{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010548523206751054,
      "grad_norm": 1.2691798210144043,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.522,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.190048336982727,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.9155,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8505806922912598,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.2513,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.8368041515350342,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.7561,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.820473849773407,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.2893,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 1.0409610271453857,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.7736,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.1511647701263428,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3722,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.1112607717514038,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.1169,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.9545691609382629,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.9197,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.5341774225234985,
      "learning_rate": 0.0015,
      "loss": 3.7698,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 1.0685570240020752,
      "learning_rate": 0.0015,
      "loss": 3.619,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.8480899333953857,
      "learning_rate": 0.0015,
      "loss": 3.5075,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.6460363864898682,
      "learning_rate": 0.0015,
      "loss": 3.3988,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.7429052591323853,
      "learning_rate": 0.0015,
      "loss": 3.3032,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.9308001399040222,
      "learning_rate": 0.0015,
      "loss": 3.2196,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.6899569630622864,
      "learning_rate": 0.0015,
      "loss": 3.1532,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.9509626626968384,
      "learning_rate": 0.0015,
      "loss": 3.0805,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.776450514793396,
      "learning_rate": 0.0015,
      "loss": 3.0375,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.7692133784294128,
      "learning_rate": 0.0015,
      "loss": 2.975,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.8211736679077148,
      "learning_rate": 0.0015,
      "loss": 2.9283,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.7810904383659363,
      "learning_rate": 0.0015,
      "loss": 2.889,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.8243034482002258,
      "learning_rate": 0.0015,
      "loss": 2.8562,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.768045961856842,
      "learning_rate": 0.0015,
      "loss": 2.8091,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.9104819297790527,
      "learning_rate": 0.0015,
      "loss": 2.7694,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 1.0683870315551758,
      "learning_rate": 0.0015,
      "loss": 2.7353,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.7473042607307434,
      "learning_rate": 0.0015,
      "loss": 2.7033,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.8680592775344849,
      "learning_rate": 0.0015,
      "loss": 2.6715,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.7758686542510986,
      "learning_rate": 0.0015,
      "loss": 2.6387,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 1.1664143800735474,
      "learning_rate": 0.0015,
      "loss": 2.613,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 1.0230486392974854,
      "learning_rate": 0.0015,
      "loss": 2.5867,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.7810431122779846,
      "learning_rate": 0.0015,
      "loss": 2.5615,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 1.0520604848861694,
      "learning_rate": 0.0015,
      "loss": 2.539,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.9819400310516357,
      "learning_rate": 0.0015,
      "loss": 2.5296,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 1.0482884645462036,
      "learning_rate": 0.0015,
      "loss": 2.5035,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.8537718653678894,
      "learning_rate": 0.0015,
      "loss": 2.4623,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.8536440134048462,
      "learning_rate": 0.0015,
      "loss": 2.4484,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.8123664259910583,
      "learning_rate": 0.0015,
      "loss": 2.4335,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.8474407196044922,
      "learning_rate": 0.0015,
      "loss": 2.4246,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.7529342174530029,
      "learning_rate": 0.0015,
      "loss": 2.3984,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.986809253692627,
      "learning_rate": 0.0015,
      "loss": 2.3729,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.8170089721679688,
      "learning_rate": 0.0015,
      "loss": 2.3652,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.7728642821311951,
      "learning_rate": 0.0015,
      "loss": 2.3442,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.9778690934181213,
      "learning_rate": 0.0015,
      "loss": 2.3212,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.7992797493934631,
      "learning_rate": 0.0015,
      "loss": 2.3086,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.9398708343505859,
      "learning_rate": 0.0015,
      "loss": 2.3091,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.8730383515357971,
      "learning_rate": 0.0015,
      "loss": 2.2856,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.8130483627319336,
      "learning_rate": 0.0015,
      "loss": 2.267,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.8602993488311768,
      "learning_rate": 0.0015,
      "loss": 2.2583,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 1.0511897802352905,
      "learning_rate": 0.0015,
      "loss": 2.2402,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.9979124069213867,
      "learning_rate": 0.0015,
      "loss": 2.2346,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.9300376772880554,
      "learning_rate": 0.0015,
      "loss": 2.2198,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.8899880647659302,
      "learning_rate": 0.0015,
      "loss": 2.206,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.9637308716773987,
      "learning_rate": 0.0015,
      "loss": 2.1915,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.8068028688430786,
      "learning_rate": 0.0015,
      "loss": 2.1731,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.8182671666145325,
      "learning_rate": 0.0015,
      "loss": 2.1743,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.704035758972168,
      "learning_rate": 0.0015,
      "loss": 2.1413,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.8451551795005798,
      "learning_rate": 0.0015,
      "loss": 2.1429,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 0.9787982106208801,
      "learning_rate": 0.0015,
      "loss": 2.1469,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.784288763999939,
      "learning_rate": 0.0015,
      "loss": 2.1196,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.7017039656639099,
      "learning_rate": 0.0015,
      "loss": 2.0997,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.7353077530860901,
      "learning_rate": 0.0015,
      "loss": 2.1061,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.7414589524269104,
      "learning_rate": 0.0015,
      "loss": 2.0927,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.8267672061920166,
      "learning_rate": 0.0015,
      "loss": 2.0754,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.7181387543678284,
      "learning_rate": 0.0015,
      "loss": 2.0883,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 1.069088101387024,
      "learning_rate": 0.0015,
      "loss": 2.0758,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.9701215624809265,
      "learning_rate": 0.0015,
      "loss": 2.0672,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.7990618348121643,
      "learning_rate": 0.0015,
      "loss": 2.0503,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.8165965676307678,
      "learning_rate": 0.0015,
      "loss": 2.0366,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7588340044021606,
      "learning_rate": 0.0015,
      "loss": 2.0473,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.7811934351921082,
      "learning_rate": 0.0015,
      "loss": 2.0292,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.8503795266151428,
      "learning_rate": 0.0015,
      "loss": 2.0208,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 1.1329801082611084,
      "learning_rate": 0.0015,
      "loss": 1.9997,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.95328688621521,
      "learning_rate": 0.0015,
      "loss": 1.9936,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.8979207277297974,
      "learning_rate": 0.0015,
      "loss": 2.0033,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.8836649656295776,
      "learning_rate": 0.0015,
      "loss": 1.988,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.0178972482681274,
      "learning_rate": 0.0015,
      "loss": 1.9851,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.8203606009483337,
      "learning_rate": 0.0015,
      "loss": 1.9739,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.776364803314209,
      "learning_rate": 0.0015,
      "loss": 1.9693,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.7987720966339111,
      "learning_rate": 0.0015,
      "loss": 1.9736,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.7319262027740479,
      "learning_rate": 0.0015,
      "loss": 1.9482,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.8773535490036011,
      "learning_rate": 0.0015,
      "loss": 1.9367,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.7392305731773376,
      "learning_rate": 0.0015,
      "loss": 1.9461,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 1.023331642150879,
      "learning_rate": 0.0015,
      "loss": 1.9494,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.9687519669532776,
      "learning_rate": 0.0015,
      "loss": 1.9237,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.9138364195823669,
      "learning_rate": 0.0015,
      "loss": 1.9254,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.7576828598976135,
      "learning_rate": 0.0015,
      "loss": 1.9254,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.7307135462760925,
      "learning_rate": 0.0015,
      "loss": 1.9204,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.738254189491272,
      "learning_rate": 0.0015,
      "loss": 1.9153,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.9347200989723206,
      "learning_rate": 0.0015,
      "loss": 1.9064,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 1.0709482431411743,
      "learning_rate": 0.0015,
      "loss": 1.9023,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.7532578706741333,
      "learning_rate": 0.0015,
      "loss": 1.9069,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.7279590964317322,
      "learning_rate": 0.0015,
      "loss": 1.8944,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.7947350144386292,
      "learning_rate": 0.0015,
      "loss": 1.8804,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.9116294384002686,
      "learning_rate": 0.0015,
      "loss": 1.8878,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.9219313263893127,
      "learning_rate": 0.0015,
      "loss": 1.8726,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.7465686798095703,
      "learning_rate": 0.0015,
      "loss": 1.8755,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 0.8785697817802429,
      "learning_rate": 0.0015,
      "loss": 1.8777,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.7366559505462646,
      "learning_rate": 0.0015,
      "loss": 1.859,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.7610734701156616,
      "learning_rate": 0.0015,
      "loss": 1.8594,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 1.1366569995880127,
      "learning_rate": 0.0015,
      "loss": 1.8687,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.7980130910873413,
      "learning_rate": 0.0015,
      "loss": 1.8575,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.887115478515625,
      "learning_rate": 0.0015,
      "loss": 1.8452,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.8321040868759155,
      "learning_rate": 0.0015,
      "loss": 1.8486,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.750380277633667,
      "learning_rate": 0.0015,
      "loss": 1.838,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.9794703125953674,
      "learning_rate": 0.0015,
      "loss": 1.8429,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.8354634642601013,
      "learning_rate": 0.0015,
      "loss": 1.8325,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.6737495064735413,
      "learning_rate": 0.0015,
      "loss": 1.8227,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.9093047380447388,
      "learning_rate": 0.0015,
      "loss": 1.8232,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 1.0107520818710327,
      "learning_rate": 0.0015,
      "loss": 1.829,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.848486602306366,
      "learning_rate": 0.0015,
      "loss": 1.8167,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.7347695231437683,
      "learning_rate": 0.0015,
      "loss": 1.8224,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.9195750951766968,
      "learning_rate": 0.0015,
      "loss": 1.8149,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.7962725758552551,
      "learning_rate": 0.0015,
      "loss": 1.8007,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.8207571506500244,
      "learning_rate": 0.0015,
      "loss": 1.8139,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.2019542455673218,
      "learning_rate": 0.0015,
      "loss": 1.7987,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.7293469309806824,
      "learning_rate": 0.0015,
      "loss": 1.8107,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.9432297348976135,
      "learning_rate": 0.0015,
      "loss": 1.7945,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.7298336625099182,
      "learning_rate": 0.0015,
      "loss": 1.7808,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.6769946217536926,
      "learning_rate": 0.0015,
      "loss": 1.7828,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.7276390194892883,
      "learning_rate": 0.0015,
      "loss": 1.7998,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.6962103247642517,
      "learning_rate": 0.0015,
      "loss": 1.789,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.6777324080467224,
      "learning_rate": 0.0015,
      "loss": 1.7748,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.6568731665611267,
      "learning_rate": 0.0015,
      "loss": 1.7628,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.7241616249084473,
      "learning_rate": 0.0015,
      "loss": 1.7801,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.7091214656829834,
      "learning_rate": 0.0015,
      "loss": 1.7821,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.9228652715682983,
      "learning_rate": 0.0015,
      "loss": 1.7668,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.7739994525909424,
      "learning_rate": 0.0015,
      "loss": 1.7643,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.7284632325172424,
      "learning_rate": 0.0015,
      "loss": 1.7771,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.9050618410110474,
      "learning_rate": 0.0015,
      "loss": 1.7646,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 1.594030499458313,
      "learning_rate": 0.0015,
      "loss": 1.7542,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 1.271827220916748,
      "learning_rate": 0.0015,
      "loss": 1.7756,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 1.489267110824585,
      "learning_rate": 0.0015,
      "loss": 1.7589,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 0.9819322824478149,
      "learning_rate": 0.0015,
      "loss": 1.7434,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.6417033076286316,
      "learning_rate": 0.0015,
      "loss": 1.7429,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.7139493823051453,
      "learning_rate": 0.0015,
      "loss": 1.7507,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.7161443829536438,
      "learning_rate": 0.0015,
      "loss": 1.7446,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.7190232276916504,
      "learning_rate": 0.0015,
      "loss": 1.7409,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.6767439842224121,
      "learning_rate": 0.0015,
      "loss": 1.7308,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.8346786499023438,
      "learning_rate": 0.0015,
      "loss": 1.7453,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.6780781149864197,
      "learning_rate": 0.0015,
      "loss": 1.7296,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.6594715118408203,
      "learning_rate": 0.0015,
      "loss": 1.7307,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.740116536617279,
      "learning_rate": 0.0015,
      "loss": 1.737,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.7160284519195557,
      "learning_rate": 0.0015,
      "loss": 1.7317,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.7835481762886047,
      "learning_rate": 0.0015,
      "loss": 1.7289,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.7016604542732239,
      "learning_rate": 0.0015,
      "loss": 1.723,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.707410991191864,
      "learning_rate": 0.0015,
      "loss": 1.721,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 1.1425987482070923,
      "learning_rate": 0.0015,
      "loss": 1.72,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.624231219291687,
      "learning_rate": 0.0015,
      "loss": 1.7235,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.8226703405380249,
      "learning_rate": 0.0015,
      "loss": 1.7204,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.6356648802757263,
      "learning_rate": 0.0015,
      "loss": 1.7213,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.7041848301887512,
      "learning_rate": 0.0015,
      "loss": 1.7161,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.6434488892555237,
      "learning_rate": 0.0015,
      "loss": 1.7198,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.6322984099388123,
      "learning_rate": 0.0015,
      "loss": 1.7017,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6440200805664062,
      "learning_rate": 0.0015,
      "loss": 1.6916,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.7490743398666382,
      "learning_rate": 0.0015,
      "loss": 1.6954,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.7250597476959229,
      "learning_rate": 0.0015,
      "loss": 1.7001,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.640522301197052,
      "learning_rate": 0.0015,
      "loss": 1.6894,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.719833493232727,
      "learning_rate": 0.0015,
      "loss": 1.7001,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 1.0009751319885254,
      "learning_rate": 0.0015,
      "loss": 1.6927,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.7644470930099487,
      "learning_rate": 0.0015,
      "loss": 1.701,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.7390194535255432,
      "learning_rate": 0.0015,
      "loss": 1.6982,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.639541506767273,
      "learning_rate": 0.0015,
      "loss": 1.6695,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.7821475267410278,
      "learning_rate": 0.0015,
      "loss": 1.6799,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.7226126790046692,
      "learning_rate": 0.0015,
      "loss": 1.6847,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.93052738904953,
      "learning_rate": 0.0015,
      "loss": 1.6786,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.7509690523147583,
      "learning_rate": 0.0015,
      "loss": 1.6768,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.7351312041282654,
      "learning_rate": 0.0015,
      "loss": 1.6843,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.383955717086792,
      "learning_rate": 0.0015,
      "loss": 1.673,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.7214679718017578,
      "learning_rate": 0.0015,
      "loss": 1.6649,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.6363106369972229,
      "learning_rate": 0.0015,
      "loss": 1.6816,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.7506694793701172,
      "learning_rate": 0.0015,
      "loss": 1.6629,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.6698363423347473,
      "learning_rate": 0.0015,
      "loss": 1.6614,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.8435531854629517,
      "learning_rate": 0.0015,
      "loss": 1.6663,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.7420846819877625,
      "learning_rate": 0.0015,
      "loss": 1.6672,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.6367708444595337,
      "learning_rate": 0.0015,
      "loss": 1.6847,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.7232658267021179,
      "learning_rate": 0.0015,
      "loss": 1.6643,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.8593874573707581,
      "learning_rate": 0.0015,
      "loss": 1.6674,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.6410152912139893,
      "learning_rate": 0.0015,
      "loss": 1.6697,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 1.2279235124588013,
      "learning_rate": 0.0015,
      "loss": 1.6695,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 1.0295931100845337,
      "learning_rate": 0.0015,
      "loss": 1.6695,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.6923174262046814,
      "learning_rate": 0.0015,
      "loss": 1.6462,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.7438858151435852,
      "learning_rate": 0.0015,
      "loss": 1.6477,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.7807824611663818,
      "learning_rate": 0.0015,
      "loss": 1.6483,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.7299506068229675,
      "learning_rate": 0.0015,
      "loss": 1.6473,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.857148289680481,
      "learning_rate": 0.0015,
      "loss": 1.6516,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.7330800294876099,
      "learning_rate": 0.0015,
      "loss": 1.6532,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.6870981454849243,
      "learning_rate": 0.0015,
      "loss": 1.6546,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.7019320130348206,
      "learning_rate": 0.0015,
      "loss": 1.6396,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.696793794631958,
      "learning_rate": 0.0015,
      "loss": 1.6491,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.9838576912879944,
      "learning_rate": 0.0015,
      "loss": 1.6449,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 1.0997356176376343,
      "learning_rate": 0.0015,
      "loss": 1.6475,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.6624078750610352,
      "learning_rate": 0.0015,
      "loss": 1.6404,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 1.037745714187622,
      "learning_rate": 0.0015,
      "loss": 1.6412,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.6833543181419373,
      "learning_rate": 0.0015,
      "loss": 1.6325,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.6265554428100586,
      "learning_rate": 0.0015,
      "loss": 1.6297,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.7579610347747803,
      "learning_rate": 0.0015,
      "loss": 1.6345,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.5994946956634521,
      "learning_rate": 0.0015,
      "loss": 1.6464,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.8627594113349915,
      "learning_rate": 0.0015,
      "loss": 1.6257,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.6660335659980774,
      "learning_rate": 0.0015,
      "loss": 1.6272,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.6796160936355591,
      "learning_rate": 0.0015,
      "loss": 1.624,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.9388406872749329,
      "learning_rate": 0.0015,
      "loss": 1.6321,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 1.0387760400772095,
      "learning_rate": 0.0015,
      "loss": 1.6426,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.6553621888160706,
      "learning_rate": 0.0015,
      "loss": 1.627,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.6387179493904114,
      "learning_rate": 0.0015,
      "loss": 1.6276,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.7289672493934631,
      "learning_rate": 0.0015,
      "loss": 1.6261,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.6390879154205322,
      "learning_rate": 0.0015,
      "loss": 1.6166,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.884262204170227,
      "learning_rate": 0.0015,
      "loss": 1.6222,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.7829119563102722,
      "learning_rate": 0.0015,
      "loss": 1.6234,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.6269692182540894,
      "learning_rate": 0.0015,
      "loss": 1.613,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.6027466654777527,
      "learning_rate": 0.0015,
      "loss": 1.6198,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 1.0918283462524414,
      "learning_rate": 0.0015,
      "loss": 1.6171,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.6074740290641785,
      "learning_rate": 0.0015,
      "loss": 1.6093,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.7111483812332153,
      "learning_rate": 0.0015,
      "loss": 1.6141,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6221132278442383,
      "learning_rate": 0.0015,
      "loss": 1.6089,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.9400107264518738,
      "learning_rate": 0.0015,
      "loss": 1.6158,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.103214144706726,
      "learning_rate": 0.0015,
      "loss": 1.6291,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.7383182644844055,
      "learning_rate": 0.0015,
      "loss": 1.6138,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.713843584060669,
      "learning_rate": 0.0015,
      "loss": 1.5957,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 1.0974149703979492,
      "learning_rate": 0.0015,
      "loss": 1.6112,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.723660945892334,
      "learning_rate": 0.0015,
      "loss": 1.6146,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.782029390335083,
      "learning_rate": 0.0015,
      "loss": 1.5973,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.6948687434196472,
      "learning_rate": 0.0015,
      "loss": 1.6127,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 1.1345460414886475,
      "learning_rate": 0.0015,
      "loss": 1.6054,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 1.0687614679336548,
      "learning_rate": 0.0015,
      "loss": 1.6031,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.6718165278434753,
      "learning_rate": 0.0015,
      "loss": 1.5958,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.6679635047912598,
      "learning_rate": 0.0015,
      "loss": 1.5853,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.8055697679519653,
      "learning_rate": 0.0015,
      "loss": 1.62,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.9656091332435608,
      "learning_rate": 0.0015,
      "loss": 1.6072,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.9608715772628784,
      "learning_rate": 0.0015,
      "loss": 1.5968,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.8238662481307983,
      "learning_rate": 0.0015,
      "loss": 1.5948,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.7834868431091309,
      "learning_rate": 0.0015,
      "loss": 1.5952,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.6714757680892944,
      "learning_rate": 0.0015,
      "loss": 1.6039,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.6760356426239014,
      "learning_rate": 0.0015,
      "loss": 1.5854,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.7482396960258484,
      "learning_rate": 0.0015,
      "loss": 1.5804,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.8608502745628357,
      "learning_rate": 0.0015,
      "loss": 1.5825,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.6435486674308777,
      "learning_rate": 0.0015,
      "loss": 1.5942,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5963234901428223,
      "learning_rate": 0.0015,
      "loss": 1.5931,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.6503880023956299,
      "learning_rate": 0.0015,
      "loss": 1.5901,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.6955623030662537,
      "learning_rate": 0.0015,
      "loss": 1.5884,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.6004531383514404,
      "learning_rate": 0.0015,
      "loss": 1.5883,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.607153594493866,
      "learning_rate": 0.0015,
      "loss": 1.5892,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.9378383755683899,
      "learning_rate": 0.0015,
      "loss": 1.5851,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.7985408902168274,
      "learning_rate": 0.0015,
      "loss": 1.583,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.8261814117431641,
      "learning_rate": 0.0015,
      "loss": 1.5893,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.6244887709617615,
      "learning_rate": 0.0015,
      "loss": 1.575,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.6061866283416748,
      "learning_rate": 0.0015,
      "loss": 1.5909,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.9484696388244629,
      "learning_rate": 0.0015,
      "loss": 1.5829,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.7839576601982117,
      "learning_rate": 0.0015,
      "loss": 1.5884,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.6155287623405457,
      "learning_rate": 0.0015,
      "loss": 1.573,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.7642000317573547,
      "learning_rate": 0.0015,
      "loss": 1.5784,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.6023858785629272,
      "learning_rate": 0.0015,
      "loss": 1.5767,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.725964367389679,
      "learning_rate": 0.0015,
      "loss": 1.5816,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.7024165987968445,
      "learning_rate": 0.0015,
      "loss": 1.5737,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.7889173626899719,
      "learning_rate": 0.0015,
      "loss": 1.5793,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.6713038086891174,
      "learning_rate": 0.0015,
      "loss": 1.5748,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.6811700463294983,
      "learning_rate": 0.0015,
      "loss": 1.5819,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.6109973192214966,
      "learning_rate": 0.0015,
      "loss": 1.5745,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.5974403619766235,
      "learning_rate": 0.0015,
      "loss": 1.5716,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.7356406450271606,
      "learning_rate": 0.0015,
      "loss": 1.582,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.6180926561355591,
      "learning_rate": 0.0015,
      "loss": 1.5783,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.5976831316947937,
      "learning_rate": 0.0015,
      "loss": 1.5754,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6535393595695496,
      "learning_rate": 0.0015,
      "loss": 1.5743,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.7285734415054321,
      "learning_rate": 0.0015,
      "loss": 1.5647,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.6646901965141296,
      "learning_rate": 0.0015,
      "loss": 1.5788,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.7660537958145142,
      "learning_rate": 0.0015,
      "loss": 1.5706,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.8967321515083313,
      "learning_rate": 0.0015,
      "loss": 1.5631,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.6177824139595032,
      "learning_rate": 0.0015,
      "loss": 1.5653,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.6162772178649902,
      "learning_rate": 0.0015,
      "loss": 1.5656,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.7549130320549011,
      "learning_rate": 0.0015,
      "loss": 1.5648,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6640269756317139,
      "learning_rate": 0.0015,
      "loss": 1.5581,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.7048502564430237,
      "learning_rate": 0.0015,
      "loss": 1.5683,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.615058958530426,
      "learning_rate": 0.0015,
      "loss": 1.5679,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.5851198434829712,
      "learning_rate": 0.0015,
      "loss": 1.546,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 1.017960548400879,
      "learning_rate": 0.0015,
      "loss": 1.5574,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.6042937636375427,
      "learning_rate": 0.0015,
      "loss": 1.5665,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.926185667514801,
      "learning_rate": 0.0015,
      "loss": 1.5707,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6197679042816162,
      "learning_rate": 0.0015,
      "loss": 1.5672,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.6802936792373657,
      "learning_rate": 0.0015,
      "loss": 1.564,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.7331050038337708,
      "learning_rate": 0.0015,
      "loss": 1.5602,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.721390962600708,
      "learning_rate": 0.0015,
      "loss": 1.5589,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.6204345226287842,
      "learning_rate": 0.0015,
      "loss": 1.5564,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6381465196609497,
      "learning_rate": 0.0015,
      "loss": 1.5651,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 1.0429388284683228,
      "learning_rate": 0.0015,
      "loss": 1.5408,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.8542698621749878,
      "learning_rate": 0.0015,
      "loss": 1.5455,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.6300585865974426,
      "learning_rate": 0.0015,
      "loss": 1.5481,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.6176901459693909,
      "learning_rate": 0.0015,
      "loss": 1.5525,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.9087796807289124,
      "learning_rate": 0.0015,
      "loss": 1.5546,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.7421715259552002,
      "learning_rate": 0.0015,
      "loss": 1.5441,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.6909385919570923,
      "learning_rate": 0.0015,
      "loss": 1.547,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.7861342430114746,
      "learning_rate": 0.0015,
      "loss": 1.5522,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.9386938214302063,
      "learning_rate": 0.0015,
      "loss": 1.5546,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.7081777453422546,
      "learning_rate": 0.0015,
      "loss": 1.5461,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.6430122256278992,
      "learning_rate": 0.0015,
      "loss": 1.5435,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.8484134078025818,
      "learning_rate": 0.0015,
      "loss": 1.5533,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.8623349666595459,
      "learning_rate": 0.0015,
      "loss": 1.5589,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.9112716317176819,
      "learning_rate": 0.0015,
      "loss": 1.55,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.8495725989341736,
      "learning_rate": 0.0015,
      "loss": 1.5305,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.5744346976280212,
      "learning_rate": 0.0015,
      "loss": 1.5354,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.6701624393463135,
      "learning_rate": 0.0015,
      "loss": 1.55,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.6425880789756775,
      "learning_rate": 0.0015,
      "loss": 1.533,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.6229080557823181,
      "learning_rate": 0.0015,
      "loss": 1.5425,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.755316972732544,
      "learning_rate": 0.0015,
      "loss": 1.5447,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6249800324440002,
      "learning_rate": 0.0015,
      "loss": 1.5311,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.7024720311164856,
      "learning_rate": 0.0015,
      "loss": 1.5398,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.6813156604766846,
      "learning_rate": 0.0015,
      "loss": 1.5375,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.6011126637458801,
      "learning_rate": 0.0015,
      "loss": 1.5471,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.5836905837059021,
      "learning_rate": 0.0015,
      "loss": 1.536,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.7479152679443359,
      "learning_rate": 0.0015,
      "loss": 1.5288,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.5793021321296692,
      "learning_rate": 0.0015,
      "loss": 1.5433,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.6929194331169128,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6408833265304565,
      "learning_rate": 0.0015,
      "loss": 1.5349,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.721777617931366,
      "learning_rate": 0.0015,
      "loss": 1.5372,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.6097099781036377,
      "learning_rate": 0.0015,
      "loss": 1.5493,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.7727136611938477,
      "learning_rate": 0.0015,
      "loss": 1.5257,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5993808507919312,
      "learning_rate": 0.0015,
      "loss": 1.5368,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.6082150340080261,
      "learning_rate": 0.0015,
      "loss": 1.5364,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.647101640701294,
      "learning_rate": 0.0015,
      "loss": 1.5358,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.6560900211334229,
      "learning_rate": 0.0015,
      "loss": 1.532,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.715570867061615,
      "learning_rate": 0.0015,
      "loss": 1.5334,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.7903317809104919,
      "learning_rate": 0.0015,
      "loss": 1.5259,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.6264176964759827,
      "learning_rate": 0.0015,
      "loss": 1.5385,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.6412289142608643,
      "learning_rate": 0.0015,
      "loss": 1.5304,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.8620365262031555,
      "learning_rate": 0.0015,
      "loss": 1.525,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.6095845103263855,
      "learning_rate": 0.0015,
      "loss": 1.5059,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.6399675607681274,
      "learning_rate": 0.0015,
      "loss": 1.5331,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.5802530646324158,
      "learning_rate": 0.0015,
      "loss": 1.5188,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.591494083404541,
      "learning_rate": 0.0015,
      "loss": 1.5383,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.665297269821167,
      "learning_rate": 0.0015,
      "loss": 1.5373,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 1.0794377326965332,
      "learning_rate": 0.0015,
      "loss": 1.5198,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.6363126039505005,
      "learning_rate": 0.0015,
      "loss": 1.5279,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.612177312374115,
      "learning_rate": 0.0015,
      "loss": 1.5271,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.748126745223999,
      "learning_rate": 0.0015,
      "loss": 1.5367,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.5856114029884338,
      "learning_rate": 0.0015,
      "loss": 1.522,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.6275367140769958,
      "learning_rate": 0.0015,
      "loss": 1.5244,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6555677056312561,
      "learning_rate": 0.0015,
      "loss": 1.5165,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.6494587063789368,
      "learning_rate": 0.0015,
      "loss": 1.5181,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.806426465511322,
      "learning_rate": 0.0015,
      "loss": 1.5264,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.6121223568916321,
      "learning_rate": 0.0015,
      "loss": 1.5209,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.7112828493118286,
      "learning_rate": 0.0015,
      "loss": 1.5158,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.5737324953079224,
      "learning_rate": 0.0015,
      "loss": 1.5272,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.6188103556632996,
      "learning_rate": 0.0015,
      "loss": 1.5238,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.7734886407852173,
      "learning_rate": 0.0015,
      "loss": 1.5043,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.6218626499176025,
      "learning_rate": 0.0015,
      "loss": 1.5183,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.6316360235214233,
      "learning_rate": 0.0015,
      "loss": 1.5152,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.5388330221176147,
      "learning_rate": 0.0015,
      "loss": 1.5151,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.6421667337417603,
      "learning_rate": 0.0015,
      "loss": 1.5065,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.8086821436882019,
      "learning_rate": 0.0015,
      "loss": 1.5244,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.5916324853897095,
      "learning_rate": 0.0015,
      "loss": 1.5269,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.5721726417541504,
      "learning_rate": 0.0015,
      "loss": 1.5208,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.6366936564445496,
      "learning_rate": 0.0015,
      "loss": 1.5009,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.7353329062461853,
      "learning_rate": 0.0015,
      "loss": 1.5208,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.6907286643981934,
      "learning_rate": 0.0015,
      "loss": 1.5211,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.6311991214752197,
      "learning_rate": 0.0015,
      "loss": 1.5299,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.7826913595199585,
      "learning_rate": 0.0015,
      "loss": 1.5213,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.7268832325935364,
      "learning_rate": 0.0015,
      "loss": 1.5108,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.6679678559303284,
      "learning_rate": 0.0015,
      "loss": 1.5111,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.9584214091300964,
      "learning_rate": 0.0015,
      "loss": 1.5064,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.5330318212509155,
      "learning_rate": 0.0015,
      "loss": 1.5075,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.6112387180328369,
      "learning_rate": 0.0015,
      "loss": 1.5196,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.55507493019104,
      "learning_rate": 0.0015,
      "loss": 1.5052,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.576727569103241,
      "learning_rate": 0.0015,
      "loss": 1.5055,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.7007277607917786,
      "learning_rate": 0.0015,
      "loss": 1.507,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.6003903746604919,
      "learning_rate": 0.0015,
      "loss": 1.5091,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.7698028087615967,
      "learning_rate": 0.0015,
      "loss": 1.5063,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.6928577423095703,
      "learning_rate": 0.0015,
      "loss": 1.5187,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.8100811243057251,
      "learning_rate": 0.0015,
      "loss": 1.5111,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.7094262838363647,
      "learning_rate": 0.0015,
      "loss": 1.5058,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.681793212890625,
      "learning_rate": 0.0015,
      "loss": 1.5039,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.6533880233764648,
      "learning_rate": 0.0015,
      "loss": 1.5156,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.6184443831443787,
      "learning_rate": 0.0015,
      "loss": 1.5196,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.6186136603355408,
      "learning_rate": 0.0015,
      "loss": 1.5064,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.6013845205307007,
      "learning_rate": 0.0015,
      "loss": 1.4854,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.7819629311561584,
      "learning_rate": 0.0015,
      "loss": 1.5063,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.5571489930152893,
      "learning_rate": 0.0015,
      "loss": 1.5021,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.7706273794174194,
      "learning_rate": 0.0015,
      "loss": 1.5108,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 1.0314881801605225,
      "learning_rate": 0.0015,
      "loss": 1.4991,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.6681764125823975,
      "learning_rate": 0.0015,
      "loss": 1.5114,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.5619609951972961,
      "learning_rate": 0.0015,
      "loss": 1.5023,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.7922562956809998,
      "learning_rate": 0.0015,
      "loss": 1.5028,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.5868403315544128,
      "learning_rate": 0.0015,
      "loss": 1.5112,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.5683006644248962,
      "learning_rate": 0.0015,
      "loss": 1.4971,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.629831850528717,
      "learning_rate": 0.0015,
      "loss": 1.4916,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.5975214242935181,
      "learning_rate": 0.0015,
      "loss": 1.4943,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.6330074071884155,
      "learning_rate": 0.0015,
      "loss": 1.5121,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.6558031439781189,
      "learning_rate": 0.0015,
      "loss": 1.4951,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.5945179462432861,
      "learning_rate": 0.0015,
      "loss": 1.5067,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.5857622623443604,
      "learning_rate": 0.0015,
      "loss": 1.4945,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.6149414777755737,
      "learning_rate": 0.0015,
      "loss": 1.4876,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.631554901599884,
      "learning_rate": 0.0015,
      "loss": 1.486,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.9400520920753479,
      "learning_rate": 0.0015,
      "loss": 1.4882,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.7286548614501953,
      "learning_rate": 0.0015,
      "loss": 1.4935,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.5828094482421875,
      "learning_rate": 0.0015,
      "loss": 1.4896,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.6287121772766113,
      "learning_rate": 0.0015,
      "loss": 1.4948,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.7357299327850342,
      "learning_rate": 0.0015,
      "loss": 1.4903,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.5453791618347168,
      "learning_rate": 0.0015,
      "loss": 1.4958,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.5658535361289978,
      "learning_rate": 0.0015,
      "loss": 1.4963,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.5815938115119934,
      "learning_rate": 0.0015,
      "loss": 1.4849,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.5606580376625061,
      "learning_rate": 0.0015,
      "loss": 1.496,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.585895836353302,
      "learning_rate": 0.0015,
      "loss": 1.4854,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.5472810864448547,
      "learning_rate": 0.0015,
      "loss": 1.503,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.6179347038269043,
      "learning_rate": 0.0015,
      "loss": 1.4866,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.6303660273551941,
      "learning_rate": 0.0015,
      "loss": 1.491,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.707388162612915,
      "learning_rate": 0.0015,
      "loss": 1.4812,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.6531679630279541,
      "learning_rate": 0.0015,
      "loss": 1.4921,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5899868011474609,
      "learning_rate": 0.0015,
      "loss": 1.4932,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.8273098468780518,
      "learning_rate": 0.0015,
      "loss": 1.4965,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.7142001986503601,
      "learning_rate": 0.0015,
      "loss": 1.4916,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6741710305213928,
      "learning_rate": 0.0015,
      "loss": 1.501,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.8062102198600769,
      "learning_rate": 0.0015,
      "loss": 1.4922,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.7213407158851624,
      "learning_rate": 0.0015,
      "loss": 1.489,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.6004040241241455,
      "learning_rate": 0.0015,
      "loss": 1.4825,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.921000063419342,
      "learning_rate": 0.0015,
      "loss": 1.4895,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.759276270866394,
      "learning_rate": 0.0015,
      "loss": 1.4822,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5667396187782288,
      "learning_rate": 0.0015,
      "loss": 1.4836,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.5702187418937683,
      "learning_rate": 0.0015,
      "loss": 1.4897,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.5534017086029053,
      "learning_rate": 0.0015,
      "loss": 1.4823,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.6381139755249023,
      "learning_rate": 0.0015,
      "loss": 1.4792,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.6257467865943909,
      "learning_rate": 0.0015,
      "loss": 1.4813,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.6904028058052063,
      "learning_rate": 0.0015,
      "loss": 1.4856,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.7427784204483032,
      "learning_rate": 0.0015,
      "loss": 1.4891,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.5929211378097534,
      "learning_rate": 0.0015,
      "loss": 1.4776,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.6614937782287598,
      "learning_rate": 0.0015,
      "loss": 1.4844,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.8259798884391785,
      "learning_rate": 0.0015,
      "loss": 1.4816,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.6505470871925354,
      "learning_rate": 0.0015,
      "loss": 1.4803,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.5709390640258789,
      "learning_rate": 0.0015,
      "loss": 1.484,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.651495635509491,
      "learning_rate": 0.0015,
      "loss": 1.4919,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.6256493926048279,
      "learning_rate": 0.0015,
      "loss": 1.4789,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.7242328524589539,
      "learning_rate": 0.0015,
      "loss": 1.475,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.7946824431419373,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.7557520270347595,
      "learning_rate": 0.0015,
      "loss": 1.4837,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.6960965394973755,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.9885152578353882,
      "learning_rate": 0.0015,
      "loss": 1.476,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.6563693881034851,
      "learning_rate": 0.0015,
      "loss": 1.4868,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.8019644021987915,
      "learning_rate": 0.0015,
      "loss": 1.4758,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.6949400901794434,
      "learning_rate": 0.0015,
      "loss": 1.4792,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.6423824429512024,
      "learning_rate": 0.0015,
      "loss": 1.4921,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.6390447616577148,
      "learning_rate": 0.0015,
      "loss": 1.4679,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5630648732185364,
      "learning_rate": 0.0015,
      "loss": 1.4748,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.669380784034729,
      "learning_rate": 0.0015,
      "loss": 1.472,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.7069663405418396,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.6067928075790405,
      "learning_rate": 0.0015,
      "loss": 1.494,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.6364792585372925,
      "learning_rate": 0.0015,
      "loss": 1.4749,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.6240522861480713,
      "learning_rate": 0.0015,
      "loss": 1.4811,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.6970255374908447,
      "learning_rate": 0.0015,
      "loss": 1.4754,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.5970498919487,
      "learning_rate": 0.0015,
      "loss": 1.4823,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.567321240901947,
      "learning_rate": 0.0015,
      "loss": 1.4697,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.6129364371299744,
      "learning_rate": 0.0015,
      "loss": 1.4778,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.6318622827529907,
      "learning_rate": 0.0015,
      "loss": 1.4832,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.6397720575332642,
      "learning_rate": 0.0015,
      "loss": 1.4717,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.7977009415626526,
      "learning_rate": 0.0015,
      "loss": 1.4705,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5972973108291626,
      "learning_rate": 0.0015,
      "loss": 1.4744,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.5378081202507019,
      "learning_rate": 0.0015,
      "loss": 1.4592,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.8785254955291748,
      "learning_rate": 0.0015,
      "loss": 1.4799,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.5821215510368347,
      "learning_rate": 0.0015,
      "loss": 1.4795,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.7573177814483643,
      "learning_rate": 0.0015,
      "loss": 1.4753,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.5414453148841858,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.5165130496025085,
      "learning_rate": 0.0015,
      "loss": 1.4481,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.7440067529678345,
      "learning_rate": 0.0015,
      "loss": 1.4733,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.7288051247596741,
      "learning_rate": 0.0015,
      "loss": 1.4609,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.6948131918907166,
      "learning_rate": 0.0015,
      "loss": 1.4802,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": Infinity,
      "learning_rate": 0.0015,
      "loss": 1.4841,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 1.2940268516540527,
      "learning_rate": 0.0015,
      "loss": 1.4655,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.6071725487709045,
      "learning_rate": 0.0015,
      "loss": 1.4767,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.5999435782432556,
      "learning_rate": 0.0015,
      "loss": 1.472,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5505561232566833,
      "learning_rate": 0.0015,
      "loss": 1.4758,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.6503564715385437,
      "learning_rate": 0.0015,
      "loss": 1.464,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.7435416579246521,
      "learning_rate": 0.0015,
      "loss": 1.465,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.5838521122932434,
      "learning_rate": 0.0015,
      "loss": 1.4636,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.5816810131072998,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.5804526209831238,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.6030544638633728,
      "learning_rate": 0.0015,
      "loss": 1.4703,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.6989827752113342,
      "learning_rate": 0.0015,
      "loss": 1.4654,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.5915131568908691,
      "learning_rate": 0.0015,
      "loss": 1.4624,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6471115946769714,
      "learning_rate": 0.0015,
      "loss": 1.4624,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.6063437461853027,
      "learning_rate": 0.0015,
      "loss": 1.4808,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.518700361251831,
      "learning_rate": 0.0015,
      "loss": 1.4661,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.543921947479248,
      "learning_rate": 0.0015,
      "loss": 1.4641,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5837245583534241,
      "learning_rate": 0.0015,
      "loss": 1.4732,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.760454535484314,
      "learning_rate": 0.0015,
      "loss": 1.4582,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.6135420799255371,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.571808397769928,
      "learning_rate": 0.0015,
      "loss": 1.4618,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.6032215356826782,
      "learning_rate": 0.0015,
      "loss": 1.4817,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.6733672618865967,
      "learning_rate": 0.0015,
      "loss": 1.4631,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.6539626121520996,
      "learning_rate": 0.0015,
      "loss": 1.458,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.5375906825065613,
      "learning_rate": 0.0015,
      "loss": 1.4696,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.6148267984390259,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.6515779495239258,
      "learning_rate": 0.0015,
      "loss": 1.4512,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.593586802482605,
      "learning_rate": 0.0015,
      "loss": 1.4719,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.5916845202445984,
      "learning_rate": 0.0015,
      "loss": 1.4612,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.5250903367996216,
      "learning_rate": 0.0015,
      "loss": 1.4677,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.6175042986869812,
      "learning_rate": 0.0015,
      "loss": 1.4622,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.6020768880844116,
      "learning_rate": 0.0015,
      "loss": 1.4574,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.6732794046401978,
      "learning_rate": 0.0015,
      "loss": 1.466,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.7224721908569336,
      "learning_rate": 0.0015,
      "loss": 1.4652,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.8541702032089233,
      "learning_rate": 0.0015,
      "loss": 1.4727,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.730940580368042,
      "learning_rate": 0.0015,
      "loss": 1.4516,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.5959996581077576,
      "learning_rate": 0.0015,
      "loss": 1.4595,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.5917717218399048,
      "learning_rate": 0.0015,
      "loss": 1.4548,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.5279277563095093,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.639276385307312,
      "learning_rate": 0.0015,
      "loss": 1.4688,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.6213117241859436,
      "learning_rate": 0.0015,
      "loss": 1.4593,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.5400778651237488,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.5232504606246948,
      "learning_rate": 0.0015,
      "loss": 1.4387,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.5442522764205933,
      "learning_rate": 0.0015,
      "loss": 1.4666,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.6611676812171936,
      "learning_rate": 0.0015,
      "loss": 1.4558,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.535550057888031,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.8105790019035339,
      "learning_rate": 0.0015,
      "loss": 1.4542,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.5619795322418213,
      "learning_rate": 0.0015,
      "loss": 1.4706,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.6469224095344543,
      "learning_rate": 0.0015,
      "loss": 1.4604,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.5533626675605774,
      "learning_rate": 0.0015,
      "loss": 1.4437,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.5283637642860413,
      "learning_rate": 0.0015,
      "loss": 1.4566,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6227394938468933,
      "learning_rate": 0.0015,
      "loss": 1.4553,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.5344183444976807,
      "learning_rate": 0.0015,
      "loss": 1.4506,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.588984489440918,
      "learning_rate": 0.0015,
      "loss": 1.4678,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.5949463844299316,
      "learning_rate": 0.0015,
      "loss": 1.464,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.557900071144104,
      "learning_rate": 0.0015,
      "loss": 1.4578,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5850265026092529,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.7179521322250366,
      "learning_rate": 0.0015,
      "loss": 1.4667,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.6477261781692505,
      "learning_rate": 0.0015,
      "loss": 1.4478,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.7852134108543396,
      "learning_rate": 0.0015,
      "loss": 1.4621,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.7401909232139587,
      "learning_rate": 0.0015,
      "loss": 1.4471,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.6288110613822937,
      "learning_rate": 0.0015,
      "loss": 1.4581,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.5337368249893188,
      "learning_rate": 0.0015,
      "loss": 1.4498,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.6970085501670837,
      "learning_rate": 0.0015,
      "loss": 1.4552,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.634314239025116,
      "learning_rate": 0.0015,
      "loss": 1.4472,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.5549997687339783,
      "learning_rate": 0.0015,
      "loss": 1.4535,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.5504727959632874,
      "learning_rate": 0.0015,
      "loss": 1.4627,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.8286926746368408,
      "learning_rate": 0.0015,
      "loss": 1.4535,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.6269571781158447,
      "learning_rate": 0.0015,
      "loss": 1.4547,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.8340073227882385,
      "learning_rate": 0.0015,
      "loss": 1.4577,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.9701064229011536,
      "learning_rate": 0.0015,
      "loss": 1.448,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.5321536660194397,
      "learning_rate": 0.0015,
      "loss": 1.4444,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.5817542672157288,
      "learning_rate": 0.0015,
      "loss": 1.4547,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.74298095703125,
      "learning_rate": 0.0015,
      "loss": 1.4479,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.5651299357414246,
      "learning_rate": 0.0015,
      "loss": 1.4435,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.723031222820282,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.6351364254951477,
      "learning_rate": 0.0015,
      "loss": 1.4532,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.5029274225234985,
      "learning_rate": 0.0015,
      "loss": 1.4441,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.5667977929115295,
      "learning_rate": 0.0015,
      "loss": 1.4432,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5874744057655334,
      "learning_rate": 0.0015,
      "loss": 1.4434,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5383337736129761,
      "learning_rate": 0.0015,
      "loss": 1.4468,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.6023250818252563,
      "learning_rate": 0.0015,
      "loss": 1.4395,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.6562930345535278,
      "learning_rate": 0.0015,
      "loss": 1.44,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.528099000453949,
      "learning_rate": 0.0015,
      "loss": 1.4375,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.6635940074920654,
      "learning_rate": 0.0015,
      "loss": 1.4468,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.5587949156761169,
      "learning_rate": 0.0015,
      "loss": 1.4493,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.6516025066375732,
      "learning_rate": 0.0015,
      "loss": 1.447,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.6126524806022644,
      "learning_rate": 0.0015,
      "loss": 1.4485,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.5466421842575073,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.5659489035606384,
      "learning_rate": 0.0015,
      "loss": 1.4445,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.6079151630401611,
      "learning_rate": 0.0015,
      "loss": 1.4508,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.5349658727645874,
      "learning_rate": 0.0015,
      "loss": 1.439,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.5877987742424011,
      "learning_rate": 0.0015,
      "loss": 1.4345,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.5495848655700684,
      "learning_rate": 0.0015,
      "loss": 1.4355,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.517038881778717,
      "learning_rate": 0.0015,
      "loss": 1.4517,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.5200017690658569,
      "learning_rate": 0.0015,
      "loss": 1.4566,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.5127320289611816,
      "learning_rate": 0.0015,
      "loss": 1.4427,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.6458799242973328,
      "learning_rate": 0.0015,
      "loss": 1.4405,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.5785557627677917,
      "learning_rate": 0.0015,
      "loss": 1.4433,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.6503692269325256,
      "learning_rate": 0.0015,
      "loss": 1.4393,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.5408360958099365,
      "learning_rate": 0.0015,
      "loss": 1.4452,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.6215591430664062,
      "learning_rate": 0.0015,
      "loss": 1.4518,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.7795698642730713,
      "learning_rate": 0.0015,
      "loss": 1.4454,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.5660273432731628,
      "learning_rate": 0.0015,
      "loss": 1.4559,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.7219056487083435,
      "learning_rate": 0.0015,
      "loss": 1.4496,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.741102397441864,
      "learning_rate": 0.0015,
      "loss": 1.4453,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.5923230051994324,
      "learning_rate": 0.0015,
      "loss": 1.451,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5904003977775574,
      "learning_rate": 0.0015,
      "loss": 1.4518,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.7221293449401855,
      "learning_rate": 0.0015,
      "loss": 1.433,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.6381110548973083,
      "learning_rate": 0.0015,
      "loss": 1.4378,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.5245643258094788,
      "learning_rate": 0.0015,
      "loss": 1.4406,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5776628851890564,
      "learning_rate": 0.0015,
      "loss": 1.4452,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.6253775954246521,
      "learning_rate": 0.0015,
      "loss": 1.4406,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.5543458461761475,
      "learning_rate": 0.0015,
      "loss": 1.4396,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.7039231657981873,
      "learning_rate": 0.0015,
      "loss": 1.4164,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5481839776039124,
      "learning_rate": 0.0015,
      "loss": 1.4243,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.6938220262527466,
      "learning_rate": 0.0015,
      "loss": 1.4292,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.5412605404853821,
      "learning_rate": 0.0015,
      "loss": 1.4366,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.7932698130607605,
      "learning_rate": 0.0015,
      "loss": 1.4333,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.539391815662384,
      "learning_rate": 0.0015,
      "loss": 1.4368,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.6700359582901001,
      "learning_rate": 0.0015,
      "loss": 1.4403,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5859553813934326,
      "learning_rate": 0.0015,
      "loss": 1.4561,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.6671823263168335,
      "learning_rate": 0.0015,
      "loss": 1.4381,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.7624210119247437,
      "learning_rate": 0.0015,
      "loss": 1.4509,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.7688072919845581,
      "learning_rate": 0.0015,
      "loss": 1.4541,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.5609461069107056,
      "learning_rate": 0.0015,
      "loss": 1.4402,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.5772584676742554,
      "learning_rate": 0.0015,
      "loss": 1.4337,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.6252260804176331,
      "learning_rate": 0.0015,
      "loss": 1.4432,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.5394101738929749,
      "learning_rate": 0.0015,
      "loss": 1.4349,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.5544285178184509,
      "learning_rate": 0.0015,
      "loss": 1.4356,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.6994200944900513,
      "learning_rate": 0.0015,
      "loss": 1.4331,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.580315887928009,
      "learning_rate": 0.0015,
      "loss": 1.4266,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.8660300374031067,
      "learning_rate": 0.0015,
      "loss": 1.4376,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.6988215446472168,
      "learning_rate": 0.0015,
      "loss": 1.4417,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.6541784405708313,
      "learning_rate": 0.0015,
      "loss": 1.4309,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.6802022457122803,
      "learning_rate": 0.0015,
      "loss": 1.4372,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5214741826057434,
      "learning_rate": 0.0015,
      "loss": 1.4372,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.5116277933120728,
      "learning_rate": 0.0015,
      "loss": 1.4204,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.5941348075866699,
      "learning_rate": 0.0015,
      "loss": 1.4408,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.54196697473526,
      "learning_rate": 0.0015,
      "loss": 1.4481,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.5527881979942322,
      "learning_rate": 0.0015,
      "loss": 1.4356,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.7087220549583435,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.547170877456665,
      "learning_rate": 0.0015,
      "loss": 1.4182,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.6261120438575745,
      "learning_rate": 0.0015,
      "loss": 1.4384,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.5071873068809509,
      "learning_rate": 0.0015,
      "loss": 1.421,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.5573424100875854,
      "learning_rate": 0.0015,
      "loss": 1.441,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.6330763101577759,
      "learning_rate": 0.0015,
      "loss": 1.43,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.584770143032074,
      "learning_rate": 0.0015,
      "loss": 1.431,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5460183620452881,
      "learning_rate": 0.0015,
      "loss": 1.4307,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.6885669231414795,
      "learning_rate": 0.0015,
      "loss": 1.4254,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.7244248986244202,
      "learning_rate": 0.0015,
      "loss": 1.4402,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.639764666557312,
      "learning_rate": 0.0015,
      "loss": 1.4336,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.6047524213790894,
      "learning_rate": 0.0015,
      "loss": 1.4299,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.9635192155838013,
      "learning_rate": 0.0015,
      "loss": 1.4311,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.604000449180603,
      "learning_rate": 0.0015,
      "loss": 1.4419,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.6478769183158875,
      "learning_rate": 0.0015,
      "loss": 1.4322,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5845364928245544,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.4125,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.9109283685684204,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.4363,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5506412386894226,
      "learning_rate": 0.001444828815847542,
      "loss": 1.4375,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.69488126039505,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.4264,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5653802752494812,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.4226,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.5073166489601135,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4305,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.5020584464073181,
      "learning_rate": 0.001366801618799797,
      "loss": 1.4168,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5227780342102051,
      "learning_rate": 0.001347962334099052,
      "loss": 1.4184,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.4819777309894562,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4247,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.5877610445022583,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4156,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.5603451132774353,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.4227,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5968028903007507,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.4207,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.5048508644104004,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.4161,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.7932562232017517,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4107,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.6583470702171326,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4086,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5934231281280518,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.4025,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.5164998173713684,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4091,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.5378324389457703,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4017,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.5045227408409119,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.4052,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.6413946747779846,
      "learning_rate": 0.001141156021915355,
      "loss": 1.4033,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.5191912055015564,
      "learning_rate": 0.001125426919103997,
      "loss": 1.3832,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.5992634892463684,
      "learning_rate": 0.001109914618088799,
      "loss": 1.3866,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.7955918312072754,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.4026,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.7773348689079285,
      "learning_rate": 0.001079528509501728,
      "loss": 1.4016,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.5571773052215576,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.3965,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.4990285038948059,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.3841,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.5601270198822021,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.3932,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.5050643682479858,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3765,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5214822888374329,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.3791,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.5144291520118713,
      "learning_rate": 0.000993270988848579,
      "loss": 1.3921,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5946586728096008,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.3861,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.7176449298858643,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.3913,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.45808374881744385,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.3844,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.5587730407714844,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.397,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.5175219774246216,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.3975,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.4646693766117096,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.3819,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.5937052965164185,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3725,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.5852144956588745,
      "learning_rate": 0.000888885722567627,
      "loss": 1.3832,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.4718400835990906,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.3758,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.5763649344444275,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3747,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.541046679019928,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.3712,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.5396777391433716,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3694,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.5645665526390076,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3736,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5329201817512512,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3778,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.47563907504081726,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.3632,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.48661863803863525,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3717,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.5526289939880371,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.366,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.6463859677314758,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.3674,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.4902038276195526,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3637,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.5209696888923645,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3631,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.5268991589546204,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.3648,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.5123580694198608,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.366,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.48749664425849915,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.3582,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.4711751639842987,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.3581,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.5960403084754944,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.3619,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.6335268616676331,
      "learning_rate": 0.000692383699170611,
      "loss": 1.3499,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.5655462145805359,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3591,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.5002266764640808,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3602,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.5185707807540894,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.3641,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.614739716053009,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3574,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.49084967374801636,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3588,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.5136387348175049,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3551,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.5223336815834045,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.3584,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.4992089867591858,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3447,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.5769894123077393,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3504,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.5668891668319702,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.3509,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.6820117235183716,
      "learning_rate": 0.000594349404202577,
      "loss": 1.3497,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.5896968245506287,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.3386,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.5106127262115479,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.3461,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5448651909828186,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3422,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.5113927125930786,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3395,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.5165741443634033,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.3465,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.4909664988517761,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3362,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.5048149824142456,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3356,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.48986268043518066,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.3586,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.5755218267440796,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.3396,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.5522173047065735,
      "learning_rate": 0.000517326298154212,
      "loss": 1.3455,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.5478484630584717,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3482,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.4900250732898712,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3445,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.473868727684021,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3346,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.5402085185050964,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.3276,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.5062165260314941,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3313,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.5080792903900146,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.3331,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.541022777557373,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3436,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.47802749276161194,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.3156,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.6114453077316284,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.3308,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.5069577693939209,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3348,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.6038907170295715,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3362,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5295746326446533,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3257,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.5159029960632324,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.3319,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.5625802874565125,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.3373,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.5231423377990723,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3331,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.5272135734558105,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3304,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.5432569980621338,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.3266,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5289936065673828,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.3385,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.5625152587890625,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3296,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.511777937412262,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.3387,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.5690923929214478,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.3217,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.46988624334335327,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.3181,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.4778980314731598,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3273,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.6266793012619019,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.3357,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.5043386816978455,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.3142,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.49880918860435486,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3154,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.5304259061813354,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.3211,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.5800580382347107,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.324,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.5011001825332642,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.3374,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.5401927828788757,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3331,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.487700492143631,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3285,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.49600911140441895,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3238,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.48997899889945984,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.3196,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.4990381896495819,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.318,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.5055996775627136,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.321,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.5182024240493774,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.3316,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.5547622442245483,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.335,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.4912651777267456,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.315,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.509490966796875,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.3106,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.5316436290740967,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3152,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.5913136601448059,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3138,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.4977995753288269,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3106,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.48362046480178833,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3176,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.5586822032928467,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3199,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.47931405901908875,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3068,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.4615863859653473,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.317,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.4860086441040039,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3204,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.4814727008342743,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.3026,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.5079910755157471,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3086,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.4841192364692688,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.295,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.47682246565818787,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.2986,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.5059146285057068,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.3018,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.4751178026199341,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.311,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.46254920959472656,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3038,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.49197128415107727,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3107,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.4641369879245758,
      "learning_rate": 0.000237799979917852,
      "loss": 1.3175,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.5277642011642456,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3112,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.4912530779838562,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.318,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.4632818102836609,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.3012,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.48173755407333374,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.3033,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.5092982053756714,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.2993,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.518355667591095,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.2987,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.4870567321777344,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.3004,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.5034220218658447,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3119,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.49648088216781616,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.3082,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.48447710275650024,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.2997,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.48734259605407715,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.3055,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.48913803696632385,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.3137,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.4835722744464874,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.2881,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.5292708873748779,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.3117,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.49038606882095337,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.3026,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.5049091577529907,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.3194,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.48894670605659485,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.2952,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.469951331615448,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.3035,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5279447436332703,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.3162,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.5621384382247925,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.2877,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.48656702041625977,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.3003,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.5730150938034058,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.2989,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.5166862607002258,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.3014,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.5328842997550964,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3076,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.46243101358413696,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3081,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.4814881384372711,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.2896,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5162342190742493,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.2925,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.46444493532180786,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.2996,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.5196223258972168,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.2922,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5249384641647339,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.3112,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.4600869417190552,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.3001,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.5109081268310547,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.2955,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.4703962206840515,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3057,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.49680525064468384,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.2954,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.4776685833930969,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.2821,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.5237215757369995,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.2897,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.45696312189102173,
      "learning_rate": 0.000142293772767289,
      "loss": 1.2787,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.5189747214317322,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.2911,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.5011923313140869,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3128,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.4645780026912689,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.3079,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.5004777312278748,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.2872,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.5973535776138306,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.2976,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.5669199824333191,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.296,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.46773242950439453,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3041,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.48976677656173706,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.2981,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.47129586338996887,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.3007,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.4685057997703552,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.2874,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.47045186161994934,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.2871,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.4687444567680359,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.2839,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.6013496518135071,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.2983,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.5217690467834473,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.2925,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.5072504281997681,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.2992,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.47519269585609436,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.2932,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.5145894289016724,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.2904,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.4669525623321533,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3018,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.4746662676334381,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.2822,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.5146121978759766,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.2903,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.465825617313385,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.2965,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4787508547306061,
      "learning_rate": 0.000104851799496331,
      "loss": 1.2973,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.512465238571167,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.2871,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.5024791955947876,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.2792,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.5367498397827148,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.2985,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.4867110848426819,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.2879,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.48658135533332825,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.2897,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.5055621266365051,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.2872,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.4670827388763428,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.2841,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.49410444498062134,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.2955,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.47974035143852234,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.2855,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.4761445224285126,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.293,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.48335903882980347,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.286,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.4784451425075531,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.2991,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.46803441643714905,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.2922,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.4703162610530853,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.2985,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.4672531485557556,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.2918,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.4863806664943695,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.2909,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.47227323055267334,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3038,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.5004388689994812,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.2945,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.5001670718193054,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.2913,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.4512779414653778,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2755,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.5033439993858337,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.2741,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.4707455039024353,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.3006,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.4731435775756836,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.2797,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.4715854823589325,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.2986,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.4568154215812683,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.2921,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.4706604480743408,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.2792,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.483224093914032,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.2834,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.4954493045806885,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.2805,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.4545811116695404,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.2882,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.49659305810928345,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.296,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.4791646897792816,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.2845,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.4629514217376709,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.2866,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.452927827835083,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.3022,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4618093967437744,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.2893,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.48160403966903687,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.2874,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.4891617000102997,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2814,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.5169182419776917,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.2893,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.5058215856552124,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.2871,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.460235059261322,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.2875,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.4831646978855133,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.2804,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.46499505639076233,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.2869,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.47788241505622864,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2784,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.46817654371261597,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.2869,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.48694300651550293,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.2842,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.4783858358860016,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.2958,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.4681575298309326,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.2928,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.5110870003700256,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.2857,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.47763773798942566,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2751,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.527594268321991,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.2961,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.45547598600387573,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.2859,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5116627812385559,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.291,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.453938752412796,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.2832,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.5115503668785095,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.2866,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.4982527494430542,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2828,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.4778030216693878,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.2913,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5285346508026123,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2752,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.4840639531612396,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2772,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.5008677840232849,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.2799,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.46262115240097046,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.2864,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.460462361574173,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.2941,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.47785452008247375,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.2868,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.45095816254615784,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.2859,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.46034693717956543,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.2843,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.4498767554759979,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.2785,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.4550316333770752,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.2917,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.4780294597148895,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2802,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.45463302731513977,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.2751,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.4685260057449341,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.2658,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4742572605609894,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.2785,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.48642367124557495,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.2836,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.44754165410995483,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2781,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.4758548438549042,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.274,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.4646531641483307,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.2955,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.4570622444152832,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.2805,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.48418715596199036,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.2977,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.4781525433063507,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.2817,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.4933951497077942,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.2811,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.5042923092842102,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.2817,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.4611527919769287,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.2844,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.48543456196784973,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.2881,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.4693308472633362,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.2877,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.4632687568664551,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2776,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.5444868206977844,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2757,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.44671541452407837,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.2827,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.46076348423957825,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.2806,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.49430784583091736,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.2814,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.45694178342819214,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.2829,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.5009331703186035,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.2863,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5236653685569763,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.2828,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5181384682655334,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.2812,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.4782514274120331,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.2875,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.4620096683502197,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.2805,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.4873718321323395,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.2786,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.48899489641189575,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.2766,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.481893390417099,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.2981,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.4656616449356079,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.2737,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.4694961607456207,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.285,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.45946142077445984,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.2851,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.46677711606025696,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.288,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.45549505949020386,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.2878,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.47640255093574524,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.2882,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.462722510099411,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2666,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.46981674432754517,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.279,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.4761793315410614,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.288,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.4618094861507416,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.2744,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.49426230788230896,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.2917,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.49430301785469055,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.2808,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.46488291025161743,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.2862,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.47296518087387085,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.2852,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.46403154730796814,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2819,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.5006576180458069,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2761,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.47130194306373596,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2807,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.46654370427131653,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.2812,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.46959954500198364,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.279,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.44776010513305664,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.2699,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.476149320602417,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.2654,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.4552365839481354,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.2774,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.46575242280960083,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.2737,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.4691290855407715,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.2821,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.47082453966140747,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2719,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.4488515853881836,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.2788,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.4794052541255951,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2789,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.4959319531917572,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2785,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.48605209589004517,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2741,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4825482964515686,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.2858,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.47821491956710815,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.2851,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.44771477580070496,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.2944,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.46373650431632996,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.2858,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.46483108401298523,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.2862,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.4672599732875824,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2671,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.4882128834724426,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2739,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.4712848365306854,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.2876,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.4877878427505493,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2627,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.4719924032688141,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.2839,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.45898473262786865,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.2839,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.46861326694488525,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.2782,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.45154401659965515,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.2865,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.4655771851539612,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.2716,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.4726909101009369,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.2829,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4056248664855957,
      "learning_rate": 1.5020833487815421e-05,
      "loss": 1.2754,
      "step": 9480
    }
  ],
  "logging_steps": 10,
  "max_steps": 9480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.832308198648013e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}