.407636260986328,
      "step": 3970
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.5760347843170166,
      "learning_rate": 0.0001409252628510894,
      "loss": 3.4377456665039063,
      "step": 3980
    },
    {
      "epoch": 0.399,
      "grad_norm": 0.5677999258041382,
      "learning_rate": 0.00014062330747096027,
      "loss": 3.4796775817871093,
      "step": 3990
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6611320376396179,
      "learning_rate": 0.00014032090784084903,
      "loss": 3.5316745758056642,
      "step": 4000
    },
    {
      "epoch": 0.401,
      "grad_norm": 0.7084479928016663,
      "learning_rate": 0.00014001806726774964,
      "loss": 3.670903778076172,
      "step": 4010
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.6611690521240234,
      "learning_rate": 0.00013971478906347806,
      "loss": 3.5300315856933593,
      "step": 4020
    },
    {
      "epoch": 0.403,
      "grad_norm": 0.6577683687210083,
      "learning_rate": 0.0001394110765446362,
      "loss": 3.4536102294921873,
      "step": 4030
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.6837309002876282,
      "learning_rate": 0.00013910693303257541,
      "loss": 3.390685272216797,
      "step": 4040
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.6647169589996338,
      "learning_rate": 0.00013880236185336044,
      "loss": 3.4458927154541015,
      "step": 4050
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.5843098759651184,
      "learning_rate": 0.00013849736633773297,
      "loss": 3.4168113708496093,
      "step": 4060
    },
    {
      "epoch": 0.407,
      "grad_norm": 0.6291281580924988,
      "learning_rate": 0.00013819194982107503,
      "loss": 3.468335723876953,
      "step": 4070
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.5572615265846252,
      "learning_rate": 0.00013788611564337277,
      "loss": 3.3637523651123047,
      "step": 4080
    },
    {
      "epoch": 0.409,
      "grad_norm": 0.6441338658332825,
      "learning_rate": 0.00013757986714917972,
      "loss": 3.5386322021484373,
      "step": 4090
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9576770067214966,
      "learning_rate": 0.00013727320768758033,
      "loss": 3.4578102111816404,
      "step": 4100
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.6570802330970764,
      "learning_rate": 0.0001369661406121534,
      "loss": 3.4554061889648438,
      "step": 4110
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.5712320804595947,
      "learning_rate": 0.0001366586692809352,
      "loss": 3.4746540069580076,
      "step": 4120
    },
    {
      "epoch": 0.413,
      "grad_norm": 0.5784115791320801,
      "learning_rate": 0.00013635079705638298,
      "loss": 3.495618438720703,
      "step": 4130
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.5961707830429077,
      "learning_rate": 0.0001360425273053381,
      "loss": 3.388241195678711,
      "step": 4140
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.5853912234306335,
      "learning_rate": 0.0001357338633989891,
      "loss": 3.3478675842285157,
      "step": 4150
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.6434437036514282,
      "learning_rate": 0.00013542480871283507,
      "loss": 3.3786964416503906,
      "step": 4160
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.7699936628341675,
      "learning_rate": 0.00013511536662664847,
      "loss": 3.3438987731933594,
      "step": 4170
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.6072028875350952,
      "learning_rate": 0.00013480554052443846,
      "loss": 3.415898895263672,
      "step": 4180
    },
    {
      "epoch": 0.419,
      "grad_norm": 0.636644721031189,
      "learning_rate": 0.0001344953337944136,
      "loss": 3.3665145874023437,
      "step": 4190
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.5747740268707275,
      "learning_rate": 0.00013418474982894498,
      "loss": 3.3864891052246096,
      "step": 4200
    },
    {
      "epoch": 0.421,
      "grad_norm": 0.5477055311203003,
      "learning_rate": 0.00013387379202452917,
      "loss": 3.3389076232910155,
      "step": 4210
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.616389811038971,
      "learning_rate": 0.0001335624637817508,
      "loss": 3.348885345458984,
      "step": 4220
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.6074135303497314,
      "learning_rate": 0.0001332507685052457,
      "loss": 3.422442626953125,
      "step": 4230
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.6009362936019897,
      "learning_rate": 0.00013293870960366346,
      "loss": 3.342233657836914,
      "step": 4240
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.6653842329978943,
      "learning_rate": 0.00013262629048963026,
      "loss": 3.3096473693847654,
      "step": 4250
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.6647281050682068,
      "learning_rate": 0.00013231351457971145,
      "loss": 3.365904998779297,
      "step": 4260
    },
    {
      "epoch": 0.427,
      "grad_norm": 0.672748327255249,
      "learning_rate": 0.00013200038529437427,
      "loss": 3.3477851867675783,
      "step": 4270
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.710106611251831,
      "learning_rate": 0.00013168690605795045,
      "loss": 3.278575897216797,
      "step": 4280
    },
    {
      "epoch": 0.429,
      "grad_norm": 0.5651130676269531,
      "learning_rate": 0.00013137308029859864,
      "loss": 3.2648223876953124,
      "step": 4290
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.582023561000824,
      "learning_rate": 0.00013105891144826715,
      "loss": 3.297943878173828,
      "step": 4300
    },
    {
      "epoch": 0.431,
      "grad_norm": 0.6400803327560425,
      "learning_rate": 0.0001307444029426561,
      "loss": 3.1890386581420898,
      "step": 4310
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.7010566592216492,
      "learning_rate": 0.00013042955822118016,
      "loss": 3.3261810302734376,
      "step": 4320
    },
    {
      "epoch": 0.433,
      "grad_norm": 0.7995615601539612,
      "learning_rate": 0.00013011438072693077,
      "loss": 3.3836013793945314,
      "step": 4330
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.5754761099815369,
      "learning_rate": 0.00012979887390663847,
      "loss": 3.3827919006347655,
      "step": 4340
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.5895978808403015,
      "learning_rate": 0.00012948304121063536,
      "loss": 3.4533233642578125,
      "step": 4350
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.5944644808769226,
      "learning_rate": 0.00012916688609281715,
      "loss": 3.3764328002929687,
      "step": 4360
    },
    {
      "epoch": 0.437,
      "grad_norm": 0.6045302152633667,
      "learning_rate": 0.0001288504120106056,
      "loss": 3.288006591796875,
      "step": 4370
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.7450554966926575,
      "learning_rate": 0.00012853362242491053,
      "loss": 3.247863006591797,
      "step": 4380
    },
    {
      "epoch": 0.439,
      "grad_norm": 0.5482000708580017,
      "learning_rate": 0.00012821652080009213,
      "loss": 3.293805694580078,
      "step": 4390
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6079279184341431,
      "learning_rate": 0.00012789911060392294,
      "loss": 3.3878684997558595,
      "step": 4400
    },
    {
      "epoch": 0.441,
      "grad_norm": 0.6110211610794067,
      "learning_rate": 0.00012758139530755,
      "loss": 3.3952926635742187,
      "step": 4410
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.6493304967880249,
      "learning_rate": 0.0001272633783854569,
      "loss": 3.335715484619141,
      "step": 4420
    },
    {
      "epoch": 0.443,
      "grad_norm": 0.6809185147285461,
      "learning_rate": 0.0001269450633154258,
      "loss": 3.307756805419922,
      "step": 4430
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.6012436747550964,
      "learning_rate": 0.00012662645357849924,
      "loss": 3.3034202575683596,
      "step": 4440
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.6538596153259277,
      "learning_rate": 0.0001263075526589423,
      "loss": 3.1933319091796877,
      "step": 4450
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.6607118844985962,
      "learning_rate": 0.0001259883640442043,
      "loss": 3.3256767272949217,
      "step": 4460
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.5607236623764038,
      "learning_rate": 0.00012566889122488084,
      "loss": 3.4197723388671877,
      "step": 4470
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.6522222757339478,
      "learning_rate": 0.0001253491376946754,
      "loss": 3.3221405029296873,
      "step": 4480
    },
    {
      "epoch": 0.449,
      "grad_norm": 0.7603774666786194,
      "learning_rate": 0.00012502910695036146,
      "loss": 3.428578186035156,
      "step": 4490
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.0236477851867676,
      "learning_rate": 0.00012470880249174397,
      "loss": 3.471366119384766,
      "step": 4500
    },
    {
      "epoch": 0.451,
      "grad_norm": 0.7241112589836121,
      "learning_rate": 0.0001243882278216211,
      "loss": 3.4871959686279297,
      "step": 4510
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.5743816494941711,
      "learning_rate": 0.00012406738644574612,
      "loss": 3.3087753295898437,
      "step": 4520
    },
    {
      "epoch": 0.453,
      "grad_norm": 0.7384122610092163,
      "learning_rate": 0.00012374628187278888,
      "loss": 3.3535835266113283,
      "step": 4530
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.5444827079772949,
      "learning_rate": 0.0001234249176142976,
      "loss": 3.4222755432128906,
      "step": 4540
    },
    {
      "epoch": 0.455,
      "grad_norm": 0.6096729636192322,
      "learning_rate": 0.00012310329718466024,
      "loss": 3.3737037658691404,
      "step": 4550
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.6477919816970825,
      "learning_rate": 0.0001227814241010664,
      "loss": 3.3756435394287108,
      "step": 4560
    },
    {
      "epoch": 0.457,
      "grad_norm": 0.9112372398376465,
      "learning_rate": 0.00012245930188346847,
      "loss": 3.4406837463378905,
      "step": 4570
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.8897560834884644,
      "learning_rate": 0.00012213693405454344,
      "loss": 3.3183822631835938,
      "step": 4580
    },
    {
      "epoch": 0.459,
      "grad_norm": 0.5743932723999023,
      "learning_rate": 0.00012181432413965428,
      "loss": 3.4875934600830076,
      "step": 4590
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5845673680305481,
      "learning_rate": 0.00012149147566681124,
      "loss": 3.4266891479492188,
      "step": 4600
    },
    {
      "epoch": 0.461,
      "grad_norm": 0.5950560569763184,
      "learning_rate": 0.00012116839216663359,
      "loss": 3.3584316253662108,
      "step": 4610
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.569413423538208,
      "learning_rate": 0.00012084507717231064,
      "loss": 3.377399444580078,
      "step": 4620
    },
    {
      "epoch": 0.463,
      "grad_norm": 0.6886783838272095,
      "learning_rate": 0.00012052153421956342,
      "loss": 3.3393524169921873,
      "step": 4630
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.6228946447372437,
      "learning_rate": 0.00012019776684660584,
      "loss": 3.4307697296142576,
      "step": 4640
    },
    {
      "epoch": 0.465,
      "grad_norm": 0.5816109776496887,
      "learning_rate": 0.00011987377859410597,
      "loss": 3.305574417114258,
      "step": 4650
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.5540332198143005,
      "learning_rate": 0.00011954957300514748,
      "loss": 3.484218215942383,
      "step": 4660
    },
    {
      "epoch": 0.467,
      "grad_norm": 0.6386479139328003,
      "learning_rate": 0.0001192251536251907,
      "loss": 3.31146125793457,
      "step": 4670
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.5721418261528015,
      "learning_rate": 0.00011890052400203404,
      "loss": 3.4248252868652345,
      "step": 4680
    },
    {
      "epoch": 0.469,
      "grad_norm": 0.6279579997062683,
      "learning_rate": 0.000118575687685775,
      "loss": 3.3772239685058594,
      "step": 4690
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6600814461708069,
      "learning_rate": 0.00011825064822877152,
      "loss": 3.4140548706054688,
      "step": 4700
    },
    {
      "epoch": 0.471,
      "grad_norm": 0.5594117045402527,
      "learning_rate": 0.00011792540918560306,
      "loss": 3.4825454711914063,
      "step": 4710
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.6652513146400452,
      "learning_rate": 0.00011759997411303161,
      "loss": 3.2841751098632814,
      "step": 4720
    },
    {
      "epoch": 0.473,
      "grad_norm": 0.6556232571601868,
      "learning_rate": 0.00011727434656996305,
      "loss": 3.413335418701172,
      "step": 4730
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.583631157875061,
      "learning_rate": 0.00011694853011740798,
      "loss": 3.3857162475585936,
      "step": 4740
    },
    {
      "epoch": 0.475,
      "grad_norm": 0.5874406099319458,
      "learning_rate": 0.00011662252831844296,
      "loss": 3.4508628845214844,
      "step": 4750
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.6048982739448547,
      "learning_rate": 0.00011629634473817137,
      "loss": 3.338873291015625,
      "step": 4760
    },
    {
      "epoch": 0.477,
      "grad_norm": 0.7144953012466431,
      "learning_rate": 0.00011596998294368466,
      "loss": 3.4625774383544923,
      "step": 4770
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.6062578558921814,
      "learning_rate": 0.0001156434465040231,
      "loss": 3.476113128662109,
      "step": 4780
    },
    {
      "epoch": 0.479,
      "grad_norm": 0.6345227360725403,
      "learning_rate": 0.00011531673899013686,
      "loss": 3.4687313079833983,
      "step": 4790
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.682437539100647,
      "learning_rate": 0.000114989863974847,
      "loss": 3.421291732788086,
      "step": 4800
    },
    {
      "epoch": 0.481,
      "grad_norm": 0.6347635388374329,
      "learning_rate": 0.0001146628250328063,
      "loss": 3.406683349609375,
      "step": 4810
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.5453388094902039,
      "learning_rate": 0.00011433562574046019,
      "loss": 3.398492431640625,
      "step": 4820
    },
    {
      "epoch": 0.483,
      "grad_norm": 0.5737029314041138,
      "learning_rate": 0.0001140082696760078,
      "loss": 3.3925594329833983,
      "step": 4830
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.5444730520248413,
      "learning_rate": 0.00011368076041936251,
      "loss": 3.4567337036132812,
      "step": 4840
    },
    {
      "epoch": 0.485,
      "grad_norm": 0.5788640379905701,
      "learning_rate": 0.00011335310155211318,
      "loss": 3.362701416015625,
      "step": 4850
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.5628058910369873,
      "learning_rate": 0.0001130252966574846,
      "loss": 3.3347236633300783,
      "step": 4860
    },
    {
      "epoch": 0.487,
      "grad_norm": 0.6594650149345398,
      "learning_rate": 0.00011269734932029863,
      "loss": 3.442543792724609,
      "step": 4870
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.7112128734588623,
      "learning_rate": 0.00011236926312693479,
      "loss": 3.394313430786133,
      "step": 4880
    },
    {
      "epoch": 0.489,
      "grad_norm": 0.6329519152641296,
      "learning_rate": 0.0001120410416652911,
      "loss": 3.4499832153320313,
      "step": 4890
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5771443247795105,
      "learning_rate": 0.00011171268852474493,
      "loss": 3.419938659667969,
      "step": 4900
    },
    {
      "epoch": 0.491,
      "grad_norm": 0.925687313079834,
      "learning_rate": 0.00011138420729611355,
      "loss": 3.482391357421875,
      "step": 4910
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.6629887819290161,
      "learning_rate": 0.00011105560157161513,
      "loss": 3.459466552734375,
      "step": 4920
    },
    {
      "epoch": 0.493,
      "grad_norm": 0.5960015654563904,
      "learning_rate": 0.00011072687494482919,
      "loss": 3.4282146453857423,
      "step": 4930
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.6006155014038086,
      "learning_rate": 0.00011039803101065742,
      "loss": 3.3696937561035156,
      "step": 4940
    },
    {
      "epoch": 0.495,
      "grad_norm": 0.61974036693573,
      "learning_rate": 0.00011006907336528445,
      "loss": 3.532196044921875,
      "step": 4950
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.5892999172210693,
      "learning_rate": 0.00010974000560613835,
      "loss": 3.468359375,
      "step": 4960
    },
    {
      "epoch": 0.497,
      "grad_norm": 0.7188873291015625,
      "learning_rate": 0.00010941083133185146,
      "loss": 3.5330413818359374,
      "step": 4970
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.5818237662315369,
      "learning_rate": 0.00010908155414222083,
      "loss": 3.3889827728271484,
      "step": 4980
    },
    {
      "epoch": 0.499,
      "grad_norm": 0.605966329574585,
      "learning_rate": 0.00010875217763816911,
      "loss": 3.3176342010498048,
      "step": 4990
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6832404136657715,
      "learning_rate": 0.00010842270542170499,
      "loss": 3.3936172485351563,
      "step": 5000
    },
    {
      "epoch": 0.501,
      "grad_norm": 0.5781661868095398,
      "learning_rate": 0.00010809314109588378,
      "loss": 3.5117469787597657,
      "step": 5010
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.6056392788887024,
      "learning_rate": 0.00010776348826476821,
      "loss": 3.3459613800048826,
      "step": 5020
    },
    {
      "epoch": 0.503,
      "grad_norm": 0.6095778346061707,
      "learning_rate": 0.00010743375053338877,
      "loss": 3.420059585571289,
      "step": 5030
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.5147314667701721,
      "learning_rate": 0.0001071039315077045,
      "loss": 3.3664085388183596,
      "step": 5040
    },
    {
      "epoch": 0.505,
      "grad_norm": 0.6032224297523499,
      "learning_rate": 0.00010677403479456341,
      "loss": 3.3725307464599608,
      "step": 5050
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.7496193051338196,
      "learning_rate": 0.00010644406400166309,
      "loss": 3.5557319641113283,
      "step": 5060
    },
    {
      "epoch": 0.507,
      "grad_norm": 0.5862683653831482,
      "learning_rate": 0.00010611402273751132,
      "loss": 3.386946105957031,
      "step": 5070
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.5528510808944702,
      "learning_rate": 0.00010578391461138641,
      "loss": 3.4095222473144533,
      "step": 5080
    },
    {
      "epoch": 0.509,
      "grad_norm": 0.9300732612609863,
      "learning_rate": 0.000105453743233298,
      "loss": 3.4315444946289064,
      "step": 5090
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.5916882157325745,
      "learning_rate": 0.00010512351221394731,
      "loss": 3.513938140869141,
      "step": 5100
    },
    {
      "epoch": 0.511,
      "grad_norm": 0.5917195677757263,
      "learning_rate": 0.00010479322516468797,
      "loss": 3.359455108642578,
      "step": 5110
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.58696049451828,
      "learning_rate": 0.00010446288569748613,
      "loss": 3.4676593780517577,
      "step": 5120
    },
    {
      "epoch": 0.513,
      "grad_norm": 0.622844934463501,
      "learning_rate": 0.00010413249742488131,
      "loss": 3.1903457641601562,
      "step": 5130
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.5546836853027344,
      "learning_rate": 0.00010380206395994679,
      "loss": 3.339521789550781,
      "step": 5140
    },
    {
      "epoch": 0.515,
      "grad_norm": 0.652741551399231,
      "learning_rate": 0.00010347158891624991,
      "loss": 3.429127883911133,
      "step": 5150
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.5863950848579407,
      "learning_rate": 0.00010314107590781284,
      "loss": 3.3279129028320313,
      "step": 5160
    },
    {
      "epoch": 0.517,
      "grad_norm": 0.6411070823669434,
      "learning_rate": 0.00010281052854907289,
      "loss": 3.4003170013427733,
      "step": 5170
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.5748854875564575,
      "learning_rate": 0.00010247995045484302,
      "loss": 3.3802757263183594,
      "step": 5180
    },
    {
      "epoch": 0.519,
      "grad_norm": 0.6322228908538818,
      "learning_rate": 0.00010214934524027228,
      "loss": 3.4053604125976564,
      "step": 5190
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.596312940120697,
      "learning_rate": 0.00010181871652080637,
      "loss": 3.3769157409667967,
      "step": 5200
    },
    {
      "epoch": 0.521,
      "grad_norm": 0.595954418182373,
      "learning_rate": 0.00010148806791214798,
      "loss": 3.396070098876953,
      "step": 5210
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.6632212996482849,
      "learning_rate": 0.00010115740303021728,
      "loss": 3.431427001953125,
      "step": 5220
    },
    {
      "epoch": 0.523,
      "grad_norm": 0.6198023557662964,
      "learning_rate": 0.0001008267254911125,
      "loss": 3.4389549255371095,
      "step": 5230
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.6012481451034546,
      "learning_rate": 0.0001004960389110702,
      "loss": 3.4386077880859376,
      "step": 5240
    },
    {
      "epoch": 0.525,
      "grad_norm": 0.5540056228637695,
      "learning_rate": 0.00010016534690642583,
      "loss": 3.4821502685546877,
      "step": 5250
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.66548091173172,
      "learning_rate": 9.983465309357418e-05,
      "loss": 3.4394546508789063,
      "step": 5260
    },
    {
      "epoch": 0.527,
      "grad_norm": 0.728541910648346,
      "learning_rate": 9.950396108892983e-05,
      "loss": 3.4411659240722656,
      "step": 5270
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.5508425235748291,
      "learning_rate": 9.917327450888751e-05,
      "loss": 3.410908508300781,
      "step": 5280
    },
    {
      "epoch": 0.529,
      "grad_norm": 0.8100053071975708,
      "learning_rate": 9.884259696978274e-05,
      "loss": 3.4032230377197266,
      "step": 5290
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6017371416091919,
      "learning_rate": 9.851193208785206e-05,
      "loss": 3.4071701049804686,
      "step": 5300
    },
    {
      "epoch": 0.531,
      "grad_norm": 0.5750782489776611,
      "learning_rate": 9.818128347919362e-05,
      "loss": 3.3987308502197267,
      "step": 5310
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.5803584456443787,
      "learning_rate": 9.785065475972774e-05,
      "loss": 3.356268310546875,
      "step": 5320
    },
    {
      "epoch": 0.533,
      "grad_norm": 0.5716253519058228,
      "learning_rate": 9.7520049545157e-05,
      "loss": 3.3503509521484376,
      "step": 5330
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.5557127594947815,
      "learning_rate": 9.718947145092713e-05,
      "loss": 3.348654937744141,
      "step": 5340
    },
    {
      "epoch": 0.535,
      "grad_norm": 0.6414467096328735,
      "learning_rate": 9.685892409218717e-05,
      "loss": 3.358196258544922,
      "step": 5350
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.5752204656600952,
      "learning_rate": 9.652841108375012e-05,
      "loss": 3.3791130065917967,
      "step": 5360
    },
    {
      "epoch": 0.537,
      "grad_norm": 0.6413206458091736,
      "learning_rate": 9.619793604005323e-05,
      "loss": 3.389569854736328,
      "step": 5370
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.5846290588378906,
      "learning_rate": 9.586750257511867e-05,
      "loss": 3.3406608581542967,
      "step": 5380
    },
    {
      "epoch": 0.539,
      "grad_norm": 0.6580010652542114,
      "learning_rate": 9.553711430251388e-05,
      "loss": 3.3569034576416015,
      "step": 5390
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7277615666389465,
      "learning_rate": 9.520677483531205e-05,
      "loss": 3.3255069732666014,
      "step": 5400
    },
    {
      "epoch": 0.541,
      "grad_norm": 0.6761912107467651,
      "learning_rate": 9.48764877860527e-05,
      "loss": 3.4096694946289063,
      "step": 5410
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.5481249094009399,
      "learning_rate": 9.454625676670203e-05,
      "loss": 3.393737030029297,
      "step": 5420
    },
    {
      "epoch": 0.543,
      "grad_norm": 0.5369051694869995,
      "learning_rate": 9.421608538861361e-05,
      "loss": 3.3308784484863283,
      "step": 5430
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.6069549322128296,
      "learning_rate": 9.38859772624887e-05,
      "loss": 3.3847000122070314,
      "step": 5440
    },
    {
      "epoch": 0.545,
      "grad_norm": 0.5608025193214417,
      "learning_rate": 9.35559359983369e-05,
      "loss": 3.392359161376953,
      "step": 5450
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.6789959669113159,
      "learning_rate": 9.32259652054366e-05,
      "loss": 3.2926254272460938,
      "step": 5460
    },
    {
      "epoch": 0.547,
      "grad_norm": 0.5733320713043213,
      "learning_rate": 9.289606849229551e-05,
      "loss": 3.352497100830078,
      "step": 5470
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.5643939971923828,
      "learning_rate": 9.256624946661125e-05,
      "loss": 3.3993389129638674,
      "step": 5480
    },
    {
      "epoch": 0.549,
      "grad_norm": 0.5881873965263367,
      "learning_rate": 9.223651173523182e-05,
      "loss": 3.296744155883789,
      "step": 5490
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.571657121181488,
      "learning_rate": 9.190685890411627e-05,
      "loss": 3.35467414855957,
      "step": 5500
    },
    {
      "epoch": 0.551,
      "grad_norm": 0.7389491200447083,
      "learning_rate": 9.157729457829503e-05,
      "loss": 3.3461029052734377,
      "step": 5510
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.6836391687393188,
      "learning_rate": 9.124782236183088e-05,
      "loss": 3.4497825622558596,
      "step": 5520
    },
    {
      "epoch": 0.553,
      "grad_norm": 0.5360568761825562,
      "learning_rate": 9.091844585777918e-05,
      "loss": 3.299654006958008,
      "step": 5530
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.5460582375526428,
      "learning_rate": 9.058916866814858e-05,
      "loss": 3.382965850830078,
      "step": 5540
    },
    {
      "epoch": 0.555,
      "grad_norm": 0.639477014541626,
      "learning_rate": 9.025999439386169e-05,
      "loss": 3.3513526916503906,
      "step": 5550
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.7162673473358154,
      "learning_rate": 8.993092663471556e-05,
      "loss": 3.3255210876464845,
      "step": 5560
    },
    {
      "epoch": 0.557,
      "grad_norm": 0.6337275505065918,
      "learning_rate": 8.960196898934263e-05,
      "loss": 3.322473907470703,
      "step": 5570
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.5694190263748169,
      "learning_rate": 8.927312505517085e-05,
      "loss": 3.363755798339844,
      "step": 5580
    },
    {
      "epoch": 0.559,
      "grad_norm": 0.7194228768348694,
      "learning_rate": 8.894439842838488e-05,
      "loss": 3.3668949127197267,
      "step": 5590
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5786919593811035,
      "learning_rate": 8.861579270388648e-05,
      "loss": 3.3777557373046876,
      "step": 5600
    },
    {
      "epoch": 0.561,
      "grad_norm": 0.8630422949790955,
      "learning_rate": 8.82873114752551e-05,
      "loss": 3.339837646484375,
      "step": 5610
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.5743309259414673,
      "learning_rate": 8.795895833470895e-05,
      "loss": 3.297917938232422,
      "step": 5620
    },
    {
      "epoch": 0.563,
      "grad_norm": 0.5526283979415894,
      "learning_rate": 8.763073687306524e-05,
      "loss": 3.1993476867675783,
      "step": 5630
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.5465028285980225,
      "learning_rate": 8.730265067970142e-05,
      "loss": 3.2985305786132812,
      "step": 5640
    },
    {
      "epoch": 0.565,
      "grad_norm": 0.8999131917953491,
      "learning_rate": 8.697470334251543e-05,
      "loss": 3.3341941833496094,
      "step": 5650
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.5656766891479492,
      "learning_rate": 8.664689844788683e-05,
      "loss": 3.282196044921875,
      "step": 5660
    },
    {
      "epoch": 0.567,
      "grad_norm": 0.5912207365036011,
      "learning_rate": 8.63192395806375e-05,
      "loss": 3.18927116394043,
      "step": 5670
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.5090358257293701,
      "learning_rate": 8.599173032399221e-05,
      "loss": 3.1961231231689453,
      "step": 5680
    },
    {
      "epoch": 0.569,
      "grad_norm": 0.6499865055084229,
      "learning_rate": 8.566437425953983e-05,
      "loss": 3.270585632324219,
      "step": 5690
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5761920213699341,
      "learning_rate": 8.533717496719373e-05,
      "loss": 3.2009956359863283,
      "step": 5700
    },
    {
      "epoch": 0.571,
      "grad_norm": 0.5735033750534058,
      "learning_rate": 8.501013602515306e-05,
      "loss": 3.2796539306640624,
      "step": 5710
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.6368776559829712,
      "learning_rate": 8.468326100986317e-05,
      "loss": 3.4039962768554686,
      "step": 5720
    },
    {
      "epoch": 0.573,
      "grad_norm": 0.6045583486557007,
      "learning_rate": 8.435655349597689e-05,
      "loss": 3.235260009765625,
      "step": 5730
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.5776654481887817,
      "learning_rate": 8.403001705631536e-05,
      "loss": 3.252074432373047,
      "step": 5740
    },
    {
      "epoch": 0.575,
      "grad_norm": 0.7625782489776611,
      "learning_rate": 8.370365526182861e-05,
      "loss": 3.2916000366210936,
      "step": 5750
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.5528947114944458,
      "learning_rate": 8.337747168155709e-05,
      "loss": 3.1939159393310548,
      "step": 5760
    },
    {
      "epoch": 0.577,
      "grad_norm": 0.7308654189109802,
      "learning_rate": 8.305146988259205e-05,
      "loss": 3.2417407989501954,
      "step": 5770
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.5406909584999084,
      "learning_rate": 8.2725653430037e-05,
      "loss": 3.2778396606445312,
      "step": 5780
    },
    {
      "epoch": 0.579,
      "grad_norm": 0.5374094843864441,
      "learning_rate": 8.240002588696842e-05,
      "loss": 3.2559051513671875,
      "step": 5790
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5618265271186829,
      "learning_rate": 8.207459081439695e-05,
      "loss": 3.2217041015625,
      "step": 5800
    },
    {
      "epoch": 0.581,
      "grad_norm": 0.5441599488258362,
      "learning_rate": 8.174935177122849e-05,
      "loss": 3.323306655883789,
      "step": 5810
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.6818861365318298,
      "learning_rate": 8.1424312314225e-05,
      "loss": 3.3473800659179687,
      "step": 5820
    },
    {
      "epoch": 0.583,
      "grad_norm": 0.6282973885536194,
      "learning_rate": 8.1099475997966e-05,
      "loss": 3.235007476806641,
      "step": 5830
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.5572187900543213,
      "learning_rate": 8.07748463748093e-05,
      "loss": 3.3852062225341797,
      "step": 5840
    },
    {
      "epoch": 0.585,
      "grad_norm": 0.6209778785705566,
      "learning_rate": 8.045042699485255e-05,
      "loss": 3.300220489501953,
      "step": 5850
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.650616466999054,
      "learning_rate": 8.012622140589404e-05,
      "loss": 3.2928466796875,
      "step": 5860
    },
    {
      "epoch": 0.587,
      "grad_norm": 0.5636923313140869,
      "learning_rate": 7.980223315339417e-05,
      "loss": 3.2997039794921874,
      "step": 5870
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.5796397924423218,
      "learning_rate": 7.947846578043659e-05,
      "loss": 3.3791275024414062,
      "step": 5880
    },
    {
      "epoch": 0.589,
      "grad_norm": 0.598293125629425,
      "learning_rate": 7.915492282768937e-05,
      "loss": 3.2516517639160156,
      "step": 5890
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5401917099952698,
      "learning_rate": 7.883160783336645e-05,
      "loss": 3.316074752807617,
      "step": 5900
    },
    {
      "epoch": 0.591,
      "grad_norm": 0.6009707450866699,
      "learning_rate": 7.850852433318877e-05,
      "loss": 3.4310035705566406,
      "step": 5910
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.5513361692428589,
      "learning_rate": 7.818567586034577e-05,
      "loss": 3.3075531005859373,
      "step": 5920
    },
    {
      "epoch": 0.593,
      "grad_norm": 0.5836353898048401,
      "learning_rate": 7.786306594545657e-05,
      "loss": 3.3529991149902343,
      "step": 5930
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.5985729694366455,
      "learning_rate": 7.754069811653154e-05,
      "loss": 3.3714271545410157,
      "step": 5940
    },
    {
      "epoch": 0.595,
      "grad_norm": 0.5671996474266052,
      "learning_rate": 7.721857589893364e-05,
      "loss": 3.322060394287109,
      "step": 5950
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.561960756778717,
      "learning_rate": 7.689670281533976e-05,
      "loss": 3.411103057861328,
      "step": 5960
    },
    {
      "epoch": 0.597,
      "grad_norm": 0.6210089921951294,
      "learning_rate": 7.657508238570246e-05,
      "loss": 3.3525543212890625,
      "step": 5970
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.6080389022827148,
      "learning_rate": 7.625371812721114e-05,
      "loss": 3.319987487792969,
      "step": 5980
    },
    {
      "epoch": 0.599,
      "grad_norm": 0.6236148476600647,
      "learning_rate": 7.593261355425394e-05,
      "loss": 3.2959564208984373,
      "step": 5990
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6411356925964355,
      "learning_rate": 7.561177217837894e-05,
      "loss": 3.3301933288574217,
      "step": 6000
    },
    {
      "epoch": 0.601,
      "grad_norm": 0.6190471053123474,
      "learning_rate": 7.529119750825606e-05,
      "loss": 3.2937397003173827,
      "step": 6010
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.5404062867164612,
      "learning_rate": 7.497089304963853e-05,
      "loss": 3.3483715057373047,
      "step": 6020
    },
    {
      "epoch": 0.603,
      "grad_norm": 0.579624354839325,
      "learning_rate": 7.46508623053246e-05,
      "loss": 3.2702259063720702,
      "step": 6030
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.6470794081687927,
      "learning_rate": 7.433110877511922e-05,
      "loss": 3.3062698364257814,
      "step": 6040
    },
    {
      "epoch": 0.605,
      "grad_norm": 0.537086546421051,
      "learning_rate": 7.401163595579572e-05,
      "loss": 3.323369598388672,
      "step": 6050
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.5717902183532715,
      "learning_rate": 7.369244734105774e-05,
      "loss": 3.3086502075195314,
      "step": 6060
    },
    {
      "epoch": 0.607,
      "grad_norm": 0.5908029675483704,
      "learning_rate": 7.337354642150077e-05,
      "loss": 3.468159484863281,
      "step": 6070
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.6381606459617615,
      "learning_rate": 7.30549366845742e-05,
      "loss": 3.3596118927001952,
      "step": 6080
    },
    {
      "epoch": 0.609,
      "grad_norm": 0.5811834931373596,
      "learning_rate": 7.273662161454309e-05,
      "loss": 3.3802398681640624,
      "step": 6090
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.5594106316566467,
      "learning_rate": 7.241860469245001e-05,
      "loss": 3.3437938690185547,
      "step": 6100
    },
    {
      "epoch": 0.611,
      "grad_norm": 0.5406367778778076,
      "learning_rate": 7.210088939607708e-05,
      "loss": 3.129538345336914,
      "step": 6110
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.6994330883026123,
      "learning_rate": 7.178347919990789e-05,
      "loss": 3.4236900329589846,
      "step": 6120
    },
    {
      "epoch": 0.613,
      "grad_norm": 0.5376851558685303,
      "learning_rate": 7.146637757508949e-05,
      "loss": 3.339020538330078,
      "step": 6130
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.6126795411109924,
      "learning_rate": 7.114958798939442e-05,
      "loss": 3.3541770935058595,
      "step": 6140
    },
    {
      "epoch": 0.615,
      "grad_norm": 0.6448752880096436,
      "learning_rate": 7.083311390718284e-05,
      "loss": 3.373933029174805,
      "step": 6150
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.5439074635505676,
      "learning_rate": 7.051695878936466e-05,
      "loss": 3.37576904296875,
      "step": 6160
    },
    {
      "epoch": 0.617,
      "grad_norm": 0.6267314553260803,
      "learning_rate": 7.020112609336153e-05,
      "loss": 3.3639251708984377,
      "step": 6170
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.5915143489837646,
      "learning_rate": 6.988561927306927e-05,
      "loss": 3.3968490600585937,
      "step": 6180
    },
    {
      "epoch": 0.619,
      "grad_norm": 0.6531071066856384,
      "learning_rate": 6.957044177881987e-05,
      "loss": 3.294902801513672,
      "step": 6190
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5973216891288757,
      "learning_rate": 6.925559705734394e-05,
      "loss": 3.344374084472656,
      "step": 6200
    },
    {
      "epoch": 0.621,
      "grad_norm": 0.5954884886741638,
      "learning_rate": 6.894108855173288e-05,
      "loss": 3.3901927947998045,
      "step": 6210
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.7344208359718323,
      "learning_rate": 6.862691970140134e-05,
      "loss": 3.2878189086914062,
      "step": 6220
    },
    {
      "epoch": 0.623,
      "grad_norm": 0.6660248637199402,
      "learning_rate": 6.831309394204957e-05,
      "loss": 3.3664417266845703,
      "step": 6230
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.6550434231758118,
      "learning_rate": 6.799961470562573e-05,
      "loss": 3.4067611694335938,
      "step": 6240
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.5701867938041687,
      "learning_rate": 6.768648542028858e-05,
      "loss": 3.326247787475586,
      "step": 6250
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.5788747072219849,
      "learning_rate": 6.737370951036976e-05,
      "loss": 3.3925094604492188,
      "step": 6260
    },
    {
      "epoch": 0.627,
      "grad_norm": 0.7393894791603088,
      "learning_rate": 6.706129039633659e-05,
      "loss": 3.316583251953125,
      "step": 6270
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.5733975172042847,
      "learning_rate": 6.674923149475432e-05,
      "loss": 3.3438907623291017,
      "step": 6280
    },
    {
      "epoch": 0.629,
      "grad_norm": 0.5896235704421997,
      "learning_rate": 6.643753621824922e-05,
      "loss": 3.333628463745117,
      "step": 6290
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6701786518096924,
      "learning_rate": 6.612620797547087e-05,
      "loss": 3.4828834533691406,
      "step": 6300
    },
    {
      "epoch": 0.631,
      "grad_norm": 0.6754198670387268,
      "learning_rate": 6.581525017105501e-05,
      "loss": 3.495597076416016,
      "step": 6310
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.5892626643180847,
      "learning_rate": 6.550466620558643e-05,
      "loss": 3.3608783721923827,
      "step": 6320
    },
    {
      "epoch": 0.633,
      "grad_norm": 0.5734609961509705,
      "learning_rate": 6.519445947556155e-05,
      "loss": 3.2244083404541017,
      "step": 6330
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.604988694190979,
      "learning_rate": 6.488463337335157e-05,
      "loss": 3.3789615631103516,
      "step": 6340
    },
    {
      "epoch": 0.635,
      "grad_norm": 0.6067793369293213,
      "learning_rate": 6.457519128716497e-05,
      "loss": 3.358556365966797,
      "step": 6350
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.5685458779335022,
      "learning_rate": 6.426613660101091e-05,
      "loss": 3.3084999084472657,
      "step": 6360
    },
    {
      "epoch": 0.637,
      "grad_norm": 0.591894805431366,
      "learning_rate": 6.395747269466192e-05,
      "loss": 3.3472694396972655,
      "step": 6370
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.6098620891571045,
      "learning_rate": 6.3649202943617e-05,
      "loss": 3.4307762145996095,
      "step": 6380
    },
    {
      "epoch": 0.639,
      "grad_norm": 0.5914278030395508,
      "learning_rate": 6.334133071906483e-05,
      "loss": 3.3197708129882812,
      "step": 6390
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5753639936447144,
      "learning_rate": 6.303385938784663e-05,
      "loss": 3.3101627349853517,
      "step": 6400
    },
    {
      "epoch": 0.641,
      "grad_norm": 0.591321587562561,
      "learning_rate": 6.27267923124197e-05,
      "loss": 3.245317840576172,
      "step": 6410
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.7738291025161743,
      "learning_rate": 6.242013285082029e-05,
      "loss": 3.5058624267578127,
      "step": 6420
    },
    {
      "epoch": 0.643,
      "grad_norm": 0.6543744206428528,
      "learning_rate": 6.211388435662721e-05,
      "loss": 3.3384780883789062,
      "step": 6430
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.5630353689193726,
      "learning_rate": 6.180805017892498e-05,
      "loss": 3.356060028076172,
      "step": 6440
    },
    {
      "epoch": 0.645,
      "grad_norm": 0.5631670951843262,
      "learning_rate": 6.150263366226703e-05,
      "loss": 3.3661167144775392,
      "step": 6450
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.6407634615898132,
      "learning_rate": 6.119763814663958e-05,
      "loss": 3.379840850830078,
      "step": 6460
    },
    {
      "epoch": 0.647,
      "grad_norm": 0.6552280187606812,
      "learning_rate": 6.089306696742461e-05,
      "loss": 3.2964599609375,
      "step": 6470
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.5933181643486023,
      "learning_rate": 6.0588923455363864e-05,
      "loss": 3.3642288208007813,
      "step": 6480
    },
    {
      "epoch": 0.649,
      "grad_norm": 0.7284210920333862,
      "learning_rate": 6.0285210936521955e-05,
      "loss": 3.4394763946533202,
      "step": 6490
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6503067016601562,
      "learning_rate": 5.998193273225036e-05,
      "loss": 3.3815567016601564,
      "step": 6500
    },
    {
      "epoch": 0.651,
      "grad_norm": 0.6401969790458679,
      "learning_rate": 5.967909215915099e-05,
      "loss": 3.3583553314208983,
      "step": 6510
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.5397207736968994,
      "learning_rate": 5.937669252903973e-05,
      "loss": 3.3225028991699217,
      "step": 6520
    },
    {
      "epoch": 0.653,
      "grad_norm": 0.5971732139587402,
      "learning_rate": 5.907473714891061e-05,
      "loss": 3.3478958129882814,
      "step": 6530
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.7859312295913696,
      "learning_rate": 5.877322932089919e-05,
      "loss": 3.3094932556152346,
      "step": 6540
    },
    {
      "epoch": 0.655,
      "grad_norm": 0.6023569703102112,
      "learning_rate": 5.847217234224681e-05,
      "loss": 3.3368064880371096,
      "step": 6550
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.6322081685066223,
      "learning_rate": 5.8171569505264314e-05,
      "loss": 3.357105255126953,
      "step": 6560
    },
    {
      "epoch": 0.657,
      "grad_norm": 0.5233018398284912,
      "learning_rate": 5.7871424097296224e-05,
      "loss": 3.3510116577148437,
      "step": 6570
    },
    {
      "epoch": 0.658,
      "grad_norm": 0.5499789118766785,
      "learning_rate": 5.757173940068464e-05,
      "loss": 3.3366062164306642,
      "step": 6580
    },
    {
      "epoch": 0.659,
      "grad_norm": 0.6057854294776917,
      "learning_rate": 5.727251869273333e-05,
      "loss": 3.3797225952148438,
      "step": 6590
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5440980195999146,
      "learning_rate": 5.697376524567205e-05,
      "loss": 3.2615066528320313,
      "step": 6600
    },
    {
      "epoch": 0.661,
      "grad_norm": 0.5871827602386475,
      "learning_rate": 5.6675482326620635e-05,
      "loss": 3.508020782470703,
      "step": 6610
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.6648843884468079,
      "learning_rate": 5.6377673197553325e-05,
      "loss": 3.4385955810546873,
      "step": 6620
    },
    {
      "epoch": 0.663,
      "grad_norm": 0.560800313949585,
      "learning_rate": 5.608034111526298e-05,
      "loss": 3.3368663787841797,
      "step": 6630
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.5215429663658142,
      "learning_rate": 5.578348933132565e-05,
      "loss": 3.283843994140625,
      "step": 6640
    },
    {
      "epoch": 0.665,
      "grad_norm": 0.5655593276023865,
      "learning_rate": 5.548712109206486e-05,
      "loss": 3.3516883850097656,
      "step": 6650
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.5367684960365295,
      "learning_rate": 5.519123963851623e-05,
      "loss": 3.2509128570556642,
      "step": 6660
    },
    {
      "epoch": 0.667,
      "grad_norm": 0.5667892694473267,
      "learning_rate": 5.489584820639195e-05,
      "loss": 3.279737091064453,
      "step": 6670
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.6646414995193481,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 3.337319564819336,
      "step": 6680
    },
    {
      "epoch": 0.669,
      "grad_norm": NaN,
      "learning_rate": 5.430654832243569e-05,
      "loss": 3.336614990234375,
      "step": 6690
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5647449493408203,
      "learning_rate": 5.404201393785122e-05,
      "loss": 3.2939689636230467,
      "step": 6700
    },
    {
      "epoch": 0.671,
      "grad_norm": 0.7248727083206177,
      "learning_rate": 5.374856440533078e-05,
      "loss": 3.2961097717285157,
      "step": 6710
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.5753278732299805,
      "learning_rate": 5.345562067109984e-05,
      "loss": 3.436513137817383,
      "step": 6720
    },
    {
      "epoch": 0.673,
      "grad_norm": 0.5696133375167847,
      "learning_rate": 5.316318593874415e-05,
      "loss": 3.3226078033447264,
      "step": 6730
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.6013801097869873,
      "learning_rate": 5.287126340628312e-05,
      "loss": 3.461760330200195,
      "step": 6740
    },
    {
      "epoch": 0.675,
      "grad_norm": 0.6747791767120361,
      "learning_rate": 5.25798562661348e-05,
      "loss": 3.2670974731445312,
      "step": 6750
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.6186642050743103,
      "learning_rate": 5.2288967705081e-05,
      "loss": 3.2816627502441404,
      "step": 6760
    },
    {
      "epoch": 0.677,
      "grad_norm": 0.5448969006538391,
      "learning_rate": 5.199860090423233e-05,
      "loss": 3.3400787353515624,
      "step": 6770
    },
    {
      "epoch": 0.678,
      "grad_norm": 0.6425849199295044,
      "learning_rate": 5.170875903899375e-05,
      "loss": 3.406647491455078,
      "step": 6780
    },
    {
      "epoch": 0.679,
      "grad_norm": 0.6630573868751526,
      "learning_rate": 5.141944527902932e-05,
      "loss": 3.3560340881347654,
      "step": 6790
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6306223273277283,
      "learning_rate": 5.113066278822807e-05,
      "loss": 3.27835693359375,
      "step": 6800
    },
    {
      "epoch": 0.681,
      "grad_norm": 0.7350035905838013,
      "learning_rate": 5.084241472466897e-05,
      "loss": 3.368451690673828,
      "step": 6810
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.6573901176452637,
      "learning_rate": 5.055470424058666e-05,
      "loss": 3.437656021118164,
      "step": 6820
    },
    {
      "epoch": 0.683,
      "grad_norm": 0.5916248559951782,
      "learning_rate": 5.026753448233703e-05,
      "loss": 3.410802459716797,
      "step": 6830
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.5933331847190857,
      "learning_rate": 4.99809085903624e-05,
      "loss": 3.3656959533691406,
      "step": 6840
    },
    {
      "epoch": 0.685,
      "grad_norm": 0.5601134300231934,
      "learning_rate": 4.9694829699157695e-05,
      "loss": 3.2687145233154298,
      "step": 6850
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.5328415632247925,
      "learning_rate": 4.940930093723578e-05,
      "loss": 3.2618335723876952,
      "step": 6860
    },
    {
      "epoch": 0.687,
      "grad_norm": 0.6314665079116821,
      "learning_rate": 4.9124325427093455e-05,
      "loss": 3.427937316894531,
      "step": 6870
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.5896925330162048,
      "learning_rate": 4.883990628517725e-05,
      "loss": 3.269194793701172,
      "step": 6880
    },
    {
      "epoch": 0.689,
      "grad_norm": 0.5918858647346497,
      "learning_rate": 4.8556046621849346e-05,
      "loss": 3.2424644470214843,
      "step": 6890
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5558356642723083,
      "learning_rate": 4.827274954135358e-05,
      "loss": 3.2991458892822267,
      "step": 6900
    },
    {
      "epoch": 0.691,
      "grad_norm": 0.6418652534484863,
      "learning_rate": 4.7990018141781344e-05,
      "loss": 3.3685096740722655,
      "step": 6910
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.772983968257904,
      "learning_rate": 4.770785551503798e-05,
      "loss": 3.2606552124023436,
      "step": 6920
    },
    {
      "epoch": 0.693,
      "grad_norm": 0.5610319972038269,
      "learning_rate": 4.7426264746808755e-05,
      "loss": 3.241876983642578,
      "step": 6930
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.5964747667312622,
      "learning_rate": 4.714524891652524e-05,
      "loss": 3.3474945068359374,
      "step": 6940
    },
    {
      "epoch": 0.695,
      "grad_norm": 0.7057782411575317,
      "learning_rate": 4.686481109733146e-05,
      "loss": 3.364746856689453,
      "step": 6950
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.5429649353027344,
      "learning_rate": 4.658495435605051e-05,
      "loss": 3.219074249267578,
      "step": 6960
    },
    {
      "epoch": 0.697,
      "grad_norm": 0.5928279161453247,
      "learning_rate": 4.630568175315088e-05,
      "loss": 3.3083282470703126,
      "step": 6970
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.5792699456214905,
      "learning_rate": 4.6026996342713e-05,
      "loss": 3.290052795410156,
      "step": 6980
    },
    {
      "epoch": 0.699,
      "grad_norm": 0.5817694664001465,
      "learning_rate": 4.574890117239592e-05,
      "loss": 3.337035369873047,
      "step": 6990
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7038497924804688,
      "learning_rate": 4.5471399283403784e-05,
      "loss": 3.2541748046875,
      "step": 7000
    },
    {
      "epoch": 0.701,
      "grad_norm": 0.7261914610862732,
      "learning_rate": 4.5194493710452825e-05,
      "loss": 3.267203521728516,
      "step": 7010
    },
    {
      "epoch": 0.702,
      "grad_norm": 0.5465496182441711,
      "learning_rate": 4.491818748173804e-05,
      "loss": 3.299674224853516,
      "step": 7020
    },
    {
      "epoch": 0.703,
      "grad_norm": 0.596564531326294,
      "learning_rate": 4.464248361890006e-05,
      "loss": 3.2302669525146483,
      "step": 7030
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.7128263115882874,
      "learning_rate": 4.43673851369922e-05,
      "loss": 3.275185394287109,
      "step": 7040
    },
    {
      "epoch": 0.705,
      "grad_norm": 0.637469470500946,
      "learning_rate": 4.409289504444732e-05,
      "loss": 3.272734451293945,
      "step": 7050
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.5867853164672852,
      "learning_rate": 4.381901634304512e-05,
      "loss": 3.266620635986328,
      "step": 7060
    },
    {
      "epoch": 0.707,
      "grad_norm": 0.5334182977676392,
      "learning_rate": 4.35457520278792e-05,
      "loss": 3.1406517028808594,
      "step": 7070
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.6416709423065186,
      "learning_rate": 4.327310508732437e-05,
      "loss": 3.283594512939453,
      "step": 7080
    },
    {
      "epoch": 0.709,
      "grad_norm": 0.5375783443450928,
      "learning_rate": 4.3001078503003825e-05,
      "loss": 3.240416717529297,
      "step": 7090
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.6176676154136658,
      "learning_rate": 4.272967524975673e-05,
      "loss": 3.09521427154541,
      "step": 7100
    },
    {
      "epoch": 0.711,
      "grad_norm": 0.5920681357383728,
      "learning_rate": 4.245889829560559e-05,
      "loss": 3.2712093353271485,
      "step": 7110
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.5535682439804077,
      "learning_rate": 4.218875060172379e-05,
      "loss": 3.16171932220459,
      "step": 7120
    },
    {
      "epoch": 0.713,
      "grad_norm": 0.6745008230209351,
      "learning_rate": 4.191923512240327e-05,
      "loss": 3.1789222717285157,
      "step": 7130
    },
    {
      "epoch": 0.714,
      "grad_norm": 0.5194398164749146,
      "learning_rate": 4.165035480502204e-05,
      "loss": 3.3986759185791016,
      "step": 7140
    },
    {
      "epoch": 0.715,
      "grad_norm": 0.5144760012626648,
      "learning_rate": 4.138211259001222e-05,
      "loss": 3.2070045471191406,
      "step": 7150
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.6644211411476135,
      "learning_rate": 4.1114511410827714e-05,
      "loss": 3.2547012329101563,
      "step": 7160
    },
    {
      "epoch": 0.717,
      "grad_norm": 0.6074444651603699,
      "learning_rate": 4.084755419391213e-05,
      "loss": 3.2359996795654298,
      "step": 7170
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.5718947649002075,
      "learning_rate": 4.058124385866685e-05,
      "loss": 3.3530162811279296,
      "step": 7180
    },
    {
      "epoch": 0.719,
      "grad_norm": 0.6775391101837158,
      "learning_rate": 4.031558331741897e-05,
      "loss": 3.3261619567871095,
      "step": 7190
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5684776306152344,
      "learning_rate": 4.005057547538964e-05,
      "loss": 3.395799255371094,
      "step": 7200
    },
    {
      "epoch": 0.721,
      "grad_norm": 0.5319148302078247,
      "learning_rate": 3.978622323066217e-05,
      "loss": 3.2445823669433596,
      "step": 7210
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.7098920941352844,
      "learning_rate": 3.952252947415038e-05,
      "loss": 3.3086097717285154,
      "step": 7220
    },
    {
      "epoch": 0.723,
      "grad_norm": 0.5932138562202454,
      "learning_rate": 3.925949708956689e-05,
      "loss": 3.2274311065673826,
      "step": 7230
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.5264906883239746,
      "learning_rate": 3.8997128953391727e-05,
      "loss": 3.2914695739746094,
      "step": 7240
    },
    {
      "epoch": 0.725,
      "grad_norm": 0.5507687330245972,
      "learning_rate": 3.873542793484081e-05,
      "loss": 3.2581558227539062,
      "step": 7250
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.5904282927513123,
      "learning_rate": 3.847439689583454e-05,
      "loss": 3.3060516357421874,
      "step": 7260
    },
    {
      "epoch": 0.727,
      "grad_norm": 0.5618739128112793,
      "learning_rate": 3.821403869096658e-05,
      "loss": 3.275089645385742,
      "step": 7270
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.657352864742279,
      "learning_rate": 3.7954356167472485e-05,
      "loss": 3.250388336181641,
      "step": 7280
    },
    {
      "epoch": 0.729,
      "grad_norm": 0.5361801981925964,
      "learning_rate": 3.7695352165198774e-05,
      "loss": 3.2737518310546876,
      "step": 7290
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7433923482894897,
      "learning_rate": 3.743702951657163e-05,
      "loss": 3.323565673828125,
      "step": 7300
    },
    {
      "epoch": 0.731,
      "grad_norm": 0.6384003162384033,
      "learning_rate": 3.717939104656626e-05,
      "loss": 3.3464263916015624,
      "step": 7310
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.5765330195426941,
      "learning_rate": 3.692243957267568e-05,
      "loss": 3.2238090515136717,
      "step": 7320
    },
    {
      "epoch": 0.733,
      "grad_norm": 0.5962571501731873,
      "learning_rate": 3.6666177904879994e-05,
      "loss": 3.338971710205078,
      "step": 7330
    },
    {
      "epoch": 0.734,
      "grad_norm": 0.5337924361228943,
      "learning_rate": 3.641060884561579e-05,
      "loss": 3.3450973510742186,
      "step": 7340
    },
    {
      "epoch": 0.735,
      "grad_norm": 0.5177887678146362,
      "learning_rate": 3.615573518974525e-05,
      "loss": 3.335235595703125,
      "step": 7350
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.6168720722198486,
      "learning_rate": 3.5901559724526e-05,
      "loss": 3.2904739379882812,
      "step": 7360
    },
    {
      "epoch": 0.737,
      "grad_norm": 0.6027379035949707,
      "learning_rate": 3.5648085229580095e-05,
      "loss": 3.3977287292480467,
      "step": 7370
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.5507408380508423,
      "learning_rate": 3.5395314476864026e-05,
      "loss": 3.442243957519531,
      "step": 7380
    },
    {
      "epoch": 0.739,
      "grad_norm": 0.5507969856262207,
      "learning_rate": 3.5143250230638245e-05,
      "loss": 3.2021720886230467,
      "step": 7390
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5718161463737488,
      "learning_rate": 3.4891895247436976e-05,
      "loss": 3.3434303283691404,
      "step": 7400
    },
    {
      "epoch": 0.741,
      "grad_norm": 0.5989702343940735,
      "learning_rate": 3.464125227603808e-05,
      "loss": 3.316603088378906,
      "step": 7410
    },
    {
      "epoch": 0.742,
      "grad_norm": 0.6042175889015198,
      "learning_rate": 3.439132405743286e-05,
      "loss": 3.341657257080078,
      "step": 7420
    },
    {
      "epoch": 0.743,
      "grad_norm": 0.672458291053772,
      "learning_rate": 3.4142113324796344e-05,
      "loss": 3.235280990600586,
      "step": 7430
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.6673842668533325,
      "learning_rate": 3.389362280345707e-05,
      "loss": 3.3190284729003907,
      "step": 7440
    },
    {
      "epoch": 0.745,
      "grad_norm": 0.6822168231010437,
      "learning_rate": 3.364585521086767e-05,
      "loss": 3.22139892578125,
      "step": 7450
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.5346335768699646,
      "learning_rate": 3.339881325657484e-05,
      "loss": 3.2692192077636717,
      "step": 7460
    },
    {
      "epoch": 0.747,
      "grad_norm": 0.5776004195213318,
      "learning_rate": 3.315249964218976e-05,
      "loss": 3.3009048461914063,
      "step": 7470
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.5709183216094971,
      "learning_rate": 3.290691706135871e-05,
      "loss": 3.375341033935547,
      "step": 7480
    },
    {
      "epoch": 0.749,
      "grad_norm": 0.5566404461860657,
      "learning_rate": 3.266206819973336e-05,
      "loss": 3.2381568908691407,
      "step": 7490
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6487706899642944,
      "learning_rate": 3.241795573494173e-05,
      "loss": 3.2587295532226563,
      "step": 7500
    },
    {
      "epoch": 0.751,
      "grad_norm": 0.5849614143371582,
      "learning_rate": 3.217458233655856e-05,
      "loss": 3.349720764160156,
      "step": 7510
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.5646069049835205,
      "learning_rate": 3.193195066607636e-05,
      "loss": 3.269457244873047,
      "step": 7520
    },
    {
      "epoch": 0.753,
      "grad_norm": 0.545831561088562,
      "learning_rate": 3.1690063376876246e-05,
      "loss": 3.329206085205078,
      "step": 7530
    },
    {
      "epoch": 0.754,
      "grad_norm": 0.5804587006568909,
      "learning_rate": 3.144892311419876e-05,
      "loss": 3.3934288024902344,
      "step": 7540
    },
    {
      "epoch": 0.755,
      "grad_norm": 0.5214906930923462,
      "learning_rate": 3.12085325151153e-05,
      "loss": 3.2741878509521483,
      "step": 7550
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.5385957360267639,
      "learning_rate": 3.096889420849886e-05,
      "loss": 3.2283992767333984,
      "step": 7560
    },
    {
      "epoch": 0.757,
      "grad_norm": 0.5632297396659851,
      "learning_rate": 3.0730010814995626e-05,
      "loss": 3.30167236328125,
      "step": 7570
    },
    {
      "epoch": 0.758,
      "grad_norm": 0.5182475447654724,
      "learning_rate": 3.0491884946996054e-05,
      "loss": 3.2385730743408203,
      "step": 7580
    },
    {
      "epoch": 0.759,
      "grad_norm": 0.5148998498916626,
      "learning_rate": 3.025451920860649e-05,
      "loss": 3.3078025817871093,
      "step": 7590
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5538557171821594,
      "learning_rate": 3.001791619562071e-05,
      "loss": 3.188834381103516,
      "step": 7600
    },
    {
      "epoch": 0.761,
      "grad_norm": 0.5732308626174927,
      "learning_rate": 2.9782078495491262e-05,
      "loss": 3.3667579650878907,
      "step": 7610
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.5839775800704956,
      "learning_rate": 2.9547008687301502e-05,
      "loss": 3.2624549865722656,
      "step": 7620
    },
    {
      "epoch": 0.763,
      "grad_norm": 0.5456337332725525,
      "learning_rate": 2.9312709341737153e-05,
      "loss": 3.182592010498047,
      "step": 7630
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.547325074672699,
      "learning_rate": 2.907918302105833e-05,
      "loss": 3.2775733947753904,
      "step": 7640
    },
    {
      "epoch": 0.765,
      "grad_norm": 0.6242472529411316,
      "learning_rate": 2.8846432279071467e-05,
      "loss": 3.281597137451172,
      "step": 7650
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.5570681095123291,
      "learning_rate": 2.8614459661101402e-05,
      "loss": 3.2667316436767577,
      "step": 7660
    },
    {
      "epoch": 0.767,
      "grad_norm": 0.5491817593574524,
      "learning_rate": 2.8383267703963522e-05,
      "loss": 3.3433570861816406,
      "step": 7670
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.5825505256652832,
      "learning_rate": 2.8152858935936e-05,
      "loss": 3.314492034912109,
      "step": 7680
    },
    {
      "epoch": 0.769,
      "grad_norm": 0.5554798245429993,
      "learning_rate": 2.7923235876732223e-05,
      "loss": 3.3332977294921875,
      "step": 7690
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5039050579071045,
      "learning_rate": 2.7694401037473185e-05,
      "loss": 3.358651351928711,
      "step": 7700
    },
    {
      "epoch": 0.771,
      "grad_norm": 0.543664813041687,
      "learning_rate": 2.746635692066004e-05,
      "loss": 3.341577911376953,
      "step": 7710
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.5973897576332092,
      "learning_rate": 2.7239106020146666e-05,
      "loss": 3.3563983917236326,
      "step": 7720
    },
    {
      "epoch": 0.773,
      "grad_norm": 0.5725898742675781,
      "learning_rate": 2.701265082111253e-05,
      "loss": 3.2859027862548826,
      "step": 7730
    },
    {
      "epoch": 0.774,
      "grad_norm": 0.6573544144630432,
      "learning_rate": 2.678699380003541e-05,
      "loss": 3.3413490295410155,
      "step": 7740
    },
    {
      "epoch": 0.775,
      "grad_norm": 0.5749434232711792,
      "learning_rate": 2.6562137424664324e-05,
      "loss": 3.348063659667969,
      "step": 7750
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.6746200323104858,
      "learning_rate": 2.6338084153992613e-05,
      "loss": 3.3089950561523436,
      "step": 7760
    },
    {
      "epoch": 0.777,
      "grad_norm": 0.573680579662323,
      "learning_rate": 2.6114836438230884e-05,
      "loss": 3.4571548461914063,
      "step": 7770
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.5210006833076477,
      "learning_rate": 2.589239671878041e-05,
      "loss": 3.35076904296875,
      "step": 7780
    },
    {
      "epoch": 0.779,
      "grad_norm": 0.5155611634254456,
      "learning_rate": 2.5670767428206334e-05,
      "loss": 3.249658966064453,
      "step": 7790
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5406736135482788,
      "learning_rate": 2.544995099021107e-05,
      "loss": 3.20379638671875,
      "step": 7800
    },
    {
      "epoch": 0.781,
      "grad_norm": 0.5363479852676392,
      "learning_rate": 2.5229949819607824e-05,
      "loss": 3.35313720703125,
      "step": 7810
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.5995455980300903,
      "learning_rate": 2.501076632229411e-05,
      "loss": 3.430755615234375,
      "step": 7820
    },
    {
      "epoch": 0.783,
      "grad_norm": 0.5543540716171265,
      "learning_rate": 2.4792402895225553e-05,
      "loss": 3.278521728515625,
      "step": 7830
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.5739737749099731,
      "learning_rate": 2.4574861926389615e-05,
      "loss": 3.2458751678466795,
      "step": 7840
    },
    {
      "epoch": 0.785,
      "grad_norm": 0.5541499257087708,
      "learning_rate": 2.4358145794779508e-05,
      "loss": 3.360853576660156,
      "step": 7850
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.5331305265426636,
      "learning_rate": 2.414225687036805e-05,
      "loss": 3.3310672760009767,
      "step": 7860
    },
    {
      "epoch": 0.787,
      "grad_norm": 0.5857387185096741,
      "learning_rate": 2.392719751408199e-05,
      "loss": 3.2632461547851563,
      "step": 7870
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.5080597400665283,
      "learning_rate": 2.3712970077775964e-05,
      "loss": 3.2718784332275392,
      "step": 7880
    },
    {
      "epoch": 0.789,
      "grad_norm": 0.5076976418495178,
      "learning_rate": 2.349957690420691e-05,
      "loss": 3.2448493957519533,
      "step": 7890
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5474519729614258,
      "learning_rate": 2.3287020327008403e-05,
      "loss": 3.2573211669921873,
      "step": 7900
    },
    {
      "epoch": 0.791,
      "grad_norm": 0.6605607867240906,
      "learning_rate": 2.3075302670665066e-05,
      "loss": 3.336904525756836,
      "step": 7910
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.5917698740959167,
      "learning_rate": 2.2864426250487302e-05,
      "loss": 3.3484943389892576,
      "step": 7920
    },
    {
      "epoch": 0.793,
      "grad_norm": 0.5651469826698303,
      "learning_rate": 2.2654393372585848e-05,
      "loss": 3.2923912048339843,
      "step": 7930
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.7481618523597717,
      "learning_rate": 2.24452063338466e-05,
      "loss": 3.3820640563964846,
      "step": 7940
    },
    {
      "epoch": 0.795,
      "grad_norm": 0.5606748461723328,
      "learning_rate": 2.223686742190555e-05,
      "loss": 3.2421913146972656,
      "step": 7950
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.5224206447601318,
      "learning_rate": 2.2029378915123587e-05,
      "loss": 3.2815006256103514,
      "step": 7960
    },
    {
      "epoch": 0.797,
      "grad_norm": 0.5654622912406921,
      "learning_rate": 2.1822743082561814e-05,
      "loss": 3.3593517303466798,
      "step": 7970
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.5794079899787903,
      "learning_rate": 2.161696218395658e-05,
      "loss": 3.4465927124023437,
      "step": 7980
    },
    {
      "epoch": 0.799,
      "grad_norm": 0.5236971378326416,
      "learning_rate": 2.141203846969486e-05,
      "loss": 3.243749237060547,
      "step": 7990
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5712748169898987,
      "learning_rate": 2.1207974180789502e-05,
      "loss": 3.343666839599609,
      "step": 8000
    },
    {
      "epoch": 0.801,
      "grad_norm": 0.7116403579711914,
      "learning_rate": 2.1004771548854907e-05,
      "loss": 3.3135520935058596,
      "step": 8010
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.6322494149208069,
      "learning_rate": 2.0802432796082493e-05,
      "loss": 3.3885589599609376,
      "step": 8020
    },
    {
      "epoch": 0.803,
      "grad_norm": 0.5378990769386292,
      "learning_rate": 2.0600960135216462e-05,
      "loss": 3.3157615661621094,
      "step": 8030
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.5889938473701477,
      "learning_rate": 2.0400355769529557e-05,
      "loss": 3.310374450683594,
      "step": 8040
    },
    {
      "epoch": 0.805,
      "grad_norm": 0.5741366147994995,
      "learning_rate": 2.0200621892798942e-05,
      "loss": 3.3729011535644533,
      "step": 8050
    },
    {
      "epoch": 0.806,
      "grad_norm": 0.6490705013275146,
      "learning_rate": 2.0001760689282323e-05,
      "loss": 3.2855796813964844,
      "step": 8060
    },
    {
      "epoch": 0.807,
      "grad_norm": 0.5436992049217224,
      "learning_rate": 1.9803774333693946e-05,
      "loss": 3.270477294921875,
      "step": 8070
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.5672959089279175,
      "learning_rate": 1.960666499118089e-05,
      "loss": 3.3581192016601564,
      "step": 8080
    },
    {
      "epoch": 0.809,
      "grad_norm": 0.5380074977874756,
      "learning_rate": 1.9410434817299363e-05,
      "loss": 3.2324256896972656,
      "step": 8090
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5884382128715515,
      "learning_rate": 1.9215085957991052e-05,
      "loss": 3.3777339935302733,
      "step": 8100
    },
    {
      "epoch": 0.811,
      "grad_norm": 0.6653020977973938,
      "learning_rate": 1.902062054955982e-05,
      "loss": 3.3532341003417967,
      "step": 8110
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.8454363942146301,
      "learning_rate": 1.8827040718648225e-05,
      "loss": 3.366088104248047,
      "step": 8120
    },
    {
      "epoch": 0.813,
      "grad_norm": 0.5459985136985779,
      "learning_rate": 1.863434858221429e-05,
      "loss": 3.323456573486328,
      "step": 8130
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.7710661888122559,
      "learning_rate": 1.8442546247508307e-05,
      "loss": 3.1747488021850585,
      "step": 8140
    },
    {
      "epoch": 0.815,
      "grad_norm": 0.5796877145767212,
      "learning_rate": 1.8251635812049896e-05,
      "loss": 3.342686080932617,
      "step": 8150
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.53633713722229,
      "learning_rate": 1.8061619363605e-05,
      "loss": 3.2689586639404298,
      "step": 8160
    },
    {
      "epoch": 0.817,
      "grad_norm": 0.6265182495117188,
      "learning_rate": 1.7872498980163067e-05,
      "loss": 3.265943908691406,
      "step": 8170
    },
    {
      "epoch": 0.818,
      "grad_norm": 0.6161515712738037,
      "learning_rate": 1.7684276729914305e-05,
      "loss": 3.292559814453125,
      "step": 8180
    },
    {
      "epoch": 0.819,
      "grad_norm": 0.5597274899482727,
      "learning_rate": 1.749695467122706e-05,
      "loss": 3.3021610260009764,
      "step": 8190
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5746941566467285,
      "learning_rate": 1.7310534852625404e-05,
      "loss": 3.236225128173828,
      "step": 8200
    },
    {
      "epoch": 0.821,
      "grad_norm": 0.6512794494628906,
      "learning_rate": 1.712501931276653e-05,
      "loss": 3.301092529296875,
      "step": 8210
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.535951554775238,
      "learning_rate": 1.6940410080418723e-05,
      "loss": 3.413623046875,
      "step": 8220
    },
    {
      "epoch": 0.823,
      "grad_norm": 0.580573320388794,
      "learning_rate": 1.6756709174438978e-05,
      "loss": 3.314319610595703,
      "step": 8230
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.5554879903793335,
      "learning_rate": 1.6573918603750926e-05,
      "loss": 3.3839969635009766,
      "step": 8240
    },
    {
      "epoch": 0.825,
      "grad_norm": 0.5198671221733093,
      "learning_rate": 1.6392040367323015e-05,
      "loss": 3.267973709106445,
      "step": 8250
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.5812584757804871,
      "learning_rate": 1.621107645414641e-05,
      "loss": 3.3751026153564454,
      "step": 8260
    },
    {
      "epoch": 0.827,
      "grad_norm": 0.7090417146682739,
      "learning_rate": 1.603102884321358e-05,
      "loss": 3.381013107299805,
      "step": 8270
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.5810737013816833,
      "learning_rate": 1.585189950349627e-05,
      "loss": 3.2726131439208985,
      "step": 8280
    },
    {
      "epoch": 0.829,
      "grad_norm": 0.5707302689552307,
      "learning_rate": 1.5673690393924312e-05,
      "loss": 3.2773696899414064,
      "step": 8290
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5487689971923828,
      "learning_rate": 1.5496403463363896e-05,
      "loss": 3.1862133026123045,
      "step": 8300
    },
    {
      "epoch": 0.831,
      "grad_norm": 0.5835641026496887,
      "learning_rate": 1.5320040650596522e-05,
      "loss": 3.2724941253662108,
      "step": 8310
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.5516979098320007,
      "learning_rate": 1.5144603884297703e-05,
      "loss": 3.249259185791016,
      "step": 8320
    },
    {
      "epoch": 0.833,
      "grad_norm": 0.6163832545280457,
      "learning_rate": 1.4970095083015757e-05,
      "loss": 3.250591278076172,
      "step": 8330
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.6584014892578125,
      "learning_rate": 1.4796516155150996e-05,
      "loss": 3.2912757873535154,
      "step": 8340
    },
    {
      "epoch": 0.835,
      "grad_norm": 0.5662320256233215,
      "learning_rate": 1.4623868998934742e-05,
      "loss": 3.2910205841064455,
      "step": 8350
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.6626244187355042,
      "learning_rate": 1.4452155502408649e-05,
      "loss": 3.2477272033691404,
      "step": 8360
    },
    {
      "epoch": 0.837,
      "grad_norm": 0.5572372674942017,
      "learning_rate": 1.4281377543404006e-05,
      "loss": 3.2527355194091796,
      "step": 8370
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.5663153529167175,
      "learning_rate": 1.4111536989521213e-05,
      "loss": 3.3268417358398437,
      "step": 8380
    },
    {
      "epoch": 0.839,
      "grad_norm": 0.6475211977958679,
      "learning_rate": 1.3942635698109375e-05,
      "loss": 3.2496551513671874,
      "step": 8390
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5726585984230042,
      "learning_rate": 1.3774675516245906e-05,
      "loss": 3.286568450927734,
      "step": 8400
    },
    {
      "epoch": 0.841,
      "grad_norm": 0.5539810061454773,
      "learning_rate": 1.3607658280716473e-05,
      "loss": 3.1763076782226562,
      "step": 8410
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.5224063992500305,
      "learning_rate": 1.3441585817994817e-05,
      "loss": 3.287486267089844,
      "step": 8420
    },
    {
      "epoch": 0.843,
      "grad_norm": 0.582256555557251,
      "learning_rate": 1.3276459944222785e-05,
      "loss": 3.2822196960449217,
      "step": 8430
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.5346740484237671,
      "learning_rate": 1.311228246519044e-05,
      "loss": 3.1780576705932617,
      "step": 8440
    },
    {
      "epoch": 0.845,
      "grad_norm": 0.5655465722084045,
      "learning_rate": 1.2949055176316416e-05,
      "loss": 3.1385475158691407,
      "step": 8450
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.514869213104248,
      "learning_rate": 1.2786779862628184e-05,
      "loss": 3.1950870513916017,
      "step": 8460
    },
    {
      "epoch": 0.847,
      "grad_norm": 0.5570836067199707,
      "learning_rate": 1.2625458298742588e-05,
      "loss": 3.193470001220703,
      "step": 8470
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.5663630366325378,
      "learning_rate": 1.2465092248846422e-05,
      "loss": 3.184743118286133,
      "step": 8480
    },
    {
      "epoch": 0.849,
      "grad_norm": 0.5246147513389587,
      "learning_rate": 1.2305683466677064e-05,
      "loss": 3.159442901611328,
      "step": 8490
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5354295969009399,
      "learning_rate": 1.2147233695503435e-05,
      "loss": 3.2410423278808596,
      "step": 8500
    },
    {
      "epoch": 0.851,
      "grad_norm": 0.5339637994766235,
      "learning_rate": 1.1989744668106861e-05,
      "loss": 3.1145458221435547,
      "step": 8510
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.5150908827781677,
      "learning_rate": 1.1833218106762111e-05,
      "loss": 3.095259094238281,
      "step": 8520
    },
    {
      "epoch": 0.853,
      "grad_norm": 0.5691410303115845,
      "learning_rate": 1.1677655723218594e-05,
      "loss": 3.2062965393066407,
      "step": 8530
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.5352501273155212,
      "learning_rate": 1.1523059218681587e-05,
      "loss": 3.1426538467407226,
      "step": 8540
    },
    {
      "epoch": 0.855,
      "grad_norm": 0.5989806652069092,
      "learning_rate": 1.1369430283793692e-05,
      "loss": 3.2057510375976563,
      "step": 8550
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.5571030974388123,
      "learning_rate": 1.1216770598616345e-05,
      "loss": 3.1466861724853517,
      "step": 8560
    },
    {
      "epoch": 0.857,
      "grad_norm": 0.5288048386573792,
      "learning_rate": 1.1065081832611423e-05,
      "loss": 3.218009185791016,
      "step": 8570
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.5213762521743774,
      "learning_rate": 1.0914365644622926e-05,
      "loss": 3.2676414489746093,
      "step": 8580
    },
    {
      "epoch": 0.859,
      "grad_norm": 0.768602728843689,
      "learning_rate": 1.0764623682858965e-05,
      "loss": 3.2673858642578124,
      "step": 8590
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.6353641152381897,
      "learning_rate": 1.0615857584873623e-05,
      "loss": 3.2108795166015627,
      "step": 8600
    },
    {
      "epoch": 0.861,
      "grad_norm": 0.6073545813560486,
      "learning_rate": 1.0468068977549138e-05,
      "loss": 3.24593505859375,
      "step": 8610
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.5335069894790649,
      "learning_rate": 1.032125947707805e-05,
      "loss": 3.240306854248047,
      "step": 8620
    },
    {
      "epoch": 0.863,
      "grad_norm": 0.5833173394203186,
      "learning_rate": 1.0175430688945486e-05,
      "loss": 3.5032169342041017,
      "step": 8630
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.5387097597122192,
      "learning_rate": 1.0030584207911753e-05,
      "loss": 3.1924625396728517,
      "step": 8640
    },
    {
      "epoch": 0.865,
      "grad_norm": 0.5405256748199463,
      "learning_rate": 9.886721617994742e-06,
      "loss": 3.1322149276733398,
      "step": 8650
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.5946422815322876,
      "learning_rate": 9.743844492452692e-06,
      "loss": 3.2625667572021486,
      "step": 8660
    },
    {
      "epoch": 0.867,
      "grad_norm": 0.5521227121353149,
      "learning_rate": 9.601954393766977e-06,
      "loss": 3.2599853515625,
      "step": 8670
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.8592839241027832,
      "learning_rate": 9.461052873624942e-06,
      "loss": 3.3097152709960938,
      "step": 8680
    },
    {
      "epoch": 0.869,
      "grad_norm": 0.5284611582756042,
      "learning_rate": 9.321141472903062e-06,
      "loss": 3.2661388397216795,
      "step": 8690
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6615422368049622,
      "learning_rate": 9.182221721649998e-06,
      "loss": 3.2094482421875,
      "step": 8700
    },
    {
      "epoch": 0.871,
      "grad_norm": 0.5840532779693604,
      "learning_rate": 9.044295139069913e-06,
      "loss": 3.2160259246826173,
      "step": 8710
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.5256271958351135,
      "learning_rate": 8.907363233505771e-06,
      "loss": 3.2145225524902346,
      "step": 8720
    },
    {
      "epoch": 0.873,
      "grad_norm": 0.6713451147079468,
      "learning_rate": 8.771427502422979e-06,
      "loss": 3.310803985595703,
      "step": 8730
    },
    {
      "epoch": 0.874,
      "grad_norm": 0.5493841767311096,
      "learning_rate": 8.636489432392914e-06,
      "loss": 3.228730010986328,
      "step": 8740
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.5322839617729187,
      "learning_rate": 8.502550499076688e-06,
      "loss": 3.2603679656982423,
      "step": 8750
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.5735679268836975,
      "learning_rate": 8.369612167209028e-06,
      "loss": 3.3153900146484374,
      "step": 8760
    },
    {
      "epoch": 0.877,
      "grad_norm": 0.5163535475730896,
      "learning_rate": 8.237675890582208e-06,
      "loss": 3.3361961364746096,
      "step": 8770
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.5604974031448364,
      "learning_rate": 8.106743112030223e-06,
      "loss": 3.248381423950195,
      "step": 8780
    },
    {
      "epoch": 0.879,
      "grad_norm": 0.5509691834449768,
      "learning_rate": 7.976815263412963e-06,
      "loss": 3.2776134490966795,
      "step": 8790
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5757589340209961,
      "learning_rate": 7.847893765600568e-06,
      "loss": 3.3460174560546876,
      "step": 8800
    },
    {
      "epoch": 0.881,
      "grad_norm": 0.5490387678146362,
      "learning_rate": 7.719980028457896e-06,
      "loss": 3.323968505859375,
      "step": 8810
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.5322655439376831,
      "learning_rate": 7.593075450829046e-06,
      "loss": 3.259532928466797,
      "step": 8820
    },
    {
      "epoch": 0.883,
      "grad_norm": 0.5910403728485107,
      "learning_rate": 7.467181420522173e-06,
      "loss": 3.35705680847168,
      "step": 8830
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.5726730227470398,
      "learning_rate": 7.34229931429421e-06,
      "loss": 3.348969268798828,
      "step": 8840
    },
    {
      "epoch": 0.885,
      "grad_norm": 0.5818085074424744,
      "learning_rate": 7.218430497835893e-06,
      "loss": 3.3548179626464845,
      "step": 8850
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.7846088409423828,
      "learning_rate": 7.0955763257567145e-06,
      "loss": 3.3330883026123046,
      "step": 8860
    },
    {
      "epoch": 0.887,
      "grad_norm": 0.5434913039207458,
      "learning_rate": 6.973738141570229e-06,
      "loss": 3.3421844482421874,
      "step": 8870
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.5919123888015747,
      "learning_rate": 6.852917277679305e-06,
      "loss": 3.3192184448242186,
      "step": 8880
    },
    {
      "epoch": 0.889,
      "grad_norm": 0.5196734070777893,
      "learning_rate": 6.733115055361538e-06,
      "loss": 3.367898941040039,
      "step": 8890
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5231317281723022,
      "learning_rate": 6.614332784754884e-06,
      "loss": 3.2795528411865233,
      "step": 8900
    },
    {
      "epoch": 0.891,
      "grad_norm": 0.6019504070281982,
      "learning_rate": 6.496571764843162e-06,
      "loss": 3.341367721557617,
      "step": 8910
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.526096761226654,
      "learning_rate": 6.379833283442061e-06,
      "loss": 3.3155712127685546,
      "step": 8920
    },
    {
      "epoch": 0.893,
      "grad_norm": 0.5712595582008362,
      "learning_rate": 6.264118617184878e-06,
      "loss": 3.3145851135253905,
      "step": 8930
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.5692469477653503,
      "learning_rate": 6.14942903150868e-06,
      "loss": 3.2928466796875,
      "step": 8940
    },
    {
      "epoch": 0.895,
      "grad_norm": 0.5580922961235046,
      "learning_rate": 6.035765780640424e-06,
      "loss": 3.2733516693115234,
      "step": 8950
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.5351900458335876,
      "learning_rate": 5.923130107583175e-06,
      "loss": 3.207528305053711,
      "step": 8960
    },
    {
      "epoch": 0.897,
      "grad_norm": 0.5516851544380188,
      "learning_rate": 5.811523244102645e-06,
      "loss": 3.2690643310546874,
      "step": 8970
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.5206120014190674,
      "learning_rate": 5.700946410713548e-06,
      "loss": 3.2711753845214844,
      "step": 8980
    },
    {
      "epoch": 0.899,
      "grad_norm": 0.592097282409668,
      "learning_rate": 5.591400816666492e-06,
      "loss": 3.2979400634765623,
      "step": 8990
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6678906083106995,
      "learning_rate": 5.482887659934499e-06,
      "loss": 3.4632102966308596,
      "step": 9000
    },
    {
      "epoch": 0.901,
      "grad_norm": 0.5334870219230652,
      "learning_rate": 5.3754081272000766e-06,
      "loss": 3.3372676849365233,
      "step": 9010
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.5410240292549133,
      "learning_rate": 5.268963393842186e-06,
      "loss": 3.372754669189453,
      "step": 9020
    },
    {
      "epoch": 0.903,
      "grad_norm": 0.5744450688362122,
      "learning_rate": 5.163554623923339e-06,
      "loss": 3.205644989013672,
      "step": 9030
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.5435152053833008,
      "learning_rate": 5.0591829701770126e-06,
      "loss": 3.4315513610839843,
      "step": 9040
    },
    {
      "epoch": 0.905,
      "grad_norm": 0.5681019425392151,
      "learning_rate": 4.955849573994853e-06,
      "loss": 3.300423431396484,
      "step": 9050
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.5253817439079285,
      "learning_rate": 4.853555565414336e-06,
      "loss": 3.2931449890136717,
      "step": 9060
    },
    {
      "epoch": 0.907,
      "grad_norm": 0.532925546169281,
      "learning_rate": 4.75230206310634e-06,
      "loss": 3.2370223999023438,
      "step": 9070
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.5376046299934387,
      "learning_rate": 4.652090174362933e-06,
      "loss": 3.304608154296875,
      "step": 9080
    },
    {
      "epoch": 0.909,
      "grad_norm": 0.5790829062461853,
      "learning_rate": 4.552920995085319e-06,
      "loss": 3.263776397705078,
      "step": 9090
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5627272129058838,
      "learning_rate": 4.454795609771722e-06,
      "loss": 3.2813758850097656,
      "step": 9100
    },
    {
      "epoch": 0.911,
      "grad_norm": 0.5572156310081482,
      "learning_rate": 4.3577150915056545e-06,
      "loss": 3.293822479248047,
      "step": 9110
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.5098896622657776,
      "learning_rate": 4.261680501944065e-06,
      "loss": 3.370766448974609,
      "step": 9120
    },
    {
      "epoch": 0.913,
      "grad_norm": 0.5891233682632446,
      "learning_rate": 4.166692891305901e-06,
      "loss": 3.292277526855469,
      "step": 9130
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.5303240418434143,
      "learning_rate": 4.072753298360421e-06,
      "loss": 3.3216793060302736,
      "step": 9140
    },
    {
      "epoch": 0.915,
      "grad_norm": 0.5739419460296631,
      "learning_rate": 3.979862750415953e-06,
      "loss": 3.270555877685547,
      "step": 9150
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.5292540788650513,
      "learning_rate": 3.888022263308677e-06,
      "loss": 3.343107986450195,
      "step": 9160
    },
    {
      "epoch": 0.917,
      "grad_norm": 0.5174182057380676,
      "learning_rate": 3.797232841391407e-06,
      "loss": 3.416941833496094,
      "step": 9170
    },
    {
      "epoch": 0.918,
      "grad_norm": 0.6690304279327393,
      "learning_rate": 3.70749547752276e-06,
      "loss": 3.2996448516845702,
      "step": 9180
    },
    {
      "epoch": 0.919,
      "grad_norm": 0.914656400680542,
      "learning_rate": 3.6188111530561342e-06,
      "loss": 3.3074714660644533,
      "step": 9190
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.543173611164093,
      "learning_rate": 3.5311808378291178e-06,
      "loss": 3.2885147094726563,
      "step": 9200
    },
    {
      "epoch": 0.921,
      "grad_norm": 0.5518679022789001,
      "learning_rate": 3.444605490152786e-06,
      "loss": 3.330775833129883,
      "step": 9210
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.5148446559906006,
      "learning_rate": 3.359086056801253e-06,
      "loss": 3.3429542541503907,
      "step": 9220
    },
    {
      "epoch": 0.923,
      "grad_norm": 0.566531240940094,
      "learning_rate": 3.2746234730013814e-06,
      "loss": 3.255440521240234,
      "step": 9230
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.5787199139595032,
      "learning_rate": 3.191218662422424e-06,
      "loss": 3.319964599609375,
      "step": 9240
    },
    {
      "epoch": 0.925,
      "grad_norm": 0.6114018559455872,
      "learning_rate": 3.108872537166041e-06,
      "loss": 3.3059253692626953,
      "step": 9250
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.5787607431411743,
      "learning_rate": 3.0275859977562547e-06,
      "loss": 3.309949493408203,
      "step": 9260
    },
    {
      "epoch": 0.927,
      "grad_norm": 0.5404456257820129,
      "learning_rate": 2.9473599331296453e-06,
      "loss": 3.2888023376464846,
      "step": 9270
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.5700089931488037,
      "learning_rate": 2.8681952206256024e-06,
      "loss": 3.246033477783203,
      "step": 9280
    },
    {
      "epoch": 0.929,
      "grad_norm": 0.5543075203895569,
      "learning_rate": 2.790092725976723e-06,
      "loss": 3.335582733154297,
      "step": 9290
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5385544300079346,
      "learning_rate": 2.7130533032994067e-06,
      "loss": 3.2566192626953123,
      "step": 9300
    },
    {
      "epoch": 0.931,
      "grad_norm": 0.5907449126243591,
      "learning_rate": 2.6446274459844712e-06,
      "loss": 3.3299945831298827,
      "step": 9310
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.5637195110321045,
      "learning_rate": 2.569610171499226e-06,
      "loss": 3.4018951416015626,
      "step": 9320
    },
    {
      "epoch": 0.933,
      "grad_norm": 0.5973566174507141,
      "learning_rate": 2.495658380147414e-06,
      "loss": 3.477225494384766,
      "step": 9330
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.5683455467224121,
      "learning_rate": 2.4227728806539672e-06,
      "loss": 3.2978351593017576,
      "step": 9340
    },
    {
      "epoch": 0.935,
      "grad_norm": 0.5308555960655212,
      "learning_rate": 2.3509544700830556e-06,
      "loss": 3.310380554199219,
      "step": 9350
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.5950163006782532,
      "learning_rate": 2.2802039338292948e-06,
      "loss": 3.338762664794922,
      "step": 9360
    },
    {
      "epoch": 0.937,
      "grad_norm": 0.5759704113006592,
      "learning_rate": 2.21052204560922e-06,
      "loss": 3.395525360107422,
      "step": 9370
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.5767008066177368,
      "learning_rate": 2.141909567452793e-06,
      "loss": 3.3470550537109376,
      "step": 9380
    },
    {
      "epoch": 0.939,
      "grad_norm": 0.5472702980041504,
      "learning_rate": 2.074367249695086e-06,
      "loss": 3.293049621582031,
      "step": 9390
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.547215461730957,
      "learning_rate": 2.007895830968054e-06,
      "loss": 3.3077880859375,
      "step": 9400
    },
    {
      "epoch": 0.941,
      "grad_norm": 0.4985172748565674,
      "learning_rate": 1.9424960381925093e-06,
      "loss": 3.285900115966797,
      "step": 9410
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.5206795930862427,
      "learning_rate": 1.8781685865700727e-06,
      "loss": 3.3347183227539063,
      "step": 9420
    },
    {
      "epoch": 0.943,
      "grad_norm": 0.5386129021644592,
      "learning_rate": 1.8149141795754776e-06,
      "loss": 3.2009525299072266,
      "step": 9430
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.5351526141166687,
      "learning_rate": 1.7527335089487673e-06,
      "loss": 3.2432479858398438,
      "step": 9440
    },
    {
      "epoch": 0.945,
      "grad_norm": 0.5193876624107361,
      "learning_rate": 1.6916272546877886e-06,
      "loss": 3.3323917388916016,
      "step": 9450
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.5275362730026245,
      "learning_rate": 1.6315960850407875e-06,
      "loss": 3.3038597106933594,
      "step": 9460
    },
    {
      "epoch": 0.947,
      "grad_norm": 0.5964645743370056,
      "learning_rate": 1.5726406564989694e-06,
      "loss": 3.3702957153320314,
      "step": 9470
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.5186903476715088,
      "learning_rate": 1.5147616137894727e-06,
      "loss": 3.335035705566406,
      "step": 9480
    },
    {
      "epoch": 0.949,
      "grad_norm": 0.5502458214759827,
      "learning_rate": 1.4579595898682409e-06,
      "loss": 3.2522850036621094,
      "step": 9490
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5138074159622192,
      "learning_rate": 1.402235205913105e-06,
      "loss": 3.2864910125732423,
      "step": 9500
    },
    {
      "epoch": 0.951,
      "grad_norm": 0.5488574504852295,
      "learning_rate": 1.3475890713170013e-06,
      "loss": 3.319361114501953,
      "step": 9510
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.5387448072433472,
      "learning_rate": 1.2940217836812984e-06,
      "loss": 3.2092208862304688,
      "step": 9520
    },
    {
      "epoch": 0.953,
      "grad_norm": 0.5514314770698547,
      "learning_rate": 1.2415339288092688e-06,
      "loss": 3.283746337890625,
      "step": 9530
    },
    {
      "epoch": 0.954,
      "grad_norm": 0.5287697911262512,
      "learning_rate": 1.1901260806996939e-06,
      "loss": 3.2056739807128904,
      "step": 9540
    },
    {
      "epoch": 0.955,
      "grad_norm": 0.5870466828346252,
      "learning_rate": 1.1397988015405591e-06,
      "loss": 3.2513351440429688,
      "step": 9550
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.5016852021217346,
      "learning_rate": 1.0905526417029243e-06,
      "loss": 3.2226863861083985,
      "step": 9560
    },
    {
      "epoch": 0.957,
      "grad_norm": 0.517249345779419,
      "learning_rate": 1.0423881397349068e-06,
      "loss": 3.289010238647461,
      "step": 9570
    },
    {
      "epoch": 0.958,
      "grad_norm": 0.5414342284202576,
      "learning_rate": 9.953058223557743e-07,
      "loss": 3.2674976348876954,
      "step": 9580
    },
    {
      "epoch": 0.959,
      "grad_norm": 0.5130435228347778,
      "learning_rate": 9.493062044502177e-07,
      "loss": 3.316796875,
      "step": 9590
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5622803568840027,
      "learning_rate": 9.043897890626984e-07,
      "loss": 3.295796203613281,
      "step": 9600
    },
    {
      "epoch": 0.961,
      "grad_norm": 0.5406859517097473,
      "learning_rate": 8.605570673919094e-07,
      "loss": 3.281194305419922,
      "step": 9610
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.5717828273773193,
      "learning_rate": 8.178085187855011e-07,
      "loss": 3.1789260864257813,
      "step": 9620
    },
    {
      "epoch": 0.963,
      "grad_norm": 0.5866690874099731,
      "learning_rate": 7.761446107347304e-07,
      "loss": 3.230789566040039,
      "step": 9630
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.5426310896873474,
      "learning_rate": 7.355657988694308e-07,
      "loss": 3.2540416717529297,
      "step": 9640
    },
    {
      "epoch": 0.965,
      "grad_norm": 0.5363496541976929,
      "learning_rate": 6.960725269529845e-07,
      "loss": 3.2461917877197264,
      "step": 9650
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.5410125851631165,
      "learning_rate": 6.576652268774908e-07,
      "loss": 3.207598114013672,
      "step": 9660
    },
    {
      "epoch": 0.967,
      "grad_norm": 0.5626195073127747,
      "learning_rate": 6.203443186590275e-07,
      "loss": 3.2666065216064455,
      "step": 9670
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.5538455247879028,
      "learning_rate": 5.841102104330643e-07,
      "loss": 3.297545623779297,
      "step": 9680
    },
    {
      "epoch": 0.969,
      "grad_norm": 0.5277932286262512,
      "learning_rate": 5.489632984500115e-07,
      "loss": 3.1660205841064455,
      "step": 9690
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5652297139167786,
      "learning_rate": 5.149039670708567e-07,
      "loss": 3.232364273071289,
      "step": 9700
    },
    {
      "epoch": 0.971,
      "grad_norm": 0.6368861794471741,
      "learning_rate": 4.819325887630011e-07,
      "loss": 3.278098297119141,
      "step": 9710
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.5569401979446411,
      "learning_rate": 4.500495240961411e-07,
      "loss": 3.294664764404297,
      "step": 9720
    },
    {
      "epoch": 0.973,
      "grad_norm": 0.5639516711235046,
      "learning_rate": 4.192551217383822e-07,
      "loss": 3.1939184188842775,
      "step": 9730
    },
    {
      "epoch": 0.974,
      "grad_norm": 0.5448621511459351,
      "learning_rate": 3.8954971845236443e-07,
      "loss": 3.3081939697265623,
      "step": 9740
    },
    {
      "epoch": 0.975,
      "grad_norm": 0.6208817362785339,
      "learning_rate": 3.609336390916207e-07,
      "loss": 3.218395233154297,
      "step": 9750
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.5700429081916809,
      "learning_rate": 3.3340719659701313e-07,
      "loss": 3.3779693603515626,
      "step": 9760
    },
    {
      "epoch": 0.977,
      "grad_norm": 0.5775555968284607,
      "learning_rate": 3.0697069199330243e-07,
      "loss": 3.2535438537597656,
      "step": 9770
    },
    {
      "epoch": 0.978,
      "grad_norm": 0.5635033845901489,
      "learning_rate": 2.816244143858615e-07,
      "loss": 3.1452919006347657,
      "step": 9780
    },
    {
      "epoch": 0.979,
      "grad_norm": 0.5793444514274597,
      "learning_rate": 2.5736864095753377e-07,
      "loss": 3.1370126724243166,
      "step": 9790
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.555690586566925,
      "learning_rate": 2.342036369655687e-07,
      "loss": 3.197304916381836,
      "step": 9800
    },
    {
      "epoch": 0.981,
      "grad_norm": 0.5603624582290649,
      "learning_rate": 2.1212965573873534e-07,
      "loss": 3.209071731567383,
      "step": 9810
    },
    {
      "epoch": 0.982,
      "grad_norm": 0.5377429127693176,
      "learning_rate": 1.911469386745468e-07,
      "loss": 3.2360294342041014,
      "step": 9820
    },
    {
      "epoch": 0.983,
      "grad_norm": 0.5872378349304199,
      "learning_rate": 1.7125571523666228e-07,
      "loss": 3.298314666748047,
      "step": 9830
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.5205584764480591,
      "learning_rate": 1.5245620295228912e-07,
      "loss": 3.1320266723632812,
      "step": 9840
    },
    {
      "epoch": 0.985,
      "grad_norm": 0.5475061535835266,
      "learning_rate": 1.347486074098847e-07,
      "loss": 3.0846954345703126,
      "step": 9850
    },
    {
      "epoch": 0.986,
      "grad_norm": 0.5434274673461914,
      "learning_rate": 1.1813312225689155e-07,
      "loss": 3.1396461486816407,
      "step": 9860
    },
    {
      "epoch": 0.987,
      "grad_norm": 0.5132783651351929,
      "learning_rate": 1.0260992919757239e-07,
      "loss": 3.2976734161376955,
      "step": 9870
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.5402784943580627,
      "learning_rate": 8.817919799108953e-08,
      "loss": 3.150349807739258,
      "step": 9880
    },
    {
      "epoch": 0.989,
      "grad_norm": 0.6735221743583679,
      "learning_rate": 7.48410864496174e-08,
      "loss": 3.1719587326049803,
      "step": 9890
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5235326886177063,
      "learning_rate": 6.259574043659954e-08,
      "loss": 3.2585243225097655,
      "step": 9900
    },
    {
      "epoch": 0.991,
      "grad_norm": 0.5635644793510437,
      "learning_rate": 5.14432938651721e-08,
      "loss": 3.1897645950317384,
      "step": 9910
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.5258952975273132,
      "learning_rate": 4.138386869670941e-08,
      "loss": 3.2563705444335938,
      "step": 9920
    },
    {
      "epoch": 0.993,
      "grad_norm": 0.5228592753410339,
      "learning_rate": 3.241757493946951e-08,
      "loss": 3.1966835021972657,
      "step": 9930
    },
    {
      "epoch": 0.994,
      "grad_norm": 0.5348578095436096,
      "learning_rate": 2.4544510647417363e-08,
      "loss": 3.240966796875,
      "step": 9940
    },
    {
      "epoch": 0.995,
      "grad_norm": 0.5407728552818298,
      "learning_rate": 1.7764761919103477e-08,
      "loss": 3.1647939682006836,
      "step": 9950
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.5441321134567261,
      "learning_rate": 1.2078402896775753e-08,
      "loss": 3.205100250244141,
      "step": 9960
    },
    {
      "epoch": 0.997,
      "grad_norm": 0.5862337350845337,
      "learning_rate": 7.485495765569005e-09,
      "loss": 3.4486141204833984,
      "step": 9970
    },
    {
      "epoch": 0.998,
      "grad_norm": 0.5468406081199646,
      "learning_rate": 3.9860907527500265e-09,
      "loss": 3.1633674621582033,
      "step": 9980
    },
    {
      "epoch": 0.999,
      "grad_norm": 0.5793742537498474,
      "learning_rate": 1.5802261272734875e-09,
      "loss": 3.155015754699707,
      "step": 9990
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.515745222568512,
      "learning_rate": 2.6792819928234127e-10,
      "loss": 3.2327392578125,
      "step": 10000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.08804499161088e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}