{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2033680145875791,
  "eval_steps": 500,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000163826998689384,
      "grad_norm": 3.011117458343506,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.9607,
      "step": 1
    },
    {
      "epoch": 0.000327653997378768,
      "grad_norm": 3.4058752059936523,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.0397,
      "step": 2
    },
    {
      "epoch": 0.000491480996068152,
      "grad_norm": 3.222395420074463,
      "learning_rate": 6e-06,
      "loss": 1.9606,
      "step": 3
    },
    {
      "epoch": 0.000655307994757536,
      "grad_norm": 2.819679021835327,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.0057,
      "step": 4
    },
    {
      "epoch": 0.00081913499344692,
      "grad_norm": 3.026033878326416,
      "learning_rate": 1e-05,
      "loss": 1.999,
      "step": 5
    },
    {
      "epoch": 0.000982961992136304,
      "grad_norm": 2.993104934692383,
      "learning_rate": 1.2e-05,
      "loss": 1.9474,
      "step": 6
    },
    {
      "epoch": 0.0011467889908256881,
      "grad_norm": 1.598062515258789,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 1.9593,
      "step": 7
    },
    {
      "epoch": 0.001310615989515072,
      "grad_norm": 1.3988381624221802,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.8104,
      "step": 8
    },
    {
      "epoch": 0.001474442988204456,
      "grad_norm": 1.1895484924316406,
      "learning_rate": 1.8e-05,
      "loss": 1.842,
      "step": 9
    },
    {
      "epoch": 0.00163826998689384,
      "grad_norm": 0.9336122870445251,
      "learning_rate": 2e-05,
      "loss": 1.8687,
      "step": 10
    },
    {
      "epoch": 0.0018020969855832242,
      "grad_norm": 0.5977860689163208,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.7648,
      "step": 11
    },
    {
      "epoch": 0.001965923984272608,
      "grad_norm": 0.48661908507347107,
      "learning_rate": 2.4e-05,
      "loss": 1.7359,
      "step": 12
    },
    {
      "epoch": 0.0021297509829619923,
      "grad_norm": 0.6308302283287048,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 1.7997,
      "step": 13
    },
    {
      "epoch": 0.0022935779816513763,
      "grad_norm": 0.427565336227417,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 1.7374,
      "step": 14
    },
    {
      "epoch": 0.0024574049803407602,
      "grad_norm": 0.43908464908599854,
      "learning_rate": 3e-05,
      "loss": 1.6684,
      "step": 15
    },
    {
      "epoch": 0.002621231979030144,
      "grad_norm": 0.44852790236473083,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.6539,
      "step": 16
    },
    {
      "epoch": 0.002785058977719528,
      "grad_norm": 0.43064501881599426,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.6744,
      "step": 17
    },
    {
      "epoch": 0.002948885976408912,
      "grad_norm": 0.3355182111263275,
      "learning_rate": 3.6e-05,
      "loss": 1.697,
      "step": 18
    },
    {
      "epoch": 0.003112712975098296,
      "grad_norm": 0.3131241202354431,
      "learning_rate": 3.8e-05,
      "loss": 1.6305,
      "step": 19
    },
    {
      "epoch": 0.00327653997378768,
      "grad_norm": 0.29811975359916687,
      "learning_rate": 4e-05,
      "loss": 1.697,
      "step": 20
    },
    {
      "epoch": 0.0034403669724770644,
      "grad_norm": 0.18077552318572998,
      "learning_rate": 4.2e-05,
      "loss": 1.6952,
      "step": 21
    },
    {
      "epoch": 0.0036041939711664484,
      "grad_norm": 0.15599356591701508,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.6369,
      "step": 22
    },
    {
      "epoch": 0.0037680209698558323,
      "grad_norm": 0.20569440722465515,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.695,
      "step": 23
    },
    {
      "epoch": 0.003931847968545216,
      "grad_norm": 0.19235457479953766,
      "learning_rate": 4.8e-05,
      "loss": 1.5212,
      "step": 24
    },
    {
      "epoch": 0.0040956749672346,
      "grad_norm": 0.2047315090894699,
      "learning_rate": 5e-05,
      "loss": 1.6077,
      "step": 25
    },
    {
      "epoch": 0.004259501965923985,
      "grad_norm": 0.21554157137870789,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 1.682,
      "step": 26
    },
    {
      "epoch": 0.004423328964613368,
      "grad_norm": 0.19817550480365753,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 1.5661,
      "step": 27
    },
    {
      "epoch": 0.0045871559633027525,
      "grad_norm": 0.21529093384742737,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 1.5741,
      "step": 28
    },
    {
      "epoch": 0.004750982961992136,
      "grad_norm": 0.19022610783576965,
      "learning_rate": 5.8e-05,
      "loss": 1.4729,
      "step": 29
    },
    {
      "epoch": 0.0049148099606815205,
      "grad_norm": 0.17255447804927826,
      "learning_rate": 6e-05,
      "loss": 1.5372,
      "step": 30
    },
    {
      "epoch": 0.005078636959370904,
      "grad_norm": 0.1521906703710556,
      "learning_rate": 6.2e-05,
      "loss": 1.4949,
      "step": 31
    },
    {
      "epoch": 0.005242463958060288,
      "grad_norm": 0.1881936937570572,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.592,
      "step": 32
    },
    {
      "epoch": 0.005406290956749673,
      "grad_norm": 0.18808823823928833,
      "learning_rate": 6.6e-05,
      "loss": 1.7965,
      "step": 33
    },
    {
      "epoch": 0.005570117955439056,
      "grad_norm": 0.16976410150527954,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.6898,
      "step": 34
    },
    {
      "epoch": 0.005733944954128441,
      "grad_norm": 0.17400828003883362,
      "learning_rate": 7e-05,
      "loss": 1.6652,
      "step": 35
    },
    {
      "epoch": 0.005897771952817824,
      "grad_norm": 0.15237464010715485,
      "learning_rate": 7.2e-05,
      "loss": 1.6014,
      "step": 36
    },
    {
      "epoch": 0.006061598951507209,
      "grad_norm": 0.14509867131710052,
      "learning_rate": 7.4e-05,
      "loss": 1.5321,
      "step": 37
    },
    {
      "epoch": 0.006225425950196592,
      "grad_norm": 0.17446769773960114,
      "learning_rate": 7.6e-05,
      "loss": 1.7037,
      "step": 38
    },
    {
      "epoch": 0.0063892529488859765,
      "grad_norm": 0.15834857523441315,
      "learning_rate": 7.800000000000001e-05,
      "loss": 1.6677,
      "step": 39
    },
    {
      "epoch": 0.00655307994757536,
      "grad_norm": 0.15981952846050262,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 40
    },
    {
      "epoch": 0.006716906946264744,
      "grad_norm": 0.18606367707252502,
      "learning_rate": 8.2e-05,
      "loss": 1.5002,
      "step": 41
    },
    {
      "epoch": 0.006880733944954129,
      "grad_norm": 0.16529963910579681,
      "learning_rate": 8.4e-05,
      "loss": 1.6444,
      "step": 42
    },
    {
      "epoch": 0.007044560943643512,
      "grad_norm": 0.167267307639122,
      "learning_rate": 8.6e-05,
      "loss": 1.5973,
      "step": 43
    },
    {
      "epoch": 0.007208387942332897,
      "grad_norm": 0.1671181619167328,
      "learning_rate": 8.800000000000001e-05,
      "loss": 1.527,
      "step": 44
    },
    {
      "epoch": 0.00737221494102228,
      "grad_norm": 0.15473029017448425,
      "learning_rate": 9e-05,
      "loss": 1.6306,
      "step": 45
    },
    {
      "epoch": 0.007536041939711665,
      "grad_norm": 0.14409710466861725,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.5615,
      "step": 46
    },
    {
      "epoch": 0.007699868938401048,
      "grad_norm": 0.15754491090774536,
      "learning_rate": 9.4e-05,
      "loss": 1.6799,
      "step": 47
    },
    {
      "epoch": 0.007863695937090432,
      "grad_norm": 0.13560158014297485,
      "learning_rate": 9.6e-05,
      "loss": 1.6155,
      "step": 48
    },
    {
      "epoch": 0.008027522935779817,
      "grad_norm": 0.15016230940818787,
      "learning_rate": 9.8e-05,
      "loss": 1.7633,
      "step": 49
    },
    {
      "epoch": 0.0081913499344692,
      "grad_norm": 0.13735011219978333,
      "learning_rate": 0.0001,
      "loss": 1.5739,
      "step": 50
    },
    {
      "epoch": 0.008355176933158584,
      "grad_norm": 0.14251820743083954,
      "learning_rate": 0.0001,
      "loss": 1.589,
      "step": 51
    },
    {
      "epoch": 0.00851900393184797,
      "grad_norm": 0.153127983212471,
      "learning_rate": 0.0001,
      "loss": 1.5962,
      "step": 52
    },
    {
      "epoch": 0.008682830930537353,
      "grad_norm": 0.14780348539352417,
      "learning_rate": 0.0001,
      "loss": 1.5557,
      "step": 53
    },
    {
      "epoch": 0.008846657929226736,
      "grad_norm": 0.14456704258918762,
      "learning_rate": 0.0001,
      "loss": 1.6047,
      "step": 54
    },
    {
      "epoch": 0.00901048492791612,
      "grad_norm": 0.1672697812318802,
      "learning_rate": 0.0001,
      "loss": 1.6097,
      "step": 55
    },
    {
      "epoch": 0.009174311926605505,
      "grad_norm": 0.15527459979057312,
      "learning_rate": 0.0001,
      "loss": 1.6285,
      "step": 56
    },
    {
      "epoch": 0.009338138925294889,
      "grad_norm": 0.14947618544101715,
      "learning_rate": 0.0001,
      "loss": 1.6069,
      "step": 57
    },
    {
      "epoch": 0.009501965923984272,
      "grad_norm": 0.13615983724594116,
      "learning_rate": 0.0001,
      "loss": 1.6083,
      "step": 58
    },
    {
      "epoch": 0.009665792922673657,
      "grad_norm": 0.13993817567825317,
      "learning_rate": 0.0001,
      "loss": 1.491,
      "step": 59
    },
    {
      "epoch": 0.009829619921363041,
      "grad_norm": 0.13932132720947266,
      "learning_rate": 0.0001,
      "loss": 1.698,
      "step": 60
    },
    {
      "epoch": 0.009993446920052424,
      "grad_norm": 0.1486559808254242,
      "learning_rate": 0.0001,
      "loss": 1.6142,
      "step": 61
    },
    {
      "epoch": 0.010157273918741808,
      "grad_norm": 0.1372678428888321,
      "learning_rate": 0.0001,
      "loss": 1.6342,
      "step": 62
    },
    {
      "epoch": 0.010321100917431193,
      "grad_norm": 0.12783578038215637,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 63
    },
    {
      "epoch": 0.010484927916120577,
      "grad_norm": 0.13788290321826935,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 64
    },
    {
      "epoch": 0.01064875491480996,
      "grad_norm": 0.13062472641468048,
      "learning_rate": 0.0001,
      "loss": 1.482,
      "step": 65
    },
    {
      "epoch": 0.010812581913499346,
      "grad_norm": 0.1475992500782013,
      "learning_rate": 0.0001,
      "loss": 1.5932,
      "step": 66
    },
    {
      "epoch": 0.010976408912188729,
      "grad_norm": 0.135057732462883,
      "learning_rate": 0.0001,
      "loss": 1.5361,
      "step": 67
    },
    {
      "epoch": 0.011140235910878113,
      "grad_norm": 0.14832264184951782,
      "learning_rate": 0.0001,
      "loss": 1.7138,
      "step": 68
    },
    {
      "epoch": 0.011304062909567496,
      "grad_norm": 0.1267196387052536,
      "learning_rate": 0.0001,
      "loss": 1.5848,
      "step": 69
    },
    {
      "epoch": 0.011467889908256881,
      "grad_norm": 0.1351178139448166,
      "learning_rate": 0.0001,
      "loss": 1.4603,
      "step": 70
    },
    {
      "epoch": 0.011631716906946265,
      "grad_norm": 0.13595224916934967,
      "learning_rate": 0.0001,
      "loss": 1.5799,
      "step": 71
    },
    {
      "epoch": 0.011795543905635648,
      "grad_norm": 0.13580814003944397,
      "learning_rate": 0.0001,
      "loss": 1.558,
      "step": 72
    },
    {
      "epoch": 0.011959370904325032,
      "grad_norm": 0.14337950944900513,
      "learning_rate": 0.0001,
      "loss": 1.553,
      "step": 73
    },
    {
      "epoch": 0.012123197903014417,
      "grad_norm": 0.14592956006526947,
      "learning_rate": 0.0001,
      "loss": 1.6583,
      "step": 74
    },
    {
      "epoch": 0.0122870249017038,
      "grad_norm": 0.12927336990833282,
      "learning_rate": 0.0001,
      "loss": 1.5457,
      "step": 75
    },
    {
      "epoch": 0.012450851900393184,
      "grad_norm": 0.15650339424610138,
      "learning_rate": 0.0001,
      "loss": 1.541,
      "step": 76
    },
    {
      "epoch": 0.01261467889908257,
      "grad_norm": 0.12211811542510986,
      "learning_rate": 0.0001,
      "loss": 1.3893,
      "step": 77
    },
    {
      "epoch": 0.012778505897771953,
      "grad_norm": 0.14160878956317902,
      "learning_rate": 0.0001,
      "loss": 1.6178,
      "step": 78
    },
    {
      "epoch": 0.012942332896461337,
      "grad_norm": 0.13554741442203522,
      "learning_rate": 0.0001,
      "loss": 1.5746,
      "step": 79
    },
    {
      "epoch": 0.01310615989515072,
      "grad_norm": 0.13684110343456268,
      "learning_rate": 0.0001,
      "loss": 1.4363,
      "step": 80
    },
    {
      "epoch": 0.013269986893840105,
      "grad_norm": 0.13343599438667297,
      "learning_rate": 0.0001,
      "loss": 1.5856,
      "step": 81
    },
    {
      "epoch": 0.013433813892529489,
      "grad_norm": 0.15745502710342407,
      "learning_rate": 0.0001,
      "loss": 1.6274,
      "step": 82
    },
    {
      "epoch": 0.013597640891218872,
      "grad_norm": 0.14018264412879944,
      "learning_rate": 0.0001,
      "loss": 1.6055,
      "step": 83
    },
    {
      "epoch": 0.013761467889908258,
      "grad_norm": 0.13848982751369476,
      "learning_rate": 0.0001,
      "loss": 1.6515,
      "step": 84
    },
    {
      "epoch": 0.013925294888597641,
      "grad_norm": 0.1455080807209015,
      "learning_rate": 0.0001,
      "loss": 1.5835,
      "step": 85
    },
    {
      "epoch": 0.014089121887287025,
      "grad_norm": 0.1383785754442215,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 86
    },
    {
      "epoch": 0.014252948885976408,
      "grad_norm": 0.13937243819236755,
      "learning_rate": 0.0001,
      "loss": 1.6273,
      "step": 87
    },
    {
      "epoch": 0.014416775884665793,
      "grad_norm": 0.14290837943553925,
      "learning_rate": 0.0001,
      "loss": 1.575,
      "step": 88
    },
    {
      "epoch": 0.014580602883355177,
      "grad_norm": 0.12148977071046829,
      "learning_rate": 0.0001,
      "loss": 1.5953,
      "step": 89
    },
    {
      "epoch": 0.01474442988204456,
      "grad_norm": 0.13915131986141205,
      "learning_rate": 0.0001,
      "loss": 1.5817,
      "step": 90
    },
    {
      "epoch": 0.014908256880733946,
      "grad_norm": 0.14456072449684143,
      "learning_rate": 0.0001,
      "loss": 1.6662,
      "step": 91
    },
    {
      "epoch": 0.01507208387942333,
      "grad_norm": 0.1521051675081253,
      "learning_rate": 0.0001,
      "loss": 1.5484,
      "step": 92
    },
    {
      "epoch": 0.015235910878112713,
      "grad_norm": 0.2531245946884155,
      "learning_rate": 0.0001,
      "loss": 1.4908,
      "step": 93
    },
    {
      "epoch": 0.015399737876802096,
      "grad_norm": 0.17517811059951782,
      "learning_rate": 0.0001,
      "loss": 1.5652,
      "step": 94
    },
    {
      "epoch": 0.015563564875491482,
      "grad_norm": 0.1450427919626236,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 95
    },
    {
      "epoch": 0.015727391874180863,
      "grad_norm": 0.13058120012283325,
      "learning_rate": 0.0001,
      "loss": 1.483,
      "step": 96
    },
    {
      "epoch": 0.01589121887287025,
      "grad_norm": 0.1680396944284439,
      "learning_rate": 0.0001,
      "loss": 1.6169,
      "step": 97
    },
    {
      "epoch": 0.016055045871559634,
      "grad_norm": 0.13577497005462646,
      "learning_rate": 0.0001,
      "loss": 1.5546,
      "step": 98
    },
    {
      "epoch": 0.016218872870249017,
      "grad_norm": 0.1347057968378067,
      "learning_rate": 0.0001,
      "loss": 1.5294,
      "step": 99
    },
    {
      "epoch": 0.0163826998689384,
      "grad_norm": 0.1565379649400711,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 100
    },
    {
      "epoch": 0.016546526867627784,
      "grad_norm": 0.13265185058116913,
      "learning_rate": 0.0001,
      "loss": 1.6414,
      "step": 101
    },
    {
      "epoch": 0.016710353866317168,
      "grad_norm": 0.12347199767827988,
      "learning_rate": 0.0001,
      "loss": 1.5819,
      "step": 102
    },
    {
      "epoch": 0.01687418086500655,
      "grad_norm": 0.14774161577224731,
      "learning_rate": 0.0001,
      "loss": 1.7008,
      "step": 103
    },
    {
      "epoch": 0.01703800786369594,
      "grad_norm": 0.1292174607515335,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 104
    },
    {
      "epoch": 0.017201834862385322,
      "grad_norm": 0.14048151671886444,
      "learning_rate": 0.0001,
      "loss": 1.4968,
      "step": 105
    },
    {
      "epoch": 0.017365661861074706,
      "grad_norm": 0.13337083160877228,
      "learning_rate": 0.0001,
      "loss": 1.5841,
      "step": 106
    },
    {
      "epoch": 0.01752948885976409,
      "grad_norm": 0.124249666929245,
      "learning_rate": 0.0001,
      "loss": 1.5548,
      "step": 107
    },
    {
      "epoch": 0.017693315858453473,
      "grad_norm": 0.12418890744447708,
      "learning_rate": 0.0001,
      "loss": 1.5109,
      "step": 108
    },
    {
      "epoch": 0.017857142857142856,
      "grad_norm": 0.1261230856180191,
      "learning_rate": 0.0001,
      "loss": 1.5647,
      "step": 109
    },
    {
      "epoch": 0.01802096985583224,
      "grad_norm": 0.12908600270748138,
      "learning_rate": 0.0001,
      "loss": 1.566,
      "step": 110
    },
    {
      "epoch": 0.018184796854521627,
      "grad_norm": 0.13716214895248413,
      "learning_rate": 0.0001,
      "loss": 1.4972,
      "step": 111
    },
    {
      "epoch": 0.01834862385321101,
      "grad_norm": 0.12899552285671234,
      "learning_rate": 0.0001,
      "loss": 1.4513,
      "step": 112
    },
    {
      "epoch": 0.018512450851900394,
      "grad_norm": 0.14466632902622223,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 113
    },
    {
      "epoch": 0.018676277850589777,
      "grad_norm": 0.13627144694328308,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 114
    },
    {
      "epoch": 0.01884010484927916,
      "grad_norm": 0.14187748730182648,
      "learning_rate": 0.0001,
      "loss": 1.5715,
      "step": 115
    },
    {
      "epoch": 0.019003931847968544,
      "grad_norm": 0.1391870677471161,
      "learning_rate": 0.0001,
      "loss": 1.5012,
      "step": 116
    },
    {
      "epoch": 0.019167758846657928,
      "grad_norm": 0.13076739013195038,
      "learning_rate": 0.0001,
      "loss": 1.482,
      "step": 117
    },
    {
      "epoch": 0.019331585845347315,
      "grad_norm": 0.13955922424793243,
      "learning_rate": 0.0001,
      "loss": 1.6041,
      "step": 118
    },
    {
      "epoch": 0.0194954128440367,
      "grad_norm": 0.13336504995822906,
      "learning_rate": 0.0001,
      "loss": 1.567,
      "step": 119
    },
    {
      "epoch": 0.019659239842726082,
      "grad_norm": 0.13444162905216217,
      "learning_rate": 0.0001,
      "loss": 1.6219,
      "step": 120
    },
    {
      "epoch": 0.019823066841415465,
      "grad_norm": 0.14157338440418243,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 121
    },
    {
      "epoch": 0.01998689384010485,
      "grad_norm": 0.13454873859882355,
      "learning_rate": 0.0001,
      "loss": 1.5664,
      "step": 122
    },
    {
      "epoch": 0.020150720838794232,
      "grad_norm": 0.13625149428844452,
      "learning_rate": 0.0001,
      "loss": 1.5915,
      "step": 123
    },
    {
      "epoch": 0.020314547837483616,
      "grad_norm": 0.12858495116233826,
      "learning_rate": 0.0001,
      "loss": 1.4959,
      "step": 124
    },
    {
      "epoch": 0.020478374836173003,
      "grad_norm": 0.13190023601055145,
      "learning_rate": 0.0001,
      "loss": 1.5109,
      "step": 125
    },
    {
      "epoch": 0.020642201834862386,
      "grad_norm": 0.13660550117492676,
      "learning_rate": 0.0001,
      "loss": 1.6081,
      "step": 126
    },
    {
      "epoch": 0.02080602883355177,
      "grad_norm": 0.1281704604625702,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 127
    },
    {
      "epoch": 0.020969855832241154,
      "grad_norm": 0.13630761206150055,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 128
    },
    {
      "epoch": 0.021133682830930537,
      "grad_norm": 0.12346865236759186,
      "learning_rate": 0.0001,
      "loss": 1.4249,
      "step": 129
    },
    {
      "epoch": 0.02129750982961992,
      "grad_norm": 0.140494704246521,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 130
    },
    {
      "epoch": 0.021461336828309304,
      "grad_norm": 0.12337122112512589,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 131
    },
    {
      "epoch": 0.02162516382699869,
      "grad_norm": 0.15968075394630432,
      "learning_rate": 0.0001,
      "loss": 1.5994,
      "step": 132
    },
    {
      "epoch": 0.021788990825688075,
      "grad_norm": 0.1251770704984665,
      "learning_rate": 0.0001,
      "loss": 1.6207,
      "step": 133
    },
    {
      "epoch": 0.021952817824377458,
      "grad_norm": 0.13994941115379333,
      "learning_rate": 0.0001,
      "loss": 1.6013,
      "step": 134
    },
    {
      "epoch": 0.02211664482306684,
      "grad_norm": 0.14387977123260498,
      "learning_rate": 0.0001,
      "loss": 1.55,
      "step": 135
    },
    {
      "epoch": 0.022280471821756225,
      "grad_norm": 0.13476435840129852,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 136
    },
    {
      "epoch": 0.02244429882044561,
      "grad_norm": 0.12425505369901657,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 137
    },
    {
      "epoch": 0.022608125819134992,
      "grad_norm": 0.1455349624156952,
      "learning_rate": 0.0001,
      "loss": 1.4925,
      "step": 138
    },
    {
      "epoch": 0.022771952817824376,
      "grad_norm": 0.1297989785671234,
      "learning_rate": 0.0001,
      "loss": 1.5655,
      "step": 139
    },
    {
      "epoch": 0.022935779816513763,
      "grad_norm": 0.12332233786582947,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 140
    },
    {
      "epoch": 0.023099606815203146,
      "grad_norm": 0.1306566745042801,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 141
    },
    {
      "epoch": 0.02326343381389253,
      "grad_norm": 0.1275010108947754,
      "learning_rate": 0.0001,
      "loss": 1.5721,
      "step": 142
    },
    {
      "epoch": 0.023427260812581913,
      "grad_norm": 0.1295231580734253,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 143
    },
    {
      "epoch": 0.023591087811271297,
      "grad_norm": 0.13065816462039948,
      "learning_rate": 0.0001,
      "loss": 1.6432,
      "step": 144
    },
    {
      "epoch": 0.02375491480996068,
      "grad_norm": 0.11832357943058014,
      "learning_rate": 0.0001,
      "loss": 1.4586,
      "step": 145
    },
    {
      "epoch": 0.023918741808650064,
      "grad_norm": 0.12194979935884476,
      "learning_rate": 0.0001,
      "loss": 1.5058,
      "step": 146
    },
    {
      "epoch": 0.02408256880733945,
      "grad_norm": 0.12897257506847382,
      "learning_rate": 0.0001,
      "loss": 1.5566,
      "step": 147
    },
    {
      "epoch": 0.024246395806028834,
      "grad_norm": 0.12578266859054565,
      "learning_rate": 0.0001,
      "loss": 1.5906,
      "step": 148
    },
    {
      "epoch": 0.024410222804718218,
      "grad_norm": 0.13263170421123505,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 149
    },
    {
      "epoch": 0.0245740498034076,
      "grad_norm": 0.12266577780246735,
      "learning_rate": 0.0001,
      "loss": 1.5252,
      "step": 150
    },
    {
      "epoch": 0.024737876802096985,
      "grad_norm": 0.12437823414802551,
      "learning_rate": 0.0001,
      "loss": 1.5858,
      "step": 151
    },
    {
      "epoch": 0.02490170380078637,
      "grad_norm": 0.13961590826511383,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 152
    },
    {
      "epoch": 0.025065530799475752,
      "grad_norm": 0.13006527721881866,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 153
    },
    {
      "epoch": 0.02522935779816514,
      "grad_norm": 0.1338171511888504,
      "learning_rate": 0.0001,
      "loss": 1.4854,
      "step": 154
    },
    {
      "epoch": 0.025393184796854523,
      "grad_norm": 0.1268617808818817,
      "learning_rate": 0.0001,
      "loss": 1.4862,
      "step": 155
    },
    {
      "epoch": 0.025557011795543906,
      "grad_norm": 0.16333773732185364,
      "learning_rate": 0.0001,
      "loss": 1.6142,
      "step": 156
    },
    {
      "epoch": 0.02572083879423329,
      "grad_norm": 0.13431358337402344,
      "learning_rate": 0.0001,
      "loss": 1.5615,
      "step": 157
    },
    {
      "epoch": 0.025884665792922673,
      "grad_norm": 0.14530155062675476,
      "learning_rate": 0.0001,
      "loss": 1.4522,
      "step": 158
    },
    {
      "epoch": 0.026048492791612057,
      "grad_norm": 0.12248417735099792,
      "learning_rate": 0.0001,
      "loss": 1.5354,
      "step": 159
    },
    {
      "epoch": 0.02621231979030144,
      "grad_norm": 0.12270587682723999,
      "learning_rate": 0.0001,
      "loss": 1.5301,
      "step": 160
    },
    {
      "epoch": 0.026376146788990827,
      "grad_norm": 0.12352810800075531,
      "learning_rate": 0.0001,
      "loss": 1.5482,
      "step": 161
    },
    {
      "epoch": 0.02653997378768021,
      "grad_norm": 0.1266404390335083,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 162
    },
    {
      "epoch": 0.026703800786369594,
      "grad_norm": 0.12465646862983704,
      "learning_rate": 0.0001,
      "loss": 1.6164,
      "step": 163
    },
    {
      "epoch": 0.026867627785058978,
      "grad_norm": 0.12986133992671967,
      "learning_rate": 0.0001,
      "loss": 1.4972,
      "step": 164
    },
    {
      "epoch": 0.02703145478374836,
      "grad_norm": 0.12310322374105453,
      "learning_rate": 0.0001,
      "loss": 1.5845,
      "step": 165
    },
    {
      "epoch": 0.027195281782437745,
      "grad_norm": 0.1234891340136528,
      "learning_rate": 0.0001,
      "loss": 1.5675,
      "step": 166
    },
    {
      "epoch": 0.02735910878112713,
      "grad_norm": 0.12272564321756363,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 167
    },
    {
      "epoch": 0.027522935779816515,
      "grad_norm": 0.13187192380428314,
      "learning_rate": 0.0001,
      "loss": 1.5584,
      "step": 168
    },
    {
      "epoch": 0.0276867627785059,
      "grad_norm": 0.13356643915176392,
      "learning_rate": 0.0001,
      "loss": 1.4894,
      "step": 169
    },
    {
      "epoch": 0.027850589777195282,
      "grad_norm": 0.1312696784734726,
      "learning_rate": 0.0001,
      "loss": 1.5677,
      "step": 170
    },
    {
      "epoch": 0.028014416775884666,
      "grad_norm": 0.1294780969619751,
      "learning_rate": 0.0001,
      "loss": 1.5876,
      "step": 171
    },
    {
      "epoch": 0.02817824377457405,
      "grad_norm": 0.11621183156967163,
      "learning_rate": 0.0001,
      "loss": 1.4188,
      "step": 172
    },
    {
      "epoch": 0.028342070773263433,
      "grad_norm": 0.14989803731441498,
      "learning_rate": 0.0001,
      "loss": 1.5644,
      "step": 173
    },
    {
      "epoch": 0.028505897771952816,
      "grad_norm": 0.12611323595046997,
      "learning_rate": 0.0001,
      "loss": 1.5026,
      "step": 174
    },
    {
      "epoch": 0.028669724770642203,
      "grad_norm": 0.12209807336330414,
      "learning_rate": 0.0001,
      "loss": 1.5755,
      "step": 175
    },
    {
      "epoch": 0.028833551769331587,
      "grad_norm": 0.1330537647008896,
      "learning_rate": 0.0001,
      "loss": 1.4403,
      "step": 176
    },
    {
      "epoch": 0.02899737876802097,
      "grad_norm": 0.1380019634962082,
      "learning_rate": 0.0001,
      "loss": 1.5888,
      "step": 177
    },
    {
      "epoch": 0.029161205766710354,
      "grad_norm": 0.1515824943780899,
      "learning_rate": 0.0001,
      "loss": 1.5832,
      "step": 178
    },
    {
      "epoch": 0.029325032765399738,
      "grad_norm": 0.13740727305412292,
      "learning_rate": 0.0001,
      "loss": 1.6936,
      "step": 179
    },
    {
      "epoch": 0.02948885976408912,
      "grad_norm": 0.1259126216173172,
      "learning_rate": 0.0001,
      "loss": 1.4353,
      "step": 180
    },
    {
      "epoch": 0.029652686762778505,
      "grad_norm": 0.13039293885231018,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 181
    },
    {
      "epoch": 0.02981651376146789,
      "grad_norm": 0.13748720288276672,
      "learning_rate": 0.0001,
      "loss": 1.539,
      "step": 182
    },
    {
      "epoch": 0.029980340760157275,
      "grad_norm": 0.13033299148082733,
      "learning_rate": 0.0001,
      "loss": 1.4259,
      "step": 183
    },
    {
      "epoch": 0.03014416775884666,
      "grad_norm": 0.1298520267009735,
      "learning_rate": 0.0001,
      "loss": 1.6139,
      "step": 184
    },
    {
      "epoch": 0.030307994757536042,
      "grad_norm": 0.1275649219751358,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 185
    },
    {
      "epoch": 0.030471821756225426,
      "grad_norm": 0.1331489086151123,
      "learning_rate": 0.0001,
      "loss": 1.5204,
      "step": 186
    },
    {
      "epoch": 0.03063564875491481,
      "grad_norm": 0.1291024088859558,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 187
    },
    {
      "epoch": 0.030799475753604193,
      "grad_norm": 0.11880326271057129,
      "learning_rate": 0.0001,
      "loss": 1.5131,
      "step": 188
    },
    {
      "epoch": 0.03096330275229358,
      "grad_norm": 0.1341976523399353,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 189
    },
    {
      "epoch": 0.031127129750982963,
      "grad_norm": 0.14086049795150757,
      "learning_rate": 0.0001,
      "loss": 1.5667,
      "step": 190
    },
    {
      "epoch": 0.03129095674967235,
      "grad_norm": 0.12789331376552582,
      "learning_rate": 0.0001,
      "loss": 1.5969,
      "step": 191
    },
    {
      "epoch": 0.03145478374836173,
      "grad_norm": 0.12503425776958466,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 192
    },
    {
      "epoch": 0.031618610747051114,
      "grad_norm": 0.12992706894874573,
      "learning_rate": 0.0001,
      "loss": 1.5911,
      "step": 193
    },
    {
      "epoch": 0.0317824377457405,
      "grad_norm": 0.13048727810382843,
      "learning_rate": 0.0001,
      "loss": 1.5736,
      "step": 194
    },
    {
      "epoch": 0.03194626474442988,
      "grad_norm": 0.127937912940979,
      "learning_rate": 0.0001,
      "loss": 1.5653,
      "step": 195
    },
    {
      "epoch": 0.03211009174311927,
      "grad_norm": 0.13289302587509155,
      "learning_rate": 0.0001,
      "loss": 1.5979,
      "step": 196
    },
    {
      "epoch": 0.03227391874180865,
      "grad_norm": 0.12686611711978912,
      "learning_rate": 0.0001,
      "loss": 1.6094,
      "step": 197
    },
    {
      "epoch": 0.032437745740498035,
      "grad_norm": 0.12489918619394302,
      "learning_rate": 0.0001,
      "loss": 1.4987,
      "step": 198
    },
    {
      "epoch": 0.032601572739187415,
      "grad_norm": 0.1302986443042755,
      "learning_rate": 0.0001,
      "loss": 1.5461,
      "step": 199
    },
    {
      "epoch": 0.0327653997378768,
      "grad_norm": 0.12994195520877838,
      "learning_rate": 0.0001,
      "loss": 1.5081,
      "step": 200
    },
    {
      "epoch": 0.03292922673656619,
      "grad_norm": 0.1251252442598343,
      "learning_rate": 0.0001,
      "loss": 1.6199,
      "step": 201
    },
    {
      "epoch": 0.03309305373525557,
      "grad_norm": 0.13519275188446045,
      "learning_rate": 0.0001,
      "loss": 1.6202,
      "step": 202
    },
    {
      "epoch": 0.033256880733944956,
      "grad_norm": 0.12701208889484406,
      "learning_rate": 0.0001,
      "loss": 1.4802,
      "step": 203
    },
    {
      "epoch": 0.033420707732634336,
      "grad_norm": 0.12712019681930542,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 204
    },
    {
      "epoch": 0.03358453473132372,
      "grad_norm": 0.12929703295230865,
      "learning_rate": 0.0001,
      "loss": 1.568,
      "step": 205
    },
    {
      "epoch": 0.0337483617300131,
      "grad_norm": 0.1312536746263504,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 206
    },
    {
      "epoch": 0.03391218872870249,
      "grad_norm": 0.16917507350444794,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 207
    },
    {
      "epoch": 0.03407601572739188,
      "grad_norm": 0.14357730746269226,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 208
    },
    {
      "epoch": 0.03423984272608126,
      "grad_norm": 0.14517861604690552,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 209
    },
    {
      "epoch": 0.034403669724770644,
      "grad_norm": 0.12322504073381424,
      "learning_rate": 0.0001,
      "loss": 1.5074,
      "step": 210
    },
    {
      "epoch": 0.034567496723460024,
      "grad_norm": 0.15104396641254425,
      "learning_rate": 0.0001,
      "loss": 1.5845,
      "step": 211
    },
    {
      "epoch": 0.03473132372214941,
      "grad_norm": 0.12808775901794434,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 212
    },
    {
      "epoch": 0.03489515072083879,
      "grad_norm": 0.14732445776462555,
      "learning_rate": 0.0001,
      "loss": 1.5922,
      "step": 213
    },
    {
      "epoch": 0.03505897771952818,
      "grad_norm": 0.13598917424678802,
      "learning_rate": 0.0001,
      "loss": 1.5768,
      "step": 214
    },
    {
      "epoch": 0.035222804718217565,
      "grad_norm": 0.1341174989938736,
      "learning_rate": 0.0001,
      "loss": 1.5422,
      "step": 215
    },
    {
      "epoch": 0.035386631716906945,
      "grad_norm": 0.1326247602701187,
      "learning_rate": 0.0001,
      "loss": 1.6768,
      "step": 216
    },
    {
      "epoch": 0.03555045871559633,
      "grad_norm": 0.14427874982357025,
      "learning_rate": 0.0001,
      "loss": 1.4822,
      "step": 217
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 0.13326330482959747,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 218
    },
    {
      "epoch": 0.0358781127129751,
      "grad_norm": 0.13189126551151276,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 219
    },
    {
      "epoch": 0.03604193971166448,
      "grad_norm": 0.12171744555234909,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 220
    },
    {
      "epoch": 0.036205766710353866,
      "grad_norm": 0.1379660665988922,
      "learning_rate": 0.0001,
      "loss": 1.7044,
      "step": 221
    },
    {
      "epoch": 0.03636959370904325,
      "grad_norm": 0.12818795442581177,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 222
    },
    {
      "epoch": 0.03653342070773263,
      "grad_norm": 0.13201676309108734,
      "learning_rate": 0.0001,
      "loss": 1.4824,
      "step": 223
    },
    {
      "epoch": 0.03669724770642202,
      "grad_norm": 0.13088124990463257,
      "learning_rate": 0.0001,
      "loss": 1.5481,
      "step": 224
    },
    {
      "epoch": 0.0368610747051114,
      "grad_norm": 0.12919820845127106,
      "learning_rate": 0.0001,
      "loss": 1.5858,
      "step": 225
    },
    {
      "epoch": 0.03702490170380079,
      "grad_norm": 0.14287778735160828,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 226
    },
    {
      "epoch": 0.03718872870249017,
      "grad_norm": 0.1261826455593109,
      "learning_rate": 0.0001,
      "loss": 1.5966,
      "step": 227
    },
    {
      "epoch": 0.037352555701179554,
      "grad_norm": 0.1233975887298584,
      "learning_rate": 0.0001,
      "loss": 1.5369,
      "step": 228
    },
    {
      "epoch": 0.03751638269986894,
      "grad_norm": 0.13325338065624237,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 229
    },
    {
      "epoch": 0.03768020969855832,
      "grad_norm": 0.12640780210494995,
      "learning_rate": 0.0001,
      "loss": 1.6503,
      "step": 230
    },
    {
      "epoch": 0.03784403669724771,
      "grad_norm": 0.13321474194526672,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 231
    },
    {
      "epoch": 0.03800786369593709,
      "grad_norm": 0.1328231245279312,
      "learning_rate": 0.0001,
      "loss": 1.4851,
      "step": 232
    },
    {
      "epoch": 0.038171690694626476,
      "grad_norm": 0.12659676373004913,
      "learning_rate": 0.0001,
      "loss": 1.4732,
      "step": 233
    },
    {
      "epoch": 0.038335517693315856,
      "grad_norm": 0.13439978659152985,
      "learning_rate": 0.0001,
      "loss": 1.6974,
      "step": 234
    },
    {
      "epoch": 0.03849934469200524,
      "grad_norm": 0.12741895020008087,
      "learning_rate": 0.0001,
      "loss": 1.5094,
      "step": 235
    },
    {
      "epoch": 0.03866317169069463,
      "grad_norm": 0.13140830397605896,
      "learning_rate": 0.0001,
      "loss": 1.5519,
      "step": 236
    },
    {
      "epoch": 0.03882699868938401,
      "grad_norm": 0.1295180469751358,
      "learning_rate": 0.0001,
      "loss": 1.5378,
      "step": 237
    },
    {
      "epoch": 0.0389908256880734,
      "grad_norm": 0.137701615691185,
      "learning_rate": 0.0001,
      "loss": 1.5957,
      "step": 238
    },
    {
      "epoch": 0.03915465268676278,
      "grad_norm": 0.12642106413841248,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 239
    },
    {
      "epoch": 0.039318479685452164,
      "grad_norm": 0.1384691447019577,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 240
    },
    {
      "epoch": 0.039482306684141544,
      "grad_norm": 0.13016806542873383,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 241
    },
    {
      "epoch": 0.03964613368283093,
      "grad_norm": 0.1230437383055687,
      "learning_rate": 0.0001,
      "loss": 1.6033,
      "step": 242
    },
    {
      "epoch": 0.03980996068152032,
      "grad_norm": 0.13529936969280243,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 243
    },
    {
      "epoch": 0.0399737876802097,
      "grad_norm": 0.13021984696388245,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 244
    },
    {
      "epoch": 0.040137614678899085,
      "grad_norm": 0.1371053159236908,
      "learning_rate": 0.0001,
      "loss": 1.4904,
      "step": 245
    },
    {
      "epoch": 0.040301441677588465,
      "grad_norm": 0.12950442731380463,
      "learning_rate": 0.0001,
      "loss": 1.6471,
      "step": 246
    },
    {
      "epoch": 0.04046526867627785,
      "grad_norm": 0.14378178119659424,
      "learning_rate": 0.0001,
      "loss": 1.6185,
      "step": 247
    },
    {
      "epoch": 0.04062909567496723,
      "grad_norm": 0.11540716141462326,
      "learning_rate": 0.0001,
      "loss": 1.4228,
      "step": 248
    },
    {
      "epoch": 0.04079292267365662,
      "grad_norm": 0.1310529112815857,
      "learning_rate": 0.0001,
      "loss": 1.6337,
      "step": 249
    },
    {
      "epoch": 0.040956749672346006,
      "grad_norm": 0.12578359246253967,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 250
    },
    {
      "epoch": 0.041120576671035386,
      "grad_norm": 0.14310301840305328,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 251
    },
    {
      "epoch": 0.04128440366972477,
      "grad_norm": 0.13070790469646454,
      "learning_rate": 0.0001,
      "loss": 1.5399,
      "step": 252
    },
    {
      "epoch": 0.04144823066841415,
      "grad_norm": 0.12864753603935242,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 253
    },
    {
      "epoch": 0.04161205766710354,
      "grad_norm": 0.13220366835594177,
      "learning_rate": 0.0001,
      "loss": 1.529,
      "step": 254
    },
    {
      "epoch": 0.04177588466579292,
      "grad_norm": 0.146741583943367,
      "learning_rate": 0.0001,
      "loss": 1.5841,
      "step": 255
    },
    {
      "epoch": 0.04193971166448231,
      "grad_norm": 0.1276448518037796,
      "learning_rate": 0.0001,
      "loss": 1.4215,
      "step": 256
    },
    {
      "epoch": 0.042103538663171694,
      "grad_norm": 0.14377842843532562,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 257
    },
    {
      "epoch": 0.042267365661861074,
      "grad_norm": 0.12483279407024384,
      "learning_rate": 0.0001,
      "loss": 1.6706,
      "step": 258
    },
    {
      "epoch": 0.04243119266055046,
      "grad_norm": 0.14348362386226654,
      "learning_rate": 0.0001,
      "loss": 1.4323,
      "step": 259
    },
    {
      "epoch": 0.04259501965923984,
      "grad_norm": 0.13376536965370178,
      "learning_rate": 0.0001,
      "loss": 1.5975,
      "step": 260
    },
    {
      "epoch": 0.04275884665792923,
      "grad_norm": 0.12588359415531158,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 261
    },
    {
      "epoch": 0.04292267365661861,
      "grad_norm": 0.12889719009399414,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 262
    },
    {
      "epoch": 0.043086500655307995,
      "grad_norm": 0.1287308782339096,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 263
    },
    {
      "epoch": 0.04325032765399738,
      "grad_norm": 0.13218294084072113,
      "learning_rate": 0.0001,
      "loss": 1.5443,
      "step": 264
    },
    {
      "epoch": 0.04341415465268676,
      "grad_norm": 0.1268392950296402,
      "learning_rate": 0.0001,
      "loss": 1.5553,
      "step": 265
    },
    {
      "epoch": 0.04357798165137615,
      "grad_norm": 0.12143880873918533,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 266
    },
    {
      "epoch": 0.04374180865006553,
      "grad_norm": 0.12715524435043335,
      "learning_rate": 0.0001,
      "loss": 1.5088,
      "step": 267
    },
    {
      "epoch": 0.043905635648754916,
      "grad_norm": 0.13102728128433228,
      "learning_rate": 0.0001,
      "loss": 1.5073,
      "step": 268
    },
    {
      "epoch": 0.044069462647444296,
      "grad_norm": 0.12840379774570465,
      "learning_rate": 0.0001,
      "loss": 1.5304,
      "step": 269
    },
    {
      "epoch": 0.04423328964613368,
      "grad_norm": 0.12590573728084564,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 270
    },
    {
      "epoch": 0.04439711664482307,
      "grad_norm": 0.1382930874824524,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 271
    },
    {
      "epoch": 0.04456094364351245,
      "grad_norm": 0.132846400141716,
      "learning_rate": 0.0001,
      "loss": 1.5762,
      "step": 272
    },
    {
      "epoch": 0.04472477064220184,
      "grad_norm": 0.14085017144680023,
      "learning_rate": 0.0001,
      "loss": 1.5998,
      "step": 273
    },
    {
      "epoch": 0.04488859764089122,
      "grad_norm": 0.1291300654411316,
      "learning_rate": 0.0001,
      "loss": 1.5968,
      "step": 274
    },
    {
      "epoch": 0.045052424639580604,
      "grad_norm": 0.1458633691072464,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 275
    },
    {
      "epoch": 0.045216251638269984,
      "grad_norm": 0.12372390180826187,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 276
    },
    {
      "epoch": 0.04538007863695937,
      "grad_norm": 0.13282449543476105,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 277
    },
    {
      "epoch": 0.04554390563564875,
      "grad_norm": 0.1416933834552765,
      "learning_rate": 0.0001,
      "loss": 1.4657,
      "step": 278
    },
    {
      "epoch": 0.04570773263433814,
      "grad_norm": 0.27011579275131226,
      "learning_rate": 0.0001,
      "loss": 1.6193,
      "step": 279
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 0.1464989334344864,
      "learning_rate": 0.0001,
      "loss": 1.4919,
      "step": 280
    },
    {
      "epoch": 0.046035386631716906,
      "grad_norm": 0.13141606748104095,
      "learning_rate": 0.0001,
      "loss": 1.5525,
      "step": 281
    },
    {
      "epoch": 0.04619921363040629,
      "grad_norm": 0.138356015086174,
      "learning_rate": 0.0001,
      "loss": 1.4969,
      "step": 282
    },
    {
      "epoch": 0.04636304062909567,
      "grad_norm": 0.12877558171749115,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 283
    },
    {
      "epoch": 0.04652686762778506,
      "grad_norm": 0.13624034821987152,
      "learning_rate": 0.0001,
      "loss": 1.4335,
      "step": 284
    },
    {
      "epoch": 0.04669069462647444,
      "grad_norm": 0.12946444749832153,
      "learning_rate": 0.0001,
      "loss": 1.4487,
      "step": 285
    },
    {
      "epoch": 0.04685452162516383,
      "grad_norm": 0.12871332466602325,
      "learning_rate": 0.0001,
      "loss": 1.5652,
      "step": 286
    },
    {
      "epoch": 0.047018348623853214,
      "grad_norm": 0.1342865526676178,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 287
    },
    {
      "epoch": 0.047182175622542594,
      "grad_norm": 0.1324579119682312,
      "learning_rate": 0.0001,
      "loss": 1.5403,
      "step": 288
    },
    {
      "epoch": 0.04734600262123198,
      "grad_norm": 0.14403870701789856,
      "learning_rate": 0.0001,
      "loss": 1.5591,
      "step": 289
    },
    {
      "epoch": 0.04750982961992136,
      "grad_norm": 0.13270540535449982,
      "learning_rate": 0.0001,
      "loss": 1.6424,
      "step": 290
    },
    {
      "epoch": 0.04767365661861075,
      "grad_norm": 0.13297078013420105,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 291
    },
    {
      "epoch": 0.04783748361730013,
      "grad_norm": 0.12434036284685135,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 292
    },
    {
      "epoch": 0.048001310615989515,
      "grad_norm": 0.13441793620586395,
      "learning_rate": 0.0001,
      "loss": 1.5097,
      "step": 293
    },
    {
      "epoch": 0.0481651376146789,
      "grad_norm": 0.12939564883708954,
      "learning_rate": 0.0001,
      "loss": 1.5746,
      "step": 294
    },
    {
      "epoch": 0.04832896461336828,
      "grad_norm": 0.13639216125011444,
      "learning_rate": 0.0001,
      "loss": 1.5926,
      "step": 295
    },
    {
      "epoch": 0.04849279161205767,
      "grad_norm": 0.12338531762361526,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 296
    },
    {
      "epoch": 0.04865661861074705,
      "grad_norm": 0.12743960320949554,
      "learning_rate": 0.0001,
      "loss": 1.5825,
      "step": 297
    },
    {
      "epoch": 0.048820445609436436,
      "grad_norm": 0.1301930844783783,
      "learning_rate": 0.0001,
      "loss": 1.5629,
      "step": 298
    },
    {
      "epoch": 0.048984272608125816,
      "grad_norm": 0.1270844042301178,
      "learning_rate": 0.0001,
      "loss": 1.4851,
      "step": 299
    },
    {
      "epoch": 0.0491480996068152,
      "grad_norm": 0.14574366807937622,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 300
    },
    {
      "epoch": 0.04931192660550459,
      "grad_norm": 0.13358139991760254,
      "learning_rate": 0.0001,
      "loss": 1.5599,
      "step": 301
    },
    {
      "epoch": 0.04947575360419397,
      "grad_norm": 0.12385711818933487,
      "learning_rate": 0.0001,
      "loss": 1.4603,
      "step": 302
    },
    {
      "epoch": 0.04963958060288336,
      "grad_norm": 0.11963044852018356,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 303
    },
    {
      "epoch": 0.04980340760157274,
      "grad_norm": 0.12476995587348938,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 304
    },
    {
      "epoch": 0.049967234600262124,
      "grad_norm": 0.12620995938777924,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 305
    },
    {
      "epoch": 0.050131061598951504,
      "grad_norm": 0.12936405837535858,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 306
    },
    {
      "epoch": 0.05029488859764089,
      "grad_norm": 0.13023820519447327,
      "learning_rate": 0.0001,
      "loss": 1.63,
      "step": 307
    },
    {
      "epoch": 0.05045871559633028,
      "grad_norm": 0.12781022489070892,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 308
    },
    {
      "epoch": 0.05062254259501966,
      "grad_norm": 0.12420178204774857,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 309
    },
    {
      "epoch": 0.050786369593709045,
      "grad_norm": 0.12907648086547852,
      "learning_rate": 0.0001,
      "loss": 1.6488,
      "step": 310
    },
    {
      "epoch": 0.050950196592398425,
      "grad_norm": 0.11746347695589066,
      "learning_rate": 0.0001,
      "loss": 1.4459,
      "step": 311
    },
    {
      "epoch": 0.05111402359108781,
      "grad_norm": 0.1358799785375595,
      "learning_rate": 0.0001,
      "loss": 1.5526,
      "step": 312
    },
    {
      "epoch": 0.05127785058977719,
      "grad_norm": 0.1279253214597702,
      "learning_rate": 0.0001,
      "loss": 1.6247,
      "step": 313
    },
    {
      "epoch": 0.05144167758846658,
      "grad_norm": 0.12622061371803284,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 314
    },
    {
      "epoch": 0.051605504587155966,
      "grad_norm": 0.12897972762584686,
      "learning_rate": 0.0001,
      "loss": 1.5246,
      "step": 315
    },
    {
      "epoch": 0.051769331585845346,
      "grad_norm": 0.13107405602931976,
      "learning_rate": 0.0001,
      "loss": 1.5498,
      "step": 316
    },
    {
      "epoch": 0.05193315858453473,
      "grad_norm": 0.12708549201488495,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 317
    },
    {
      "epoch": 0.05209698558322411,
      "grad_norm": 0.1240776851773262,
      "learning_rate": 0.0001,
      "loss": 1.4711,
      "step": 318
    },
    {
      "epoch": 0.0522608125819135,
      "grad_norm": 0.12508228421211243,
      "learning_rate": 0.0001,
      "loss": 1.5003,
      "step": 319
    },
    {
      "epoch": 0.05242463958060288,
      "grad_norm": 0.12555113434791565,
      "learning_rate": 0.0001,
      "loss": 1.4059,
      "step": 320
    },
    {
      "epoch": 0.05258846657929227,
      "grad_norm": 0.12630921602249146,
      "learning_rate": 0.0001,
      "loss": 1.6321,
      "step": 321
    },
    {
      "epoch": 0.052752293577981654,
      "grad_norm": 0.13323728740215302,
      "learning_rate": 0.0001,
      "loss": 1.552,
      "step": 322
    },
    {
      "epoch": 0.052916120576671034,
      "grad_norm": 0.12522096931934357,
      "learning_rate": 0.0001,
      "loss": 1.5048,
      "step": 323
    },
    {
      "epoch": 0.05307994757536042,
      "grad_norm": 0.13109713792800903,
      "learning_rate": 0.0001,
      "loss": 1.5941,
      "step": 324
    },
    {
      "epoch": 0.0532437745740498,
      "grad_norm": 0.11887567490339279,
      "learning_rate": 0.0001,
      "loss": 1.4313,
      "step": 325
    },
    {
      "epoch": 0.05340760157273919,
      "grad_norm": 0.1254127472639084,
      "learning_rate": 0.0001,
      "loss": 1.486,
      "step": 326
    },
    {
      "epoch": 0.05357142857142857,
      "grad_norm": 0.12283234298229218,
      "learning_rate": 0.0001,
      "loss": 1.5422,
      "step": 327
    },
    {
      "epoch": 0.053735255570117955,
      "grad_norm": 0.1256858855485916,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 328
    },
    {
      "epoch": 0.05389908256880734,
      "grad_norm": 0.1313122808933258,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 329
    },
    {
      "epoch": 0.05406290956749672,
      "grad_norm": 0.12590482831001282,
      "learning_rate": 0.0001,
      "loss": 1.6124,
      "step": 330
    },
    {
      "epoch": 0.05422673656618611,
      "grad_norm": 0.13533324003219604,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 331
    },
    {
      "epoch": 0.05439056356487549,
      "grad_norm": 0.12202730029821396,
      "learning_rate": 0.0001,
      "loss": 1.4703,
      "step": 332
    },
    {
      "epoch": 0.05455439056356488,
      "grad_norm": 0.13683480024337769,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 333
    },
    {
      "epoch": 0.05471821756225426,
      "grad_norm": 0.12281996011734009,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 334
    },
    {
      "epoch": 0.054882044560943644,
      "grad_norm": 0.13574418425559998,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 335
    },
    {
      "epoch": 0.05504587155963303,
      "grad_norm": 0.12132158130407333,
      "learning_rate": 0.0001,
      "loss": 1.4477,
      "step": 336
    },
    {
      "epoch": 0.05520969855832241,
      "grad_norm": 0.11583832651376724,
      "learning_rate": 0.0001,
      "loss": 1.5914,
      "step": 337
    },
    {
      "epoch": 0.0553735255570118,
      "grad_norm": 0.1344902664422989,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 338
    },
    {
      "epoch": 0.05553735255570118,
      "grad_norm": 0.13589662313461304,
      "learning_rate": 0.0001,
      "loss": 1.5994,
      "step": 339
    },
    {
      "epoch": 0.055701179554390565,
      "grad_norm": 0.13395152986049652,
      "learning_rate": 0.0001,
      "loss": 1.4322,
      "step": 340
    },
    {
      "epoch": 0.055865006553079945,
      "grad_norm": 0.13249383866786957,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 341
    },
    {
      "epoch": 0.05602883355176933,
      "grad_norm": 0.12041502445936203,
      "learning_rate": 0.0001,
      "loss": 1.4047,
      "step": 342
    },
    {
      "epoch": 0.05619266055045872,
      "grad_norm": 0.1256500482559204,
      "learning_rate": 0.0001,
      "loss": 1.4798,
      "step": 343
    },
    {
      "epoch": 0.0563564875491481,
      "grad_norm": 0.124946229159832,
      "learning_rate": 0.0001,
      "loss": 1.4943,
      "step": 344
    },
    {
      "epoch": 0.056520314547837486,
      "grad_norm": 0.12998640537261963,
      "learning_rate": 0.0001,
      "loss": 1.556,
      "step": 345
    },
    {
      "epoch": 0.056684141546526866,
      "grad_norm": 0.12739400565624237,
      "learning_rate": 0.0001,
      "loss": 1.5846,
      "step": 346
    },
    {
      "epoch": 0.05684796854521625,
      "grad_norm": 0.12163793295621872,
      "learning_rate": 0.0001,
      "loss": 1.501,
      "step": 347
    },
    {
      "epoch": 0.05701179554390563,
      "grad_norm": 0.13316822052001953,
      "learning_rate": 0.0001,
      "loss": 1.5408,
      "step": 348
    },
    {
      "epoch": 0.05717562254259502,
      "grad_norm": 0.12592287361621857,
      "learning_rate": 0.0001,
      "loss": 1.5247,
      "step": 349
    },
    {
      "epoch": 0.05733944954128441,
      "grad_norm": 0.1332852840423584,
      "learning_rate": 0.0001,
      "loss": 1.4538,
      "step": 350
    },
    {
      "epoch": 0.05750327653997379,
      "grad_norm": 0.12046833336353302,
      "learning_rate": 0.0001,
      "loss": 1.476,
      "step": 351
    },
    {
      "epoch": 0.057667103538663174,
      "grad_norm": 0.13134358823299408,
      "learning_rate": 0.0001,
      "loss": 1.4384,
      "step": 352
    },
    {
      "epoch": 0.057830930537352554,
      "grad_norm": 0.13023953139781952,
      "learning_rate": 0.0001,
      "loss": 1.487,
      "step": 353
    },
    {
      "epoch": 0.05799475753604194,
      "grad_norm": 0.12965761125087738,
      "learning_rate": 0.0001,
      "loss": 1.484,
      "step": 354
    },
    {
      "epoch": 0.05815858453473132,
      "grad_norm": 0.12695816159248352,
      "learning_rate": 0.0001,
      "loss": 1.3359,
      "step": 355
    },
    {
      "epoch": 0.05832241153342071,
      "grad_norm": 0.1261056661605835,
      "learning_rate": 0.0001,
      "loss": 1.4521,
      "step": 356
    },
    {
      "epoch": 0.058486238532110095,
      "grad_norm": 0.12789607048034668,
      "learning_rate": 0.0001,
      "loss": 1.5449,
      "step": 357
    },
    {
      "epoch": 0.058650065530799475,
      "grad_norm": 0.12544968724250793,
      "learning_rate": 0.0001,
      "loss": 1.575,
      "step": 358
    },
    {
      "epoch": 0.05881389252948886,
      "grad_norm": 0.12247415632009506,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 359
    },
    {
      "epoch": 0.05897771952817824,
      "grad_norm": 0.12501631677150726,
      "learning_rate": 0.0001,
      "loss": 1.4995,
      "step": 360
    },
    {
      "epoch": 0.05914154652686763,
      "grad_norm": 0.13352151215076447,
      "learning_rate": 0.0001,
      "loss": 1.4006,
      "step": 361
    },
    {
      "epoch": 0.05930537352555701,
      "grad_norm": 0.1242838203907013,
      "learning_rate": 0.0001,
      "loss": 1.56,
      "step": 362
    },
    {
      "epoch": 0.059469200524246396,
      "grad_norm": 0.1233491525053978,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 363
    },
    {
      "epoch": 0.05963302752293578,
      "grad_norm": 0.1260528266429901,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 364
    },
    {
      "epoch": 0.05979685452162516,
      "grad_norm": 0.1240929439663887,
      "learning_rate": 0.0001,
      "loss": 1.5176,
      "step": 365
    },
    {
      "epoch": 0.05996068152031455,
      "grad_norm": 0.13003942370414734,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 366
    },
    {
      "epoch": 0.06012450851900393,
      "grad_norm": 0.12103753536939621,
      "learning_rate": 0.0001,
      "loss": 1.4288,
      "step": 367
    },
    {
      "epoch": 0.06028833551769332,
      "grad_norm": 0.1274299919605255,
      "learning_rate": 0.0001,
      "loss": 1.4751,
      "step": 368
    },
    {
      "epoch": 0.0604521625163827,
      "grad_norm": 0.13269151747226715,
      "learning_rate": 0.0001,
      "loss": 1.5963,
      "step": 369
    },
    {
      "epoch": 0.060615989515072084,
      "grad_norm": 0.13308580219745636,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 370
    },
    {
      "epoch": 0.06077981651376147,
      "grad_norm": 0.1303136795759201,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 371
    },
    {
      "epoch": 0.06094364351245085,
      "grad_norm": 0.12432117015123367,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 372
    },
    {
      "epoch": 0.06110747051114024,
      "grad_norm": 0.1350603997707367,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 373
    },
    {
      "epoch": 0.06127129750982962,
      "grad_norm": 0.1269674301147461,
      "learning_rate": 0.0001,
      "loss": 1.6526,
      "step": 374
    },
    {
      "epoch": 0.061435124508519005,
      "grad_norm": 0.14046020805835724,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 375
    },
    {
      "epoch": 0.061598951507208385,
      "grad_norm": 0.12981869280338287,
      "learning_rate": 0.0001,
      "loss": 1.4626,
      "step": 376
    },
    {
      "epoch": 0.06176277850589777,
      "grad_norm": 0.12370241433382034,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 377
    },
    {
      "epoch": 0.06192660550458716,
      "grad_norm": 0.12999741733074188,
      "learning_rate": 0.0001,
      "loss": 1.5382,
      "step": 378
    },
    {
      "epoch": 0.06209043250327654,
      "grad_norm": 0.12526722252368927,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 379
    },
    {
      "epoch": 0.062254259501965926,
      "grad_norm": 0.12394998967647552,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 380
    },
    {
      "epoch": 0.062418086500655307,
      "grad_norm": 0.12774012982845306,
      "learning_rate": 0.0001,
      "loss": 1.504,
      "step": 381
    },
    {
      "epoch": 0.0625819134993447,
      "grad_norm": 0.1275997906923294,
      "learning_rate": 0.0001,
      "loss": 1.4786,
      "step": 382
    },
    {
      "epoch": 0.06274574049803408,
      "grad_norm": 0.1321539282798767,
      "learning_rate": 0.0001,
      "loss": 1.5128,
      "step": 383
    },
    {
      "epoch": 0.06290956749672345,
      "grad_norm": 0.123343825340271,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 384
    },
    {
      "epoch": 0.06307339449541284,
      "grad_norm": 0.12264310568571091,
      "learning_rate": 0.0001,
      "loss": 1.5865,
      "step": 385
    },
    {
      "epoch": 0.06323722149410223,
      "grad_norm": 0.1272432804107666,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 386
    },
    {
      "epoch": 0.06340104849279161,
      "grad_norm": 0.12889088690280914,
      "learning_rate": 0.0001,
      "loss": 1.5877,
      "step": 387
    },
    {
      "epoch": 0.063564875491481,
      "grad_norm": 0.1353144347667694,
      "learning_rate": 0.0001,
      "loss": 1.3997,
      "step": 388
    },
    {
      "epoch": 0.06372870249017037,
      "grad_norm": 0.12980051338672638,
      "learning_rate": 0.0001,
      "loss": 1.4331,
      "step": 389
    },
    {
      "epoch": 0.06389252948885976,
      "grad_norm": 0.1314554214477539,
      "learning_rate": 0.0001,
      "loss": 1.4958,
      "step": 390
    },
    {
      "epoch": 0.06405635648754915,
      "grad_norm": 0.13064256310462952,
      "learning_rate": 0.0001,
      "loss": 1.4101,
      "step": 391
    },
    {
      "epoch": 0.06422018348623854,
      "grad_norm": 0.1208432987332344,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 392
    },
    {
      "epoch": 0.06438401048492792,
      "grad_norm": 0.13642841577529907,
      "learning_rate": 0.0001,
      "loss": 1.496,
      "step": 393
    },
    {
      "epoch": 0.0645478374836173,
      "grad_norm": 0.1234004944562912,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 394
    },
    {
      "epoch": 0.06471166448230668,
      "grad_norm": 0.11878219246864319,
      "learning_rate": 0.0001,
      "loss": 1.483,
      "step": 395
    },
    {
      "epoch": 0.06487549148099607,
      "grad_norm": 0.13111650943756104,
      "learning_rate": 0.0001,
      "loss": 1.4455,
      "step": 396
    },
    {
      "epoch": 0.06503931847968546,
      "grad_norm": 0.132078155875206,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 397
    },
    {
      "epoch": 0.06520314547837483,
      "grad_norm": 0.16390706598758698,
      "learning_rate": 0.0001,
      "loss": 1.5262,
      "step": 398
    },
    {
      "epoch": 0.06536697247706422,
      "grad_norm": 0.12536875903606415,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 399
    },
    {
      "epoch": 0.0655307994757536,
      "grad_norm": 0.13276083767414093,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 400
    },
    {
      "epoch": 0.06569462647444299,
      "grad_norm": 0.12888388335704803,
      "learning_rate": 0.0001,
      "loss": 1.4404,
      "step": 401
    },
    {
      "epoch": 0.06585845347313238,
      "grad_norm": 0.14632686972618103,
      "learning_rate": 0.0001,
      "loss": 1.5609,
      "step": 402
    },
    {
      "epoch": 0.06602228047182175,
      "grad_norm": 0.13215306401252747,
      "learning_rate": 0.0001,
      "loss": 1.4955,
      "step": 403
    },
    {
      "epoch": 0.06618610747051114,
      "grad_norm": 0.1260053962469101,
      "learning_rate": 0.0001,
      "loss": 1.5119,
      "step": 404
    },
    {
      "epoch": 0.06634993446920052,
      "grad_norm": 0.12198645621538162,
      "learning_rate": 0.0001,
      "loss": 1.5604,
      "step": 405
    },
    {
      "epoch": 0.06651376146788991,
      "grad_norm": 0.14916189014911652,
      "learning_rate": 0.0001,
      "loss": 1.3984,
      "step": 406
    },
    {
      "epoch": 0.0666775884665793,
      "grad_norm": 0.12111485004425049,
      "learning_rate": 0.0001,
      "loss": 1.3488,
      "step": 407
    },
    {
      "epoch": 0.06684141546526867,
      "grad_norm": 0.12623156607151031,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 408
    },
    {
      "epoch": 0.06700524246395806,
      "grad_norm": 0.11741374433040619,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 409
    },
    {
      "epoch": 0.06716906946264745,
      "grad_norm": 0.1475224494934082,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 410
    },
    {
      "epoch": 0.06733289646133683,
      "grad_norm": 0.12895812094211578,
      "learning_rate": 0.0001,
      "loss": 1.4604,
      "step": 411
    },
    {
      "epoch": 0.0674967234600262,
      "grad_norm": 0.1348087191581726,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 412
    },
    {
      "epoch": 0.0676605504587156,
      "grad_norm": 0.1191665381193161,
      "learning_rate": 0.0001,
      "loss": 1.4763,
      "step": 413
    },
    {
      "epoch": 0.06782437745740498,
      "grad_norm": 0.13580629229545593,
      "learning_rate": 0.0001,
      "loss": 1.4251,
      "step": 414
    },
    {
      "epoch": 0.06798820445609437,
      "grad_norm": 0.12614792585372925,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 415
    },
    {
      "epoch": 0.06815203145478375,
      "grad_norm": 0.13216400146484375,
      "learning_rate": 0.0001,
      "loss": 1.5681,
      "step": 416
    },
    {
      "epoch": 0.06831585845347313,
      "grad_norm": 0.12403993308544159,
      "learning_rate": 0.0001,
      "loss": 1.5248,
      "step": 417
    },
    {
      "epoch": 0.06847968545216251,
      "grad_norm": 0.12538757920265198,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 418
    },
    {
      "epoch": 0.0686435124508519,
      "grad_norm": 0.13892193138599396,
      "learning_rate": 0.0001,
      "loss": 1.5317,
      "step": 419
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 0.12235568463802338,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 420
    },
    {
      "epoch": 0.06897116644823068,
      "grad_norm": 0.13059476017951965,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 421
    },
    {
      "epoch": 0.06913499344692005,
      "grad_norm": 0.13674752414226532,
      "learning_rate": 0.0001,
      "loss": 1.4694,
      "step": 422
    },
    {
      "epoch": 0.06929882044560944,
      "grad_norm": 0.12781813740730286,
      "learning_rate": 0.0001,
      "loss": 1.5908,
      "step": 423
    },
    {
      "epoch": 0.06946264744429882,
      "grad_norm": 0.13811762630939484,
      "learning_rate": 0.0001,
      "loss": 1.4026,
      "step": 424
    },
    {
      "epoch": 0.06962647444298821,
      "grad_norm": 0.1292458027601242,
      "learning_rate": 0.0001,
      "loss": 1.6433,
      "step": 425
    },
    {
      "epoch": 0.06979030144167758,
      "grad_norm": 0.14235101640224457,
      "learning_rate": 0.0001,
      "loss": 1.5836,
      "step": 426
    },
    {
      "epoch": 0.06995412844036697,
      "grad_norm": 0.12941884994506836,
      "learning_rate": 0.0001,
      "loss": 1.4691,
      "step": 427
    },
    {
      "epoch": 0.07011795543905636,
      "grad_norm": 0.13243472576141357,
      "learning_rate": 0.0001,
      "loss": 1.5644,
      "step": 428
    },
    {
      "epoch": 0.07028178243774574,
      "grad_norm": 0.1248096227645874,
      "learning_rate": 0.0001,
      "loss": 1.6077,
      "step": 429
    },
    {
      "epoch": 0.07044560943643513,
      "grad_norm": 0.12996190786361694,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 430
    },
    {
      "epoch": 0.0706094364351245,
      "grad_norm": 0.13066492974758148,
      "learning_rate": 0.0001,
      "loss": 1.5334,
      "step": 431
    },
    {
      "epoch": 0.07077326343381389,
      "grad_norm": 0.15567709505558014,
      "learning_rate": 0.0001,
      "loss": 1.3533,
      "step": 432
    },
    {
      "epoch": 0.07093709043250328,
      "grad_norm": 0.12287536263465881,
      "learning_rate": 0.0001,
      "loss": 1.4279,
      "step": 433
    },
    {
      "epoch": 0.07110091743119266,
      "grad_norm": 0.13063016533851624,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 434
    },
    {
      "epoch": 0.07126474442988205,
      "grad_norm": 0.1260465681552887,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 435
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 0.13714338839054108,
      "learning_rate": 0.0001,
      "loss": 1.6151,
      "step": 436
    },
    {
      "epoch": 0.07159239842726081,
      "grad_norm": 0.13997219502925873,
      "learning_rate": 0.0001,
      "loss": 1.5689,
      "step": 437
    },
    {
      "epoch": 0.0717562254259502,
      "grad_norm": 0.12384165078401566,
      "learning_rate": 0.0001,
      "loss": 1.495,
      "step": 438
    },
    {
      "epoch": 0.07192005242463959,
      "grad_norm": 0.1350708305835724,
      "learning_rate": 0.0001,
      "loss": 1.6009,
      "step": 439
    },
    {
      "epoch": 0.07208387942332896,
      "grad_norm": 0.1288231611251831,
      "learning_rate": 0.0001,
      "loss": 1.4216,
      "step": 440
    },
    {
      "epoch": 0.07224770642201835,
      "grad_norm": 0.13040326535701752,
      "learning_rate": 0.0001,
      "loss": 1.5065,
      "step": 441
    },
    {
      "epoch": 0.07241153342070773,
      "grad_norm": 0.12293791770935059,
      "learning_rate": 0.0001,
      "loss": 1.5192,
      "step": 442
    },
    {
      "epoch": 0.07257536041939712,
      "grad_norm": 0.13307669758796692,
      "learning_rate": 0.0001,
      "loss": 1.5356,
      "step": 443
    },
    {
      "epoch": 0.0727391874180865,
      "grad_norm": 0.12706497311592102,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 444
    },
    {
      "epoch": 0.07290301441677588,
      "grad_norm": 0.12654677033424377,
      "learning_rate": 0.0001,
      "loss": 1.4968,
      "step": 445
    },
    {
      "epoch": 0.07306684141546527,
      "grad_norm": 0.1276216208934784,
      "learning_rate": 0.0001,
      "loss": 1.4886,
      "step": 446
    },
    {
      "epoch": 0.07323066841415465,
      "grad_norm": 0.13371863961219788,
      "learning_rate": 0.0001,
      "loss": 1.5419,
      "step": 447
    },
    {
      "epoch": 0.07339449541284404,
      "grad_norm": 0.1271762251853943,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 448
    },
    {
      "epoch": 0.07355832241153341,
      "grad_norm": 0.1183372437953949,
      "learning_rate": 0.0001,
      "loss": 1.3492,
      "step": 449
    },
    {
      "epoch": 0.0737221494102228,
      "grad_norm": 0.1318168044090271,
      "learning_rate": 0.0001,
      "loss": 1.5524,
      "step": 450
    },
    {
      "epoch": 0.07388597640891219,
      "grad_norm": 0.1291465312242508,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 451
    },
    {
      "epoch": 0.07404980340760157,
      "grad_norm": 0.12119657546281815,
      "learning_rate": 0.0001,
      "loss": 1.5095,
      "step": 452
    },
    {
      "epoch": 0.07421363040629096,
      "grad_norm": 0.12941871583461761,
      "learning_rate": 0.0001,
      "loss": 1.3606,
      "step": 453
    },
    {
      "epoch": 0.07437745740498033,
      "grad_norm": 0.12342154234647751,
      "learning_rate": 0.0001,
      "loss": 1.5716,
      "step": 454
    },
    {
      "epoch": 0.07454128440366972,
      "grad_norm": 0.1351212114095688,
      "learning_rate": 0.0001,
      "loss": 1.4705,
      "step": 455
    },
    {
      "epoch": 0.07470511140235911,
      "grad_norm": 0.12705810368061066,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 456
    },
    {
      "epoch": 0.0748689384010485,
      "grad_norm": 0.13370715081691742,
      "learning_rate": 0.0001,
      "loss": 1.4872,
      "step": 457
    },
    {
      "epoch": 0.07503276539973788,
      "grad_norm": 0.1277092546224594,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 458
    },
    {
      "epoch": 0.07519659239842726,
      "grad_norm": 0.12582367658615112,
      "learning_rate": 0.0001,
      "loss": 1.437,
      "step": 459
    },
    {
      "epoch": 0.07536041939711664,
      "grad_norm": 0.14314083755016327,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 460
    },
    {
      "epoch": 0.07552424639580603,
      "grad_norm": 0.13048389554023743,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 461
    },
    {
      "epoch": 0.07568807339449542,
      "grad_norm": 0.13759669661521912,
      "learning_rate": 0.0001,
      "loss": 1.6332,
      "step": 462
    },
    {
      "epoch": 0.07585190039318479,
      "grad_norm": 0.12615306675434113,
      "learning_rate": 0.0001,
      "loss": 1.3955,
      "step": 463
    },
    {
      "epoch": 0.07601572739187418,
      "grad_norm": 0.13040898740291595,
      "learning_rate": 0.0001,
      "loss": 1.4901,
      "step": 464
    },
    {
      "epoch": 0.07617955439056356,
      "grad_norm": 0.13577300310134888,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 465
    },
    {
      "epoch": 0.07634338138925295,
      "grad_norm": 0.1453324854373932,
      "learning_rate": 0.0001,
      "loss": 1.4204,
      "step": 466
    },
    {
      "epoch": 0.07650720838794234,
      "grad_norm": 0.1309783160686493,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 467
    },
    {
      "epoch": 0.07667103538663171,
      "grad_norm": 0.13509559631347656,
      "learning_rate": 0.0001,
      "loss": 1.3766,
      "step": 468
    },
    {
      "epoch": 0.0768348623853211,
      "grad_norm": 0.13670599460601807,
      "learning_rate": 0.0001,
      "loss": 1.5007,
      "step": 469
    },
    {
      "epoch": 0.07699868938401049,
      "grad_norm": 0.13084541261196136,
      "learning_rate": 0.0001,
      "loss": 1.3875,
      "step": 470
    },
    {
      "epoch": 0.07716251638269987,
      "grad_norm": 0.13244254887104034,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 471
    },
    {
      "epoch": 0.07732634338138926,
      "grad_norm": 0.13520269095897675,
      "learning_rate": 0.0001,
      "loss": 1.5128,
      "step": 472
    },
    {
      "epoch": 0.07749017038007863,
      "grad_norm": 0.13036954402923584,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 473
    },
    {
      "epoch": 0.07765399737876802,
      "grad_norm": 0.1432826817035675,
      "learning_rate": 0.0001,
      "loss": 1.4477,
      "step": 474
    },
    {
      "epoch": 0.0778178243774574,
      "grad_norm": 0.12765368819236755,
      "learning_rate": 0.0001,
      "loss": 1.581,
      "step": 475
    },
    {
      "epoch": 0.0779816513761468,
      "grad_norm": 0.14131535589694977,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 476
    },
    {
      "epoch": 0.07814547837483617,
      "grad_norm": 0.1357503980398178,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 477
    },
    {
      "epoch": 0.07830930537352555,
      "grad_norm": 0.13135850429534912,
      "learning_rate": 0.0001,
      "loss": 1.5958,
      "step": 478
    },
    {
      "epoch": 0.07847313237221494,
      "grad_norm": 0.13799071311950684,
      "learning_rate": 0.0001,
      "loss": 1.5866,
      "step": 479
    },
    {
      "epoch": 0.07863695937090433,
      "grad_norm": 0.12825989723205566,
      "learning_rate": 0.0001,
      "loss": 1.5579,
      "step": 480
    },
    {
      "epoch": 0.07880078636959371,
      "grad_norm": 0.12298025190830231,
      "learning_rate": 0.0001,
      "loss": 1.3793,
      "step": 481
    },
    {
      "epoch": 0.07896461336828309,
      "grad_norm": 0.15233029425144196,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 482
    },
    {
      "epoch": 0.07912844036697247,
      "grad_norm": 0.1285896599292755,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 483
    },
    {
      "epoch": 0.07929226736566186,
      "grad_norm": 0.13491550087928772,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 484
    },
    {
      "epoch": 0.07945609436435125,
      "grad_norm": 0.12800095975399017,
      "learning_rate": 0.0001,
      "loss": 1.5514,
      "step": 485
    },
    {
      "epoch": 0.07961992136304064,
      "grad_norm": 0.12830448150634766,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 486
    },
    {
      "epoch": 0.07978374836173001,
      "grad_norm": 0.1268654763698578,
      "learning_rate": 0.0001,
      "loss": 1.3371,
      "step": 487
    },
    {
      "epoch": 0.0799475753604194,
      "grad_norm": 0.12257809937000275,
      "learning_rate": 0.0001,
      "loss": 1.4651,
      "step": 488
    },
    {
      "epoch": 0.08011140235910878,
      "grad_norm": 0.14243540167808533,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 489
    },
    {
      "epoch": 0.08027522935779817,
      "grad_norm": 0.14031371474266052,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 490
    },
    {
      "epoch": 0.08043905635648754,
      "grad_norm": 0.12618736922740936,
      "learning_rate": 0.0001,
      "loss": 1.5658,
      "step": 491
    },
    {
      "epoch": 0.08060288335517693,
      "grad_norm": 0.1309361755847931,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 492
    },
    {
      "epoch": 0.08076671035386632,
      "grad_norm": 0.12347295135259628,
      "learning_rate": 0.0001,
      "loss": 1.4248,
      "step": 493
    },
    {
      "epoch": 0.0809305373525557,
      "grad_norm": 0.13087432086467743,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 494
    },
    {
      "epoch": 0.08109436435124509,
      "grad_norm": 0.1344584822654724,
      "learning_rate": 0.0001,
      "loss": 1.568,
      "step": 495
    },
    {
      "epoch": 0.08125819134993446,
      "grad_norm": 0.12378037720918655,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 496
    },
    {
      "epoch": 0.08142201834862385,
      "grad_norm": 0.11846324056386948,
      "learning_rate": 0.0001,
      "loss": 1.4089,
      "step": 497
    },
    {
      "epoch": 0.08158584534731324,
      "grad_norm": 0.14966660737991333,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 498
    },
    {
      "epoch": 0.08174967234600262,
      "grad_norm": 0.13066788017749786,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 499
    },
    {
      "epoch": 0.08191349934469201,
      "grad_norm": 0.12721063196659088,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 500
    },
    {
      "epoch": 0.08207732634338138,
      "grad_norm": 0.12411727011203766,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 501
    },
    {
      "epoch": 0.08224115334207077,
      "grad_norm": 0.12929244339466095,
      "learning_rate": 0.0001,
      "loss": 1.5134,
      "step": 502
    },
    {
      "epoch": 0.08240498034076016,
      "grad_norm": 0.13284482061862946,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 503
    },
    {
      "epoch": 0.08256880733944955,
      "grad_norm": 0.12866252660751343,
      "learning_rate": 0.0001,
      "loss": 1.3575,
      "step": 504
    },
    {
      "epoch": 0.08273263433813892,
      "grad_norm": 0.12941104173660278,
      "learning_rate": 0.0001,
      "loss": 1.6105,
      "step": 505
    },
    {
      "epoch": 0.0828964613368283,
      "grad_norm": 0.12711840867996216,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 506
    },
    {
      "epoch": 0.08306028833551769,
      "grad_norm": 0.11537956446409225,
      "learning_rate": 0.0001,
      "loss": 1.4302,
      "step": 507
    },
    {
      "epoch": 0.08322411533420708,
      "grad_norm": 0.11822990328073502,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 508
    },
    {
      "epoch": 0.08338794233289647,
      "grad_norm": 0.13282987475395203,
      "learning_rate": 0.0001,
      "loss": 1.5638,
      "step": 509
    },
    {
      "epoch": 0.08355176933158584,
      "grad_norm": 0.13540470600128174,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 510
    },
    {
      "epoch": 0.08371559633027523,
      "grad_norm": 0.13063937425613403,
      "learning_rate": 0.0001,
      "loss": 1.5249,
      "step": 511
    },
    {
      "epoch": 0.08387942332896461,
      "grad_norm": 0.1296718865633011,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 512
    },
    {
      "epoch": 0.084043250327654,
      "grad_norm": 0.12861867249011993,
      "learning_rate": 0.0001,
      "loss": 1.4767,
      "step": 513
    },
    {
      "epoch": 0.08420707732634339,
      "grad_norm": 0.12838983535766602,
      "learning_rate": 0.0001,
      "loss": 1.3857,
      "step": 514
    },
    {
      "epoch": 0.08437090432503276,
      "grad_norm": 0.13112609088420868,
      "learning_rate": 0.0001,
      "loss": 1.4075,
      "step": 515
    },
    {
      "epoch": 0.08453473132372215,
      "grad_norm": 0.12255129218101501,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 516
    },
    {
      "epoch": 0.08469855832241154,
      "grad_norm": 0.12763561308383942,
      "learning_rate": 0.0001,
      "loss": 1.515,
      "step": 517
    },
    {
      "epoch": 0.08486238532110092,
      "grad_norm": 0.1200021505355835,
      "learning_rate": 0.0001,
      "loss": 1.3781,
      "step": 518
    },
    {
      "epoch": 0.0850262123197903,
      "grad_norm": 0.13018867373466492,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 519
    },
    {
      "epoch": 0.08519003931847968,
      "grad_norm": 0.128676638007164,
      "learning_rate": 0.0001,
      "loss": 1.5516,
      "step": 520
    },
    {
      "epoch": 0.08535386631716907,
      "grad_norm": 0.13627661764621735,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 521
    },
    {
      "epoch": 0.08551769331585846,
      "grad_norm": 0.1386180967092514,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 522
    },
    {
      "epoch": 0.08568152031454784,
      "grad_norm": 0.12752613425254822,
      "learning_rate": 0.0001,
      "loss": 1.4231,
      "step": 523
    },
    {
      "epoch": 0.08584534731323722,
      "grad_norm": 0.12926934659481049,
      "learning_rate": 0.0001,
      "loss": 1.4819,
      "step": 524
    },
    {
      "epoch": 0.0860091743119266,
      "grad_norm": 0.12614765763282776,
      "learning_rate": 0.0001,
      "loss": 1.5547,
      "step": 525
    },
    {
      "epoch": 0.08617300131061599,
      "grad_norm": 0.13577258586883545,
      "learning_rate": 0.0001,
      "loss": 1.5285,
      "step": 526
    },
    {
      "epoch": 0.08633682830930538,
      "grad_norm": 0.13787762820720673,
      "learning_rate": 0.0001,
      "loss": 1.5744,
      "step": 527
    },
    {
      "epoch": 0.08650065530799476,
      "grad_norm": 0.12209629267454147,
      "learning_rate": 0.0001,
      "loss": 1.46,
      "step": 528
    },
    {
      "epoch": 0.08666448230668414,
      "grad_norm": 0.123148612678051,
      "learning_rate": 0.0001,
      "loss": 1.4668,
      "step": 529
    },
    {
      "epoch": 0.08682830930537352,
      "grad_norm": 0.1355384737253189,
      "learning_rate": 0.0001,
      "loss": 1.6057,
      "step": 530
    },
    {
      "epoch": 0.08699213630406291,
      "grad_norm": 0.12739501893520355,
      "learning_rate": 0.0001,
      "loss": 1.4657,
      "step": 531
    },
    {
      "epoch": 0.0871559633027523,
      "grad_norm": 0.1396438181400299,
      "learning_rate": 0.0001,
      "loss": 1.4894,
      "step": 532
    },
    {
      "epoch": 0.08731979030144167,
      "grad_norm": 0.12382902950048447,
      "learning_rate": 0.0001,
      "loss": 1.4257,
      "step": 533
    },
    {
      "epoch": 0.08748361730013106,
      "grad_norm": 0.1240193247795105,
      "learning_rate": 0.0001,
      "loss": 1.4322,
      "step": 534
    },
    {
      "epoch": 0.08764744429882045,
      "grad_norm": 0.12545058131217957,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 535
    },
    {
      "epoch": 0.08781127129750983,
      "grad_norm": 0.13388679921627045,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 536
    },
    {
      "epoch": 0.08797509829619922,
      "grad_norm": 0.1312274932861328,
      "learning_rate": 0.0001,
      "loss": 1.4975,
      "step": 537
    },
    {
      "epoch": 0.08813892529488859,
      "grad_norm": 0.1342012882232666,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 538
    },
    {
      "epoch": 0.08830275229357798,
      "grad_norm": 0.135207399725914,
      "learning_rate": 0.0001,
      "loss": 1.4805,
      "step": 539
    },
    {
      "epoch": 0.08846657929226737,
      "grad_norm": 0.1313864141702652,
      "learning_rate": 0.0001,
      "loss": 1.4644,
      "step": 540
    },
    {
      "epoch": 0.08863040629095675,
      "grad_norm": 0.12351572513580322,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 541
    },
    {
      "epoch": 0.08879423328964614,
      "grad_norm": 0.12391596287488937,
      "learning_rate": 0.0001,
      "loss": 1.535,
      "step": 542
    },
    {
      "epoch": 0.08895806028833551,
      "grad_norm": 0.12310027331113815,
      "learning_rate": 0.0001,
      "loss": 1.3704,
      "step": 543
    },
    {
      "epoch": 0.0891218872870249,
      "grad_norm": 0.13538172841072083,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 544
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 0.11908945441246033,
      "learning_rate": 0.0001,
      "loss": 1.4053,
      "step": 545
    },
    {
      "epoch": 0.08944954128440367,
      "grad_norm": 0.1210799366235733,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 546
    },
    {
      "epoch": 0.08961336828309305,
      "grad_norm": 0.13151605427265167,
      "learning_rate": 0.0001,
      "loss": 1.4598,
      "step": 547
    },
    {
      "epoch": 0.08977719528178243,
      "grad_norm": 0.12175597250461578,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 548
    },
    {
      "epoch": 0.08994102228047182,
      "grad_norm": 0.12706248462200165,
      "learning_rate": 0.0001,
      "loss": 1.4151,
      "step": 549
    },
    {
      "epoch": 0.09010484927916121,
      "grad_norm": 0.13427045941352844,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 550
    },
    {
      "epoch": 0.0902686762778506,
      "grad_norm": 0.1334429532289505,
      "learning_rate": 0.0001,
      "loss": 1.6033,
      "step": 551
    },
    {
      "epoch": 0.09043250327653997,
      "grad_norm": 0.1250360906124115,
      "learning_rate": 0.0001,
      "loss": 1.4365,
      "step": 552
    },
    {
      "epoch": 0.09059633027522936,
      "grad_norm": 0.12384143471717834,
      "learning_rate": 0.0001,
      "loss": 1.5383,
      "step": 553
    },
    {
      "epoch": 0.09076015727391874,
      "grad_norm": 0.12594933807849884,
      "learning_rate": 0.0001,
      "loss": 1.4636,
      "step": 554
    },
    {
      "epoch": 0.09092398427260813,
      "grad_norm": 0.13059823215007782,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 555
    },
    {
      "epoch": 0.0910878112712975,
      "grad_norm": 0.13235454261302948,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 556
    },
    {
      "epoch": 0.09125163826998689,
      "grad_norm": 0.1220584288239479,
      "learning_rate": 0.0001,
      "loss": 1.3681,
      "step": 557
    },
    {
      "epoch": 0.09141546526867628,
      "grad_norm": 0.1357879787683487,
      "learning_rate": 0.0001,
      "loss": 1.5094,
      "step": 558
    },
    {
      "epoch": 0.09157929226736566,
      "grad_norm": 0.1270623803138733,
      "learning_rate": 0.0001,
      "loss": 1.4125,
      "step": 559
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 0.14505165815353394,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 560
    },
    {
      "epoch": 0.09190694626474442,
      "grad_norm": 0.12570977210998535,
      "learning_rate": 0.0001,
      "loss": 1.4555,
      "step": 561
    },
    {
      "epoch": 0.09207077326343381,
      "grad_norm": 0.12332609295845032,
      "learning_rate": 0.0001,
      "loss": 1.4514,
      "step": 562
    },
    {
      "epoch": 0.0922346002621232,
      "grad_norm": 0.14275409281253815,
      "learning_rate": 0.0001,
      "loss": 1.3801,
      "step": 563
    },
    {
      "epoch": 0.09239842726081259,
      "grad_norm": 0.12130515277385712,
      "learning_rate": 0.0001,
      "loss": 1.4815,
      "step": 564
    },
    {
      "epoch": 0.09256225425950197,
      "grad_norm": 0.1338832974433899,
      "learning_rate": 0.0001,
      "loss": 1.6048,
      "step": 565
    },
    {
      "epoch": 0.09272608125819135,
      "grad_norm": 0.13224637508392334,
      "learning_rate": 0.0001,
      "loss": 1.4451,
      "step": 566
    },
    {
      "epoch": 0.09288990825688073,
      "grad_norm": 0.12537266314029694,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 567
    },
    {
      "epoch": 0.09305373525557012,
      "grad_norm": 0.1278243511915207,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 568
    },
    {
      "epoch": 0.0932175622542595,
      "grad_norm": 0.12537169456481934,
      "learning_rate": 0.0001,
      "loss": 1.4802,
      "step": 569
    },
    {
      "epoch": 0.09338138925294888,
      "grad_norm": 0.1262512058019638,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 570
    },
    {
      "epoch": 0.09354521625163827,
      "grad_norm": 0.1289786696434021,
      "learning_rate": 0.0001,
      "loss": 1.5536,
      "step": 571
    },
    {
      "epoch": 0.09370904325032765,
      "grad_norm": 0.1348441243171692,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 572
    },
    {
      "epoch": 0.09387287024901704,
      "grad_norm": 0.1269592046737671,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 573
    },
    {
      "epoch": 0.09403669724770643,
      "grad_norm": 0.12503275275230408,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 574
    },
    {
      "epoch": 0.0942005242463958,
      "grad_norm": 0.12443274259567261,
      "learning_rate": 0.0001,
      "loss": 1.4209,
      "step": 575
    },
    {
      "epoch": 0.09436435124508519,
      "grad_norm": 0.12385372817516327,
      "learning_rate": 0.0001,
      "loss": 1.4122,
      "step": 576
    },
    {
      "epoch": 0.09452817824377457,
      "grad_norm": 0.12416256219148636,
      "learning_rate": 0.0001,
      "loss": 1.3935,
      "step": 577
    },
    {
      "epoch": 0.09469200524246396,
      "grad_norm": 0.1326628476381302,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 578
    },
    {
      "epoch": 0.09485583224115335,
      "grad_norm": 0.12143007665872574,
      "learning_rate": 0.0001,
      "loss": 1.4476,
      "step": 579
    },
    {
      "epoch": 0.09501965923984272,
      "grad_norm": 0.1283188760280609,
      "learning_rate": 0.0001,
      "loss": 1.4569,
      "step": 580
    },
    {
      "epoch": 0.09518348623853211,
      "grad_norm": 0.12522418797016144,
      "learning_rate": 0.0001,
      "loss": 1.5934,
      "step": 581
    },
    {
      "epoch": 0.0953473132372215,
      "grad_norm": 0.12706248462200165,
      "learning_rate": 0.0001,
      "loss": 1.568,
      "step": 582
    },
    {
      "epoch": 0.09551114023591088,
      "grad_norm": 0.13070838153362274,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 583
    },
    {
      "epoch": 0.09567496723460026,
      "grad_norm": 0.12980811297893524,
      "learning_rate": 0.0001,
      "loss": 1.5717,
      "step": 584
    },
    {
      "epoch": 0.09583879423328964,
      "grad_norm": 0.12547041475772858,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 585
    },
    {
      "epoch": 0.09600262123197903,
      "grad_norm": 0.11894677579402924,
      "learning_rate": 0.0001,
      "loss": 1.3879,
      "step": 586
    },
    {
      "epoch": 0.09616644823066842,
      "grad_norm": 0.11892498284578323,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 587
    },
    {
      "epoch": 0.0963302752293578,
      "grad_norm": 0.12646342813968658,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 588
    },
    {
      "epoch": 0.09649410222804718,
      "grad_norm": 0.12537556886672974,
      "learning_rate": 0.0001,
      "loss": 1.4067,
      "step": 589
    },
    {
      "epoch": 0.09665792922673656,
      "grad_norm": 0.13702069222927094,
      "learning_rate": 0.0001,
      "loss": 1.5596,
      "step": 590
    },
    {
      "epoch": 0.09682175622542595,
      "grad_norm": 0.12882833182811737,
      "learning_rate": 0.0001,
      "loss": 1.4899,
      "step": 591
    },
    {
      "epoch": 0.09698558322411534,
      "grad_norm": 0.12854504585266113,
      "learning_rate": 0.0001,
      "loss": 1.503,
      "step": 592
    },
    {
      "epoch": 0.09714941022280472,
      "grad_norm": 0.1309109479188919,
      "learning_rate": 0.0001,
      "loss": 1.5227,
      "step": 593
    },
    {
      "epoch": 0.0973132372214941,
      "grad_norm": 0.12623263895511627,
      "learning_rate": 0.0001,
      "loss": 1.439,
      "step": 594
    },
    {
      "epoch": 0.09747706422018348,
      "grad_norm": 0.12573908269405365,
      "learning_rate": 0.0001,
      "loss": 1.3876,
      "step": 595
    },
    {
      "epoch": 0.09764089121887287,
      "grad_norm": 0.12299245595932007,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 596
    },
    {
      "epoch": 0.09780471821756226,
      "grad_norm": 0.1270495355129242,
      "learning_rate": 0.0001,
      "loss": 1.4209,
      "step": 597
    },
    {
      "epoch": 0.09796854521625163,
      "grad_norm": 0.1374630331993103,
      "learning_rate": 0.0001,
      "loss": 1.5659,
      "step": 598
    },
    {
      "epoch": 0.09813237221494102,
      "grad_norm": 0.1296689659357071,
      "learning_rate": 0.0001,
      "loss": 1.448,
      "step": 599
    },
    {
      "epoch": 0.0982961992136304,
      "grad_norm": 0.12338189035654068,
      "learning_rate": 0.0001,
      "loss": 1.4545,
      "step": 600
    },
    {
      "epoch": 0.09846002621231979,
      "grad_norm": 0.11994987726211548,
      "learning_rate": 0.0001,
      "loss": 1.4275,
      "step": 601
    },
    {
      "epoch": 0.09862385321100918,
      "grad_norm": 0.1280028074979782,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 602
    },
    {
      "epoch": 0.09878768020969855,
      "grad_norm": 0.12570776045322418,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 603
    },
    {
      "epoch": 0.09895150720838794,
      "grad_norm": 0.13548217713832855,
      "learning_rate": 0.0001,
      "loss": 1.5359,
      "step": 604
    },
    {
      "epoch": 0.09911533420707733,
      "grad_norm": 0.12489759176969528,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 605
    },
    {
      "epoch": 0.09927916120576671,
      "grad_norm": 0.13694241642951965,
      "learning_rate": 0.0001,
      "loss": 1.3721,
      "step": 606
    },
    {
      "epoch": 0.0994429882044561,
      "grad_norm": 0.12032367289066315,
      "learning_rate": 0.0001,
      "loss": 1.3815,
      "step": 607
    },
    {
      "epoch": 0.09960681520314547,
      "grad_norm": 0.12383434176445007,
      "learning_rate": 0.0001,
      "loss": 1.3034,
      "step": 608
    },
    {
      "epoch": 0.09977064220183486,
      "grad_norm": 0.1330319195985794,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 609
    },
    {
      "epoch": 0.09993446920052425,
      "grad_norm": 0.13346469402313232,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 610
    },
    {
      "epoch": 0.10009829619921363,
      "grad_norm": 0.13877597451210022,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 611
    },
    {
      "epoch": 0.10026212319790301,
      "grad_norm": 0.12128439545631409,
      "learning_rate": 0.0001,
      "loss": 1.4691,
      "step": 612
    },
    {
      "epoch": 0.1004259501965924,
      "grad_norm": 0.1459330916404724,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 613
    },
    {
      "epoch": 0.10058977719528178,
      "grad_norm": 0.13037164509296417,
      "learning_rate": 0.0001,
      "loss": 1.4912,
      "step": 614
    },
    {
      "epoch": 0.10075360419397117,
      "grad_norm": 0.12892304360866547,
      "learning_rate": 0.0001,
      "loss": 1.487,
      "step": 615
    },
    {
      "epoch": 0.10091743119266056,
      "grad_norm": 0.14817756414413452,
      "learning_rate": 0.0001,
      "loss": 1.5623,
      "step": 616
    },
    {
      "epoch": 0.10108125819134993,
      "grad_norm": 0.12962795794010162,
      "learning_rate": 0.0001,
      "loss": 1.5387,
      "step": 617
    },
    {
      "epoch": 0.10124508519003932,
      "grad_norm": 0.13264529407024384,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 618
    },
    {
      "epoch": 0.1014089121887287,
      "grad_norm": 0.1367076188325882,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 619
    },
    {
      "epoch": 0.10157273918741809,
      "grad_norm": 0.12670248746871948,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 620
    },
    {
      "epoch": 0.10173656618610748,
      "grad_norm": 0.14859408140182495,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 621
    },
    {
      "epoch": 0.10190039318479685,
      "grad_norm": 0.12892693281173706,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 622
    },
    {
      "epoch": 0.10206422018348624,
      "grad_norm": 0.13376861810684204,
      "learning_rate": 0.0001,
      "loss": 1.5072,
      "step": 623
    },
    {
      "epoch": 0.10222804718217562,
      "grad_norm": 0.14278852939605713,
      "learning_rate": 0.0001,
      "loss": 1.5104,
      "step": 624
    },
    {
      "epoch": 0.10239187418086501,
      "grad_norm": 0.13521432876586914,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 625
    },
    {
      "epoch": 0.10255570117955438,
      "grad_norm": 0.14631788432598114,
      "learning_rate": 0.0001,
      "loss": 1.5672,
      "step": 626
    },
    {
      "epoch": 0.10271952817824377,
      "grad_norm": 0.13398370146751404,
      "learning_rate": 0.0001,
      "loss": 1.4631,
      "step": 627
    },
    {
      "epoch": 0.10288335517693316,
      "grad_norm": 0.12100734561681747,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 628
    },
    {
      "epoch": 0.10304718217562255,
      "grad_norm": 0.14890453219413757,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 629
    },
    {
      "epoch": 0.10321100917431193,
      "grad_norm": 0.13400933146476746,
      "learning_rate": 0.0001,
      "loss": 1.683,
      "step": 630
    },
    {
      "epoch": 0.1033748361730013,
      "grad_norm": 0.14169012010097504,
      "learning_rate": 0.0001,
      "loss": 1.491,
      "step": 631
    },
    {
      "epoch": 0.10353866317169069,
      "grad_norm": 0.13695670664310455,
      "learning_rate": 0.0001,
      "loss": 1.43,
      "step": 632
    },
    {
      "epoch": 0.10370249017038008,
      "grad_norm": 0.12358171492815018,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 633
    },
    {
      "epoch": 0.10386631716906947,
      "grad_norm": 0.14155566692352295,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 634
    },
    {
      "epoch": 0.10403014416775885,
      "grad_norm": 0.13217955827713013,
      "learning_rate": 0.0001,
      "loss": 1.5643,
      "step": 635
    },
    {
      "epoch": 0.10419397116644823,
      "grad_norm": 0.13999274373054504,
      "learning_rate": 0.0001,
      "loss": 1.4283,
      "step": 636
    },
    {
      "epoch": 0.10435779816513761,
      "grad_norm": 0.1391279101371765,
      "learning_rate": 0.0001,
      "loss": 1.4906,
      "step": 637
    },
    {
      "epoch": 0.104521625163827,
      "grad_norm": 0.1385086327791214,
      "learning_rate": 0.0001,
      "loss": 1.631,
      "step": 638
    },
    {
      "epoch": 0.10468545216251639,
      "grad_norm": 0.12770938873291016,
      "learning_rate": 0.0001,
      "loss": 1.4991,
      "step": 639
    },
    {
      "epoch": 0.10484927916120576,
      "grad_norm": 0.14131084084510803,
      "learning_rate": 0.0001,
      "loss": 1.3986,
      "step": 640
    },
    {
      "epoch": 0.10501310615989515,
      "grad_norm": 0.13531000912189484,
      "learning_rate": 0.0001,
      "loss": 1.5176,
      "step": 641
    },
    {
      "epoch": 0.10517693315858453,
      "grad_norm": 0.13043974339962006,
      "learning_rate": 0.0001,
      "loss": 1.546,
      "step": 642
    },
    {
      "epoch": 0.10534076015727392,
      "grad_norm": 0.12293804436922073,
      "learning_rate": 0.0001,
      "loss": 1.466,
      "step": 643
    },
    {
      "epoch": 0.10550458715596331,
      "grad_norm": 0.1367322951555252,
      "learning_rate": 0.0001,
      "loss": 1.5514,
      "step": 644
    },
    {
      "epoch": 0.10566841415465268,
      "grad_norm": 0.12427351623773575,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 645
    },
    {
      "epoch": 0.10583224115334207,
      "grad_norm": 0.12683941423892975,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 646
    },
    {
      "epoch": 0.10599606815203146,
      "grad_norm": 0.1280469447374344,
      "learning_rate": 0.0001,
      "loss": 1.4054,
      "step": 647
    },
    {
      "epoch": 0.10615989515072084,
      "grad_norm": 0.12691210210323334,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 648
    },
    {
      "epoch": 0.10632372214941023,
      "grad_norm": 0.13379907608032227,
      "learning_rate": 0.0001,
      "loss": 1.4951,
      "step": 649
    },
    {
      "epoch": 0.1064875491480996,
      "grad_norm": 0.1315588653087616,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 650
    },
    {
      "epoch": 0.10665137614678899,
      "grad_norm": 0.13715428113937378,
      "learning_rate": 0.0001,
      "loss": 1.4425,
      "step": 651
    },
    {
      "epoch": 0.10681520314547838,
      "grad_norm": 0.14846813678741455,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 652
    },
    {
      "epoch": 0.10697903014416776,
      "grad_norm": 0.12357223033905029,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 653
    },
    {
      "epoch": 0.10714285714285714,
      "grad_norm": 0.139797180891037,
      "learning_rate": 0.0001,
      "loss": 1.3992,
      "step": 654
    },
    {
      "epoch": 0.10730668414154652,
      "grad_norm": 0.12418066710233688,
      "learning_rate": 0.0001,
      "loss": 1.4735,
      "step": 655
    },
    {
      "epoch": 0.10747051114023591,
      "grad_norm": 0.14065606892108917,
      "learning_rate": 0.0001,
      "loss": 1.5774,
      "step": 656
    },
    {
      "epoch": 0.1076343381389253,
      "grad_norm": 0.13426558673381805,
      "learning_rate": 0.0001,
      "loss": 1.4832,
      "step": 657
    },
    {
      "epoch": 0.10779816513761468,
      "grad_norm": 0.13940469920635223,
      "learning_rate": 0.0001,
      "loss": 1.5211,
      "step": 658
    },
    {
      "epoch": 0.10796199213630406,
      "grad_norm": 0.12870129942893982,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 659
    },
    {
      "epoch": 0.10812581913499344,
      "grad_norm": 0.13486787676811218,
      "learning_rate": 0.0001,
      "loss": 1.4604,
      "step": 660
    },
    {
      "epoch": 0.10828964613368283,
      "grad_norm": 0.1372421681880951,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 661
    },
    {
      "epoch": 0.10845347313237222,
      "grad_norm": 0.14406070113182068,
      "learning_rate": 0.0001,
      "loss": 1.4243,
      "step": 662
    },
    {
      "epoch": 0.10861730013106159,
      "grad_norm": 0.1320434808731079,
      "learning_rate": 0.0001,
      "loss": 1.613,
      "step": 663
    },
    {
      "epoch": 0.10878112712975098,
      "grad_norm": 0.13125257194042206,
      "learning_rate": 0.0001,
      "loss": 1.4329,
      "step": 664
    },
    {
      "epoch": 0.10894495412844037,
      "grad_norm": 0.1250775009393692,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 665
    },
    {
      "epoch": 0.10910878112712975,
      "grad_norm": 0.13515514135360718,
      "learning_rate": 0.0001,
      "loss": 1.3995,
      "step": 666
    },
    {
      "epoch": 0.10927260812581914,
      "grad_norm": 0.1347736418247223,
      "learning_rate": 0.0001,
      "loss": 1.4524,
      "step": 667
    },
    {
      "epoch": 0.10943643512450851,
      "grad_norm": 0.11915424466133118,
      "learning_rate": 0.0001,
      "loss": 1.5793,
      "step": 668
    },
    {
      "epoch": 0.1096002621231979,
      "grad_norm": 0.15728433430194855,
      "learning_rate": 0.0001,
      "loss": 1.5087,
      "step": 669
    },
    {
      "epoch": 0.10976408912188729,
      "grad_norm": 0.13275671005249023,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 670
    },
    {
      "epoch": 0.10992791612057667,
      "grad_norm": 0.1422632336616516,
      "learning_rate": 0.0001,
      "loss": 1.4929,
      "step": 671
    },
    {
      "epoch": 0.11009174311926606,
      "grad_norm": 0.13993944227695465,
      "learning_rate": 0.0001,
      "loss": 1.4189,
      "step": 672
    },
    {
      "epoch": 0.11025557011795543,
      "grad_norm": 0.13283543288707733,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 673
    },
    {
      "epoch": 0.11041939711664482,
      "grad_norm": 0.14838266372680664,
      "learning_rate": 0.0001,
      "loss": 1.4697,
      "step": 674
    },
    {
      "epoch": 0.11058322411533421,
      "grad_norm": 0.1331229954957962,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 675
    },
    {
      "epoch": 0.1107470511140236,
      "grad_norm": 0.13245533406734467,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 676
    },
    {
      "epoch": 0.11091087811271297,
      "grad_norm": 0.13154149055480957,
      "learning_rate": 0.0001,
      "loss": 1.4434,
      "step": 677
    },
    {
      "epoch": 0.11107470511140236,
      "grad_norm": 0.12860529124736786,
      "learning_rate": 0.0001,
      "loss": 1.3993,
      "step": 678
    },
    {
      "epoch": 0.11123853211009174,
      "grad_norm": 0.13766147196292877,
      "learning_rate": 0.0001,
      "loss": 1.4765,
      "step": 679
    },
    {
      "epoch": 0.11140235910878113,
      "grad_norm": 0.1406853199005127,
      "learning_rate": 0.0001,
      "loss": 1.4333,
      "step": 680
    },
    {
      "epoch": 0.11156618610747052,
      "grad_norm": 0.12667514383792877,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 681
    },
    {
      "epoch": 0.11173001310615989,
      "grad_norm": 0.14250241219997406,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 682
    },
    {
      "epoch": 0.11189384010484928,
      "grad_norm": 0.1283908486366272,
      "learning_rate": 0.0001,
      "loss": 1.4456,
      "step": 683
    },
    {
      "epoch": 0.11205766710353866,
      "grad_norm": 0.13275974988937378,
      "learning_rate": 0.0001,
      "loss": 1.3679,
      "step": 684
    },
    {
      "epoch": 0.11222149410222805,
      "grad_norm": 0.12075935304164886,
      "learning_rate": 0.0001,
      "loss": 1.4635,
      "step": 685
    },
    {
      "epoch": 0.11238532110091744,
      "grad_norm": 0.12809894979000092,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 686
    },
    {
      "epoch": 0.11254914809960681,
      "grad_norm": 0.1315552145242691,
      "learning_rate": 0.0001,
      "loss": 1.4746,
      "step": 687
    },
    {
      "epoch": 0.1127129750982962,
      "grad_norm": 0.1346770077943802,
      "learning_rate": 0.0001,
      "loss": 1.5511,
      "step": 688
    },
    {
      "epoch": 0.11287680209698558,
      "grad_norm": 0.1336110532283783,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 689
    },
    {
      "epoch": 0.11304062909567497,
      "grad_norm": 0.13231389224529266,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 690
    },
    {
      "epoch": 0.11320445609436434,
      "grad_norm": 0.13388662040233612,
      "learning_rate": 0.0001,
      "loss": 1.5073,
      "step": 691
    },
    {
      "epoch": 0.11336828309305373,
      "grad_norm": 0.1264558881521225,
      "learning_rate": 0.0001,
      "loss": 1.3739,
      "step": 692
    },
    {
      "epoch": 0.11353211009174312,
      "grad_norm": 0.13136084377765656,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 693
    },
    {
      "epoch": 0.1136959370904325,
      "grad_norm": 0.1264805793762207,
      "learning_rate": 0.0001,
      "loss": 1.4853,
      "step": 694
    },
    {
      "epoch": 0.11385976408912189,
      "grad_norm": 0.12551257014274597,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 695
    },
    {
      "epoch": 0.11402359108781127,
      "grad_norm": 0.12379380315542221,
      "learning_rate": 0.0001,
      "loss": 1.3927,
      "step": 696
    },
    {
      "epoch": 0.11418741808650065,
      "grad_norm": 0.1331804394721985,
      "learning_rate": 0.0001,
      "loss": 1.5155,
      "step": 697
    },
    {
      "epoch": 0.11435124508519004,
      "grad_norm": 0.12703682482242584,
      "learning_rate": 0.0001,
      "loss": 1.4094,
      "step": 698
    },
    {
      "epoch": 0.11451507208387943,
      "grad_norm": 0.13457874953746796,
      "learning_rate": 0.0001,
      "loss": 1.6352,
      "step": 699
    },
    {
      "epoch": 0.11467889908256881,
      "grad_norm": 0.12837709486484528,
      "learning_rate": 0.0001,
      "loss": 1.3613,
      "step": 700
    },
    {
      "epoch": 0.11484272608125819,
      "grad_norm": 0.1268714964389801,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 701
    },
    {
      "epoch": 0.11500655307994757,
      "grad_norm": 0.12463714182376862,
      "learning_rate": 0.0001,
      "loss": 1.4517,
      "step": 702
    },
    {
      "epoch": 0.11517038007863696,
      "grad_norm": 0.12422896921634674,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 703
    },
    {
      "epoch": 0.11533420707732635,
      "grad_norm": 0.12679609656333923,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 704
    },
    {
      "epoch": 0.11549803407601572,
      "grad_norm": 0.12883196771144867,
      "learning_rate": 0.0001,
      "loss": 1.4611,
      "step": 705
    },
    {
      "epoch": 0.11566186107470511,
      "grad_norm": 0.12408079206943512,
      "learning_rate": 0.0001,
      "loss": 1.4696,
      "step": 706
    },
    {
      "epoch": 0.1158256880733945,
      "grad_norm": 0.12930727005004883,
      "learning_rate": 0.0001,
      "loss": 1.5412,
      "step": 707
    },
    {
      "epoch": 0.11598951507208388,
      "grad_norm": 0.1308346837759018,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 708
    },
    {
      "epoch": 0.11615334207077327,
      "grad_norm": 0.1333078145980835,
      "learning_rate": 0.0001,
      "loss": 1.4485,
      "step": 709
    },
    {
      "epoch": 0.11631716906946264,
      "grad_norm": 0.1255452185869217,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 710
    },
    {
      "epoch": 0.11648099606815203,
      "grad_norm": 0.12969857454299927,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 711
    },
    {
      "epoch": 0.11664482306684142,
      "grad_norm": 0.1274283081293106,
      "learning_rate": 0.0001,
      "loss": 1.4999,
      "step": 712
    },
    {
      "epoch": 0.1168086500655308,
      "grad_norm": 0.1273488998413086,
      "learning_rate": 0.0001,
      "loss": 1.5266,
      "step": 713
    },
    {
      "epoch": 0.11697247706422019,
      "grad_norm": 0.12344162911176682,
      "learning_rate": 0.0001,
      "loss": 1.3889,
      "step": 714
    },
    {
      "epoch": 0.11713630406290956,
      "grad_norm": 0.12548457086086273,
      "learning_rate": 0.0001,
      "loss": 1.4846,
      "step": 715
    },
    {
      "epoch": 0.11730013106159895,
      "grad_norm": 0.12746374309062958,
      "learning_rate": 0.0001,
      "loss": 1.475,
      "step": 716
    },
    {
      "epoch": 0.11746395806028834,
      "grad_norm": 0.13075010478496552,
      "learning_rate": 0.0001,
      "loss": 1.4512,
      "step": 717
    },
    {
      "epoch": 0.11762778505897772,
      "grad_norm": 0.12803134322166443,
      "learning_rate": 0.0001,
      "loss": 1.5982,
      "step": 718
    },
    {
      "epoch": 0.1177916120576671,
      "grad_norm": 0.12973140180110931,
      "learning_rate": 0.0001,
      "loss": 1.4999,
      "step": 719
    },
    {
      "epoch": 0.11795543905635648,
      "grad_norm": 0.13038846850395203,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 720
    },
    {
      "epoch": 0.11811926605504587,
      "grad_norm": 0.12627002596855164,
      "learning_rate": 0.0001,
      "loss": 1.4614,
      "step": 721
    },
    {
      "epoch": 0.11828309305373526,
      "grad_norm": 0.1276407241821289,
      "learning_rate": 0.0001,
      "loss": 1.5237,
      "step": 722
    },
    {
      "epoch": 0.11844692005242465,
      "grad_norm": 0.12358659505844116,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 723
    },
    {
      "epoch": 0.11861074705111402,
      "grad_norm": 0.12839065492153168,
      "learning_rate": 0.0001,
      "loss": 1.5227,
      "step": 724
    },
    {
      "epoch": 0.1187745740498034,
      "grad_norm": 0.12353626638650894,
      "learning_rate": 0.0001,
      "loss": 1.4146,
      "step": 725
    },
    {
      "epoch": 0.11893840104849279,
      "grad_norm": 0.14218755066394806,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 726
    },
    {
      "epoch": 0.11910222804718218,
      "grad_norm": 0.13292694091796875,
      "learning_rate": 0.0001,
      "loss": 1.5364,
      "step": 727
    },
    {
      "epoch": 0.11926605504587157,
      "grad_norm": 0.13555757701396942,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 728
    },
    {
      "epoch": 0.11942988204456094,
      "grad_norm": 0.1301213800907135,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 729
    },
    {
      "epoch": 0.11959370904325033,
      "grad_norm": 0.13762345910072327,
      "learning_rate": 0.0001,
      "loss": 1.5625,
      "step": 730
    },
    {
      "epoch": 0.11975753604193971,
      "grad_norm": 0.13400858640670776,
      "learning_rate": 0.0001,
      "loss": 1.5382,
      "step": 731
    },
    {
      "epoch": 0.1199213630406291,
      "grad_norm": 0.13771311938762665,
      "learning_rate": 0.0001,
      "loss": 1.3003,
      "step": 732
    },
    {
      "epoch": 0.12008519003931847,
      "grad_norm": 0.14105655252933502,
      "learning_rate": 0.0001,
      "loss": 1.3943,
      "step": 733
    },
    {
      "epoch": 0.12024901703800786,
      "grad_norm": 0.15480032563209534,
      "learning_rate": 0.0001,
      "loss": 1.5608,
      "step": 734
    },
    {
      "epoch": 0.12041284403669725,
      "grad_norm": 0.140495166182518,
      "learning_rate": 0.0001,
      "loss": 1.5726,
      "step": 735
    },
    {
      "epoch": 0.12057667103538663,
      "grad_norm": 0.15563678741455078,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 736
    },
    {
      "epoch": 0.12074049803407602,
      "grad_norm": 0.13888481259346008,
      "learning_rate": 0.0001,
      "loss": 1.5549,
      "step": 737
    },
    {
      "epoch": 0.1209043250327654,
      "grad_norm": 0.14686883985996246,
      "learning_rate": 0.0001,
      "loss": 1.4967,
      "step": 738
    },
    {
      "epoch": 0.12106815203145478,
      "grad_norm": 0.1358942687511444,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 739
    },
    {
      "epoch": 0.12123197903014417,
      "grad_norm": 0.1292220652103424,
      "learning_rate": 0.0001,
      "loss": 1.5109,
      "step": 740
    },
    {
      "epoch": 0.12139580602883356,
      "grad_norm": 0.13149432837963104,
      "learning_rate": 0.0001,
      "loss": 1.3725,
      "step": 741
    },
    {
      "epoch": 0.12155963302752294,
      "grad_norm": 0.13636918365955353,
      "learning_rate": 0.0001,
      "loss": 1.5104,
      "step": 742
    },
    {
      "epoch": 0.12172346002621232,
      "grad_norm": 0.1292741894721985,
      "learning_rate": 0.0001,
      "loss": 1.4215,
      "step": 743
    },
    {
      "epoch": 0.1218872870249017,
      "grad_norm": 0.1364440619945526,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 744
    },
    {
      "epoch": 0.12205111402359109,
      "grad_norm": 0.14230547845363617,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 745
    },
    {
      "epoch": 0.12221494102228048,
      "grad_norm": 0.13091309368610382,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 746
    },
    {
      "epoch": 0.12237876802096985,
      "grad_norm": 0.14762283861637115,
      "learning_rate": 0.0001,
      "loss": 1.4201,
      "step": 747
    },
    {
      "epoch": 0.12254259501965924,
      "grad_norm": 0.1322590559720993,
      "learning_rate": 0.0001,
      "loss": 1.4545,
      "step": 748
    },
    {
      "epoch": 0.12270642201834862,
      "grad_norm": 0.13209043443202972,
      "learning_rate": 0.0001,
      "loss": 1.4652,
      "step": 749
    },
    {
      "epoch": 0.12287024901703801,
      "grad_norm": 0.12821365892887115,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 750
    },
    {
      "epoch": 0.1230340760157274,
      "grad_norm": 0.13409477472305298,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 751
    },
    {
      "epoch": 0.12319790301441677,
      "grad_norm": 0.12697286903858185,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 752
    },
    {
      "epoch": 0.12336173001310616,
      "grad_norm": 0.13978971540927887,
      "learning_rate": 0.0001,
      "loss": 1.4572,
      "step": 753
    },
    {
      "epoch": 0.12352555701179554,
      "grad_norm": 0.13238103687763214,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 754
    },
    {
      "epoch": 0.12368938401048493,
      "grad_norm": 0.1310109943151474,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 755
    },
    {
      "epoch": 0.12385321100917432,
      "grad_norm": 0.1307615488767624,
      "learning_rate": 0.0001,
      "loss": 1.4938,
      "step": 756
    },
    {
      "epoch": 0.12401703800786369,
      "grad_norm": 0.12742164731025696,
      "learning_rate": 0.0001,
      "loss": 1.4416,
      "step": 757
    },
    {
      "epoch": 0.12418086500655308,
      "grad_norm": 0.13729506731033325,
      "learning_rate": 0.0001,
      "loss": 1.4101,
      "step": 758
    },
    {
      "epoch": 0.12434469200524247,
      "grad_norm": 0.12506775557994843,
      "learning_rate": 0.0001,
      "loss": 1.4612,
      "step": 759
    },
    {
      "epoch": 0.12450851900393185,
      "grad_norm": 0.1453479677438736,
      "learning_rate": 0.0001,
      "loss": 1.4454,
      "step": 760
    },
    {
      "epoch": 0.12467234600262123,
      "grad_norm": 0.12722153961658478,
      "learning_rate": 0.0001,
      "loss": 1.4318,
      "step": 761
    },
    {
      "epoch": 0.12483617300131061,
      "grad_norm": 0.13778507709503174,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 762
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.1326015740633011,
      "learning_rate": 0.0001,
      "loss": 1.4106,
      "step": 763
    },
    {
      "epoch": 0.1251638269986894,
      "grad_norm": 0.12171121686697006,
      "learning_rate": 0.0001,
      "loss": 1.4055,
      "step": 764
    },
    {
      "epoch": 0.12532765399737877,
      "grad_norm": 0.12851622700691223,
      "learning_rate": 0.0001,
      "loss": 1.3769,
      "step": 765
    },
    {
      "epoch": 0.12549148099606816,
      "grad_norm": 0.13635750114917755,
      "learning_rate": 0.0001,
      "loss": 1.5809,
      "step": 766
    },
    {
      "epoch": 0.12565530799475755,
      "grad_norm": 0.12070872634649277,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 767
    },
    {
      "epoch": 0.1258191349934469,
      "grad_norm": 0.13348183035850525,
      "learning_rate": 0.0001,
      "loss": 1.4608,
      "step": 768
    },
    {
      "epoch": 0.1259829619921363,
      "grad_norm": 0.12588299810886383,
      "learning_rate": 0.0001,
      "loss": 1.6346,
      "step": 769
    },
    {
      "epoch": 0.12614678899082568,
      "grad_norm": 0.12883533537387848,
      "learning_rate": 0.0001,
      "loss": 1.4195,
      "step": 770
    },
    {
      "epoch": 0.12631061598951507,
      "grad_norm": 0.12913841009140015,
      "learning_rate": 0.0001,
      "loss": 1.4629,
      "step": 771
    },
    {
      "epoch": 0.12647444298820446,
      "grad_norm": 0.13032715022563934,
      "learning_rate": 0.0001,
      "loss": 1.4873,
      "step": 772
    },
    {
      "epoch": 0.12663826998689384,
      "grad_norm": 0.12646310031414032,
      "learning_rate": 0.0001,
      "loss": 1.4534,
      "step": 773
    },
    {
      "epoch": 0.12680209698558323,
      "grad_norm": 0.12511228024959564,
      "learning_rate": 0.0001,
      "loss": 1.3246,
      "step": 774
    },
    {
      "epoch": 0.12696592398427262,
      "grad_norm": 0.13833408057689667,
      "learning_rate": 0.0001,
      "loss": 1.5557,
      "step": 775
    },
    {
      "epoch": 0.127129750982962,
      "grad_norm": 0.1325414627790451,
      "learning_rate": 0.0001,
      "loss": 1.4,
      "step": 776
    },
    {
      "epoch": 0.12729357798165136,
      "grad_norm": 0.12797212600708008,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 777
    },
    {
      "epoch": 0.12745740498034075,
      "grad_norm": 0.13081598281860352,
      "learning_rate": 0.0001,
      "loss": 1.477,
      "step": 778
    },
    {
      "epoch": 0.12762123197903014,
      "grad_norm": 0.12568604946136475,
      "learning_rate": 0.0001,
      "loss": 1.4083,
      "step": 779
    },
    {
      "epoch": 0.12778505897771952,
      "grad_norm": 0.12645204365253448,
      "learning_rate": 0.0001,
      "loss": 1.4522,
      "step": 780
    },
    {
      "epoch": 0.1279488859764089,
      "grad_norm": 0.13151831924915314,
      "learning_rate": 0.0001,
      "loss": 1.4745,
      "step": 781
    },
    {
      "epoch": 0.1281127129750983,
      "grad_norm": 0.13152624666690826,
      "learning_rate": 0.0001,
      "loss": 1.5535,
      "step": 782
    },
    {
      "epoch": 0.12827653997378768,
      "grad_norm": 0.12534844875335693,
      "learning_rate": 0.0001,
      "loss": 1.4665,
      "step": 783
    },
    {
      "epoch": 0.12844036697247707,
      "grad_norm": 0.133204847574234,
      "learning_rate": 0.0001,
      "loss": 1.5534,
      "step": 784
    },
    {
      "epoch": 0.12860419397116646,
      "grad_norm": 0.13000473380088806,
      "learning_rate": 0.0001,
      "loss": 1.3859,
      "step": 785
    },
    {
      "epoch": 0.12876802096985585,
      "grad_norm": 0.1267244666814804,
      "learning_rate": 0.0001,
      "loss": 1.5292,
      "step": 786
    },
    {
      "epoch": 0.1289318479685452,
      "grad_norm": 0.13117781281471252,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 787
    },
    {
      "epoch": 0.1290956749672346,
      "grad_norm": 0.1268448531627655,
      "learning_rate": 0.0001,
      "loss": 1.4242,
      "step": 788
    },
    {
      "epoch": 0.12925950196592398,
      "grad_norm": 0.1264076977968216,
      "learning_rate": 0.0001,
      "loss": 1.4227,
      "step": 789
    },
    {
      "epoch": 0.12942332896461337,
      "grad_norm": 0.139515683054924,
      "learning_rate": 0.0001,
      "loss": 1.4427,
      "step": 790
    },
    {
      "epoch": 0.12958715596330275,
      "grad_norm": 0.1312001496553421,
      "learning_rate": 0.0001,
      "loss": 1.5742,
      "step": 791
    },
    {
      "epoch": 0.12975098296199214,
      "grad_norm": 0.1295720785856247,
      "learning_rate": 0.0001,
      "loss": 1.4833,
      "step": 792
    },
    {
      "epoch": 0.12991480996068153,
      "grad_norm": 0.1356278359889984,
      "learning_rate": 0.0001,
      "loss": 1.5197,
      "step": 793
    },
    {
      "epoch": 0.1300786369593709,
      "grad_norm": 0.14651133120059967,
      "learning_rate": 0.0001,
      "loss": 1.5408,
      "step": 794
    },
    {
      "epoch": 0.1302424639580603,
      "grad_norm": 0.1278732270002365,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 795
    },
    {
      "epoch": 0.13040629095674966,
      "grad_norm": 0.1295413374900818,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 796
    },
    {
      "epoch": 0.13057011795543905,
      "grad_norm": 0.13473984599113464,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 797
    },
    {
      "epoch": 0.13073394495412843,
      "grad_norm": 0.12724176049232483,
      "learning_rate": 0.0001,
      "loss": 1.4768,
      "step": 798
    },
    {
      "epoch": 0.13089777195281782,
      "grad_norm": 0.13682308793067932,
      "learning_rate": 0.0001,
      "loss": 1.4854,
      "step": 799
    },
    {
      "epoch": 0.1310615989515072,
      "grad_norm": 0.13282006978988647,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 800
    },
    {
      "epoch": 0.1312254259501966,
      "grad_norm": 0.13964909315109253,
      "learning_rate": 0.0001,
      "loss": 1.5548,
      "step": 801
    },
    {
      "epoch": 0.13138925294888598,
      "grad_norm": 0.1477719247341156,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 802
    },
    {
      "epoch": 0.13155307994757537,
      "grad_norm": 0.13212165236473083,
      "learning_rate": 0.0001,
      "loss": 1.533,
      "step": 803
    },
    {
      "epoch": 0.13171690694626476,
      "grad_norm": 0.12764780223369598,
      "learning_rate": 0.0001,
      "loss": 1.435,
      "step": 804
    },
    {
      "epoch": 0.13188073394495411,
      "grad_norm": 0.1317799985408783,
      "learning_rate": 0.0001,
      "loss": 1.3789,
      "step": 805
    },
    {
      "epoch": 0.1320445609436435,
      "grad_norm": 0.11748173832893372,
      "learning_rate": 0.0001,
      "loss": 1.3426,
      "step": 806
    },
    {
      "epoch": 0.1322083879423329,
      "grad_norm": 0.13320861756801605,
      "learning_rate": 0.0001,
      "loss": 1.5095,
      "step": 807
    },
    {
      "epoch": 0.13237221494102228,
      "grad_norm": 0.13587425649166107,
      "learning_rate": 0.0001,
      "loss": 1.4879,
      "step": 808
    },
    {
      "epoch": 0.13253604193971166,
      "grad_norm": 0.12064733356237411,
      "learning_rate": 0.0001,
      "loss": 1.4287,
      "step": 809
    },
    {
      "epoch": 0.13269986893840105,
      "grad_norm": 0.1295618712902069,
      "learning_rate": 0.0001,
      "loss": 1.5026,
      "step": 810
    },
    {
      "epoch": 0.13286369593709044,
      "grad_norm": 0.1342451423406601,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 811
    },
    {
      "epoch": 0.13302752293577982,
      "grad_norm": 0.13342107832431793,
      "learning_rate": 0.0001,
      "loss": 1.4809,
      "step": 812
    },
    {
      "epoch": 0.1331913499344692,
      "grad_norm": 0.13591350615024567,
      "learning_rate": 0.0001,
      "loss": 1.5605,
      "step": 813
    },
    {
      "epoch": 0.1333551769331586,
      "grad_norm": 0.12232061475515366,
      "learning_rate": 0.0001,
      "loss": 1.34,
      "step": 814
    },
    {
      "epoch": 0.13351900393184796,
      "grad_norm": 0.12476851046085358,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 815
    },
    {
      "epoch": 0.13368283093053734,
      "grad_norm": 0.15262094140052795,
      "learning_rate": 0.0001,
      "loss": 1.4905,
      "step": 816
    },
    {
      "epoch": 0.13384665792922673,
      "grad_norm": 0.12979242205619812,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 817
    },
    {
      "epoch": 0.13401048492791612,
      "grad_norm": 0.1288214921951294,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 818
    },
    {
      "epoch": 0.1341743119266055,
      "grad_norm": 0.1359139084815979,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 819
    },
    {
      "epoch": 0.1343381389252949,
      "grad_norm": 0.15711714327335358,
      "learning_rate": 0.0001,
      "loss": 1.4289,
      "step": 820
    },
    {
      "epoch": 0.13450196592398428,
      "grad_norm": 0.12752211093902588,
      "learning_rate": 0.0001,
      "loss": 1.468,
      "step": 821
    },
    {
      "epoch": 0.13466579292267367,
      "grad_norm": 0.13990715146064758,
      "learning_rate": 0.0001,
      "loss": 1.4808,
      "step": 822
    },
    {
      "epoch": 0.13482961992136305,
      "grad_norm": 0.12881919741630554,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 823
    },
    {
      "epoch": 0.1349934469200524,
      "grad_norm": 0.13479682803153992,
      "learning_rate": 0.0001,
      "loss": 1.5067,
      "step": 824
    },
    {
      "epoch": 0.1351572739187418,
      "grad_norm": 0.12457506358623505,
      "learning_rate": 0.0001,
      "loss": 1.3518,
      "step": 825
    },
    {
      "epoch": 0.1353211009174312,
      "grad_norm": 0.12842564284801483,
      "learning_rate": 0.0001,
      "loss": 1.4679,
      "step": 826
    },
    {
      "epoch": 0.13548492791612057,
      "grad_norm": 0.1355355829000473,
      "learning_rate": 0.0001,
      "loss": 1.394,
      "step": 827
    },
    {
      "epoch": 0.13564875491480996,
      "grad_norm": 0.12799072265625,
      "learning_rate": 0.0001,
      "loss": 1.404,
      "step": 828
    },
    {
      "epoch": 0.13581258191349935,
      "grad_norm": 0.13437612354755402,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 829
    },
    {
      "epoch": 0.13597640891218873,
      "grad_norm": 0.13197094202041626,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 830
    },
    {
      "epoch": 0.13614023591087812,
      "grad_norm": 0.1354922652244568,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 831
    },
    {
      "epoch": 0.1363040629095675,
      "grad_norm": 0.13365304470062256,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 832
    },
    {
      "epoch": 0.13646788990825687,
      "grad_norm": 0.1301252543926239,
      "learning_rate": 0.0001,
      "loss": 1.495,
      "step": 833
    },
    {
      "epoch": 0.13663171690694625,
      "grad_norm": 0.13223440945148468,
      "learning_rate": 0.0001,
      "loss": 1.4115,
      "step": 834
    },
    {
      "epoch": 0.13679554390563564,
      "grad_norm": 0.12345802783966064,
      "learning_rate": 0.0001,
      "loss": 1.3902,
      "step": 835
    },
    {
      "epoch": 0.13695937090432503,
      "grad_norm": 0.12329761683940887,
      "learning_rate": 0.0001,
      "loss": 1.3567,
      "step": 836
    },
    {
      "epoch": 0.13712319790301442,
      "grad_norm": 0.14092130959033966,
      "learning_rate": 0.0001,
      "loss": 1.4886,
      "step": 837
    },
    {
      "epoch": 0.1372870249017038,
      "grad_norm": 0.12496673315763474,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 838
    },
    {
      "epoch": 0.1374508519003932,
      "grad_norm": 0.13800324499607086,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 839
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 0.13516810536384583,
      "learning_rate": 0.0001,
      "loss": 1.5372,
      "step": 840
    },
    {
      "epoch": 0.13777850589777196,
      "grad_norm": 0.13210734724998474,
      "learning_rate": 0.0001,
      "loss": 1.4232,
      "step": 841
    },
    {
      "epoch": 0.13794233289646135,
      "grad_norm": 0.13791127502918243,
      "learning_rate": 0.0001,
      "loss": 1.506,
      "step": 842
    },
    {
      "epoch": 0.1381061598951507,
      "grad_norm": 0.13421420753002167,
      "learning_rate": 0.0001,
      "loss": 1.468,
      "step": 843
    },
    {
      "epoch": 0.1382699868938401,
      "grad_norm": 0.1336262971162796,
      "learning_rate": 0.0001,
      "loss": 1.5667,
      "step": 844
    },
    {
      "epoch": 0.13843381389252948,
      "grad_norm": 0.13383284211158752,
      "learning_rate": 0.0001,
      "loss": 1.4432,
      "step": 845
    },
    {
      "epoch": 0.13859764089121887,
      "grad_norm": 0.13057659566402435,
      "learning_rate": 0.0001,
      "loss": 1.4829,
      "step": 846
    },
    {
      "epoch": 0.13876146788990826,
      "grad_norm": 0.12703344225883484,
      "learning_rate": 0.0001,
      "loss": 1.4495,
      "step": 847
    },
    {
      "epoch": 0.13892529488859764,
      "grad_norm": 0.14052940905094147,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 848
    },
    {
      "epoch": 0.13908912188728703,
      "grad_norm": 0.13233421742916107,
      "learning_rate": 0.0001,
      "loss": 1.5691,
      "step": 849
    },
    {
      "epoch": 0.13925294888597642,
      "grad_norm": 0.13677503168582916,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 850
    },
    {
      "epoch": 0.1394167758846658,
      "grad_norm": 0.13155722618103027,
      "learning_rate": 0.0001,
      "loss": 1.4596,
      "step": 851
    },
    {
      "epoch": 0.13958060288335516,
      "grad_norm": 0.1304759681224823,
      "learning_rate": 0.0001,
      "loss": 1.4295,
      "step": 852
    },
    {
      "epoch": 0.13974442988204455,
      "grad_norm": 0.14475083351135254,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 853
    },
    {
      "epoch": 0.13990825688073394,
      "grad_norm": 0.14297665655612946,
      "learning_rate": 0.0001,
      "loss": 1.5004,
      "step": 854
    },
    {
      "epoch": 0.14007208387942333,
      "grad_norm": 0.13002531230449677,
      "learning_rate": 0.0001,
      "loss": 1.5091,
      "step": 855
    },
    {
      "epoch": 0.1402359108781127,
      "grad_norm": 0.13651664555072784,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 856
    },
    {
      "epoch": 0.1403997378768021,
      "grad_norm": 0.13938665390014648,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 857
    },
    {
      "epoch": 0.1405635648754915,
      "grad_norm": 0.12868976593017578,
      "learning_rate": 0.0001,
      "loss": 1.6073,
      "step": 858
    },
    {
      "epoch": 0.14072739187418087,
      "grad_norm": 0.13151785731315613,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 859
    },
    {
      "epoch": 0.14089121887287026,
      "grad_norm": 0.13438887894153595,
      "learning_rate": 0.0001,
      "loss": 1.4377,
      "step": 860
    },
    {
      "epoch": 0.14105504587155962,
      "grad_norm": 0.1386173516511917,
      "learning_rate": 0.0001,
      "loss": 1.3486,
      "step": 861
    },
    {
      "epoch": 0.141218872870249,
      "grad_norm": 0.13364125788211823,
      "learning_rate": 0.0001,
      "loss": 1.4862,
      "step": 862
    },
    {
      "epoch": 0.1413826998689384,
      "grad_norm": 0.12615139782428741,
      "learning_rate": 0.0001,
      "loss": 1.3719,
      "step": 863
    },
    {
      "epoch": 0.14154652686762778,
      "grad_norm": 0.13253378868103027,
      "learning_rate": 0.0001,
      "loss": 1.4436,
      "step": 864
    },
    {
      "epoch": 0.14171035386631717,
      "grad_norm": 0.13049408793449402,
      "learning_rate": 0.0001,
      "loss": 1.4918,
      "step": 865
    },
    {
      "epoch": 0.14187418086500656,
      "grad_norm": 0.12856425344944,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 866
    },
    {
      "epoch": 0.14203800786369594,
      "grad_norm": 0.14721113443374634,
      "learning_rate": 0.0001,
      "loss": 1.3768,
      "step": 867
    },
    {
      "epoch": 0.14220183486238533,
      "grad_norm": 0.13527396321296692,
      "learning_rate": 0.0001,
      "loss": 1.4982,
      "step": 868
    },
    {
      "epoch": 0.14236566186107472,
      "grad_norm": 0.12970054149627686,
      "learning_rate": 0.0001,
      "loss": 1.4868,
      "step": 869
    },
    {
      "epoch": 0.1425294888597641,
      "grad_norm": 0.13485200703144073,
      "learning_rate": 0.0001,
      "loss": 1.4829,
      "step": 870
    },
    {
      "epoch": 0.14269331585845346,
      "grad_norm": 0.14117906987667084,
      "learning_rate": 0.0001,
      "loss": 1.5834,
      "step": 871
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.13866063952445984,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 872
    },
    {
      "epoch": 0.14302096985583224,
      "grad_norm": 0.141671285033226,
      "learning_rate": 0.0001,
      "loss": 1.5829,
      "step": 873
    },
    {
      "epoch": 0.14318479685452162,
      "grad_norm": 0.1414012908935547,
      "learning_rate": 0.0001,
      "loss": 1.406,
      "step": 874
    },
    {
      "epoch": 0.143348623853211,
      "grad_norm": 0.13079500198364258,
      "learning_rate": 0.0001,
      "loss": 1.4723,
      "step": 875
    },
    {
      "epoch": 0.1435124508519004,
      "grad_norm": 0.13441041111946106,
      "learning_rate": 0.0001,
      "loss": 1.3974,
      "step": 876
    },
    {
      "epoch": 0.14367627785058978,
      "grad_norm": 0.138494074344635,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 877
    },
    {
      "epoch": 0.14384010484927917,
      "grad_norm": 0.13074208796024323,
      "learning_rate": 0.0001,
      "loss": 1.229,
      "step": 878
    },
    {
      "epoch": 0.14400393184796856,
      "grad_norm": 0.1328742802143097,
      "learning_rate": 0.0001,
      "loss": 1.4382,
      "step": 879
    },
    {
      "epoch": 0.14416775884665792,
      "grad_norm": 0.1312500238418579,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 880
    },
    {
      "epoch": 0.1443315858453473,
      "grad_norm": 0.13210885226726532,
      "learning_rate": 0.0001,
      "loss": 1.4584,
      "step": 881
    },
    {
      "epoch": 0.1444954128440367,
      "grad_norm": 0.133746936917305,
      "learning_rate": 0.0001,
      "loss": 1.5722,
      "step": 882
    },
    {
      "epoch": 0.14465923984272608,
      "grad_norm": 0.1264539510011673,
      "learning_rate": 0.0001,
      "loss": 1.4937,
      "step": 883
    },
    {
      "epoch": 0.14482306684141547,
      "grad_norm": 0.12479538470506668,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 884
    },
    {
      "epoch": 0.14498689384010485,
      "grad_norm": 0.12786085903644562,
      "learning_rate": 0.0001,
      "loss": 1.476,
      "step": 885
    },
    {
      "epoch": 0.14515072083879424,
      "grad_norm": 0.13730409741401672,
      "learning_rate": 0.0001,
      "loss": 1.5254,
      "step": 886
    },
    {
      "epoch": 0.14531454783748363,
      "grad_norm": 0.13699358701705933,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 887
    },
    {
      "epoch": 0.145478374836173,
      "grad_norm": 0.12211377918720245,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 888
    },
    {
      "epoch": 0.14564220183486237,
      "grad_norm": 0.12985435128211975,
      "learning_rate": 0.0001,
      "loss": 1.3977,
      "step": 889
    },
    {
      "epoch": 0.14580602883355176,
      "grad_norm": 0.14085176587104797,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 890
    },
    {
      "epoch": 0.14596985583224115,
      "grad_norm": 0.13322080671787262,
      "learning_rate": 0.0001,
      "loss": 1.453,
      "step": 891
    },
    {
      "epoch": 0.14613368283093053,
      "grad_norm": 0.14627189934253693,
      "learning_rate": 0.0001,
      "loss": 1.5252,
      "step": 892
    },
    {
      "epoch": 0.14629750982961992,
      "grad_norm": 0.12851673364639282,
      "learning_rate": 0.0001,
      "loss": 1.5168,
      "step": 893
    },
    {
      "epoch": 0.1464613368283093,
      "grad_norm": 0.12940853834152222,
      "learning_rate": 0.0001,
      "loss": 1.4601,
      "step": 894
    },
    {
      "epoch": 0.1466251638269987,
      "grad_norm": 0.14077062904834747,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 895
    },
    {
      "epoch": 0.14678899082568808,
      "grad_norm": 0.12612144649028778,
      "learning_rate": 0.0001,
      "loss": 1.4486,
      "step": 896
    },
    {
      "epoch": 0.14695281782437747,
      "grad_norm": 0.12915703654289246,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 897
    },
    {
      "epoch": 0.14711664482306683,
      "grad_norm": 0.137868732213974,
      "learning_rate": 0.0001,
      "loss": 1.431,
      "step": 898
    },
    {
      "epoch": 0.14728047182175621,
      "grad_norm": 0.13154515624046326,
      "learning_rate": 0.0001,
      "loss": 1.4666,
      "step": 899
    },
    {
      "epoch": 0.1474442988204456,
      "grad_norm": 0.14809870719909668,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 900
    },
    {
      "epoch": 0.147608125819135,
      "grad_norm": 0.13884691894054413,
      "learning_rate": 0.0001,
      "loss": 1.5071,
      "step": 901
    },
    {
      "epoch": 0.14777195281782438,
      "grad_norm": 0.13934724032878876,
      "learning_rate": 0.0001,
      "loss": 1.5397,
      "step": 902
    },
    {
      "epoch": 0.14793577981651376,
      "grad_norm": 0.14282019436359406,
      "learning_rate": 0.0001,
      "loss": 1.44,
      "step": 903
    },
    {
      "epoch": 0.14809960681520315,
      "grad_norm": 0.1297287791967392,
      "learning_rate": 0.0001,
      "loss": 1.4534,
      "step": 904
    },
    {
      "epoch": 0.14826343381389254,
      "grad_norm": 0.1263768970966339,
      "learning_rate": 0.0001,
      "loss": 1.3634,
      "step": 905
    },
    {
      "epoch": 0.14842726081258192,
      "grad_norm": 0.14507046341896057,
      "learning_rate": 0.0001,
      "loss": 1.5649,
      "step": 906
    },
    {
      "epoch": 0.1485910878112713,
      "grad_norm": 0.1251538097858429,
      "learning_rate": 0.0001,
      "loss": 1.4264,
      "step": 907
    },
    {
      "epoch": 0.14875491480996067,
      "grad_norm": 0.1384262591600418,
      "learning_rate": 0.0001,
      "loss": 1.484,
      "step": 908
    },
    {
      "epoch": 0.14891874180865006,
      "grad_norm": 0.14661893248558044,
      "learning_rate": 0.0001,
      "loss": 1.4911,
      "step": 909
    },
    {
      "epoch": 0.14908256880733944,
      "grad_norm": 0.13075251877307892,
      "learning_rate": 0.0001,
      "loss": 1.4844,
      "step": 910
    },
    {
      "epoch": 0.14924639580602883,
      "grad_norm": 0.13334473967552185,
      "learning_rate": 0.0001,
      "loss": 1.4273,
      "step": 911
    },
    {
      "epoch": 0.14941022280471822,
      "grad_norm": 0.12892960011959076,
      "learning_rate": 0.0001,
      "loss": 1.4376,
      "step": 912
    },
    {
      "epoch": 0.1495740498034076,
      "grad_norm": 0.1310575306415558,
      "learning_rate": 0.0001,
      "loss": 1.4002,
      "step": 913
    },
    {
      "epoch": 0.149737876802097,
      "grad_norm": 0.13606613874435425,
      "learning_rate": 0.0001,
      "loss": 1.4738,
      "step": 914
    },
    {
      "epoch": 0.14990170380078638,
      "grad_norm": 0.12975278496742249,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 915
    },
    {
      "epoch": 0.15006553079947577,
      "grad_norm": 0.13267037272453308,
      "learning_rate": 0.0001,
      "loss": 1.442,
      "step": 916
    },
    {
      "epoch": 0.15022935779816513,
      "grad_norm": 0.12309186160564423,
      "learning_rate": 0.0001,
      "loss": 1.4223,
      "step": 917
    },
    {
      "epoch": 0.1503931847968545,
      "grad_norm": 0.13569849729537964,
      "learning_rate": 0.0001,
      "loss": 1.428,
      "step": 918
    },
    {
      "epoch": 0.1505570117955439,
      "grad_norm": 0.12518325448036194,
      "learning_rate": 0.0001,
      "loss": 1.3323,
      "step": 919
    },
    {
      "epoch": 0.15072083879423329,
      "grad_norm": 0.1262001395225525,
      "learning_rate": 0.0001,
      "loss": 1.4641,
      "step": 920
    },
    {
      "epoch": 0.15088466579292267,
      "grad_norm": 0.13454599678516388,
      "learning_rate": 0.0001,
      "loss": 1.4664,
      "step": 921
    },
    {
      "epoch": 0.15104849279161206,
      "grad_norm": 0.13165174424648285,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 922
    },
    {
      "epoch": 0.15121231979030145,
      "grad_norm": 0.13380499184131622,
      "learning_rate": 0.0001,
      "loss": 1.4244,
      "step": 923
    },
    {
      "epoch": 0.15137614678899083,
      "grad_norm": 0.13461482524871826,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 924
    },
    {
      "epoch": 0.15153997378768022,
      "grad_norm": 0.12390399724245071,
      "learning_rate": 0.0001,
      "loss": 1.3683,
      "step": 925
    },
    {
      "epoch": 0.15170380078636958,
      "grad_norm": 0.13173697888851166,
      "learning_rate": 0.0001,
      "loss": 1.3705,
      "step": 926
    },
    {
      "epoch": 0.15186762778505897,
      "grad_norm": 0.13951408863067627,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 927
    },
    {
      "epoch": 0.15203145478374835,
      "grad_norm": 0.13069495558738708,
      "learning_rate": 0.0001,
      "loss": 1.4044,
      "step": 928
    },
    {
      "epoch": 0.15219528178243774,
      "grad_norm": 0.14366430044174194,
      "learning_rate": 0.0001,
      "loss": 1.4489,
      "step": 929
    },
    {
      "epoch": 0.15235910878112713,
      "grad_norm": 0.13332362473011017,
      "learning_rate": 0.0001,
      "loss": 1.6005,
      "step": 930
    },
    {
      "epoch": 0.15252293577981652,
      "grad_norm": 0.13039685785770416,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 931
    },
    {
      "epoch": 0.1526867627785059,
      "grad_norm": 0.1323046088218689,
      "learning_rate": 0.0001,
      "loss": 1.5461,
      "step": 932
    },
    {
      "epoch": 0.1528505897771953,
      "grad_norm": 0.12273120880126953,
      "learning_rate": 0.0001,
      "loss": 1.4354,
      "step": 933
    },
    {
      "epoch": 0.15301441677588468,
      "grad_norm": 0.12755340337753296,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 934
    },
    {
      "epoch": 0.15317824377457406,
      "grad_norm": 0.12971536815166473,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 935
    },
    {
      "epoch": 0.15334207077326342,
      "grad_norm": 0.13140687346458435,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 936
    },
    {
      "epoch": 0.1535058977719528,
      "grad_norm": 0.12588505446910858,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 937
    },
    {
      "epoch": 0.1536697247706422,
      "grad_norm": 0.13628828525543213,
      "learning_rate": 0.0001,
      "loss": 1.4544,
      "step": 938
    },
    {
      "epoch": 0.15383355176933158,
      "grad_norm": 0.1333438605070114,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 939
    },
    {
      "epoch": 0.15399737876802097,
      "grad_norm": 0.1360521763563156,
      "learning_rate": 0.0001,
      "loss": 1.4314,
      "step": 940
    },
    {
      "epoch": 0.15416120576671036,
      "grad_norm": 0.14650259912014008,
      "learning_rate": 0.0001,
      "loss": 1.5672,
      "step": 941
    },
    {
      "epoch": 0.15432503276539974,
      "grad_norm": 0.139370396733284,
      "learning_rate": 0.0001,
      "loss": 1.4718,
      "step": 942
    },
    {
      "epoch": 0.15448885976408913,
      "grad_norm": 0.14333318173885345,
      "learning_rate": 0.0001,
      "loss": 1.5515,
      "step": 943
    },
    {
      "epoch": 0.15465268676277852,
      "grad_norm": 0.13650213181972504,
      "learning_rate": 0.0001,
      "loss": 1.5011,
      "step": 944
    },
    {
      "epoch": 0.15481651376146788,
      "grad_norm": 0.1390881985425949,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 945
    },
    {
      "epoch": 0.15498034076015726,
      "grad_norm": 0.1302831470966339,
      "learning_rate": 0.0001,
      "loss": 1.5472,
      "step": 946
    },
    {
      "epoch": 0.15514416775884665,
      "grad_norm": 0.1375759094953537,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 947
    },
    {
      "epoch": 0.15530799475753604,
      "grad_norm": 0.13592498004436493,
      "learning_rate": 0.0001,
      "loss": 1.5062,
      "step": 948
    },
    {
      "epoch": 0.15547182175622543,
      "grad_norm": 0.12605321407318115,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 949
    },
    {
      "epoch": 0.1556356487549148,
      "grad_norm": 0.1253965049982071,
      "learning_rate": 0.0001,
      "loss": 1.498,
      "step": 950
    },
    {
      "epoch": 0.1557994757536042,
      "grad_norm": 0.12434224039316177,
      "learning_rate": 0.0001,
      "loss": 1.4305,
      "step": 951
    },
    {
      "epoch": 0.1559633027522936,
      "grad_norm": 0.1299263834953308,
      "learning_rate": 0.0001,
      "loss": 1.4292,
      "step": 952
    },
    {
      "epoch": 0.15612712975098297,
      "grad_norm": 0.13775578141212463,
      "learning_rate": 0.0001,
      "loss": 1.4728,
      "step": 953
    },
    {
      "epoch": 0.15629095674967233,
      "grad_norm": 0.13631080090999603,
      "learning_rate": 0.0001,
      "loss": 1.5141,
      "step": 954
    },
    {
      "epoch": 0.15645478374836172,
      "grad_norm": 0.13672727346420288,
      "learning_rate": 0.0001,
      "loss": 1.4934,
      "step": 955
    },
    {
      "epoch": 0.1566186107470511,
      "grad_norm": 0.13358145952224731,
      "learning_rate": 0.0001,
      "loss": 1.3875,
      "step": 956
    },
    {
      "epoch": 0.1567824377457405,
      "grad_norm": 0.12535107135772705,
      "learning_rate": 0.0001,
      "loss": 1.4186,
      "step": 957
    },
    {
      "epoch": 0.15694626474442988,
      "grad_norm": 0.13820543885231018,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 958
    },
    {
      "epoch": 0.15711009174311927,
      "grad_norm": 0.13554927706718445,
      "learning_rate": 0.0001,
      "loss": 1.5568,
      "step": 959
    },
    {
      "epoch": 0.15727391874180865,
      "grad_norm": 0.1274915486574173,
      "learning_rate": 0.0001,
      "loss": 1.4286,
      "step": 960
    },
    {
      "epoch": 0.15743774574049804,
      "grad_norm": 0.14633774757385254,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 961
    },
    {
      "epoch": 0.15760157273918743,
      "grad_norm": 0.13605555891990662,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 962
    },
    {
      "epoch": 0.15776539973787682,
      "grad_norm": 0.14636871218681335,
      "learning_rate": 0.0001,
      "loss": 1.5165,
      "step": 963
    },
    {
      "epoch": 0.15792922673656618,
      "grad_norm": 0.1509140431880951,
      "learning_rate": 0.0001,
      "loss": 1.4761,
      "step": 964
    },
    {
      "epoch": 0.15809305373525556,
      "grad_norm": 0.1349087953567505,
      "learning_rate": 0.0001,
      "loss": 1.5018,
      "step": 965
    },
    {
      "epoch": 0.15825688073394495,
      "grad_norm": 0.1653502732515335,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 966
    },
    {
      "epoch": 0.15842070773263434,
      "grad_norm": 0.13891054689884186,
      "learning_rate": 0.0001,
      "loss": 1.4277,
      "step": 967
    },
    {
      "epoch": 0.15858453473132372,
      "grad_norm": 0.1348383128643036,
      "learning_rate": 0.0001,
      "loss": 1.4954,
      "step": 968
    },
    {
      "epoch": 0.1587483617300131,
      "grad_norm": 0.148629829287529,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 969
    },
    {
      "epoch": 0.1589121887287025,
      "grad_norm": 0.13720032572746277,
      "learning_rate": 0.0001,
      "loss": 1.3854,
      "step": 970
    },
    {
      "epoch": 0.15907601572739188,
      "grad_norm": 0.13478314876556396,
      "learning_rate": 0.0001,
      "loss": 1.3843,
      "step": 971
    },
    {
      "epoch": 0.15923984272608127,
      "grad_norm": 0.14525622129440308,
      "learning_rate": 0.0001,
      "loss": 1.3814,
      "step": 972
    },
    {
      "epoch": 0.15940366972477063,
      "grad_norm": 0.13203337788581848,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 973
    },
    {
      "epoch": 0.15956749672346002,
      "grad_norm": 0.1290595680475235,
      "learning_rate": 0.0001,
      "loss": 1.412,
      "step": 974
    },
    {
      "epoch": 0.1597313237221494,
      "grad_norm": 0.1589415818452835,
      "learning_rate": 0.0001,
      "loss": 1.4829,
      "step": 975
    },
    {
      "epoch": 0.1598951507208388,
      "grad_norm": 0.13755802810192108,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 976
    },
    {
      "epoch": 0.16005897771952818,
      "grad_norm": 0.13499528169631958,
      "learning_rate": 0.0001,
      "loss": 1.4973,
      "step": 977
    },
    {
      "epoch": 0.16022280471821757,
      "grad_norm": 0.15408873558044434,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 978
    },
    {
      "epoch": 0.16038663171690695,
      "grad_norm": 0.1363677978515625,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 979
    },
    {
      "epoch": 0.16055045871559634,
      "grad_norm": 0.12958399951457977,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 980
    },
    {
      "epoch": 0.16071428571428573,
      "grad_norm": 0.14144963026046753,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 981
    },
    {
      "epoch": 0.16087811271297509,
      "grad_norm": 0.1355958878993988,
      "learning_rate": 0.0001,
      "loss": 1.4042,
      "step": 982
    },
    {
      "epoch": 0.16104193971166447,
      "grad_norm": 0.14460384845733643,
      "learning_rate": 0.0001,
      "loss": 1.476,
      "step": 983
    },
    {
      "epoch": 0.16120576671035386,
      "grad_norm": 0.1358080953359604,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 984
    },
    {
      "epoch": 0.16136959370904325,
      "grad_norm": 0.13358932733535767,
      "learning_rate": 0.0001,
      "loss": 1.5164,
      "step": 985
    },
    {
      "epoch": 0.16153342070773263,
      "grad_norm": 0.1432204693555832,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 986
    },
    {
      "epoch": 0.16169724770642202,
      "grad_norm": 0.13913975656032562,
      "learning_rate": 0.0001,
      "loss": 1.4079,
      "step": 987
    },
    {
      "epoch": 0.1618610747051114,
      "grad_norm": 0.14018113911151886,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 988
    },
    {
      "epoch": 0.1620249017038008,
      "grad_norm": 0.14030413329601288,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 989
    },
    {
      "epoch": 0.16218872870249018,
      "grad_norm": 0.1372368037700653,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 990
    },
    {
      "epoch": 0.16235255570117954,
      "grad_norm": 0.12873098254203796,
      "learning_rate": 0.0001,
      "loss": 1.4525,
      "step": 991
    },
    {
      "epoch": 0.16251638269986893,
      "grad_norm": 0.13500003516674042,
      "learning_rate": 0.0001,
      "loss": 1.4959,
      "step": 992
    },
    {
      "epoch": 0.16268020969855831,
      "grad_norm": 0.1478472650051117,
      "learning_rate": 0.0001,
      "loss": 1.4127,
      "step": 993
    },
    {
      "epoch": 0.1628440366972477,
      "grad_norm": 0.13208359479904175,
      "learning_rate": 0.0001,
      "loss": 1.5329,
      "step": 994
    },
    {
      "epoch": 0.1630078636959371,
      "grad_norm": 0.148567795753479,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 995
    },
    {
      "epoch": 0.16317169069462648,
      "grad_norm": 0.1376078575849533,
      "learning_rate": 0.0001,
      "loss": 1.5318,
      "step": 996
    },
    {
      "epoch": 0.16333551769331586,
      "grad_norm": 0.13221588730812073,
      "learning_rate": 0.0001,
      "loss": 1.4663,
      "step": 997
    },
    {
      "epoch": 0.16349934469200525,
      "grad_norm": 0.13471892476081848,
      "learning_rate": 0.0001,
      "loss": 1.4353,
      "step": 998
    },
    {
      "epoch": 0.16366317169069464,
      "grad_norm": 0.13675783574581146,
      "learning_rate": 0.0001,
      "loss": 1.4894,
      "step": 999
    },
    {
      "epoch": 0.16382699868938402,
      "grad_norm": 0.12859280407428741,
      "learning_rate": 0.0001,
      "loss": 1.4109,
      "step": 1000
    },
    {
      "epoch": 0.16399082568807338,
      "grad_norm": 0.1379777193069458,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 1001
    },
    {
      "epoch": 0.16415465268676277,
      "grad_norm": 0.1350652575492859,
      "learning_rate": 0.0001,
      "loss": 1.407,
      "step": 1002
    },
    {
      "epoch": 0.16431847968545216,
      "grad_norm": 0.1357770562171936,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 1003
    },
    {
      "epoch": 0.16448230668414154,
      "grad_norm": 0.15089915692806244,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 1004
    },
    {
      "epoch": 0.16464613368283093,
      "grad_norm": 0.13771754503250122,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 1005
    },
    {
      "epoch": 0.16480996068152032,
      "grad_norm": 0.14095044136047363,
      "learning_rate": 0.0001,
      "loss": 1.504,
      "step": 1006
    },
    {
      "epoch": 0.1649737876802097,
      "grad_norm": 0.14101985096931458,
      "learning_rate": 0.0001,
      "loss": 1.5046,
      "step": 1007
    },
    {
      "epoch": 0.1651376146788991,
      "grad_norm": 0.14022016525268555,
      "learning_rate": 0.0001,
      "loss": 1.4019,
      "step": 1008
    },
    {
      "epoch": 0.16530144167758848,
      "grad_norm": 0.1469026356935501,
      "learning_rate": 0.0001,
      "loss": 1.4083,
      "step": 1009
    },
    {
      "epoch": 0.16546526867627784,
      "grad_norm": 0.12376164644956589,
      "learning_rate": 0.0001,
      "loss": 1.4088,
      "step": 1010
    },
    {
      "epoch": 0.16562909567496722,
      "grad_norm": 0.1330767273902893,
      "learning_rate": 0.0001,
      "loss": 1.4736,
      "step": 1011
    },
    {
      "epoch": 0.1657929226736566,
      "grad_norm": 0.14418423175811768,
      "learning_rate": 0.0001,
      "loss": 1.4995,
      "step": 1012
    },
    {
      "epoch": 0.165956749672346,
      "grad_norm": 0.1518072634935379,
      "learning_rate": 0.0001,
      "loss": 1.4433,
      "step": 1013
    },
    {
      "epoch": 0.16612057667103539,
      "grad_norm": 0.13048358261585236,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 1014
    },
    {
      "epoch": 0.16628440366972477,
      "grad_norm": 0.1434691995382309,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 1015
    },
    {
      "epoch": 0.16644823066841416,
      "grad_norm": 0.13723652064800262,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 1016
    },
    {
      "epoch": 0.16661205766710355,
      "grad_norm": 0.13145390152931213,
      "learning_rate": 0.0001,
      "loss": 1.5212,
      "step": 1017
    },
    {
      "epoch": 0.16677588466579293,
      "grad_norm": 0.14107628166675568,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 1018
    },
    {
      "epoch": 0.1669397116644823,
      "grad_norm": 0.14621242880821228,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 1019
    },
    {
      "epoch": 0.16710353866317168,
      "grad_norm": 0.13182333111763,
      "learning_rate": 0.0001,
      "loss": 1.4389,
      "step": 1020
    },
    {
      "epoch": 0.16726736566186107,
      "grad_norm": 0.14174515008926392,
      "learning_rate": 0.0001,
      "loss": 1.5264,
      "step": 1021
    },
    {
      "epoch": 0.16743119266055045,
      "grad_norm": 0.13635191321372986,
      "learning_rate": 0.0001,
      "loss": 1.4764,
      "step": 1022
    },
    {
      "epoch": 0.16759501965923984,
      "grad_norm": 0.1316136121749878,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 1023
    },
    {
      "epoch": 0.16775884665792923,
      "grad_norm": 0.13220228254795074,
      "learning_rate": 0.0001,
      "loss": 1.4698,
      "step": 1024
    },
    {
      "epoch": 0.16792267365661862,
      "grad_norm": 0.13377471268177032,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 1025
    },
    {
      "epoch": 0.168086500655308,
      "grad_norm": 0.12811405956745148,
      "learning_rate": 0.0001,
      "loss": 1.4658,
      "step": 1026
    },
    {
      "epoch": 0.1682503276539974,
      "grad_norm": 0.13001900911331177,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 1027
    },
    {
      "epoch": 0.16841415465268678,
      "grad_norm": 0.1447964310646057,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 1028
    },
    {
      "epoch": 0.16857798165137614,
      "grad_norm": 0.13122352957725525,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 1029
    },
    {
      "epoch": 0.16874180865006552,
      "grad_norm": 0.14353182911872864,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 1030
    },
    {
      "epoch": 0.1689056356487549,
      "grad_norm": 0.1465579718351364,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 1031
    },
    {
      "epoch": 0.1690694626474443,
      "grad_norm": 0.12737677991390228,
      "learning_rate": 0.0001,
      "loss": 1.3653,
      "step": 1032
    },
    {
      "epoch": 0.16923328964613368,
      "grad_norm": 0.13926975429058075,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 1033
    },
    {
      "epoch": 0.16939711664482307,
      "grad_norm": 0.1398305594921112,
      "learning_rate": 0.0001,
      "loss": 1.5044,
      "step": 1034
    },
    {
      "epoch": 0.16956094364351246,
      "grad_norm": 0.1347191035747528,
      "learning_rate": 0.0001,
      "loss": 1.4907,
      "step": 1035
    },
    {
      "epoch": 0.16972477064220184,
      "grad_norm": 0.14173738658428192,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 1036
    },
    {
      "epoch": 0.16988859764089123,
      "grad_norm": 0.1442819982767105,
      "learning_rate": 0.0001,
      "loss": 1.4702,
      "step": 1037
    },
    {
      "epoch": 0.1700524246395806,
      "grad_norm": 0.1271684318780899,
      "learning_rate": 0.0001,
      "loss": 1.5085,
      "step": 1038
    },
    {
      "epoch": 0.17021625163826998,
      "grad_norm": 0.13438351452350616,
      "learning_rate": 0.0001,
      "loss": 1.5264,
      "step": 1039
    },
    {
      "epoch": 0.17038007863695936,
      "grad_norm": 0.13932441174983978,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 1040
    },
    {
      "epoch": 0.17054390563564875,
      "grad_norm": 0.12472131848335266,
      "learning_rate": 0.0001,
      "loss": 1.4211,
      "step": 1041
    },
    {
      "epoch": 0.17070773263433814,
      "grad_norm": 0.13076287508010864,
      "learning_rate": 0.0001,
      "loss": 1.4268,
      "step": 1042
    },
    {
      "epoch": 0.17087155963302753,
      "grad_norm": 0.13389906287193298,
      "learning_rate": 0.0001,
      "loss": 1.4196,
      "step": 1043
    },
    {
      "epoch": 0.1710353866317169,
      "grad_norm": 0.12706439197063446,
      "learning_rate": 0.0001,
      "loss": 1.4856,
      "step": 1044
    },
    {
      "epoch": 0.1711992136304063,
      "grad_norm": 0.13442771136760712,
      "learning_rate": 0.0001,
      "loss": 1.4476,
      "step": 1045
    },
    {
      "epoch": 0.1713630406290957,
      "grad_norm": 0.1368313729763031,
      "learning_rate": 0.0001,
      "loss": 1.4564,
      "step": 1046
    },
    {
      "epoch": 0.17152686762778505,
      "grad_norm": 0.12584026157855988,
      "learning_rate": 0.0001,
      "loss": 1.3767,
      "step": 1047
    },
    {
      "epoch": 0.17169069462647443,
      "grad_norm": 0.13403959572315216,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 1048
    },
    {
      "epoch": 0.17185452162516382,
      "grad_norm": 0.13692210614681244,
      "learning_rate": 0.0001,
      "loss": 1.4548,
      "step": 1049
    },
    {
      "epoch": 0.1720183486238532,
      "grad_norm": 0.13215316832065582,
      "learning_rate": 0.0001,
      "loss": 1.3974,
      "step": 1050
    },
    {
      "epoch": 0.1721821756225426,
      "grad_norm": 0.12955117225646973,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 1051
    },
    {
      "epoch": 0.17234600262123198,
      "grad_norm": 0.13104918599128723,
      "learning_rate": 0.0001,
      "loss": 1.4636,
      "step": 1052
    },
    {
      "epoch": 0.17250982961992137,
      "grad_norm": 0.13202208280563354,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 1053
    },
    {
      "epoch": 0.17267365661861075,
      "grad_norm": 0.1320265233516693,
      "learning_rate": 0.0001,
      "loss": 1.3351,
      "step": 1054
    },
    {
      "epoch": 0.17283748361730014,
      "grad_norm": 0.1233760192990303,
      "learning_rate": 0.0001,
      "loss": 1.3325,
      "step": 1055
    },
    {
      "epoch": 0.17300131061598953,
      "grad_norm": 0.1297565996646881,
      "learning_rate": 0.0001,
      "loss": 1.3724,
      "step": 1056
    },
    {
      "epoch": 0.1731651376146789,
      "grad_norm": 0.13623128831386566,
      "learning_rate": 0.0001,
      "loss": 1.4593,
      "step": 1057
    },
    {
      "epoch": 0.17332896461336827,
      "grad_norm": 0.13244757056236267,
      "learning_rate": 0.0001,
      "loss": 1.3416,
      "step": 1058
    },
    {
      "epoch": 0.17349279161205766,
      "grad_norm": 0.12845009565353394,
      "learning_rate": 0.0001,
      "loss": 1.395,
      "step": 1059
    },
    {
      "epoch": 0.17365661861074705,
      "grad_norm": 0.16922105848789215,
      "learning_rate": 0.0001,
      "loss": 1.5124,
      "step": 1060
    },
    {
      "epoch": 0.17382044560943644,
      "grad_norm": 0.13284632563591003,
      "learning_rate": 0.0001,
      "loss": 1.4295,
      "step": 1061
    },
    {
      "epoch": 0.17398427260812582,
      "grad_norm": 0.13081838190555573,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 1062
    },
    {
      "epoch": 0.1741480996068152,
      "grad_norm": 0.13468126952648163,
      "learning_rate": 0.0001,
      "loss": 1.4296,
      "step": 1063
    },
    {
      "epoch": 0.1743119266055046,
      "grad_norm": 0.13577158749103546,
      "learning_rate": 0.0001,
      "loss": 1.5244,
      "step": 1064
    },
    {
      "epoch": 0.17447575360419398,
      "grad_norm": 0.1312638372182846,
      "learning_rate": 0.0001,
      "loss": 1.5744,
      "step": 1065
    },
    {
      "epoch": 0.17463958060288334,
      "grad_norm": 0.13969166576862335,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 1066
    },
    {
      "epoch": 0.17480340760157273,
      "grad_norm": 0.13546361029148102,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 1067
    },
    {
      "epoch": 0.17496723460026212,
      "grad_norm": 0.12968793511390686,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 1068
    },
    {
      "epoch": 0.1751310615989515,
      "grad_norm": 0.13104459643363953,
      "learning_rate": 0.0001,
      "loss": 1.4442,
      "step": 1069
    },
    {
      "epoch": 0.1752948885976409,
      "grad_norm": 0.12666799128055573,
      "learning_rate": 0.0001,
      "loss": 1.4323,
      "step": 1070
    },
    {
      "epoch": 0.17545871559633028,
      "grad_norm": 0.1334940493106842,
      "learning_rate": 0.0001,
      "loss": 1.4139,
      "step": 1071
    },
    {
      "epoch": 0.17562254259501967,
      "grad_norm": 0.13696043193340302,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 1072
    },
    {
      "epoch": 0.17578636959370905,
      "grad_norm": 0.13326141238212585,
      "learning_rate": 0.0001,
      "loss": 1.4095,
      "step": 1073
    },
    {
      "epoch": 0.17595019659239844,
      "grad_norm": 0.13667449355125427,
      "learning_rate": 0.0001,
      "loss": 1.3927,
      "step": 1074
    },
    {
      "epoch": 0.1761140235910878,
      "grad_norm": 0.1324305236339569,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 1075
    },
    {
      "epoch": 0.17627785058977719,
      "grad_norm": 0.12715396285057068,
      "learning_rate": 0.0001,
      "loss": 1.4208,
      "step": 1076
    },
    {
      "epoch": 0.17644167758846657,
      "grad_norm": 0.14142371714115143,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 1077
    },
    {
      "epoch": 0.17660550458715596,
      "grad_norm": 0.14664088189601898,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 1078
    },
    {
      "epoch": 0.17676933158584535,
      "grad_norm": 0.13482631742954254,
      "learning_rate": 0.0001,
      "loss": 1.3785,
      "step": 1079
    },
    {
      "epoch": 0.17693315858453473,
      "grad_norm": 0.1382308155298233,
      "learning_rate": 0.0001,
      "loss": 1.485,
      "step": 1080
    },
    {
      "epoch": 0.17709698558322412,
      "grad_norm": 0.13963602483272552,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 1081
    },
    {
      "epoch": 0.1772608125819135,
      "grad_norm": 0.1407417505979538,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 1082
    },
    {
      "epoch": 0.1774246395806029,
      "grad_norm": 0.12424825131893158,
      "learning_rate": 0.0001,
      "loss": 1.4345,
      "step": 1083
    },
    {
      "epoch": 0.17758846657929228,
      "grad_norm": 0.13410243391990662,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 1084
    },
    {
      "epoch": 0.17775229357798164,
      "grad_norm": 0.14197877049446106,
      "learning_rate": 0.0001,
      "loss": 1.6068,
      "step": 1085
    },
    {
      "epoch": 0.17791612057667103,
      "grad_norm": 0.1369401067495346,
      "learning_rate": 0.0001,
      "loss": 1.4707,
      "step": 1086
    },
    {
      "epoch": 0.17807994757536041,
      "grad_norm": 0.12968944013118744,
      "learning_rate": 0.0001,
      "loss": 1.3657,
      "step": 1087
    },
    {
      "epoch": 0.1782437745740498,
      "grad_norm": 0.13015304505825043,
      "learning_rate": 0.0001,
      "loss": 1.5266,
      "step": 1088
    },
    {
      "epoch": 0.1784076015727392,
      "grad_norm": 0.13442043960094452,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 1089
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.13031138479709625,
      "learning_rate": 0.0001,
      "loss": 1.5138,
      "step": 1090
    },
    {
      "epoch": 0.17873525557011796,
      "grad_norm": 0.12648873031139374,
      "learning_rate": 0.0001,
      "loss": 1.5297,
      "step": 1091
    },
    {
      "epoch": 0.17889908256880735,
      "grad_norm": 0.13075079023838043,
      "learning_rate": 0.0001,
      "loss": 1.5276,
      "step": 1092
    },
    {
      "epoch": 0.17906290956749674,
      "grad_norm": 0.1304907351732254,
      "learning_rate": 0.0001,
      "loss": 1.4694,
      "step": 1093
    },
    {
      "epoch": 0.1792267365661861,
      "grad_norm": 0.13130216300487518,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 1094
    },
    {
      "epoch": 0.17939056356487548,
      "grad_norm": 0.14008961617946625,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 1095
    },
    {
      "epoch": 0.17955439056356487,
      "grad_norm": 0.13090933859348297,
      "learning_rate": 0.0001,
      "loss": 1.4976,
      "step": 1096
    },
    {
      "epoch": 0.17971821756225426,
      "grad_norm": 0.12943311035633087,
      "learning_rate": 0.0001,
      "loss": 1.3661,
      "step": 1097
    },
    {
      "epoch": 0.17988204456094364,
      "grad_norm": 0.12928925454616547,
      "learning_rate": 0.0001,
      "loss": 1.4545,
      "step": 1098
    },
    {
      "epoch": 0.18004587155963303,
      "grad_norm": 0.14555864036083221,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 1099
    },
    {
      "epoch": 0.18020969855832242,
      "grad_norm": 0.14300112426280975,
      "learning_rate": 0.0001,
      "loss": 1.566,
      "step": 1100
    },
    {
      "epoch": 0.1803735255570118,
      "grad_norm": 0.13661816716194153,
      "learning_rate": 0.0001,
      "loss": 1.4174,
      "step": 1101
    },
    {
      "epoch": 0.1805373525557012,
      "grad_norm": 0.13011403381824493,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 1102
    },
    {
      "epoch": 0.18070117955439055,
      "grad_norm": 0.1336241066455841,
      "learning_rate": 0.0001,
      "loss": 1.4719,
      "step": 1103
    },
    {
      "epoch": 0.18086500655307994,
      "grad_norm": 0.13280604779720306,
      "learning_rate": 0.0001,
      "loss": 1.6085,
      "step": 1104
    },
    {
      "epoch": 0.18102883355176932,
      "grad_norm": 0.1405499130487442,
      "learning_rate": 0.0001,
      "loss": 1.4025,
      "step": 1105
    },
    {
      "epoch": 0.1811926605504587,
      "grad_norm": 0.1323344111442566,
      "learning_rate": 0.0001,
      "loss": 1.4317,
      "step": 1106
    },
    {
      "epoch": 0.1813564875491481,
      "grad_norm": 0.13078348338603973,
      "learning_rate": 0.0001,
      "loss": 1.4018,
      "step": 1107
    },
    {
      "epoch": 0.18152031454783749,
      "grad_norm": 0.13212835788726807,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 1108
    },
    {
      "epoch": 0.18168414154652687,
      "grad_norm": 0.1362771838903427,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 1109
    },
    {
      "epoch": 0.18184796854521626,
      "grad_norm": 0.1315169334411621,
      "learning_rate": 0.0001,
      "loss": 1.3008,
      "step": 1110
    },
    {
      "epoch": 0.18201179554390565,
      "grad_norm": 0.13282889127731323,
      "learning_rate": 0.0001,
      "loss": 1.4146,
      "step": 1111
    },
    {
      "epoch": 0.182175622542595,
      "grad_norm": 0.1358182728290558,
      "learning_rate": 0.0001,
      "loss": 1.5212,
      "step": 1112
    },
    {
      "epoch": 0.1823394495412844,
      "grad_norm": 0.13243716955184937,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 1113
    },
    {
      "epoch": 0.18250327653997378,
      "grad_norm": 0.1370578557252884,
      "learning_rate": 0.0001,
      "loss": 1.4226,
      "step": 1114
    },
    {
      "epoch": 0.18266710353866317,
      "grad_norm": 0.13979144394397736,
      "learning_rate": 0.0001,
      "loss": 1.4619,
      "step": 1115
    },
    {
      "epoch": 0.18283093053735255,
      "grad_norm": 0.13293658196926117,
      "learning_rate": 0.0001,
      "loss": 1.45,
      "step": 1116
    },
    {
      "epoch": 0.18299475753604194,
      "grad_norm": 0.1357840597629547,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 1117
    },
    {
      "epoch": 0.18315858453473133,
      "grad_norm": 0.13536308705806732,
      "learning_rate": 0.0001,
      "loss": 1.4765,
      "step": 1118
    },
    {
      "epoch": 0.18332241153342071,
      "grad_norm": 0.14107313752174377,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 1119
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 0.1317204385995865,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 1120
    },
    {
      "epoch": 0.1836500655307995,
      "grad_norm": 0.13170745968818665,
      "learning_rate": 0.0001,
      "loss": 1.6213,
      "step": 1121
    },
    {
      "epoch": 0.18381389252948885,
      "grad_norm": 0.13820359110832214,
      "learning_rate": 0.0001,
      "loss": 1.4429,
      "step": 1122
    },
    {
      "epoch": 0.18397771952817824,
      "grad_norm": 0.13258500397205353,
      "learning_rate": 0.0001,
      "loss": 1.5109,
      "step": 1123
    },
    {
      "epoch": 0.18414154652686762,
      "grad_norm": 0.13229259848594666,
      "learning_rate": 0.0001,
      "loss": 1.4485,
      "step": 1124
    },
    {
      "epoch": 0.184305373525557,
      "grad_norm": 0.13898524641990662,
      "learning_rate": 0.0001,
      "loss": 1.4566,
      "step": 1125
    },
    {
      "epoch": 0.1844692005242464,
      "grad_norm": 0.13584959506988525,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 1126
    },
    {
      "epoch": 0.18463302752293578,
      "grad_norm": 0.13885585963726044,
      "learning_rate": 0.0001,
      "loss": 1.4751,
      "step": 1127
    },
    {
      "epoch": 0.18479685452162517,
      "grad_norm": 0.12894929945468903,
      "learning_rate": 0.0001,
      "loss": 1.4771,
      "step": 1128
    },
    {
      "epoch": 0.18496068152031456,
      "grad_norm": 0.12987957894802094,
      "learning_rate": 0.0001,
      "loss": 1.4913,
      "step": 1129
    },
    {
      "epoch": 0.18512450851900394,
      "grad_norm": 0.12785962224006653,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 1130
    },
    {
      "epoch": 0.1852883355176933,
      "grad_norm": 0.14019078016281128,
      "learning_rate": 0.0001,
      "loss": 1.4927,
      "step": 1131
    },
    {
      "epoch": 0.1854521625163827,
      "grad_norm": 0.13319732248783112,
      "learning_rate": 0.0001,
      "loss": 1.4636,
      "step": 1132
    },
    {
      "epoch": 0.18561598951507208,
      "grad_norm": 0.13098987936973572,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 1133
    },
    {
      "epoch": 0.18577981651376146,
      "grad_norm": 0.15145766735076904,
      "learning_rate": 0.0001,
      "loss": 1.3895,
      "step": 1134
    },
    {
      "epoch": 0.18594364351245085,
      "grad_norm": 0.12862655520439148,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 1135
    },
    {
      "epoch": 0.18610747051114024,
      "grad_norm": 0.13998766243457794,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 1136
    },
    {
      "epoch": 0.18627129750982963,
      "grad_norm": 0.1369950920343399,
      "learning_rate": 0.0001,
      "loss": 1.4909,
      "step": 1137
    },
    {
      "epoch": 0.186435124508519,
      "grad_norm": 0.13350030779838562,
      "learning_rate": 0.0001,
      "loss": 1.4468,
      "step": 1138
    },
    {
      "epoch": 0.1865989515072084,
      "grad_norm": 0.1353130340576172,
      "learning_rate": 0.0001,
      "loss": 1.4431,
      "step": 1139
    },
    {
      "epoch": 0.18676277850589776,
      "grad_norm": 0.13228526711463928,
      "learning_rate": 0.0001,
      "loss": 1.3561,
      "step": 1140
    },
    {
      "epoch": 0.18692660550458715,
      "grad_norm": 0.1259518414735794,
      "learning_rate": 0.0001,
      "loss": 1.4241,
      "step": 1141
    },
    {
      "epoch": 0.18709043250327653,
      "grad_norm": 0.1277199685573578,
      "learning_rate": 0.0001,
      "loss": 1.3435,
      "step": 1142
    },
    {
      "epoch": 0.18725425950196592,
      "grad_norm": 0.1333499401807785,
      "learning_rate": 0.0001,
      "loss": 1.4362,
      "step": 1143
    },
    {
      "epoch": 0.1874180865006553,
      "grad_norm": 0.1267353743314743,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 1144
    },
    {
      "epoch": 0.1875819134993447,
      "grad_norm": 0.13085569441318512,
      "learning_rate": 0.0001,
      "loss": 1.3293,
      "step": 1145
    },
    {
      "epoch": 0.18774574049803408,
      "grad_norm": 0.1395970582962036,
      "learning_rate": 0.0001,
      "loss": 1.5399,
      "step": 1146
    },
    {
      "epoch": 0.18790956749672347,
      "grad_norm": 0.13815650343894958,
      "learning_rate": 0.0001,
      "loss": 1.4243,
      "step": 1147
    },
    {
      "epoch": 0.18807339449541285,
      "grad_norm": 0.13011090457439423,
      "learning_rate": 0.0001,
      "loss": 1.3479,
      "step": 1148
    },
    {
      "epoch": 0.18823722149410224,
      "grad_norm": 0.13208620250225067,
      "learning_rate": 0.0001,
      "loss": 1.4795,
      "step": 1149
    },
    {
      "epoch": 0.1884010484927916,
      "grad_norm": 0.13524389266967773,
      "learning_rate": 0.0001,
      "loss": 1.4405,
      "step": 1150
    },
    {
      "epoch": 0.188564875491481,
      "grad_norm": 0.12808232009410858,
      "learning_rate": 0.0001,
      "loss": 1.3812,
      "step": 1151
    },
    {
      "epoch": 0.18872870249017037,
      "grad_norm": 0.14272338151931763,
      "learning_rate": 0.0001,
      "loss": 1.6341,
      "step": 1152
    },
    {
      "epoch": 0.18889252948885976,
      "grad_norm": 0.1286228746175766,
      "learning_rate": 0.0001,
      "loss": 1.4629,
      "step": 1153
    },
    {
      "epoch": 0.18905635648754915,
      "grad_norm": 0.13578060269355774,
      "learning_rate": 0.0001,
      "loss": 1.4727,
      "step": 1154
    },
    {
      "epoch": 0.18922018348623854,
      "grad_norm": 0.13113774359226227,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 1155
    },
    {
      "epoch": 0.18938401048492792,
      "grad_norm": 0.1411672681570053,
      "learning_rate": 0.0001,
      "loss": 1.5828,
      "step": 1156
    },
    {
      "epoch": 0.1895478374836173,
      "grad_norm": 0.1338953673839569,
      "learning_rate": 0.0001,
      "loss": 1.4608,
      "step": 1157
    },
    {
      "epoch": 0.1897116644823067,
      "grad_norm": 0.13650624454021454,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 1158
    },
    {
      "epoch": 0.18987549148099606,
      "grad_norm": 0.13545076549053192,
      "learning_rate": 0.0001,
      "loss": 1.4746,
      "step": 1159
    },
    {
      "epoch": 0.19003931847968544,
      "grad_norm": 0.1349044293165207,
      "learning_rate": 0.0001,
      "loss": 1.4026,
      "step": 1160
    },
    {
      "epoch": 0.19020314547837483,
      "grad_norm": 0.1391029953956604,
      "learning_rate": 0.0001,
      "loss": 1.3136,
      "step": 1161
    },
    {
      "epoch": 0.19036697247706422,
      "grad_norm": 0.13191692531108856,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 1162
    },
    {
      "epoch": 0.1905307994757536,
      "grad_norm": 0.12929551303386688,
      "learning_rate": 0.0001,
      "loss": 1.4726,
      "step": 1163
    },
    {
      "epoch": 0.190694626474443,
      "grad_norm": 0.1321699470281601,
      "learning_rate": 0.0001,
      "loss": 1.39,
      "step": 1164
    },
    {
      "epoch": 0.19085845347313238,
      "grad_norm": 0.1465764045715332,
      "learning_rate": 0.0001,
      "loss": 1.5085,
      "step": 1165
    },
    {
      "epoch": 0.19102228047182176,
      "grad_norm": 0.12846218049526215,
      "learning_rate": 0.0001,
      "loss": 1.3781,
      "step": 1166
    },
    {
      "epoch": 0.19118610747051115,
      "grad_norm": 0.12856018543243408,
      "learning_rate": 0.0001,
      "loss": 1.2938,
      "step": 1167
    },
    {
      "epoch": 0.1913499344692005,
      "grad_norm": 0.13824020326137543,
      "learning_rate": 0.0001,
      "loss": 1.4708,
      "step": 1168
    },
    {
      "epoch": 0.1915137614678899,
      "grad_norm": 0.13236665725708008,
      "learning_rate": 0.0001,
      "loss": 1.4119,
      "step": 1169
    },
    {
      "epoch": 0.19167758846657929,
      "grad_norm": 0.13543397188186646,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 1170
    },
    {
      "epoch": 0.19184141546526867,
      "grad_norm": 0.13141369819641113,
      "learning_rate": 0.0001,
      "loss": 1.4026,
      "step": 1171
    },
    {
      "epoch": 0.19200524246395806,
      "grad_norm": 0.1298513561487198,
      "learning_rate": 0.0001,
      "loss": 1.321,
      "step": 1172
    },
    {
      "epoch": 0.19216906946264745,
      "grad_norm": 0.13754335045814514,
      "learning_rate": 0.0001,
      "loss": 1.6623,
      "step": 1173
    },
    {
      "epoch": 0.19233289646133683,
      "grad_norm": 0.13040773570537567,
      "learning_rate": 0.0001,
      "loss": 1.3383,
      "step": 1174
    },
    {
      "epoch": 0.19249672346002622,
      "grad_norm": 0.1342795044183731,
      "learning_rate": 0.0001,
      "loss": 1.4587,
      "step": 1175
    },
    {
      "epoch": 0.1926605504587156,
      "grad_norm": 0.13794533908367157,
      "learning_rate": 0.0001,
      "loss": 1.4828,
      "step": 1176
    },
    {
      "epoch": 0.192824377457405,
      "grad_norm": 0.1319655179977417,
      "learning_rate": 0.0001,
      "loss": 1.4534,
      "step": 1177
    },
    {
      "epoch": 0.19298820445609435,
      "grad_norm": 0.13661973178386688,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 1178
    },
    {
      "epoch": 0.19315203145478374,
      "grad_norm": 0.12856581807136536,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 1179
    },
    {
      "epoch": 0.19331585845347313,
      "grad_norm": 0.13203103840351105,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 1180
    },
    {
      "epoch": 0.19347968545216251,
      "grad_norm": 0.14129862189292908,
      "learning_rate": 0.0001,
      "loss": 1.4696,
      "step": 1181
    },
    {
      "epoch": 0.1936435124508519,
      "grad_norm": 0.15406079590320587,
      "learning_rate": 0.0001,
      "loss": 1.4744,
      "step": 1182
    },
    {
      "epoch": 0.1938073394495413,
      "grad_norm": 0.1553330421447754,
      "learning_rate": 0.0001,
      "loss": 1.3747,
      "step": 1183
    },
    {
      "epoch": 0.19397116644823068,
      "grad_norm": 0.13692615926265717,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 1184
    },
    {
      "epoch": 0.19413499344692006,
      "grad_norm": 0.13345380127429962,
      "learning_rate": 0.0001,
      "loss": 1.5074,
      "step": 1185
    },
    {
      "epoch": 0.19429882044560945,
      "grad_norm": 0.1286674290895462,
      "learning_rate": 0.0001,
      "loss": 1.3398,
      "step": 1186
    },
    {
      "epoch": 0.1944626474442988,
      "grad_norm": 0.14007093012332916,
      "learning_rate": 0.0001,
      "loss": 1.4753,
      "step": 1187
    },
    {
      "epoch": 0.1946264744429882,
      "grad_norm": 0.130698099732399,
      "learning_rate": 0.0001,
      "loss": 1.3859,
      "step": 1188
    },
    {
      "epoch": 0.19479030144167758,
      "grad_norm": 0.15968690812587738,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 1189
    },
    {
      "epoch": 0.19495412844036697,
      "grad_norm": 0.13254328072071075,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 1190
    },
    {
      "epoch": 0.19511795543905636,
      "grad_norm": 0.13009296357631683,
      "learning_rate": 0.0001,
      "loss": 1.39,
      "step": 1191
    },
    {
      "epoch": 0.19528178243774574,
      "grad_norm": 0.13056106865406036,
      "learning_rate": 0.0001,
      "loss": 1.4289,
      "step": 1192
    },
    {
      "epoch": 0.19544560943643513,
      "grad_norm": 0.13280987739562988,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 1193
    },
    {
      "epoch": 0.19560943643512452,
      "grad_norm": 0.14135167002677917,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 1194
    },
    {
      "epoch": 0.1957732634338139,
      "grad_norm": 0.12770439684391022,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 1195
    },
    {
      "epoch": 0.19593709043250326,
      "grad_norm": 0.13908319175243378,
      "learning_rate": 0.0001,
      "loss": 1.5318,
      "step": 1196
    },
    {
      "epoch": 0.19610091743119265,
      "grad_norm": 0.13284587860107422,
      "learning_rate": 0.0001,
      "loss": 1.4091,
      "step": 1197
    },
    {
      "epoch": 0.19626474442988204,
      "grad_norm": 0.13405296206474304,
      "learning_rate": 0.0001,
      "loss": 1.3898,
      "step": 1198
    },
    {
      "epoch": 0.19642857142857142,
      "grad_norm": 0.16391359269618988,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 1199
    },
    {
      "epoch": 0.1965923984272608,
      "grad_norm": 0.13392792642116547,
      "learning_rate": 0.0001,
      "loss": 1.4934,
      "step": 1200
    },
    {
      "epoch": 0.1967562254259502,
      "grad_norm": 0.1348104178905487,
      "learning_rate": 0.0001,
      "loss": 1.4122,
      "step": 1201
    },
    {
      "epoch": 0.19692005242463959,
      "grad_norm": 0.13676144182682037,
      "learning_rate": 0.0001,
      "loss": 1.4567,
      "step": 1202
    },
    {
      "epoch": 0.19708387942332897,
      "grad_norm": 0.1341894567012787,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 1203
    },
    {
      "epoch": 0.19724770642201836,
      "grad_norm": 0.14481893181800842,
      "learning_rate": 0.0001,
      "loss": 1.4868,
      "step": 1204
    },
    {
      "epoch": 0.19741153342070772,
      "grad_norm": 0.13542009890079498,
      "learning_rate": 0.0001,
      "loss": 1.4117,
      "step": 1205
    },
    {
      "epoch": 0.1975753604193971,
      "grad_norm": 0.1270584613084793,
      "learning_rate": 0.0001,
      "loss": 1.2937,
      "step": 1206
    },
    {
      "epoch": 0.1977391874180865,
      "grad_norm": 0.17376403510570526,
      "learning_rate": 0.0001,
      "loss": 1.4607,
      "step": 1207
    },
    {
      "epoch": 0.19790301441677588,
      "grad_norm": 0.14347633719444275,
      "learning_rate": 0.0001,
      "loss": 1.5299,
      "step": 1208
    },
    {
      "epoch": 0.19806684141546527,
      "grad_norm": 0.13140593469142914,
      "learning_rate": 0.0001,
      "loss": 1.4422,
      "step": 1209
    },
    {
      "epoch": 0.19823066841415465,
      "grad_norm": 0.1349049210548401,
      "learning_rate": 0.0001,
      "loss": 1.3748,
      "step": 1210
    },
    {
      "epoch": 0.19839449541284404,
      "grad_norm": 0.13718320429325104,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 1211
    },
    {
      "epoch": 0.19855832241153343,
      "grad_norm": 0.1304779052734375,
      "learning_rate": 0.0001,
      "loss": 1.3887,
      "step": 1212
    },
    {
      "epoch": 0.19872214941022281,
      "grad_norm": 0.14168348908424377,
      "learning_rate": 0.0001,
      "loss": 1.5596,
      "step": 1213
    },
    {
      "epoch": 0.1988859764089122,
      "grad_norm": 0.14601846039295197,
      "learning_rate": 0.0001,
      "loss": 1.5625,
      "step": 1214
    },
    {
      "epoch": 0.19904980340760156,
      "grad_norm": 0.13818511366844177,
      "learning_rate": 0.0001,
      "loss": 1.389,
      "step": 1215
    },
    {
      "epoch": 0.19921363040629095,
      "grad_norm": 0.14145509898662567,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 1216
    },
    {
      "epoch": 0.19937745740498033,
      "grad_norm": 0.1367284506559372,
      "learning_rate": 0.0001,
      "loss": 1.4333,
      "step": 1217
    },
    {
      "epoch": 0.19954128440366972,
      "grad_norm": 0.14505602419376373,
      "learning_rate": 0.0001,
      "loss": 1.5537,
      "step": 1218
    },
    {
      "epoch": 0.1997051114023591,
      "grad_norm": 0.12982140481472015,
      "learning_rate": 0.0001,
      "loss": 1.437,
      "step": 1219
    },
    {
      "epoch": 0.1998689384010485,
      "grad_norm": 0.1406346559524536,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 1220
    },
    {
      "epoch": 0.20003276539973788,
      "grad_norm": 0.1415746510028839,
      "learning_rate": 0.0001,
      "loss": 1.4442,
      "step": 1221
    },
    {
      "epoch": 0.20019659239842727,
      "grad_norm": 0.13162486255168915,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 1222
    },
    {
      "epoch": 0.20036041939711666,
      "grad_norm": 0.13078632950782776,
      "learning_rate": 0.0001,
      "loss": 1.426,
      "step": 1223
    },
    {
      "epoch": 0.20052424639580602,
      "grad_norm": 0.136398583650589,
      "learning_rate": 0.0001,
      "loss": 1.3768,
      "step": 1224
    },
    {
      "epoch": 0.2006880733944954,
      "grad_norm": 0.12791553139686584,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 1225
    },
    {
      "epoch": 0.2008519003931848,
      "grad_norm": 0.13692991435527802,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 1226
    },
    {
      "epoch": 0.20101572739187418,
      "grad_norm": 0.13194362819194794,
      "learning_rate": 0.0001,
      "loss": 1.4549,
      "step": 1227
    },
    {
      "epoch": 0.20117955439056356,
      "grad_norm": 0.13717131316661835,
      "learning_rate": 0.0001,
      "loss": 1.5743,
      "step": 1228
    },
    {
      "epoch": 0.20134338138925295,
      "grad_norm": 0.13729619979858398,
      "learning_rate": 0.0001,
      "loss": 1.3404,
      "step": 1229
    },
    {
      "epoch": 0.20150720838794234,
      "grad_norm": 0.13153815269470215,
      "learning_rate": 0.0001,
      "loss": 1.3151,
      "step": 1230
    },
    {
      "epoch": 0.20167103538663173,
      "grad_norm": 0.13249318301677704,
      "learning_rate": 0.0001,
      "loss": 1.4614,
      "step": 1231
    },
    {
      "epoch": 0.2018348623853211,
      "grad_norm": 0.1341630220413208,
      "learning_rate": 0.0001,
      "loss": 1.4976,
      "step": 1232
    },
    {
      "epoch": 0.20199868938401047,
      "grad_norm": 0.14622019231319427,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 1233
    },
    {
      "epoch": 0.20216251638269986,
      "grad_norm": 0.13434013724327087,
      "learning_rate": 0.0001,
      "loss": 1.5371,
      "step": 1234
    },
    {
      "epoch": 0.20232634338138925,
      "grad_norm": 0.14392277598381042,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 1235
    },
    {
      "epoch": 0.20249017038007863,
      "grad_norm": 0.14191731810569763,
      "learning_rate": 0.0001,
      "loss": 1.5487,
      "step": 1236
    },
    {
      "epoch": 0.20265399737876802,
      "grad_norm": 0.1336645931005478,
      "learning_rate": 0.0001,
      "loss": 1.3828,
      "step": 1237
    },
    {
      "epoch": 0.2028178243774574,
      "grad_norm": 0.14240789413452148,
      "learning_rate": 0.0001,
      "loss": 1.5068,
      "step": 1238
    },
    {
      "epoch": 0.2029816513761468,
      "grad_norm": 0.12849019467830658,
      "learning_rate": 0.0001,
      "loss": 1.3897,
      "step": 1239
    },
    {
      "epoch": 0.20314547837483618,
      "grad_norm": 0.1314997375011444,
      "learning_rate": 0.0001,
      "loss": 1.4438,
      "step": 1240
    },
    {
      "epoch": 0.20330930537352557,
      "grad_norm": 0.13564425706863403,
      "learning_rate": 0.0001,
      "loss": 1.457,
      "step": 1241
    },
    {
      "epoch": 0.20347313237221495,
      "grad_norm": 0.14633633196353912,
      "learning_rate": 0.0001,
      "loss": 1.5139,
      "step": 1242
    },
    {
      "epoch": 0.2036369593709043,
      "grad_norm": 0.13303284347057343,
      "learning_rate": 0.0001,
      "loss": 1.5087,
      "step": 1243
    },
    {
      "epoch": 0.2038007863695937,
      "grad_norm": 0.1403651237487793,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 1244
    },
    {
      "epoch": 0.2039646133682831,
      "grad_norm": 0.12879008054733276,
      "learning_rate": 0.0001,
      "loss": 1.3451,
      "step": 1245
    },
    {
      "epoch": 0.20412844036697247,
      "grad_norm": 0.12900450825691223,
      "learning_rate": 0.0001,
      "loss": 1.3495,
      "step": 1246
    },
    {
      "epoch": 0.20429226736566186,
      "grad_norm": 0.13349464535713196,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 1247
    },
    {
      "epoch": 0.20445609436435125,
      "grad_norm": 0.13309772312641144,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 1248
    },
    {
      "epoch": 0.20461992136304064,
      "grad_norm": 0.13345350325107574,
      "learning_rate": 0.0001,
      "loss": 1.4912,
      "step": 1249
    },
    {
      "epoch": 0.20478374836173002,
      "grad_norm": 0.1339840441942215,
      "learning_rate": 0.0001,
      "loss": 1.4639,
      "step": 1250
    },
    {
      "epoch": 0.2049475753604194,
      "grad_norm": 0.1412966251373291,
      "learning_rate": 0.0001,
      "loss": 1.489,
      "step": 1251
    },
    {
      "epoch": 0.20511140235910877,
      "grad_norm": 0.1435202956199646,
      "learning_rate": 0.0001,
      "loss": 1.4827,
      "step": 1252
    },
    {
      "epoch": 0.20527522935779816,
      "grad_norm": 0.1327572464942932,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 1253
    },
    {
      "epoch": 0.20543905635648754,
      "grad_norm": 0.13939295709133148,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 1254
    },
    {
      "epoch": 0.20560288335517693,
      "grad_norm": 0.14035841822624207,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 1255
    },
    {
      "epoch": 0.20576671035386632,
      "grad_norm": 0.1361629068851471,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 1256
    },
    {
      "epoch": 0.2059305373525557,
      "grad_norm": 0.13688170909881592,
      "learning_rate": 0.0001,
      "loss": 1.4941,
      "step": 1257
    },
    {
      "epoch": 0.2060943643512451,
      "grad_norm": 0.1429302841424942,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 1258
    },
    {
      "epoch": 0.20625819134993448,
      "grad_norm": 0.13294102251529694,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 1259
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 0.14095330238342285,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 1260
    },
    {
      "epoch": 0.20658584534731322,
      "grad_norm": 0.14034394919872284,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 1261
    },
    {
      "epoch": 0.2067496723460026,
      "grad_norm": 0.1465233862400055,
      "learning_rate": 0.0001,
      "loss": 1.4393,
      "step": 1262
    },
    {
      "epoch": 0.206913499344692,
      "grad_norm": 0.13921520113945007,
      "learning_rate": 0.0001,
      "loss": 1.4221,
      "step": 1263
    },
    {
      "epoch": 0.20707732634338138,
      "grad_norm": 0.13950198888778687,
      "learning_rate": 0.0001,
      "loss": 1.4469,
      "step": 1264
    },
    {
      "epoch": 0.20724115334207077,
      "grad_norm": 0.1396608203649521,
      "learning_rate": 0.0001,
      "loss": 1.4249,
      "step": 1265
    },
    {
      "epoch": 0.20740498034076016,
      "grad_norm": 0.13926605880260468,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 1266
    },
    {
      "epoch": 0.20756880733944955,
      "grad_norm": 0.13196703791618347,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 1267
    },
    {
      "epoch": 0.20773263433813893,
      "grad_norm": 0.13104552030563354,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 1268
    },
    {
      "epoch": 0.20789646133682832,
      "grad_norm": 0.128145232796669,
      "learning_rate": 0.0001,
      "loss": 1.4321,
      "step": 1269
    },
    {
      "epoch": 0.2080602883355177,
      "grad_norm": 0.1395496428012848,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 1270
    },
    {
      "epoch": 0.20822411533420707,
      "grad_norm": 0.14076267182826996,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 1271
    },
    {
      "epoch": 0.20838794233289645,
      "grad_norm": 0.1419232040643692,
      "learning_rate": 0.0001,
      "loss": 1.4339,
      "step": 1272
    },
    {
      "epoch": 0.20855176933158584,
      "grad_norm": 0.14380769431591034,
      "learning_rate": 0.0001,
      "loss": 1.4525,
      "step": 1273
    },
    {
      "epoch": 0.20871559633027523,
      "grad_norm": 0.13371752202510834,
      "learning_rate": 0.0001,
      "loss": 1.4024,
      "step": 1274
    },
    {
      "epoch": 0.20887942332896461,
      "grad_norm": 0.13915200531482697,
      "learning_rate": 0.0001,
      "loss": 1.4934,
      "step": 1275
    },
    {
      "epoch": 0.209043250327654,
      "grad_norm": 0.13129346072673798,
      "learning_rate": 0.0001,
      "loss": 1.3742,
      "step": 1276
    },
    {
      "epoch": 0.2092070773263434,
      "grad_norm": 0.142393559217453,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 1277
    },
    {
      "epoch": 0.20937090432503278,
      "grad_norm": 0.1356535404920578,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 1278
    },
    {
      "epoch": 0.20953473132372216,
      "grad_norm": 0.13274087011814117,
      "learning_rate": 0.0001,
      "loss": 1.3502,
      "step": 1279
    },
    {
      "epoch": 0.20969855832241152,
      "grad_norm": 0.14267927408218384,
      "learning_rate": 0.0001,
      "loss": 1.4418,
      "step": 1280
    },
    {
      "epoch": 0.2098623853211009,
      "grad_norm": 0.12689326703548431,
      "learning_rate": 0.0001,
      "loss": 1.307,
      "step": 1281
    },
    {
      "epoch": 0.2100262123197903,
      "grad_norm": 0.1329113394021988,
      "learning_rate": 0.0001,
      "loss": 1.4136,
      "step": 1282
    },
    {
      "epoch": 0.21019003931847968,
      "grad_norm": 0.13643690943717957,
      "learning_rate": 0.0001,
      "loss": 1.4268,
      "step": 1283
    },
    {
      "epoch": 0.21035386631716907,
      "grad_norm": 0.1485663205385208,
      "learning_rate": 0.0001,
      "loss": 1.3784,
      "step": 1284
    },
    {
      "epoch": 0.21051769331585846,
      "grad_norm": 0.13952384889125824,
      "learning_rate": 0.0001,
      "loss": 1.4926,
      "step": 1285
    },
    {
      "epoch": 0.21068152031454784,
      "grad_norm": 0.1298101842403412,
      "learning_rate": 0.0001,
      "loss": 1.2845,
      "step": 1286
    },
    {
      "epoch": 0.21084534731323723,
      "grad_norm": 0.14213290810585022,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 1287
    },
    {
      "epoch": 0.21100917431192662,
      "grad_norm": 0.13690057396888733,
      "learning_rate": 0.0001,
      "loss": 1.4445,
      "step": 1288
    },
    {
      "epoch": 0.21117300131061598,
      "grad_norm": 0.14599868655204773,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 1289
    },
    {
      "epoch": 0.21133682830930536,
      "grad_norm": 0.1367422342300415,
      "learning_rate": 0.0001,
      "loss": 1.4605,
      "step": 1290
    },
    {
      "epoch": 0.21150065530799475,
      "grad_norm": 0.13341045379638672,
      "learning_rate": 0.0001,
      "loss": 1.3628,
      "step": 1291
    },
    {
      "epoch": 0.21166448230668414,
      "grad_norm": 0.13219159841537476,
      "learning_rate": 0.0001,
      "loss": 1.423,
      "step": 1292
    },
    {
      "epoch": 0.21182830930537352,
      "grad_norm": 0.1352115422487259,
      "learning_rate": 0.0001,
      "loss": 1.4586,
      "step": 1293
    },
    {
      "epoch": 0.2119921363040629,
      "grad_norm": 0.15225926041603088,
      "learning_rate": 0.0001,
      "loss": 1.4042,
      "step": 1294
    },
    {
      "epoch": 0.2121559633027523,
      "grad_norm": 0.1353004276752472,
      "learning_rate": 0.0001,
      "loss": 1.4209,
      "step": 1295
    },
    {
      "epoch": 0.21231979030144169,
      "grad_norm": 0.13987170159816742,
      "learning_rate": 0.0001,
      "loss": 1.4883,
      "step": 1296
    },
    {
      "epoch": 0.21248361730013107,
      "grad_norm": 0.13903464376926422,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 1297
    },
    {
      "epoch": 0.21264744429882046,
      "grad_norm": 0.14167137444019318,
      "learning_rate": 0.0001,
      "loss": 1.4047,
      "step": 1298
    },
    {
      "epoch": 0.21281127129750982,
      "grad_norm": 0.13759171962738037,
      "learning_rate": 0.0001,
      "loss": 1.3961,
      "step": 1299
    },
    {
      "epoch": 0.2129750982961992,
      "grad_norm": 0.1359858363866806,
      "learning_rate": 0.0001,
      "loss": 1.3766,
      "step": 1300
    },
    {
      "epoch": 0.2131389252948886,
      "grad_norm": 0.14181122183799744,
      "learning_rate": 0.0001,
      "loss": 1.4325,
      "step": 1301
    },
    {
      "epoch": 0.21330275229357798,
      "grad_norm": 0.14149852097034454,
      "learning_rate": 0.0001,
      "loss": 1.4452,
      "step": 1302
    },
    {
      "epoch": 0.21346657929226737,
      "grad_norm": 0.14338012039661407,
      "learning_rate": 0.0001,
      "loss": 1.4145,
      "step": 1303
    },
    {
      "epoch": 0.21363040629095675,
      "grad_norm": 0.13402484357357025,
      "learning_rate": 0.0001,
      "loss": 1.3893,
      "step": 1304
    },
    {
      "epoch": 0.21379423328964614,
      "grad_norm": 0.1326330453157425,
      "learning_rate": 0.0001,
      "loss": 1.4499,
      "step": 1305
    },
    {
      "epoch": 0.21395806028833553,
      "grad_norm": 0.13813242316246033,
      "learning_rate": 0.0001,
      "loss": 1.4973,
      "step": 1306
    },
    {
      "epoch": 0.21412188728702491,
      "grad_norm": 0.16084395349025726,
      "learning_rate": 0.0001,
      "loss": 1.4242,
      "step": 1307
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 0.13317394256591797,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 1308
    },
    {
      "epoch": 0.21444954128440366,
      "grad_norm": 0.14121295511722565,
      "learning_rate": 0.0001,
      "loss": 1.4929,
      "step": 1309
    },
    {
      "epoch": 0.21461336828309305,
      "grad_norm": 0.14856991171836853,
      "learning_rate": 0.0001,
      "loss": 1.3886,
      "step": 1310
    },
    {
      "epoch": 0.21477719528178243,
      "grad_norm": 0.15215080976486206,
      "learning_rate": 0.0001,
      "loss": 1.5066,
      "step": 1311
    },
    {
      "epoch": 0.21494102228047182,
      "grad_norm": 0.13939781486988068,
      "learning_rate": 0.0001,
      "loss": 1.3555,
      "step": 1312
    },
    {
      "epoch": 0.2151048492791612,
      "grad_norm": 0.13751883804798126,
      "learning_rate": 0.0001,
      "loss": 1.3311,
      "step": 1313
    },
    {
      "epoch": 0.2152686762778506,
      "grad_norm": 0.13903993368148804,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 1314
    },
    {
      "epoch": 0.21543250327653998,
      "grad_norm": 0.14972324669361115,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 1315
    },
    {
      "epoch": 0.21559633027522937,
      "grad_norm": 0.13693812489509583,
      "learning_rate": 0.0001,
      "loss": 1.4441,
      "step": 1316
    },
    {
      "epoch": 0.21576015727391873,
      "grad_norm": 0.13806460797786713,
      "learning_rate": 0.0001,
      "loss": 1.5461,
      "step": 1317
    },
    {
      "epoch": 0.21592398427260812,
      "grad_norm": 0.14102841913700104,
      "learning_rate": 0.0001,
      "loss": 1.4689,
      "step": 1318
    },
    {
      "epoch": 0.2160878112712975,
      "grad_norm": 0.1362563967704773,
      "learning_rate": 0.0001,
      "loss": 1.4206,
      "step": 1319
    },
    {
      "epoch": 0.2162516382699869,
      "grad_norm": 0.13347607851028442,
      "learning_rate": 0.0001,
      "loss": 1.2938,
      "step": 1320
    },
    {
      "epoch": 0.21641546526867628,
      "grad_norm": 0.14789579808712006,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 1321
    },
    {
      "epoch": 0.21657929226736566,
      "grad_norm": 0.12767274677753448,
      "learning_rate": 0.0001,
      "loss": 1.3321,
      "step": 1322
    },
    {
      "epoch": 0.21674311926605505,
      "grad_norm": 0.13166604936122894,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 1323
    },
    {
      "epoch": 0.21690694626474444,
      "grad_norm": 0.13658079504966736,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 1324
    },
    {
      "epoch": 0.21707077326343382,
      "grad_norm": 0.13495700061321259,
      "learning_rate": 0.0001,
      "loss": 1.4297,
      "step": 1325
    },
    {
      "epoch": 0.21723460026212318,
      "grad_norm": 0.14436089992523193,
      "learning_rate": 0.0001,
      "loss": 1.469,
      "step": 1326
    },
    {
      "epoch": 0.21739842726081257,
      "grad_norm": 0.1370379626750946,
      "learning_rate": 0.0001,
      "loss": 1.4751,
      "step": 1327
    },
    {
      "epoch": 0.21756225425950196,
      "grad_norm": 0.13644836843013763,
      "learning_rate": 0.0001,
      "loss": 1.4312,
      "step": 1328
    },
    {
      "epoch": 0.21772608125819135,
      "grad_norm": 0.13595281541347504,
      "learning_rate": 0.0001,
      "loss": 1.4683,
      "step": 1329
    },
    {
      "epoch": 0.21788990825688073,
      "grad_norm": 0.14060191810131073,
      "learning_rate": 0.0001,
      "loss": 1.4461,
      "step": 1330
    },
    {
      "epoch": 0.21805373525557012,
      "grad_norm": 0.13694505393505096,
      "learning_rate": 0.0001,
      "loss": 1.3706,
      "step": 1331
    },
    {
      "epoch": 0.2182175622542595,
      "grad_norm": 0.13774947822093964,
      "learning_rate": 0.0001,
      "loss": 1.4284,
      "step": 1332
    },
    {
      "epoch": 0.2183813892529489,
      "grad_norm": 0.13363972306251526,
      "learning_rate": 0.0001,
      "loss": 1.4856,
      "step": 1333
    },
    {
      "epoch": 0.21854521625163828,
      "grad_norm": 0.1323726773262024,
      "learning_rate": 0.0001,
      "loss": 1.4328,
      "step": 1334
    },
    {
      "epoch": 0.21870904325032767,
      "grad_norm": 0.12876828014850616,
      "learning_rate": 0.0001,
      "loss": 1.3524,
      "step": 1335
    },
    {
      "epoch": 0.21887287024901703,
      "grad_norm": 0.13660110533237457,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 1336
    },
    {
      "epoch": 0.2190366972477064,
      "grad_norm": 0.12864382565021515,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 1337
    },
    {
      "epoch": 0.2192005242463958,
      "grad_norm": 0.13516047596931458,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 1338
    },
    {
      "epoch": 0.2193643512450852,
      "grad_norm": 0.13409137725830078,
      "learning_rate": 0.0001,
      "loss": 1.4573,
      "step": 1339
    },
    {
      "epoch": 0.21952817824377457,
      "grad_norm": 0.13804766535758972,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 1340
    },
    {
      "epoch": 0.21969200524246396,
      "grad_norm": 0.13269712030887604,
      "learning_rate": 0.0001,
      "loss": 1.4864,
      "step": 1341
    },
    {
      "epoch": 0.21985583224115335,
      "grad_norm": 0.13185594975948334,
      "learning_rate": 0.0001,
      "loss": 1.448,
      "step": 1342
    },
    {
      "epoch": 0.22001965923984274,
      "grad_norm": 0.1517920047044754,
      "learning_rate": 0.0001,
      "loss": 1.4726,
      "step": 1343
    },
    {
      "epoch": 0.22018348623853212,
      "grad_norm": 0.13250653445720673,
      "learning_rate": 0.0001,
      "loss": 1.4245,
      "step": 1344
    },
    {
      "epoch": 0.22034731323722148,
      "grad_norm": 0.14555874466896057,
      "learning_rate": 0.0001,
      "loss": 1.4517,
      "step": 1345
    },
    {
      "epoch": 0.22051114023591087,
      "grad_norm": 0.14072683453559875,
      "learning_rate": 0.0001,
      "loss": 1.5458,
      "step": 1346
    },
    {
      "epoch": 0.22067496723460026,
      "grad_norm": 0.1440013200044632,
      "learning_rate": 0.0001,
      "loss": 1.4389,
      "step": 1347
    },
    {
      "epoch": 0.22083879423328964,
      "grad_norm": 0.15016330778598785,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 1348
    },
    {
      "epoch": 0.22100262123197903,
      "grad_norm": 0.13981382548809052,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 1349
    },
    {
      "epoch": 0.22116644823066842,
      "grad_norm": 0.14187388122081757,
      "learning_rate": 0.0001,
      "loss": 1.5155,
      "step": 1350
    },
    {
      "epoch": 0.2213302752293578,
      "grad_norm": 0.13154274225234985,
      "learning_rate": 0.0001,
      "loss": 1.4893,
      "step": 1351
    },
    {
      "epoch": 0.2214941022280472,
      "grad_norm": 0.13985154032707214,
      "learning_rate": 0.0001,
      "loss": 1.4474,
      "step": 1352
    },
    {
      "epoch": 0.22165792922673658,
      "grad_norm": 0.14237110316753387,
      "learning_rate": 0.0001,
      "loss": 1.3323,
      "step": 1353
    },
    {
      "epoch": 0.22182175622542594,
      "grad_norm": 0.13007678091526031,
      "learning_rate": 0.0001,
      "loss": 1.4654,
      "step": 1354
    },
    {
      "epoch": 0.22198558322411532,
      "grad_norm": 0.13674074411392212,
      "learning_rate": 0.0001,
      "loss": 1.4445,
      "step": 1355
    },
    {
      "epoch": 0.2221494102228047,
      "grad_norm": 0.14455245435237885,
      "learning_rate": 0.0001,
      "loss": 1.4919,
      "step": 1356
    },
    {
      "epoch": 0.2223132372214941,
      "grad_norm": 0.13116657733917236,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 1357
    },
    {
      "epoch": 0.22247706422018348,
      "grad_norm": 0.14610256254673004,
      "learning_rate": 0.0001,
      "loss": 1.4937,
      "step": 1358
    },
    {
      "epoch": 0.22264089121887287,
      "grad_norm": 0.1368582844734192,
      "learning_rate": 0.0001,
      "loss": 1.438,
      "step": 1359
    },
    {
      "epoch": 0.22280471821756226,
      "grad_norm": 0.12760044634342194,
      "learning_rate": 0.0001,
      "loss": 1.3695,
      "step": 1360
    },
    {
      "epoch": 0.22296854521625165,
      "grad_norm": 0.13207577168941498,
      "learning_rate": 0.0001,
      "loss": 1.4649,
      "step": 1361
    },
    {
      "epoch": 0.22313237221494103,
      "grad_norm": 0.14041396975517273,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 1362
    },
    {
      "epoch": 0.22329619921363042,
      "grad_norm": 0.13654224574565887,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 1363
    },
    {
      "epoch": 0.22346002621231978,
      "grad_norm": 0.1417308747768402,
      "learning_rate": 0.0001,
      "loss": 1.4751,
      "step": 1364
    },
    {
      "epoch": 0.22362385321100917,
      "grad_norm": 0.13906514644622803,
      "learning_rate": 0.0001,
      "loss": 1.5627,
      "step": 1365
    },
    {
      "epoch": 0.22378768020969855,
      "grad_norm": 0.12665614485740662,
      "learning_rate": 0.0001,
      "loss": 1.3388,
      "step": 1366
    },
    {
      "epoch": 0.22395150720838794,
      "grad_norm": 0.12871922552585602,
      "learning_rate": 0.0001,
      "loss": 1.4553,
      "step": 1367
    },
    {
      "epoch": 0.22411533420707733,
      "grad_norm": 0.12979231774806976,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 1368
    },
    {
      "epoch": 0.22427916120576671,
      "grad_norm": 0.13382403552532196,
      "learning_rate": 0.0001,
      "loss": 1.4018,
      "step": 1369
    },
    {
      "epoch": 0.2244429882044561,
      "grad_norm": 0.15614357590675354,
      "learning_rate": 0.0001,
      "loss": 1.4325,
      "step": 1370
    },
    {
      "epoch": 0.2246068152031455,
      "grad_norm": 0.13874006271362305,
      "learning_rate": 0.0001,
      "loss": 1.4486,
      "step": 1371
    },
    {
      "epoch": 0.22477064220183487,
      "grad_norm": 0.14173908531665802,
      "learning_rate": 0.0001,
      "loss": 1.4374,
      "step": 1372
    },
    {
      "epoch": 0.22493446920052423,
      "grad_norm": 0.14517882466316223,
      "learning_rate": 0.0001,
      "loss": 1.4602,
      "step": 1373
    },
    {
      "epoch": 0.22509829619921362,
      "grad_norm": 0.1352497935295105,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 1374
    },
    {
      "epoch": 0.225262123197903,
      "grad_norm": 0.1397503763437271,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 1375
    },
    {
      "epoch": 0.2254259501965924,
      "grad_norm": 0.1324470490217209,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 1376
    },
    {
      "epoch": 0.22558977719528178,
      "grad_norm": 0.13713940978050232,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 1377
    },
    {
      "epoch": 0.22575360419397117,
      "grad_norm": 0.13393084704875946,
      "learning_rate": 0.0001,
      "loss": 1.4772,
      "step": 1378
    },
    {
      "epoch": 0.22591743119266056,
      "grad_norm": 0.135294571518898,
      "learning_rate": 0.0001,
      "loss": 1.4093,
      "step": 1379
    },
    {
      "epoch": 0.22608125819134994,
      "grad_norm": 0.13887521624565125,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 1380
    },
    {
      "epoch": 0.22624508519003933,
      "grad_norm": 0.13540856540203094,
      "learning_rate": 0.0001,
      "loss": 1.4466,
      "step": 1381
    },
    {
      "epoch": 0.2264089121887287,
      "grad_norm": 0.13495634496212006,
      "learning_rate": 0.0001,
      "loss": 1.432,
      "step": 1382
    },
    {
      "epoch": 0.22657273918741808,
      "grad_norm": 0.13949793577194214,
      "learning_rate": 0.0001,
      "loss": 1.4075,
      "step": 1383
    },
    {
      "epoch": 0.22673656618610746,
      "grad_norm": 0.13397884368896484,
      "learning_rate": 0.0001,
      "loss": 1.4165,
      "step": 1384
    },
    {
      "epoch": 0.22690039318479685,
      "grad_norm": 0.1284884810447693,
      "learning_rate": 0.0001,
      "loss": 1.2922,
      "step": 1385
    },
    {
      "epoch": 0.22706422018348624,
      "grad_norm": 0.14472176134586334,
      "learning_rate": 0.0001,
      "loss": 1.4776,
      "step": 1386
    },
    {
      "epoch": 0.22722804718217562,
      "grad_norm": 0.1433040201663971,
      "learning_rate": 0.0001,
      "loss": 1.4869,
      "step": 1387
    },
    {
      "epoch": 0.227391874180865,
      "grad_norm": 0.14214088022708893,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 1388
    },
    {
      "epoch": 0.2275557011795544,
      "grad_norm": 0.13230592012405396,
      "learning_rate": 0.0001,
      "loss": 1.5357,
      "step": 1389
    },
    {
      "epoch": 0.22771952817824379,
      "grad_norm": 0.1370100975036621,
      "learning_rate": 0.0001,
      "loss": 1.5322,
      "step": 1390
    },
    {
      "epoch": 0.22788335517693317,
      "grad_norm": 0.1402907371520996,
      "learning_rate": 0.0001,
      "loss": 1.4514,
      "step": 1391
    },
    {
      "epoch": 0.22804718217562253,
      "grad_norm": 0.1313382089138031,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 1392
    },
    {
      "epoch": 0.22821100917431192,
      "grad_norm": 0.14005379378795624,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 1393
    },
    {
      "epoch": 0.2283748361730013,
      "grad_norm": 0.1383068561553955,
      "learning_rate": 0.0001,
      "loss": 1.3816,
      "step": 1394
    },
    {
      "epoch": 0.2285386631716907,
      "grad_norm": 0.13818033039569855,
      "learning_rate": 0.0001,
      "loss": 1.4274,
      "step": 1395
    },
    {
      "epoch": 0.22870249017038008,
      "grad_norm": 0.1472204327583313,
      "learning_rate": 0.0001,
      "loss": 1.5189,
      "step": 1396
    },
    {
      "epoch": 0.22886631716906947,
      "grad_norm": 0.12967702746391296,
      "learning_rate": 0.0001,
      "loss": 1.3633,
      "step": 1397
    },
    {
      "epoch": 0.22903014416775885,
      "grad_norm": 0.13917693495750427,
      "learning_rate": 0.0001,
      "loss": 1.5429,
      "step": 1398
    },
    {
      "epoch": 0.22919397116644824,
      "grad_norm": 0.13525858521461487,
      "learning_rate": 0.0001,
      "loss": 1.4359,
      "step": 1399
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 0.13839469850063324,
      "learning_rate": 0.0001,
      "loss": 1.4368,
      "step": 1400
    },
    {
      "epoch": 0.229521625163827,
      "grad_norm": 0.13541020452976227,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 1401
    },
    {
      "epoch": 0.22968545216251637,
      "grad_norm": 0.13844071328639984,
      "learning_rate": 0.0001,
      "loss": 1.5,
      "step": 1402
    },
    {
      "epoch": 0.22984927916120576,
      "grad_norm": 0.13042961061000824,
      "learning_rate": 0.0001,
      "loss": 1.3804,
      "step": 1403
    },
    {
      "epoch": 0.23001310615989515,
      "grad_norm": 0.13901501893997192,
      "learning_rate": 0.0001,
      "loss": 1.5422,
      "step": 1404
    },
    {
      "epoch": 0.23017693315858453,
      "grad_norm": 0.1413167119026184,
      "learning_rate": 0.0001,
      "loss": 1.5077,
      "step": 1405
    },
    {
      "epoch": 0.23034076015727392,
      "grad_norm": 0.14506112039089203,
      "learning_rate": 0.0001,
      "loss": 1.4799,
      "step": 1406
    },
    {
      "epoch": 0.2305045871559633,
      "grad_norm": 0.1390533149242401,
      "learning_rate": 0.0001,
      "loss": 1.4885,
      "step": 1407
    },
    {
      "epoch": 0.2306684141546527,
      "grad_norm": 0.13211926817893982,
      "learning_rate": 0.0001,
      "loss": 1.3577,
      "step": 1408
    },
    {
      "epoch": 0.23083224115334208,
      "grad_norm": 0.14011211693286896,
      "learning_rate": 0.0001,
      "loss": 1.4502,
      "step": 1409
    },
    {
      "epoch": 0.23099606815203144,
      "grad_norm": 0.13427332043647766,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 1410
    },
    {
      "epoch": 0.23115989515072083,
      "grad_norm": 0.13233742117881775,
      "learning_rate": 0.0001,
      "loss": 1.4485,
      "step": 1411
    },
    {
      "epoch": 0.23132372214941022,
      "grad_norm": 0.1297065168619156,
      "learning_rate": 0.0001,
      "loss": 1.4689,
      "step": 1412
    },
    {
      "epoch": 0.2314875491480996,
      "grad_norm": 0.14859548211097717,
      "learning_rate": 0.0001,
      "loss": 1.3496,
      "step": 1413
    },
    {
      "epoch": 0.231651376146789,
      "grad_norm": 0.12970107793807983,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 1414
    },
    {
      "epoch": 0.23181520314547838,
      "grad_norm": 0.14277444779872894,
      "learning_rate": 0.0001,
      "loss": 1.4954,
      "step": 1415
    },
    {
      "epoch": 0.23197903014416776,
      "grad_norm": 0.14695708453655243,
      "learning_rate": 0.0001,
      "loss": 1.365,
      "step": 1416
    },
    {
      "epoch": 0.23214285714285715,
      "grad_norm": 0.1348785161972046,
      "learning_rate": 0.0001,
      "loss": 1.5089,
      "step": 1417
    },
    {
      "epoch": 0.23230668414154654,
      "grad_norm": 0.13616888225078583,
      "learning_rate": 0.0001,
      "loss": 1.408,
      "step": 1418
    },
    {
      "epoch": 0.2324705111402359,
      "grad_norm": 0.14663852751255035,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 1419
    },
    {
      "epoch": 0.23263433813892528,
      "grad_norm": 0.1358736753463745,
      "learning_rate": 0.0001,
      "loss": 1.3898,
      "step": 1420
    },
    {
      "epoch": 0.23279816513761467,
      "grad_norm": 0.1402503103017807,
      "learning_rate": 0.0001,
      "loss": 1.4967,
      "step": 1421
    },
    {
      "epoch": 0.23296199213630406,
      "grad_norm": 0.1389833688735962,
      "learning_rate": 0.0001,
      "loss": 1.4278,
      "step": 1422
    },
    {
      "epoch": 0.23312581913499344,
      "grad_norm": 0.13778497278690338,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 1423
    },
    {
      "epoch": 0.23328964613368283,
      "grad_norm": 0.14516301453113556,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 1424
    },
    {
      "epoch": 0.23345347313237222,
      "grad_norm": 0.13570809364318848,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 1425
    },
    {
      "epoch": 0.2336173001310616,
      "grad_norm": 0.13807594776153564,
      "learning_rate": 0.0001,
      "loss": 1.4602,
      "step": 1426
    },
    {
      "epoch": 0.233781127129751,
      "grad_norm": 0.14611394703388214,
      "learning_rate": 0.0001,
      "loss": 1.4714,
      "step": 1427
    },
    {
      "epoch": 0.23394495412844038,
      "grad_norm": 0.13596495985984802,
      "learning_rate": 0.0001,
      "loss": 1.3632,
      "step": 1428
    },
    {
      "epoch": 0.23410878112712974,
      "grad_norm": 0.13584259152412415,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 1429
    },
    {
      "epoch": 0.23427260812581913,
      "grad_norm": 0.13976898789405823,
      "learning_rate": 0.0001,
      "loss": 1.5413,
      "step": 1430
    },
    {
      "epoch": 0.2344364351245085,
      "grad_norm": 0.13303270936012268,
      "learning_rate": 0.0001,
      "loss": 1.4024,
      "step": 1431
    },
    {
      "epoch": 0.2346002621231979,
      "grad_norm": 0.1382950395345688,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 1432
    },
    {
      "epoch": 0.2347640891218873,
      "grad_norm": 0.1444646716117859,
      "learning_rate": 0.0001,
      "loss": 1.5451,
      "step": 1433
    },
    {
      "epoch": 0.23492791612057667,
      "grad_norm": 0.1351502537727356,
      "learning_rate": 0.0001,
      "loss": 1.4358,
      "step": 1434
    },
    {
      "epoch": 0.23509174311926606,
      "grad_norm": 0.13853132724761963,
      "learning_rate": 0.0001,
      "loss": 1.5042,
      "step": 1435
    },
    {
      "epoch": 0.23525557011795545,
      "grad_norm": 0.1416371464729309,
      "learning_rate": 0.0001,
      "loss": 1.4677,
      "step": 1436
    },
    {
      "epoch": 0.23541939711664484,
      "grad_norm": 0.15358391404151917,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 1437
    },
    {
      "epoch": 0.2355832241153342,
      "grad_norm": 0.14408761262893677,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 1438
    },
    {
      "epoch": 0.23574705111402358,
      "grad_norm": 0.14212678372859955,
      "learning_rate": 0.0001,
      "loss": 1.5354,
      "step": 1439
    },
    {
      "epoch": 0.23591087811271297,
      "grad_norm": 0.13998210430145264,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 1440
    },
    {
      "epoch": 0.23607470511140236,
      "grad_norm": 0.1367165446281433,
      "learning_rate": 0.0001,
      "loss": 1.556,
      "step": 1441
    },
    {
      "epoch": 0.23623853211009174,
      "grad_norm": 0.14183135330677032,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 1442
    },
    {
      "epoch": 0.23640235910878113,
      "grad_norm": 0.14272549748420715,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 1443
    },
    {
      "epoch": 0.23656618610747052,
      "grad_norm": 0.1381421685218811,
      "learning_rate": 0.0001,
      "loss": 1.5021,
      "step": 1444
    },
    {
      "epoch": 0.2367300131061599,
      "grad_norm": 0.13008996844291687,
      "learning_rate": 0.0001,
      "loss": 1.2937,
      "step": 1445
    },
    {
      "epoch": 0.2368938401048493,
      "grad_norm": 0.14003001153469086,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 1446
    },
    {
      "epoch": 0.23705766710353865,
      "grad_norm": 0.13653084635734558,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 1447
    },
    {
      "epoch": 0.23722149410222804,
      "grad_norm": 0.13896872103214264,
      "learning_rate": 0.0001,
      "loss": 1.3841,
      "step": 1448
    },
    {
      "epoch": 0.23738532110091742,
      "grad_norm": 0.15635424852371216,
      "learning_rate": 0.0001,
      "loss": 1.4365,
      "step": 1449
    },
    {
      "epoch": 0.2375491480996068,
      "grad_norm": 0.13395565748214722,
      "learning_rate": 0.0001,
      "loss": 1.4887,
      "step": 1450
    },
    {
      "epoch": 0.2377129750982962,
      "grad_norm": 0.12725184857845306,
      "learning_rate": 0.0001,
      "loss": 1.3476,
      "step": 1451
    },
    {
      "epoch": 0.23787680209698558,
      "grad_norm": 0.13641944527626038,
      "learning_rate": 0.0001,
      "loss": 1.4304,
      "step": 1452
    },
    {
      "epoch": 0.23804062909567497,
      "grad_norm": 0.13047680258750916,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 1453
    },
    {
      "epoch": 0.23820445609436436,
      "grad_norm": 0.1361381560564041,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 1454
    },
    {
      "epoch": 0.23836828309305375,
      "grad_norm": 0.12961244583129883,
      "learning_rate": 0.0001,
      "loss": 1.4781,
      "step": 1455
    },
    {
      "epoch": 0.23853211009174313,
      "grad_norm": 0.13643904030323029,
      "learning_rate": 0.0001,
      "loss": 1.443,
      "step": 1456
    },
    {
      "epoch": 0.2386959370904325,
      "grad_norm": 0.13938497006893158,
      "learning_rate": 0.0001,
      "loss": 1.4412,
      "step": 1457
    },
    {
      "epoch": 0.23885976408912188,
      "grad_norm": 0.14360737800598145,
      "learning_rate": 0.0001,
      "loss": 1.4832,
      "step": 1458
    },
    {
      "epoch": 0.23902359108781127,
      "grad_norm": 0.1427478939294815,
      "learning_rate": 0.0001,
      "loss": 1.3777,
      "step": 1459
    },
    {
      "epoch": 0.23918741808650065,
      "grad_norm": 0.13589389622211456,
      "learning_rate": 0.0001,
      "loss": 1.379,
      "step": 1460
    },
    {
      "epoch": 0.23935124508519004,
      "grad_norm": 0.1351812183856964,
      "learning_rate": 0.0001,
      "loss": 1.3996,
      "step": 1461
    },
    {
      "epoch": 0.23951507208387943,
      "grad_norm": 0.13299618661403656,
      "learning_rate": 0.0001,
      "loss": 1.4482,
      "step": 1462
    },
    {
      "epoch": 0.2396788990825688,
      "grad_norm": 0.13931900262832642,
      "learning_rate": 0.0001,
      "loss": 1.5004,
      "step": 1463
    },
    {
      "epoch": 0.2398427260812582,
      "grad_norm": 0.1386769860982895,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 1464
    },
    {
      "epoch": 0.2400065530799476,
      "grad_norm": 0.14558853209018707,
      "learning_rate": 0.0001,
      "loss": 1.4098,
      "step": 1465
    },
    {
      "epoch": 0.24017038007863695,
      "grad_norm": 0.14158041775226593,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 1466
    },
    {
      "epoch": 0.24033420707732633,
      "grad_norm": 0.15180878341197968,
      "learning_rate": 0.0001,
      "loss": 1.4631,
      "step": 1467
    },
    {
      "epoch": 0.24049803407601572,
      "grad_norm": 0.15271572768688202,
      "learning_rate": 0.0001,
      "loss": 1.5225,
      "step": 1468
    },
    {
      "epoch": 0.2406618610747051,
      "grad_norm": 0.13806262612342834,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 1469
    },
    {
      "epoch": 0.2408256880733945,
      "grad_norm": 0.134982168674469,
      "learning_rate": 0.0001,
      "loss": 1.4106,
      "step": 1470
    },
    {
      "epoch": 0.24098951507208388,
      "grad_norm": 0.13995356857776642,
      "learning_rate": 0.0001,
      "loss": 1.3826,
      "step": 1471
    },
    {
      "epoch": 0.24115334207077327,
      "grad_norm": 0.13023018836975098,
      "learning_rate": 0.0001,
      "loss": 1.3777,
      "step": 1472
    },
    {
      "epoch": 0.24131716906946266,
      "grad_norm": 0.14182351529598236,
      "learning_rate": 0.0001,
      "loss": 1.4735,
      "step": 1473
    },
    {
      "epoch": 0.24148099606815204,
      "grad_norm": 0.13691093027591705,
      "learning_rate": 0.0001,
      "loss": 1.4285,
      "step": 1474
    },
    {
      "epoch": 0.2416448230668414,
      "grad_norm": 0.13922372460365295,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 1475
    },
    {
      "epoch": 0.2418086500655308,
      "grad_norm": 0.1379704773426056,
      "learning_rate": 0.0001,
      "loss": 1.4393,
      "step": 1476
    },
    {
      "epoch": 0.24197247706422018,
      "grad_norm": 0.17447306215763092,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 1477
    },
    {
      "epoch": 0.24213630406290956,
      "grad_norm": 0.13045382499694824,
      "learning_rate": 0.0001,
      "loss": 1.3718,
      "step": 1478
    },
    {
      "epoch": 0.24230013106159895,
      "grad_norm": 0.1481386423110962,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 1479
    },
    {
      "epoch": 0.24246395806028834,
      "grad_norm": 0.1347479373216629,
      "learning_rate": 0.0001,
      "loss": 1.4236,
      "step": 1480
    },
    {
      "epoch": 0.24262778505897772,
      "grad_norm": 0.13309594988822937,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 1481
    },
    {
      "epoch": 0.2427916120576671,
      "grad_norm": 0.141520693898201,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 1482
    },
    {
      "epoch": 0.2429554390563565,
      "grad_norm": 0.1428850293159485,
      "learning_rate": 0.0001,
      "loss": 1.4145,
      "step": 1483
    },
    {
      "epoch": 0.24311926605504589,
      "grad_norm": 0.143016517162323,
      "learning_rate": 0.0001,
      "loss": 1.5516,
      "step": 1484
    },
    {
      "epoch": 0.24328309305373524,
      "grad_norm": 0.14390859007835388,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 1485
    },
    {
      "epoch": 0.24344692005242463,
      "grad_norm": 0.14119261503219604,
      "learning_rate": 0.0001,
      "loss": 1.372,
      "step": 1486
    },
    {
      "epoch": 0.24361074705111402,
      "grad_norm": 0.13439373672008514,
      "learning_rate": 0.0001,
      "loss": 1.3882,
      "step": 1487
    },
    {
      "epoch": 0.2437745740498034,
      "grad_norm": 0.13351856172084808,
      "learning_rate": 0.0001,
      "loss": 1.4752,
      "step": 1488
    },
    {
      "epoch": 0.2439384010484928,
      "grad_norm": 0.14050473272800446,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 1489
    },
    {
      "epoch": 0.24410222804718218,
      "grad_norm": 0.13997386395931244,
      "learning_rate": 0.0001,
      "loss": 1.4066,
      "step": 1490
    },
    {
      "epoch": 0.24426605504587157,
      "grad_norm": 0.14412151277065277,
      "learning_rate": 0.0001,
      "loss": 1.3553,
      "step": 1491
    },
    {
      "epoch": 0.24442988204456095,
      "grad_norm": 0.14444701373577118,
      "learning_rate": 0.0001,
      "loss": 1.4218,
      "step": 1492
    },
    {
      "epoch": 0.24459370904325034,
      "grad_norm": 0.15384599566459656,
      "learning_rate": 0.0001,
      "loss": 1.4217,
      "step": 1493
    },
    {
      "epoch": 0.2447575360419397,
      "grad_norm": 0.14263832569122314,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 1494
    },
    {
      "epoch": 0.2449213630406291,
      "grad_norm": 0.14328421652317047,
      "learning_rate": 0.0001,
      "loss": 1.4459,
      "step": 1495
    },
    {
      "epoch": 0.24508519003931847,
      "grad_norm": 0.14401395618915558,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 1496
    },
    {
      "epoch": 0.24524901703800786,
      "grad_norm": 0.13899309933185577,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 1497
    },
    {
      "epoch": 0.24541284403669725,
      "grad_norm": 0.14180311560630798,
      "learning_rate": 0.0001,
      "loss": 1.5036,
      "step": 1498
    },
    {
      "epoch": 0.24557667103538663,
      "grad_norm": 0.14008383452892303,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 1499
    },
    {
      "epoch": 0.24574049803407602,
      "grad_norm": 0.1384759396314621,
      "learning_rate": 0.0001,
      "loss": 1.4761,
      "step": 1500
    },
    {
      "epoch": 0.2459043250327654,
      "grad_norm": 0.1370176076889038,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 1501
    },
    {
      "epoch": 0.2460681520314548,
      "grad_norm": 0.1446456015110016,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 1502
    },
    {
      "epoch": 0.24623197903014415,
      "grad_norm": 0.13142569363117218,
      "learning_rate": 0.0001,
      "loss": 1.4189,
      "step": 1503
    },
    {
      "epoch": 0.24639580602883354,
      "grad_norm": 0.13795821368694305,
      "learning_rate": 0.0001,
      "loss": 1.5342,
      "step": 1504
    },
    {
      "epoch": 0.24655963302752293,
      "grad_norm": 0.1430029720067978,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 1505
    },
    {
      "epoch": 0.24672346002621232,
      "grad_norm": 0.14699003100395203,
      "learning_rate": 0.0001,
      "loss": 1.4299,
      "step": 1506
    },
    {
      "epoch": 0.2468872870249017,
      "grad_norm": 0.14670325815677643,
      "learning_rate": 0.0001,
      "loss": 1.4808,
      "step": 1507
    },
    {
      "epoch": 0.2470511140235911,
      "grad_norm": 0.14122462272644043,
      "learning_rate": 0.0001,
      "loss": 1.3567,
      "step": 1508
    },
    {
      "epoch": 0.24721494102228048,
      "grad_norm": 0.13726849853992462,
      "learning_rate": 0.0001,
      "loss": 1.3488,
      "step": 1509
    },
    {
      "epoch": 0.24737876802096986,
      "grad_norm": 0.14140696823596954,
      "learning_rate": 0.0001,
      "loss": 1.3858,
      "step": 1510
    },
    {
      "epoch": 0.24754259501965925,
      "grad_norm": 0.13517622649669647,
      "learning_rate": 0.0001,
      "loss": 1.4251,
      "step": 1511
    },
    {
      "epoch": 0.24770642201834864,
      "grad_norm": 0.1349184811115265,
      "learning_rate": 0.0001,
      "loss": 1.4568,
      "step": 1512
    },
    {
      "epoch": 0.247870249017038,
      "grad_norm": 0.13556483387947083,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 1513
    },
    {
      "epoch": 0.24803407601572738,
      "grad_norm": 0.13869108259677887,
      "learning_rate": 0.0001,
      "loss": 1.5015,
      "step": 1514
    },
    {
      "epoch": 0.24819790301441677,
      "grad_norm": 0.14262007176876068,
      "learning_rate": 0.0001,
      "loss": 1.4288,
      "step": 1515
    },
    {
      "epoch": 0.24836173001310616,
      "grad_norm": 0.1380031704902649,
      "learning_rate": 0.0001,
      "loss": 1.4261,
      "step": 1516
    },
    {
      "epoch": 0.24852555701179554,
      "grad_norm": 0.1386934071779251,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 1517
    },
    {
      "epoch": 0.24868938401048493,
      "grad_norm": 0.14692693948745728,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 1518
    },
    {
      "epoch": 0.24885321100917432,
      "grad_norm": 0.1514987051486969,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 1519
    },
    {
      "epoch": 0.2490170380078637,
      "grad_norm": 0.13201722502708435,
      "learning_rate": 0.0001,
      "loss": 1.3928,
      "step": 1520
    },
    {
      "epoch": 0.2491808650065531,
      "grad_norm": 0.13618551194667816,
      "learning_rate": 0.0001,
      "loss": 1.4101,
      "step": 1521
    },
    {
      "epoch": 0.24934469200524245,
      "grad_norm": 0.1422361135482788,
      "learning_rate": 0.0001,
      "loss": 1.496,
      "step": 1522
    },
    {
      "epoch": 0.24950851900393184,
      "grad_norm": 0.1431334912776947,
      "learning_rate": 0.0001,
      "loss": 1.4864,
      "step": 1523
    },
    {
      "epoch": 0.24967234600262123,
      "grad_norm": 0.12649454176425934,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 1524
    },
    {
      "epoch": 0.2498361730013106,
      "grad_norm": 0.15334337949752808,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 1525
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.14438626170158386,
      "learning_rate": 0.0001,
      "loss": 1.5647,
      "step": 1526
    },
    {
      "epoch": 0.2501638269986894,
      "grad_norm": 0.13284343481063843,
      "learning_rate": 0.0001,
      "loss": 1.3474,
      "step": 1527
    },
    {
      "epoch": 0.2503276539973788,
      "grad_norm": 0.13586591184139252,
      "learning_rate": 0.0001,
      "loss": 1.3755,
      "step": 1528
    },
    {
      "epoch": 0.25049148099606816,
      "grad_norm": 0.1405128836631775,
      "learning_rate": 0.0001,
      "loss": 1.5143,
      "step": 1529
    },
    {
      "epoch": 0.25065530799475755,
      "grad_norm": 0.14128229022026062,
      "learning_rate": 0.0001,
      "loss": 1.3705,
      "step": 1530
    },
    {
      "epoch": 0.25081913499344693,
      "grad_norm": 0.1326082944869995,
      "learning_rate": 0.0001,
      "loss": 1.3684,
      "step": 1531
    },
    {
      "epoch": 0.2509829619921363,
      "grad_norm": 0.13387435674667358,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 1532
    },
    {
      "epoch": 0.2511467889908257,
      "grad_norm": 0.13642969727516174,
      "learning_rate": 0.0001,
      "loss": 1.4601,
      "step": 1533
    },
    {
      "epoch": 0.2513106159895151,
      "grad_norm": 0.14183510839939117,
      "learning_rate": 0.0001,
      "loss": 1.5144,
      "step": 1534
    },
    {
      "epoch": 0.2514744429882045,
      "grad_norm": 0.1381664276123047,
      "learning_rate": 0.0001,
      "loss": 1.2728,
      "step": 1535
    },
    {
      "epoch": 0.2516382699868938,
      "grad_norm": 0.13913817703723907,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 1536
    },
    {
      "epoch": 0.2518020969855832,
      "grad_norm": 0.1312909722328186,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 1537
    },
    {
      "epoch": 0.2519659239842726,
      "grad_norm": 0.13945162296295166,
      "learning_rate": 0.0001,
      "loss": 1.4574,
      "step": 1538
    },
    {
      "epoch": 0.252129750982962,
      "grad_norm": 0.14941301941871643,
      "learning_rate": 0.0001,
      "loss": 1.5297,
      "step": 1539
    },
    {
      "epoch": 0.25229357798165136,
      "grad_norm": 0.1355467140674591,
      "learning_rate": 0.0001,
      "loss": 1.4273,
      "step": 1540
    },
    {
      "epoch": 0.25245740498034075,
      "grad_norm": 0.13217385113239288,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 1541
    },
    {
      "epoch": 0.25262123197903014,
      "grad_norm": 0.14712102711200714,
      "learning_rate": 0.0001,
      "loss": 1.6568,
      "step": 1542
    },
    {
      "epoch": 0.2527850589777195,
      "grad_norm": 0.14134183526039124,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 1543
    },
    {
      "epoch": 0.2529488859764089,
      "grad_norm": 0.14736267924308777,
      "learning_rate": 0.0001,
      "loss": 1.4097,
      "step": 1544
    },
    {
      "epoch": 0.2531127129750983,
      "grad_norm": 0.13848388195037842,
      "learning_rate": 0.0001,
      "loss": 1.4857,
      "step": 1545
    },
    {
      "epoch": 0.2532765399737877,
      "grad_norm": 0.13585376739501953,
      "learning_rate": 0.0001,
      "loss": 1.3547,
      "step": 1546
    },
    {
      "epoch": 0.25344036697247707,
      "grad_norm": 0.14260517060756683,
      "learning_rate": 0.0001,
      "loss": 1.4387,
      "step": 1547
    },
    {
      "epoch": 0.25360419397116646,
      "grad_norm": 0.14070190489292145,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 1548
    },
    {
      "epoch": 0.25376802096985585,
      "grad_norm": 0.14917384088039398,
      "learning_rate": 0.0001,
      "loss": 1.5158,
      "step": 1549
    },
    {
      "epoch": 0.25393184796854523,
      "grad_norm": 0.13900260627269745,
      "learning_rate": 0.0001,
      "loss": 1.4361,
      "step": 1550
    },
    {
      "epoch": 0.2540956749672346,
      "grad_norm": 0.1413586139678955,
      "learning_rate": 0.0001,
      "loss": 1.4906,
      "step": 1551
    },
    {
      "epoch": 0.254259501965924,
      "grad_norm": 0.13651412725448608,
      "learning_rate": 0.0001,
      "loss": 1.468,
      "step": 1552
    },
    {
      "epoch": 0.2544233289646134,
      "grad_norm": 0.1342843621969223,
      "learning_rate": 0.0001,
      "loss": 1.4607,
      "step": 1553
    },
    {
      "epoch": 0.2545871559633027,
      "grad_norm": 0.1421276330947876,
      "learning_rate": 0.0001,
      "loss": 1.4657,
      "step": 1554
    },
    {
      "epoch": 0.2547509829619921,
      "grad_norm": 0.13576820492744446,
      "learning_rate": 0.0001,
      "loss": 1.4391,
      "step": 1555
    },
    {
      "epoch": 0.2549148099606815,
      "grad_norm": 0.14635837078094482,
      "learning_rate": 0.0001,
      "loss": 1.4208,
      "step": 1556
    },
    {
      "epoch": 0.2550786369593709,
      "grad_norm": 0.13740237057209015,
      "learning_rate": 0.0001,
      "loss": 1.3042,
      "step": 1557
    },
    {
      "epoch": 0.2552424639580603,
      "grad_norm": 0.13988633453845978,
      "learning_rate": 0.0001,
      "loss": 1.3972,
      "step": 1558
    },
    {
      "epoch": 0.25540629095674966,
      "grad_norm": 0.1397271454334259,
      "learning_rate": 0.0001,
      "loss": 1.417,
      "step": 1559
    },
    {
      "epoch": 0.25557011795543905,
      "grad_norm": 0.14033791422843933,
      "learning_rate": 0.0001,
      "loss": 1.4499,
      "step": 1560
    },
    {
      "epoch": 0.25573394495412843,
      "grad_norm": 0.13488472998142242,
      "learning_rate": 0.0001,
      "loss": 1.4477,
      "step": 1561
    },
    {
      "epoch": 0.2558977719528178,
      "grad_norm": 0.1352391391992569,
      "learning_rate": 0.0001,
      "loss": 1.4702,
      "step": 1562
    },
    {
      "epoch": 0.2560615989515072,
      "grad_norm": 0.13397885859012604,
      "learning_rate": 0.0001,
      "loss": 1.3917,
      "step": 1563
    },
    {
      "epoch": 0.2562254259501966,
      "grad_norm": 0.13955478370189667,
      "learning_rate": 0.0001,
      "loss": 1.4805,
      "step": 1564
    },
    {
      "epoch": 0.256389252948886,
      "grad_norm": 0.1380913108587265,
      "learning_rate": 0.0001,
      "loss": 1.3344,
      "step": 1565
    },
    {
      "epoch": 0.25655307994757537,
      "grad_norm": 0.1406307965517044,
      "learning_rate": 0.0001,
      "loss": 1.4118,
      "step": 1566
    },
    {
      "epoch": 0.25671690694626476,
      "grad_norm": 0.13944190740585327,
      "learning_rate": 0.0001,
      "loss": 1.4414,
      "step": 1567
    },
    {
      "epoch": 0.25688073394495414,
      "grad_norm": 0.13672661781311035,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 1568
    },
    {
      "epoch": 0.25704456094364353,
      "grad_norm": 0.1393662542104721,
      "learning_rate": 0.0001,
      "loss": 1.3592,
      "step": 1569
    },
    {
      "epoch": 0.2572083879423329,
      "grad_norm": 0.13855676352977753,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 1570
    },
    {
      "epoch": 0.2573722149410223,
      "grad_norm": 0.13432767987251282,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 1571
    },
    {
      "epoch": 0.2575360419397117,
      "grad_norm": 0.1463669389486313,
      "learning_rate": 0.0001,
      "loss": 1.4332,
      "step": 1572
    },
    {
      "epoch": 0.257699868938401,
      "grad_norm": 0.14090314507484436,
      "learning_rate": 0.0001,
      "loss": 1.4968,
      "step": 1573
    },
    {
      "epoch": 0.2578636959370904,
      "grad_norm": 0.13676491379737854,
      "learning_rate": 0.0001,
      "loss": 1.4425,
      "step": 1574
    },
    {
      "epoch": 0.2580275229357798,
      "grad_norm": 0.14262230694293976,
      "learning_rate": 0.0001,
      "loss": 1.4658,
      "step": 1575
    },
    {
      "epoch": 0.2581913499344692,
      "grad_norm": 0.14057207107543945,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 1576
    },
    {
      "epoch": 0.25835517693315857,
      "grad_norm": 0.13464364409446716,
      "learning_rate": 0.0001,
      "loss": 1.3459,
      "step": 1577
    },
    {
      "epoch": 0.25851900393184796,
      "grad_norm": 0.14135457575321198,
      "learning_rate": 0.0001,
      "loss": 1.3802,
      "step": 1578
    },
    {
      "epoch": 0.25868283093053734,
      "grad_norm": 0.13926689326763153,
      "learning_rate": 0.0001,
      "loss": 1.4077,
      "step": 1579
    },
    {
      "epoch": 0.25884665792922673,
      "grad_norm": 0.1284532994031906,
      "learning_rate": 0.0001,
      "loss": 1.3968,
      "step": 1580
    },
    {
      "epoch": 0.2590104849279161,
      "grad_norm": 0.14124466478824615,
      "learning_rate": 0.0001,
      "loss": 1.4041,
      "step": 1581
    },
    {
      "epoch": 0.2591743119266055,
      "grad_norm": 0.1419697105884552,
      "learning_rate": 0.0001,
      "loss": 1.5238,
      "step": 1582
    },
    {
      "epoch": 0.2593381389252949,
      "grad_norm": 0.1319558024406433,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 1583
    },
    {
      "epoch": 0.2595019659239843,
      "grad_norm": 0.14164187014102936,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 1584
    },
    {
      "epoch": 0.25966579292267367,
      "grad_norm": 0.13898637890815735,
      "learning_rate": 0.0001,
      "loss": 1.3836,
      "step": 1585
    },
    {
      "epoch": 0.25982961992136305,
      "grad_norm": 0.1410815715789795,
      "learning_rate": 0.0001,
      "loss": 1.3435,
      "step": 1586
    },
    {
      "epoch": 0.25999344692005244,
      "grad_norm": 0.13889847695827484,
      "learning_rate": 0.0001,
      "loss": 1.4519,
      "step": 1587
    },
    {
      "epoch": 0.2601572739187418,
      "grad_norm": 0.13758698105812073,
      "learning_rate": 0.0001,
      "loss": 1.4185,
      "step": 1588
    },
    {
      "epoch": 0.2603211009174312,
      "grad_norm": 0.13902972638607025,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 1589
    },
    {
      "epoch": 0.2604849279161206,
      "grad_norm": 0.13187755644321442,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 1590
    },
    {
      "epoch": 0.26064875491481,
      "grad_norm": 0.14192380011081696,
      "learning_rate": 0.0001,
      "loss": 1.4713,
      "step": 1591
    },
    {
      "epoch": 0.2608125819134993,
      "grad_norm": 0.13667398691177368,
      "learning_rate": 0.0001,
      "loss": 1.292,
      "step": 1592
    },
    {
      "epoch": 0.2609764089121887,
      "grad_norm": 0.13098962604999542,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 1593
    },
    {
      "epoch": 0.2611402359108781,
      "grad_norm": 0.1406790316104889,
      "learning_rate": 0.0001,
      "loss": 1.4633,
      "step": 1594
    },
    {
      "epoch": 0.2613040629095675,
      "grad_norm": 0.14410783350467682,
      "learning_rate": 0.0001,
      "loss": 1.366,
      "step": 1595
    },
    {
      "epoch": 0.26146788990825687,
      "grad_norm": 0.14813190698623657,
      "learning_rate": 0.0001,
      "loss": 1.4403,
      "step": 1596
    },
    {
      "epoch": 0.26163171690694625,
      "grad_norm": 0.13872158527374268,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 1597
    },
    {
      "epoch": 0.26179554390563564,
      "grad_norm": 0.14901527762413025,
      "learning_rate": 0.0001,
      "loss": 1.3757,
      "step": 1598
    },
    {
      "epoch": 0.26195937090432503,
      "grad_norm": 0.13365288078784943,
      "learning_rate": 0.0001,
      "loss": 1.3781,
      "step": 1599
    },
    {
      "epoch": 0.2621231979030144,
      "grad_norm": 0.1368209570646286,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 1600
    },
    {
      "epoch": 0.2622870249017038,
      "grad_norm": 0.14010444283485413,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 1601
    },
    {
      "epoch": 0.2624508519003932,
      "grad_norm": 0.14879627525806427,
      "learning_rate": 0.0001,
      "loss": 1.4923,
      "step": 1602
    },
    {
      "epoch": 0.2626146788990826,
      "grad_norm": 0.1497354358434677,
      "learning_rate": 0.0001,
      "loss": 1.4933,
      "step": 1603
    },
    {
      "epoch": 0.26277850589777196,
      "grad_norm": 0.14126631617546082,
      "learning_rate": 0.0001,
      "loss": 1.4028,
      "step": 1604
    },
    {
      "epoch": 0.26294233289646135,
      "grad_norm": 0.1474335491657257,
      "learning_rate": 0.0001,
      "loss": 1.5105,
      "step": 1605
    },
    {
      "epoch": 0.26310615989515074,
      "grad_norm": 0.14736789464950562,
      "learning_rate": 0.0001,
      "loss": 1.4395,
      "step": 1606
    },
    {
      "epoch": 0.2632699868938401,
      "grad_norm": 0.14721505343914032,
      "learning_rate": 0.0001,
      "loss": 1.4742,
      "step": 1607
    },
    {
      "epoch": 0.2634338138925295,
      "grad_norm": 0.13304905593395233,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 1608
    },
    {
      "epoch": 0.2635976408912189,
      "grad_norm": 0.14149995148181915,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 1609
    },
    {
      "epoch": 0.26376146788990823,
      "grad_norm": 0.14642231166362762,
      "learning_rate": 0.0001,
      "loss": 1.3705,
      "step": 1610
    },
    {
      "epoch": 0.2639252948885976,
      "grad_norm": 0.13726823031902313,
      "learning_rate": 0.0001,
      "loss": 1.4663,
      "step": 1611
    },
    {
      "epoch": 0.264089121887287,
      "grad_norm": 0.1410422921180725,
      "learning_rate": 0.0001,
      "loss": 1.3514,
      "step": 1612
    },
    {
      "epoch": 0.2642529488859764,
      "grad_norm": 0.14905300736427307,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 1613
    },
    {
      "epoch": 0.2644167758846658,
      "grad_norm": 0.14630122482776642,
      "learning_rate": 0.0001,
      "loss": 1.4475,
      "step": 1614
    },
    {
      "epoch": 0.26458060288335516,
      "grad_norm": 0.1399676501750946,
      "learning_rate": 0.0001,
      "loss": 1.4346,
      "step": 1615
    },
    {
      "epoch": 0.26474442988204455,
      "grad_norm": 0.14270730316638947,
      "learning_rate": 0.0001,
      "loss": 1.4562,
      "step": 1616
    },
    {
      "epoch": 0.26490825688073394,
      "grad_norm": 0.13685283064842224,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 1617
    },
    {
      "epoch": 0.2650720838794233,
      "grad_norm": 0.14679796993732452,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 1618
    },
    {
      "epoch": 0.2652359108781127,
      "grad_norm": 0.13302841782569885,
      "learning_rate": 0.0001,
      "loss": 1.4365,
      "step": 1619
    },
    {
      "epoch": 0.2653997378768021,
      "grad_norm": 0.14465032517910004,
      "learning_rate": 0.0001,
      "loss": 1.4073,
      "step": 1620
    },
    {
      "epoch": 0.2655635648754915,
      "grad_norm": 0.13588382303714752,
      "learning_rate": 0.0001,
      "loss": 1.3714,
      "step": 1621
    },
    {
      "epoch": 0.2657273918741809,
      "grad_norm": 0.13521866500377655,
      "learning_rate": 0.0001,
      "loss": 1.3248,
      "step": 1622
    },
    {
      "epoch": 0.26589121887287026,
      "grad_norm": 0.13451997935771942,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 1623
    },
    {
      "epoch": 0.26605504587155965,
      "grad_norm": 0.1409652978181839,
      "learning_rate": 0.0001,
      "loss": 1.4999,
      "step": 1624
    },
    {
      "epoch": 0.26621887287024903,
      "grad_norm": 0.14567911624908447,
      "learning_rate": 0.0001,
      "loss": 1.4967,
      "step": 1625
    },
    {
      "epoch": 0.2663826998689384,
      "grad_norm": 0.14251235127449036,
      "learning_rate": 0.0001,
      "loss": 1.4475,
      "step": 1626
    },
    {
      "epoch": 0.2665465268676278,
      "grad_norm": 0.15833978354930878,
      "learning_rate": 0.0001,
      "loss": 1.2865,
      "step": 1627
    },
    {
      "epoch": 0.2667103538663172,
      "grad_norm": 0.1417645812034607,
      "learning_rate": 0.0001,
      "loss": 1.4349,
      "step": 1628
    },
    {
      "epoch": 0.2668741808650065,
      "grad_norm": 0.1332193911075592,
      "learning_rate": 0.0001,
      "loss": 1.3956,
      "step": 1629
    },
    {
      "epoch": 0.2670380078636959,
      "grad_norm": 0.13508488237857819,
      "learning_rate": 0.0001,
      "loss": 1.4477,
      "step": 1630
    },
    {
      "epoch": 0.2672018348623853,
      "grad_norm": 0.132359579205513,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 1631
    },
    {
      "epoch": 0.2673656618610747,
      "grad_norm": 0.14752821624279022,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 1632
    },
    {
      "epoch": 0.2675294888597641,
      "grad_norm": 0.1507118195295334,
      "learning_rate": 0.0001,
      "loss": 1.4054,
      "step": 1633
    },
    {
      "epoch": 0.26769331585845346,
      "grad_norm": 0.13890908658504486,
      "learning_rate": 0.0001,
      "loss": 1.4823,
      "step": 1634
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 0.1319638341665268,
      "learning_rate": 0.0001,
      "loss": 1.3462,
      "step": 1635
    },
    {
      "epoch": 0.26802096985583224,
      "grad_norm": 0.13547824323177338,
      "learning_rate": 0.0001,
      "loss": 1.3576,
      "step": 1636
    },
    {
      "epoch": 0.2681847968545216,
      "grad_norm": 0.13890932500362396,
      "learning_rate": 0.0001,
      "loss": 1.4746,
      "step": 1637
    },
    {
      "epoch": 0.268348623853211,
      "grad_norm": 0.14714154601097107,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 1638
    },
    {
      "epoch": 0.2685124508519004,
      "grad_norm": 0.14010606706142426,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 1639
    },
    {
      "epoch": 0.2686762778505898,
      "grad_norm": 0.1396878957748413,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 1640
    },
    {
      "epoch": 0.26884010484927917,
      "grad_norm": 0.14947780966758728,
      "learning_rate": 0.0001,
      "loss": 1.5036,
      "step": 1641
    },
    {
      "epoch": 0.26900393184796856,
      "grad_norm": 0.1384734958410263,
      "learning_rate": 0.0001,
      "loss": 1.3906,
      "step": 1642
    },
    {
      "epoch": 0.26916775884665795,
      "grad_norm": 0.1421780288219452,
      "learning_rate": 0.0001,
      "loss": 1.4512,
      "step": 1643
    },
    {
      "epoch": 0.26933158584534733,
      "grad_norm": 0.1543324589729309,
      "learning_rate": 0.0001,
      "loss": 1.4334,
      "step": 1644
    },
    {
      "epoch": 0.2694954128440367,
      "grad_norm": 0.13510698080062866,
      "learning_rate": 0.0001,
      "loss": 1.364,
      "step": 1645
    },
    {
      "epoch": 0.2696592398427261,
      "grad_norm": 0.14592459797859192,
      "learning_rate": 0.0001,
      "loss": 1.4827,
      "step": 1646
    },
    {
      "epoch": 0.26982306684141544,
      "grad_norm": 0.1456330418586731,
      "learning_rate": 0.0001,
      "loss": 1.3981,
      "step": 1647
    },
    {
      "epoch": 0.2699868938401048,
      "grad_norm": 0.1435934156179428,
      "learning_rate": 0.0001,
      "loss": 1.5534,
      "step": 1648
    },
    {
      "epoch": 0.2701507208387942,
      "grad_norm": 0.14720387756824493,
      "learning_rate": 0.0001,
      "loss": 1.58,
      "step": 1649
    },
    {
      "epoch": 0.2703145478374836,
      "grad_norm": 0.14271391928195953,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 1650
    },
    {
      "epoch": 0.270478374836173,
      "grad_norm": 0.15668486058712006,
      "learning_rate": 0.0001,
      "loss": 1.4844,
      "step": 1651
    },
    {
      "epoch": 0.2706422018348624,
      "grad_norm": 0.1387423425912857,
      "learning_rate": 0.0001,
      "loss": 1.5678,
      "step": 1652
    },
    {
      "epoch": 0.27080602883355176,
      "grad_norm": 0.16873414814472198,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 1653
    },
    {
      "epoch": 0.27096985583224115,
      "grad_norm": 0.14510564506053925,
      "learning_rate": 0.0001,
      "loss": 1.5344,
      "step": 1654
    },
    {
      "epoch": 0.27113368283093053,
      "grad_norm": 0.13739588856697083,
      "learning_rate": 0.0001,
      "loss": 1.3477,
      "step": 1655
    },
    {
      "epoch": 0.2712975098296199,
      "grad_norm": 0.14241552352905273,
      "learning_rate": 0.0001,
      "loss": 1.5261,
      "step": 1656
    },
    {
      "epoch": 0.2714613368283093,
      "grad_norm": 0.14121893048286438,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 1657
    },
    {
      "epoch": 0.2716251638269987,
      "grad_norm": 0.14170095324516296,
      "learning_rate": 0.0001,
      "loss": 1.4507,
      "step": 1658
    },
    {
      "epoch": 0.2717889908256881,
      "grad_norm": 0.13418777287006378,
      "learning_rate": 0.0001,
      "loss": 1.3651,
      "step": 1659
    },
    {
      "epoch": 0.27195281782437747,
      "grad_norm": 0.14088474214076996,
      "learning_rate": 0.0001,
      "loss": 1.4444,
      "step": 1660
    },
    {
      "epoch": 0.27211664482306686,
      "grad_norm": 0.1361752301454544,
      "learning_rate": 0.0001,
      "loss": 1.4624,
      "step": 1661
    },
    {
      "epoch": 0.27228047182175624,
      "grad_norm": 0.13655591011047363,
      "learning_rate": 0.0001,
      "loss": 1.3612,
      "step": 1662
    },
    {
      "epoch": 0.27244429882044563,
      "grad_norm": 0.1363121122121811,
      "learning_rate": 0.0001,
      "loss": 1.3417,
      "step": 1663
    },
    {
      "epoch": 0.272608125819135,
      "grad_norm": 0.13503505289554596,
      "learning_rate": 0.0001,
      "loss": 1.3534,
      "step": 1664
    },
    {
      "epoch": 0.2727719528178244,
      "grad_norm": 0.13399475812911987,
      "learning_rate": 0.0001,
      "loss": 1.418,
      "step": 1665
    },
    {
      "epoch": 0.27293577981651373,
      "grad_norm": 0.14262616634368896,
      "learning_rate": 0.0001,
      "loss": 1.4331,
      "step": 1666
    },
    {
      "epoch": 0.2730996068152031,
      "grad_norm": 0.14565593004226685,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 1667
    },
    {
      "epoch": 0.2732634338138925,
      "grad_norm": 0.14346784353256226,
      "learning_rate": 0.0001,
      "loss": 1.5264,
      "step": 1668
    },
    {
      "epoch": 0.2734272608125819,
      "grad_norm": 0.13593865931034088,
      "learning_rate": 0.0001,
      "loss": 1.4575,
      "step": 1669
    },
    {
      "epoch": 0.2735910878112713,
      "grad_norm": 0.1438358724117279,
      "learning_rate": 0.0001,
      "loss": 1.4588,
      "step": 1670
    },
    {
      "epoch": 0.27375491480996067,
      "grad_norm": 0.13479255139827728,
      "learning_rate": 0.0001,
      "loss": 1.3899,
      "step": 1671
    },
    {
      "epoch": 0.27391874180865006,
      "grad_norm": 0.14612439274787903,
      "learning_rate": 0.0001,
      "loss": 1.5032,
      "step": 1672
    },
    {
      "epoch": 0.27408256880733944,
      "grad_norm": 0.13931678235530853,
      "learning_rate": 0.0001,
      "loss": 1.381,
      "step": 1673
    },
    {
      "epoch": 0.27424639580602883,
      "grad_norm": 0.1649237722158432,
      "learning_rate": 0.0001,
      "loss": 1.5694,
      "step": 1674
    },
    {
      "epoch": 0.2744102228047182,
      "grad_norm": 0.13513094186782837,
      "learning_rate": 0.0001,
      "loss": 1.4135,
      "step": 1675
    },
    {
      "epoch": 0.2745740498034076,
      "grad_norm": 0.15425032377243042,
      "learning_rate": 0.0001,
      "loss": 1.5305,
      "step": 1676
    },
    {
      "epoch": 0.274737876802097,
      "grad_norm": 0.16422727704048157,
      "learning_rate": 0.0001,
      "loss": 1.3914,
      "step": 1677
    },
    {
      "epoch": 0.2749017038007864,
      "grad_norm": 0.1385728120803833,
      "learning_rate": 0.0001,
      "loss": 1.3814,
      "step": 1678
    },
    {
      "epoch": 0.27506553079947577,
      "grad_norm": 0.14296138286590576,
      "learning_rate": 0.0001,
      "loss": 1.4838,
      "step": 1679
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 0.15012051165103912,
      "learning_rate": 0.0001,
      "loss": 1.5332,
      "step": 1680
    },
    {
      "epoch": 0.27539318479685454,
      "grad_norm": 0.14276817440986633,
      "learning_rate": 0.0001,
      "loss": 1.3789,
      "step": 1681
    },
    {
      "epoch": 0.2755570117955439,
      "grad_norm": 0.15215058624744415,
      "learning_rate": 0.0001,
      "loss": 1.457,
      "step": 1682
    },
    {
      "epoch": 0.2757208387942333,
      "grad_norm": 0.14970549941062927,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 1683
    },
    {
      "epoch": 0.2758846657929227,
      "grad_norm": 0.15210726857185364,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 1684
    },
    {
      "epoch": 0.27604849279161203,
      "grad_norm": 0.15899045765399933,
      "learning_rate": 0.0001,
      "loss": 1.4981,
      "step": 1685
    },
    {
      "epoch": 0.2762123197903014,
      "grad_norm": 0.1448175460100174,
      "learning_rate": 0.0001,
      "loss": 1.3954,
      "step": 1686
    },
    {
      "epoch": 0.2763761467889908,
      "grad_norm": 0.1376006156206131,
      "learning_rate": 0.0001,
      "loss": 1.3375,
      "step": 1687
    },
    {
      "epoch": 0.2765399737876802,
      "grad_norm": 0.15372176468372345,
      "learning_rate": 0.0001,
      "loss": 1.5501,
      "step": 1688
    },
    {
      "epoch": 0.2767038007863696,
      "grad_norm": 0.14108850061893463,
      "learning_rate": 0.0001,
      "loss": 1.3177,
      "step": 1689
    },
    {
      "epoch": 0.27686762778505897,
      "grad_norm": 0.14824871718883514,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 1690
    },
    {
      "epoch": 0.27703145478374835,
      "grad_norm": 0.15435978770256042,
      "learning_rate": 0.0001,
      "loss": 1.4356,
      "step": 1691
    },
    {
      "epoch": 0.27719528178243774,
      "grad_norm": 0.13472141325473785,
      "learning_rate": 0.0001,
      "loss": 1.3308,
      "step": 1692
    },
    {
      "epoch": 0.27735910878112713,
      "grad_norm": 0.1392168253660202,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 1693
    },
    {
      "epoch": 0.2775229357798165,
      "grad_norm": 0.14961545169353485,
      "learning_rate": 0.0001,
      "loss": 1.3839,
      "step": 1694
    },
    {
      "epoch": 0.2776867627785059,
      "grad_norm": 0.13778327405452728,
      "learning_rate": 0.0001,
      "loss": 1.389,
      "step": 1695
    },
    {
      "epoch": 0.2778505897771953,
      "grad_norm": 0.14622870087623596,
      "learning_rate": 0.0001,
      "loss": 1.393,
      "step": 1696
    },
    {
      "epoch": 0.2780144167758847,
      "grad_norm": 0.1490260362625122,
      "learning_rate": 0.0001,
      "loss": 1.4196,
      "step": 1697
    },
    {
      "epoch": 0.27817824377457406,
      "grad_norm": 0.1408802568912506,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 1698
    },
    {
      "epoch": 0.27834207077326345,
      "grad_norm": 0.1436590999364853,
      "learning_rate": 0.0001,
      "loss": 1.4465,
      "step": 1699
    },
    {
      "epoch": 0.27850589777195284,
      "grad_norm": 0.15496467053890228,
      "learning_rate": 0.0001,
      "loss": 1.4978,
      "step": 1700
    },
    {
      "epoch": 0.2786697247706422,
      "grad_norm": 0.1438951939344406,
      "learning_rate": 0.0001,
      "loss": 1.4814,
      "step": 1701
    },
    {
      "epoch": 0.2788335517693316,
      "grad_norm": 0.14717042446136475,
      "learning_rate": 0.0001,
      "loss": 1.4739,
      "step": 1702
    },
    {
      "epoch": 0.27899737876802094,
      "grad_norm": 0.14863371849060059,
      "learning_rate": 0.0001,
      "loss": 1.4294,
      "step": 1703
    },
    {
      "epoch": 0.27916120576671033,
      "grad_norm": 0.15955810248851776,
      "learning_rate": 0.0001,
      "loss": 1.4043,
      "step": 1704
    },
    {
      "epoch": 0.2793250327653997,
      "grad_norm": 0.14562343060970306,
      "learning_rate": 0.0001,
      "loss": 1.429,
      "step": 1705
    },
    {
      "epoch": 0.2794888597640891,
      "grad_norm": 0.1441631317138672,
      "learning_rate": 0.0001,
      "loss": 1.3415,
      "step": 1706
    },
    {
      "epoch": 0.2796526867627785,
      "grad_norm": 0.14941725134849548,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 1707
    },
    {
      "epoch": 0.2798165137614679,
      "grad_norm": 0.14250634610652924,
      "learning_rate": 0.0001,
      "loss": 1.4019,
      "step": 1708
    },
    {
      "epoch": 0.27998034076015726,
      "grad_norm": 0.14455091953277588,
      "learning_rate": 0.0001,
      "loss": 1.4061,
      "step": 1709
    },
    {
      "epoch": 0.28014416775884665,
      "grad_norm": 0.17768818140029907,
      "learning_rate": 0.0001,
      "loss": 1.6065,
      "step": 1710
    },
    {
      "epoch": 0.28030799475753604,
      "grad_norm": 0.14666207134723663,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 1711
    },
    {
      "epoch": 0.2804718217562254,
      "grad_norm": 0.14580278098583221,
      "learning_rate": 0.0001,
      "loss": 1.3022,
      "step": 1712
    },
    {
      "epoch": 0.2806356487549148,
      "grad_norm": 0.16146087646484375,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 1713
    },
    {
      "epoch": 0.2807994757536042,
      "grad_norm": 0.14011487364768982,
      "learning_rate": 0.0001,
      "loss": 1.5138,
      "step": 1714
    },
    {
      "epoch": 0.2809633027522936,
      "grad_norm": 0.15292246639728546,
      "learning_rate": 0.0001,
      "loss": 1.545,
      "step": 1715
    },
    {
      "epoch": 0.281127129750983,
      "grad_norm": 0.14729154109954834,
      "learning_rate": 0.0001,
      "loss": 1.3862,
      "step": 1716
    },
    {
      "epoch": 0.28129095674967236,
      "grad_norm": 0.13655667006969452,
      "learning_rate": 0.0001,
      "loss": 1.4082,
      "step": 1717
    },
    {
      "epoch": 0.28145478374836175,
      "grad_norm": 0.14140574634075165,
      "learning_rate": 0.0001,
      "loss": 1.3989,
      "step": 1718
    },
    {
      "epoch": 0.28161861074705113,
      "grad_norm": 0.14591026306152344,
      "learning_rate": 0.0001,
      "loss": 1.447,
      "step": 1719
    },
    {
      "epoch": 0.2817824377457405,
      "grad_norm": 0.14258135855197906,
      "learning_rate": 0.0001,
      "loss": 1.4105,
      "step": 1720
    },
    {
      "epoch": 0.2819462647444299,
      "grad_norm": 0.14185772836208344,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 1721
    },
    {
      "epoch": 0.28211009174311924,
      "grad_norm": 0.14729367196559906,
      "learning_rate": 0.0001,
      "loss": 1.5384,
      "step": 1722
    },
    {
      "epoch": 0.2822739187418086,
      "grad_norm": 0.13823837041854858,
      "learning_rate": 0.0001,
      "loss": 1.4742,
      "step": 1723
    },
    {
      "epoch": 0.282437745740498,
      "grad_norm": 0.14388912916183472,
      "learning_rate": 0.0001,
      "loss": 1.3444,
      "step": 1724
    },
    {
      "epoch": 0.2826015727391874,
      "grad_norm": 0.13775673508644104,
      "learning_rate": 0.0001,
      "loss": 1.3882,
      "step": 1725
    },
    {
      "epoch": 0.2827653997378768,
      "grad_norm": 0.13593755662441254,
      "learning_rate": 0.0001,
      "loss": 1.4102,
      "step": 1726
    },
    {
      "epoch": 0.2829292267365662,
      "grad_norm": 0.13951152563095093,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 1727
    },
    {
      "epoch": 0.28309305373525556,
      "grad_norm": 0.13914059102535248,
      "learning_rate": 0.0001,
      "loss": 1.3736,
      "step": 1728
    },
    {
      "epoch": 0.28325688073394495,
      "grad_norm": 0.14618508517742157,
      "learning_rate": 0.0001,
      "loss": 1.4489,
      "step": 1729
    },
    {
      "epoch": 0.28342070773263434,
      "grad_norm": 0.13436877727508545,
      "learning_rate": 0.0001,
      "loss": 1.3481,
      "step": 1730
    },
    {
      "epoch": 0.2835845347313237,
      "grad_norm": 0.14852552115917206,
      "learning_rate": 0.0001,
      "loss": 1.4219,
      "step": 1731
    },
    {
      "epoch": 0.2837483617300131,
      "grad_norm": 0.14195145666599274,
      "learning_rate": 0.0001,
      "loss": 1.3445,
      "step": 1732
    },
    {
      "epoch": 0.2839121887287025,
      "grad_norm": 0.147239550948143,
      "learning_rate": 0.0001,
      "loss": 1.5578,
      "step": 1733
    },
    {
      "epoch": 0.2840760157273919,
      "grad_norm": 0.14333771169185638,
      "learning_rate": 0.0001,
      "loss": 1.327,
      "step": 1734
    },
    {
      "epoch": 0.28423984272608127,
      "grad_norm": 0.14393383264541626,
      "learning_rate": 0.0001,
      "loss": 1.3502,
      "step": 1735
    },
    {
      "epoch": 0.28440366972477066,
      "grad_norm": 0.14481110870838165,
      "learning_rate": 0.0001,
      "loss": 1.4312,
      "step": 1736
    },
    {
      "epoch": 0.28456749672346004,
      "grad_norm": 0.13373464345932007,
      "learning_rate": 0.0001,
      "loss": 1.3818,
      "step": 1737
    },
    {
      "epoch": 0.28473132372214943,
      "grad_norm": 0.13585199415683746,
      "learning_rate": 0.0001,
      "loss": 1.3944,
      "step": 1738
    },
    {
      "epoch": 0.2848951507208388,
      "grad_norm": 0.13818001747131348,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 1739
    },
    {
      "epoch": 0.2850589777195282,
      "grad_norm": 0.1355072408914566,
      "learning_rate": 0.0001,
      "loss": 1.3791,
      "step": 1740
    },
    {
      "epoch": 0.28522280471821754,
      "grad_norm": 0.14079520106315613,
      "learning_rate": 0.0001,
      "loss": 1.4886,
      "step": 1741
    },
    {
      "epoch": 0.2853866317169069,
      "grad_norm": 0.14542478322982788,
      "learning_rate": 0.0001,
      "loss": 1.4814,
      "step": 1742
    },
    {
      "epoch": 0.2855504587155963,
      "grad_norm": 0.1447782963514328,
      "learning_rate": 0.0001,
      "loss": 1.4303,
      "step": 1743
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.14225538074970245,
      "learning_rate": 0.0001,
      "loss": 1.5031,
      "step": 1744
    },
    {
      "epoch": 0.2858781127129751,
      "grad_norm": 0.13870565593242645,
      "learning_rate": 0.0001,
      "loss": 1.3878,
      "step": 1745
    },
    {
      "epoch": 0.2860419397116645,
      "grad_norm": 0.14566750824451447,
      "learning_rate": 0.0001,
      "loss": 1.4474,
      "step": 1746
    },
    {
      "epoch": 0.28620576671035386,
      "grad_norm": 0.1367940753698349,
      "learning_rate": 0.0001,
      "loss": 1.3689,
      "step": 1747
    },
    {
      "epoch": 0.28636959370904325,
      "grad_norm": 0.14396508038043976,
      "learning_rate": 0.0001,
      "loss": 1.4223,
      "step": 1748
    },
    {
      "epoch": 0.28653342070773263,
      "grad_norm": 0.13565029203891754,
      "learning_rate": 0.0001,
      "loss": 1.3873,
      "step": 1749
    },
    {
      "epoch": 0.286697247706422,
      "grad_norm": 0.1355242282152176,
      "learning_rate": 0.0001,
      "loss": 1.4343,
      "step": 1750
    },
    {
      "epoch": 0.2868610747051114,
      "grad_norm": 0.1470026820898056,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 1751
    },
    {
      "epoch": 0.2870249017038008,
      "grad_norm": 0.14111322164535522,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 1752
    },
    {
      "epoch": 0.2871887287024902,
      "grad_norm": 0.14506760239601135,
      "learning_rate": 0.0001,
      "loss": 1.4462,
      "step": 1753
    },
    {
      "epoch": 0.28735255570117957,
      "grad_norm": 0.14999398589134216,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 1754
    },
    {
      "epoch": 0.28751638269986896,
      "grad_norm": 0.1457619071006775,
      "learning_rate": 0.0001,
      "loss": 1.4778,
      "step": 1755
    },
    {
      "epoch": 0.28768020969855834,
      "grad_norm": 0.1569737046957016,
      "learning_rate": 0.0001,
      "loss": 1.4688,
      "step": 1756
    },
    {
      "epoch": 0.28784403669724773,
      "grad_norm": 0.1441645324230194,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 1757
    },
    {
      "epoch": 0.2880078636959371,
      "grad_norm": 0.1445416361093521,
      "learning_rate": 0.0001,
      "loss": 1.4647,
      "step": 1758
    },
    {
      "epoch": 0.28817169069462645,
      "grad_norm": 0.13890433311462402,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 1759
    },
    {
      "epoch": 0.28833551769331583,
      "grad_norm": 0.13177503645420074,
      "learning_rate": 0.0001,
      "loss": 1.4654,
      "step": 1760
    },
    {
      "epoch": 0.2884993446920052,
      "grad_norm": 0.1394852101802826,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 1761
    },
    {
      "epoch": 0.2886631716906946,
      "grad_norm": 0.14440110325813293,
      "learning_rate": 0.0001,
      "loss": 1.4576,
      "step": 1762
    },
    {
      "epoch": 0.288826998689384,
      "grad_norm": 0.1464320719242096,
      "learning_rate": 0.0001,
      "loss": 1.4542,
      "step": 1763
    },
    {
      "epoch": 0.2889908256880734,
      "grad_norm": 0.13260774314403534,
      "learning_rate": 0.0001,
      "loss": 1.3972,
      "step": 1764
    },
    {
      "epoch": 0.28915465268676277,
      "grad_norm": 0.13891860842704773,
      "learning_rate": 0.0001,
      "loss": 1.3835,
      "step": 1765
    },
    {
      "epoch": 0.28931847968545216,
      "grad_norm": 0.14175495505332947,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 1766
    },
    {
      "epoch": 0.28948230668414154,
      "grad_norm": 0.14236563444137573,
      "learning_rate": 0.0001,
      "loss": 1.4913,
      "step": 1767
    },
    {
      "epoch": 0.28964613368283093,
      "grad_norm": 0.13660913705825806,
      "learning_rate": 0.0001,
      "loss": 1.4306,
      "step": 1768
    },
    {
      "epoch": 0.2898099606815203,
      "grad_norm": 0.14282740652561188,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 1769
    },
    {
      "epoch": 0.2899737876802097,
      "grad_norm": 0.14561021327972412,
      "learning_rate": 0.0001,
      "loss": 1.3563,
      "step": 1770
    },
    {
      "epoch": 0.2901376146788991,
      "grad_norm": 0.13896389305591583,
      "learning_rate": 0.0001,
      "loss": 1.3657,
      "step": 1771
    },
    {
      "epoch": 0.2903014416775885,
      "grad_norm": 0.13997530937194824,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 1772
    },
    {
      "epoch": 0.29046526867627787,
      "grad_norm": 0.14409686625003815,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 1773
    },
    {
      "epoch": 0.29062909567496725,
      "grad_norm": 0.14051014184951782,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 1774
    },
    {
      "epoch": 0.29079292267365664,
      "grad_norm": 0.13707299530506134,
      "learning_rate": 0.0001,
      "loss": 1.4747,
      "step": 1775
    },
    {
      "epoch": 0.290956749672346,
      "grad_norm": 0.14451894164085388,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 1776
    },
    {
      "epoch": 0.2911205766710354,
      "grad_norm": 0.13883142173290253,
      "learning_rate": 0.0001,
      "loss": 1.3728,
      "step": 1777
    },
    {
      "epoch": 0.29128440366972475,
      "grad_norm": 0.14165426790714264,
      "learning_rate": 0.0001,
      "loss": 1.4628,
      "step": 1778
    },
    {
      "epoch": 0.29144823066841413,
      "grad_norm": 0.13663309812545776,
      "learning_rate": 0.0001,
      "loss": 1.4109,
      "step": 1779
    },
    {
      "epoch": 0.2916120576671035,
      "grad_norm": 0.14142049849033356,
      "learning_rate": 0.0001,
      "loss": 1.4688,
      "step": 1780
    },
    {
      "epoch": 0.2917758846657929,
      "grad_norm": 0.13765829801559448,
      "learning_rate": 0.0001,
      "loss": 1.5076,
      "step": 1781
    },
    {
      "epoch": 0.2919397116644823,
      "grad_norm": 0.14551730453968048,
      "learning_rate": 0.0001,
      "loss": 1.4088,
      "step": 1782
    },
    {
      "epoch": 0.2921035386631717,
      "grad_norm": 0.14052410423755646,
      "learning_rate": 0.0001,
      "loss": 1.4077,
      "step": 1783
    },
    {
      "epoch": 0.29226736566186107,
      "grad_norm": 0.14411698281764984,
      "learning_rate": 0.0001,
      "loss": 1.4629,
      "step": 1784
    },
    {
      "epoch": 0.29243119266055045,
      "grad_norm": 0.1491684764623642,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 1785
    },
    {
      "epoch": 0.29259501965923984,
      "grad_norm": 0.1399351954460144,
      "learning_rate": 0.0001,
      "loss": 1.4487,
      "step": 1786
    },
    {
      "epoch": 0.29275884665792923,
      "grad_norm": 0.13687637448310852,
      "learning_rate": 0.0001,
      "loss": 1.3117,
      "step": 1787
    },
    {
      "epoch": 0.2929226736566186,
      "grad_norm": 0.1467146873474121,
      "learning_rate": 0.0001,
      "loss": 1.3787,
      "step": 1788
    },
    {
      "epoch": 0.293086500655308,
      "grad_norm": 0.14483210444450378,
      "learning_rate": 0.0001,
      "loss": 1.4105,
      "step": 1789
    },
    {
      "epoch": 0.2932503276539974,
      "grad_norm": 0.13824933767318726,
      "learning_rate": 0.0001,
      "loss": 1.4815,
      "step": 1790
    },
    {
      "epoch": 0.2934141546526868,
      "grad_norm": 0.1455807089805603,
      "learning_rate": 0.0001,
      "loss": 1.3944,
      "step": 1791
    },
    {
      "epoch": 0.29357798165137616,
      "grad_norm": 0.14608709514141083,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 1792
    },
    {
      "epoch": 0.29374180865006555,
      "grad_norm": 0.15210823714733124,
      "learning_rate": 0.0001,
      "loss": 1.51,
      "step": 1793
    },
    {
      "epoch": 0.29390563564875494,
      "grad_norm": 0.14679822325706482,
      "learning_rate": 0.0001,
      "loss": 1.4229,
      "step": 1794
    },
    {
      "epoch": 0.2940694626474443,
      "grad_norm": 0.14654439687728882,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 1795
    },
    {
      "epoch": 0.29423328964613366,
      "grad_norm": 0.14774249494075775,
      "learning_rate": 0.0001,
      "loss": 1.4476,
      "step": 1796
    },
    {
      "epoch": 0.29439711664482304,
      "grad_norm": 0.14294812083244324,
      "learning_rate": 0.0001,
      "loss": 1.4031,
      "step": 1797
    },
    {
      "epoch": 0.29456094364351243,
      "grad_norm": 0.1465403288602829,
      "learning_rate": 0.0001,
      "loss": 1.4608,
      "step": 1798
    },
    {
      "epoch": 0.2947247706422018,
      "grad_norm": 0.144893616437912,
      "learning_rate": 0.0001,
      "loss": 1.4246,
      "step": 1799
    },
    {
      "epoch": 0.2948885976408912,
      "grad_norm": 0.1358432173728943,
      "learning_rate": 0.0001,
      "loss": 1.3463,
      "step": 1800
    },
    {
      "epoch": 0.2950524246395806,
      "grad_norm": 0.14153340458869934,
      "learning_rate": 0.0001,
      "loss": 1.4622,
      "step": 1801
    },
    {
      "epoch": 0.29521625163827,
      "grad_norm": 0.14054498076438904,
      "learning_rate": 0.0001,
      "loss": 1.4593,
      "step": 1802
    },
    {
      "epoch": 0.29538007863695936,
      "grad_norm": 0.13735920190811157,
      "learning_rate": 0.0001,
      "loss": 1.485,
      "step": 1803
    },
    {
      "epoch": 0.29554390563564875,
      "grad_norm": 0.1454010009765625,
      "learning_rate": 0.0001,
      "loss": 1.363,
      "step": 1804
    },
    {
      "epoch": 0.29570773263433814,
      "grad_norm": 0.14070762693881989,
      "learning_rate": 0.0001,
      "loss": 1.4824,
      "step": 1805
    },
    {
      "epoch": 0.2958715596330275,
      "grad_norm": 0.17583850026130676,
      "learning_rate": 0.0001,
      "loss": 1.4725,
      "step": 1806
    },
    {
      "epoch": 0.2960353866317169,
      "grad_norm": 0.1433008313179016,
      "learning_rate": 0.0001,
      "loss": 1.411,
      "step": 1807
    },
    {
      "epoch": 0.2961992136304063,
      "grad_norm": 0.13722997903823853,
      "learning_rate": 0.0001,
      "loss": 1.3567,
      "step": 1808
    },
    {
      "epoch": 0.2963630406290957,
      "grad_norm": 0.13441462814807892,
      "learning_rate": 0.0001,
      "loss": 1.3496,
      "step": 1809
    },
    {
      "epoch": 0.2965268676277851,
      "grad_norm": 0.15105511248111725,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 1810
    },
    {
      "epoch": 0.29669069462647446,
      "grad_norm": 0.1411907821893692,
      "learning_rate": 0.0001,
      "loss": 1.4726,
      "step": 1811
    },
    {
      "epoch": 0.29685452162516385,
      "grad_norm": 0.15442931652069092,
      "learning_rate": 0.0001,
      "loss": 1.4739,
      "step": 1812
    },
    {
      "epoch": 0.29701834862385323,
      "grad_norm": 0.14104041457176208,
      "learning_rate": 0.0001,
      "loss": 1.4569,
      "step": 1813
    },
    {
      "epoch": 0.2971821756225426,
      "grad_norm": 0.13553574681282043,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 1814
    },
    {
      "epoch": 0.29734600262123195,
      "grad_norm": 0.1500244438648224,
      "learning_rate": 0.0001,
      "loss": 1.3941,
      "step": 1815
    },
    {
      "epoch": 0.29750982961992134,
      "grad_norm": 0.1362808495759964,
      "learning_rate": 0.0001,
      "loss": 1.4203,
      "step": 1816
    },
    {
      "epoch": 0.2976736566186107,
      "grad_norm": 0.14288361370563507,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 1817
    },
    {
      "epoch": 0.2978374836173001,
      "grad_norm": 0.152119979262352,
      "learning_rate": 0.0001,
      "loss": 1.4119,
      "step": 1818
    },
    {
      "epoch": 0.2980013106159895,
      "grad_norm": 0.14612405002117157,
      "learning_rate": 0.0001,
      "loss": 1.4667,
      "step": 1819
    },
    {
      "epoch": 0.2981651376146789,
      "grad_norm": 0.14316248893737793,
      "learning_rate": 0.0001,
      "loss": 1.4765,
      "step": 1820
    },
    {
      "epoch": 0.2983289646133683,
      "grad_norm": 0.14498436450958252,
      "learning_rate": 0.0001,
      "loss": 1.4523,
      "step": 1821
    },
    {
      "epoch": 0.29849279161205766,
      "grad_norm": 0.14548833668231964,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 1822
    },
    {
      "epoch": 0.29865661861074705,
      "grad_norm": 0.13803046941757202,
      "learning_rate": 0.0001,
      "loss": 1.3739,
      "step": 1823
    },
    {
      "epoch": 0.29882044560943644,
      "grad_norm": 0.1518692970275879,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 1824
    },
    {
      "epoch": 0.2989842726081258,
      "grad_norm": 0.14503788948059082,
      "learning_rate": 0.0001,
      "loss": 1.4732,
      "step": 1825
    },
    {
      "epoch": 0.2991480996068152,
      "grad_norm": 0.13530954718589783,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 1826
    },
    {
      "epoch": 0.2993119266055046,
      "grad_norm": 0.14279820024967194,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 1827
    },
    {
      "epoch": 0.299475753604194,
      "grad_norm": 0.1414223313331604,
      "learning_rate": 0.0001,
      "loss": 1.4257,
      "step": 1828
    },
    {
      "epoch": 0.29963958060288337,
      "grad_norm": 0.14515173435211182,
      "learning_rate": 0.0001,
      "loss": 1.5488,
      "step": 1829
    },
    {
      "epoch": 0.29980340760157276,
      "grad_norm": 0.1509760171175003,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 1830
    },
    {
      "epoch": 0.29996723460026214,
      "grad_norm": 0.14276790618896484,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 1831
    },
    {
      "epoch": 0.30013106159895153,
      "grad_norm": 0.15405097603797913,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 1832
    },
    {
      "epoch": 0.3002948885976409,
      "grad_norm": 0.15489919483661652,
      "learning_rate": 0.0001,
      "loss": 1.4982,
      "step": 1833
    },
    {
      "epoch": 0.30045871559633025,
      "grad_norm": 0.1478624939918518,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 1834
    },
    {
      "epoch": 0.30062254259501964,
      "grad_norm": 0.1457146853208542,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 1835
    },
    {
      "epoch": 0.300786369593709,
      "grad_norm": 0.14380784332752228,
      "learning_rate": 0.0001,
      "loss": 1.6017,
      "step": 1836
    },
    {
      "epoch": 0.3009501965923984,
      "grad_norm": 0.1274847537279129,
      "learning_rate": 0.0001,
      "loss": 1.2083,
      "step": 1837
    },
    {
      "epoch": 0.3011140235910878,
      "grad_norm": 0.1382535994052887,
      "learning_rate": 0.0001,
      "loss": 1.4269,
      "step": 1838
    },
    {
      "epoch": 0.3012778505897772,
      "grad_norm": 0.1420762687921524,
      "learning_rate": 0.0001,
      "loss": 1.3796,
      "step": 1839
    },
    {
      "epoch": 0.30144167758846657,
      "grad_norm": 0.134435772895813,
      "learning_rate": 0.0001,
      "loss": 1.341,
      "step": 1840
    },
    {
      "epoch": 0.30160550458715596,
      "grad_norm": 0.1356593817472458,
      "learning_rate": 0.0001,
      "loss": 1.4392,
      "step": 1841
    },
    {
      "epoch": 0.30176933158584535,
      "grad_norm": 0.1421826332807541,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 1842
    },
    {
      "epoch": 0.30193315858453473,
      "grad_norm": 0.1485128104686737,
      "learning_rate": 0.0001,
      "loss": 1.4957,
      "step": 1843
    },
    {
      "epoch": 0.3020969855832241,
      "grad_norm": 0.14132030308246613,
      "learning_rate": 0.0001,
      "loss": 1.4059,
      "step": 1844
    },
    {
      "epoch": 0.3022608125819135,
      "grad_norm": 0.14259353280067444,
      "learning_rate": 0.0001,
      "loss": 1.4537,
      "step": 1845
    },
    {
      "epoch": 0.3024246395806029,
      "grad_norm": 0.1419815868139267,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 1846
    },
    {
      "epoch": 0.3025884665792923,
      "grad_norm": 0.14728984236717224,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 1847
    },
    {
      "epoch": 0.30275229357798167,
      "grad_norm": 0.1427375078201294,
      "learning_rate": 0.0001,
      "loss": 1.391,
      "step": 1848
    },
    {
      "epoch": 0.30291612057667106,
      "grad_norm": 0.149126797914505,
      "learning_rate": 0.0001,
      "loss": 1.4564,
      "step": 1849
    },
    {
      "epoch": 0.30307994757536044,
      "grad_norm": 0.14444901049137115,
      "learning_rate": 0.0001,
      "loss": 1.349,
      "step": 1850
    },
    {
      "epoch": 0.30324377457404983,
      "grad_norm": 0.14822527766227722,
      "learning_rate": 0.0001,
      "loss": 1.5117,
      "step": 1851
    },
    {
      "epoch": 0.30340760157273916,
      "grad_norm": 0.15321433544158936,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 1852
    },
    {
      "epoch": 0.30357142857142855,
      "grad_norm": 0.13699325919151306,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 1853
    },
    {
      "epoch": 0.30373525557011793,
      "grad_norm": 0.15328937768936157,
      "learning_rate": 0.0001,
      "loss": 1.5033,
      "step": 1854
    },
    {
      "epoch": 0.3038990825688073,
      "grad_norm": 0.14392180740833282,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 1855
    },
    {
      "epoch": 0.3040629095674967,
      "grad_norm": 0.13604719936847687,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 1856
    },
    {
      "epoch": 0.3042267365661861,
      "grad_norm": 0.14484088122844696,
      "learning_rate": 0.0001,
      "loss": 1.4469,
      "step": 1857
    },
    {
      "epoch": 0.3043905635648755,
      "grad_norm": 0.13545481860637665,
      "learning_rate": 0.0001,
      "loss": 1.3769,
      "step": 1858
    },
    {
      "epoch": 0.30455439056356487,
      "grad_norm": 0.14406892657279968,
      "learning_rate": 0.0001,
      "loss": 1.4623,
      "step": 1859
    },
    {
      "epoch": 0.30471821756225426,
      "grad_norm": 0.14074856042861938,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 1860
    },
    {
      "epoch": 0.30488204456094364,
      "grad_norm": 0.1432824432849884,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 1861
    },
    {
      "epoch": 0.30504587155963303,
      "grad_norm": 0.13405080139636993,
      "learning_rate": 0.0001,
      "loss": 1.3447,
      "step": 1862
    },
    {
      "epoch": 0.3052096985583224,
      "grad_norm": 0.1520107239484787,
      "learning_rate": 0.0001,
      "loss": 1.4711,
      "step": 1863
    },
    {
      "epoch": 0.3053735255570118,
      "grad_norm": 0.14694997668266296,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 1864
    },
    {
      "epoch": 0.3055373525557012,
      "grad_norm": 0.15092605352401733,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 1865
    },
    {
      "epoch": 0.3057011795543906,
      "grad_norm": 0.1388293206691742,
      "learning_rate": 0.0001,
      "loss": 1.3065,
      "step": 1866
    },
    {
      "epoch": 0.30586500655307997,
      "grad_norm": 0.1440192312002182,
      "learning_rate": 0.0001,
      "loss": 1.4823,
      "step": 1867
    },
    {
      "epoch": 0.30602883355176935,
      "grad_norm": 0.18856140971183777,
      "learning_rate": 0.0001,
      "loss": 1.5619,
      "step": 1868
    },
    {
      "epoch": 0.30619266055045874,
      "grad_norm": 0.1487552672624588,
      "learning_rate": 0.0001,
      "loss": 1.4379,
      "step": 1869
    },
    {
      "epoch": 0.3063564875491481,
      "grad_norm": 0.13507327437400818,
      "learning_rate": 0.0001,
      "loss": 1.3627,
      "step": 1870
    },
    {
      "epoch": 0.30652031454783746,
      "grad_norm": 0.13469074666500092,
      "learning_rate": 0.0001,
      "loss": 1.4014,
      "step": 1871
    },
    {
      "epoch": 0.30668414154652685,
      "grad_norm": 0.14463138580322266,
      "learning_rate": 0.0001,
      "loss": 1.4529,
      "step": 1872
    },
    {
      "epoch": 0.30684796854521623,
      "grad_norm": 0.14003746211528778,
      "learning_rate": 0.0001,
      "loss": 1.4601,
      "step": 1873
    },
    {
      "epoch": 0.3070117955439056,
      "grad_norm": 0.14733673632144928,
      "learning_rate": 0.0001,
      "loss": 1.4352,
      "step": 1874
    },
    {
      "epoch": 0.307175622542595,
      "grad_norm": 0.1437702775001526,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 1875
    },
    {
      "epoch": 0.3073394495412844,
      "grad_norm": 0.1330818086862564,
      "learning_rate": 0.0001,
      "loss": 1.3835,
      "step": 1876
    },
    {
      "epoch": 0.3075032765399738,
      "grad_norm": 0.14641380310058594,
      "learning_rate": 0.0001,
      "loss": 1.3727,
      "step": 1877
    },
    {
      "epoch": 0.30766710353866317,
      "grad_norm": 0.1498032957315445,
      "learning_rate": 0.0001,
      "loss": 1.3637,
      "step": 1878
    },
    {
      "epoch": 0.30783093053735255,
      "grad_norm": 0.14560869336128235,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 1879
    },
    {
      "epoch": 0.30799475753604194,
      "grad_norm": 0.14332719147205353,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 1880
    },
    {
      "epoch": 0.30815858453473133,
      "grad_norm": 0.14901988208293915,
      "learning_rate": 0.0001,
      "loss": 1.3555,
      "step": 1881
    },
    {
      "epoch": 0.3083224115334207,
      "grad_norm": 0.1452895849943161,
      "learning_rate": 0.0001,
      "loss": 1.38,
      "step": 1882
    },
    {
      "epoch": 0.3084862385321101,
      "grad_norm": 0.14490410685539246,
      "learning_rate": 0.0001,
      "loss": 1.4227,
      "step": 1883
    },
    {
      "epoch": 0.3086500655307995,
      "grad_norm": 0.14232099056243896,
      "learning_rate": 0.0001,
      "loss": 1.3436,
      "step": 1884
    },
    {
      "epoch": 0.3088138925294889,
      "grad_norm": 0.14323531091213226,
      "learning_rate": 0.0001,
      "loss": 1.3626,
      "step": 1885
    },
    {
      "epoch": 0.30897771952817826,
      "grad_norm": 0.14542457461357117,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 1886
    },
    {
      "epoch": 0.30914154652686765,
      "grad_norm": 0.13863718509674072,
      "learning_rate": 0.0001,
      "loss": 1.3651,
      "step": 1887
    },
    {
      "epoch": 0.30930537352555704,
      "grad_norm": 0.1510324776172638,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 1888
    },
    {
      "epoch": 0.30946920052424637,
      "grad_norm": 0.14582128822803497,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 1889
    },
    {
      "epoch": 0.30963302752293576,
      "grad_norm": 0.14115621149539948,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 1890
    },
    {
      "epoch": 0.30979685452162514,
      "grad_norm": 0.136497363448143,
      "learning_rate": 0.0001,
      "loss": 1.4744,
      "step": 1891
    },
    {
      "epoch": 0.30996068152031453,
      "grad_norm": 0.14306990802288055,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 1892
    },
    {
      "epoch": 0.3101245085190039,
      "grad_norm": 0.14317022264003754,
      "learning_rate": 0.0001,
      "loss": 1.4534,
      "step": 1893
    },
    {
      "epoch": 0.3102883355176933,
      "grad_norm": 0.14112688601016998,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 1894
    },
    {
      "epoch": 0.3104521625163827,
      "grad_norm": 0.15025830268859863,
      "learning_rate": 0.0001,
      "loss": 1.561,
      "step": 1895
    },
    {
      "epoch": 0.3106159895150721,
      "grad_norm": 0.1421600878238678,
      "learning_rate": 0.0001,
      "loss": 1.5463,
      "step": 1896
    },
    {
      "epoch": 0.31077981651376146,
      "grad_norm": 0.13857091963291168,
      "learning_rate": 0.0001,
      "loss": 1.3195,
      "step": 1897
    },
    {
      "epoch": 0.31094364351245085,
      "grad_norm": 0.14120925962924957,
      "learning_rate": 0.0001,
      "loss": 1.3887,
      "step": 1898
    },
    {
      "epoch": 0.31110747051114024,
      "grad_norm": 0.14421389997005463,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 1899
    },
    {
      "epoch": 0.3112712975098296,
      "grad_norm": 0.14588037133216858,
      "learning_rate": 0.0001,
      "loss": 1.513,
      "step": 1900
    },
    {
      "epoch": 0.311435124508519,
      "grad_norm": 0.14421522617340088,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 1901
    },
    {
      "epoch": 0.3115989515072084,
      "grad_norm": 0.134551540017128,
      "learning_rate": 0.0001,
      "loss": 1.4785,
      "step": 1902
    },
    {
      "epoch": 0.3117627785058978,
      "grad_norm": 0.15922300517559052,
      "learning_rate": 0.0001,
      "loss": 1.4745,
      "step": 1903
    },
    {
      "epoch": 0.3119266055045872,
      "grad_norm": 0.14235961437225342,
      "learning_rate": 0.0001,
      "loss": 1.4022,
      "step": 1904
    },
    {
      "epoch": 0.31209043250327656,
      "grad_norm": 0.13776586949825287,
      "learning_rate": 0.0001,
      "loss": 1.4112,
      "step": 1905
    },
    {
      "epoch": 0.31225425950196595,
      "grad_norm": 0.13686664402484894,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 1906
    },
    {
      "epoch": 0.31241808650065533,
      "grad_norm": 0.14437633752822876,
      "learning_rate": 0.0001,
      "loss": 1.4294,
      "step": 1907
    },
    {
      "epoch": 0.31258191349934467,
      "grad_norm": 0.1464550644159317,
      "learning_rate": 0.0001,
      "loss": 1.4853,
      "step": 1908
    },
    {
      "epoch": 0.31274574049803405,
      "grad_norm": 0.14967568218708038,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 1909
    },
    {
      "epoch": 0.31290956749672344,
      "grad_norm": 0.14646577835083008,
      "learning_rate": 0.0001,
      "loss": 1.4955,
      "step": 1910
    },
    {
      "epoch": 0.3130733944954128,
      "grad_norm": 0.14183512330055237,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 1911
    },
    {
      "epoch": 0.3132372214941022,
      "grad_norm": 0.13609865307807922,
      "learning_rate": 0.0001,
      "loss": 1.3936,
      "step": 1912
    },
    {
      "epoch": 0.3134010484927916,
      "grad_norm": 0.14022505283355713,
      "learning_rate": 0.0001,
      "loss": 1.3819,
      "step": 1913
    },
    {
      "epoch": 0.313564875491481,
      "grad_norm": 0.13743114471435547,
      "learning_rate": 0.0001,
      "loss": 1.4972,
      "step": 1914
    },
    {
      "epoch": 0.3137287024901704,
      "grad_norm": 0.13810060918331146,
      "learning_rate": 0.0001,
      "loss": 1.5092,
      "step": 1915
    },
    {
      "epoch": 0.31389252948885976,
      "grad_norm": 0.14669278264045715,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 1916
    },
    {
      "epoch": 0.31405635648754915,
      "grad_norm": 0.15147681534290314,
      "learning_rate": 0.0001,
      "loss": 1.5038,
      "step": 1917
    },
    {
      "epoch": 0.31422018348623854,
      "grad_norm": 0.1394781619310379,
      "learning_rate": 0.0001,
      "loss": 1.4927,
      "step": 1918
    },
    {
      "epoch": 0.3143840104849279,
      "grad_norm": 0.14344221353530884,
      "learning_rate": 0.0001,
      "loss": 1.3503,
      "step": 1919
    },
    {
      "epoch": 0.3145478374836173,
      "grad_norm": 0.1432821899652481,
      "learning_rate": 0.0001,
      "loss": 1.4559,
      "step": 1920
    },
    {
      "epoch": 0.3147116644823067,
      "grad_norm": 0.14150455594062805,
      "learning_rate": 0.0001,
      "loss": 1.4218,
      "step": 1921
    },
    {
      "epoch": 0.3148754914809961,
      "grad_norm": 0.15129810571670532,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 1922
    },
    {
      "epoch": 0.31503931847968547,
      "grad_norm": 0.14512160420417786,
      "learning_rate": 0.0001,
      "loss": 1.4155,
      "step": 1923
    },
    {
      "epoch": 0.31520314547837486,
      "grad_norm": 0.1422361582517624,
      "learning_rate": 0.0001,
      "loss": 1.4635,
      "step": 1924
    },
    {
      "epoch": 0.31536697247706424,
      "grad_norm": 0.14217732846736908,
      "learning_rate": 0.0001,
      "loss": 1.3694,
      "step": 1925
    },
    {
      "epoch": 0.31553079947575363,
      "grad_norm": 0.1426618993282318,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 1926
    },
    {
      "epoch": 0.31569462647444296,
      "grad_norm": 0.14184679090976715,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 1927
    },
    {
      "epoch": 0.31585845347313235,
      "grad_norm": 0.14089162647724152,
      "learning_rate": 0.0001,
      "loss": 1.497,
      "step": 1928
    },
    {
      "epoch": 0.31602228047182174,
      "grad_norm": 0.13724742829799652,
      "learning_rate": 0.0001,
      "loss": 1.2558,
      "step": 1929
    },
    {
      "epoch": 0.3161861074705111,
      "grad_norm": 0.16532741487026215,
      "learning_rate": 0.0001,
      "loss": 1.338,
      "step": 1930
    },
    {
      "epoch": 0.3163499344692005,
      "grad_norm": 0.13555799424648285,
      "learning_rate": 0.0001,
      "loss": 1.28,
      "step": 1931
    },
    {
      "epoch": 0.3165137614678899,
      "grad_norm": 0.14544038474559784,
      "learning_rate": 0.0001,
      "loss": 1.504,
      "step": 1932
    },
    {
      "epoch": 0.3166775884665793,
      "grad_norm": 0.13327200710773468,
      "learning_rate": 0.0001,
      "loss": 1.4189,
      "step": 1933
    },
    {
      "epoch": 0.31684141546526867,
      "grad_norm": 0.14231349527835846,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 1934
    },
    {
      "epoch": 0.31700524246395806,
      "grad_norm": 0.139101043343544,
      "learning_rate": 0.0001,
      "loss": 1.4509,
      "step": 1935
    },
    {
      "epoch": 0.31716906946264745,
      "grad_norm": 0.13938872516155243,
      "learning_rate": 0.0001,
      "loss": 1.4226,
      "step": 1936
    },
    {
      "epoch": 0.31733289646133683,
      "grad_norm": 0.1485563963651657,
      "learning_rate": 0.0001,
      "loss": 1.3891,
      "step": 1937
    },
    {
      "epoch": 0.3174967234600262,
      "grad_norm": 0.13743644952774048,
      "learning_rate": 0.0001,
      "loss": 1.3662,
      "step": 1938
    },
    {
      "epoch": 0.3176605504587156,
      "grad_norm": 0.14009110629558563,
      "learning_rate": 0.0001,
      "loss": 1.3627,
      "step": 1939
    },
    {
      "epoch": 0.317824377457405,
      "grad_norm": 0.1419152468442917,
      "learning_rate": 0.0001,
      "loss": 1.5333,
      "step": 1940
    },
    {
      "epoch": 0.3179882044560944,
      "grad_norm": 0.1541091948747635,
      "learning_rate": 0.0001,
      "loss": 1.4406,
      "step": 1941
    },
    {
      "epoch": 0.31815203145478377,
      "grad_norm": 0.14419621229171753,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 1942
    },
    {
      "epoch": 0.31831585845347315,
      "grad_norm": 0.14475376904010773,
      "learning_rate": 0.0001,
      "loss": 1.4706,
      "step": 1943
    },
    {
      "epoch": 0.31847968545216254,
      "grad_norm": 0.14800912141799927,
      "learning_rate": 0.0001,
      "loss": 1.4265,
      "step": 1944
    },
    {
      "epoch": 0.3186435124508519,
      "grad_norm": 0.15022395551204681,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 1945
    },
    {
      "epoch": 0.31880733944954126,
      "grad_norm": 0.14572060108184814,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 1946
    },
    {
      "epoch": 0.31897116644823065,
      "grad_norm": 0.1381460726261139,
      "learning_rate": 0.0001,
      "loss": 1.3732,
      "step": 1947
    },
    {
      "epoch": 0.31913499344692003,
      "grad_norm": 0.14553400874137878,
      "learning_rate": 0.0001,
      "loss": 1.3623,
      "step": 1948
    },
    {
      "epoch": 0.3192988204456094,
      "grad_norm": 0.13992440700531006,
      "learning_rate": 0.0001,
      "loss": 1.3972,
      "step": 1949
    },
    {
      "epoch": 0.3194626474442988,
      "grad_norm": 0.1395554393529892,
      "learning_rate": 0.0001,
      "loss": 1.4271,
      "step": 1950
    },
    {
      "epoch": 0.3196264744429882,
      "grad_norm": 0.13878998160362244,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 1951
    },
    {
      "epoch": 0.3197903014416776,
      "grad_norm": 0.1444920003414154,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 1952
    },
    {
      "epoch": 0.31995412844036697,
      "grad_norm": 0.1468057483434677,
      "learning_rate": 0.0001,
      "loss": 1.4324,
      "step": 1953
    },
    {
      "epoch": 0.32011795543905636,
      "grad_norm": 0.13308681547641754,
      "learning_rate": 0.0001,
      "loss": 1.4181,
      "step": 1954
    },
    {
      "epoch": 0.32028178243774574,
      "grad_norm": 0.1399073302745819,
      "learning_rate": 0.0001,
      "loss": 1.3963,
      "step": 1955
    },
    {
      "epoch": 0.32044560943643513,
      "grad_norm": 0.14922945201396942,
      "learning_rate": 0.0001,
      "loss": 1.4567,
      "step": 1956
    },
    {
      "epoch": 0.3206094364351245,
      "grad_norm": 0.14971600472927094,
      "learning_rate": 0.0001,
      "loss": 1.5371,
      "step": 1957
    },
    {
      "epoch": 0.3207732634338139,
      "grad_norm": 0.15927162766456604,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 1958
    },
    {
      "epoch": 0.3209370904325033,
      "grad_norm": 0.1403357982635498,
      "learning_rate": 0.0001,
      "loss": 1.4109,
      "step": 1959
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 0.14407004415988922,
      "learning_rate": 0.0001,
      "loss": 1.4053,
      "step": 1960
    },
    {
      "epoch": 0.32126474442988207,
      "grad_norm": 0.14436763525009155,
      "learning_rate": 0.0001,
      "loss": 1.3187,
      "step": 1961
    },
    {
      "epoch": 0.32142857142857145,
      "grad_norm": 0.1395435780286789,
      "learning_rate": 0.0001,
      "loss": 1.3505,
      "step": 1962
    },
    {
      "epoch": 0.32159239842726084,
      "grad_norm": 0.16961164772510529,
      "learning_rate": 0.0001,
      "loss": 1.4159,
      "step": 1963
    },
    {
      "epoch": 0.32175622542595017,
      "grad_norm": 0.14652565121650696,
      "learning_rate": 0.0001,
      "loss": 1.4077,
      "step": 1964
    },
    {
      "epoch": 0.32192005242463956,
      "grad_norm": 0.14651401340961456,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 1965
    },
    {
      "epoch": 0.32208387942332894,
      "grad_norm": 0.13701294362545013,
      "learning_rate": 0.0001,
      "loss": 1.4202,
      "step": 1966
    },
    {
      "epoch": 0.32224770642201833,
      "grad_norm": 0.14735658466815948,
      "learning_rate": 0.0001,
      "loss": 1.5738,
      "step": 1967
    },
    {
      "epoch": 0.3224115334207077,
      "grad_norm": 0.1457902193069458,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 1968
    },
    {
      "epoch": 0.3225753604193971,
      "grad_norm": 0.147698774933815,
      "learning_rate": 0.0001,
      "loss": 1.4502,
      "step": 1969
    },
    {
      "epoch": 0.3227391874180865,
      "grad_norm": 0.14751797914505005,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 1970
    },
    {
      "epoch": 0.3229030144167759,
      "grad_norm": 0.1453886479139328,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 1971
    },
    {
      "epoch": 0.32306684141546527,
      "grad_norm": 0.14671465754508972,
      "learning_rate": 0.0001,
      "loss": 1.4269,
      "step": 1972
    },
    {
      "epoch": 0.32323066841415465,
      "grad_norm": 0.14755694568157196,
      "learning_rate": 0.0001,
      "loss": 1.5363,
      "step": 1973
    },
    {
      "epoch": 0.32339449541284404,
      "grad_norm": 0.14367824792861938,
      "learning_rate": 0.0001,
      "loss": 1.3263,
      "step": 1974
    },
    {
      "epoch": 0.32355832241153343,
      "grad_norm": 0.14053252339363098,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 1975
    },
    {
      "epoch": 0.3237221494102228,
      "grad_norm": 0.15858472883701324,
      "learning_rate": 0.0001,
      "loss": 1.481,
      "step": 1976
    },
    {
      "epoch": 0.3238859764089122,
      "grad_norm": 0.1399313062429428,
      "learning_rate": 0.0001,
      "loss": 1.4119,
      "step": 1977
    },
    {
      "epoch": 0.3240498034076016,
      "grad_norm": 0.13843382894992828,
      "learning_rate": 0.0001,
      "loss": 1.4426,
      "step": 1978
    },
    {
      "epoch": 0.324213630406291,
      "grad_norm": 0.14544877409934998,
      "learning_rate": 0.0001,
      "loss": 1.4288,
      "step": 1979
    },
    {
      "epoch": 0.32437745740498036,
      "grad_norm": 0.14690731465816498,
      "learning_rate": 0.0001,
      "loss": 1.3415,
      "step": 1980
    },
    {
      "epoch": 0.32454128440366975,
      "grad_norm": 0.14837132394313812,
      "learning_rate": 0.0001,
      "loss": 1.4893,
      "step": 1981
    },
    {
      "epoch": 0.3247051114023591,
      "grad_norm": 0.14288003742694855,
      "learning_rate": 0.0001,
      "loss": 1.3678,
      "step": 1982
    },
    {
      "epoch": 0.32486893840104847,
      "grad_norm": 0.1401015669107437,
      "learning_rate": 0.0001,
      "loss": 1.4306,
      "step": 1983
    },
    {
      "epoch": 0.32503276539973786,
      "grad_norm": 0.14192764461040497,
      "learning_rate": 0.0001,
      "loss": 1.4705,
      "step": 1984
    },
    {
      "epoch": 0.32519659239842724,
      "grad_norm": 0.14742350578308105,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 1985
    },
    {
      "epoch": 0.32536041939711663,
      "grad_norm": 0.14103855192661285,
      "learning_rate": 0.0001,
      "loss": 1.343,
      "step": 1986
    },
    {
      "epoch": 0.325524246395806,
      "grad_norm": 0.14006884396076202,
      "learning_rate": 0.0001,
      "loss": 1.3683,
      "step": 1987
    },
    {
      "epoch": 0.3256880733944954,
      "grad_norm": 0.14080122113227844,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 1988
    },
    {
      "epoch": 0.3258519003931848,
      "grad_norm": 0.14507192373275757,
      "learning_rate": 0.0001,
      "loss": 1.3555,
      "step": 1989
    },
    {
      "epoch": 0.3260157273918742,
      "grad_norm": 0.1395675241947174,
      "learning_rate": 0.0001,
      "loss": 1.2812,
      "step": 1990
    },
    {
      "epoch": 0.32617955439056356,
      "grad_norm": 0.15542568266391754,
      "learning_rate": 0.0001,
      "loss": 1.4452,
      "step": 1991
    },
    {
      "epoch": 0.32634338138925295,
      "grad_norm": 0.14061613380908966,
      "learning_rate": 0.0001,
      "loss": 1.3924,
      "step": 1992
    },
    {
      "epoch": 0.32650720838794234,
      "grad_norm": 0.1470685750246048,
      "learning_rate": 0.0001,
      "loss": 1.3802,
      "step": 1993
    },
    {
      "epoch": 0.3266710353866317,
      "grad_norm": 0.1451471447944641,
      "learning_rate": 0.0001,
      "loss": 1.4439,
      "step": 1994
    },
    {
      "epoch": 0.3268348623853211,
      "grad_norm": 0.1415880173444748,
      "learning_rate": 0.0001,
      "loss": 1.4155,
      "step": 1995
    },
    {
      "epoch": 0.3269986893840105,
      "grad_norm": 0.14750997722148895,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 1996
    },
    {
      "epoch": 0.3271625163826999,
      "grad_norm": 0.13700968027114868,
      "learning_rate": 0.0001,
      "loss": 1.51,
      "step": 1997
    },
    {
      "epoch": 0.3273263433813893,
      "grad_norm": 0.1536913812160492,
      "learning_rate": 0.0001,
      "loss": 1.4752,
      "step": 1998
    },
    {
      "epoch": 0.32749017038007866,
      "grad_norm": 0.15308912098407745,
      "learning_rate": 0.0001,
      "loss": 1.5418,
      "step": 1999
    },
    {
      "epoch": 0.32765399737876805,
      "grad_norm": 0.1324336975812912,
      "learning_rate": 0.0001,
      "loss": 1.3722,
      "step": 2000
    },
    {
      "epoch": 0.3278178243774574,
      "grad_norm": 0.1536037027835846,
      "learning_rate": 0.0001,
      "loss": 1.4495,
      "step": 2001
    },
    {
      "epoch": 0.32798165137614677,
      "grad_norm": 0.14633017778396606,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 2002
    },
    {
      "epoch": 0.32814547837483615,
      "grad_norm": 0.14599743485450745,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 2003
    },
    {
      "epoch": 0.32830930537352554,
      "grad_norm": 0.15829786658287048,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 2004
    },
    {
      "epoch": 0.3284731323722149,
      "grad_norm": 0.13929295539855957,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 2005
    },
    {
      "epoch": 0.3286369593709043,
      "grad_norm": 0.1558154672384262,
      "learning_rate": 0.0001,
      "loss": 1.4519,
      "step": 2006
    },
    {
      "epoch": 0.3288007863695937,
      "grad_norm": 0.14866873621940613,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 2007
    },
    {
      "epoch": 0.3289646133682831,
      "grad_norm": 0.14736075699329376,
      "learning_rate": 0.0001,
      "loss": 1.3938,
      "step": 2008
    },
    {
      "epoch": 0.3291284403669725,
      "grad_norm": 0.1597878485918045,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 2009
    },
    {
      "epoch": 0.32929226736566186,
      "grad_norm": 0.1557389348745346,
      "learning_rate": 0.0001,
      "loss": 1.4634,
      "step": 2010
    },
    {
      "epoch": 0.32945609436435125,
      "grad_norm": 0.1399179846048355,
      "learning_rate": 0.0001,
      "loss": 1.3534,
      "step": 2011
    },
    {
      "epoch": 0.32961992136304064,
      "grad_norm": 0.15231259167194366,
      "learning_rate": 0.0001,
      "loss": 1.2911,
      "step": 2012
    },
    {
      "epoch": 0.32978374836173,
      "grad_norm": 0.18055400252342224,
      "learning_rate": 0.0001,
      "loss": 1.5412,
      "step": 2013
    },
    {
      "epoch": 0.3299475753604194,
      "grad_norm": 0.1381160169839859,
      "learning_rate": 0.0001,
      "loss": 1.4479,
      "step": 2014
    },
    {
      "epoch": 0.3301114023591088,
      "grad_norm": 0.15641556680202484,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 2015
    },
    {
      "epoch": 0.3302752293577982,
      "grad_norm": 0.15361428260803223,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 2016
    },
    {
      "epoch": 0.33043905635648757,
      "grad_norm": 0.1451871544122696,
      "learning_rate": 0.0001,
      "loss": 1.3747,
      "step": 2017
    },
    {
      "epoch": 0.33060288335517696,
      "grad_norm": 0.15201956033706665,
      "learning_rate": 0.0001,
      "loss": 1.3568,
      "step": 2018
    },
    {
      "epoch": 0.33076671035386634,
      "grad_norm": 0.14211899042129517,
      "learning_rate": 0.0001,
      "loss": 1.3484,
      "step": 2019
    },
    {
      "epoch": 0.3309305373525557,
      "grad_norm": 0.13947446644306183,
      "learning_rate": 0.0001,
      "loss": 1.3039,
      "step": 2020
    },
    {
      "epoch": 0.33109436435124506,
      "grad_norm": 0.15693870186805725,
      "learning_rate": 0.0001,
      "loss": 1.4098,
      "step": 2021
    },
    {
      "epoch": 0.33125819134993445,
      "grad_norm": 0.16319498419761658,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 2022
    },
    {
      "epoch": 0.33142201834862384,
      "grad_norm": 0.14558222889900208,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 2023
    },
    {
      "epoch": 0.3315858453473132,
      "grad_norm": 0.15057902038097382,
      "learning_rate": 0.0001,
      "loss": 1.4664,
      "step": 2024
    },
    {
      "epoch": 0.3317496723460026,
      "grad_norm": 0.16414856910705566,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 2025
    },
    {
      "epoch": 0.331913499344692,
      "grad_norm": 0.1322239637374878,
      "learning_rate": 0.0001,
      "loss": 1.2117,
      "step": 2026
    },
    {
      "epoch": 0.3320773263433814,
      "grad_norm": 0.1481717973947525,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 2027
    },
    {
      "epoch": 0.33224115334207077,
      "grad_norm": 0.14869138598442078,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 2028
    },
    {
      "epoch": 0.33240498034076016,
      "grad_norm": 0.1422293335199356,
      "learning_rate": 0.0001,
      "loss": 1.432,
      "step": 2029
    },
    {
      "epoch": 0.33256880733944955,
      "grad_norm": 0.14378374814987183,
      "learning_rate": 0.0001,
      "loss": 1.4487,
      "step": 2030
    },
    {
      "epoch": 0.33273263433813893,
      "grad_norm": 0.1472582370042801,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 2031
    },
    {
      "epoch": 0.3328964613368283,
      "grad_norm": 0.14162804186344147,
      "learning_rate": 0.0001,
      "loss": 1.4584,
      "step": 2032
    },
    {
      "epoch": 0.3330602883355177,
      "grad_norm": 0.1420387625694275,
      "learning_rate": 0.0001,
      "loss": 1.4669,
      "step": 2033
    },
    {
      "epoch": 0.3332241153342071,
      "grad_norm": 0.16325341165065765,
      "learning_rate": 0.0001,
      "loss": 1.4761,
      "step": 2034
    },
    {
      "epoch": 0.3333879423328965,
      "grad_norm": 0.14162102341651917,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 2035
    },
    {
      "epoch": 0.33355176933158587,
      "grad_norm": 0.14334933459758759,
      "learning_rate": 0.0001,
      "loss": 1.5623,
      "step": 2036
    },
    {
      "epoch": 0.33371559633027525,
      "grad_norm": 0.14544324576854706,
      "learning_rate": 0.0001,
      "loss": 1.4241,
      "step": 2037
    },
    {
      "epoch": 0.3338794233289646,
      "grad_norm": 0.14132477343082428,
      "learning_rate": 0.0001,
      "loss": 1.4221,
      "step": 2038
    },
    {
      "epoch": 0.334043250327654,
      "grad_norm": 0.14273221790790558,
      "learning_rate": 0.0001,
      "loss": 1.4264,
      "step": 2039
    },
    {
      "epoch": 0.33420707732634336,
      "grad_norm": 0.15094201266765594,
      "learning_rate": 0.0001,
      "loss": 1.48,
      "step": 2040
    },
    {
      "epoch": 0.33437090432503275,
      "grad_norm": 0.1487542986869812,
      "learning_rate": 0.0001,
      "loss": 1.4151,
      "step": 2041
    },
    {
      "epoch": 0.33453473132372213,
      "grad_norm": 0.15940453112125397,
      "learning_rate": 0.0001,
      "loss": 1.5247,
      "step": 2042
    },
    {
      "epoch": 0.3346985583224115,
      "grad_norm": 0.14899078011512756,
      "learning_rate": 0.0001,
      "loss": 1.413,
      "step": 2043
    },
    {
      "epoch": 0.3348623853211009,
      "grad_norm": 0.14664438366889954,
      "learning_rate": 0.0001,
      "loss": 1.5331,
      "step": 2044
    },
    {
      "epoch": 0.3350262123197903,
      "grad_norm": 0.14876455068588257,
      "learning_rate": 0.0001,
      "loss": 1.3482,
      "step": 2045
    },
    {
      "epoch": 0.3351900393184797,
      "grad_norm": 0.15632794797420502,
      "learning_rate": 0.0001,
      "loss": 1.3993,
      "step": 2046
    },
    {
      "epoch": 0.33535386631716907,
      "grad_norm": 0.15133985877037048,
      "learning_rate": 0.0001,
      "loss": 1.5156,
      "step": 2047
    },
    {
      "epoch": 0.33551769331585846,
      "grad_norm": 0.14279407262802124,
      "learning_rate": 0.0001,
      "loss": 1.4146,
      "step": 2048
    },
    {
      "epoch": 0.33568152031454784,
      "grad_norm": 0.1527835726737976,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 2049
    },
    {
      "epoch": 0.33584534731323723,
      "grad_norm": 0.14239096641540527,
      "learning_rate": 0.0001,
      "loss": 1.432,
      "step": 2050
    },
    {
      "epoch": 0.3360091743119266,
      "grad_norm": 0.15610910952091217,
      "learning_rate": 0.0001,
      "loss": 1.5108,
      "step": 2051
    },
    {
      "epoch": 0.336173001310616,
      "grad_norm": 0.15075740218162537,
      "learning_rate": 0.0001,
      "loss": 1.3813,
      "step": 2052
    },
    {
      "epoch": 0.3363368283093054,
      "grad_norm": 0.15795326232910156,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 2053
    },
    {
      "epoch": 0.3365006553079948,
      "grad_norm": 0.15337443351745605,
      "learning_rate": 0.0001,
      "loss": 1.4441,
      "step": 2054
    },
    {
      "epoch": 0.33666448230668417,
      "grad_norm": 0.1428510695695877,
      "learning_rate": 0.0001,
      "loss": 1.3351,
      "step": 2055
    },
    {
      "epoch": 0.33682830930537355,
      "grad_norm": 0.1447797268629074,
      "learning_rate": 0.0001,
      "loss": 1.404,
      "step": 2056
    },
    {
      "epoch": 0.3369921363040629,
      "grad_norm": 0.13884468376636505,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 2057
    },
    {
      "epoch": 0.33715596330275227,
      "grad_norm": 0.14656689763069153,
      "learning_rate": 0.0001,
      "loss": 1.4608,
      "step": 2058
    },
    {
      "epoch": 0.33731979030144166,
      "grad_norm": 0.1395944356918335,
      "learning_rate": 0.0001,
      "loss": 1.43,
      "step": 2059
    },
    {
      "epoch": 0.33748361730013104,
      "grad_norm": 0.14766384661197662,
      "learning_rate": 0.0001,
      "loss": 1.4229,
      "step": 2060
    },
    {
      "epoch": 0.33764744429882043,
      "grad_norm": 0.15360026061534882,
      "learning_rate": 0.0001,
      "loss": 1.4087,
      "step": 2061
    },
    {
      "epoch": 0.3378112712975098,
      "grad_norm": 0.1387275755405426,
      "learning_rate": 0.0001,
      "loss": 1.3685,
      "step": 2062
    },
    {
      "epoch": 0.3379750982961992,
      "grad_norm": 0.1443026065826416,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 2063
    },
    {
      "epoch": 0.3381389252948886,
      "grad_norm": 0.14369140565395355,
      "learning_rate": 0.0001,
      "loss": 1.4279,
      "step": 2064
    },
    {
      "epoch": 0.338302752293578,
      "grad_norm": 0.15171192586421967,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 2065
    },
    {
      "epoch": 0.33846657929226737,
      "grad_norm": 0.15206143260002136,
      "learning_rate": 0.0001,
      "loss": 1.4239,
      "step": 2066
    },
    {
      "epoch": 0.33863040629095675,
      "grad_norm": 0.14777450263500214,
      "learning_rate": 0.0001,
      "loss": 1.4514,
      "step": 2067
    },
    {
      "epoch": 0.33879423328964614,
      "grad_norm": 0.14483733475208282,
      "learning_rate": 0.0001,
      "loss": 1.384,
      "step": 2068
    },
    {
      "epoch": 0.3389580602883355,
      "grad_norm": 0.1416064351797104,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 2069
    },
    {
      "epoch": 0.3391218872870249,
      "grad_norm": 0.13789865374565125,
      "learning_rate": 0.0001,
      "loss": 1.3808,
      "step": 2070
    },
    {
      "epoch": 0.3392857142857143,
      "grad_norm": 0.1641058623790741,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 2071
    },
    {
      "epoch": 0.3394495412844037,
      "grad_norm": 0.1497175544500351,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 2072
    },
    {
      "epoch": 0.3396133682830931,
      "grad_norm": 0.1419355273246765,
      "learning_rate": 0.0001,
      "loss": 1.3855,
      "step": 2073
    },
    {
      "epoch": 0.33977719528178246,
      "grad_norm": 0.14612577855587006,
      "learning_rate": 0.0001,
      "loss": 1.3687,
      "step": 2074
    },
    {
      "epoch": 0.3399410222804718,
      "grad_norm": 0.14056460559368134,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 2075
    },
    {
      "epoch": 0.3401048492791612,
      "grad_norm": 0.14071188867092133,
      "learning_rate": 0.0001,
      "loss": 1.407,
      "step": 2076
    },
    {
      "epoch": 0.34026867627785057,
      "grad_norm": 0.1433897316455841,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 2077
    },
    {
      "epoch": 0.34043250327653996,
      "grad_norm": 0.13452444970607758,
      "learning_rate": 0.0001,
      "loss": 1.375,
      "step": 2078
    },
    {
      "epoch": 0.34059633027522934,
      "grad_norm": 0.13843946158885956,
      "learning_rate": 0.0001,
      "loss": 1.2775,
      "step": 2079
    },
    {
      "epoch": 0.34076015727391873,
      "grad_norm": 0.14785663783550262,
      "learning_rate": 0.0001,
      "loss": 1.4302,
      "step": 2080
    },
    {
      "epoch": 0.3409239842726081,
      "grad_norm": 0.14319287240505219,
      "learning_rate": 0.0001,
      "loss": 1.3806,
      "step": 2081
    },
    {
      "epoch": 0.3410878112712975,
      "grad_norm": 0.1385897397994995,
      "learning_rate": 0.0001,
      "loss": 1.4064,
      "step": 2082
    },
    {
      "epoch": 0.3412516382699869,
      "grad_norm": 0.14391832053661346,
      "learning_rate": 0.0001,
      "loss": 1.3207,
      "step": 2083
    },
    {
      "epoch": 0.3414154652686763,
      "grad_norm": 0.1569412499666214,
      "learning_rate": 0.0001,
      "loss": 1.4348,
      "step": 2084
    },
    {
      "epoch": 0.34157929226736566,
      "grad_norm": 0.1446705460548401,
      "learning_rate": 0.0001,
      "loss": 1.451,
      "step": 2085
    },
    {
      "epoch": 0.34174311926605505,
      "grad_norm": 0.1566343605518341,
      "learning_rate": 0.0001,
      "loss": 1.4079,
      "step": 2086
    },
    {
      "epoch": 0.34190694626474444,
      "grad_norm": 0.14230947196483612,
      "learning_rate": 0.0001,
      "loss": 1.442,
      "step": 2087
    },
    {
      "epoch": 0.3420707732634338,
      "grad_norm": 0.151717409491539,
      "learning_rate": 0.0001,
      "loss": 1.4214,
      "step": 2088
    },
    {
      "epoch": 0.3422346002621232,
      "grad_norm": 0.14131182432174683,
      "learning_rate": 0.0001,
      "loss": 1.3705,
      "step": 2089
    },
    {
      "epoch": 0.3423984272608126,
      "grad_norm": 0.14744548499584198,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 2090
    },
    {
      "epoch": 0.342562254259502,
      "grad_norm": 0.14364808797836304,
      "learning_rate": 0.0001,
      "loss": 1.398,
      "step": 2091
    },
    {
      "epoch": 0.3427260812581914,
      "grad_norm": 0.15598343312740326,
      "learning_rate": 0.0001,
      "loss": 1.447,
      "step": 2092
    },
    {
      "epoch": 0.34288990825688076,
      "grad_norm": 0.13761240243911743,
      "learning_rate": 0.0001,
      "loss": 1.2675,
      "step": 2093
    },
    {
      "epoch": 0.3430537352555701,
      "grad_norm": 0.14402559399604797,
      "learning_rate": 0.0001,
      "loss": 1.4161,
      "step": 2094
    },
    {
      "epoch": 0.3432175622542595,
      "grad_norm": 0.14696772396564484,
      "learning_rate": 0.0001,
      "loss": 1.5062,
      "step": 2095
    },
    {
      "epoch": 0.34338138925294887,
      "grad_norm": 0.14174583554267883,
      "learning_rate": 0.0001,
      "loss": 1.4569,
      "step": 2096
    },
    {
      "epoch": 0.34354521625163825,
      "grad_norm": 0.14388883113861084,
      "learning_rate": 0.0001,
      "loss": 1.3671,
      "step": 2097
    },
    {
      "epoch": 0.34370904325032764,
      "grad_norm": 0.1443578451871872,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 2098
    },
    {
      "epoch": 0.343872870249017,
      "grad_norm": 0.1589295119047165,
      "learning_rate": 0.0001,
      "loss": 1.4197,
      "step": 2099
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 0.1511935591697693,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 2100
    },
    {
      "epoch": 0.3442005242463958,
      "grad_norm": 0.14415985345840454,
      "learning_rate": 0.0001,
      "loss": 1.3179,
      "step": 2101
    },
    {
      "epoch": 0.3443643512450852,
      "grad_norm": 0.14061856269836426,
      "learning_rate": 0.0001,
      "loss": 1.4901,
      "step": 2102
    },
    {
      "epoch": 0.3445281782437746,
      "grad_norm": 0.14574812352657318,
      "learning_rate": 0.0001,
      "loss": 1.3902,
      "step": 2103
    },
    {
      "epoch": 0.34469200524246396,
      "grad_norm": 0.15109269320964813,
      "learning_rate": 0.0001,
      "loss": 1.371,
      "step": 2104
    },
    {
      "epoch": 0.34485583224115335,
      "grad_norm": 0.14203903079032898,
      "learning_rate": 0.0001,
      "loss": 1.4396,
      "step": 2105
    },
    {
      "epoch": 0.34501965923984274,
      "grad_norm": 0.14307141304016113,
      "learning_rate": 0.0001,
      "loss": 1.387,
      "step": 2106
    },
    {
      "epoch": 0.3451834862385321,
      "grad_norm": 0.1440681517124176,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 2107
    },
    {
      "epoch": 0.3453473132372215,
      "grad_norm": 0.14319375157356262,
      "learning_rate": 0.0001,
      "loss": 1.4282,
      "step": 2108
    },
    {
      "epoch": 0.3455111402359109,
      "grad_norm": 0.14295312762260437,
      "learning_rate": 0.0001,
      "loss": 1.4171,
      "step": 2109
    },
    {
      "epoch": 0.3456749672346003,
      "grad_norm": 0.16266761720180511,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 2110
    },
    {
      "epoch": 0.34583879423328967,
      "grad_norm": 0.15184083580970764,
      "learning_rate": 0.0001,
      "loss": 1.5263,
      "step": 2111
    },
    {
      "epoch": 0.34600262123197906,
      "grad_norm": 0.14476868510246277,
      "learning_rate": 0.0001,
      "loss": 1.393,
      "step": 2112
    },
    {
      "epoch": 0.3461664482306684,
      "grad_norm": 0.15266399085521698,
      "learning_rate": 0.0001,
      "loss": 1.4112,
      "step": 2113
    },
    {
      "epoch": 0.3463302752293578,
      "grad_norm": 0.13821552693843842,
      "learning_rate": 0.0001,
      "loss": 1.3875,
      "step": 2114
    },
    {
      "epoch": 0.34649410222804716,
      "grad_norm": 0.14559254050254822,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 2115
    },
    {
      "epoch": 0.34665792922673655,
      "grad_norm": 0.1382431834936142,
      "learning_rate": 0.0001,
      "loss": 1.4026,
      "step": 2116
    },
    {
      "epoch": 0.34682175622542594,
      "grad_norm": 0.1418442279100418,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 2117
    },
    {
      "epoch": 0.3469855832241153,
      "grad_norm": 0.14912022650241852,
      "learning_rate": 0.0001,
      "loss": 1.4195,
      "step": 2118
    },
    {
      "epoch": 0.3471494102228047,
      "grad_norm": 0.14608030021190643,
      "learning_rate": 0.0001,
      "loss": 1.4358,
      "step": 2119
    },
    {
      "epoch": 0.3473132372214941,
      "grad_norm": 0.1371443122625351,
      "learning_rate": 0.0001,
      "loss": 1.3297,
      "step": 2120
    },
    {
      "epoch": 0.3474770642201835,
      "grad_norm": 0.13694873452186584,
      "learning_rate": 0.0001,
      "loss": 1.3533,
      "step": 2121
    },
    {
      "epoch": 0.34764089121887287,
      "grad_norm": 0.14296211302280426,
      "learning_rate": 0.0001,
      "loss": 1.4928,
      "step": 2122
    },
    {
      "epoch": 0.34780471821756226,
      "grad_norm": 0.1434214860200882,
      "learning_rate": 0.0001,
      "loss": 1.4414,
      "step": 2123
    },
    {
      "epoch": 0.34796854521625165,
      "grad_norm": 0.1428804248571396,
      "learning_rate": 0.0001,
      "loss": 1.3825,
      "step": 2124
    },
    {
      "epoch": 0.34813237221494103,
      "grad_norm": 0.14406083524227142,
      "learning_rate": 0.0001,
      "loss": 1.4566,
      "step": 2125
    },
    {
      "epoch": 0.3482961992136304,
      "grad_norm": 0.139811709523201,
      "learning_rate": 0.0001,
      "loss": 1.4135,
      "step": 2126
    },
    {
      "epoch": 0.3484600262123198,
      "grad_norm": 0.14062584936618805,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 2127
    },
    {
      "epoch": 0.3486238532110092,
      "grad_norm": 0.1465856432914734,
      "learning_rate": 0.0001,
      "loss": 1.4256,
      "step": 2128
    },
    {
      "epoch": 0.3487876802096986,
      "grad_norm": 0.13778959214687347,
      "learning_rate": 0.0001,
      "loss": 1.5509,
      "step": 2129
    },
    {
      "epoch": 0.34895150720838797,
      "grad_norm": 0.13611343502998352,
      "learning_rate": 0.0001,
      "loss": 1.4011,
      "step": 2130
    },
    {
      "epoch": 0.3491153342070773,
      "grad_norm": 0.14628565311431885,
      "learning_rate": 0.0001,
      "loss": 1.4594,
      "step": 2131
    },
    {
      "epoch": 0.3492791612057667,
      "grad_norm": 0.14621125161647797,
      "learning_rate": 0.0001,
      "loss": 1.4561,
      "step": 2132
    },
    {
      "epoch": 0.3494429882044561,
      "grad_norm": 0.14127273857593536,
      "learning_rate": 0.0001,
      "loss": 1.4559,
      "step": 2133
    },
    {
      "epoch": 0.34960681520314546,
      "grad_norm": 0.15840531885623932,
      "learning_rate": 0.0001,
      "loss": 1.4108,
      "step": 2134
    },
    {
      "epoch": 0.34977064220183485,
      "grad_norm": 0.15365763008594513,
      "learning_rate": 0.0001,
      "loss": 1.3757,
      "step": 2135
    },
    {
      "epoch": 0.34993446920052423,
      "grad_norm": 0.1464042216539383,
      "learning_rate": 0.0001,
      "loss": 1.3675,
      "step": 2136
    },
    {
      "epoch": 0.3500982961992136,
      "grad_norm": 0.14569227397441864,
      "learning_rate": 0.0001,
      "loss": 1.296,
      "step": 2137
    },
    {
      "epoch": 0.350262123197903,
      "grad_norm": 0.14873871207237244,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 2138
    },
    {
      "epoch": 0.3504259501965924,
      "grad_norm": 0.15090104937553406,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 2139
    },
    {
      "epoch": 0.3505897771952818,
      "grad_norm": 0.14021719992160797,
      "learning_rate": 0.0001,
      "loss": 1.3828,
      "step": 2140
    },
    {
      "epoch": 0.35075360419397117,
      "grad_norm": 0.1510886251926422,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 2141
    },
    {
      "epoch": 0.35091743119266056,
      "grad_norm": 0.1524900197982788,
      "learning_rate": 0.0001,
      "loss": 1.4553,
      "step": 2142
    },
    {
      "epoch": 0.35108125819134994,
      "grad_norm": 0.1466847062110901,
      "learning_rate": 0.0001,
      "loss": 1.4767,
      "step": 2143
    },
    {
      "epoch": 0.35124508519003933,
      "grad_norm": 0.1502920538187027,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 2144
    },
    {
      "epoch": 0.3514089121887287,
      "grad_norm": 0.14443401992321014,
      "learning_rate": 0.0001,
      "loss": 1.4991,
      "step": 2145
    },
    {
      "epoch": 0.3515727391874181,
      "grad_norm": 0.1436457335948944,
      "learning_rate": 0.0001,
      "loss": 1.5054,
      "step": 2146
    },
    {
      "epoch": 0.3517365661861075,
      "grad_norm": 0.14530134201049805,
      "learning_rate": 0.0001,
      "loss": 1.3054,
      "step": 2147
    },
    {
      "epoch": 0.3519003931847969,
      "grad_norm": 0.1558932662010193,
      "learning_rate": 0.0001,
      "loss": 1.3719,
      "step": 2148
    },
    {
      "epoch": 0.35206422018348627,
      "grad_norm": 0.13937313854694366,
      "learning_rate": 0.0001,
      "loss": 1.3025,
      "step": 2149
    },
    {
      "epoch": 0.3522280471821756,
      "grad_norm": 0.1465480923652649,
      "learning_rate": 0.0001,
      "loss": 1.376,
      "step": 2150
    },
    {
      "epoch": 0.352391874180865,
      "grad_norm": 0.14736413955688477,
      "learning_rate": 0.0001,
      "loss": 1.4243,
      "step": 2151
    },
    {
      "epoch": 0.35255570117955437,
      "grad_norm": 0.13905519247055054,
      "learning_rate": 0.0001,
      "loss": 1.3259,
      "step": 2152
    },
    {
      "epoch": 0.35271952817824376,
      "grad_norm": 0.1456368863582611,
      "learning_rate": 0.0001,
      "loss": 1.379,
      "step": 2153
    },
    {
      "epoch": 0.35288335517693314,
      "grad_norm": 0.15476089715957642,
      "learning_rate": 0.0001,
      "loss": 1.3877,
      "step": 2154
    },
    {
      "epoch": 0.35304718217562253,
      "grad_norm": 0.1425773650407791,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 2155
    },
    {
      "epoch": 0.3532110091743119,
      "grad_norm": 0.1540389508008957,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 2156
    },
    {
      "epoch": 0.3533748361730013,
      "grad_norm": 0.15003181993961334,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 2157
    },
    {
      "epoch": 0.3535386631716907,
      "grad_norm": 0.1425335854291916,
      "learning_rate": 0.0001,
      "loss": 1.3852,
      "step": 2158
    },
    {
      "epoch": 0.3537024901703801,
      "grad_norm": 0.14714473485946655,
      "learning_rate": 0.0001,
      "loss": 1.421,
      "step": 2159
    },
    {
      "epoch": 0.35386631716906947,
      "grad_norm": 0.14304111897945404,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 2160
    },
    {
      "epoch": 0.35403014416775885,
      "grad_norm": 0.14743739366531372,
      "learning_rate": 0.0001,
      "loss": 1.3291,
      "step": 2161
    },
    {
      "epoch": 0.35419397116644824,
      "grad_norm": 0.14665457606315613,
      "learning_rate": 0.0001,
      "loss": 1.4244,
      "step": 2162
    },
    {
      "epoch": 0.3543577981651376,
      "grad_norm": 0.1486324965953827,
      "learning_rate": 0.0001,
      "loss": 1.4425,
      "step": 2163
    },
    {
      "epoch": 0.354521625163827,
      "grad_norm": 0.15409113466739655,
      "learning_rate": 0.0001,
      "loss": 1.3378,
      "step": 2164
    },
    {
      "epoch": 0.3546854521625164,
      "grad_norm": 0.14587226510047913,
      "learning_rate": 0.0001,
      "loss": 1.3673,
      "step": 2165
    },
    {
      "epoch": 0.3548492791612058,
      "grad_norm": 0.14995211362838745,
      "learning_rate": 0.0001,
      "loss": 1.44,
      "step": 2166
    },
    {
      "epoch": 0.3550131061598952,
      "grad_norm": 0.14917004108428955,
      "learning_rate": 0.0001,
      "loss": 1.4419,
      "step": 2167
    },
    {
      "epoch": 0.35517693315858456,
      "grad_norm": 0.14261211454868317,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 2168
    },
    {
      "epoch": 0.3553407601572739,
      "grad_norm": 0.14451386034488678,
      "learning_rate": 0.0001,
      "loss": 1.3907,
      "step": 2169
    },
    {
      "epoch": 0.3555045871559633,
      "grad_norm": 0.14966602623462677,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 2170
    },
    {
      "epoch": 0.35566841415465267,
      "grad_norm": 0.14322564005851746,
      "learning_rate": 0.0001,
      "loss": 1.3601,
      "step": 2171
    },
    {
      "epoch": 0.35583224115334205,
      "grad_norm": 0.15526942908763885,
      "learning_rate": 0.0001,
      "loss": 1.4307,
      "step": 2172
    },
    {
      "epoch": 0.35599606815203144,
      "grad_norm": 0.15131491422653198,
      "learning_rate": 0.0001,
      "loss": 1.4258,
      "step": 2173
    },
    {
      "epoch": 0.35615989515072083,
      "grad_norm": 0.15047062933444977,
      "learning_rate": 0.0001,
      "loss": 1.4093,
      "step": 2174
    },
    {
      "epoch": 0.3563237221494102,
      "grad_norm": 0.14163531363010406,
      "learning_rate": 0.0001,
      "loss": 1.3445,
      "step": 2175
    },
    {
      "epoch": 0.3564875491480996,
      "grad_norm": 0.15000802278518677,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 2176
    },
    {
      "epoch": 0.356651376146789,
      "grad_norm": 0.14793910086154938,
      "learning_rate": 0.0001,
      "loss": 1.4157,
      "step": 2177
    },
    {
      "epoch": 0.3568152031454784,
      "grad_norm": 0.14880335330963135,
      "learning_rate": 0.0001,
      "loss": 1.4634,
      "step": 2178
    },
    {
      "epoch": 0.35697903014416776,
      "grad_norm": 0.14408732950687408,
      "learning_rate": 0.0001,
      "loss": 1.4524,
      "step": 2179
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.14633090794086456,
      "learning_rate": 0.0001,
      "loss": 1.5005,
      "step": 2180
    },
    {
      "epoch": 0.35730668414154654,
      "grad_norm": 0.15080860257148743,
      "learning_rate": 0.0001,
      "loss": 1.3565,
      "step": 2181
    },
    {
      "epoch": 0.3574705111402359,
      "grad_norm": 0.14752033352851868,
      "learning_rate": 0.0001,
      "loss": 1.4163,
      "step": 2182
    },
    {
      "epoch": 0.3576343381389253,
      "grad_norm": 0.15626530349254608,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 2183
    },
    {
      "epoch": 0.3577981651376147,
      "grad_norm": 0.13993461430072784,
      "learning_rate": 0.0001,
      "loss": 1.3485,
      "step": 2184
    },
    {
      "epoch": 0.3579619921363041,
      "grad_norm": 0.14450356364250183,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 2185
    },
    {
      "epoch": 0.3581258191349935,
      "grad_norm": 0.14498627185821533,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 2186
    },
    {
      "epoch": 0.3582896461336828,
      "grad_norm": 0.14986784756183624,
      "learning_rate": 0.0001,
      "loss": 1.4746,
      "step": 2187
    },
    {
      "epoch": 0.3584534731323722,
      "grad_norm": 0.15267504751682281,
      "learning_rate": 0.0001,
      "loss": 1.362,
      "step": 2188
    },
    {
      "epoch": 0.3586173001310616,
      "grad_norm": 0.14109204709529877,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 2189
    },
    {
      "epoch": 0.35878112712975097,
      "grad_norm": 0.14485153555870056,
      "learning_rate": 0.0001,
      "loss": 1.371,
      "step": 2190
    },
    {
      "epoch": 0.35894495412844035,
      "grad_norm": 0.15188652276992798,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 2191
    },
    {
      "epoch": 0.35910878112712974,
      "grad_norm": 0.14243943989276886,
      "learning_rate": 0.0001,
      "loss": 1.3846,
      "step": 2192
    },
    {
      "epoch": 0.3592726081258191,
      "grad_norm": 0.1480424851179123,
      "learning_rate": 0.0001,
      "loss": 1.5444,
      "step": 2193
    },
    {
      "epoch": 0.3594364351245085,
      "grad_norm": 0.14340557157993317,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 2194
    },
    {
      "epoch": 0.3596002621231979,
      "grad_norm": 0.14653831720352173,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 2195
    },
    {
      "epoch": 0.3597640891218873,
      "grad_norm": 0.15590567886829376,
      "learning_rate": 0.0001,
      "loss": 1.502,
      "step": 2196
    },
    {
      "epoch": 0.3599279161205767,
      "grad_norm": 0.14722971618175507,
      "learning_rate": 0.0001,
      "loss": 1.4687,
      "step": 2197
    },
    {
      "epoch": 0.36009174311926606,
      "grad_norm": 0.15116336941719055,
      "learning_rate": 0.0001,
      "loss": 1.3911,
      "step": 2198
    },
    {
      "epoch": 0.36025557011795545,
      "grad_norm": 0.1449773758649826,
      "learning_rate": 0.0001,
      "loss": 1.3267,
      "step": 2199
    },
    {
      "epoch": 0.36041939711664484,
      "grad_norm": 0.14571118354797363,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 2200
    },
    {
      "epoch": 0.3605832241153342,
      "grad_norm": 0.13657960295677185,
      "learning_rate": 0.0001,
      "loss": 1.4588,
      "step": 2201
    },
    {
      "epoch": 0.3607470511140236,
      "grad_norm": 0.13872499763965607,
      "learning_rate": 0.0001,
      "loss": 1.3151,
      "step": 2202
    },
    {
      "epoch": 0.360910878112713,
      "grad_norm": 0.1488146334886551,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 2203
    },
    {
      "epoch": 0.3610747051114024,
      "grad_norm": 0.14146222174167633,
      "learning_rate": 0.0001,
      "loss": 1.4432,
      "step": 2204
    },
    {
      "epoch": 0.36123853211009177,
      "grad_norm": 0.14561352133750916,
      "learning_rate": 0.0001,
      "loss": 1.4587,
      "step": 2205
    },
    {
      "epoch": 0.3614023591087811,
      "grad_norm": 0.14963263273239136,
      "learning_rate": 0.0001,
      "loss": 1.5036,
      "step": 2206
    },
    {
      "epoch": 0.3615661861074705,
      "grad_norm": 0.13940243422985077,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 2207
    },
    {
      "epoch": 0.3617300131061599,
      "grad_norm": 0.14050929248332977,
      "learning_rate": 0.0001,
      "loss": 1.4149,
      "step": 2208
    },
    {
      "epoch": 0.36189384010484926,
      "grad_norm": 0.14976058900356293,
      "learning_rate": 0.0001,
      "loss": 1.3882,
      "step": 2209
    },
    {
      "epoch": 0.36205766710353865,
      "grad_norm": 0.15308433771133423,
      "learning_rate": 0.0001,
      "loss": 1.3855,
      "step": 2210
    },
    {
      "epoch": 0.36222149410222804,
      "grad_norm": 0.14142969250679016,
      "learning_rate": 0.0001,
      "loss": 1.3786,
      "step": 2211
    },
    {
      "epoch": 0.3623853211009174,
      "grad_norm": 0.15605521202087402,
      "learning_rate": 0.0001,
      "loss": 1.3527,
      "step": 2212
    },
    {
      "epoch": 0.3625491480996068,
      "grad_norm": 0.14440365135669708,
      "learning_rate": 0.0001,
      "loss": 1.4373,
      "step": 2213
    },
    {
      "epoch": 0.3627129750982962,
      "grad_norm": 0.14356395602226257,
      "learning_rate": 0.0001,
      "loss": 1.5012,
      "step": 2214
    },
    {
      "epoch": 0.3628768020969856,
      "grad_norm": 0.14800292253494263,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 2215
    },
    {
      "epoch": 0.36304062909567497,
      "grad_norm": 0.14084097743034363,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 2216
    },
    {
      "epoch": 0.36320445609436436,
      "grad_norm": 0.14456844329833984,
      "learning_rate": 0.0001,
      "loss": 1.4093,
      "step": 2217
    },
    {
      "epoch": 0.36336828309305375,
      "grad_norm": 0.1409318745136261,
      "learning_rate": 0.0001,
      "loss": 1.4565,
      "step": 2218
    },
    {
      "epoch": 0.36353211009174313,
      "grad_norm": 0.14670254290103912,
      "learning_rate": 0.0001,
      "loss": 1.4648,
      "step": 2219
    },
    {
      "epoch": 0.3636959370904325,
      "grad_norm": 0.14027443528175354,
      "learning_rate": 0.0001,
      "loss": 1.3419,
      "step": 2220
    },
    {
      "epoch": 0.3638597640891219,
      "grad_norm": 0.1406625658273697,
      "learning_rate": 0.0001,
      "loss": 1.3574,
      "step": 2221
    },
    {
      "epoch": 0.3640235910878113,
      "grad_norm": 0.14566479623317719,
      "learning_rate": 0.0001,
      "loss": 1.4042,
      "step": 2222
    },
    {
      "epoch": 0.3641874180865007,
      "grad_norm": 0.14777342975139618,
      "learning_rate": 0.0001,
      "loss": 1.4367,
      "step": 2223
    },
    {
      "epoch": 0.36435124508519,
      "grad_norm": 0.1547519713640213,
      "learning_rate": 0.0001,
      "loss": 1.2338,
      "step": 2224
    },
    {
      "epoch": 0.3645150720838794,
      "grad_norm": 0.15779384970664978,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 2225
    },
    {
      "epoch": 0.3646788990825688,
      "grad_norm": 0.1543739289045334,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 2226
    },
    {
      "epoch": 0.3648427260812582,
      "grad_norm": 0.14658623933792114,
      "learning_rate": 0.0001,
      "loss": 1.3853,
      "step": 2227
    },
    {
      "epoch": 0.36500655307994756,
      "grad_norm": 0.14946866035461426,
      "learning_rate": 0.0001,
      "loss": 1.4438,
      "step": 2228
    },
    {
      "epoch": 0.36517038007863695,
      "grad_norm": 0.15614977478981018,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 2229
    },
    {
      "epoch": 0.36533420707732633,
      "grad_norm": 0.14375586807727814,
      "learning_rate": 0.0001,
      "loss": 1.4466,
      "step": 2230
    },
    {
      "epoch": 0.3654980340760157,
      "grad_norm": 0.14945584535598755,
      "learning_rate": 0.0001,
      "loss": 1.434,
      "step": 2231
    },
    {
      "epoch": 0.3656618610747051,
      "grad_norm": 0.14992086589336395,
      "learning_rate": 0.0001,
      "loss": 1.2663,
      "step": 2232
    },
    {
      "epoch": 0.3658256880733945,
      "grad_norm": 0.14407996833324432,
      "learning_rate": 0.0001,
      "loss": 1.3421,
      "step": 2233
    },
    {
      "epoch": 0.3659895150720839,
      "grad_norm": 0.13913407921791077,
      "learning_rate": 0.0001,
      "loss": 1.4845,
      "step": 2234
    },
    {
      "epoch": 0.36615334207077327,
      "grad_norm": 0.15160208940505981,
      "learning_rate": 0.0001,
      "loss": 1.4115,
      "step": 2235
    },
    {
      "epoch": 0.36631716906946266,
      "grad_norm": 0.14019736647605896,
      "learning_rate": 0.0001,
      "loss": 1.2844,
      "step": 2236
    },
    {
      "epoch": 0.36648099606815204,
      "grad_norm": 0.1508610099554062,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 2237
    },
    {
      "epoch": 0.36664482306684143,
      "grad_norm": 0.14075900614261627,
      "learning_rate": 0.0001,
      "loss": 1.3991,
      "step": 2238
    },
    {
      "epoch": 0.3668086500655308,
      "grad_norm": 0.13968797028064728,
      "learning_rate": 0.0001,
      "loss": 1.3649,
      "step": 2239
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 0.14620724320411682,
      "learning_rate": 0.0001,
      "loss": 1.3475,
      "step": 2240
    },
    {
      "epoch": 0.3671363040629096,
      "grad_norm": 0.1488366723060608,
      "learning_rate": 0.0001,
      "loss": 1.451,
      "step": 2241
    },
    {
      "epoch": 0.367300131061599,
      "grad_norm": 0.14609704911708832,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 2242
    },
    {
      "epoch": 0.3674639580602883,
      "grad_norm": 0.1498280167579651,
      "learning_rate": 0.0001,
      "loss": 1.4359,
      "step": 2243
    },
    {
      "epoch": 0.3676277850589777,
      "grad_norm": 0.1425849348306656,
      "learning_rate": 0.0001,
      "loss": 1.3788,
      "step": 2244
    },
    {
      "epoch": 0.3677916120576671,
      "grad_norm": 0.1405731737613678,
      "learning_rate": 0.0001,
      "loss": 1.4008,
      "step": 2245
    },
    {
      "epoch": 0.36795543905635647,
      "grad_norm": 0.15471498668193817,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 2246
    },
    {
      "epoch": 0.36811926605504586,
      "grad_norm": 0.13937480747699738,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 2247
    },
    {
      "epoch": 0.36828309305373524,
      "grad_norm": 0.1387612223625183,
      "learning_rate": 0.0001,
      "loss": 1.3865,
      "step": 2248
    },
    {
      "epoch": 0.36844692005242463,
      "grad_norm": 0.1614387035369873,
      "learning_rate": 0.0001,
      "loss": 1.3758,
      "step": 2249
    },
    {
      "epoch": 0.368610747051114,
      "grad_norm": 0.167240172624588,
      "learning_rate": 0.0001,
      "loss": 1.4497,
      "step": 2250
    },
    {
      "epoch": 0.3687745740498034,
      "grad_norm": 0.13892097771167755,
      "learning_rate": 0.0001,
      "loss": 1.3712,
      "step": 2251
    },
    {
      "epoch": 0.3689384010484928,
      "grad_norm": 0.14175273478031158,
      "learning_rate": 0.0001,
      "loss": 1.4234,
      "step": 2252
    },
    {
      "epoch": 0.3691022280471822,
      "grad_norm": 0.1475791484117508,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 2253
    },
    {
      "epoch": 0.36926605504587157,
      "grad_norm": 0.14201723039150238,
      "learning_rate": 0.0001,
      "loss": 1.4082,
      "step": 2254
    },
    {
      "epoch": 0.36942988204456095,
      "grad_norm": 0.13479918241500854,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 2255
    },
    {
      "epoch": 0.36959370904325034,
      "grad_norm": 0.14680060744285583,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 2256
    },
    {
      "epoch": 0.3697575360419397,
      "grad_norm": 0.14775186777114868,
      "learning_rate": 0.0001,
      "loss": 1.3687,
      "step": 2257
    },
    {
      "epoch": 0.3699213630406291,
      "grad_norm": 0.15924414992332458,
      "learning_rate": 0.0001,
      "loss": 1.4133,
      "step": 2258
    },
    {
      "epoch": 0.3700851900393185,
      "grad_norm": 0.14137621223926544,
      "learning_rate": 0.0001,
      "loss": 1.3811,
      "step": 2259
    },
    {
      "epoch": 0.3702490170380079,
      "grad_norm": 0.1476057916879654,
      "learning_rate": 0.0001,
      "loss": 1.3839,
      "step": 2260
    },
    {
      "epoch": 0.3704128440366973,
      "grad_norm": 0.15009431540966034,
      "learning_rate": 0.0001,
      "loss": 1.4365,
      "step": 2261
    },
    {
      "epoch": 0.3705766710353866,
      "grad_norm": 0.15836067497730255,
      "learning_rate": 0.0001,
      "loss": 1.3796,
      "step": 2262
    },
    {
      "epoch": 0.370740498034076,
      "grad_norm": 0.1457124650478363,
      "learning_rate": 0.0001,
      "loss": 1.2619,
      "step": 2263
    },
    {
      "epoch": 0.3709043250327654,
      "grad_norm": 0.145296111702919,
      "learning_rate": 0.0001,
      "loss": 1.3643,
      "step": 2264
    },
    {
      "epoch": 0.37106815203145477,
      "grad_norm": 0.15316180884838104,
      "learning_rate": 0.0001,
      "loss": 1.5221,
      "step": 2265
    },
    {
      "epoch": 0.37123197903014415,
      "grad_norm": 0.1430603563785553,
      "learning_rate": 0.0001,
      "loss": 1.3545,
      "step": 2266
    },
    {
      "epoch": 0.37139580602883354,
      "grad_norm": 0.14833450317382812,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 2267
    },
    {
      "epoch": 0.37155963302752293,
      "grad_norm": 0.14156807959079742,
      "learning_rate": 0.0001,
      "loss": 1.4029,
      "step": 2268
    },
    {
      "epoch": 0.3717234600262123,
      "grad_norm": 0.1419268399477005,
      "learning_rate": 0.0001,
      "loss": 1.4189,
      "step": 2269
    },
    {
      "epoch": 0.3718872870249017,
      "grad_norm": 0.150970920920372,
      "learning_rate": 0.0001,
      "loss": 1.4739,
      "step": 2270
    },
    {
      "epoch": 0.3720511140235911,
      "grad_norm": 0.1400698870420456,
      "learning_rate": 0.0001,
      "loss": 1.4851,
      "step": 2271
    },
    {
      "epoch": 0.3722149410222805,
      "grad_norm": 0.13560213148593903,
      "learning_rate": 0.0001,
      "loss": 1.3918,
      "step": 2272
    },
    {
      "epoch": 0.37237876802096986,
      "grad_norm": 0.15459400415420532,
      "learning_rate": 0.0001,
      "loss": 1.4938,
      "step": 2273
    },
    {
      "epoch": 0.37254259501965925,
      "grad_norm": 0.145756796002388,
      "learning_rate": 0.0001,
      "loss": 1.3807,
      "step": 2274
    },
    {
      "epoch": 0.37270642201834864,
      "grad_norm": 0.13817551732063293,
      "learning_rate": 0.0001,
      "loss": 1.4111,
      "step": 2275
    },
    {
      "epoch": 0.372870249017038,
      "grad_norm": 0.13877052068710327,
      "learning_rate": 0.0001,
      "loss": 1.367,
      "step": 2276
    },
    {
      "epoch": 0.3730340760157274,
      "grad_norm": 0.13835711777210236,
      "learning_rate": 0.0001,
      "loss": 1.3814,
      "step": 2277
    },
    {
      "epoch": 0.3731979030144168,
      "grad_norm": 0.1502508521080017,
      "learning_rate": 0.0001,
      "loss": 1.428,
      "step": 2278
    },
    {
      "epoch": 0.3733617300131062,
      "grad_norm": 0.1391015350818634,
      "learning_rate": 0.0001,
      "loss": 1.3775,
      "step": 2279
    },
    {
      "epoch": 0.3735255570117955,
      "grad_norm": 0.1494036316871643,
      "learning_rate": 0.0001,
      "loss": 1.4529,
      "step": 2280
    },
    {
      "epoch": 0.3736893840104849,
      "grad_norm": 0.1392051726579666,
      "learning_rate": 0.0001,
      "loss": 1.2963,
      "step": 2281
    },
    {
      "epoch": 0.3738532110091743,
      "grad_norm": 0.14664766192436218,
      "learning_rate": 0.0001,
      "loss": 1.3779,
      "step": 2282
    },
    {
      "epoch": 0.3740170380078637,
      "grad_norm": 0.15027619898319244,
      "learning_rate": 0.0001,
      "loss": 1.4598,
      "step": 2283
    },
    {
      "epoch": 0.37418086500655307,
      "grad_norm": 0.15694065392017365,
      "learning_rate": 0.0001,
      "loss": 1.4691,
      "step": 2284
    },
    {
      "epoch": 0.37434469200524245,
      "grad_norm": 0.14994128048419952,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 2285
    },
    {
      "epoch": 0.37450851900393184,
      "grad_norm": 0.14165563881397247,
      "learning_rate": 0.0001,
      "loss": 1.4087,
      "step": 2286
    },
    {
      "epoch": 0.3746723460026212,
      "grad_norm": 0.15309903025627136,
      "learning_rate": 0.0001,
      "loss": 1.3759,
      "step": 2287
    },
    {
      "epoch": 0.3748361730013106,
      "grad_norm": 0.1563466638326645,
      "learning_rate": 0.0001,
      "loss": 1.4625,
      "step": 2288
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.14989209175109863,
      "learning_rate": 0.0001,
      "loss": 1.3414,
      "step": 2289
    },
    {
      "epoch": 0.3751638269986894,
      "grad_norm": 0.15851755440235138,
      "learning_rate": 0.0001,
      "loss": 1.4349,
      "step": 2290
    },
    {
      "epoch": 0.3753276539973788,
      "grad_norm": 0.14244037866592407,
      "learning_rate": 0.0001,
      "loss": 1.2484,
      "step": 2291
    },
    {
      "epoch": 0.37549148099606816,
      "grad_norm": 0.1517471820116043,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 2292
    },
    {
      "epoch": 0.37565530799475755,
      "grad_norm": 0.15731287002563477,
      "learning_rate": 0.0001,
      "loss": 1.4508,
      "step": 2293
    },
    {
      "epoch": 0.37581913499344693,
      "grad_norm": 0.145687997341156,
      "learning_rate": 0.0001,
      "loss": 1.3861,
      "step": 2294
    },
    {
      "epoch": 0.3759829619921363,
      "grad_norm": 0.14887650310993195,
      "learning_rate": 0.0001,
      "loss": 1.3935,
      "step": 2295
    },
    {
      "epoch": 0.3761467889908257,
      "grad_norm": 0.15117675065994263,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 2296
    },
    {
      "epoch": 0.3763106159895151,
      "grad_norm": 0.1471964567899704,
      "learning_rate": 0.0001,
      "loss": 1.3905,
      "step": 2297
    },
    {
      "epoch": 0.3764744429882045,
      "grad_norm": 0.145172581076622,
      "learning_rate": 0.0001,
      "loss": 1.4672,
      "step": 2298
    },
    {
      "epoch": 0.3766382699868938,
      "grad_norm": 0.15228424966335297,
      "learning_rate": 0.0001,
      "loss": 1.338,
      "step": 2299
    },
    {
      "epoch": 0.3768020969855832,
      "grad_norm": 0.15388502180576324,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 2300
    },
    {
      "epoch": 0.3769659239842726,
      "grad_norm": 0.14778517186641693,
      "learning_rate": 0.0001,
      "loss": 1.3325,
      "step": 2301
    },
    {
      "epoch": 0.377129750982962,
      "grad_norm": 0.14890705049037933,
      "learning_rate": 0.0001,
      "loss": 1.3716,
      "step": 2302
    },
    {
      "epoch": 0.37729357798165136,
      "grad_norm": 0.15932151675224304,
      "learning_rate": 0.0001,
      "loss": 1.6192,
      "step": 2303
    },
    {
      "epoch": 0.37745740498034075,
      "grad_norm": 0.1481349766254425,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 2304
    },
    {
      "epoch": 0.37762123197903014,
      "grad_norm": 0.16594953835010529,
      "learning_rate": 0.0001,
      "loss": 1.5524,
      "step": 2305
    },
    {
      "epoch": 0.3777850589777195,
      "grad_norm": 0.14525631070137024,
      "learning_rate": 0.0001,
      "loss": 1.4958,
      "step": 2306
    },
    {
      "epoch": 0.3779488859764089,
      "grad_norm": 0.15038231015205383,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 2307
    },
    {
      "epoch": 0.3781127129750983,
      "grad_norm": 0.14878684282302856,
      "learning_rate": 0.0001,
      "loss": 1.3857,
      "step": 2308
    },
    {
      "epoch": 0.3782765399737877,
      "grad_norm": 0.14681436121463776,
      "learning_rate": 0.0001,
      "loss": 1.3185,
      "step": 2309
    },
    {
      "epoch": 0.37844036697247707,
      "grad_norm": 0.14739233255386353,
      "learning_rate": 0.0001,
      "loss": 1.3809,
      "step": 2310
    },
    {
      "epoch": 0.37860419397116646,
      "grad_norm": 0.15819987654685974,
      "learning_rate": 0.0001,
      "loss": 1.4345,
      "step": 2311
    },
    {
      "epoch": 0.37876802096985585,
      "grad_norm": 0.1491282731294632,
      "learning_rate": 0.0001,
      "loss": 1.3678,
      "step": 2312
    },
    {
      "epoch": 0.37893184796854523,
      "grad_norm": 0.1535070687532425,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 2313
    },
    {
      "epoch": 0.3790956749672346,
      "grad_norm": 0.1399320513010025,
      "learning_rate": 0.0001,
      "loss": 1.3058,
      "step": 2314
    },
    {
      "epoch": 0.379259501965924,
      "grad_norm": 0.15362682938575745,
      "learning_rate": 0.0001,
      "loss": 1.4034,
      "step": 2315
    },
    {
      "epoch": 0.3794233289646134,
      "grad_norm": 0.15242721140384674,
      "learning_rate": 0.0001,
      "loss": 1.4217,
      "step": 2316
    },
    {
      "epoch": 0.3795871559633027,
      "grad_norm": 0.14554694294929504,
      "learning_rate": 0.0001,
      "loss": 1.3818,
      "step": 2317
    },
    {
      "epoch": 0.3797509829619921,
      "grad_norm": 0.1517591029405594,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 2318
    },
    {
      "epoch": 0.3799148099606815,
      "grad_norm": 0.146103173494339,
      "learning_rate": 0.0001,
      "loss": 1.3095,
      "step": 2319
    },
    {
      "epoch": 0.3800786369593709,
      "grad_norm": 0.1482011079788208,
      "learning_rate": 0.0001,
      "loss": 1.4974,
      "step": 2320
    },
    {
      "epoch": 0.3802424639580603,
      "grad_norm": 0.14017827808856964,
      "learning_rate": 0.0001,
      "loss": 1.3396,
      "step": 2321
    },
    {
      "epoch": 0.38040629095674966,
      "grad_norm": 0.15246710181236267,
      "learning_rate": 0.0001,
      "loss": 1.4004,
      "step": 2322
    },
    {
      "epoch": 0.38057011795543905,
      "grad_norm": 0.14677226543426514,
      "learning_rate": 0.0001,
      "loss": 1.4004,
      "step": 2323
    },
    {
      "epoch": 0.38073394495412843,
      "grad_norm": 0.15274712443351746,
      "learning_rate": 0.0001,
      "loss": 1.4426,
      "step": 2324
    },
    {
      "epoch": 0.3808977719528178,
      "grad_norm": 0.16095827519893646,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 2325
    },
    {
      "epoch": 0.3810615989515072,
      "grad_norm": 0.14807507395744324,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 2326
    },
    {
      "epoch": 0.3812254259501966,
      "grad_norm": 0.14514195919036865,
      "learning_rate": 0.0001,
      "loss": 1.4814,
      "step": 2327
    },
    {
      "epoch": 0.381389252948886,
      "grad_norm": 0.15995891392230988,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 2328
    },
    {
      "epoch": 0.38155307994757537,
      "grad_norm": 0.14471562206745148,
      "learning_rate": 0.0001,
      "loss": 1.4111,
      "step": 2329
    },
    {
      "epoch": 0.38171690694626476,
      "grad_norm": 0.14767248928546906,
      "learning_rate": 0.0001,
      "loss": 1.2871,
      "step": 2330
    },
    {
      "epoch": 0.38188073394495414,
      "grad_norm": 0.1574028730392456,
      "learning_rate": 0.0001,
      "loss": 1.3722,
      "step": 2331
    },
    {
      "epoch": 0.38204456094364353,
      "grad_norm": 0.14769522845745087,
      "learning_rate": 0.0001,
      "loss": 1.3551,
      "step": 2332
    },
    {
      "epoch": 0.3822083879423329,
      "grad_norm": 0.15305721759796143,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 2333
    },
    {
      "epoch": 0.3823722149410223,
      "grad_norm": 0.15262024104595184,
      "learning_rate": 0.0001,
      "loss": 1.3537,
      "step": 2334
    },
    {
      "epoch": 0.3825360419397117,
      "grad_norm": 0.15360096096992493,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 2335
    },
    {
      "epoch": 0.382699868938401,
      "grad_norm": 0.16709662973880768,
      "learning_rate": 0.0001,
      "loss": 1.3598,
      "step": 2336
    },
    {
      "epoch": 0.3828636959370904,
      "grad_norm": 0.15259000658988953,
      "learning_rate": 0.0001,
      "loss": 1.402,
      "step": 2337
    },
    {
      "epoch": 0.3830275229357798,
      "grad_norm": 0.15057791769504547,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 2338
    },
    {
      "epoch": 0.3831913499344692,
      "grad_norm": 0.1460309773683548,
      "learning_rate": 0.0001,
      "loss": 1.4568,
      "step": 2339
    },
    {
      "epoch": 0.38335517693315857,
      "grad_norm": 0.15133918821811676,
      "learning_rate": 0.0001,
      "loss": 1.37,
      "step": 2340
    },
    {
      "epoch": 0.38351900393184796,
      "grad_norm": 0.14783698320388794,
      "learning_rate": 0.0001,
      "loss": 1.4871,
      "step": 2341
    },
    {
      "epoch": 0.38368283093053734,
      "grad_norm": 0.14270305633544922,
      "learning_rate": 0.0001,
      "loss": 1.3928,
      "step": 2342
    },
    {
      "epoch": 0.38384665792922673,
      "grad_norm": 0.15686197578907013,
      "learning_rate": 0.0001,
      "loss": 1.4137,
      "step": 2343
    },
    {
      "epoch": 0.3840104849279161,
      "grad_norm": 0.147307351231575,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 2344
    },
    {
      "epoch": 0.3841743119266055,
      "grad_norm": 0.15715551376342773,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 2345
    },
    {
      "epoch": 0.3843381389252949,
      "grad_norm": 0.1652243435382843,
      "learning_rate": 0.0001,
      "loss": 1.4389,
      "step": 2346
    },
    {
      "epoch": 0.3845019659239843,
      "grad_norm": 0.149471253156662,
      "learning_rate": 0.0001,
      "loss": 1.3732,
      "step": 2347
    },
    {
      "epoch": 0.38466579292267367,
      "grad_norm": 0.15216177701950073,
      "learning_rate": 0.0001,
      "loss": 1.2866,
      "step": 2348
    },
    {
      "epoch": 0.38482961992136305,
      "grad_norm": 0.14930860698223114,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 2349
    },
    {
      "epoch": 0.38499344692005244,
      "grad_norm": 0.1449468731880188,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 2350
    },
    {
      "epoch": 0.3851572739187418,
      "grad_norm": 0.15307986736297607,
      "learning_rate": 0.0001,
      "loss": 1.529,
      "step": 2351
    },
    {
      "epoch": 0.3853211009174312,
      "grad_norm": 0.13854564726352692,
      "learning_rate": 0.0001,
      "loss": 1.263,
      "step": 2352
    },
    {
      "epoch": 0.3854849279161206,
      "grad_norm": 0.14291517436504364,
      "learning_rate": 0.0001,
      "loss": 1.4283,
      "step": 2353
    },
    {
      "epoch": 0.38564875491481,
      "grad_norm": 0.15100780129432678,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 2354
    },
    {
      "epoch": 0.3858125819134993,
      "grad_norm": 0.13911060988903046,
      "learning_rate": 0.0001,
      "loss": 1.3973,
      "step": 2355
    },
    {
      "epoch": 0.3859764089121887,
      "grad_norm": 0.14976902306079865,
      "learning_rate": 0.0001,
      "loss": 1.3573,
      "step": 2356
    },
    {
      "epoch": 0.3861402359108781,
      "grad_norm": 0.14963756501674652,
      "learning_rate": 0.0001,
      "loss": 1.3824,
      "step": 2357
    },
    {
      "epoch": 0.3863040629095675,
      "grad_norm": 0.1444290280342102,
      "learning_rate": 0.0001,
      "loss": 1.4154,
      "step": 2358
    },
    {
      "epoch": 0.38646788990825687,
      "grad_norm": 0.1462053656578064,
      "learning_rate": 0.0001,
      "loss": 1.4613,
      "step": 2359
    },
    {
      "epoch": 0.38663171690694625,
      "grad_norm": 0.1459377408027649,
      "learning_rate": 0.0001,
      "loss": 1.4554,
      "step": 2360
    },
    {
      "epoch": 0.38679554390563564,
      "grad_norm": 0.14414416253566742,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 2361
    },
    {
      "epoch": 0.38695937090432503,
      "grad_norm": 0.14800980687141418,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 2362
    },
    {
      "epoch": 0.3871231979030144,
      "grad_norm": 0.1530102640390396,
      "learning_rate": 0.0001,
      "loss": 1.4334,
      "step": 2363
    },
    {
      "epoch": 0.3872870249017038,
      "grad_norm": 0.14771685004234314,
      "learning_rate": 0.0001,
      "loss": 1.4772,
      "step": 2364
    },
    {
      "epoch": 0.3874508519003932,
      "grad_norm": 0.14525920152664185,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 2365
    },
    {
      "epoch": 0.3876146788990826,
      "grad_norm": 0.1427529752254486,
      "learning_rate": 0.0001,
      "loss": 1.3548,
      "step": 2366
    },
    {
      "epoch": 0.38777850589777196,
      "grad_norm": 0.15163177251815796,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 2367
    },
    {
      "epoch": 0.38794233289646135,
      "grad_norm": 0.15674318373203278,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 2368
    },
    {
      "epoch": 0.38810615989515074,
      "grad_norm": 0.1617388278245926,
      "learning_rate": 0.0001,
      "loss": 1.5021,
      "step": 2369
    },
    {
      "epoch": 0.3882699868938401,
      "grad_norm": 0.14243784546852112,
      "learning_rate": 0.0001,
      "loss": 1.3764,
      "step": 2370
    },
    {
      "epoch": 0.3884338138925295,
      "grad_norm": 0.1520567387342453,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 2371
    },
    {
      "epoch": 0.3885976408912189,
      "grad_norm": 0.15338481962680817,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 2372
    },
    {
      "epoch": 0.38876146788990823,
      "grad_norm": 0.16232556104660034,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 2373
    },
    {
      "epoch": 0.3889252948885976,
      "grad_norm": 0.15680763125419617,
      "learning_rate": 0.0001,
      "loss": 1.3558,
      "step": 2374
    },
    {
      "epoch": 0.389089121887287,
      "grad_norm": 0.14583218097686768,
      "learning_rate": 0.0001,
      "loss": 1.4086,
      "step": 2375
    },
    {
      "epoch": 0.3892529488859764,
      "grad_norm": 0.1546075940132141,
      "learning_rate": 0.0001,
      "loss": 1.5424,
      "step": 2376
    },
    {
      "epoch": 0.3894167758846658,
      "grad_norm": 0.14731718599796295,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 2377
    },
    {
      "epoch": 0.38958060288335516,
      "grad_norm": 0.15872612595558167,
      "learning_rate": 0.0001,
      "loss": 1.3013,
      "step": 2378
    },
    {
      "epoch": 0.38974442988204455,
      "grad_norm": 0.1605895459651947,
      "learning_rate": 0.0001,
      "loss": 1.4006,
      "step": 2379
    },
    {
      "epoch": 0.38990825688073394,
      "grad_norm": 0.13742883503437042,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 2380
    },
    {
      "epoch": 0.3900720838794233,
      "grad_norm": 0.14324870705604553,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 2381
    },
    {
      "epoch": 0.3902359108781127,
      "grad_norm": 0.1513160914182663,
      "learning_rate": 0.0001,
      "loss": 1.4083,
      "step": 2382
    },
    {
      "epoch": 0.3903997378768021,
      "grad_norm": 0.14188188314437866,
      "learning_rate": 0.0001,
      "loss": 1.4329,
      "step": 2383
    },
    {
      "epoch": 0.3905635648754915,
      "grad_norm": 0.1505788415670395,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 2384
    },
    {
      "epoch": 0.3907273918741809,
      "grad_norm": 0.143506720662117,
      "learning_rate": 0.0001,
      "loss": 1.4928,
      "step": 2385
    },
    {
      "epoch": 0.39089121887287026,
      "grad_norm": 0.15057533979415894,
      "learning_rate": 0.0001,
      "loss": 1.4234,
      "step": 2386
    },
    {
      "epoch": 0.39105504587155965,
      "grad_norm": 0.14766007661819458,
      "learning_rate": 0.0001,
      "loss": 1.4372,
      "step": 2387
    },
    {
      "epoch": 0.39121887287024903,
      "grad_norm": 0.15161535143852234,
      "learning_rate": 0.0001,
      "loss": 1.3494,
      "step": 2388
    },
    {
      "epoch": 0.3913826998689384,
      "grad_norm": 0.145694300532341,
      "learning_rate": 0.0001,
      "loss": 1.409,
      "step": 2389
    },
    {
      "epoch": 0.3915465268676278,
      "grad_norm": 0.14679044485092163,
      "learning_rate": 0.0001,
      "loss": 1.2692,
      "step": 2390
    },
    {
      "epoch": 0.3917103538663172,
      "grad_norm": 0.14385615289211273,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 2391
    },
    {
      "epoch": 0.3918741808650065,
      "grad_norm": 0.1581970453262329,
      "learning_rate": 0.0001,
      "loss": 1.4397,
      "step": 2392
    },
    {
      "epoch": 0.3920380078636959,
      "grad_norm": 0.1468660831451416,
      "learning_rate": 0.0001,
      "loss": 1.3164,
      "step": 2393
    },
    {
      "epoch": 0.3922018348623853,
      "grad_norm": 0.15065287053585052,
      "learning_rate": 0.0001,
      "loss": 1.4281,
      "step": 2394
    },
    {
      "epoch": 0.3923656618610747,
      "grad_norm": 0.15431161224842072,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 2395
    },
    {
      "epoch": 0.3925294888597641,
      "grad_norm": 0.14657315611839294,
      "learning_rate": 0.0001,
      "loss": 1.3921,
      "step": 2396
    },
    {
      "epoch": 0.39269331585845346,
      "grad_norm": 0.15412791073322296,
      "learning_rate": 0.0001,
      "loss": 1.4435,
      "step": 2397
    },
    {
      "epoch": 0.39285714285714285,
      "grad_norm": 0.1454719454050064,
      "learning_rate": 0.0001,
      "loss": 1.3547,
      "step": 2398
    },
    {
      "epoch": 0.39302096985583224,
      "grad_norm": 0.14396053552627563,
      "learning_rate": 0.0001,
      "loss": 1.3044,
      "step": 2399
    },
    {
      "epoch": 0.3931847968545216,
      "grad_norm": 0.14520706236362457,
      "learning_rate": 0.0001,
      "loss": 1.3636,
      "step": 2400
    },
    {
      "epoch": 0.393348623853211,
      "grad_norm": 0.1545875370502472,
      "learning_rate": 0.0001,
      "loss": 1.4172,
      "step": 2401
    },
    {
      "epoch": 0.3935124508519004,
      "grad_norm": 0.1519928276538849,
      "learning_rate": 0.0001,
      "loss": 1.4914,
      "step": 2402
    },
    {
      "epoch": 0.3936762778505898,
      "grad_norm": 0.14023195207118988,
      "learning_rate": 0.0001,
      "loss": 1.2776,
      "step": 2403
    },
    {
      "epoch": 0.39384010484927917,
      "grad_norm": 0.15426473319530487,
      "learning_rate": 0.0001,
      "loss": 1.3388,
      "step": 2404
    },
    {
      "epoch": 0.39400393184796856,
      "grad_norm": 0.14434881508350372,
      "learning_rate": 0.0001,
      "loss": 1.457,
      "step": 2405
    },
    {
      "epoch": 0.39416775884665795,
      "grad_norm": 0.14144235849380493,
      "learning_rate": 0.0001,
      "loss": 1.5034,
      "step": 2406
    },
    {
      "epoch": 0.39433158584534733,
      "grad_norm": 0.14857922494411469,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 2407
    },
    {
      "epoch": 0.3944954128440367,
      "grad_norm": 0.15059998631477356,
      "learning_rate": 0.0001,
      "loss": 1.4,
      "step": 2408
    },
    {
      "epoch": 0.3946592398427261,
      "grad_norm": 0.15325510501861572,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 2409
    },
    {
      "epoch": 0.39482306684141544,
      "grad_norm": 0.16302776336669922,
      "learning_rate": 0.0001,
      "loss": 1.374,
      "step": 2410
    },
    {
      "epoch": 0.3949868938401048,
      "grad_norm": 0.15149818360805511,
      "learning_rate": 0.0001,
      "loss": 1.4132,
      "step": 2411
    },
    {
      "epoch": 0.3951507208387942,
      "grad_norm": 0.15564587712287903,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 2412
    },
    {
      "epoch": 0.3953145478374836,
      "grad_norm": 0.15537816286087036,
      "learning_rate": 0.0001,
      "loss": 1.5212,
      "step": 2413
    },
    {
      "epoch": 0.395478374836173,
      "grad_norm": 0.14151465892791748,
      "learning_rate": 0.0001,
      "loss": 1.3651,
      "step": 2414
    },
    {
      "epoch": 0.3956422018348624,
      "grad_norm": 0.1554529219865799,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 2415
    },
    {
      "epoch": 0.39580602883355176,
      "grad_norm": 0.15614469349384308,
      "learning_rate": 0.0001,
      "loss": 1.4213,
      "step": 2416
    },
    {
      "epoch": 0.39596985583224115,
      "grad_norm": 0.14715610444545746,
      "learning_rate": 0.0001,
      "loss": 1.3551,
      "step": 2417
    },
    {
      "epoch": 0.39613368283093053,
      "grad_norm": 0.14572781324386597,
      "learning_rate": 0.0001,
      "loss": 1.4273,
      "step": 2418
    },
    {
      "epoch": 0.3962975098296199,
      "grad_norm": 0.1541505604982376,
      "learning_rate": 0.0001,
      "loss": 1.3379,
      "step": 2419
    },
    {
      "epoch": 0.3964613368283093,
      "grad_norm": 0.14806011319160461,
      "learning_rate": 0.0001,
      "loss": 1.4687,
      "step": 2420
    },
    {
      "epoch": 0.3966251638269987,
      "grad_norm": 0.15307824313640594,
      "learning_rate": 0.0001,
      "loss": 1.4069,
      "step": 2421
    },
    {
      "epoch": 0.3967889908256881,
      "grad_norm": 0.15404696762561798,
      "learning_rate": 0.0001,
      "loss": 1.4176,
      "step": 2422
    },
    {
      "epoch": 0.39695281782437747,
      "grad_norm": 0.14859159290790558,
      "learning_rate": 0.0001,
      "loss": 1.3776,
      "step": 2423
    },
    {
      "epoch": 0.39711664482306686,
      "grad_norm": 0.15018390119075775,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 2424
    },
    {
      "epoch": 0.39728047182175624,
      "grad_norm": 0.14426064491271973,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 2425
    },
    {
      "epoch": 0.39744429882044563,
      "grad_norm": 0.14562033116817474,
      "learning_rate": 0.0001,
      "loss": 1.4419,
      "step": 2426
    },
    {
      "epoch": 0.397608125819135,
      "grad_norm": 0.14631861448287964,
      "learning_rate": 0.0001,
      "loss": 1.3199,
      "step": 2427
    },
    {
      "epoch": 0.3977719528178244,
      "grad_norm": 0.15547482669353485,
      "learning_rate": 0.0001,
      "loss": 1.3636,
      "step": 2428
    },
    {
      "epoch": 0.39793577981651373,
      "grad_norm": 0.14649444818496704,
      "learning_rate": 0.0001,
      "loss": 1.3841,
      "step": 2429
    },
    {
      "epoch": 0.3980996068152031,
      "grad_norm": 0.1481143832206726,
      "learning_rate": 0.0001,
      "loss": 1.3765,
      "step": 2430
    },
    {
      "epoch": 0.3982634338138925,
      "grad_norm": 0.15687237679958344,
      "learning_rate": 0.0001,
      "loss": 1.4635,
      "step": 2431
    },
    {
      "epoch": 0.3984272608125819,
      "grad_norm": 0.1513882875442505,
      "learning_rate": 0.0001,
      "loss": 1.4716,
      "step": 2432
    },
    {
      "epoch": 0.3985910878112713,
      "grad_norm": 0.14148876070976257,
      "learning_rate": 0.0001,
      "loss": 1.4448,
      "step": 2433
    },
    {
      "epoch": 0.39875491480996067,
      "grad_norm": 0.1497381180524826,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 2434
    },
    {
      "epoch": 0.39891874180865006,
      "grad_norm": 0.1550208479166031,
      "learning_rate": 0.0001,
      "loss": 1.3683,
      "step": 2435
    },
    {
      "epoch": 0.39908256880733944,
      "grad_norm": 0.16146330535411835,
      "learning_rate": 0.0001,
      "loss": 1.4203,
      "step": 2436
    },
    {
      "epoch": 0.39924639580602883,
      "grad_norm": 0.15215200185775757,
      "learning_rate": 0.0001,
      "loss": 1.4441,
      "step": 2437
    },
    {
      "epoch": 0.3994102228047182,
      "grad_norm": 0.14610891044139862,
      "learning_rate": 0.0001,
      "loss": 1.4483,
      "step": 2438
    },
    {
      "epoch": 0.3995740498034076,
      "grad_norm": 0.14670845866203308,
      "learning_rate": 0.0001,
      "loss": 1.3754,
      "step": 2439
    },
    {
      "epoch": 0.399737876802097,
      "grad_norm": 0.1424649953842163,
      "learning_rate": 0.0001,
      "loss": 1.2214,
      "step": 2440
    },
    {
      "epoch": 0.3999017038007864,
      "grad_norm": 0.13959206640720367,
      "learning_rate": 0.0001,
      "loss": 1.3313,
      "step": 2441
    },
    {
      "epoch": 0.40006553079947577,
      "grad_norm": 0.15221761167049408,
      "learning_rate": 0.0001,
      "loss": 1.4365,
      "step": 2442
    },
    {
      "epoch": 0.40022935779816515,
      "grad_norm": 0.14928589761257172,
      "learning_rate": 0.0001,
      "loss": 1.4447,
      "step": 2443
    },
    {
      "epoch": 0.40039318479685454,
      "grad_norm": 0.15439894795417786,
      "learning_rate": 0.0001,
      "loss": 1.4047,
      "step": 2444
    },
    {
      "epoch": 0.4005570117955439,
      "grad_norm": 0.15295375883579254,
      "learning_rate": 0.0001,
      "loss": 1.4097,
      "step": 2445
    },
    {
      "epoch": 0.4007208387942333,
      "grad_norm": 0.1555979698896408,
      "learning_rate": 0.0001,
      "loss": 1.4871,
      "step": 2446
    },
    {
      "epoch": 0.4008846657929227,
      "grad_norm": 0.1434689611196518,
      "learning_rate": 0.0001,
      "loss": 1.4393,
      "step": 2447
    },
    {
      "epoch": 0.40104849279161203,
      "grad_norm": 0.1425488442182541,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 2448
    },
    {
      "epoch": 0.4012123197903014,
      "grad_norm": 0.1400412619113922,
      "learning_rate": 0.0001,
      "loss": 1.4181,
      "step": 2449
    },
    {
      "epoch": 0.4013761467889908,
      "grad_norm": 0.15059661865234375,
      "learning_rate": 0.0001,
      "loss": 1.4296,
      "step": 2450
    },
    {
      "epoch": 0.4015399737876802,
      "grad_norm": 0.14877402782440186,
      "learning_rate": 0.0001,
      "loss": 1.4137,
      "step": 2451
    },
    {
      "epoch": 0.4017038007863696,
      "grad_norm": 0.1415184736251831,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 2452
    },
    {
      "epoch": 0.40186762778505897,
      "grad_norm": 0.15345197916030884,
      "learning_rate": 0.0001,
      "loss": 1.417,
      "step": 2453
    },
    {
      "epoch": 0.40203145478374835,
      "grad_norm": 0.16188839077949524,
      "learning_rate": 0.0001,
      "loss": 1.3789,
      "step": 2454
    },
    {
      "epoch": 0.40219528178243774,
      "grad_norm": 0.14279690384864807,
      "learning_rate": 0.0001,
      "loss": 1.3503,
      "step": 2455
    },
    {
      "epoch": 0.40235910878112713,
      "grad_norm": 0.14649631083011627,
      "learning_rate": 0.0001,
      "loss": 1.477,
      "step": 2456
    },
    {
      "epoch": 0.4025229357798165,
      "grad_norm": 0.14913256466388702,
      "learning_rate": 0.0001,
      "loss": 1.2895,
      "step": 2457
    },
    {
      "epoch": 0.4026867627785059,
      "grad_norm": 0.15162895619869232,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 2458
    },
    {
      "epoch": 0.4028505897771953,
      "grad_norm": 0.14950183033943176,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 2459
    },
    {
      "epoch": 0.4030144167758847,
      "grad_norm": 0.17304813861846924,
      "learning_rate": 0.0001,
      "loss": 1.4261,
      "step": 2460
    },
    {
      "epoch": 0.40317824377457406,
      "grad_norm": 0.1398339718580246,
      "learning_rate": 0.0001,
      "loss": 1.3244,
      "step": 2461
    },
    {
      "epoch": 0.40334207077326345,
      "grad_norm": 0.14538155496120453,
      "learning_rate": 0.0001,
      "loss": 1.3733,
      "step": 2462
    },
    {
      "epoch": 0.40350589777195284,
      "grad_norm": 0.14969469606876373,
      "learning_rate": 0.0001,
      "loss": 1.4397,
      "step": 2463
    },
    {
      "epoch": 0.4036697247706422,
      "grad_norm": 0.1534261703491211,
      "learning_rate": 0.0001,
      "loss": 1.3632,
      "step": 2464
    },
    {
      "epoch": 0.4038335517693316,
      "grad_norm": 0.14824901521205902,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 2465
    },
    {
      "epoch": 0.40399737876802094,
      "grad_norm": 0.14226961135864258,
      "learning_rate": 0.0001,
      "loss": 1.4653,
      "step": 2466
    },
    {
      "epoch": 0.40416120576671033,
      "grad_norm": 0.14850479364395142,
      "learning_rate": 0.0001,
      "loss": 1.3665,
      "step": 2467
    },
    {
      "epoch": 0.4043250327653997,
      "grad_norm": 0.1527779996395111,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 2468
    },
    {
      "epoch": 0.4044888597640891,
      "grad_norm": 0.14922605454921722,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 2469
    },
    {
      "epoch": 0.4046526867627785,
      "grad_norm": 0.14513222873210907,
      "learning_rate": 0.0001,
      "loss": 1.4281,
      "step": 2470
    },
    {
      "epoch": 0.4048165137614679,
      "grad_norm": 0.1459590047597885,
      "learning_rate": 0.0001,
      "loss": 1.3929,
      "step": 2471
    },
    {
      "epoch": 0.40498034076015726,
      "grad_norm": 0.14371156692504883,
      "learning_rate": 0.0001,
      "loss": 1.411,
      "step": 2472
    },
    {
      "epoch": 0.40514416775884665,
      "grad_norm": 0.1455857753753662,
      "learning_rate": 0.0001,
      "loss": 1.4175,
      "step": 2473
    },
    {
      "epoch": 0.40530799475753604,
      "grad_norm": 0.146856427192688,
      "learning_rate": 0.0001,
      "loss": 1.4175,
      "step": 2474
    },
    {
      "epoch": 0.4054718217562254,
      "grad_norm": 0.16053740680217743,
      "learning_rate": 0.0001,
      "loss": 1.4314,
      "step": 2475
    },
    {
      "epoch": 0.4056356487549148,
      "grad_norm": 0.1496027261018753,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 2476
    },
    {
      "epoch": 0.4057994757536042,
      "grad_norm": 0.14471876621246338,
      "learning_rate": 0.0001,
      "loss": 1.36,
      "step": 2477
    },
    {
      "epoch": 0.4059633027522936,
      "grad_norm": 0.15499937534332275,
      "learning_rate": 0.0001,
      "loss": 1.479,
      "step": 2478
    },
    {
      "epoch": 0.406127129750983,
      "grad_norm": 0.1471986025571823,
      "learning_rate": 0.0001,
      "loss": 1.3983,
      "step": 2479
    },
    {
      "epoch": 0.40629095674967236,
      "grad_norm": 0.15037569403648376,
      "learning_rate": 0.0001,
      "loss": 1.5,
      "step": 2480
    },
    {
      "epoch": 0.40645478374836175,
      "grad_norm": 0.159732386469841,
      "learning_rate": 0.0001,
      "loss": 1.4672,
      "step": 2481
    },
    {
      "epoch": 0.40661861074705113,
      "grad_norm": 0.145772784948349,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 2482
    },
    {
      "epoch": 0.4067824377457405,
      "grad_norm": 0.1439577341079712,
      "learning_rate": 0.0001,
      "loss": 1.4496,
      "step": 2483
    },
    {
      "epoch": 0.4069462647444299,
      "grad_norm": 0.14422014355659485,
      "learning_rate": 0.0001,
      "loss": 1.38,
      "step": 2484
    },
    {
      "epoch": 0.40711009174311924,
      "grad_norm": 0.14966712892055511,
      "learning_rate": 0.0001,
      "loss": 1.3997,
      "step": 2485
    },
    {
      "epoch": 0.4072739187418086,
      "grad_norm": 0.15341031551361084,
      "learning_rate": 0.0001,
      "loss": 1.5188,
      "step": 2486
    },
    {
      "epoch": 0.407437745740498,
      "grad_norm": 0.1598280966281891,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 2487
    },
    {
      "epoch": 0.4076015727391874,
      "grad_norm": 0.14850689470767975,
      "learning_rate": 0.0001,
      "loss": 1.3857,
      "step": 2488
    },
    {
      "epoch": 0.4077653997378768,
      "grad_norm": 0.16239039599895477,
      "learning_rate": 0.0001,
      "loss": 1.5007,
      "step": 2489
    },
    {
      "epoch": 0.4079292267365662,
      "grad_norm": 0.15081410109996796,
      "learning_rate": 0.0001,
      "loss": 1.4733,
      "step": 2490
    },
    {
      "epoch": 0.40809305373525556,
      "grad_norm": 0.14787894487380981,
      "learning_rate": 0.0001,
      "loss": 1.4216,
      "step": 2491
    },
    {
      "epoch": 0.40825688073394495,
      "grad_norm": 0.14062444865703583,
      "learning_rate": 0.0001,
      "loss": 1.2971,
      "step": 2492
    },
    {
      "epoch": 0.40842070773263434,
      "grad_norm": 0.1481916606426239,
      "learning_rate": 0.0001,
      "loss": 1.4414,
      "step": 2493
    },
    {
      "epoch": 0.4085845347313237,
      "grad_norm": 0.13879208266735077,
      "learning_rate": 0.0001,
      "loss": 1.2144,
      "step": 2494
    },
    {
      "epoch": 0.4087483617300131,
      "grad_norm": 0.14813703298568726,
      "learning_rate": 0.0001,
      "loss": 1.4759,
      "step": 2495
    },
    {
      "epoch": 0.4089121887287025,
      "grad_norm": 0.14978714287281036,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 2496
    },
    {
      "epoch": 0.4090760157273919,
      "grad_norm": 0.14595647156238556,
      "learning_rate": 0.0001,
      "loss": 1.4703,
      "step": 2497
    },
    {
      "epoch": 0.40923984272608127,
      "grad_norm": 0.14923658967018127,
      "learning_rate": 0.0001,
      "loss": 1.3903,
      "step": 2498
    },
    {
      "epoch": 0.40940366972477066,
      "grad_norm": 0.14398391544818878,
      "learning_rate": 0.0001,
      "loss": 1.4574,
      "step": 2499
    },
    {
      "epoch": 0.40956749672346004,
      "grad_norm": 0.14416387677192688,
      "learning_rate": 0.0001,
      "loss": 1.4605,
      "step": 2500
    },
    {
      "epoch": 0.40973132372214943,
      "grad_norm": 0.14303745329380035,
      "learning_rate": 0.0001,
      "loss": 1.3988,
      "step": 2501
    },
    {
      "epoch": 0.4098951507208388,
      "grad_norm": 0.14431026577949524,
      "learning_rate": 0.0001,
      "loss": 1.2912,
      "step": 2502
    },
    {
      "epoch": 0.4100589777195282,
      "grad_norm": 0.15125082433223724,
      "learning_rate": 0.0001,
      "loss": 1.4806,
      "step": 2503
    },
    {
      "epoch": 0.41022280471821754,
      "grad_norm": 0.141354039311409,
      "learning_rate": 0.0001,
      "loss": 1.3903,
      "step": 2504
    },
    {
      "epoch": 0.4103866317169069,
      "grad_norm": 0.14696793258190155,
      "learning_rate": 0.0001,
      "loss": 1.2979,
      "step": 2505
    },
    {
      "epoch": 0.4105504587155963,
      "grad_norm": 0.15625838935375214,
      "learning_rate": 0.0001,
      "loss": 1.4381,
      "step": 2506
    },
    {
      "epoch": 0.4107142857142857,
      "grad_norm": 0.14602528512477875,
      "learning_rate": 0.0001,
      "loss": 1.4255,
      "step": 2507
    },
    {
      "epoch": 0.4108781127129751,
      "grad_norm": 0.1524679809808731,
      "learning_rate": 0.0001,
      "loss": 1.4355,
      "step": 2508
    },
    {
      "epoch": 0.4110419397116645,
      "grad_norm": 0.14499841630458832,
      "learning_rate": 0.0001,
      "loss": 1.3647,
      "step": 2509
    },
    {
      "epoch": 0.41120576671035386,
      "grad_norm": 0.14647114276885986,
      "learning_rate": 0.0001,
      "loss": 1.4447,
      "step": 2510
    },
    {
      "epoch": 0.41136959370904325,
      "grad_norm": 0.14921912550926208,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 2511
    },
    {
      "epoch": 0.41153342070773263,
      "grad_norm": 0.15089498460292816,
      "learning_rate": 0.0001,
      "loss": 1.3384,
      "step": 2512
    },
    {
      "epoch": 0.411697247706422,
      "grad_norm": 0.1443488597869873,
      "learning_rate": 0.0001,
      "loss": 1.3654,
      "step": 2513
    },
    {
      "epoch": 0.4118610747051114,
      "grad_norm": 0.146475151181221,
      "learning_rate": 0.0001,
      "loss": 1.3078,
      "step": 2514
    },
    {
      "epoch": 0.4120249017038008,
      "grad_norm": 0.153826043009758,
      "learning_rate": 0.0001,
      "loss": 1.4668,
      "step": 2515
    },
    {
      "epoch": 0.4121887287024902,
      "grad_norm": 0.14650121331214905,
      "learning_rate": 0.0001,
      "loss": 1.37,
      "step": 2516
    },
    {
      "epoch": 0.41235255570117957,
      "grad_norm": 0.14384005963802338,
      "learning_rate": 0.0001,
      "loss": 1.381,
      "step": 2517
    },
    {
      "epoch": 0.41251638269986896,
      "grad_norm": 0.15534716844558716,
      "learning_rate": 0.0001,
      "loss": 1.3867,
      "step": 2518
    },
    {
      "epoch": 0.41268020969855834,
      "grad_norm": 0.15433864295482635,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 2519
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 0.154121533036232,
      "learning_rate": 0.0001,
      "loss": 1.4489,
      "step": 2520
    },
    {
      "epoch": 0.4130078636959371,
      "grad_norm": 0.14587262272834778,
      "learning_rate": 0.0001,
      "loss": 1.36,
      "step": 2521
    },
    {
      "epoch": 0.41317169069462645,
      "grad_norm": 0.14514100551605225,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 2522
    },
    {
      "epoch": 0.41333551769331583,
      "grad_norm": 0.1536257416009903,
      "learning_rate": 0.0001,
      "loss": 1.4025,
      "step": 2523
    },
    {
      "epoch": 0.4134993446920052,
      "grad_norm": 0.1486305147409439,
      "learning_rate": 0.0001,
      "loss": 1.4268,
      "step": 2524
    },
    {
      "epoch": 0.4136631716906946,
      "grad_norm": 0.1445842683315277,
      "learning_rate": 0.0001,
      "loss": 1.3893,
      "step": 2525
    },
    {
      "epoch": 0.413826998689384,
      "grad_norm": 0.14322592318058014,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 2526
    },
    {
      "epoch": 0.4139908256880734,
      "grad_norm": 0.1454222947359085,
      "learning_rate": 0.0001,
      "loss": 1.3344,
      "step": 2527
    },
    {
      "epoch": 0.41415465268676277,
      "grad_norm": 0.14990738034248352,
      "learning_rate": 0.0001,
      "loss": 1.4991,
      "step": 2528
    },
    {
      "epoch": 0.41431847968545216,
      "grad_norm": 0.146295428276062,
      "learning_rate": 0.0001,
      "loss": 1.5328,
      "step": 2529
    },
    {
      "epoch": 0.41448230668414154,
      "grad_norm": 0.15165507793426514,
      "learning_rate": 0.0001,
      "loss": 1.4344,
      "step": 2530
    },
    {
      "epoch": 0.41464613368283093,
      "grad_norm": 0.14422905445098877,
      "learning_rate": 0.0001,
      "loss": 1.3851,
      "step": 2531
    },
    {
      "epoch": 0.4148099606815203,
      "grad_norm": 0.1507553905248642,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 2532
    },
    {
      "epoch": 0.4149737876802097,
      "grad_norm": 0.15098853409290314,
      "learning_rate": 0.0001,
      "loss": 1.4025,
      "step": 2533
    },
    {
      "epoch": 0.4151376146788991,
      "grad_norm": 0.15409541130065918,
      "learning_rate": 0.0001,
      "loss": 1.4616,
      "step": 2534
    },
    {
      "epoch": 0.4153014416775885,
      "grad_norm": 0.1479106843471527,
      "learning_rate": 0.0001,
      "loss": 1.4338,
      "step": 2535
    },
    {
      "epoch": 0.41546526867627787,
      "grad_norm": 0.15023742616176605,
      "learning_rate": 0.0001,
      "loss": 1.328,
      "step": 2536
    },
    {
      "epoch": 0.41562909567496725,
      "grad_norm": 0.140737384557724,
      "learning_rate": 0.0001,
      "loss": 1.3515,
      "step": 2537
    },
    {
      "epoch": 0.41579292267365664,
      "grad_norm": 0.1549571454524994,
      "learning_rate": 0.0001,
      "loss": 1.476,
      "step": 2538
    },
    {
      "epoch": 0.415956749672346,
      "grad_norm": 0.1529311090707779,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 2539
    },
    {
      "epoch": 0.4161205766710354,
      "grad_norm": 0.15314261615276337,
      "learning_rate": 0.0001,
      "loss": 1.4425,
      "step": 2540
    },
    {
      "epoch": 0.41628440366972475,
      "grad_norm": 0.15346045792102814,
      "learning_rate": 0.0001,
      "loss": 1.4256,
      "step": 2541
    },
    {
      "epoch": 0.41644823066841413,
      "grad_norm": 0.14649097621440887,
      "learning_rate": 0.0001,
      "loss": 1.3767,
      "step": 2542
    },
    {
      "epoch": 0.4166120576671035,
      "grad_norm": 0.13621006906032562,
      "learning_rate": 0.0001,
      "loss": 1.3928,
      "step": 2543
    },
    {
      "epoch": 0.4167758846657929,
      "grad_norm": 0.15840837359428406,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 2544
    },
    {
      "epoch": 0.4169397116644823,
      "grad_norm": 0.14628297090530396,
      "learning_rate": 0.0001,
      "loss": 1.3695,
      "step": 2545
    },
    {
      "epoch": 0.4171035386631717,
      "grad_norm": 0.15389777719974518,
      "learning_rate": 0.0001,
      "loss": 1.3781,
      "step": 2546
    },
    {
      "epoch": 0.41726736566186107,
      "grad_norm": 0.1524418592453003,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 2547
    },
    {
      "epoch": 0.41743119266055045,
      "grad_norm": 0.15184010565280914,
      "learning_rate": 0.0001,
      "loss": 1.4228,
      "step": 2548
    },
    {
      "epoch": 0.41759501965923984,
      "grad_norm": 0.14787764847278595,
      "learning_rate": 0.0001,
      "loss": 1.3934,
      "step": 2549
    },
    {
      "epoch": 0.41775884665792923,
      "grad_norm": 0.15565109252929688,
      "learning_rate": 0.0001,
      "loss": 1.4586,
      "step": 2550
    },
    {
      "epoch": 0.4179226736566186,
      "grad_norm": 0.15104781091213226,
      "learning_rate": 0.0001,
      "loss": 1.4252,
      "step": 2551
    },
    {
      "epoch": 0.418086500655308,
      "grad_norm": 0.15219812095165253,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 2552
    },
    {
      "epoch": 0.4182503276539974,
      "grad_norm": 0.1460484266281128,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 2553
    },
    {
      "epoch": 0.4184141546526868,
      "grad_norm": 0.15429821610450745,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 2554
    },
    {
      "epoch": 0.41857798165137616,
      "grad_norm": 0.14035376906394958,
      "learning_rate": 0.0001,
      "loss": 1.2773,
      "step": 2555
    },
    {
      "epoch": 0.41874180865006555,
      "grad_norm": 0.16053830087184906,
      "learning_rate": 0.0001,
      "loss": 1.5589,
      "step": 2556
    },
    {
      "epoch": 0.41890563564875494,
      "grad_norm": 0.15910805761814117,
      "learning_rate": 0.0001,
      "loss": 1.4453,
      "step": 2557
    },
    {
      "epoch": 0.4190694626474443,
      "grad_norm": 0.14605949819087982,
      "learning_rate": 0.0001,
      "loss": 1.3143,
      "step": 2558
    },
    {
      "epoch": 0.41923328964613366,
      "grad_norm": 0.14192338287830353,
      "learning_rate": 0.0001,
      "loss": 1.2988,
      "step": 2559
    },
    {
      "epoch": 0.41939711664482304,
      "grad_norm": 0.14896000921726227,
      "learning_rate": 0.0001,
      "loss": 1.4167,
      "step": 2560
    },
    {
      "epoch": 0.41956094364351243,
      "grad_norm": 0.15654434263706207,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 2561
    },
    {
      "epoch": 0.4197247706422018,
      "grad_norm": 0.15375560522079468,
      "learning_rate": 0.0001,
      "loss": 1.403,
      "step": 2562
    },
    {
      "epoch": 0.4198885976408912,
      "grad_norm": 0.14713813364505768,
      "learning_rate": 0.0001,
      "loss": 1.3468,
      "step": 2563
    },
    {
      "epoch": 0.4200524246395806,
      "grad_norm": 0.15281374752521515,
      "learning_rate": 0.0001,
      "loss": 1.4501,
      "step": 2564
    },
    {
      "epoch": 0.42021625163827,
      "grad_norm": 0.14667485654354095,
      "learning_rate": 0.0001,
      "loss": 1.3437,
      "step": 2565
    },
    {
      "epoch": 0.42038007863695936,
      "grad_norm": 0.1496182382106781,
      "learning_rate": 0.0001,
      "loss": 1.3653,
      "step": 2566
    },
    {
      "epoch": 0.42054390563564875,
      "grad_norm": 0.1426096260547638,
      "learning_rate": 0.0001,
      "loss": 1.3719,
      "step": 2567
    },
    {
      "epoch": 0.42070773263433814,
      "grad_norm": 0.1473112404346466,
      "learning_rate": 0.0001,
      "loss": 1.4412,
      "step": 2568
    },
    {
      "epoch": 0.4208715596330275,
      "grad_norm": 0.148676335811615,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 2569
    },
    {
      "epoch": 0.4210353866317169,
      "grad_norm": 0.16264712810516357,
      "learning_rate": 0.0001,
      "loss": 1.4821,
      "step": 2570
    },
    {
      "epoch": 0.4211992136304063,
      "grad_norm": 0.15442661941051483,
      "learning_rate": 0.0001,
      "loss": 1.3881,
      "step": 2571
    },
    {
      "epoch": 0.4213630406290957,
      "grad_norm": 0.15022100508213043,
      "learning_rate": 0.0001,
      "loss": 1.5661,
      "step": 2572
    },
    {
      "epoch": 0.4215268676277851,
      "grad_norm": 0.15201519429683685,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 2573
    },
    {
      "epoch": 0.42169069462647446,
      "grad_norm": 0.14038915932178497,
      "learning_rate": 0.0001,
      "loss": 1.4057,
      "step": 2574
    },
    {
      "epoch": 0.42185452162516385,
      "grad_norm": 0.14548785984516144,
      "learning_rate": 0.0001,
      "loss": 1.3569,
      "step": 2575
    },
    {
      "epoch": 0.42201834862385323,
      "grad_norm": 0.14441239833831787,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 2576
    },
    {
      "epoch": 0.4221821756225426,
      "grad_norm": 0.14475178718566895,
      "learning_rate": 0.0001,
      "loss": 1.3883,
      "step": 2577
    },
    {
      "epoch": 0.42234600262123195,
      "grad_norm": 0.1422284096479416,
      "learning_rate": 0.0001,
      "loss": 1.3574,
      "step": 2578
    },
    {
      "epoch": 0.42250982961992134,
      "grad_norm": 0.14153461158275604,
      "learning_rate": 0.0001,
      "loss": 1.3536,
      "step": 2579
    },
    {
      "epoch": 0.4226736566186107,
      "grad_norm": 0.15431278944015503,
      "learning_rate": 0.0001,
      "loss": 1.3128,
      "step": 2580
    },
    {
      "epoch": 0.4228374836173001,
      "grad_norm": 0.15002141892910004,
      "learning_rate": 0.0001,
      "loss": 1.4435,
      "step": 2581
    },
    {
      "epoch": 0.4230013106159895,
      "grad_norm": 0.16794338822364807,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 2582
    },
    {
      "epoch": 0.4231651376146789,
      "grad_norm": 0.13860325515270233,
      "learning_rate": 0.0001,
      "loss": 1.3152,
      "step": 2583
    },
    {
      "epoch": 0.4233289646133683,
      "grad_norm": 0.1479691118001938,
      "learning_rate": 0.0001,
      "loss": 1.4074,
      "step": 2584
    },
    {
      "epoch": 0.42349279161205766,
      "grad_norm": 0.15616071224212646,
      "learning_rate": 0.0001,
      "loss": 1.4791,
      "step": 2585
    },
    {
      "epoch": 0.42365661861074705,
      "grad_norm": 0.14320647716522217,
      "learning_rate": 0.0001,
      "loss": 1.3969,
      "step": 2586
    },
    {
      "epoch": 0.42382044560943644,
      "grad_norm": 0.1588820219039917,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 2587
    },
    {
      "epoch": 0.4239842726081258,
      "grad_norm": 0.15694963932037354,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 2588
    },
    {
      "epoch": 0.4241480996068152,
      "grad_norm": 0.1579674929380417,
      "learning_rate": 0.0001,
      "loss": 1.5769,
      "step": 2589
    },
    {
      "epoch": 0.4243119266055046,
      "grad_norm": 0.1527164876461029,
      "learning_rate": 0.0001,
      "loss": 1.373,
      "step": 2590
    },
    {
      "epoch": 0.424475753604194,
      "grad_norm": 0.15409746766090393,
      "learning_rate": 0.0001,
      "loss": 1.4602,
      "step": 2591
    },
    {
      "epoch": 0.42463958060288337,
      "grad_norm": 0.1446455717086792,
      "learning_rate": 0.0001,
      "loss": 1.4191,
      "step": 2592
    },
    {
      "epoch": 0.42480340760157276,
      "grad_norm": 0.1459273397922516,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 2593
    },
    {
      "epoch": 0.42496723460026214,
      "grad_norm": 0.15392424166202545,
      "learning_rate": 0.0001,
      "loss": 1.3224,
      "step": 2594
    },
    {
      "epoch": 0.42513106159895153,
      "grad_norm": 0.15042242407798767,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 2595
    },
    {
      "epoch": 0.4252948885976409,
      "grad_norm": 0.14808039367198944,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 2596
    },
    {
      "epoch": 0.42545871559633025,
      "grad_norm": 0.14730265736579895,
      "learning_rate": 0.0001,
      "loss": 1.4226,
      "step": 2597
    },
    {
      "epoch": 0.42562254259501964,
      "grad_norm": 0.149749755859375,
      "learning_rate": 0.0001,
      "loss": 1.4574,
      "step": 2598
    },
    {
      "epoch": 0.425786369593709,
      "grad_norm": 0.15158767998218536,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 2599
    },
    {
      "epoch": 0.4259501965923984,
      "grad_norm": 0.1462148129940033,
      "learning_rate": 0.0001,
      "loss": 1.2777,
      "step": 2600
    },
    {
      "epoch": 0.4261140235910878,
      "grad_norm": 0.15216006338596344,
      "learning_rate": 0.0001,
      "loss": 1.3585,
      "step": 2601
    },
    {
      "epoch": 0.4262778505897772,
      "grad_norm": 0.14426353573799133,
      "learning_rate": 0.0001,
      "loss": 1.3038,
      "step": 2602
    },
    {
      "epoch": 0.42644167758846657,
      "grad_norm": 0.15028896927833557,
      "learning_rate": 0.0001,
      "loss": 1.4349,
      "step": 2603
    },
    {
      "epoch": 0.42660550458715596,
      "grad_norm": 0.15098179876804352,
      "learning_rate": 0.0001,
      "loss": 1.4739,
      "step": 2604
    },
    {
      "epoch": 0.42676933158584535,
      "grad_norm": 0.15313886106014252,
      "learning_rate": 0.0001,
      "loss": 1.3485,
      "step": 2605
    },
    {
      "epoch": 0.42693315858453473,
      "grad_norm": 0.1447848081588745,
      "learning_rate": 0.0001,
      "loss": 1.3158,
      "step": 2606
    },
    {
      "epoch": 0.4270969855832241,
      "grad_norm": 0.14611642062664032,
      "learning_rate": 0.0001,
      "loss": 1.4164,
      "step": 2607
    },
    {
      "epoch": 0.4272608125819135,
      "grad_norm": 0.16929404437541962,
      "learning_rate": 0.0001,
      "loss": 1.4181,
      "step": 2608
    },
    {
      "epoch": 0.4274246395806029,
      "grad_norm": 0.14770162105560303,
      "learning_rate": 0.0001,
      "loss": 1.4113,
      "step": 2609
    },
    {
      "epoch": 0.4275884665792923,
      "grad_norm": 0.14940275251865387,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 2610
    },
    {
      "epoch": 0.42775229357798167,
      "grad_norm": 0.14538142085075378,
      "learning_rate": 0.0001,
      "loss": 1.4468,
      "step": 2611
    },
    {
      "epoch": 0.42791612057667106,
      "grad_norm": 0.14893975853919983,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 2612
    },
    {
      "epoch": 0.42807994757536044,
      "grad_norm": 0.15316927433013916,
      "learning_rate": 0.0001,
      "loss": 1.457,
      "step": 2613
    },
    {
      "epoch": 0.42824377457404983,
      "grad_norm": 0.15047654509544373,
      "learning_rate": 0.0001,
      "loss": 1.4183,
      "step": 2614
    },
    {
      "epoch": 0.42840760157273916,
      "grad_norm": 0.1527281403541565,
      "learning_rate": 0.0001,
      "loss": 1.4508,
      "step": 2615
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.14294451475143433,
      "learning_rate": 0.0001,
      "loss": 1.4126,
      "step": 2616
    },
    {
      "epoch": 0.42873525557011793,
      "grad_norm": 0.149276465177536,
      "learning_rate": 0.0001,
      "loss": 1.3145,
      "step": 2617
    },
    {
      "epoch": 0.4288990825688073,
      "grad_norm": 0.15978938341140747,
      "learning_rate": 0.0001,
      "loss": 1.4927,
      "step": 2618
    },
    {
      "epoch": 0.4290629095674967,
      "grad_norm": 0.1468295454978943,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 2619
    },
    {
      "epoch": 0.4292267365661861,
      "grad_norm": 0.1527453362941742,
      "learning_rate": 0.0001,
      "loss": 1.416,
      "step": 2620
    },
    {
      "epoch": 0.4293905635648755,
      "grad_norm": 0.15051667392253876,
      "learning_rate": 0.0001,
      "loss": 1.4325,
      "step": 2621
    },
    {
      "epoch": 0.42955439056356487,
      "grad_norm": 0.1456899642944336,
      "learning_rate": 0.0001,
      "loss": 1.3556,
      "step": 2622
    },
    {
      "epoch": 0.42971821756225426,
      "grad_norm": 0.15149158239364624,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 2623
    },
    {
      "epoch": 0.42988204456094364,
      "grad_norm": 0.15178030729293823,
      "learning_rate": 0.0001,
      "loss": 1.4146,
      "step": 2624
    },
    {
      "epoch": 0.43004587155963303,
      "grad_norm": 0.15012226998806,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 2625
    },
    {
      "epoch": 0.4302096985583224,
      "grad_norm": 0.1515461951494217,
      "learning_rate": 0.0001,
      "loss": 1.3706,
      "step": 2626
    },
    {
      "epoch": 0.4303735255570118,
      "grad_norm": 0.18924525380134583,
      "learning_rate": 0.0001,
      "loss": 1.5713,
      "step": 2627
    },
    {
      "epoch": 0.4305373525557012,
      "grad_norm": 0.17960534989833832,
      "learning_rate": 0.0001,
      "loss": 1.3471,
      "step": 2628
    },
    {
      "epoch": 0.4307011795543906,
      "grad_norm": 0.1479925811290741,
      "learning_rate": 0.0001,
      "loss": 1.4095,
      "step": 2629
    },
    {
      "epoch": 0.43086500655307997,
      "grad_norm": 0.14526301622390747,
      "learning_rate": 0.0001,
      "loss": 1.3929,
      "step": 2630
    },
    {
      "epoch": 0.43102883355176935,
      "grad_norm": 0.1389060914516449,
      "learning_rate": 0.0001,
      "loss": 1.4002,
      "step": 2631
    },
    {
      "epoch": 0.43119266055045874,
      "grad_norm": 0.15334342420101166,
      "learning_rate": 0.0001,
      "loss": 1.4001,
      "step": 2632
    },
    {
      "epoch": 0.4313564875491481,
      "grad_norm": 0.16405658423900604,
      "learning_rate": 0.0001,
      "loss": 1.3667,
      "step": 2633
    },
    {
      "epoch": 0.43152031454783746,
      "grad_norm": 0.15125282108783722,
      "learning_rate": 0.0001,
      "loss": 1.4415,
      "step": 2634
    },
    {
      "epoch": 0.43168414154652685,
      "grad_norm": 0.14406701922416687,
      "learning_rate": 0.0001,
      "loss": 1.3981,
      "step": 2635
    },
    {
      "epoch": 0.43184796854521623,
      "grad_norm": 0.15437628328800201,
      "learning_rate": 0.0001,
      "loss": 1.43,
      "step": 2636
    },
    {
      "epoch": 0.4320117955439056,
      "grad_norm": 0.15079239010810852,
      "learning_rate": 0.0001,
      "loss": 1.407,
      "step": 2637
    },
    {
      "epoch": 0.432175622542595,
      "grad_norm": 0.1522735357284546,
      "learning_rate": 0.0001,
      "loss": 1.3563,
      "step": 2638
    },
    {
      "epoch": 0.4323394495412844,
      "grad_norm": 0.16190075874328613,
      "learning_rate": 0.0001,
      "loss": 1.4007,
      "step": 2639
    },
    {
      "epoch": 0.4325032765399738,
      "grad_norm": 0.1493666023015976,
      "learning_rate": 0.0001,
      "loss": 1.4456,
      "step": 2640
    },
    {
      "epoch": 0.43266710353866317,
      "grad_norm": 0.14663782715797424,
      "learning_rate": 0.0001,
      "loss": 1.4122,
      "step": 2641
    },
    {
      "epoch": 0.43283093053735255,
      "grad_norm": 0.14942991733551025,
      "learning_rate": 0.0001,
      "loss": 1.3935,
      "step": 2642
    },
    {
      "epoch": 0.43299475753604194,
      "grad_norm": 0.1567724198102951,
      "learning_rate": 0.0001,
      "loss": 1.3689,
      "step": 2643
    },
    {
      "epoch": 0.43315858453473133,
      "grad_norm": 0.14458702504634857,
      "learning_rate": 0.0001,
      "loss": 1.3995,
      "step": 2644
    },
    {
      "epoch": 0.4333224115334207,
      "grad_norm": 0.14349272847175598,
      "learning_rate": 0.0001,
      "loss": 1.292,
      "step": 2645
    },
    {
      "epoch": 0.4334862385321101,
      "grad_norm": 0.1508297324180603,
      "learning_rate": 0.0001,
      "loss": 1.3975,
      "step": 2646
    },
    {
      "epoch": 0.4336500655307995,
      "grad_norm": 0.13707351684570312,
      "learning_rate": 0.0001,
      "loss": 1.2788,
      "step": 2647
    },
    {
      "epoch": 0.4338138925294889,
      "grad_norm": 0.1515365093946457,
      "learning_rate": 0.0001,
      "loss": 1.4848,
      "step": 2648
    },
    {
      "epoch": 0.43397771952817826,
      "grad_norm": 0.14938156306743622,
      "learning_rate": 0.0001,
      "loss": 1.3838,
      "step": 2649
    },
    {
      "epoch": 0.43414154652686765,
      "grad_norm": 0.14953160285949707,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 2650
    },
    {
      "epoch": 0.43430537352555704,
      "grad_norm": 0.15143249928951263,
      "learning_rate": 0.0001,
      "loss": 1.4044,
      "step": 2651
    },
    {
      "epoch": 0.43446920052424637,
      "grad_norm": 0.15271644294261932,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 2652
    },
    {
      "epoch": 0.43463302752293576,
      "grad_norm": 0.143209770321846,
      "learning_rate": 0.0001,
      "loss": 1.4177,
      "step": 2653
    },
    {
      "epoch": 0.43479685452162514,
      "grad_norm": 0.145667165517807,
      "learning_rate": 0.0001,
      "loss": 1.4036,
      "step": 2654
    },
    {
      "epoch": 0.43496068152031453,
      "grad_norm": 0.15391546487808228,
      "learning_rate": 0.0001,
      "loss": 1.3794,
      "step": 2655
    },
    {
      "epoch": 0.4351245085190039,
      "grad_norm": 0.15667040646076202,
      "learning_rate": 0.0001,
      "loss": 1.3773,
      "step": 2656
    },
    {
      "epoch": 0.4352883355176933,
      "grad_norm": 0.15473094582557678,
      "learning_rate": 0.0001,
      "loss": 1.3259,
      "step": 2657
    },
    {
      "epoch": 0.4354521625163827,
      "grad_norm": 0.15375083684921265,
      "learning_rate": 0.0001,
      "loss": 1.5131,
      "step": 2658
    },
    {
      "epoch": 0.4356159895150721,
      "grad_norm": 0.15488772094249725,
      "learning_rate": 0.0001,
      "loss": 1.3896,
      "step": 2659
    },
    {
      "epoch": 0.43577981651376146,
      "grad_norm": 0.1565067172050476,
      "learning_rate": 0.0001,
      "loss": 1.4367,
      "step": 2660
    },
    {
      "epoch": 0.43594364351245085,
      "grad_norm": 0.14635507762432098,
      "learning_rate": 0.0001,
      "loss": 1.314,
      "step": 2661
    },
    {
      "epoch": 0.43610747051114024,
      "grad_norm": 0.15672610700130463,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 2662
    },
    {
      "epoch": 0.4362712975098296,
      "grad_norm": 0.14590367674827576,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 2663
    },
    {
      "epoch": 0.436435124508519,
      "grad_norm": 0.15392853319644928,
      "learning_rate": 0.0001,
      "loss": 1.4462,
      "step": 2664
    },
    {
      "epoch": 0.4365989515072084,
      "grad_norm": 0.14886927604675293,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 2665
    },
    {
      "epoch": 0.4367627785058978,
      "grad_norm": 0.15773288905620575,
      "learning_rate": 0.0001,
      "loss": 1.3794,
      "step": 2666
    },
    {
      "epoch": 0.4369266055045872,
      "grad_norm": 0.15171948075294495,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 2667
    },
    {
      "epoch": 0.43709043250327656,
      "grad_norm": 0.15263362228870392,
      "learning_rate": 0.0001,
      "loss": 1.414,
      "step": 2668
    },
    {
      "epoch": 0.43725425950196595,
      "grad_norm": 0.1433008462190628,
      "learning_rate": 0.0001,
      "loss": 1.4169,
      "step": 2669
    },
    {
      "epoch": 0.43741808650065533,
      "grad_norm": 0.14350047707557678,
      "learning_rate": 0.0001,
      "loss": 1.3536,
      "step": 2670
    },
    {
      "epoch": 0.43758191349934467,
      "grad_norm": 0.15575262904167175,
      "learning_rate": 0.0001,
      "loss": 1.3345,
      "step": 2671
    },
    {
      "epoch": 0.43774574049803405,
      "grad_norm": 0.1586143523454666,
      "learning_rate": 0.0001,
      "loss": 1.4584,
      "step": 2672
    },
    {
      "epoch": 0.43790956749672344,
      "grad_norm": 0.1563752442598343,
      "learning_rate": 0.0001,
      "loss": 1.2566,
      "step": 2673
    },
    {
      "epoch": 0.4380733944954128,
      "grad_norm": 0.15361732244491577,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 2674
    },
    {
      "epoch": 0.4382372214941022,
      "grad_norm": 0.15846513211727142,
      "learning_rate": 0.0001,
      "loss": 1.431,
      "step": 2675
    },
    {
      "epoch": 0.4384010484927916,
      "grad_norm": 0.15070180594921112,
      "learning_rate": 0.0001,
      "loss": 1.367,
      "step": 2676
    },
    {
      "epoch": 0.438564875491481,
      "grad_norm": 0.14872772991657257,
      "learning_rate": 0.0001,
      "loss": 1.3298,
      "step": 2677
    },
    {
      "epoch": 0.4387287024901704,
      "grad_norm": 0.15224657952785492,
      "learning_rate": 0.0001,
      "loss": 1.441,
      "step": 2678
    },
    {
      "epoch": 0.43889252948885976,
      "grad_norm": 0.1700867861509323,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 2679
    },
    {
      "epoch": 0.43905635648754915,
      "grad_norm": 0.1594332605600357,
      "learning_rate": 0.0001,
      "loss": 1.3272,
      "step": 2680
    },
    {
      "epoch": 0.43922018348623854,
      "grad_norm": 0.16196642816066742,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 2681
    },
    {
      "epoch": 0.4393840104849279,
      "grad_norm": 0.15482528507709503,
      "learning_rate": 0.0001,
      "loss": 1.4392,
      "step": 2682
    },
    {
      "epoch": 0.4395478374836173,
      "grad_norm": 0.15452559292316437,
      "learning_rate": 0.0001,
      "loss": 1.3814,
      "step": 2683
    },
    {
      "epoch": 0.4397116644823067,
      "grad_norm": 0.14549362659454346,
      "learning_rate": 0.0001,
      "loss": 1.3832,
      "step": 2684
    },
    {
      "epoch": 0.4398754914809961,
      "grad_norm": 0.14843954145908356,
      "learning_rate": 0.0001,
      "loss": 1.4329,
      "step": 2685
    },
    {
      "epoch": 0.44003931847968547,
      "grad_norm": 0.15530537068843842,
      "learning_rate": 0.0001,
      "loss": 1.4741,
      "step": 2686
    },
    {
      "epoch": 0.44020314547837486,
      "grad_norm": 0.15287451446056366,
      "learning_rate": 0.0001,
      "loss": 1.4144,
      "step": 2687
    },
    {
      "epoch": 0.44036697247706424,
      "grad_norm": 0.15381677448749542,
      "learning_rate": 0.0001,
      "loss": 1.4009,
      "step": 2688
    },
    {
      "epoch": 0.44053079947575363,
      "grad_norm": 0.1467972844839096,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 2689
    },
    {
      "epoch": 0.44069462647444296,
      "grad_norm": 0.15593591332435608,
      "learning_rate": 0.0001,
      "loss": 1.4009,
      "step": 2690
    },
    {
      "epoch": 0.44085845347313235,
      "grad_norm": 0.15335127711296082,
      "learning_rate": 0.0001,
      "loss": 1.3754,
      "step": 2691
    },
    {
      "epoch": 0.44102228047182174,
      "grad_norm": 0.1419028341770172,
      "learning_rate": 0.0001,
      "loss": 1.3701,
      "step": 2692
    },
    {
      "epoch": 0.4411861074705111,
      "grad_norm": 0.15178701281547546,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 2693
    },
    {
      "epoch": 0.4413499344692005,
      "grad_norm": 0.14611297845840454,
      "learning_rate": 0.0001,
      "loss": 1.4036,
      "step": 2694
    },
    {
      "epoch": 0.4415137614678899,
      "grad_norm": 0.1434759646654129,
      "learning_rate": 0.0001,
      "loss": 1.2741,
      "step": 2695
    },
    {
      "epoch": 0.4416775884665793,
      "grad_norm": 0.14939868450164795,
      "learning_rate": 0.0001,
      "loss": 1.3631,
      "step": 2696
    },
    {
      "epoch": 0.44184141546526867,
      "grad_norm": 0.1499973088502884,
      "learning_rate": 0.0001,
      "loss": 1.3661,
      "step": 2697
    },
    {
      "epoch": 0.44200524246395806,
      "grad_norm": 0.1476907581090927,
      "learning_rate": 0.0001,
      "loss": 1.3677,
      "step": 2698
    },
    {
      "epoch": 0.44216906946264745,
      "grad_norm": 0.15661941468715668,
      "learning_rate": 0.0001,
      "loss": 1.409,
      "step": 2699
    },
    {
      "epoch": 0.44233289646133683,
      "grad_norm": 0.14343388378620148,
      "learning_rate": 0.0001,
      "loss": 1.3594,
      "step": 2700
    },
    {
      "epoch": 0.4424967234600262,
      "grad_norm": 0.14405149221420288,
      "learning_rate": 0.0001,
      "loss": 1.4085,
      "step": 2701
    },
    {
      "epoch": 0.4426605504587156,
      "grad_norm": 0.15001584589481354,
      "learning_rate": 0.0001,
      "loss": 1.3473,
      "step": 2702
    },
    {
      "epoch": 0.442824377457405,
      "grad_norm": 0.14773418009281158,
      "learning_rate": 0.0001,
      "loss": 1.3564,
      "step": 2703
    },
    {
      "epoch": 0.4429882044560944,
      "grad_norm": 0.14525334537029266,
      "learning_rate": 0.0001,
      "loss": 1.3538,
      "step": 2704
    },
    {
      "epoch": 0.44315203145478377,
      "grad_norm": 0.15260881185531616,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 2705
    },
    {
      "epoch": 0.44331585845347315,
      "grad_norm": 0.14584553241729736,
      "learning_rate": 0.0001,
      "loss": 1.3353,
      "step": 2706
    },
    {
      "epoch": 0.44347968545216254,
      "grad_norm": 0.17194898426532745,
      "learning_rate": 0.0001,
      "loss": 1.444,
      "step": 2707
    },
    {
      "epoch": 0.4436435124508519,
      "grad_norm": 0.15776653587818146,
      "learning_rate": 0.0001,
      "loss": 1.5321,
      "step": 2708
    },
    {
      "epoch": 0.44380733944954126,
      "grad_norm": 0.1515415906906128,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 2709
    },
    {
      "epoch": 0.44397116644823065,
      "grad_norm": 0.15061801671981812,
      "learning_rate": 0.0001,
      "loss": 1.3404,
      "step": 2710
    },
    {
      "epoch": 0.44413499344692003,
      "grad_norm": 0.14227119088172913,
      "learning_rate": 0.0001,
      "loss": 1.3858,
      "step": 2711
    },
    {
      "epoch": 0.4442988204456094,
      "grad_norm": 0.1537904590368271,
      "learning_rate": 0.0001,
      "loss": 1.3895,
      "step": 2712
    },
    {
      "epoch": 0.4444626474442988,
      "grad_norm": 0.1429193615913391,
      "learning_rate": 0.0001,
      "loss": 1.3166,
      "step": 2713
    },
    {
      "epoch": 0.4446264744429882,
      "grad_norm": 0.15175671875476837,
      "learning_rate": 0.0001,
      "loss": 1.4187,
      "step": 2714
    },
    {
      "epoch": 0.4447903014416776,
      "grad_norm": 0.15123675763607025,
      "learning_rate": 0.0001,
      "loss": 1.3556,
      "step": 2715
    },
    {
      "epoch": 0.44495412844036697,
      "grad_norm": 0.1667240858078003,
      "learning_rate": 0.0001,
      "loss": 1.4685,
      "step": 2716
    },
    {
      "epoch": 0.44511795543905636,
      "grad_norm": 0.15473119914531708,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 2717
    },
    {
      "epoch": 0.44528178243774574,
      "grad_norm": 0.1497371941804886,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 2718
    },
    {
      "epoch": 0.44544560943643513,
      "grad_norm": 0.1500352919101715,
      "learning_rate": 0.0001,
      "loss": 1.4981,
      "step": 2719
    },
    {
      "epoch": 0.4456094364351245,
      "grad_norm": 0.15625731647014618,
      "learning_rate": 0.0001,
      "loss": 1.3097,
      "step": 2720
    },
    {
      "epoch": 0.4457732634338139,
      "grad_norm": 0.15550707280635834,
      "learning_rate": 0.0001,
      "loss": 1.4223,
      "step": 2721
    },
    {
      "epoch": 0.4459370904325033,
      "grad_norm": 0.15178601443767548,
      "learning_rate": 0.0001,
      "loss": 1.5307,
      "step": 2722
    },
    {
      "epoch": 0.4461009174311927,
      "grad_norm": 0.14721369743347168,
      "learning_rate": 0.0001,
      "loss": 1.3454,
      "step": 2723
    },
    {
      "epoch": 0.44626474442988207,
      "grad_norm": 0.1522952765226364,
      "learning_rate": 0.0001,
      "loss": 1.4002,
      "step": 2724
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 0.16285108029842377,
      "learning_rate": 0.0001,
      "loss": 1.3832,
      "step": 2725
    },
    {
      "epoch": 0.44659239842726084,
      "grad_norm": 0.15608865022659302,
      "learning_rate": 0.0001,
      "loss": 1.374,
      "step": 2726
    },
    {
      "epoch": 0.44675622542595017,
      "grad_norm": 0.14484867453575134,
      "learning_rate": 0.0001,
      "loss": 1.4303,
      "step": 2727
    },
    {
      "epoch": 0.44692005242463956,
      "grad_norm": 0.15807731449604034,
      "learning_rate": 0.0001,
      "loss": 1.3619,
      "step": 2728
    },
    {
      "epoch": 0.44708387942332894,
      "grad_norm": 0.14891058206558228,
      "learning_rate": 0.0001,
      "loss": 1.3483,
      "step": 2729
    },
    {
      "epoch": 0.44724770642201833,
      "grad_norm": 0.1417301595211029,
      "learning_rate": 0.0001,
      "loss": 1.3049,
      "step": 2730
    },
    {
      "epoch": 0.4474115334207077,
      "grad_norm": 0.14088264107704163,
      "learning_rate": 0.0001,
      "loss": 1.2869,
      "step": 2731
    },
    {
      "epoch": 0.4475753604193971,
      "grad_norm": 0.14769864082336426,
      "learning_rate": 0.0001,
      "loss": 1.4612,
      "step": 2732
    },
    {
      "epoch": 0.4477391874180865,
      "grad_norm": 0.150100976228714,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 2733
    },
    {
      "epoch": 0.4479030144167759,
      "grad_norm": 0.14317584037780762,
      "learning_rate": 0.0001,
      "loss": 1.3995,
      "step": 2734
    },
    {
      "epoch": 0.44806684141546527,
      "grad_norm": 0.1527661681175232,
      "learning_rate": 0.0001,
      "loss": 1.4195,
      "step": 2735
    },
    {
      "epoch": 0.44823066841415465,
      "grad_norm": 0.14834155142307281,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 2736
    },
    {
      "epoch": 0.44839449541284404,
      "grad_norm": 0.14483408629894257,
      "learning_rate": 0.0001,
      "loss": 1.4015,
      "step": 2737
    },
    {
      "epoch": 0.44855832241153343,
      "grad_norm": 0.14627015590667725,
      "learning_rate": 0.0001,
      "loss": 1.3949,
      "step": 2738
    },
    {
      "epoch": 0.4487221494102228,
      "grad_norm": 0.14937055110931396,
      "learning_rate": 0.0001,
      "loss": 1.3488,
      "step": 2739
    },
    {
      "epoch": 0.4488859764089122,
      "grad_norm": 0.1410013735294342,
      "learning_rate": 0.0001,
      "loss": 1.4209,
      "step": 2740
    },
    {
      "epoch": 0.4490498034076016,
      "grad_norm": 0.15267659723758698,
      "learning_rate": 0.0001,
      "loss": 1.4555,
      "step": 2741
    },
    {
      "epoch": 0.449213630406291,
      "grad_norm": 0.16109752655029297,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 2742
    },
    {
      "epoch": 0.44937745740498036,
      "grad_norm": 0.1497953236103058,
      "learning_rate": 0.0001,
      "loss": 1.419,
      "step": 2743
    },
    {
      "epoch": 0.44954128440366975,
      "grad_norm": 0.14426235854625702,
      "learning_rate": 0.0001,
      "loss": 1.439,
      "step": 2744
    },
    {
      "epoch": 0.4497051114023591,
      "grad_norm": 0.1537465900182724,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 2745
    },
    {
      "epoch": 0.44986893840104847,
      "grad_norm": 0.15299992263317108,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 2746
    },
    {
      "epoch": 0.45003276539973786,
      "grad_norm": 0.14978784322738647,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 2747
    },
    {
      "epoch": 0.45019659239842724,
      "grad_norm": 0.14273345470428467,
      "learning_rate": 0.0001,
      "loss": 1.3443,
      "step": 2748
    },
    {
      "epoch": 0.45036041939711663,
      "grad_norm": 0.14928992092609406,
      "learning_rate": 0.0001,
      "loss": 1.3824,
      "step": 2749
    },
    {
      "epoch": 0.450524246395806,
      "grad_norm": 0.14938078820705414,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 2750
    },
    {
      "epoch": 0.4506880733944954,
      "grad_norm": 0.1474013477563858,
      "learning_rate": 0.0001,
      "loss": 1.3653,
      "step": 2751
    },
    {
      "epoch": 0.4508519003931848,
      "grad_norm": 0.1551920622587204,
      "learning_rate": 0.0001,
      "loss": 1.4858,
      "step": 2752
    },
    {
      "epoch": 0.4510157273918742,
      "grad_norm": 0.143799290060997,
      "learning_rate": 0.0001,
      "loss": 1.341,
      "step": 2753
    },
    {
      "epoch": 0.45117955439056356,
      "grad_norm": 0.1717778593301773,
      "learning_rate": 0.0001,
      "loss": 1.368,
      "step": 2754
    },
    {
      "epoch": 0.45134338138925295,
      "grad_norm": 0.15061365067958832,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 2755
    },
    {
      "epoch": 0.45150720838794234,
      "grad_norm": 0.1556115448474884,
      "learning_rate": 0.0001,
      "loss": 1.4303,
      "step": 2756
    },
    {
      "epoch": 0.4516710353866317,
      "grad_norm": 0.14946947991847992,
      "learning_rate": 0.0001,
      "loss": 1.3705,
      "step": 2757
    },
    {
      "epoch": 0.4518348623853211,
      "grad_norm": 0.15048848092556,
      "learning_rate": 0.0001,
      "loss": 1.5067,
      "step": 2758
    },
    {
      "epoch": 0.4519986893840105,
      "grad_norm": 0.15582504868507385,
      "learning_rate": 0.0001,
      "loss": 1.4504,
      "step": 2759
    },
    {
      "epoch": 0.4521625163826999,
      "grad_norm": 0.1470409482717514,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 2760
    },
    {
      "epoch": 0.4523263433813893,
      "grad_norm": 0.14376655220985413,
      "learning_rate": 0.0001,
      "loss": 1.3772,
      "step": 2761
    },
    {
      "epoch": 0.45249017038007866,
      "grad_norm": 0.1441483199596405,
      "learning_rate": 0.0001,
      "loss": 1.4066,
      "step": 2762
    },
    {
      "epoch": 0.45265399737876805,
      "grad_norm": 0.15086331963539124,
      "learning_rate": 0.0001,
      "loss": 1.3982,
      "step": 2763
    },
    {
      "epoch": 0.4528178243774574,
      "grad_norm": 0.14702710509300232,
      "learning_rate": 0.0001,
      "loss": 1.4137,
      "step": 2764
    },
    {
      "epoch": 0.45298165137614677,
      "grad_norm": 0.15796594321727753,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 2765
    },
    {
      "epoch": 0.45314547837483615,
      "grad_norm": 0.14866989850997925,
      "learning_rate": 0.0001,
      "loss": 1.3781,
      "step": 2766
    },
    {
      "epoch": 0.45330930537352554,
      "grad_norm": 0.14238415658473969,
      "learning_rate": 0.0001,
      "loss": 1.3927,
      "step": 2767
    },
    {
      "epoch": 0.4534731323722149,
      "grad_norm": 0.14579299092292786,
      "learning_rate": 0.0001,
      "loss": 1.3894,
      "step": 2768
    },
    {
      "epoch": 0.4536369593709043,
      "grad_norm": 0.15196356177330017,
      "learning_rate": 0.0001,
      "loss": 1.3214,
      "step": 2769
    },
    {
      "epoch": 0.4538007863695937,
      "grad_norm": 0.1558418720960617,
      "learning_rate": 0.0001,
      "loss": 1.494,
      "step": 2770
    },
    {
      "epoch": 0.4539646133682831,
      "grad_norm": 0.1478327065706253,
      "learning_rate": 0.0001,
      "loss": 1.4032,
      "step": 2771
    },
    {
      "epoch": 0.4541284403669725,
      "grad_norm": 0.1463167816400528,
      "learning_rate": 0.0001,
      "loss": 1.279,
      "step": 2772
    },
    {
      "epoch": 0.45429226736566186,
      "grad_norm": 0.15193568170070648,
      "learning_rate": 0.0001,
      "loss": 1.4206,
      "step": 2773
    },
    {
      "epoch": 0.45445609436435125,
      "grad_norm": 0.15836988389492035,
      "learning_rate": 0.0001,
      "loss": 1.434,
      "step": 2774
    },
    {
      "epoch": 0.45461992136304064,
      "grad_norm": 0.16741791367530823,
      "learning_rate": 0.0001,
      "loss": 1.3779,
      "step": 2775
    },
    {
      "epoch": 0.45478374836173,
      "grad_norm": 0.157470703125,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 2776
    },
    {
      "epoch": 0.4549475753604194,
      "grad_norm": 0.14494067430496216,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 2777
    },
    {
      "epoch": 0.4551114023591088,
      "grad_norm": 0.15011489391326904,
      "learning_rate": 0.0001,
      "loss": 1.5079,
      "step": 2778
    },
    {
      "epoch": 0.4552752293577982,
      "grad_norm": 0.1589943766593933,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 2779
    },
    {
      "epoch": 0.45543905635648757,
      "grad_norm": 0.15203316509723663,
      "learning_rate": 0.0001,
      "loss": 1.4951,
      "step": 2780
    },
    {
      "epoch": 0.45560288335517696,
      "grad_norm": 0.14968493580818176,
      "learning_rate": 0.0001,
      "loss": 1.3105,
      "step": 2781
    },
    {
      "epoch": 0.45576671035386634,
      "grad_norm": 0.14979058504104614,
      "learning_rate": 0.0001,
      "loss": 1.3821,
      "step": 2782
    },
    {
      "epoch": 0.4559305373525557,
      "grad_norm": 0.1454300433397293,
      "learning_rate": 0.0001,
      "loss": 1.332,
      "step": 2783
    },
    {
      "epoch": 0.45609436435124506,
      "grad_norm": 0.1549302488565445,
      "learning_rate": 0.0001,
      "loss": 1.4843,
      "step": 2784
    },
    {
      "epoch": 0.45625819134993445,
      "grad_norm": 0.14488856494426727,
      "learning_rate": 0.0001,
      "loss": 1.3755,
      "step": 2785
    },
    {
      "epoch": 0.45642201834862384,
      "grad_norm": 0.14653430879116058,
      "learning_rate": 0.0001,
      "loss": 1.3747,
      "step": 2786
    },
    {
      "epoch": 0.4565858453473132,
      "grad_norm": 0.15173301100730896,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 2787
    },
    {
      "epoch": 0.4567496723460026,
      "grad_norm": 0.15079468488693237,
      "learning_rate": 0.0001,
      "loss": 1.4096,
      "step": 2788
    },
    {
      "epoch": 0.456913499344692,
      "grad_norm": 0.1516498178243637,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 2789
    },
    {
      "epoch": 0.4570773263433814,
      "grad_norm": 0.1605330854654312,
      "learning_rate": 0.0001,
      "loss": 1.4271,
      "step": 2790
    },
    {
      "epoch": 0.45724115334207077,
      "grad_norm": 0.15372034907341003,
      "learning_rate": 0.0001,
      "loss": 1.4625,
      "step": 2791
    },
    {
      "epoch": 0.45740498034076016,
      "grad_norm": 0.15202003717422485,
      "learning_rate": 0.0001,
      "loss": 1.3462,
      "step": 2792
    },
    {
      "epoch": 0.45756880733944955,
      "grad_norm": 0.150900736451149,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 2793
    },
    {
      "epoch": 0.45773263433813893,
      "grad_norm": 0.15131469070911407,
      "learning_rate": 0.0001,
      "loss": 1.3727,
      "step": 2794
    },
    {
      "epoch": 0.4578964613368283,
      "grad_norm": 0.15648141503334045,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 2795
    },
    {
      "epoch": 0.4580602883355177,
      "grad_norm": 0.16622097790241241,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 2796
    },
    {
      "epoch": 0.4582241153342071,
      "grad_norm": 0.1510777324438095,
      "learning_rate": 0.0001,
      "loss": 1.4368,
      "step": 2797
    },
    {
      "epoch": 0.4583879423328965,
      "grad_norm": 0.16560021042823792,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 2798
    },
    {
      "epoch": 0.45855176933158587,
      "grad_norm": 0.15049675107002258,
      "learning_rate": 0.0001,
      "loss": 1.4043,
      "step": 2799
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 0.15842902660369873,
      "learning_rate": 0.0001,
      "loss": 1.4936,
      "step": 2800
    },
    {
      "epoch": 0.4588794233289646,
      "grad_norm": 0.1640348583459854,
      "learning_rate": 0.0001,
      "loss": 1.4595,
      "step": 2801
    },
    {
      "epoch": 0.459043250327654,
      "grad_norm": 0.14995834231376648,
      "learning_rate": 0.0001,
      "loss": 1.3605,
      "step": 2802
    },
    {
      "epoch": 0.45920707732634336,
      "grad_norm": 0.1430896669626236,
      "learning_rate": 0.0001,
      "loss": 1.3898,
      "step": 2803
    },
    {
      "epoch": 0.45937090432503275,
      "grad_norm": 0.14878322184085846,
      "learning_rate": 0.0001,
      "loss": 1.3865,
      "step": 2804
    },
    {
      "epoch": 0.45953473132372213,
      "grad_norm": 0.15232087671756744,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 2805
    },
    {
      "epoch": 0.4596985583224115,
      "grad_norm": 0.17467397451400757,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 2806
    },
    {
      "epoch": 0.4598623853211009,
      "grad_norm": 0.16983136534690857,
      "learning_rate": 0.0001,
      "loss": 1.4745,
      "step": 2807
    },
    {
      "epoch": 0.4600262123197903,
      "grad_norm": 0.151775062084198,
      "learning_rate": 0.0001,
      "loss": 1.4257,
      "step": 2808
    },
    {
      "epoch": 0.4601900393184797,
      "grad_norm": 0.1479480266571045,
      "learning_rate": 0.0001,
      "loss": 1.4287,
      "step": 2809
    },
    {
      "epoch": 0.46035386631716907,
      "grad_norm": 0.1575033813714981,
      "learning_rate": 0.0001,
      "loss": 1.4388,
      "step": 2810
    },
    {
      "epoch": 0.46051769331585846,
      "grad_norm": 0.14360104501247406,
      "learning_rate": 0.0001,
      "loss": 1.2587,
      "step": 2811
    },
    {
      "epoch": 0.46068152031454784,
      "grad_norm": 0.15519778430461884,
      "learning_rate": 0.0001,
      "loss": 1.362,
      "step": 2812
    },
    {
      "epoch": 0.46084534731323723,
      "grad_norm": 0.15464933216571808,
      "learning_rate": 0.0001,
      "loss": 1.447,
      "step": 2813
    },
    {
      "epoch": 0.4610091743119266,
      "grad_norm": 0.1479187309741974,
      "learning_rate": 0.0001,
      "loss": 1.4651,
      "step": 2814
    },
    {
      "epoch": 0.461173001310616,
      "grad_norm": 0.1518986076116562,
      "learning_rate": 0.0001,
      "loss": 1.3693,
      "step": 2815
    },
    {
      "epoch": 0.4613368283093054,
      "grad_norm": 0.14522962272167206,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 2816
    },
    {
      "epoch": 0.4615006553079948,
      "grad_norm": 0.16142447292804718,
      "learning_rate": 0.0001,
      "loss": 1.4081,
      "step": 2817
    },
    {
      "epoch": 0.46166448230668417,
      "grad_norm": 0.14992940425872803,
      "learning_rate": 0.0001,
      "loss": 1.4566,
      "step": 2818
    },
    {
      "epoch": 0.46182830930537355,
      "grad_norm": 0.170796737074852,
      "learning_rate": 0.0001,
      "loss": 1.3486,
      "step": 2819
    },
    {
      "epoch": 0.4619921363040629,
      "grad_norm": 0.14631879329681396,
      "learning_rate": 0.0001,
      "loss": 1.2865,
      "step": 2820
    },
    {
      "epoch": 0.46215596330275227,
      "grad_norm": 0.15747059881687164,
      "learning_rate": 0.0001,
      "loss": 1.2795,
      "step": 2821
    },
    {
      "epoch": 0.46231979030144166,
      "grad_norm": 0.14774009585380554,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 2822
    },
    {
      "epoch": 0.46248361730013104,
      "grad_norm": 0.15515677630901337,
      "learning_rate": 0.0001,
      "loss": 1.4361,
      "step": 2823
    },
    {
      "epoch": 0.46264744429882043,
      "grad_norm": 0.15783177316188812,
      "learning_rate": 0.0001,
      "loss": 1.3847,
      "step": 2824
    },
    {
      "epoch": 0.4628112712975098,
      "grad_norm": 0.14745734632015228,
      "learning_rate": 0.0001,
      "loss": 1.3345,
      "step": 2825
    },
    {
      "epoch": 0.4629750982961992,
      "grad_norm": 0.16706588864326477,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 2826
    },
    {
      "epoch": 0.4631389252948886,
      "grad_norm": 0.1541149616241455,
      "learning_rate": 0.0001,
      "loss": 1.4286,
      "step": 2827
    },
    {
      "epoch": 0.463302752293578,
      "grad_norm": 0.14662304520606995,
      "learning_rate": 0.0001,
      "loss": 1.3571,
      "step": 2828
    },
    {
      "epoch": 0.46346657929226737,
      "grad_norm": 0.15111735463142395,
      "learning_rate": 0.0001,
      "loss": 1.4417,
      "step": 2829
    },
    {
      "epoch": 0.46363040629095675,
      "grad_norm": 0.1472122222185135,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 2830
    },
    {
      "epoch": 0.46379423328964614,
      "grad_norm": 0.15674792230129242,
      "learning_rate": 0.0001,
      "loss": 1.3245,
      "step": 2831
    },
    {
      "epoch": 0.4639580602883355,
      "grad_norm": 0.15326640009880066,
      "learning_rate": 0.0001,
      "loss": 1.3764,
      "step": 2832
    },
    {
      "epoch": 0.4641218872870249,
      "grad_norm": 0.14782960712909698,
      "learning_rate": 0.0001,
      "loss": 1.4016,
      "step": 2833
    },
    {
      "epoch": 0.4642857142857143,
      "grad_norm": 0.15786212682724,
      "learning_rate": 0.0001,
      "loss": 1.4245,
      "step": 2834
    },
    {
      "epoch": 0.4644495412844037,
      "grad_norm": 0.1590116322040558,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 2835
    },
    {
      "epoch": 0.4646133682830931,
      "grad_norm": 0.14608407020568848,
      "learning_rate": 0.0001,
      "loss": 1.3666,
      "step": 2836
    },
    {
      "epoch": 0.46477719528178246,
      "grad_norm": 0.16565439105033875,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 2837
    },
    {
      "epoch": 0.4649410222804718,
      "grad_norm": 0.153768390417099,
      "learning_rate": 0.0001,
      "loss": 1.4129,
      "step": 2838
    },
    {
      "epoch": 0.4651048492791612,
      "grad_norm": 0.1535470336675644,
      "learning_rate": 0.0001,
      "loss": 1.4631,
      "step": 2839
    },
    {
      "epoch": 0.46526867627785057,
      "grad_norm": 0.14971940219402313,
      "learning_rate": 0.0001,
      "loss": 1.2893,
      "step": 2840
    },
    {
      "epoch": 0.46543250327653996,
      "grad_norm": 0.16715340316295624,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 2841
    },
    {
      "epoch": 0.46559633027522934,
      "grad_norm": 0.14928478002548218,
      "learning_rate": 0.0001,
      "loss": 1.3664,
      "step": 2842
    },
    {
      "epoch": 0.46576015727391873,
      "grad_norm": 0.1532152146100998,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 2843
    },
    {
      "epoch": 0.4659239842726081,
      "grad_norm": 0.1546378880739212,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 2844
    },
    {
      "epoch": 0.4660878112712975,
      "grad_norm": 0.15134795010089874,
      "learning_rate": 0.0001,
      "loss": 1.4108,
      "step": 2845
    },
    {
      "epoch": 0.4662516382699869,
      "grad_norm": 0.16447478532791138,
      "learning_rate": 0.0001,
      "loss": 1.4929,
      "step": 2846
    },
    {
      "epoch": 0.4664154652686763,
      "grad_norm": 0.15888527035713196,
      "learning_rate": 0.0001,
      "loss": 1.3122,
      "step": 2847
    },
    {
      "epoch": 0.46657929226736566,
      "grad_norm": 0.16069747507572174,
      "learning_rate": 0.0001,
      "loss": 1.4395,
      "step": 2848
    },
    {
      "epoch": 0.46674311926605505,
      "grad_norm": 0.17832712829113007,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 2849
    },
    {
      "epoch": 0.46690694626474444,
      "grad_norm": 0.1650201827287674,
      "learning_rate": 0.0001,
      "loss": 1.4041,
      "step": 2850
    },
    {
      "epoch": 0.4670707732634338,
      "grad_norm": 0.15952470898628235,
      "learning_rate": 0.0001,
      "loss": 1.442,
      "step": 2851
    },
    {
      "epoch": 0.4672346002621232,
      "grad_norm": 0.1452435553073883,
      "learning_rate": 0.0001,
      "loss": 1.2689,
      "step": 2852
    },
    {
      "epoch": 0.4673984272608126,
      "grad_norm": 0.15601693093776703,
      "learning_rate": 0.0001,
      "loss": 1.4316,
      "step": 2853
    },
    {
      "epoch": 0.467562254259502,
      "grad_norm": 0.15825258195400238,
      "learning_rate": 0.0001,
      "loss": 1.4352,
      "step": 2854
    },
    {
      "epoch": 0.4677260812581914,
      "grad_norm": 0.14795643091201782,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 2855
    },
    {
      "epoch": 0.46788990825688076,
      "grad_norm": 0.15065020322799683,
      "learning_rate": 0.0001,
      "loss": 1.2905,
      "step": 2856
    },
    {
      "epoch": 0.4680537352555701,
      "grad_norm": 0.15842494368553162,
      "learning_rate": 0.0001,
      "loss": 1.4376,
      "step": 2857
    },
    {
      "epoch": 0.4682175622542595,
      "grad_norm": 0.15567688643932343,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 2858
    },
    {
      "epoch": 0.46838138925294887,
      "grad_norm": 0.15315809845924377,
      "learning_rate": 0.0001,
      "loss": 1.3179,
      "step": 2859
    },
    {
      "epoch": 0.46854521625163825,
      "grad_norm": 0.15036451816558838,
      "learning_rate": 0.0001,
      "loss": 1.3414,
      "step": 2860
    },
    {
      "epoch": 0.46870904325032764,
      "grad_norm": 0.15350577235221863,
      "learning_rate": 0.0001,
      "loss": 1.4549,
      "step": 2861
    },
    {
      "epoch": 0.468872870249017,
      "grad_norm": 0.14832402765750885,
      "learning_rate": 0.0001,
      "loss": 1.3242,
      "step": 2862
    },
    {
      "epoch": 0.4690366972477064,
      "grad_norm": 0.17031781375408173,
      "learning_rate": 0.0001,
      "loss": 1.4408,
      "step": 2863
    },
    {
      "epoch": 0.4692005242463958,
      "grad_norm": 0.16387024521827698,
      "learning_rate": 0.0001,
      "loss": 1.3916,
      "step": 2864
    },
    {
      "epoch": 0.4693643512450852,
      "grad_norm": 0.15795935690402985,
      "learning_rate": 0.0001,
      "loss": 1.3301,
      "step": 2865
    },
    {
      "epoch": 0.4695281782437746,
      "grad_norm": 0.15711253881454468,
      "learning_rate": 0.0001,
      "loss": 1.3444,
      "step": 2866
    },
    {
      "epoch": 0.46969200524246396,
      "grad_norm": 0.1490054428577423,
      "learning_rate": 0.0001,
      "loss": 1.4012,
      "step": 2867
    },
    {
      "epoch": 0.46985583224115335,
      "grad_norm": 0.15198402106761932,
      "learning_rate": 0.0001,
      "loss": 1.4456,
      "step": 2868
    },
    {
      "epoch": 0.47001965923984274,
      "grad_norm": 0.14076882600784302,
      "learning_rate": 0.0001,
      "loss": 1.2633,
      "step": 2869
    },
    {
      "epoch": 0.4701834862385321,
      "grad_norm": 0.1446600705385208,
      "learning_rate": 0.0001,
      "loss": 1.3881,
      "step": 2870
    },
    {
      "epoch": 0.4703473132372215,
      "grad_norm": 0.14590482413768768,
      "learning_rate": 0.0001,
      "loss": 1.3518,
      "step": 2871
    },
    {
      "epoch": 0.4705111402359109,
      "grad_norm": 0.15855517983436584,
      "learning_rate": 0.0001,
      "loss": 1.4368,
      "step": 2872
    },
    {
      "epoch": 0.4706749672346003,
      "grad_norm": 0.1560964733362198,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 2873
    },
    {
      "epoch": 0.47083879423328967,
      "grad_norm": 0.15739256143569946,
      "learning_rate": 0.0001,
      "loss": 1.38,
      "step": 2874
    },
    {
      "epoch": 0.47100262123197906,
      "grad_norm": 0.1442287713289261,
      "learning_rate": 0.0001,
      "loss": 1.4136,
      "step": 2875
    },
    {
      "epoch": 0.4711664482306684,
      "grad_norm": 0.1519680619239807,
      "learning_rate": 0.0001,
      "loss": 1.4412,
      "step": 2876
    },
    {
      "epoch": 0.4713302752293578,
      "grad_norm": 0.15007776021957397,
      "learning_rate": 0.0001,
      "loss": 1.4569,
      "step": 2877
    },
    {
      "epoch": 0.47149410222804716,
      "grad_norm": 0.15005183219909668,
      "learning_rate": 0.0001,
      "loss": 1.3316,
      "step": 2878
    },
    {
      "epoch": 0.47165792922673655,
      "grad_norm": 0.15564210712909698,
      "learning_rate": 0.0001,
      "loss": 1.3919,
      "step": 2879
    },
    {
      "epoch": 0.47182175622542594,
      "grad_norm": 0.17125126719474792,
      "learning_rate": 0.0001,
      "loss": 1.4242,
      "step": 2880
    },
    {
      "epoch": 0.4719855832241153,
      "grad_norm": 0.15470801293849945,
      "learning_rate": 0.0001,
      "loss": 1.3959,
      "step": 2881
    },
    {
      "epoch": 0.4721494102228047,
      "grad_norm": 0.1483890265226364,
      "learning_rate": 0.0001,
      "loss": 1.3413,
      "step": 2882
    },
    {
      "epoch": 0.4723132372214941,
      "grad_norm": 0.1552909016609192,
      "learning_rate": 0.0001,
      "loss": 1.3661,
      "step": 2883
    },
    {
      "epoch": 0.4724770642201835,
      "grad_norm": 0.1500934660434723,
      "learning_rate": 0.0001,
      "loss": 1.3182,
      "step": 2884
    },
    {
      "epoch": 0.47264089121887287,
      "grad_norm": 0.15689034759998322,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 2885
    },
    {
      "epoch": 0.47280471821756226,
      "grad_norm": 0.15194161236286163,
      "learning_rate": 0.0001,
      "loss": 1.3828,
      "step": 2886
    },
    {
      "epoch": 0.47296854521625165,
      "grad_norm": 0.15582576394081116,
      "learning_rate": 0.0001,
      "loss": 1.4565,
      "step": 2887
    },
    {
      "epoch": 0.47313237221494103,
      "grad_norm": 0.15578047931194305,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 2888
    },
    {
      "epoch": 0.4732961992136304,
      "grad_norm": 0.15130116045475006,
      "learning_rate": 0.0001,
      "loss": 1.4481,
      "step": 2889
    },
    {
      "epoch": 0.4734600262123198,
      "grad_norm": 0.15306441485881805,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 2890
    },
    {
      "epoch": 0.4736238532110092,
      "grad_norm": 0.14770367741584778,
      "learning_rate": 0.0001,
      "loss": 1.3618,
      "step": 2891
    },
    {
      "epoch": 0.4737876802096986,
      "grad_norm": 0.1660355180501938,
      "learning_rate": 0.0001,
      "loss": 1.3931,
      "step": 2892
    },
    {
      "epoch": 0.47395150720838797,
      "grad_norm": 0.16208788752555847,
      "learning_rate": 0.0001,
      "loss": 1.3963,
      "step": 2893
    },
    {
      "epoch": 0.4741153342070773,
      "grad_norm": 0.1571948379278183,
      "learning_rate": 0.0001,
      "loss": 1.3332,
      "step": 2894
    },
    {
      "epoch": 0.4742791612057667,
      "grad_norm": 0.15170185267925262,
      "learning_rate": 0.0001,
      "loss": 1.3828,
      "step": 2895
    },
    {
      "epoch": 0.4744429882044561,
      "grad_norm": 0.16049572825431824,
      "learning_rate": 0.0001,
      "loss": 1.454,
      "step": 2896
    },
    {
      "epoch": 0.47460681520314546,
      "grad_norm": 0.153058260679245,
      "learning_rate": 0.0001,
      "loss": 1.4388,
      "step": 2897
    },
    {
      "epoch": 0.47477064220183485,
      "grad_norm": 0.17114223539829254,
      "learning_rate": 0.0001,
      "loss": 1.4807,
      "step": 2898
    },
    {
      "epoch": 0.47493446920052423,
      "grad_norm": 0.20762021839618683,
      "learning_rate": 0.0001,
      "loss": 1.431,
      "step": 2899
    },
    {
      "epoch": 0.4750982961992136,
      "grad_norm": 0.16266217827796936,
      "learning_rate": 0.0001,
      "loss": 1.4817,
      "step": 2900
    },
    {
      "epoch": 0.475262123197903,
      "grad_norm": 0.15626725554466248,
      "learning_rate": 0.0001,
      "loss": 1.3664,
      "step": 2901
    },
    {
      "epoch": 0.4754259501965924,
      "grad_norm": 0.1546672135591507,
      "learning_rate": 0.0001,
      "loss": 1.3303,
      "step": 2902
    },
    {
      "epoch": 0.4755897771952818,
      "grad_norm": 0.1507069617509842,
      "learning_rate": 0.0001,
      "loss": 1.3821,
      "step": 2903
    },
    {
      "epoch": 0.47575360419397117,
      "grad_norm": 0.14807890355587006,
      "learning_rate": 0.0001,
      "loss": 1.3173,
      "step": 2904
    },
    {
      "epoch": 0.47591743119266056,
      "grad_norm": 0.14697569608688354,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 2905
    },
    {
      "epoch": 0.47608125819134994,
      "grad_norm": 0.16383841633796692,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 2906
    },
    {
      "epoch": 0.47624508519003933,
      "grad_norm": 0.15728062391281128,
      "learning_rate": 0.0001,
      "loss": 1.4725,
      "step": 2907
    },
    {
      "epoch": 0.4764089121887287,
      "grad_norm": 0.14489291608333588,
      "learning_rate": 0.0001,
      "loss": 1.3066,
      "step": 2908
    },
    {
      "epoch": 0.4765727391874181,
      "grad_norm": 0.14415235817432404,
      "learning_rate": 0.0001,
      "loss": 1.4454,
      "step": 2909
    },
    {
      "epoch": 0.4767365661861075,
      "grad_norm": 0.1404748558998108,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 2910
    },
    {
      "epoch": 0.4769003931847969,
      "grad_norm": 0.14049337804317474,
      "learning_rate": 0.0001,
      "loss": 1.2949,
      "step": 2911
    },
    {
      "epoch": 0.47706422018348627,
      "grad_norm": 0.16024267673492432,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 2912
    },
    {
      "epoch": 0.4772280471821756,
      "grad_norm": 0.1485019326210022,
      "learning_rate": 0.0001,
      "loss": 1.3752,
      "step": 2913
    },
    {
      "epoch": 0.477391874180865,
      "grad_norm": 0.1517840027809143,
      "learning_rate": 0.0001,
      "loss": 1.3493,
      "step": 2914
    },
    {
      "epoch": 0.47755570117955437,
      "grad_norm": 0.15312868356704712,
      "learning_rate": 0.0001,
      "loss": 1.3805,
      "step": 2915
    },
    {
      "epoch": 0.47771952817824376,
      "grad_norm": 0.14931176602840424,
      "learning_rate": 0.0001,
      "loss": 1.3263,
      "step": 2916
    },
    {
      "epoch": 0.47788335517693314,
      "grad_norm": 0.15898966789245605,
      "learning_rate": 0.0001,
      "loss": 1.3063,
      "step": 2917
    },
    {
      "epoch": 0.47804718217562253,
      "grad_norm": 0.16123491525650024,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 2918
    },
    {
      "epoch": 0.4782110091743119,
      "grad_norm": 0.16017596423625946,
      "learning_rate": 0.0001,
      "loss": 1.4715,
      "step": 2919
    },
    {
      "epoch": 0.4783748361730013,
      "grad_norm": 0.16287241876125336,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 2920
    },
    {
      "epoch": 0.4785386631716907,
      "grad_norm": 0.16179607808589935,
      "learning_rate": 0.0001,
      "loss": 1.4197,
      "step": 2921
    },
    {
      "epoch": 0.4787024901703801,
      "grad_norm": 0.15580348670482635,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 2922
    },
    {
      "epoch": 0.47886631716906947,
      "grad_norm": 0.15198993682861328,
      "learning_rate": 0.0001,
      "loss": 1.4036,
      "step": 2923
    },
    {
      "epoch": 0.47903014416775885,
      "grad_norm": 0.17467933893203735,
      "learning_rate": 0.0001,
      "loss": 1.2823,
      "step": 2924
    },
    {
      "epoch": 0.47919397116644824,
      "grad_norm": 0.14492981135845184,
      "learning_rate": 0.0001,
      "loss": 1.3472,
      "step": 2925
    },
    {
      "epoch": 0.4793577981651376,
      "grad_norm": 0.1489858329296112,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 2926
    },
    {
      "epoch": 0.479521625163827,
      "grad_norm": 0.15924425423145294,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 2927
    },
    {
      "epoch": 0.4796854521625164,
      "grad_norm": 0.18542274832725525,
      "learning_rate": 0.0001,
      "loss": 1.4783,
      "step": 2928
    },
    {
      "epoch": 0.4798492791612058,
      "grad_norm": 0.14781491458415985,
      "learning_rate": 0.0001,
      "loss": 1.3181,
      "step": 2929
    },
    {
      "epoch": 0.4800131061598952,
      "grad_norm": 0.14725933969020844,
      "learning_rate": 0.0001,
      "loss": 1.4326,
      "step": 2930
    },
    {
      "epoch": 0.48017693315858456,
      "grad_norm": 0.15494024753570557,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 2931
    },
    {
      "epoch": 0.4803407601572739,
      "grad_norm": 0.14421315491199493,
      "learning_rate": 0.0001,
      "loss": 1.39,
      "step": 2932
    },
    {
      "epoch": 0.4805045871559633,
      "grad_norm": 0.18007425963878632,
      "learning_rate": 0.0001,
      "loss": 1.4478,
      "step": 2933
    },
    {
      "epoch": 0.48066841415465267,
      "grad_norm": 0.15778987109661102,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 2934
    },
    {
      "epoch": 0.48083224115334205,
      "grad_norm": 0.15611118078231812,
      "learning_rate": 0.0001,
      "loss": 1.3042,
      "step": 2935
    },
    {
      "epoch": 0.48099606815203144,
      "grad_norm": 0.15939703583717346,
      "learning_rate": 0.0001,
      "loss": 1.4485,
      "step": 2936
    },
    {
      "epoch": 0.48115989515072083,
      "grad_norm": 0.1475849747657776,
      "learning_rate": 0.0001,
      "loss": 1.3405,
      "step": 2937
    },
    {
      "epoch": 0.4813237221494102,
      "grad_norm": 0.15312089025974274,
      "learning_rate": 0.0001,
      "loss": 1.4009,
      "step": 2938
    },
    {
      "epoch": 0.4814875491480996,
      "grad_norm": 0.15505310893058777,
      "learning_rate": 0.0001,
      "loss": 1.3857,
      "step": 2939
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 0.1460728496313095,
      "learning_rate": 0.0001,
      "loss": 1.249,
      "step": 2940
    },
    {
      "epoch": 0.4818152031454784,
      "grad_norm": 0.157216414809227,
      "learning_rate": 0.0001,
      "loss": 1.4477,
      "step": 2941
    },
    {
      "epoch": 0.48197903014416776,
      "grad_norm": 0.15528973937034607,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 2942
    },
    {
      "epoch": 0.48214285714285715,
      "grad_norm": 0.15298044681549072,
      "learning_rate": 0.0001,
      "loss": 1.4224,
      "step": 2943
    },
    {
      "epoch": 0.48230668414154654,
      "grad_norm": 0.1730509251356125,
      "learning_rate": 0.0001,
      "loss": 1.3557,
      "step": 2944
    },
    {
      "epoch": 0.4824705111402359,
      "grad_norm": 0.14963556826114655,
      "learning_rate": 0.0001,
      "loss": 1.3549,
      "step": 2945
    },
    {
      "epoch": 0.4826343381389253,
      "grad_norm": 0.15217813849449158,
      "learning_rate": 0.0001,
      "loss": 1.351,
      "step": 2946
    },
    {
      "epoch": 0.4827981651376147,
      "grad_norm": 0.15214675664901733,
      "learning_rate": 0.0001,
      "loss": 1.3642,
      "step": 2947
    },
    {
      "epoch": 0.4829619921363041,
      "grad_norm": 0.18419131636619568,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 2948
    },
    {
      "epoch": 0.4831258191349935,
      "grad_norm": 0.15233394503593445,
      "learning_rate": 0.0001,
      "loss": 1.3361,
      "step": 2949
    },
    {
      "epoch": 0.4832896461336828,
      "grad_norm": 0.16016237437725067,
      "learning_rate": 0.0001,
      "loss": 1.388,
      "step": 2950
    },
    {
      "epoch": 0.4834534731323722,
      "grad_norm": 0.15365566313266754,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 2951
    },
    {
      "epoch": 0.4836173001310616,
      "grad_norm": 0.1545860320329666,
      "learning_rate": 0.0001,
      "loss": 1.4081,
      "step": 2952
    },
    {
      "epoch": 0.48378112712975097,
      "grad_norm": 0.16356691718101501,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 2953
    },
    {
      "epoch": 0.48394495412844035,
      "grad_norm": 0.15595607459545135,
      "learning_rate": 0.0001,
      "loss": 1.4378,
      "step": 2954
    },
    {
      "epoch": 0.48410878112712974,
      "grad_norm": 0.15363968908786774,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 2955
    },
    {
      "epoch": 0.4842726081258191,
      "grad_norm": 0.15118688344955444,
      "learning_rate": 0.0001,
      "loss": 1.5248,
      "step": 2956
    },
    {
      "epoch": 0.4844364351245085,
      "grad_norm": 0.16183845698833466,
      "learning_rate": 0.0001,
      "loss": 1.4636,
      "step": 2957
    },
    {
      "epoch": 0.4846002621231979,
      "grad_norm": 0.15984989702701569,
      "learning_rate": 0.0001,
      "loss": 1.4392,
      "step": 2958
    },
    {
      "epoch": 0.4847640891218873,
      "grad_norm": 0.15033628046512604,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 2959
    },
    {
      "epoch": 0.4849279161205767,
      "grad_norm": 0.1563023179769516,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 2960
    },
    {
      "epoch": 0.48509174311926606,
      "grad_norm": 0.15950509905815125,
      "learning_rate": 0.0001,
      "loss": 1.4631,
      "step": 2961
    },
    {
      "epoch": 0.48525557011795545,
      "grad_norm": 0.15301191806793213,
      "learning_rate": 0.0001,
      "loss": 1.3755,
      "step": 2962
    },
    {
      "epoch": 0.48541939711664484,
      "grad_norm": 0.16213494539260864,
      "learning_rate": 0.0001,
      "loss": 1.4561,
      "step": 2963
    },
    {
      "epoch": 0.4855832241153342,
      "grad_norm": 0.1556222289800644,
      "learning_rate": 0.0001,
      "loss": 1.3706,
      "step": 2964
    },
    {
      "epoch": 0.4857470511140236,
      "grad_norm": 0.14962467551231384,
      "learning_rate": 0.0001,
      "loss": 1.3162,
      "step": 2965
    },
    {
      "epoch": 0.485910878112713,
      "grad_norm": 0.1609780341386795,
      "learning_rate": 0.0001,
      "loss": 1.4768,
      "step": 2966
    },
    {
      "epoch": 0.4860747051114024,
      "grad_norm": 0.1612870693206787,
      "learning_rate": 0.0001,
      "loss": 1.453,
      "step": 2967
    },
    {
      "epoch": 0.48623853211009177,
      "grad_norm": 0.15300489962100983,
      "learning_rate": 0.0001,
      "loss": 1.454,
      "step": 2968
    },
    {
      "epoch": 0.4864023591087811,
      "grad_norm": 0.1595587134361267,
      "learning_rate": 0.0001,
      "loss": 1.2706,
      "step": 2969
    },
    {
      "epoch": 0.4865661861074705,
      "grad_norm": 0.15943259000778198,
      "learning_rate": 0.0001,
      "loss": 1.3932,
      "step": 2970
    },
    {
      "epoch": 0.4867300131061599,
      "grad_norm": 0.15282800793647766,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 2971
    },
    {
      "epoch": 0.48689384010484926,
      "grad_norm": 0.1605437695980072,
      "learning_rate": 0.0001,
      "loss": 1.5538,
      "step": 2972
    },
    {
      "epoch": 0.48705766710353865,
      "grad_norm": 0.14568151533603668,
      "learning_rate": 0.0001,
      "loss": 1.3665,
      "step": 2973
    },
    {
      "epoch": 0.48722149410222804,
      "grad_norm": 0.14743110537528992,
      "learning_rate": 0.0001,
      "loss": 1.3748,
      "step": 2974
    },
    {
      "epoch": 0.4873853211009174,
      "grad_norm": 0.15203672647476196,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 2975
    },
    {
      "epoch": 0.4875491480996068,
      "grad_norm": 0.16308948397636414,
      "learning_rate": 0.0001,
      "loss": 1.4289,
      "step": 2976
    },
    {
      "epoch": 0.4877129750982962,
      "grad_norm": 0.15495145320892334,
      "learning_rate": 0.0001,
      "loss": 1.402,
      "step": 2977
    },
    {
      "epoch": 0.4878768020969856,
      "grad_norm": 0.15486732125282288,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 2978
    },
    {
      "epoch": 0.48804062909567497,
      "grad_norm": 0.1599281281232834,
      "learning_rate": 0.0001,
      "loss": 1.3625,
      "step": 2979
    },
    {
      "epoch": 0.48820445609436436,
      "grad_norm": 0.14305248856544495,
      "learning_rate": 0.0001,
      "loss": 1.415,
      "step": 2980
    },
    {
      "epoch": 0.48836828309305375,
      "grad_norm": 0.15756875276565552,
      "learning_rate": 0.0001,
      "loss": 1.3769,
      "step": 2981
    },
    {
      "epoch": 0.48853211009174313,
      "grad_norm": 0.159278005361557,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 2982
    },
    {
      "epoch": 0.4886959370904325,
      "grad_norm": 0.15054531395435333,
      "learning_rate": 0.0001,
      "loss": 1.2685,
      "step": 2983
    },
    {
      "epoch": 0.4888597640891219,
      "grad_norm": 0.1550297737121582,
      "learning_rate": 0.0001,
      "loss": 1.4664,
      "step": 2984
    },
    {
      "epoch": 0.4890235910878113,
      "grad_norm": 0.15383832156658173,
      "learning_rate": 0.0001,
      "loss": 1.3943,
      "step": 2985
    },
    {
      "epoch": 0.4891874180865007,
      "grad_norm": 0.14948131144046783,
      "learning_rate": 0.0001,
      "loss": 1.2789,
      "step": 2986
    },
    {
      "epoch": 0.48935124508519,
      "grad_norm": 0.15293541550636292,
      "learning_rate": 0.0001,
      "loss": 1.3327,
      "step": 2987
    },
    {
      "epoch": 0.4895150720838794,
      "grad_norm": 0.15307463705539703,
      "learning_rate": 0.0001,
      "loss": 1.446,
      "step": 2988
    },
    {
      "epoch": 0.4896788990825688,
      "grad_norm": 0.14482532441616058,
      "learning_rate": 0.0001,
      "loss": 1.3214,
      "step": 2989
    },
    {
      "epoch": 0.4898427260812582,
      "grad_norm": 0.15790291130542755,
      "learning_rate": 0.0001,
      "loss": 1.5021,
      "step": 2990
    },
    {
      "epoch": 0.49000655307994756,
      "grad_norm": 0.1617862433195114,
      "learning_rate": 0.0001,
      "loss": 1.4457,
      "step": 2991
    },
    {
      "epoch": 0.49017038007863695,
      "grad_norm": 0.15118522942066193,
      "learning_rate": 0.0001,
      "loss": 1.4665,
      "step": 2992
    },
    {
      "epoch": 0.49033420707732633,
      "grad_norm": 0.1606876403093338,
      "learning_rate": 0.0001,
      "loss": 1.2286,
      "step": 2993
    },
    {
      "epoch": 0.4904980340760157,
      "grad_norm": 0.1526545137166977,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 2994
    },
    {
      "epoch": 0.4906618610747051,
      "grad_norm": 0.14919738471508026,
      "learning_rate": 0.0001,
      "loss": 1.3943,
      "step": 2995
    },
    {
      "epoch": 0.4908256880733945,
      "grad_norm": 0.1602771282196045,
      "learning_rate": 0.0001,
      "loss": 1.4531,
      "step": 2996
    },
    {
      "epoch": 0.4909895150720839,
      "grad_norm": 0.148070827126503,
      "learning_rate": 0.0001,
      "loss": 1.3823,
      "step": 2997
    },
    {
      "epoch": 0.49115334207077327,
      "grad_norm": 0.1599067896604538,
      "learning_rate": 0.0001,
      "loss": 1.5047,
      "step": 2998
    },
    {
      "epoch": 0.49131716906946266,
      "grad_norm": 0.15539811551570892,
      "learning_rate": 0.0001,
      "loss": 1.3919,
      "step": 2999
    },
    {
      "epoch": 0.49148099606815204,
      "grad_norm": 0.14468371868133545,
      "learning_rate": 0.0001,
      "loss": 1.3104,
      "step": 3000
    },
    {
      "epoch": 0.49164482306684143,
      "grad_norm": 0.15892009437084198,
      "learning_rate": 0.0001,
      "loss": 1.3674,
      "step": 3001
    },
    {
      "epoch": 0.4918086500655308,
      "grad_norm": 0.146945983171463,
      "learning_rate": 0.0001,
      "loss": 1.3328,
      "step": 3002
    },
    {
      "epoch": 0.4919724770642202,
      "grad_norm": 0.14460937678813934,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 3003
    },
    {
      "epoch": 0.4921363040629096,
      "grad_norm": 0.15507204830646515,
      "learning_rate": 0.0001,
      "loss": 1.3677,
      "step": 3004
    },
    {
      "epoch": 0.492300131061599,
      "grad_norm": 0.16117526590824127,
      "learning_rate": 0.0001,
      "loss": 1.3726,
      "step": 3005
    },
    {
      "epoch": 0.4924639580602883,
      "grad_norm": 0.15077729523181915,
      "learning_rate": 0.0001,
      "loss": 1.3832,
      "step": 3006
    },
    {
      "epoch": 0.4926277850589777,
      "grad_norm": 0.15698429942131042,
      "learning_rate": 0.0001,
      "loss": 1.4229,
      "step": 3007
    },
    {
      "epoch": 0.4927916120576671,
      "grad_norm": 0.1528143584728241,
      "learning_rate": 0.0001,
      "loss": 1.2681,
      "step": 3008
    },
    {
      "epoch": 0.49295543905635647,
      "grad_norm": 0.1545693725347519,
      "learning_rate": 0.0001,
      "loss": 1.3459,
      "step": 3009
    },
    {
      "epoch": 0.49311926605504586,
      "grad_norm": 0.15384678542613983,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 3010
    },
    {
      "epoch": 0.49328309305373524,
      "grad_norm": 0.1542295664548874,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 3011
    },
    {
      "epoch": 0.49344692005242463,
      "grad_norm": 0.14715562760829926,
      "learning_rate": 0.0001,
      "loss": 1.235,
      "step": 3012
    },
    {
      "epoch": 0.493610747051114,
      "grad_norm": 0.15726236999034882,
      "learning_rate": 0.0001,
      "loss": 1.4275,
      "step": 3013
    },
    {
      "epoch": 0.4937745740498034,
      "grad_norm": 0.14913862943649292,
      "learning_rate": 0.0001,
      "loss": 1.3383,
      "step": 3014
    },
    {
      "epoch": 0.4939384010484928,
      "grad_norm": 0.1605687290430069,
      "learning_rate": 0.0001,
      "loss": 1.462,
      "step": 3015
    },
    {
      "epoch": 0.4941022280471822,
      "grad_norm": 0.15534794330596924,
      "learning_rate": 0.0001,
      "loss": 1.382,
      "step": 3016
    },
    {
      "epoch": 0.49426605504587157,
      "grad_norm": 0.14940853416919708,
      "learning_rate": 0.0001,
      "loss": 1.4075,
      "step": 3017
    },
    {
      "epoch": 0.49442988204456095,
      "grad_norm": 0.1539759337902069,
      "learning_rate": 0.0001,
      "loss": 1.3455,
      "step": 3018
    },
    {
      "epoch": 0.49459370904325034,
      "grad_norm": 0.15780489146709442,
      "learning_rate": 0.0001,
      "loss": 1.4531,
      "step": 3019
    },
    {
      "epoch": 0.4947575360419397,
      "grad_norm": 0.14943566918373108,
      "learning_rate": 0.0001,
      "loss": 1.3647,
      "step": 3020
    },
    {
      "epoch": 0.4949213630406291,
      "grad_norm": 0.1576208621263504,
      "learning_rate": 0.0001,
      "loss": 1.4047,
      "step": 3021
    },
    {
      "epoch": 0.4950851900393185,
      "grad_norm": 0.164854496717453,
      "learning_rate": 0.0001,
      "loss": 1.4923,
      "step": 3022
    },
    {
      "epoch": 0.4952490170380079,
      "grad_norm": 0.16228598356246948,
      "learning_rate": 0.0001,
      "loss": 1.3989,
      "step": 3023
    },
    {
      "epoch": 0.4954128440366973,
      "grad_norm": 0.1933896541595459,
      "learning_rate": 0.0001,
      "loss": 1.4294,
      "step": 3024
    },
    {
      "epoch": 0.4955766710353866,
      "grad_norm": 0.16874301433563232,
      "learning_rate": 0.0001,
      "loss": 1.4548,
      "step": 3025
    },
    {
      "epoch": 0.495740498034076,
      "grad_norm": 0.15777797996997833,
      "learning_rate": 0.0001,
      "loss": 1.4269,
      "step": 3026
    },
    {
      "epoch": 0.4959043250327654,
      "grad_norm": 0.15633895993232727,
      "learning_rate": 0.0001,
      "loss": 1.4504,
      "step": 3027
    },
    {
      "epoch": 0.49606815203145477,
      "grad_norm": 0.1542440503835678,
      "learning_rate": 0.0001,
      "loss": 1.4726,
      "step": 3028
    },
    {
      "epoch": 0.49623197903014415,
      "grad_norm": 0.154279887676239,
      "learning_rate": 0.0001,
      "loss": 1.3178,
      "step": 3029
    },
    {
      "epoch": 0.49639580602883354,
      "grad_norm": 0.16426262259483337,
      "learning_rate": 0.0001,
      "loss": 1.4263,
      "step": 3030
    },
    {
      "epoch": 0.49655963302752293,
      "grad_norm": 0.1656256765127182,
      "learning_rate": 0.0001,
      "loss": 1.3904,
      "step": 3031
    },
    {
      "epoch": 0.4967234600262123,
      "grad_norm": 0.16086459159851074,
      "learning_rate": 0.0001,
      "loss": 1.2786,
      "step": 3032
    },
    {
      "epoch": 0.4968872870249017,
      "grad_norm": 0.15522882342338562,
      "learning_rate": 0.0001,
      "loss": 1.3404,
      "step": 3033
    },
    {
      "epoch": 0.4970511140235911,
      "grad_norm": 0.16215865314006805,
      "learning_rate": 0.0001,
      "loss": 1.4217,
      "step": 3034
    },
    {
      "epoch": 0.4972149410222805,
      "grad_norm": 0.1429782211780548,
      "learning_rate": 0.0001,
      "loss": 1.3446,
      "step": 3035
    },
    {
      "epoch": 0.49737876802096986,
      "grad_norm": 0.17472237348556519,
      "learning_rate": 0.0001,
      "loss": 1.3811,
      "step": 3036
    },
    {
      "epoch": 0.49754259501965925,
      "grad_norm": 0.1539379358291626,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 3037
    },
    {
      "epoch": 0.49770642201834864,
      "grad_norm": 0.1563873291015625,
      "learning_rate": 0.0001,
      "loss": 1.4733,
      "step": 3038
    },
    {
      "epoch": 0.497870249017038,
      "grad_norm": 0.15746092796325684,
      "learning_rate": 0.0001,
      "loss": 1.3953,
      "step": 3039
    },
    {
      "epoch": 0.4980340760157274,
      "grad_norm": 0.16261902451515198,
      "learning_rate": 0.0001,
      "loss": 1.4412,
      "step": 3040
    },
    {
      "epoch": 0.4981979030144168,
      "grad_norm": 0.1575150489807129,
      "learning_rate": 0.0001,
      "loss": 1.4573,
      "step": 3041
    },
    {
      "epoch": 0.4983617300131062,
      "grad_norm": 0.15553095936775208,
      "learning_rate": 0.0001,
      "loss": 1.4598,
      "step": 3042
    },
    {
      "epoch": 0.4985255570117955,
      "grad_norm": 0.15945817530155182,
      "learning_rate": 0.0001,
      "loss": 1.2992,
      "step": 3043
    },
    {
      "epoch": 0.4986893840104849,
      "grad_norm": 0.15544143319129944,
      "learning_rate": 0.0001,
      "loss": 1.3341,
      "step": 3044
    },
    {
      "epoch": 0.4988532110091743,
      "grad_norm": 0.15558116137981415,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 3045
    },
    {
      "epoch": 0.4990170380078637,
      "grad_norm": 0.1623867154121399,
      "learning_rate": 0.0001,
      "loss": 1.3725,
      "step": 3046
    },
    {
      "epoch": 0.49918086500655307,
      "grad_norm": 0.16694751381874084,
      "learning_rate": 0.0001,
      "loss": 1.4064,
      "step": 3047
    },
    {
      "epoch": 0.49934469200524245,
      "grad_norm": 0.1566053032875061,
      "learning_rate": 0.0001,
      "loss": 1.3548,
      "step": 3048
    },
    {
      "epoch": 0.49950851900393184,
      "grad_norm": 0.15899112820625305,
      "learning_rate": 0.0001,
      "loss": 1.2453,
      "step": 3049
    },
    {
      "epoch": 0.4996723460026212,
      "grad_norm": 0.1867550015449524,
      "learning_rate": 0.0001,
      "loss": 1.4082,
      "step": 3050
    },
    {
      "epoch": 0.4998361730013106,
      "grad_norm": 0.15618321299552917,
      "learning_rate": 0.0001,
      "loss": 1.2885,
      "step": 3051
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1471213698387146,
      "learning_rate": 0.0001,
      "loss": 1.4351,
      "step": 3052
    },
    {
      "epoch": 0.5001638269986893,
      "grad_norm": 0.15953463315963745,
      "learning_rate": 0.0001,
      "loss": 1.3673,
      "step": 3053
    },
    {
      "epoch": 0.5003276539973788,
      "grad_norm": 0.16781817376613617,
      "learning_rate": 0.0001,
      "loss": 1.4162,
      "step": 3054
    },
    {
      "epoch": 0.5004914809960681,
      "grad_norm": 0.16014081239700317,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 3055
    },
    {
      "epoch": 0.5006553079947575,
      "grad_norm": 0.15358495712280273,
      "learning_rate": 0.0001,
      "loss": 1.4248,
      "step": 3056
    },
    {
      "epoch": 0.5008191349934469,
      "grad_norm": 0.15473811328411102,
      "learning_rate": 0.0001,
      "loss": 1.4155,
      "step": 3057
    },
    {
      "epoch": 0.5009829619921363,
      "grad_norm": 0.15000563859939575,
      "learning_rate": 0.0001,
      "loss": 1.2463,
      "step": 3058
    },
    {
      "epoch": 0.5011467889908257,
      "grad_norm": 0.16117426753044128,
      "learning_rate": 0.0001,
      "loss": 1.4184,
      "step": 3059
    },
    {
      "epoch": 0.5013106159895151,
      "grad_norm": 0.1521613597869873,
      "learning_rate": 0.0001,
      "loss": 1.3555,
      "step": 3060
    },
    {
      "epoch": 0.5014744429882044,
      "grad_norm": 0.1606511026620865,
      "learning_rate": 0.0001,
      "loss": 1.4174,
      "step": 3061
    },
    {
      "epoch": 0.5016382699868939,
      "grad_norm": 0.1626928746700287,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 3062
    },
    {
      "epoch": 0.5018020969855832,
      "grad_norm": 0.15740859508514404,
      "learning_rate": 0.0001,
      "loss": 1.4662,
      "step": 3063
    },
    {
      "epoch": 0.5019659239842726,
      "grad_norm": 0.15318380296230316,
      "learning_rate": 0.0001,
      "loss": 1.3463,
      "step": 3064
    },
    {
      "epoch": 0.502129750982962,
      "grad_norm": 0.15881314873695374,
      "learning_rate": 0.0001,
      "loss": 1.4196,
      "step": 3065
    },
    {
      "epoch": 0.5022935779816514,
      "grad_norm": 0.15824539959430695,
      "learning_rate": 0.0001,
      "loss": 1.442,
      "step": 3066
    },
    {
      "epoch": 0.5024574049803407,
      "grad_norm": 0.1616899073123932,
      "learning_rate": 0.0001,
      "loss": 1.44,
      "step": 3067
    },
    {
      "epoch": 0.5026212319790302,
      "grad_norm": 0.15551972389221191,
      "learning_rate": 0.0001,
      "loss": 1.4773,
      "step": 3068
    },
    {
      "epoch": 0.5027850589777195,
      "grad_norm": 0.1551944613456726,
      "learning_rate": 0.0001,
      "loss": 1.4325,
      "step": 3069
    },
    {
      "epoch": 0.502948885976409,
      "grad_norm": 0.1586531698703766,
      "learning_rate": 0.0001,
      "loss": 1.3098,
      "step": 3070
    },
    {
      "epoch": 0.5031127129750983,
      "grad_norm": 0.15662261843681335,
      "learning_rate": 0.0001,
      "loss": 1.3894,
      "step": 3071
    },
    {
      "epoch": 0.5032765399737876,
      "grad_norm": 0.16216717660427094,
      "learning_rate": 0.0001,
      "loss": 1.3486,
      "step": 3072
    },
    {
      "epoch": 0.5034403669724771,
      "grad_norm": 0.155051127076149,
      "learning_rate": 0.0001,
      "loss": 1.3919,
      "step": 3073
    },
    {
      "epoch": 0.5036041939711664,
      "grad_norm": 0.15467730164527893,
      "learning_rate": 0.0001,
      "loss": 1.3783,
      "step": 3074
    },
    {
      "epoch": 0.5037680209698558,
      "grad_norm": 0.15726430714130402,
      "learning_rate": 0.0001,
      "loss": 1.3681,
      "step": 3075
    },
    {
      "epoch": 0.5039318479685452,
      "grad_norm": 0.16560597717761993,
      "learning_rate": 0.0001,
      "loss": 1.4423,
      "step": 3076
    },
    {
      "epoch": 0.5040956749672346,
      "grad_norm": 0.15500755608081818,
      "learning_rate": 0.0001,
      "loss": 1.387,
      "step": 3077
    },
    {
      "epoch": 0.504259501965924,
      "grad_norm": 0.15731605887413025,
      "learning_rate": 0.0001,
      "loss": 1.4004,
      "step": 3078
    },
    {
      "epoch": 0.5044233289646134,
      "grad_norm": 0.16936881840229034,
      "learning_rate": 0.0001,
      "loss": 1.4145,
      "step": 3079
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 0.15650883316993713,
      "learning_rate": 0.0001,
      "loss": 1.3256,
      "step": 3080
    },
    {
      "epoch": 0.5047509829619922,
      "grad_norm": 0.1538839191198349,
      "learning_rate": 0.0001,
      "loss": 1.3559,
      "step": 3081
    },
    {
      "epoch": 0.5049148099606815,
      "grad_norm": 0.14922861754894257,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 3082
    },
    {
      "epoch": 0.5050786369593709,
      "grad_norm": 0.16253449022769928,
      "learning_rate": 0.0001,
      "loss": 1.3895,
      "step": 3083
    },
    {
      "epoch": 0.5052424639580603,
      "grad_norm": 0.14972150325775146,
      "learning_rate": 0.0001,
      "loss": 1.281,
      "step": 3084
    },
    {
      "epoch": 0.5054062909567497,
      "grad_norm": 0.14757566154003143,
      "learning_rate": 0.0001,
      "loss": 1.2838,
      "step": 3085
    },
    {
      "epoch": 0.505570117955439,
      "grad_norm": 0.1727486103773117,
      "learning_rate": 0.0001,
      "loss": 1.3751,
      "step": 3086
    },
    {
      "epoch": 0.5057339449541285,
      "grad_norm": 0.1658698469400406,
      "learning_rate": 0.0001,
      "loss": 1.469,
      "step": 3087
    },
    {
      "epoch": 0.5058977719528178,
      "grad_norm": 0.15303896367549896,
      "learning_rate": 0.0001,
      "loss": 1.3625,
      "step": 3088
    },
    {
      "epoch": 0.5060615989515073,
      "grad_norm": 0.15903042256832123,
      "learning_rate": 0.0001,
      "loss": 1.5018,
      "step": 3089
    },
    {
      "epoch": 0.5062254259501966,
      "grad_norm": 0.16517114639282227,
      "learning_rate": 0.0001,
      "loss": 1.4822,
      "step": 3090
    },
    {
      "epoch": 0.5063892529488859,
      "grad_norm": 0.15502913296222687,
      "learning_rate": 0.0001,
      "loss": 1.5093,
      "step": 3091
    },
    {
      "epoch": 0.5065530799475754,
      "grad_norm": 0.15142430365085602,
      "learning_rate": 0.0001,
      "loss": 1.4418,
      "step": 3092
    },
    {
      "epoch": 0.5067169069462647,
      "grad_norm": 0.16117416322231293,
      "learning_rate": 0.0001,
      "loss": 1.4231,
      "step": 3093
    },
    {
      "epoch": 0.5068807339449541,
      "grad_norm": 0.16816459596157074,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 3094
    },
    {
      "epoch": 0.5070445609436435,
      "grad_norm": 0.14710134267807007,
      "learning_rate": 0.0001,
      "loss": 1.316,
      "step": 3095
    },
    {
      "epoch": 0.5072083879423329,
      "grad_norm": 0.1576150357723236,
      "learning_rate": 0.0001,
      "loss": 1.3692,
      "step": 3096
    },
    {
      "epoch": 0.5073722149410222,
      "grad_norm": 0.15747767686843872,
      "learning_rate": 0.0001,
      "loss": 1.393,
      "step": 3097
    },
    {
      "epoch": 0.5075360419397117,
      "grad_norm": 0.15026503801345825,
      "learning_rate": 0.0001,
      "loss": 1.3681,
      "step": 3098
    },
    {
      "epoch": 0.507699868938401,
      "grad_norm": 0.164973646402359,
      "learning_rate": 0.0001,
      "loss": 1.3843,
      "step": 3099
    },
    {
      "epoch": 0.5078636959370905,
      "grad_norm": 0.1605931669473648,
      "learning_rate": 0.0001,
      "loss": 1.4536,
      "step": 3100
    },
    {
      "epoch": 0.5080275229357798,
      "grad_norm": 0.16269828379154205,
      "learning_rate": 0.0001,
      "loss": 1.466,
      "step": 3101
    },
    {
      "epoch": 0.5081913499344692,
      "grad_norm": 0.17048552632331848,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 3102
    },
    {
      "epoch": 0.5083551769331586,
      "grad_norm": 0.16445574164390564,
      "learning_rate": 0.0001,
      "loss": 1.433,
      "step": 3103
    },
    {
      "epoch": 0.508519003931848,
      "grad_norm": 0.14948908984661102,
      "learning_rate": 0.0001,
      "loss": 1.2822,
      "step": 3104
    },
    {
      "epoch": 0.5086828309305373,
      "grad_norm": 0.15310120582580566,
      "learning_rate": 0.0001,
      "loss": 1.4172,
      "step": 3105
    },
    {
      "epoch": 0.5088466579292268,
      "grad_norm": 0.15388435125350952,
      "learning_rate": 0.0001,
      "loss": 1.3,
      "step": 3106
    },
    {
      "epoch": 0.5090104849279161,
      "grad_norm": 0.15065962076187134,
      "learning_rate": 0.0001,
      "loss": 1.3889,
      "step": 3107
    },
    {
      "epoch": 0.5091743119266054,
      "grad_norm": 0.15441220998764038,
      "learning_rate": 0.0001,
      "loss": 1.4727,
      "step": 3108
    },
    {
      "epoch": 0.5093381389252949,
      "grad_norm": 0.1520213782787323,
      "learning_rate": 0.0001,
      "loss": 1.373,
      "step": 3109
    },
    {
      "epoch": 0.5095019659239842,
      "grad_norm": 0.15781615674495697,
      "learning_rate": 0.0001,
      "loss": 1.4614,
      "step": 3110
    },
    {
      "epoch": 0.5096657929226737,
      "grad_norm": 0.15888218581676483,
      "learning_rate": 0.0001,
      "loss": 1.3671,
      "step": 3111
    },
    {
      "epoch": 0.509829619921363,
      "grad_norm": 0.1517275869846344,
      "learning_rate": 0.0001,
      "loss": 1.314,
      "step": 3112
    },
    {
      "epoch": 0.5099934469200524,
      "grad_norm": 0.15322531759738922,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 3113
    },
    {
      "epoch": 0.5101572739187418,
      "grad_norm": 0.15064217150211334,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 3114
    },
    {
      "epoch": 0.5103211009174312,
      "grad_norm": 0.15288230776786804,
      "learning_rate": 0.0001,
      "loss": 1.3221,
      "step": 3115
    },
    {
      "epoch": 0.5104849279161205,
      "grad_norm": 0.15563905239105225,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 3116
    },
    {
      "epoch": 0.51064875491481,
      "grad_norm": 0.14718469977378845,
      "learning_rate": 0.0001,
      "loss": 1.2861,
      "step": 3117
    },
    {
      "epoch": 0.5108125819134993,
      "grad_norm": 0.1559775471687317,
      "learning_rate": 0.0001,
      "loss": 1.4441,
      "step": 3118
    },
    {
      "epoch": 0.5109764089121888,
      "grad_norm": 0.1604371964931488,
      "learning_rate": 0.0001,
      "loss": 1.4546,
      "step": 3119
    },
    {
      "epoch": 0.5111402359108781,
      "grad_norm": 0.15143728256225586,
      "learning_rate": 0.0001,
      "loss": 1.3521,
      "step": 3120
    },
    {
      "epoch": 0.5113040629095675,
      "grad_norm": 0.15726228058338165,
      "learning_rate": 0.0001,
      "loss": 1.4497,
      "step": 3121
    },
    {
      "epoch": 0.5114678899082569,
      "grad_norm": 0.15338033437728882,
      "learning_rate": 0.0001,
      "loss": 1.3985,
      "step": 3122
    },
    {
      "epoch": 0.5116317169069463,
      "grad_norm": 0.15044595301151276,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 3123
    },
    {
      "epoch": 0.5117955439056356,
      "grad_norm": 0.15393128991127014,
      "learning_rate": 0.0001,
      "loss": 1.3634,
      "step": 3124
    },
    {
      "epoch": 0.5119593709043251,
      "grad_norm": 0.15505900979042053,
      "learning_rate": 0.0001,
      "loss": 1.4256,
      "step": 3125
    },
    {
      "epoch": 0.5121231979030144,
      "grad_norm": 0.15079718828201294,
      "learning_rate": 0.0001,
      "loss": 1.3329,
      "step": 3126
    },
    {
      "epoch": 0.5122870249017037,
      "grad_norm": 0.1613977998495102,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 3127
    },
    {
      "epoch": 0.5124508519003932,
      "grad_norm": 0.17279723286628723,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 3128
    },
    {
      "epoch": 0.5126146788990825,
      "grad_norm": 0.15811389684677124,
      "learning_rate": 0.0001,
      "loss": 1.4022,
      "step": 3129
    },
    {
      "epoch": 0.512778505897772,
      "grad_norm": 0.15464888513088226,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 3130
    },
    {
      "epoch": 0.5129423328964613,
      "grad_norm": 0.1523447483778,
      "learning_rate": 0.0001,
      "loss": 1.3007,
      "step": 3131
    },
    {
      "epoch": 0.5131061598951507,
      "grad_norm": 0.15364260971546173,
      "learning_rate": 0.0001,
      "loss": 1.3872,
      "step": 3132
    },
    {
      "epoch": 0.5132699868938401,
      "grad_norm": 0.14683347940444946,
      "learning_rate": 0.0001,
      "loss": 1.4025,
      "step": 3133
    },
    {
      "epoch": 0.5134338138925295,
      "grad_norm": 0.16101793944835663,
      "learning_rate": 0.0001,
      "loss": 1.3995,
      "step": 3134
    },
    {
      "epoch": 0.5135976408912188,
      "grad_norm": 0.15288811922073364,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 3135
    },
    {
      "epoch": 0.5137614678899083,
      "grad_norm": 0.15206459164619446,
      "learning_rate": 0.0001,
      "loss": 1.4045,
      "step": 3136
    },
    {
      "epoch": 0.5139252948885976,
      "grad_norm": 0.16388779878616333,
      "learning_rate": 0.0001,
      "loss": 1.2678,
      "step": 3137
    },
    {
      "epoch": 0.5140891218872871,
      "grad_norm": 0.15736539661884308,
      "learning_rate": 0.0001,
      "loss": 1.4848,
      "step": 3138
    },
    {
      "epoch": 0.5142529488859764,
      "grad_norm": 0.16876675188541412,
      "learning_rate": 0.0001,
      "loss": 1.3797,
      "step": 3139
    },
    {
      "epoch": 0.5144167758846658,
      "grad_norm": 0.14263050258159637,
      "learning_rate": 0.0001,
      "loss": 1.1985,
      "step": 3140
    },
    {
      "epoch": 0.5145806028833552,
      "grad_norm": 0.15652675926685333,
      "learning_rate": 0.0001,
      "loss": 1.458,
      "step": 3141
    },
    {
      "epoch": 0.5147444298820446,
      "grad_norm": 0.1560826301574707,
      "learning_rate": 0.0001,
      "loss": 1.4079,
      "step": 3142
    },
    {
      "epoch": 0.5149082568807339,
      "grad_norm": 0.1563560962677002,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 3143
    },
    {
      "epoch": 0.5150720838794234,
      "grad_norm": 0.16428299248218536,
      "learning_rate": 0.0001,
      "loss": 1.4075,
      "step": 3144
    },
    {
      "epoch": 0.5152359108781127,
      "grad_norm": 0.17014352977275848,
      "learning_rate": 0.0001,
      "loss": 1.4052,
      "step": 3145
    },
    {
      "epoch": 0.515399737876802,
      "grad_norm": 0.15248160064220428,
      "learning_rate": 0.0001,
      "loss": 1.4136,
      "step": 3146
    },
    {
      "epoch": 0.5155635648754915,
      "grad_norm": 0.15941667556762695,
      "learning_rate": 0.0001,
      "loss": 1.4326,
      "step": 3147
    },
    {
      "epoch": 0.5157273918741808,
      "grad_norm": 0.15303952991962433,
      "learning_rate": 0.0001,
      "loss": 1.3303,
      "step": 3148
    },
    {
      "epoch": 0.5158912188728703,
      "grad_norm": 0.16013024747371674,
      "learning_rate": 0.0001,
      "loss": 1.375,
      "step": 3149
    },
    {
      "epoch": 0.5160550458715596,
      "grad_norm": 0.15850119292736053,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 3150
    },
    {
      "epoch": 0.516218872870249,
      "grad_norm": 0.15732532739639282,
      "learning_rate": 0.0001,
      "loss": 1.3983,
      "step": 3151
    },
    {
      "epoch": 0.5163826998689384,
      "grad_norm": 0.1587899774312973,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 3152
    },
    {
      "epoch": 0.5165465268676278,
      "grad_norm": 0.15562982857227325,
      "learning_rate": 0.0001,
      "loss": 1.4008,
      "step": 3153
    },
    {
      "epoch": 0.5167103538663171,
      "grad_norm": 0.15390262007713318,
      "learning_rate": 0.0001,
      "loss": 1.4635,
      "step": 3154
    },
    {
      "epoch": 0.5168741808650066,
      "grad_norm": 0.15481238067150116,
      "learning_rate": 0.0001,
      "loss": 1.3048,
      "step": 3155
    },
    {
      "epoch": 0.5170380078636959,
      "grad_norm": 0.15978506207466125,
      "learning_rate": 0.0001,
      "loss": 1.3435,
      "step": 3156
    },
    {
      "epoch": 0.5172018348623854,
      "grad_norm": 0.16426631808280945,
      "learning_rate": 0.0001,
      "loss": 1.4937,
      "step": 3157
    },
    {
      "epoch": 0.5173656618610747,
      "grad_norm": 0.15610362589359283,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 3158
    },
    {
      "epoch": 0.5175294888597641,
      "grad_norm": 0.16073034703731537,
      "learning_rate": 0.0001,
      "loss": 1.3893,
      "step": 3159
    },
    {
      "epoch": 0.5176933158584535,
      "grad_norm": 0.15471351146697998,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 3160
    },
    {
      "epoch": 0.5178571428571429,
      "grad_norm": 0.1522933691740036,
      "learning_rate": 0.0001,
      "loss": 1.3519,
      "step": 3161
    },
    {
      "epoch": 0.5180209698558322,
      "grad_norm": 0.16093988716602325,
      "learning_rate": 0.0001,
      "loss": 1.3543,
      "step": 3162
    },
    {
      "epoch": 0.5181847968545217,
      "grad_norm": 0.1656874120235443,
      "learning_rate": 0.0001,
      "loss": 1.4295,
      "step": 3163
    },
    {
      "epoch": 0.518348623853211,
      "grad_norm": 0.14752846956253052,
      "learning_rate": 0.0001,
      "loss": 1.3519,
      "step": 3164
    },
    {
      "epoch": 0.5185124508519003,
      "grad_norm": 0.16837842762470245,
      "learning_rate": 0.0001,
      "loss": 1.433,
      "step": 3165
    },
    {
      "epoch": 0.5186762778505898,
      "grad_norm": 0.15857119858264923,
      "learning_rate": 0.0001,
      "loss": 1.437,
      "step": 3166
    },
    {
      "epoch": 0.5188401048492791,
      "grad_norm": 0.16640691459178925,
      "learning_rate": 0.0001,
      "loss": 1.3128,
      "step": 3167
    },
    {
      "epoch": 0.5190039318479686,
      "grad_norm": 0.1615755259990692,
      "learning_rate": 0.0001,
      "loss": 1.427,
      "step": 3168
    },
    {
      "epoch": 0.5191677588466579,
      "grad_norm": 0.1519969254732132,
      "learning_rate": 0.0001,
      "loss": 1.3182,
      "step": 3169
    },
    {
      "epoch": 0.5193315858453473,
      "grad_norm": 0.15201102197170258,
      "learning_rate": 0.0001,
      "loss": 1.4445,
      "step": 3170
    },
    {
      "epoch": 0.5194954128440367,
      "grad_norm": 0.1495627611875534,
      "learning_rate": 0.0001,
      "loss": 1.3859,
      "step": 3171
    },
    {
      "epoch": 0.5196592398427261,
      "grad_norm": 0.15822002291679382,
      "learning_rate": 0.0001,
      "loss": 1.3478,
      "step": 3172
    },
    {
      "epoch": 0.5198230668414154,
      "grad_norm": 0.15454402565956116,
      "learning_rate": 0.0001,
      "loss": 1.3716,
      "step": 3173
    },
    {
      "epoch": 0.5199868938401049,
      "grad_norm": 0.15244844555854797,
      "learning_rate": 0.0001,
      "loss": 1.3208,
      "step": 3174
    },
    {
      "epoch": 0.5201507208387942,
      "grad_norm": 0.15430091321468353,
      "learning_rate": 0.0001,
      "loss": 1.4281,
      "step": 3175
    },
    {
      "epoch": 0.5203145478374837,
      "grad_norm": 0.15329822897911072,
      "learning_rate": 0.0001,
      "loss": 1.325,
      "step": 3176
    },
    {
      "epoch": 0.520478374836173,
      "grad_norm": 0.1603015810251236,
      "learning_rate": 0.0001,
      "loss": 1.3665,
      "step": 3177
    },
    {
      "epoch": 0.5206422018348624,
      "grad_norm": 0.15734179317951202,
      "learning_rate": 0.0001,
      "loss": 1.4264,
      "step": 3178
    },
    {
      "epoch": 0.5208060288335518,
      "grad_norm": 0.15321259200572968,
      "learning_rate": 0.0001,
      "loss": 1.4033,
      "step": 3179
    },
    {
      "epoch": 0.5209698558322412,
      "grad_norm": 0.16317708790302277,
      "learning_rate": 0.0001,
      "loss": 1.3913,
      "step": 3180
    },
    {
      "epoch": 0.5211336828309305,
      "grad_norm": 0.1517869234085083,
      "learning_rate": 0.0001,
      "loss": 1.3626,
      "step": 3181
    },
    {
      "epoch": 0.52129750982962,
      "grad_norm": 0.1680460423231125,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 3182
    },
    {
      "epoch": 0.5214613368283093,
      "grad_norm": 0.16683177649974823,
      "learning_rate": 0.0001,
      "loss": 1.4216,
      "step": 3183
    },
    {
      "epoch": 0.5216251638269986,
      "grad_norm": 0.1604691445827484,
      "learning_rate": 0.0001,
      "loss": 1.4651,
      "step": 3184
    },
    {
      "epoch": 0.5217889908256881,
      "grad_norm": 0.16058731079101562,
      "learning_rate": 0.0001,
      "loss": 1.3766,
      "step": 3185
    },
    {
      "epoch": 0.5219528178243774,
      "grad_norm": 0.14968913793563843,
      "learning_rate": 0.0001,
      "loss": 1.4178,
      "step": 3186
    },
    {
      "epoch": 0.5221166448230669,
      "grad_norm": 0.16270941495895386,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 3187
    },
    {
      "epoch": 0.5222804718217562,
      "grad_norm": 0.15457259118556976,
      "learning_rate": 0.0001,
      "loss": 1.2611,
      "step": 3188
    },
    {
      "epoch": 0.5224442988204456,
      "grad_norm": 0.15791574120521545,
      "learning_rate": 0.0001,
      "loss": 1.3592,
      "step": 3189
    },
    {
      "epoch": 0.522608125819135,
      "grad_norm": 0.16487433016300201,
      "learning_rate": 0.0001,
      "loss": 1.3231,
      "step": 3190
    },
    {
      "epoch": 0.5227719528178244,
      "grad_norm": 0.15879109501838684,
      "learning_rate": 0.0001,
      "loss": 1.4623,
      "step": 3191
    },
    {
      "epoch": 0.5229357798165137,
      "grad_norm": 0.16021595895290375,
      "learning_rate": 0.0001,
      "loss": 1.3813,
      "step": 3192
    },
    {
      "epoch": 0.5230996068152032,
      "grad_norm": 0.16911016404628754,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 3193
    },
    {
      "epoch": 0.5232634338138925,
      "grad_norm": 0.16091054677963257,
      "learning_rate": 0.0001,
      "loss": 1.3215,
      "step": 3194
    },
    {
      "epoch": 0.523427260812582,
      "grad_norm": 0.166085347533226,
      "learning_rate": 0.0001,
      "loss": 1.309,
      "step": 3195
    },
    {
      "epoch": 0.5235910878112713,
      "grad_norm": 0.16262876987457275,
      "learning_rate": 0.0001,
      "loss": 1.4096,
      "step": 3196
    },
    {
      "epoch": 0.5237549148099607,
      "grad_norm": 0.1599142998456955,
      "learning_rate": 0.0001,
      "loss": 1.3466,
      "step": 3197
    },
    {
      "epoch": 0.5239187418086501,
      "grad_norm": 0.15633155405521393,
      "learning_rate": 0.0001,
      "loss": 1.405,
      "step": 3198
    },
    {
      "epoch": 0.5240825688073395,
      "grad_norm": 0.16193996369838715,
      "learning_rate": 0.0001,
      "loss": 1.325,
      "step": 3199
    },
    {
      "epoch": 0.5242463958060288,
      "grad_norm": 0.15419933199882507,
      "learning_rate": 0.0001,
      "loss": 1.4738,
      "step": 3200
    },
    {
      "epoch": 0.5244102228047182,
      "grad_norm": 0.16103234887123108,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 3201
    },
    {
      "epoch": 0.5245740498034076,
      "grad_norm": 0.16306355595588684,
      "learning_rate": 0.0001,
      "loss": 1.3747,
      "step": 3202
    },
    {
      "epoch": 0.5247378768020969,
      "grad_norm": 0.1698606312274933,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 3203
    },
    {
      "epoch": 0.5249017038007864,
      "grad_norm": 0.15497884154319763,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 3204
    },
    {
      "epoch": 0.5250655307994757,
      "grad_norm": 0.15337225794792175,
      "learning_rate": 0.0001,
      "loss": 1.3706,
      "step": 3205
    },
    {
      "epoch": 0.5252293577981652,
      "grad_norm": 0.15548992156982422,
      "learning_rate": 0.0001,
      "loss": 1.3003,
      "step": 3206
    },
    {
      "epoch": 0.5253931847968545,
      "grad_norm": 0.154807910323143,
      "learning_rate": 0.0001,
      "loss": 1.359,
      "step": 3207
    },
    {
      "epoch": 0.5255570117955439,
      "grad_norm": 0.15528562664985657,
      "learning_rate": 0.0001,
      "loss": 1.3825,
      "step": 3208
    },
    {
      "epoch": 0.5257208387942333,
      "grad_norm": 0.1626846194267273,
      "learning_rate": 0.0001,
      "loss": 1.4283,
      "step": 3209
    },
    {
      "epoch": 0.5258846657929227,
      "grad_norm": 0.1565113514661789,
      "learning_rate": 0.0001,
      "loss": 1.3536,
      "step": 3210
    },
    {
      "epoch": 0.526048492791612,
      "grad_norm": 0.17426101863384247,
      "learning_rate": 0.0001,
      "loss": 1.3934,
      "step": 3211
    },
    {
      "epoch": 0.5262123197903015,
      "grad_norm": 0.16140159964561462,
      "learning_rate": 0.0001,
      "loss": 1.4627,
      "step": 3212
    },
    {
      "epoch": 0.5263761467889908,
      "grad_norm": 0.1598568558692932,
      "learning_rate": 0.0001,
      "loss": 1.4282,
      "step": 3213
    },
    {
      "epoch": 0.5265399737876802,
      "grad_norm": 0.15218938887119293,
      "learning_rate": 0.0001,
      "loss": 1.4254,
      "step": 3214
    },
    {
      "epoch": 0.5267038007863696,
      "grad_norm": 0.146793395280838,
      "learning_rate": 0.0001,
      "loss": 1.3744,
      "step": 3215
    },
    {
      "epoch": 0.526867627785059,
      "grad_norm": 0.15167008340358734,
      "learning_rate": 0.0001,
      "loss": 1.3716,
      "step": 3216
    },
    {
      "epoch": 0.5270314547837484,
      "grad_norm": 0.15793649852275848,
      "learning_rate": 0.0001,
      "loss": 1.3954,
      "step": 3217
    },
    {
      "epoch": 0.5271952817824378,
      "grad_norm": 0.1791059821844101,
      "learning_rate": 0.0001,
      "loss": 1.4091,
      "step": 3218
    },
    {
      "epoch": 0.5273591087811271,
      "grad_norm": 0.1579255312681198,
      "learning_rate": 0.0001,
      "loss": 1.3151,
      "step": 3219
    },
    {
      "epoch": 0.5275229357798165,
      "grad_norm": 0.16419333219528198,
      "learning_rate": 0.0001,
      "loss": 1.3273,
      "step": 3220
    },
    {
      "epoch": 0.5276867627785059,
      "grad_norm": 0.16219273209571838,
      "learning_rate": 0.0001,
      "loss": 1.3023,
      "step": 3221
    },
    {
      "epoch": 0.5278505897771952,
      "grad_norm": 0.16126276552677155,
      "learning_rate": 0.0001,
      "loss": 1.3551,
      "step": 3222
    },
    {
      "epoch": 0.5280144167758847,
      "grad_norm": 0.16006450355052948,
      "learning_rate": 0.0001,
      "loss": 1.4805,
      "step": 3223
    },
    {
      "epoch": 0.528178243774574,
      "grad_norm": 0.16199712455272675,
      "learning_rate": 0.0001,
      "loss": 1.4972,
      "step": 3224
    },
    {
      "epoch": 0.5283420707732635,
      "grad_norm": 0.1577083170413971,
      "learning_rate": 0.0001,
      "loss": 1.428,
      "step": 3225
    },
    {
      "epoch": 0.5285058977719528,
      "grad_norm": 0.15229931473731995,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 3226
    },
    {
      "epoch": 0.5286697247706422,
      "grad_norm": 0.1588270366191864,
      "learning_rate": 0.0001,
      "loss": 1.4828,
      "step": 3227
    },
    {
      "epoch": 0.5288335517693316,
      "grad_norm": 0.15171857178211212,
      "learning_rate": 0.0001,
      "loss": 1.3208,
      "step": 3228
    },
    {
      "epoch": 0.528997378768021,
      "grad_norm": 0.16161440312862396,
      "learning_rate": 0.0001,
      "loss": 1.4363,
      "step": 3229
    },
    {
      "epoch": 0.5291612057667103,
      "grad_norm": 0.15563997626304626,
      "learning_rate": 0.0001,
      "loss": 1.4354,
      "step": 3230
    },
    {
      "epoch": 0.5293250327653998,
      "grad_norm": 0.15709270536899567,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 3231
    },
    {
      "epoch": 0.5294888597640891,
      "grad_norm": 0.15863892436027527,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 3232
    },
    {
      "epoch": 0.5296526867627785,
      "grad_norm": 0.1501876562833786,
      "learning_rate": 0.0001,
      "loss": 1.4037,
      "step": 3233
    },
    {
      "epoch": 0.5298165137614679,
      "grad_norm": 0.14905911684036255,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 3234
    },
    {
      "epoch": 0.5299803407601573,
      "grad_norm": 0.15317484736442566,
      "learning_rate": 0.0001,
      "loss": 1.3443,
      "step": 3235
    },
    {
      "epoch": 0.5301441677588467,
      "grad_norm": 0.15272928774356842,
      "learning_rate": 0.0001,
      "loss": 1.4159,
      "step": 3236
    },
    {
      "epoch": 0.5303079947575361,
      "grad_norm": 0.1601707488298416,
      "learning_rate": 0.0001,
      "loss": 1.406,
      "step": 3237
    },
    {
      "epoch": 0.5304718217562254,
      "grad_norm": 0.15514793992042542,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 3238
    },
    {
      "epoch": 0.5306356487549148,
      "grad_norm": 0.16160978376865387,
      "learning_rate": 0.0001,
      "loss": 1.3699,
      "step": 3239
    },
    {
      "epoch": 0.5307994757536042,
      "grad_norm": 0.16021312773227692,
      "learning_rate": 0.0001,
      "loss": 1.4583,
      "step": 3240
    },
    {
      "epoch": 0.5309633027522935,
      "grad_norm": 0.15305666625499725,
      "learning_rate": 0.0001,
      "loss": 1.2855,
      "step": 3241
    },
    {
      "epoch": 0.531127129750983,
      "grad_norm": 0.14988014101982117,
      "learning_rate": 0.0001,
      "loss": 1.3575,
      "step": 3242
    },
    {
      "epoch": 0.5312909567496723,
      "grad_norm": 0.1536436378955841,
      "learning_rate": 0.0001,
      "loss": 1.3895,
      "step": 3243
    },
    {
      "epoch": 0.5314547837483617,
      "grad_norm": 0.15701675415039062,
      "learning_rate": 0.0001,
      "loss": 1.3924,
      "step": 3244
    },
    {
      "epoch": 0.5316186107470511,
      "grad_norm": 0.16096170246601105,
      "learning_rate": 0.0001,
      "loss": 1.4302,
      "step": 3245
    },
    {
      "epoch": 0.5317824377457405,
      "grad_norm": 0.1713494211435318,
      "learning_rate": 0.0001,
      "loss": 1.3237,
      "step": 3246
    },
    {
      "epoch": 0.5319462647444299,
      "grad_norm": 0.16022519767284393,
      "learning_rate": 0.0001,
      "loss": 1.408,
      "step": 3247
    },
    {
      "epoch": 0.5321100917431193,
      "grad_norm": 0.15929211676120758,
      "learning_rate": 0.0001,
      "loss": 1.4658,
      "step": 3248
    },
    {
      "epoch": 0.5322739187418086,
      "grad_norm": 0.16144756972789764,
      "learning_rate": 0.0001,
      "loss": 1.3836,
      "step": 3249
    },
    {
      "epoch": 0.5324377457404981,
      "grad_norm": 0.1554626226425171,
      "learning_rate": 0.0001,
      "loss": 1.3894,
      "step": 3250
    },
    {
      "epoch": 0.5326015727391874,
      "grad_norm": 0.14822205901145935,
      "learning_rate": 0.0001,
      "loss": 1.3942,
      "step": 3251
    },
    {
      "epoch": 0.5327653997378768,
      "grad_norm": 0.1719445288181305,
      "learning_rate": 0.0001,
      "loss": 1.4307,
      "step": 3252
    },
    {
      "epoch": 0.5329292267365662,
      "grad_norm": 0.15101566910743713,
      "learning_rate": 0.0001,
      "loss": 1.3539,
      "step": 3253
    },
    {
      "epoch": 0.5330930537352556,
      "grad_norm": 0.15372300148010254,
      "learning_rate": 0.0001,
      "loss": 1.3979,
      "step": 3254
    },
    {
      "epoch": 0.533256880733945,
      "grad_norm": 0.15655173361301422,
      "learning_rate": 0.0001,
      "loss": 1.4642,
      "step": 3255
    },
    {
      "epoch": 0.5334207077326344,
      "grad_norm": 0.14959654211997986,
      "learning_rate": 0.0001,
      "loss": 1.3794,
      "step": 3256
    },
    {
      "epoch": 0.5335845347313237,
      "grad_norm": 0.15584081411361694,
      "learning_rate": 0.0001,
      "loss": 1.364,
      "step": 3257
    },
    {
      "epoch": 0.533748361730013,
      "grad_norm": 0.15809716284275055,
      "learning_rate": 0.0001,
      "loss": 1.4549,
      "step": 3258
    },
    {
      "epoch": 0.5339121887287025,
      "grad_norm": 0.15315701067447662,
      "learning_rate": 0.0001,
      "loss": 1.3821,
      "step": 3259
    },
    {
      "epoch": 0.5340760157273918,
      "grad_norm": 0.15030339360237122,
      "learning_rate": 0.0001,
      "loss": 1.2375,
      "step": 3260
    },
    {
      "epoch": 0.5342398427260813,
      "grad_norm": 0.15454158186912537,
      "learning_rate": 0.0001,
      "loss": 1.3489,
      "step": 3261
    },
    {
      "epoch": 0.5344036697247706,
      "grad_norm": 0.15768571197986603,
      "learning_rate": 0.0001,
      "loss": 1.3798,
      "step": 3262
    },
    {
      "epoch": 0.53456749672346,
      "grad_norm": 0.16063253581523895,
      "learning_rate": 0.0001,
      "loss": 1.4127,
      "step": 3263
    },
    {
      "epoch": 0.5347313237221494,
      "grad_norm": 0.1569528877735138,
      "learning_rate": 0.0001,
      "loss": 1.273,
      "step": 3264
    },
    {
      "epoch": 0.5348951507208388,
      "grad_norm": 0.1541382372379303,
      "learning_rate": 0.0001,
      "loss": 1.364,
      "step": 3265
    },
    {
      "epoch": 0.5350589777195282,
      "grad_norm": 0.1545758843421936,
      "learning_rate": 0.0001,
      "loss": 1.3668,
      "step": 3266
    },
    {
      "epoch": 0.5352228047182176,
      "grad_norm": 0.15677297115325928,
      "learning_rate": 0.0001,
      "loss": 1.3885,
      "step": 3267
    },
    {
      "epoch": 0.5353866317169069,
      "grad_norm": 0.16021591424942017,
      "learning_rate": 0.0001,
      "loss": 1.3752,
      "step": 3268
    },
    {
      "epoch": 0.5355504587155964,
      "grad_norm": 0.15175354480743408,
      "learning_rate": 0.0001,
      "loss": 1.2866,
      "step": 3269
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.15019062161445618,
      "learning_rate": 0.0001,
      "loss": 1.2861,
      "step": 3270
    },
    {
      "epoch": 0.5358781127129751,
      "grad_norm": 0.15766297280788422,
      "learning_rate": 0.0001,
      "loss": 1.3494,
      "step": 3271
    },
    {
      "epoch": 0.5360419397116645,
      "grad_norm": 0.1622101366519928,
      "learning_rate": 0.0001,
      "loss": 1.4358,
      "step": 3272
    },
    {
      "epoch": 0.5362057667103539,
      "grad_norm": 0.15352067351341248,
      "learning_rate": 0.0001,
      "loss": 1.3157,
      "step": 3273
    },
    {
      "epoch": 0.5363695937090432,
      "grad_norm": 0.15343217551708221,
      "learning_rate": 0.0001,
      "loss": 1.2823,
      "step": 3274
    },
    {
      "epoch": 0.5365334207077327,
      "grad_norm": 0.15676215291023254,
      "learning_rate": 0.0001,
      "loss": 1.4411,
      "step": 3275
    },
    {
      "epoch": 0.536697247706422,
      "grad_norm": 0.16070674359798431,
      "learning_rate": 0.0001,
      "loss": 1.4192,
      "step": 3276
    },
    {
      "epoch": 0.5368610747051114,
      "grad_norm": 0.1497662514448166,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 3277
    },
    {
      "epoch": 0.5370249017038008,
      "grad_norm": 0.15998876094818115,
      "learning_rate": 0.0001,
      "loss": 1.3653,
      "step": 3278
    },
    {
      "epoch": 0.5371887287024901,
      "grad_norm": 0.15945836901664734,
      "learning_rate": 0.0001,
      "loss": 1.3666,
      "step": 3279
    },
    {
      "epoch": 0.5373525557011796,
      "grad_norm": 0.16298797726631165,
      "learning_rate": 0.0001,
      "loss": 1.3438,
      "step": 3280
    },
    {
      "epoch": 0.5375163826998689,
      "grad_norm": 0.16425496339797974,
      "learning_rate": 0.0001,
      "loss": 1.3936,
      "step": 3281
    },
    {
      "epoch": 0.5376802096985583,
      "grad_norm": 0.1534551978111267,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 3282
    },
    {
      "epoch": 0.5378440366972477,
      "grad_norm": 0.15849551558494568,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 3283
    },
    {
      "epoch": 0.5380078636959371,
      "grad_norm": 0.14928393065929413,
      "learning_rate": 0.0001,
      "loss": 1.2628,
      "step": 3284
    },
    {
      "epoch": 0.5381716906946264,
      "grad_norm": 0.15841825306415558,
      "learning_rate": 0.0001,
      "loss": 1.335,
      "step": 3285
    },
    {
      "epoch": 0.5383355176933159,
      "grad_norm": 0.16684173047542572,
      "learning_rate": 0.0001,
      "loss": 1.4536,
      "step": 3286
    },
    {
      "epoch": 0.5384993446920052,
      "grad_norm": 0.1678474098443985,
      "learning_rate": 0.0001,
      "loss": 1.3568,
      "step": 3287
    },
    {
      "epoch": 0.5386631716906947,
      "grad_norm": 0.15575949847698212,
      "learning_rate": 0.0001,
      "loss": 1.3562,
      "step": 3288
    },
    {
      "epoch": 0.538826998689384,
      "grad_norm": 0.16336071491241455,
      "learning_rate": 0.0001,
      "loss": 1.4214,
      "step": 3289
    },
    {
      "epoch": 0.5389908256880734,
      "grad_norm": 0.19757138192653656,
      "learning_rate": 0.0001,
      "loss": 1.2907,
      "step": 3290
    },
    {
      "epoch": 0.5391546526867628,
      "grad_norm": 0.15097877383232117,
      "learning_rate": 0.0001,
      "loss": 1.3623,
      "step": 3291
    },
    {
      "epoch": 0.5393184796854522,
      "grad_norm": 0.15759052336215973,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 3292
    },
    {
      "epoch": 0.5394823066841415,
      "grad_norm": 0.16638770699501038,
      "learning_rate": 0.0001,
      "loss": 1.4517,
      "step": 3293
    },
    {
      "epoch": 0.5396461336828309,
      "grad_norm": 0.1652650535106659,
      "learning_rate": 0.0001,
      "loss": 1.5143,
      "step": 3294
    },
    {
      "epoch": 0.5398099606815203,
      "grad_norm": 0.16483528912067413,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 3295
    },
    {
      "epoch": 0.5399737876802096,
      "grad_norm": 0.16274838149547577,
      "learning_rate": 0.0001,
      "loss": 1.4162,
      "step": 3296
    },
    {
      "epoch": 0.5401376146788991,
      "grad_norm": 0.14565317332744598,
      "learning_rate": 0.0001,
      "loss": 1.2349,
      "step": 3297
    },
    {
      "epoch": 0.5403014416775884,
      "grad_norm": 0.16340258717536926,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 3298
    },
    {
      "epoch": 0.5404652686762779,
      "grad_norm": 0.1581149846315384,
      "learning_rate": 0.0001,
      "loss": 1.3613,
      "step": 3299
    },
    {
      "epoch": 0.5406290956749672,
      "grad_norm": 0.16416418552398682,
      "learning_rate": 0.0001,
      "loss": 1.4606,
      "step": 3300
    },
    {
      "epoch": 0.5407929226736566,
      "grad_norm": 0.15771958231925964,
      "learning_rate": 0.0001,
      "loss": 1.2827,
      "step": 3301
    },
    {
      "epoch": 0.540956749672346,
      "grad_norm": 0.17104551196098328,
      "learning_rate": 0.0001,
      "loss": 1.4538,
      "step": 3302
    },
    {
      "epoch": 0.5411205766710354,
      "grad_norm": 0.15599371492862701,
      "learning_rate": 0.0001,
      "loss": 1.4014,
      "step": 3303
    },
    {
      "epoch": 0.5412844036697247,
      "grad_norm": 0.16465947031974792,
      "learning_rate": 0.0001,
      "loss": 1.2862,
      "step": 3304
    },
    {
      "epoch": 0.5414482306684142,
      "grad_norm": 0.15466371178627014,
      "learning_rate": 0.0001,
      "loss": 1.4169,
      "step": 3305
    },
    {
      "epoch": 0.5416120576671035,
      "grad_norm": 0.1553994119167328,
      "learning_rate": 0.0001,
      "loss": 1.2662,
      "step": 3306
    },
    {
      "epoch": 0.541775884665793,
      "grad_norm": 0.16551703214645386,
      "learning_rate": 0.0001,
      "loss": 1.2796,
      "step": 3307
    },
    {
      "epoch": 0.5419397116644823,
      "grad_norm": 0.15938936173915863,
      "learning_rate": 0.0001,
      "loss": 1.4215,
      "step": 3308
    },
    {
      "epoch": 0.5421035386631717,
      "grad_norm": 0.16510869562625885,
      "learning_rate": 0.0001,
      "loss": 1.3661,
      "step": 3309
    },
    {
      "epoch": 0.5422673656618611,
      "grad_norm": 0.1556432545185089,
      "learning_rate": 0.0001,
      "loss": 1.2859,
      "step": 3310
    },
    {
      "epoch": 0.5424311926605505,
      "grad_norm": 0.1583060622215271,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 3311
    },
    {
      "epoch": 0.5425950196592398,
      "grad_norm": 0.16682517528533936,
      "learning_rate": 0.0001,
      "loss": 1.314,
      "step": 3312
    },
    {
      "epoch": 0.5427588466579292,
      "grad_norm": 0.15238280594348907,
      "learning_rate": 0.0001,
      "loss": 1.4421,
      "step": 3313
    },
    {
      "epoch": 0.5429226736566186,
      "grad_norm": 0.1597900241613388,
      "learning_rate": 0.0001,
      "loss": 1.2957,
      "step": 3314
    },
    {
      "epoch": 0.543086500655308,
      "grad_norm": 0.15793073177337646,
      "learning_rate": 0.0001,
      "loss": 1.2618,
      "step": 3315
    },
    {
      "epoch": 0.5432503276539974,
      "grad_norm": 0.15332724153995514,
      "learning_rate": 0.0001,
      "loss": 1.3195,
      "step": 3316
    },
    {
      "epoch": 0.5434141546526867,
      "grad_norm": 0.15713465213775635,
      "learning_rate": 0.0001,
      "loss": 1.3004,
      "step": 3317
    },
    {
      "epoch": 0.5435779816513762,
      "grad_norm": 0.16022226214408875,
      "learning_rate": 0.0001,
      "loss": 1.4155,
      "step": 3318
    },
    {
      "epoch": 0.5437418086500655,
      "grad_norm": 0.14790241420269012,
      "learning_rate": 0.0001,
      "loss": 1.329,
      "step": 3319
    },
    {
      "epoch": 0.5439056356487549,
      "grad_norm": 0.15472353994846344,
      "learning_rate": 0.0001,
      "loss": 1.3511,
      "step": 3320
    },
    {
      "epoch": 0.5440694626474443,
      "grad_norm": 0.16504500806331635,
      "learning_rate": 0.0001,
      "loss": 1.4097,
      "step": 3321
    },
    {
      "epoch": 0.5442332896461337,
      "grad_norm": 0.15848831832408905,
      "learning_rate": 0.0001,
      "loss": 1.4062,
      "step": 3322
    },
    {
      "epoch": 0.544397116644823,
      "grad_norm": 0.15800471603870392,
      "learning_rate": 0.0001,
      "loss": 1.4096,
      "step": 3323
    },
    {
      "epoch": 0.5445609436435125,
      "grad_norm": 0.15390364825725555,
      "learning_rate": 0.0001,
      "loss": 1.4156,
      "step": 3324
    },
    {
      "epoch": 0.5447247706422018,
      "grad_norm": 0.1534392386674881,
      "learning_rate": 0.0001,
      "loss": 1.3142,
      "step": 3325
    },
    {
      "epoch": 0.5448885976408913,
      "grad_norm": 0.14755360782146454,
      "learning_rate": 0.0001,
      "loss": 1.3018,
      "step": 3326
    },
    {
      "epoch": 0.5450524246395806,
      "grad_norm": 0.150929793715477,
      "learning_rate": 0.0001,
      "loss": 1.3264,
      "step": 3327
    },
    {
      "epoch": 0.54521625163827,
      "grad_norm": 0.1628980189561844,
      "learning_rate": 0.0001,
      "loss": 1.3324,
      "step": 3328
    },
    {
      "epoch": 0.5453800786369594,
      "grad_norm": 0.15837039053440094,
      "learning_rate": 0.0001,
      "loss": 1.3819,
      "step": 3329
    },
    {
      "epoch": 0.5455439056356488,
      "grad_norm": 0.1531239151954651,
      "learning_rate": 0.0001,
      "loss": 1.406,
      "step": 3330
    },
    {
      "epoch": 0.5457077326343381,
      "grad_norm": 0.15548843145370483,
      "learning_rate": 0.0001,
      "loss": 1.4053,
      "step": 3331
    },
    {
      "epoch": 0.5458715596330275,
      "grad_norm": 0.15069670975208282,
      "learning_rate": 0.0001,
      "loss": 1.2612,
      "step": 3332
    },
    {
      "epoch": 0.5460353866317169,
      "grad_norm": 0.1582380086183548,
      "learning_rate": 0.0001,
      "loss": 1.3402,
      "step": 3333
    },
    {
      "epoch": 0.5461992136304062,
      "grad_norm": 0.15135011076927185,
      "learning_rate": 0.0001,
      "loss": 1.3592,
      "step": 3334
    },
    {
      "epoch": 0.5463630406290957,
      "grad_norm": 0.14737272262573242,
      "learning_rate": 0.0001,
      "loss": 1.3533,
      "step": 3335
    },
    {
      "epoch": 0.546526867627785,
      "grad_norm": 0.15167444944381714,
      "learning_rate": 0.0001,
      "loss": 1.3229,
      "step": 3336
    },
    {
      "epoch": 0.5466906946264745,
      "grad_norm": 0.161701500415802,
      "learning_rate": 0.0001,
      "loss": 1.3245,
      "step": 3337
    },
    {
      "epoch": 0.5468545216251638,
      "grad_norm": 0.16168445348739624,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 3338
    },
    {
      "epoch": 0.5470183486238532,
      "grad_norm": 0.15793639421463013,
      "learning_rate": 0.0001,
      "loss": 1.4132,
      "step": 3339
    },
    {
      "epoch": 0.5471821756225426,
      "grad_norm": 0.16673880815505981,
      "learning_rate": 0.0001,
      "loss": 1.509,
      "step": 3340
    },
    {
      "epoch": 0.547346002621232,
      "grad_norm": 0.15035691857337952,
      "learning_rate": 0.0001,
      "loss": 1.3131,
      "step": 3341
    },
    {
      "epoch": 0.5475098296199213,
      "grad_norm": 0.1490635871887207,
      "learning_rate": 0.0001,
      "loss": 1.3947,
      "step": 3342
    },
    {
      "epoch": 0.5476736566186108,
      "grad_norm": 0.15526551008224487,
      "learning_rate": 0.0001,
      "loss": 1.3073,
      "step": 3343
    },
    {
      "epoch": 0.5478374836173001,
      "grad_norm": 0.1599830836057663,
      "learning_rate": 0.0001,
      "loss": 1.439,
      "step": 3344
    },
    {
      "epoch": 0.5480013106159896,
      "grad_norm": 0.159134641289711,
      "learning_rate": 0.0001,
      "loss": 1.3906,
      "step": 3345
    },
    {
      "epoch": 0.5481651376146789,
      "grad_norm": 0.15922236442565918,
      "learning_rate": 0.0001,
      "loss": 1.3721,
      "step": 3346
    },
    {
      "epoch": 0.5483289646133683,
      "grad_norm": 0.15562568604946136,
      "learning_rate": 0.0001,
      "loss": 1.3775,
      "step": 3347
    },
    {
      "epoch": 0.5484927916120577,
      "grad_norm": 0.1517964005470276,
      "learning_rate": 0.0001,
      "loss": 1.306,
      "step": 3348
    },
    {
      "epoch": 0.5486566186107471,
      "grad_norm": 0.16020222008228302,
      "learning_rate": 0.0001,
      "loss": 1.4049,
      "step": 3349
    },
    {
      "epoch": 0.5488204456094364,
      "grad_norm": 0.15934665501117706,
      "learning_rate": 0.0001,
      "loss": 1.3383,
      "step": 3350
    },
    {
      "epoch": 0.5489842726081258,
      "grad_norm": 0.15489742159843445,
      "learning_rate": 0.0001,
      "loss": 1.4444,
      "step": 3351
    },
    {
      "epoch": 0.5491480996068152,
      "grad_norm": 0.16741733253002167,
      "learning_rate": 0.0001,
      "loss": 1.3612,
      "step": 3352
    },
    {
      "epoch": 0.5493119266055045,
      "grad_norm": 0.16490508615970612,
      "learning_rate": 0.0001,
      "loss": 1.3487,
      "step": 3353
    },
    {
      "epoch": 0.549475753604194,
      "grad_norm": 0.1504100114107132,
      "learning_rate": 0.0001,
      "loss": 1.3507,
      "step": 3354
    },
    {
      "epoch": 0.5496395806028833,
      "grad_norm": 0.1538248509168625,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 3355
    },
    {
      "epoch": 0.5498034076015728,
      "grad_norm": 0.16300545632839203,
      "learning_rate": 0.0001,
      "loss": 1.3937,
      "step": 3356
    },
    {
      "epoch": 0.5499672346002621,
      "grad_norm": 0.1669493466615677,
      "learning_rate": 0.0001,
      "loss": 1.4463,
      "step": 3357
    },
    {
      "epoch": 0.5501310615989515,
      "grad_norm": 0.15277504920959473,
      "learning_rate": 0.0001,
      "loss": 1.2184,
      "step": 3358
    },
    {
      "epoch": 0.5502948885976409,
      "grad_norm": 0.15649913251399994,
      "learning_rate": 0.0001,
      "loss": 1.3866,
      "step": 3359
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 0.15929396450519562,
      "learning_rate": 0.0001,
      "loss": 1.3291,
      "step": 3360
    },
    {
      "epoch": 0.5506225425950196,
      "grad_norm": 0.15467378497123718,
      "learning_rate": 0.0001,
      "loss": 1.4604,
      "step": 3361
    },
    {
      "epoch": 0.5507863695937091,
      "grad_norm": 0.15455207228660583,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 3362
    },
    {
      "epoch": 0.5509501965923984,
      "grad_norm": 0.15765371918678284,
      "learning_rate": 0.0001,
      "loss": 1.3639,
      "step": 3363
    },
    {
      "epoch": 0.5511140235910879,
      "grad_norm": 0.1846201866865158,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 3364
    },
    {
      "epoch": 0.5512778505897772,
      "grad_norm": 0.1665903925895691,
      "learning_rate": 0.0001,
      "loss": 1.4629,
      "step": 3365
    },
    {
      "epoch": 0.5514416775884666,
      "grad_norm": 0.17558623850345612,
      "learning_rate": 0.0001,
      "loss": 1.3801,
      "step": 3366
    },
    {
      "epoch": 0.551605504587156,
      "grad_norm": 0.15337231755256653,
      "learning_rate": 0.0001,
      "loss": 1.3449,
      "step": 3367
    },
    {
      "epoch": 0.5517693315858454,
      "grad_norm": 0.15861354768276215,
      "learning_rate": 0.0001,
      "loss": 1.3555,
      "step": 3368
    },
    {
      "epoch": 0.5519331585845347,
      "grad_norm": 0.16071724891662598,
      "learning_rate": 0.0001,
      "loss": 1.328,
      "step": 3369
    },
    {
      "epoch": 0.5520969855832241,
      "grad_norm": 0.16052602231502533,
      "learning_rate": 0.0001,
      "loss": 1.3579,
      "step": 3370
    },
    {
      "epoch": 0.5522608125819135,
      "grad_norm": 0.1543329507112503,
      "learning_rate": 0.0001,
      "loss": 1.3894,
      "step": 3371
    },
    {
      "epoch": 0.5524246395806028,
      "grad_norm": 0.16040904819965363,
      "learning_rate": 0.0001,
      "loss": 1.4216,
      "step": 3372
    },
    {
      "epoch": 0.5525884665792923,
      "grad_norm": 0.15939216315746307,
      "learning_rate": 0.0001,
      "loss": 1.4278,
      "step": 3373
    },
    {
      "epoch": 0.5527522935779816,
      "grad_norm": 0.16522258520126343,
      "learning_rate": 0.0001,
      "loss": 1.3851,
      "step": 3374
    },
    {
      "epoch": 0.552916120576671,
      "grad_norm": 0.14980298280715942,
      "learning_rate": 0.0001,
      "loss": 1.3143,
      "step": 3375
    },
    {
      "epoch": 0.5530799475753604,
      "grad_norm": 0.15851494669914246,
      "learning_rate": 0.0001,
      "loss": 1.2865,
      "step": 3376
    },
    {
      "epoch": 0.5532437745740498,
      "grad_norm": 0.157485231757164,
      "learning_rate": 0.0001,
      "loss": 1.4352,
      "step": 3377
    },
    {
      "epoch": 0.5534076015727392,
      "grad_norm": 0.14806874096393585,
      "learning_rate": 0.0001,
      "loss": 1.2399,
      "step": 3378
    },
    {
      "epoch": 0.5535714285714286,
      "grad_norm": 0.15862835943698883,
      "learning_rate": 0.0001,
      "loss": 1.5005,
      "step": 3379
    },
    {
      "epoch": 0.5537352555701179,
      "grad_norm": 0.16386578977108002,
      "learning_rate": 0.0001,
      "loss": 1.4736,
      "step": 3380
    },
    {
      "epoch": 0.5538990825688074,
      "grad_norm": 0.15073326230049133,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 3381
    },
    {
      "epoch": 0.5540629095674967,
      "grad_norm": 0.15296901762485504,
      "learning_rate": 0.0001,
      "loss": 1.3135,
      "step": 3382
    },
    {
      "epoch": 0.5542267365661862,
      "grad_norm": 0.1505075842142105,
      "learning_rate": 0.0001,
      "loss": 1.304,
      "step": 3383
    },
    {
      "epoch": 0.5543905635648755,
      "grad_norm": 0.14934860169887543,
      "learning_rate": 0.0001,
      "loss": 1.3873,
      "step": 3384
    },
    {
      "epoch": 0.5545543905635649,
      "grad_norm": 0.16022281348705292,
      "learning_rate": 0.0001,
      "loss": 1.3786,
      "step": 3385
    },
    {
      "epoch": 0.5547182175622543,
      "grad_norm": 0.17381316423416138,
      "learning_rate": 0.0001,
      "loss": 1.3911,
      "step": 3386
    },
    {
      "epoch": 0.5548820445609436,
      "grad_norm": 0.16852542757987976,
      "learning_rate": 0.0001,
      "loss": 1.4317,
      "step": 3387
    },
    {
      "epoch": 0.555045871559633,
      "grad_norm": 0.16083727777004242,
      "learning_rate": 0.0001,
      "loss": 1.4015,
      "step": 3388
    },
    {
      "epoch": 0.5552096985583224,
      "grad_norm": 0.14883548021316528,
      "learning_rate": 0.0001,
      "loss": 1.2419,
      "step": 3389
    },
    {
      "epoch": 0.5553735255570118,
      "grad_norm": 0.15846002101898193,
      "learning_rate": 0.0001,
      "loss": 1.3618,
      "step": 3390
    },
    {
      "epoch": 0.5555373525557011,
      "grad_norm": 0.15605978667736053,
      "learning_rate": 0.0001,
      "loss": 1.3823,
      "step": 3391
    },
    {
      "epoch": 0.5557011795543906,
      "grad_norm": 0.16352106630802155,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 3392
    },
    {
      "epoch": 0.5558650065530799,
      "grad_norm": 0.1569790542125702,
      "learning_rate": 0.0001,
      "loss": 1.393,
      "step": 3393
    },
    {
      "epoch": 0.5560288335517694,
      "grad_norm": 0.17054183781147003,
      "learning_rate": 0.0001,
      "loss": 1.3977,
      "step": 3394
    },
    {
      "epoch": 0.5561926605504587,
      "grad_norm": 0.1623123735189438,
      "learning_rate": 0.0001,
      "loss": 1.4007,
      "step": 3395
    },
    {
      "epoch": 0.5563564875491481,
      "grad_norm": 0.14856109023094177,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 3396
    },
    {
      "epoch": 0.5565203145478375,
      "grad_norm": 0.1560380458831787,
      "learning_rate": 0.0001,
      "loss": 1.3288,
      "step": 3397
    },
    {
      "epoch": 0.5566841415465269,
      "grad_norm": 0.16775888204574585,
      "learning_rate": 0.0001,
      "loss": 1.3918,
      "step": 3398
    },
    {
      "epoch": 0.5568479685452162,
      "grad_norm": 0.16493381559848785,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 3399
    },
    {
      "epoch": 0.5570117955439057,
      "grad_norm": 0.14690376818180084,
      "learning_rate": 0.0001,
      "loss": 1.3042,
      "step": 3400
    },
    {
      "epoch": 0.557175622542595,
      "grad_norm": 0.16124671697616577,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 3401
    },
    {
      "epoch": 0.5573394495412844,
      "grad_norm": 0.1655091643333435,
      "learning_rate": 0.0001,
      "loss": 1.4133,
      "step": 3402
    },
    {
      "epoch": 0.5575032765399738,
      "grad_norm": 0.15246884524822235,
      "learning_rate": 0.0001,
      "loss": 1.384,
      "step": 3403
    },
    {
      "epoch": 0.5576671035386632,
      "grad_norm": 0.15544387698173523,
      "learning_rate": 0.0001,
      "loss": 1.3825,
      "step": 3404
    },
    {
      "epoch": 0.5578309305373526,
      "grad_norm": 0.15238462388515472,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 3405
    },
    {
      "epoch": 0.5579947575360419,
      "grad_norm": 0.15365029871463776,
      "learning_rate": 0.0001,
      "loss": 1.3805,
      "step": 3406
    },
    {
      "epoch": 0.5581585845347313,
      "grad_norm": 0.1541081964969635,
      "learning_rate": 0.0001,
      "loss": 1.3839,
      "step": 3407
    },
    {
      "epoch": 0.5583224115334207,
      "grad_norm": 0.16321957111358643,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 3408
    },
    {
      "epoch": 0.5584862385321101,
      "grad_norm": 0.14900849759578705,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 3409
    },
    {
      "epoch": 0.5586500655307994,
      "grad_norm": 0.15165677666664124,
      "learning_rate": 0.0001,
      "loss": 1.2751,
      "step": 3410
    },
    {
      "epoch": 0.5588138925294889,
      "grad_norm": 0.1472042351961136,
      "learning_rate": 0.0001,
      "loss": 1.298,
      "step": 3411
    },
    {
      "epoch": 0.5589777195281782,
      "grad_norm": 0.15495003759860992,
      "learning_rate": 0.0001,
      "loss": 1.4136,
      "step": 3412
    },
    {
      "epoch": 0.5591415465268676,
      "grad_norm": 0.15345583856105804,
      "learning_rate": 0.0001,
      "loss": 1.2776,
      "step": 3413
    },
    {
      "epoch": 0.559305373525557,
      "grad_norm": 0.15667136013507843,
      "learning_rate": 0.0001,
      "loss": 1.3184,
      "step": 3414
    },
    {
      "epoch": 0.5594692005242464,
      "grad_norm": 0.16785891354084015,
      "learning_rate": 0.0001,
      "loss": 1.3886,
      "step": 3415
    },
    {
      "epoch": 0.5596330275229358,
      "grad_norm": 0.15290211141109467,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 3416
    },
    {
      "epoch": 0.5597968545216252,
      "grad_norm": 0.1584518551826477,
      "learning_rate": 0.0001,
      "loss": 1.3445,
      "step": 3417
    },
    {
      "epoch": 0.5599606815203145,
      "grad_norm": 0.15512225031852722,
      "learning_rate": 0.0001,
      "loss": 1.4243,
      "step": 3418
    },
    {
      "epoch": 0.560124508519004,
      "grad_norm": 0.1540573686361313,
      "learning_rate": 0.0001,
      "loss": 1.3752,
      "step": 3419
    },
    {
      "epoch": 0.5602883355176933,
      "grad_norm": 0.161785289645195,
      "learning_rate": 0.0001,
      "loss": 1.4531,
      "step": 3420
    },
    {
      "epoch": 0.5604521625163827,
      "grad_norm": 0.16067829728126526,
      "learning_rate": 0.0001,
      "loss": 1.3897,
      "step": 3421
    },
    {
      "epoch": 0.5606159895150721,
      "grad_norm": 0.167903870344162,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 3422
    },
    {
      "epoch": 0.5607798165137615,
      "grad_norm": 0.152497336268425,
      "learning_rate": 0.0001,
      "loss": 1.3168,
      "step": 3423
    },
    {
      "epoch": 0.5609436435124509,
      "grad_norm": 0.16577547788619995,
      "learning_rate": 0.0001,
      "loss": 1.4054,
      "step": 3424
    },
    {
      "epoch": 0.5611074705111402,
      "grad_norm": 0.161859929561615,
      "learning_rate": 0.0001,
      "loss": 1.4499,
      "step": 3425
    },
    {
      "epoch": 0.5612712975098296,
      "grad_norm": 0.14946794509887695,
      "learning_rate": 0.0001,
      "loss": 1.3847,
      "step": 3426
    },
    {
      "epoch": 0.561435124508519,
      "grad_norm": 0.1513720452785492,
      "learning_rate": 0.0001,
      "loss": 1.3242,
      "step": 3427
    },
    {
      "epoch": 0.5615989515072084,
      "grad_norm": 0.16106495261192322,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 3428
    },
    {
      "epoch": 0.5617627785058977,
      "grad_norm": 0.15070554614067078,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 3429
    },
    {
      "epoch": 0.5619266055045872,
      "grad_norm": 0.19005869328975677,
      "learning_rate": 0.0001,
      "loss": 1.4174,
      "step": 3430
    },
    {
      "epoch": 0.5620904325032765,
      "grad_norm": 0.1551756113767624,
      "learning_rate": 0.0001,
      "loss": 1.4707,
      "step": 3431
    },
    {
      "epoch": 0.562254259501966,
      "grad_norm": 0.15510272979736328,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 3432
    },
    {
      "epoch": 0.5624180865006553,
      "grad_norm": 0.15601950883865356,
      "learning_rate": 0.0001,
      "loss": 1.337,
      "step": 3433
    },
    {
      "epoch": 0.5625819134993447,
      "grad_norm": 0.1582876592874527,
      "learning_rate": 0.0001,
      "loss": 1.4502,
      "step": 3434
    },
    {
      "epoch": 0.562745740498034,
      "grad_norm": 0.1620786190032959,
      "learning_rate": 0.0001,
      "loss": 1.4195,
      "step": 3435
    },
    {
      "epoch": 0.5629095674967235,
      "grad_norm": 0.16836808621883392,
      "learning_rate": 0.0001,
      "loss": 1.4175,
      "step": 3436
    },
    {
      "epoch": 0.5630733944954128,
      "grad_norm": 0.1606140285730362,
      "learning_rate": 0.0001,
      "loss": 1.3584,
      "step": 3437
    },
    {
      "epoch": 0.5632372214941023,
      "grad_norm": 0.1609199047088623,
      "learning_rate": 0.0001,
      "loss": 1.3542,
      "step": 3438
    },
    {
      "epoch": 0.5634010484927916,
      "grad_norm": 0.1557263880968094,
      "learning_rate": 0.0001,
      "loss": 1.2814,
      "step": 3439
    },
    {
      "epoch": 0.563564875491481,
      "grad_norm": 0.1543954759836197,
      "learning_rate": 0.0001,
      "loss": 1.3609,
      "step": 3440
    },
    {
      "epoch": 0.5637287024901704,
      "grad_norm": 0.16354897618293762,
      "learning_rate": 0.0001,
      "loss": 1.265,
      "step": 3441
    },
    {
      "epoch": 0.5638925294888598,
      "grad_norm": 0.16517752408981323,
      "learning_rate": 0.0001,
      "loss": 1.3784,
      "step": 3442
    },
    {
      "epoch": 0.5640563564875491,
      "grad_norm": 0.14962682127952576,
      "learning_rate": 0.0001,
      "loss": 1.361,
      "step": 3443
    },
    {
      "epoch": 0.5642201834862385,
      "grad_norm": 0.16009026765823364,
      "learning_rate": 0.0001,
      "loss": 1.4135,
      "step": 3444
    },
    {
      "epoch": 0.5643840104849279,
      "grad_norm": 0.1559453010559082,
      "learning_rate": 0.0001,
      "loss": 1.3328,
      "step": 3445
    },
    {
      "epoch": 0.5645478374836173,
      "grad_norm": 0.1671561449766159,
      "learning_rate": 0.0001,
      "loss": 1.4049,
      "step": 3446
    },
    {
      "epoch": 0.5647116644823067,
      "grad_norm": 0.15488769114017487,
      "learning_rate": 0.0001,
      "loss": 1.3246,
      "step": 3447
    },
    {
      "epoch": 0.564875491480996,
      "grad_norm": 0.1541290581226349,
      "learning_rate": 0.0001,
      "loss": 1.3897,
      "step": 3448
    },
    {
      "epoch": 0.5650393184796855,
      "grad_norm": 0.15666979551315308,
      "learning_rate": 0.0001,
      "loss": 1.2645,
      "step": 3449
    },
    {
      "epoch": 0.5652031454783748,
      "grad_norm": 0.16123558580875397,
      "learning_rate": 0.0001,
      "loss": 1.3893,
      "step": 3450
    },
    {
      "epoch": 0.5653669724770642,
      "grad_norm": 0.16170500218868256,
      "learning_rate": 0.0001,
      "loss": 1.4269,
      "step": 3451
    },
    {
      "epoch": 0.5655307994757536,
      "grad_norm": 0.1569894701242447,
      "learning_rate": 0.0001,
      "loss": 1.3888,
      "step": 3452
    },
    {
      "epoch": 0.565694626474443,
      "grad_norm": 0.15133146941661835,
      "learning_rate": 0.0001,
      "loss": 1.3732,
      "step": 3453
    },
    {
      "epoch": 0.5658584534731324,
      "grad_norm": 0.1543206125497818,
      "learning_rate": 0.0001,
      "loss": 1.273,
      "step": 3454
    },
    {
      "epoch": 0.5660222804718218,
      "grad_norm": 0.1677992343902588,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 3455
    },
    {
      "epoch": 0.5661861074705111,
      "grad_norm": 0.15964125096797943,
      "learning_rate": 0.0001,
      "loss": 1.4747,
      "step": 3456
    },
    {
      "epoch": 0.5663499344692006,
      "grad_norm": 0.1665022224187851,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 3457
    },
    {
      "epoch": 0.5665137614678899,
      "grad_norm": 0.1511824131011963,
      "learning_rate": 0.0001,
      "loss": 1.2317,
      "step": 3458
    },
    {
      "epoch": 0.5666775884665793,
      "grad_norm": 0.1574319452047348,
      "learning_rate": 0.0001,
      "loss": 1.4095,
      "step": 3459
    },
    {
      "epoch": 0.5668414154652687,
      "grad_norm": 0.16297291219234467,
      "learning_rate": 0.0001,
      "loss": 1.4064,
      "step": 3460
    },
    {
      "epoch": 0.5670052424639581,
      "grad_norm": 0.15802359580993652,
      "learning_rate": 0.0001,
      "loss": 1.3313,
      "step": 3461
    },
    {
      "epoch": 0.5671690694626474,
      "grad_norm": 0.15139855444431305,
      "learning_rate": 0.0001,
      "loss": 1.3931,
      "step": 3462
    },
    {
      "epoch": 0.5673328964613368,
      "grad_norm": 0.15268079936504364,
      "learning_rate": 0.0001,
      "loss": 1.3188,
      "step": 3463
    },
    {
      "epoch": 0.5674967234600262,
      "grad_norm": 0.15900036692619324,
      "learning_rate": 0.0001,
      "loss": 1.3888,
      "step": 3464
    },
    {
      "epoch": 0.5676605504587156,
      "grad_norm": 0.1613226681947708,
      "learning_rate": 0.0001,
      "loss": 1.4051,
      "step": 3465
    },
    {
      "epoch": 0.567824377457405,
      "grad_norm": 0.15469303727149963,
      "learning_rate": 0.0001,
      "loss": 1.3852,
      "step": 3466
    },
    {
      "epoch": 0.5679882044560943,
      "grad_norm": 0.16046836972236633,
      "learning_rate": 0.0001,
      "loss": 1.3617,
      "step": 3467
    },
    {
      "epoch": 0.5681520314547838,
      "grad_norm": 0.1629977971315384,
      "learning_rate": 0.0001,
      "loss": 1.3523,
      "step": 3468
    },
    {
      "epoch": 0.5683158584534731,
      "grad_norm": 0.16343943774700165,
      "learning_rate": 0.0001,
      "loss": 1.3807,
      "step": 3469
    },
    {
      "epoch": 0.5684796854521625,
      "grad_norm": 0.16251356899738312,
      "learning_rate": 0.0001,
      "loss": 1.4727,
      "step": 3470
    },
    {
      "epoch": 0.5686435124508519,
      "grad_norm": 0.1631448119878769,
      "learning_rate": 0.0001,
      "loss": 1.2811,
      "step": 3471
    },
    {
      "epoch": 0.5688073394495413,
      "grad_norm": 0.17452719807624817,
      "learning_rate": 0.0001,
      "loss": 1.4685,
      "step": 3472
    },
    {
      "epoch": 0.5689711664482306,
      "grad_norm": 0.15673619508743286,
      "learning_rate": 0.0001,
      "loss": 1.4154,
      "step": 3473
    },
    {
      "epoch": 0.5691349934469201,
      "grad_norm": 0.15408119559288025,
      "learning_rate": 0.0001,
      "loss": 1.3062,
      "step": 3474
    },
    {
      "epoch": 0.5692988204456094,
      "grad_norm": 0.15874545276165009,
      "learning_rate": 0.0001,
      "loss": 1.3341,
      "step": 3475
    },
    {
      "epoch": 0.5694626474442989,
      "grad_norm": 0.15042144060134888,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 3476
    },
    {
      "epoch": 0.5696264744429882,
      "grad_norm": 0.1547611951828003,
      "learning_rate": 0.0001,
      "loss": 1.3007,
      "step": 3477
    },
    {
      "epoch": 0.5697903014416776,
      "grad_norm": 0.16040532290935516,
      "learning_rate": 0.0001,
      "loss": 1.3023,
      "step": 3478
    },
    {
      "epoch": 0.569954128440367,
      "grad_norm": 0.15636515617370605,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 3479
    },
    {
      "epoch": 0.5701179554390564,
      "grad_norm": 0.15088489651679993,
      "learning_rate": 0.0001,
      "loss": 1.329,
      "step": 3480
    },
    {
      "epoch": 0.5702817824377457,
      "grad_norm": 0.1651688665151596,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 3481
    },
    {
      "epoch": 0.5704456094364351,
      "grad_norm": 0.15982520580291748,
      "learning_rate": 0.0001,
      "loss": 1.4449,
      "step": 3482
    },
    {
      "epoch": 0.5706094364351245,
      "grad_norm": 0.15730883181095123,
      "learning_rate": 0.0001,
      "loss": 1.3478,
      "step": 3483
    },
    {
      "epoch": 0.5707732634338138,
      "grad_norm": 0.15306681394577026,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 3484
    },
    {
      "epoch": 0.5709370904325033,
      "grad_norm": 0.16608117520809174,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 3485
    },
    {
      "epoch": 0.5711009174311926,
      "grad_norm": 0.16280698776245117,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 3486
    },
    {
      "epoch": 0.5712647444298821,
      "grad_norm": 0.14539968967437744,
      "learning_rate": 0.0001,
      "loss": 1.2464,
      "step": 3487
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.1629723757505417,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 3488
    },
    {
      "epoch": 0.5715923984272608,
      "grad_norm": 0.16686461865901947,
      "learning_rate": 0.0001,
      "loss": 1.4368,
      "step": 3489
    },
    {
      "epoch": 0.5717562254259502,
      "grad_norm": 0.16477103531360626,
      "learning_rate": 0.0001,
      "loss": 1.311,
      "step": 3490
    },
    {
      "epoch": 0.5719200524246396,
      "grad_norm": 0.1625678390264511,
      "learning_rate": 0.0001,
      "loss": 1.3935,
      "step": 3491
    },
    {
      "epoch": 0.572083879423329,
      "grad_norm": 0.15831215679645538,
      "learning_rate": 0.0001,
      "loss": 1.3248,
      "step": 3492
    },
    {
      "epoch": 0.5722477064220184,
      "grad_norm": 0.15909035503864288,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 3493
    },
    {
      "epoch": 0.5724115334207077,
      "grad_norm": 0.15749938786029816,
      "learning_rate": 0.0001,
      "loss": 1.3509,
      "step": 3494
    },
    {
      "epoch": 0.5725753604193972,
      "grad_norm": 0.16211825609207153,
      "learning_rate": 0.0001,
      "loss": 1.3098,
      "step": 3495
    },
    {
      "epoch": 0.5727391874180865,
      "grad_norm": 0.15274444222450256,
      "learning_rate": 0.0001,
      "loss": 1.3287,
      "step": 3496
    },
    {
      "epoch": 0.5729030144167759,
      "grad_norm": 0.15736955404281616,
      "learning_rate": 0.0001,
      "loss": 1.3813,
      "step": 3497
    },
    {
      "epoch": 0.5730668414154653,
      "grad_norm": 0.1627904325723648,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 3498
    },
    {
      "epoch": 0.5732306684141546,
      "grad_norm": 0.15620312094688416,
      "learning_rate": 0.0001,
      "loss": 1.4203,
      "step": 3499
    },
    {
      "epoch": 0.573394495412844,
      "grad_norm": 0.16375941038131714,
      "learning_rate": 0.0001,
      "loss": 1.371,
      "step": 3500
    },
    {
      "epoch": 0.5735583224115334,
      "grad_norm": 0.15363997220993042,
      "learning_rate": 0.0001,
      "loss": 1.364,
      "step": 3501
    },
    {
      "epoch": 0.5737221494102228,
      "grad_norm": 0.1579175591468811,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 3502
    },
    {
      "epoch": 0.5738859764089121,
      "grad_norm": 0.16805057227611542,
      "learning_rate": 0.0001,
      "loss": 1.3031,
      "step": 3503
    },
    {
      "epoch": 0.5740498034076016,
      "grad_norm": 0.16268852353096008,
      "learning_rate": 0.0001,
      "loss": 1.3834,
      "step": 3504
    },
    {
      "epoch": 0.5742136304062909,
      "grad_norm": 0.14674551784992218,
      "learning_rate": 0.0001,
      "loss": 1.3568,
      "step": 3505
    },
    {
      "epoch": 0.5743774574049804,
      "grad_norm": 0.16726160049438477,
      "learning_rate": 0.0001,
      "loss": 1.3573,
      "step": 3506
    },
    {
      "epoch": 0.5745412844036697,
      "grad_norm": 0.16070163249969482,
      "learning_rate": 0.0001,
      "loss": 1.3424,
      "step": 3507
    },
    {
      "epoch": 0.5747051114023591,
      "grad_norm": 0.1531953066587448,
      "learning_rate": 0.0001,
      "loss": 1.3896,
      "step": 3508
    },
    {
      "epoch": 0.5748689384010485,
      "grad_norm": 0.1657990962266922,
      "learning_rate": 0.0001,
      "loss": 1.439,
      "step": 3509
    },
    {
      "epoch": 0.5750327653997379,
      "grad_norm": 0.1607123464345932,
      "learning_rate": 0.0001,
      "loss": 1.42,
      "step": 3510
    },
    {
      "epoch": 0.5751965923984272,
      "grad_norm": 0.15924879908561707,
      "learning_rate": 0.0001,
      "loss": 1.4654,
      "step": 3511
    },
    {
      "epoch": 0.5753604193971167,
      "grad_norm": 0.15966913104057312,
      "learning_rate": 0.0001,
      "loss": 1.3207,
      "step": 3512
    },
    {
      "epoch": 0.575524246395806,
      "grad_norm": 0.15538710355758667,
      "learning_rate": 0.0001,
      "loss": 1.3845,
      "step": 3513
    },
    {
      "epoch": 0.5756880733944955,
      "grad_norm": 0.1467670202255249,
      "learning_rate": 0.0001,
      "loss": 1.3419,
      "step": 3514
    },
    {
      "epoch": 0.5758519003931848,
      "grad_norm": 0.16106076538562775,
      "learning_rate": 0.0001,
      "loss": 1.3829,
      "step": 3515
    },
    {
      "epoch": 0.5760157273918742,
      "grad_norm": 0.14968623220920563,
      "learning_rate": 0.0001,
      "loss": 1.2971,
      "step": 3516
    },
    {
      "epoch": 0.5761795543905636,
      "grad_norm": 0.15859009325504303,
      "learning_rate": 0.0001,
      "loss": 1.3584,
      "step": 3517
    },
    {
      "epoch": 0.5763433813892529,
      "grad_norm": 0.16297845542430878,
      "learning_rate": 0.0001,
      "loss": 1.4021,
      "step": 3518
    },
    {
      "epoch": 0.5765072083879423,
      "grad_norm": 0.16121050715446472,
      "learning_rate": 0.0001,
      "loss": 1.3759,
      "step": 3519
    },
    {
      "epoch": 0.5766710353866317,
      "grad_norm": 0.1657809168100357,
      "learning_rate": 0.0001,
      "loss": 1.398,
      "step": 3520
    },
    {
      "epoch": 0.5768348623853211,
      "grad_norm": 0.1724223792552948,
      "learning_rate": 0.0001,
      "loss": 1.2764,
      "step": 3521
    },
    {
      "epoch": 0.5769986893840104,
      "grad_norm": 0.15964095294475555,
      "learning_rate": 0.0001,
      "loss": 1.4627,
      "step": 3522
    },
    {
      "epoch": 0.5771625163826999,
      "grad_norm": 0.17176350951194763,
      "learning_rate": 0.0001,
      "loss": 1.3842,
      "step": 3523
    },
    {
      "epoch": 0.5773263433813892,
      "grad_norm": 0.169412761926651,
      "learning_rate": 0.0001,
      "loss": 1.3488,
      "step": 3524
    },
    {
      "epoch": 0.5774901703800787,
      "grad_norm": 0.1617916226387024,
      "learning_rate": 0.0001,
      "loss": 1.342,
      "step": 3525
    },
    {
      "epoch": 0.577653997378768,
      "grad_norm": 0.15583698451519012,
      "learning_rate": 0.0001,
      "loss": 1.4389,
      "step": 3526
    },
    {
      "epoch": 0.5778178243774574,
      "grad_norm": 0.1542515754699707,
      "learning_rate": 0.0001,
      "loss": 1.3862,
      "step": 3527
    },
    {
      "epoch": 0.5779816513761468,
      "grad_norm": 0.1638570874929428,
      "learning_rate": 0.0001,
      "loss": 1.4785,
      "step": 3528
    },
    {
      "epoch": 0.5781454783748362,
      "grad_norm": 0.16037791967391968,
      "learning_rate": 0.0001,
      "loss": 1.3866,
      "step": 3529
    },
    {
      "epoch": 0.5783093053735255,
      "grad_norm": 0.1604183465242386,
      "learning_rate": 0.0001,
      "loss": 1.332,
      "step": 3530
    },
    {
      "epoch": 0.578473132372215,
      "grad_norm": 0.14827312529087067,
      "learning_rate": 0.0001,
      "loss": 1.3593,
      "step": 3531
    },
    {
      "epoch": 0.5786369593709043,
      "grad_norm": 0.16556337475776672,
      "learning_rate": 0.0001,
      "loss": 1.3733,
      "step": 3532
    },
    {
      "epoch": 0.5788007863695938,
      "grad_norm": 0.16403087973594666,
      "learning_rate": 0.0001,
      "loss": 1.3242,
      "step": 3533
    },
    {
      "epoch": 0.5789646133682831,
      "grad_norm": 0.16672582924365997,
      "learning_rate": 0.0001,
      "loss": 1.3787,
      "step": 3534
    },
    {
      "epoch": 0.5791284403669725,
      "grad_norm": 0.17280122637748718,
      "learning_rate": 0.0001,
      "loss": 1.4032,
      "step": 3535
    },
    {
      "epoch": 0.5792922673656619,
      "grad_norm": 0.15663942694664001,
      "learning_rate": 0.0001,
      "loss": 1.3701,
      "step": 3536
    },
    {
      "epoch": 0.5794560943643512,
      "grad_norm": 0.16001290082931519,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 3537
    },
    {
      "epoch": 0.5796199213630406,
      "grad_norm": 0.15999019145965576,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 3538
    },
    {
      "epoch": 0.57978374836173,
      "grad_norm": 0.1533561646938324,
      "learning_rate": 0.0001,
      "loss": 1.4049,
      "step": 3539
    },
    {
      "epoch": 0.5799475753604194,
      "grad_norm": 0.16965171694755554,
      "learning_rate": 0.0001,
      "loss": 1.416,
      "step": 3540
    },
    {
      "epoch": 0.5801114023591087,
      "grad_norm": 0.1560611128807068,
      "learning_rate": 0.0001,
      "loss": 1.4351,
      "step": 3541
    },
    {
      "epoch": 0.5802752293577982,
      "grad_norm": 0.15770047903060913,
      "learning_rate": 0.0001,
      "loss": 1.2829,
      "step": 3542
    },
    {
      "epoch": 0.5804390563564875,
      "grad_norm": 0.16708381474018097,
      "learning_rate": 0.0001,
      "loss": 1.3516,
      "step": 3543
    },
    {
      "epoch": 0.580602883355177,
      "grad_norm": 0.16080616414546967,
      "learning_rate": 0.0001,
      "loss": 1.4364,
      "step": 3544
    },
    {
      "epoch": 0.5807667103538663,
      "grad_norm": 0.16277213394641876,
      "learning_rate": 0.0001,
      "loss": 1.4124,
      "step": 3545
    },
    {
      "epoch": 0.5809305373525557,
      "grad_norm": 0.15627549588680267,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 3546
    },
    {
      "epoch": 0.5810943643512451,
      "grad_norm": 0.16253221035003662,
      "learning_rate": 0.0001,
      "loss": 1.3742,
      "step": 3547
    },
    {
      "epoch": 0.5812581913499345,
      "grad_norm": 0.15959689021110535,
      "learning_rate": 0.0001,
      "loss": 1.4005,
      "step": 3548
    },
    {
      "epoch": 0.5814220183486238,
      "grad_norm": 0.16817806661128998,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 3549
    },
    {
      "epoch": 0.5815858453473133,
      "grad_norm": 0.15889140963554382,
      "learning_rate": 0.0001,
      "loss": 1.365,
      "step": 3550
    },
    {
      "epoch": 0.5817496723460026,
      "grad_norm": 0.16217680275440216,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 3551
    },
    {
      "epoch": 0.581913499344692,
      "grad_norm": 0.17311523854732513,
      "learning_rate": 0.0001,
      "loss": 1.2984,
      "step": 3552
    },
    {
      "epoch": 0.5820773263433814,
      "grad_norm": 0.16049879789352417,
      "learning_rate": 0.0001,
      "loss": 1.3418,
      "step": 3553
    },
    {
      "epoch": 0.5822411533420708,
      "grad_norm": 0.16331975162029266,
      "learning_rate": 0.0001,
      "loss": 1.3326,
      "step": 3554
    },
    {
      "epoch": 0.5824049803407602,
      "grad_norm": 0.1607385277748108,
      "learning_rate": 0.0001,
      "loss": 1.362,
      "step": 3555
    },
    {
      "epoch": 0.5825688073394495,
      "grad_norm": 0.16142834722995758,
      "learning_rate": 0.0001,
      "loss": 1.388,
      "step": 3556
    },
    {
      "epoch": 0.5827326343381389,
      "grad_norm": 0.16241391003131866,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 3557
    },
    {
      "epoch": 0.5828964613368283,
      "grad_norm": 0.15964794158935547,
      "learning_rate": 0.0001,
      "loss": 1.3713,
      "step": 3558
    },
    {
      "epoch": 0.5830602883355177,
      "grad_norm": 0.1550179421901703,
      "learning_rate": 0.0001,
      "loss": 1.337,
      "step": 3559
    },
    {
      "epoch": 0.583224115334207,
      "grad_norm": 0.15312352776527405,
      "learning_rate": 0.0001,
      "loss": 1.3506,
      "step": 3560
    },
    {
      "epoch": 0.5833879423328965,
      "grad_norm": 0.16736988723278046,
      "learning_rate": 0.0001,
      "loss": 1.2996,
      "step": 3561
    },
    {
      "epoch": 0.5835517693315858,
      "grad_norm": 0.15781083703041077,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 3562
    },
    {
      "epoch": 0.5837155963302753,
      "grad_norm": 0.15708714723587036,
      "learning_rate": 0.0001,
      "loss": 1.4033,
      "step": 3563
    },
    {
      "epoch": 0.5838794233289646,
      "grad_norm": 0.15661822259426117,
      "learning_rate": 0.0001,
      "loss": 1.3843,
      "step": 3564
    },
    {
      "epoch": 0.584043250327654,
      "grad_norm": 0.1678229123353958,
      "learning_rate": 0.0001,
      "loss": 1.4635,
      "step": 3565
    },
    {
      "epoch": 0.5842070773263434,
      "grad_norm": 0.15326310694217682,
      "learning_rate": 0.0001,
      "loss": 1.3099,
      "step": 3566
    },
    {
      "epoch": 0.5843709043250328,
      "grad_norm": 0.15655772387981415,
      "learning_rate": 0.0001,
      "loss": 1.3051,
      "step": 3567
    },
    {
      "epoch": 0.5845347313237221,
      "grad_norm": 0.1547226905822754,
      "learning_rate": 0.0001,
      "loss": 1.3371,
      "step": 3568
    },
    {
      "epoch": 0.5846985583224116,
      "grad_norm": 0.16027086973190308,
      "learning_rate": 0.0001,
      "loss": 1.3092,
      "step": 3569
    },
    {
      "epoch": 0.5848623853211009,
      "grad_norm": 0.153000146150589,
      "learning_rate": 0.0001,
      "loss": 1.3328,
      "step": 3570
    },
    {
      "epoch": 0.5850262123197904,
      "grad_norm": 0.16237233579158783,
      "learning_rate": 0.0001,
      "loss": 1.4298,
      "step": 3571
    },
    {
      "epoch": 0.5851900393184797,
      "grad_norm": 0.15854713320732117,
      "learning_rate": 0.0001,
      "loss": 1.403,
      "step": 3572
    },
    {
      "epoch": 0.5853538663171691,
      "grad_norm": 0.15837900340557098,
      "learning_rate": 0.0001,
      "loss": 1.2978,
      "step": 3573
    },
    {
      "epoch": 0.5855176933158585,
      "grad_norm": 0.16156664490699768,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 3574
    },
    {
      "epoch": 0.5856815203145478,
      "grad_norm": 0.1571023017168045,
      "learning_rate": 0.0001,
      "loss": 1.3362,
      "step": 3575
    },
    {
      "epoch": 0.5858453473132372,
      "grad_norm": 0.1617726981639862,
      "learning_rate": 0.0001,
      "loss": 1.442,
      "step": 3576
    },
    {
      "epoch": 0.5860091743119266,
      "grad_norm": 0.1627570390701294,
      "learning_rate": 0.0001,
      "loss": 1.3947,
      "step": 3577
    },
    {
      "epoch": 0.586173001310616,
      "grad_norm": 0.15581537783145905,
      "learning_rate": 0.0001,
      "loss": 1.3687,
      "step": 3578
    },
    {
      "epoch": 0.5863368283093053,
      "grad_norm": 0.1639634668827057,
      "learning_rate": 0.0001,
      "loss": 1.3821,
      "step": 3579
    },
    {
      "epoch": 0.5865006553079948,
      "grad_norm": 0.16096170246601105,
      "learning_rate": 0.0001,
      "loss": 1.3975,
      "step": 3580
    },
    {
      "epoch": 0.5866644823066841,
      "grad_norm": 0.1617354303598404,
      "learning_rate": 0.0001,
      "loss": 1.2984,
      "step": 3581
    },
    {
      "epoch": 0.5868283093053736,
      "grad_norm": 0.17029917240142822,
      "learning_rate": 0.0001,
      "loss": 1.4587,
      "step": 3582
    },
    {
      "epoch": 0.5869921363040629,
      "grad_norm": 0.1646362841129303,
      "learning_rate": 0.0001,
      "loss": 1.3191,
      "step": 3583
    },
    {
      "epoch": 0.5871559633027523,
      "grad_norm": 0.1806247979402542,
      "learning_rate": 0.0001,
      "loss": 1.3665,
      "step": 3584
    },
    {
      "epoch": 0.5873197903014417,
      "grad_norm": 0.1631089150905609,
      "learning_rate": 0.0001,
      "loss": 1.396,
      "step": 3585
    },
    {
      "epoch": 0.5874836173001311,
      "grad_norm": 0.15390431880950928,
      "learning_rate": 0.0001,
      "loss": 1.3008,
      "step": 3586
    },
    {
      "epoch": 0.5876474442988204,
      "grad_norm": 0.14781595766544342,
      "learning_rate": 0.0001,
      "loss": 1.294,
      "step": 3587
    },
    {
      "epoch": 0.5878112712975099,
      "grad_norm": 0.16499091684818268,
      "learning_rate": 0.0001,
      "loss": 1.3694,
      "step": 3588
    },
    {
      "epoch": 0.5879750982961992,
      "grad_norm": 0.16214604675769806,
      "learning_rate": 0.0001,
      "loss": 1.3248,
      "step": 3589
    },
    {
      "epoch": 0.5881389252948886,
      "grad_norm": 0.16219963133335114,
      "learning_rate": 0.0001,
      "loss": 1.4364,
      "step": 3590
    },
    {
      "epoch": 0.588302752293578,
      "grad_norm": 0.17594745755195618,
      "learning_rate": 0.0001,
      "loss": 1.4756,
      "step": 3591
    },
    {
      "epoch": 0.5884665792922673,
      "grad_norm": 0.16997532546520233,
      "learning_rate": 0.0001,
      "loss": 1.3423,
      "step": 3592
    },
    {
      "epoch": 0.5886304062909568,
      "grad_norm": 0.14852267503738403,
      "learning_rate": 0.0001,
      "loss": 1.2388,
      "step": 3593
    },
    {
      "epoch": 0.5887942332896461,
      "grad_norm": 0.1651380956172943,
      "learning_rate": 0.0001,
      "loss": 1.4277,
      "step": 3594
    },
    {
      "epoch": 0.5889580602883355,
      "grad_norm": 0.153810515999794,
      "learning_rate": 0.0001,
      "loss": 1.2899,
      "step": 3595
    },
    {
      "epoch": 0.5891218872870249,
      "grad_norm": 0.1786041259765625,
      "learning_rate": 0.0001,
      "loss": 1.4423,
      "step": 3596
    },
    {
      "epoch": 0.5892857142857143,
      "grad_norm": 0.16300401091575623,
      "learning_rate": 0.0001,
      "loss": 1.3314,
      "step": 3597
    },
    {
      "epoch": 0.5894495412844036,
      "grad_norm": 0.16128112375736237,
      "learning_rate": 0.0001,
      "loss": 1.4009,
      "step": 3598
    },
    {
      "epoch": 0.5896133682830931,
      "grad_norm": 0.15305539965629578,
      "learning_rate": 0.0001,
      "loss": 1.3661,
      "step": 3599
    },
    {
      "epoch": 0.5897771952817824,
      "grad_norm": 0.15958505868911743,
      "learning_rate": 0.0001,
      "loss": 1.4953,
      "step": 3600
    },
    {
      "epoch": 0.5899410222804718,
      "grad_norm": 0.16082057356834412,
      "learning_rate": 0.0001,
      "loss": 1.337,
      "step": 3601
    },
    {
      "epoch": 0.5901048492791612,
      "grad_norm": 0.16116824746131897,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 3602
    },
    {
      "epoch": 0.5902686762778506,
      "grad_norm": 0.16205045580863953,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 3603
    },
    {
      "epoch": 0.59043250327654,
      "grad_norm": 0.15429453551769257,
      "learning_rate": 0.0001,
      "loss": 1.3607,
      "step": 3604
    },
    {
      "epoch": 0.5905963302752294,
      "grad_norm": 0.15900547802448273,
      "learning_rate": 0.0001,
      "loss": 1.2912,
      "step": 3605
    },
    {
      "epoch": 0.5907601572739187,
      "grad_norm": 0.15585291385650635,
      "learning_rate": 0.0001,
      "loss": 1.3346,
      "step": 3606
    },
    {
      "epoch": 0.5909239842726082,
      "grad_norm": 0.15767419338226318,
      "learning_rate": 0.0001,
      "loss": 1.298,
      "step": 3607
    },
    {
      "epoch": 0.5910878112712975,
      "grad_norm": 0.18253372609615326,
      "learning_rate": 0.0001,
      "loss": 1.4171,
      "step": 3608
    },
    {
      "epoch": 0.591251638269987,
      "grad_norm": 0.15812955796718597,
      "learning_rate": 0.0001,
      "loss": 1.4007,
      "step": 3609
    },
    {
      "epoch": 0.5914154652686763,
      "grad_norm": 0.15635864436626434,
      "learning_rate": 0.0001,
      "loss": 1.3358,
      "step": 3610
    },
    {
      "epoch": 0.5915792922673656,
      "grad_norm": 0.15530209243297577,
      "learning_rate": 0.0001,
      "loss": 1.4199,
      "step": 3611
    },
    {
      "epoch": 0.591743119266055,
      "grad_norm": 0.1671627014875412,
      "learning_rate": 0.0001,
      "loss": 1.4374,
      "step": 3612
    },
    {
      "epoch": 0.5919069462647444,
      "grad_norm": 0.16412293910980225,
      "learning_rate": 0.0001,
      "loss": 1.3698,
      "step": 3613
    },
    {
      "epoch": 0.5920707732634338,
      "grad_norm": 0.162236288189888,
      "learning_rate": 0.0001,
      "loss": 1.3878,
      "step": 3614
    },
    {
      "epoch": 0.5922346002621232,
      "grad_norm": 0.1575198918581009,
      "learning_rate": 0.0001,
      "loss": 1.4249,
      "step": 3615
    },
    {
      "epoch": 0.5923984272608126,
      "grad_norm": 0.15733551979064941,
      "learning_rate": 0.0001,
      "loss": 1.3796,
      "step": 3616
    },
    {
      "epoch": 0.5925622542595019,
      "grad_norm": 0.16507184505462646,
      "learning_rate": 0.0001,
      "loss": 1.414,
      "step": 3617
    },
    {
      "epoch": 0.5927260812581914,
      "grad_norm": 0.16516627371311188,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 3618
    },
    {
      "epoch": 0.5928899082568807,
      "grad_norm": 0.16318774223327637,
      "learning_rate": 0.0001,
      "loss": 1.3368,
      "step": 3619
    },
    {
      "epoch": 0.5930537352555701,
      "grad_norm": 0.16132527589797974,
      "learning_rate": 0.0001,
      "loss": 1.3838,
      "step": 3620
    },
    {
      "epoch": 0.5932175622542595,
      "grad_norm": 0.16594849526882172,
      "learning_rate": 0.0001,
      "loss": 1.328,
      "step": 3621
    },
    {
      "epoch": 0.5933813892529489,
      "grad_norm": 0.160101518034935,
      "learning_rate": 0.0001,
      "loss": 1.4789,
      "step": 3622
    },
    {
      "epoch": 0.5935452162516383,
      "grad_norm": 0.16618841886520386,
      "learning_rate": 0.0001,
      "loss": 1.3615,
      "step": 3623
    },
    {
      "epoch": 0.5937090432503277,
      "grad_norm": 0.16838882863521576,
      "learning_rate": 0.0001,
      "loss": 1.4511,
      "step": 3624
    },
    {
      "epoch": 0.593872870249017,
      "grad_norm": 0.17587871849536896,
      "learning_rate": 0.0001,
      "loss": 1.3718,
      "step": 3625
    },
    {
      "epoch": 0.5940366972477065,
      "grad_norm": 0.1663283407688141,
      "learning_rate": 0.0001,
      "loss": 1.4323,
      "step": 3626
    },
    {
      "epoch": 0.5942005242463958,
      "grad_norm": 0.1637425422668457,
      "learning_rate": 0.0001,
      "loss": 1.4532,
      "step": 3627
    },
    {
      "epoch": 0.5943643512450852,
      "grad_norm": 0.16072359681129456,
      "learning_rate": 0.0001,
      "loss": 1.3668,
      "step": 3628
    },
    {
      "epoch": 0.5945281782437746,
      "grad_norm": 0.1670219898223877,
      "learning_rate": 0.0001,
      "loss": 1.3957,
      "step": 3629
    },
    {
      "epoch": 0.5946920052424639,
      "grad_norm": 0.17261886596679688,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 3630
    },
    {
      "epoch": 0.5948558322411533,
      "grad_norm": 0.15730971097946167,
      "learning_rate": 0.0001,
      "loss": 1.3177,
      "step": 3631
    },
    {
      "epoch": 0.5950196592398427,
      "grad_norm": 0.16900749504566193,
      "learning_rate": 0.0001,
      "loss": 1.3771,
      "step": 3632
    },
    {
      "epoch": 0.5951834862385321,
      "grad_norm": 0.16630108654499054,
      "learning_rate": 0.0001,
      "loss": 1.3983,
      "step": 3633
    },
    {
      "epoch": 0.5953473132372215,
      "grad_norm": 0.15692882239818573,
      "learning_rate": 0.0001,
      "loss": 1.3699,
      "step": 3634
    },
    {
      "epoch": 0.5955111402359109,
      "grad_norm": 0.1689680963754654,
      "learning_rate": 0.0001,
      "loss": 1.3699,
      "step": 3635
    },
    {
      "epoch": 0.5956749672346002,
      "grad_norm": 0.165337935090065,
      "learning_rate": 0.0001,
      "loss": 1.4091,
      "step": 3636
    },
    {
      "epoch": 0.5958387942332897,
      "grad_norm": 0.16824886202812195,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 3637
    },
    {
      "epoch": 0.596002621231979,
      "grad_norm": 0.16711395978927612,
      "learning_rate": 0.0001,
      "loss": 1.4654,
      "step": 3638
    },
    {
      "epoch": 0.5961664482306684,
      "grad_norm": 0.16922485828399658,
      "learning_rate": 0.0001,
      "loss": 1.3326,
      "step": 3639
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 0.16210010647773743,
      "learning_rate": 0.0001,
      "loss": 1.331,
      "step": 3640
    },
    {
      "epoch": 0.5964941022280472,
      "grad_norm": 0.15186208486557007,
      "learning_rate": 0.0001,
      "loss": 1.2971,
      "step": 3641
    },
    {
      "epoch": 0.5966579292267365,
      "grad_norm": 0.1555698961019516,
      "learning_rate": 0.0001,
      "loss": 1.3235,
      "step": 3642
    },
    {
      "epoch": 0.596821756225426,
      "grad_norm": 0.164046049118042,
      "learning_rate": 0.0001,
      "loss": 1.363,
      "step": 3643
    },
    {
      "epoch": 0.5969855832241153,
      "grad_norm": 0.16057837009429932,
      "learning_rate": 0.0001,
      "loss": 1.4443,
      "step": 3644
    },
    {
      "epoch": 0.5971494102228048,
      "grad_norm": 0.16355347633361816,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 3645
    },
    {
      "epoch": 0.5973132372214941,
      "grad_norm": 0.16517293453216553,
      "learning_rate": 0.0001,
      "loss": 1.4029,
      "step": 3646
    },
    {
      "epoch": 0.5974770642201835,
      "grad_norm": 0.15456578135490417,
      "learning_rate": 0.0001,
      "loss": 1.3808,
      "step": 3647
    },
    {
      "epoch": 0.5976408912188729,
      "grad_norm": 0.16439150273799896,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 3648
    },
    {
      "epoch": 0.5978047182175622,
      "grad_norm": 0.1582053005695343,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 3649
    },
    {
      "epoch": 0.5979685452162516,
      "grad_norm": 0.1605139970779419,
      "learning_rate": 0.0001,
      "loss": 1.3482,
      "step": 3650
    },
    {
      "epoch": 0.598132372214941,
      "grad_norm": 0.17071224749088287,
      "learning_rate": 0.0001,
      "loss": 1.4413,
      "step": 3651
    },
    {
      "epoch": 0.5982961992136304,
      "grad_norm": 0.1619299352169037,
      "learning_rate": 0.0001,
      "loss": 1.4265,
      "step": 3652
    },
    {
      "epoch": 0.5984600262123198,
      "grad_norm": 0.1714302897453308,
      "learning_rate": 0.0001,
      "loss": 1.412,
      "step": 3653
    },
    {
      "epoch": 0.5986238532110092,
      "grad_norm": 0.1685044765472412,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 3654
    },
    {
      "epoch": 0.5987876802096985,
      "grad_norm": 0.1597050428390503,
      "learning_rate": 0.0001,
      "loss": 1.3776,
      "step": 3655
    },
    {
      "epoch": 0.598951507208388,
      "grad_norm": 0.16280773282051086,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 3656
    },
    {
      "epoch": 0.5991153342070773,
      "grad_norm": 0.15168964862823486,
      "learning_rate": 0.0001,
      "loss": 1.2688,
      "step": 3657
    },
    {
      "epoch": 0.5992791612057667,
      "grad_norm": 0.16271471977233887,
      "learning_rate": 0.0001,
      "loss": 1.3286,
      "step": 3658
    },
    {
      "epoch": 0.5994429882044561,
      "grad_norm": 0.15777462720870972,
      "learning_rate": 0.0001,
      "loss": 1.4365,
      "step": 3659
    },
    {
      "epoch": 0.5996068152031455,
      "grad_norm": 0.1667216271162033,
      "learning_rate": 0.0001,
      "loss": 1.4854,
      "step": 3660
    },
    {
      "epoch": 0.5997706422018348,
      "grad_norm": 0.15806108713150024,
      "learning_rate": 0.0001,
      "loss": 1.3208,
      "step": 3661
    },
    {
      "epoch": 0.5999344692005243,
      "grad_norm": 0.162210151553154,
      "learning_rate": 0.0001,
      "loss": 1.4743,
      "step": 3662
    },
    {
      "epoch": 0.6000982961992136,
      "grad_norm": 0.15225562453269958,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 3663
    },
    {
      "epoch": 0.6002621231979031,
      "grad_norm": 0.16304782032966614,
      "learning_rate": 0.0001,
      "loss": 1.3382,
      "step": 3664
    },
    {
      "epoch": 0.6004259501965924,
      "grad_norm": 0.16091269254684448,
      "learning_rate": 0.0001,
      "loss": 1.4059,
      "step": 3665
    },
    {
      "epoch": 0.6005897771952818,
      "grad_norm": 0.18380992114543915,
      "learning_rate": 0.0001,
      "loss": 1.37,
      "step": 3666
    },
    {
      "epoch": 0.6007536041939712,
      "grad_norm": 0.15399113297462463,
      "learning_rate": 0.0001,
      "loss": 1.3189,
      "step": 3667
    },
    {
      "epoch": 0.6009174311926605,
      "grad_norm": 0.15323664247989655,
      "learning_rate": 0.0001,
      "loss": 1.27,
      "step": 3668
    },
    {
      "epoch": 0.6010812581913499,
      "grad_norm": 0.15504182875156403,
      "learning_rate": 0.0001,
      "loss": 1.3972,
      "step": 3669
    },
    {
      "epoch": 0.6012450851900393,
      "grad_norm": 0.16522178053855896,
      "learning_rate": 0.0001,
      "loss": 1.477,
      "step": 3670
    },
    {
      "epoch": 0.6014089121887287,
      "grad_norm": 0.16428308188915253,
      "learning_rate": 0.0001,
      "loss": 1.4094,
      "step": 3671
    },
    {
      "epoch": 0.601572739187418,
      "grad_norm": 0.15573526918888092,
      "learning_rate": 0.0001,
      "loss": 1.3117,
      "step": 3672
    },
    {
      "epoch": 0.6017365661861075,
      "grad_norm": 0.1822320520877838,
      "learning_rate": 0.0001,
      "loss": 1.2705,
      "step": 3673
    },
    {
      "epoch": 0.6019003931847968,
      "grad_norm": 0.1637672334909439,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 3674
    },
    {
      "epoch": 0.6020642201834863,
      "grad_norm": 0.1546907126903534,
      "learning_rate": 0.0001,
      "loss": 1.4238,
      "step": 3675
    },
    {
      "epoch": 0.6022280471821756,
      "grad_norm": 0.15344905853271484,
      "learning_rate": 0.0001,
      "loss": 1.2289,
      "step": 3676
    },
    {
      "epoch": 0.602391874180865,
      "grad_norm": 0.16311241686344147,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 3677
    },
    {
      "epoch": 0.6025557011795544,
      "grad_norm": 0.16107681393623352,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 3678
    },
    {
      "epoch": 0.6027195281782438,
      "grad_norm": 0.15223529934883118,
      "learning_rate": 0.0001,
      "loss": 1.3275,
      "step": 3679
    },
    {
      "epoch": 0.6028833551769331,
      "grad_norm": 0.16483132541179657,
      "learning_rate": 0.0001,
      "loss": 1.3682,
      "step": 3680
    },
    {
      "epoch": 0.6030471821756226,
      "grad_norm": 0.1691090613603592,
      "learning_rate": 0.0001,
      "loss": 1.3919,
      "step": 3681
    },
    {
      "epoch": 0.6032110091743119,
      "grad_norm": 0.1574559360742569,
      "learning_rate": 0.0001,
      "loss": 1.4415,
      "step": 3682
    },
    {
      "epoch": 0.6033748361730014,
      "grad_norm": 0.16925790905952454,
      "learning_rate": 0.0001,
      "loss": 1.401,
      "step": 3683
    },
    {
      "epoch": 0.6035386631716907,
      "grad_norm": 0.17385122179985046,
      "learning_rate": 0.0001,
      "loss": 1.3799,
      "step": 3684
    },
    {
      "epoch": 0.60370249017038,
      "grad_norm": 0.16531826555728912,
      "learning_rate": 0.0001,
      "loss": 1.4126,
      "step": 3685
    },
    {
      "epoch": 0.6038663171690695,
      "grad_norm": 0.17547836899757385,
      "learning_rate": 0.0001,
      "loss": 1.4922,
      "step": 3686
    },
    {
      "epoch": 0.6040301441677588,
      "grad_norm": 0.16164405643939972,
      "learning_rate": 0.0001,
      "loss": 1.3847,
      "step": 3687
    },
    {
      "epoch": 0.6041939711664482,
      "grad_norm": 0.16635198891162872,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 3688
    },
    {
      "epoch": 0.6043577981651376,
      "grad_norm": 0.16672852635383606,
      "learning_rate": 0.0001,
      "loss": 1.3703,
      "step": 3689
    },
    {
      "epoch": 0.604521625163827,
      "grad_norm": 0.1582021564245224,
      "learning_rate": 0.0001,
      "loss": 1.3242,
      "step": 3690
    },
    {
      "epoch": 0.6046854521625163,
      "grad_norm": 0.16189585626125336,
      "learning_rate": 0.0001,
      "loss": 1.347,
      "step": 3691
    },
    {
      "epoch": 0.6048492791612058,
      "grad_norm": 0.17320604622364044,
      "learning_rate": 0.0001,
      "loss": 1.3836,
      "step": 3692
    },
    {
      "epoch": 0.6050131061598951,
      "grad_norm": 0.15775029361248016,
      "learning_rate": 0.0001,
      "loss": 1.4028,
      "step": 3693
    },
    {
      "epoch": 0.6051769331585846,
      "grad_norm": 0.1627461314201355,
      "learning_rate": 0.0001,
      "loss": 1.4527,
      "step": 3694
    },
    {
      "epoch": 0.6053407601572739,
      "grad_norm": 0.20082619786262512,
      "learning_rate": 0.0001,
      "loss": 1.3079,
      "step": 3695
    },
    {
      "epoch": 0.6055045871559633,
      "grad_norm": 0.16805824637413025,
      "learning_rate": 0.0001,
      "loss": 1.412,
      "step": 3696
    },
    {
      "epoch": 0.6056684141546527,
      "grad_norm": 0.1676231026649475,
      "learning_rate": 0.0001,
      "loss": 1.3422,
      "step": 3697
    },
    {
      "epoch": 0.6058322411533421,
      "grad_norm": 0.16322234272956848,
      "learning_rate": 0.0001,
      "loss": 1.3694,
      "step": 3698
    },
    {
      "epoch": 0.6059960681520314,
      "grad_norm": 0.1663224995136261,
      "learning_rate": 0.0001,
      "loss": 1.3345,
      "step": 3699
    },
    {
      "epoch": 0.6061598951507209,
      "grad_norm": 0.15938784182071686,
      "learning_rate": 0.0001,
      "loss": 1.4022,
      "step": 3700
    },
    {
      "epoch": 0.6063237221494102,
      "grad_norm": 0.15183177590370178,
      "learning_rate": 0.0001,
      "loss": 1.3208,
      "step": 3701
    },
    {
      "epoch": 0.6064875491480997,
      "grad_norm": 0.1668749749660492,
      "learning_rate": 0.0001,
      "loss": 1.3955,
      "step": 3702
    },
    {
      "epoch": 0.606651376146789,
      "grad_norm": 0.1630232185125351,
      "learning_rate": 0.0001,
      "loss": 1.2804,
      "step": 3703
    },
    {
      "epoch": 0.6068152031454783,
      "grad_norm": 0.15355725586414337,
      "learning_rate": 0.0001,
      "loss": 1.3961,
      "step": 3704
    },
    {
      "epoch": 0.6069790301441678,
      "grad_norm": 0.17119291424751282,
      "learning_rate": 0.0001,
      "loss": 1.4759,
      "step": 3705
    },
    {
      "epoch": 0.6071428571428571,
      "grad_norm": 0.16861537098884583,
      "learning_rate": 0.0001,
      "loss": 1.3855,
      "step": 3706
    },
    {
      "epoch": 0.6073066841415465,
      "grad_norm": 0.16313588619232178,
      "learning_rate": 0.0001,
      "loss": 1.3928,
      "step": 3707
    },
    {
      "epoch": 0.6074705111402359,
      "grad_norm": 0.18417809903621674,
      "learning_rate": 0.0001,
      "loss": 1.4426,
      "step": 3708
    },
    {
      "epoch": 0.6076343381389253,
      "grad_norm": 0.16348007321357727,
      "learning_rate": 0.0001,
      "loss": 1.308,
      "step": 3709
    },
    {
      "epoch": 0.6077981651376146,
      "grad_norm": 0.16263221204280853,
      "learning_rate": 0.0001,
      "loss": 1.3937,
      "step": 3710
    },
    {
      "epoch": 0.6079619921363041,
      "grad_norm": 0.15671859681606293,
      "learning_rate": 0.0001,
      "loss": 1.3787,
      "step": 3711
    },
    {
      "epoch": 0.6081258191349934,
      "grad_norm": 0.15477213263511658,
      "learning_rate": 0.0001,
      "loss": 1.3873,
      "step": 3712
    },
    {
      "epoch": 0.6082896461336829,
      "grad_norm": 0.15409733355045319,
      "learning_rate": 0.0001,
      "loss": 1.2688,
      "step": 3713
    },
    {
      "epoch": 0.6084534731323722,
      "grad_norm": 0.16555291414260864,
      "learning_rate": 0.0001,
      "loss": 1.4287,
      "step": 3714
    },
    {
      "epoch": 0.6086173001310616,
      "grad_norm": 0.16315937042236328,
      "learning_rate": 0.0001,
      "loss": 1.4625,
      "step": 3715
    },
    {
      "epoch": 0.608781127129751,
      "grad_norm": 0.16837675869464874,
      "learning_rate": 0.0001,
      "loss": 1.4403,
      "step": 3716
    },
    {
      "epoch": 0.6089449541284404,
      "grad_norm": 0.16073638200759888,
      "learning_rate": 0.0001,
      "loss": 1.2886,
      "step": 3717
    },
    {
      "epoch": 0.6091087811271297,
      "grad_norm": 0.15579545497894287,
      "learning_rate": 0.0001,
      "loss": 1.3593,
      "step": 3718
    },
    {
      "epoch": 0.6092726081258192,
      "grad_norm": 0.15356241166591644,
      "learning_rate": 0.0001,
      "loss": 1.2614,
      "step": 3719
    },
    {
      "epoch": 0.6094364351245085,
      "grad_norm": 0.16653624176979065,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 3720
    },
    {
      "epoch": 0.609600262123198,
      "grad_norm": 0.16003188490867615,
      "learning_rate": 0.0001,
      "loss": 1.395,
      "step": 3721
    },
    {
      "epoch": 0.6097640891218873,
      "grad_norm": 0.1592954844236374,
      "learning_rate": 0.0001,
      "loss": 1.4293,
      "step": 3722
    },
    {
      "epoch": 0.6099279161205766,
      "grad_norm": 0.15839873254299164,
      "learning_rate": 0.0001,
      "loss": 1.4451,
      "step": 3723
    },
    {
      "epoch": 0.6100917431192661,
      "grad_norm": 0.1585712432861328,
      "learning_rate": 0.0001,
      "loss": 1.3963,
      "step": 3724
    },
    {
      "epoch": 0.6102555701179554,
      "grad_norm": 0.15933899581432343,
      "learning_rate": 0.0001,
      "loss": 1.3828,
      "step": 3725
    },
    {
      "epoch": 0.6104193971166448,
      "grad_norm": 0.15352033078670502,
      "learning_rate": 0.0001,
      "loss": 1.2876,
      "step": 3726
    },
    {
      "epoch": 0.6105832241153342,
      "grad_norm": 0.16043050587177277,
      "learning_rate": 0.0001,
      "loss": 1.3851,
      "step": 3727
    },
    {
      "epoch": 0.6107470511140236,
      "grad_norm": 0.1746024489402771,
      "learning_rate": 0.0001,
      "loss": 1.3842,
      "step": 3728
    },
    {
      "epoch": 0.6109108781127129,
      "grad_norm": 0.1769411563873291,
      "learning_rate": 0.0001,
      "loss": 1.4131,
      "step": 3729
    },
    {
      "epoch": 0.6110747051114024,
      "grad_norm": 0.1708727478981018,
      "learning_rate": 0.0001,
      "loss": 1.3924,
      "step": 3730
    },
    {
      "epoch": 0.6112385321100917,
      "grad_norm": 0.15929819643497467,
      "learning_rate": 0.0001,
      "loss": 1.2023,
      "step": 3731
    },
    {
      "epoch": 0.6114023591087812,
      "grad_norm": 0.1577625721693039,
      "learning_rate": 0.0001,
      "loss": 1.2481,
      "step": 3732
    },
    {
      "epoch": 0.6115661861074705,
      "grad_norm": 0.16204862296581268,
      "learning_rate": 0.0001,
      "loss": 1.3161,
      "step": 3733
    },
    {
      "epoch": 0.6117300131061599,
      "grad_norm": 0.16914305090904236,
      "learning_rate": 0.0001,
      "loss": 1.3321,
      "step": 3734
    },
    {
      "epoch": 0.6118938401048493,
      "grad_norm": 0.15865738689899445,
      "learning_rate": 0.0001,
      "loss": 1.3133,
      "step": 3735
    },
    {
      "epoch": 0.6120576671035387,
      "grad_norm": 0.1666552722454071,
      "learning_rate": 0.0001,
      "loss": 1.2602,
      "step": 3736
    },
    {
      "epoch": 0.612221494102228,
      "grad_norm": 0.15728642046451569,
      "learning_rate": 0.0001,
      "loss": 1.3303,
      "step": 3737
    },
    {
      "epoch": 0.6123853211009175,
      "grad_norm": 0.18357910215854645,
      "learning_rate": 0.0001,
      "loss": 1.5437,
      "step": 3738
    },
    {
      "epoch": 0.6125491480996068,
      "grad_norm": 0.15752705931663513,
      "learning_rate": 0.0001,
      "loss": 1.3344,
      "step": 3739
    },
    {
      "epoch": 0.6127129750982963,
      "grad_norm": 0.15462592244148254,
      "learning_rate": 0.0001,
      "loss": 1.3547,
      "step": 3740
    },
    {
      "epoch": 0.6128768020969856,
      "grad_norm": 0.16800232231616974,
      "learning_rate": 0.0001,
      "loss": 1.3813,
      "step": 3741
    },
    {
      "epoch": 0.6130406290956749,
      "grad_norm": 0.1608123779296875,
      "learning_rate": 0.0001,
      "loss": 1.4042,
      "step": 3742
    },
    {
      "epoch": 0.6132044560943644,
      "grad_norm": 0.1649903655052185,
      "learning_rate": 0.0001,
      "loss": 1.3941,
      "step": 3743
    },
    {
      "epoch": 0.6133682830930537,
      "grad_norm": 0.15392906963825226,
      "learning_rate": 0.0001,
      "loss": 1.2976,
      "step": 3744
    },
    {
      "epoch": 0.6135321100917431,
      "grad_norm": 0.1638864427804947,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 3745
    },
    {
      "epoch": 0.6136959370904325,
      "grad_norm": 0.1607896387577057,
      "learning_rate": 0.0001,
      "loss": 1.3509,
      "step": 3746
    },
    {
      "epoch": 0.6138597640891219,
      "grad_norm": 0.1577974259853363,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 3747
    },
    {
      "epoch": 0.6140235910878112,
      "grad_norm": 0.16560427844524384,
      "learning_rate": 0.0001,
      "loss": 1.2897,
      "step": 3748
    },
    {
      "epoch": 0.6141874180865007,
      "grad_norm": 0.17071999609470367,
      "learning_rate": 0.0001,
      "loss": 1.4076,
      "step": 3749
    },
    {
      "epoch": 0.61435124508519,
      "grad_norm": 0.17851218581199646,
      "learning_rate": 0.0001,
      "loss": 1.4434,
      "step": 3750
    },
    {
      "epoch": 0.6145150720838795,
      "grad_norm": 0.15679679811000824,
      "learning_rate": 0.0001,
      "loss": 1.3734,
      "step": 3751
    },
    {
      "epoch": 0.6146788990825688,
      "grad_norm": 0.16832293570041656,
      "learning_rate": 0.0001,
      "loss": 1.3805,
      "step": 3752
    },
    {
      "epoch": 0.6148427260812582,
      "grad_norm": 0.15970708429813385,
      "learning_rate": 0.0001,
      "loss": 1.373,
      "step": 3753
    },
    {
      "epoch": 0.6150065530799476,
      "grad_norm": 0.15697123110294342,
      "learning_rate": 0.0001,
      "loss": 1.3997,
      "step": 3754
    },
    {
      "epoch": 0.615170380078637,
      "grad_norm": 0.15812325477600098,
      "learning_rate": 0.0001,
      "loss": 1.2754,
      "step": 3755
    },
    {
      "epoch": 0.6153342070773263,
      "grad_norm": 0.15327519178390503,
      "learning_rate": 0.0001,
      "loss": 1.3007,
      "step": 3756
    },
    {
      "epoch": 0.6154980340760158,
      "grad_norm": 0.1835162192583084,
      "learning_rate": 0.0001,
      "loss": 1.3598,
      "step": 3757
    },
    {
      "epoch": 0.6156618610747051,
      "grad_norm": 0.16579587757587433,
      "learning_rate": 0.0001,
      "loss": 1.4288,
      "step": 3758
    },
    {
      "epoch": 0.6158256880733946,
      "grad_norm": 0.1570982038974762,
      "learning_rate": 0.0001,
      "loss": 1.4257,
      "step": 3759
    },
    {
      "epoch": 0.6159895150720839,
      "grad_norm": 0.16818179190158844,
      "learning_rate": 0.0001,
      "loss": 1.3124,
      "step": 3760
    },
    {
      "epoch": 0.6161533420707732,
      "grad_norm": 0.16095368564128876,
      "learning_rate": 0.0001,
      "loss": 1.4513,
      "step": 3761
    },
    {
      "epoch": 0.6163171690694627,
      "grad_norm": 0.16379491984844208,
      "learning_rate": 0.0001,
      "loss": 1.3802,
      "step": 3762
    },
    {
      "epoch": 0.616480996068152,
      "grad_norm": 0.17470639944076538,
      "learning_rate": 0.0001,
      "loss": 1.376,
      "step": 3763
    },
    {
      "epoch": 0.6166448230668414,
      "grad_norm": 0.1616041511297226,
      "learning_rate": 0.0001,
      "loss": 1.3035,
      "step": 3764
    },
    {
      "epoch": 0.6168086500655308,
      "grad_norm": 0.16732747852802277,
      "learning_rate": 0.0001,
      "loss": 1.3809,
      "step": 3765
    },
    {
      "epoch": 0.6169724770642202,
      "grad_norm": 0.17258885502815247,
      "learning_rate": 0.0001,
      "loss": 1.332,
      "step": 3766
    },
    {
      "epoch": 0.6171363040629095,
      "grad_norm": 0.16323073208332062,
      "learning_rate": 0.0001,
      "loss": 1.3295,
      "step": 3767
    },
    {
      "epoch": 0.617300131061599,
      "grad_norm": 0.1554909199476242,
      "learning_rate": 0.0001,
      "loss": 1.2879,
      "step": 3768
    },
    {
      "epoch": 0.6174639580602883,
      "grad_norm": 0.16617965698242188,
      "learning_rate": 0.0001,
      "loss": 1.3552,
      "step": 3769
    },
    {
      "epoch": 0.6176277850589778,
      "grad_norm": 0.1558120995759964,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 3770
    },
    {
      "epoch": 0.6177916120576671,
      "grad_norm": 0.15738679468631744,
      "learning_rate": 0.0001,
      "loss": 1.3895,
      "step": 3771
    },
    {
      "epoch": 0.6179554390563565,
      "grad_norm": 0.17592395842075348,
      "learning_rate": 0.0001,
      "loss": 1.3567,
      "step": 3772
    },
    {
      "epoch": 0.6181192660550459,
      "grad_norm": 0.16305215656757355,
      "learning_rate": 0.0001,
      "loss": 1.4742,
      "step": 3773
    },
    {
      "epoch": 0.6182830930537353,
      "grad_norm": 0.16653096675872803,
      "learning_rate": 0.0001,
      "loss": 1.3061,
      "step": 3774
    },
    {
      "epoch": 0.6184469200524246,
      "grad_norm": 0.15758787095546722,
      "learning_rate": 0.0001,
      "loss": 1.3159,
      "step": 3775
    },
    {
      "epoch": 0.6186107470511141,
      "grad_norm": 0.17045369744300842,
      "learning_rate": 0.0001,
      "loss": 1.4267,
      "step": 3776
    },
    {
      "epoch": 0.6187745740498034,
      "grad_norm": 0.15648996829986572,
      "learning_rate": 0.0001,
      "loss": 1.358,
      "step": 3777
    },
    {
      "epoch": 0.6189384010484927,
      "grad_norm": 0.1617046445608139,
      "learning_rate": 0.0001,
      "loss": 1.3502,
      "step": 3778
    },
    {
      "epoch": 0.6191022280471822,
      "grad_norm": 0.16152484714984894,
      "learning_rate": 0.0001,
      "loss": 1.3602,
      "step": 3779
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 0.168503075838089,
      "learning_rate": 0.0001,
      "loss": 1.397,
      "step": 3780
    },
    {
      "epoch": 0.619429882044561,
      "grad_norm": 0.1608007252216339,
      "learning_rate": 0.0001,
      "loss": 1.3764,
      "step": 3781
    },
    {
      "epoch": 0.6195937090432503,
      "grad_norm": 0.16987119615077972,
      "learning_rate": 0.0001,
      "loss": 1.388,
      "step": 3782
    },
    {
      "epoch": 0.6197575360419397,
      "grad_norm": 0.16456356644630432,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 3783
    },
    {
      "epoch": 0.6199213630406291,
      "grad_norm": 0.1657218039035797,
      "learning_rate": 0.0001,
      "loss": 1.3924,
      "step": 3784
    },
    {
      "epoch": 0.6200851900393185,
      "grad_norm": 0.16665807366371155,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 3785
    },
    {
      "epoch": 0.6202490170380078,
      "grad_norm": 0.1639365702867508,
      "learning_rate": 0.0001,
      "loss": 1.387,
      "step": 3786
    },
    {
      "epoch": 0.6204128440366973,
      "grad_norm": 0.1643993854522705,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 3787
    },
    {
      "epoch": 0.6205766710353866,
      "grad_norm": 0.1584642082452774,
      "learning_rate": 0.0001,
      "loss": 1.3743,
      "step": 3788
    },
    {
      "epoch": 0.620740498034076,
      "grad_norm": 0.15493004024028778,
      "learning_rate": 0.0001,
      "loss": 1.3396,
      "step": 3789
    },
    {
      "epoch": 0.6209043250327654,
      "grad_norm": 0.16875435411930084,
      "learning_rate": 0.0001,
      "loss": 1.2648,
      "step": 3790
    },
    {
      "epoch": 0.6210681520314548,
      "grad_norm": 0.16324414312839508,
      "learning_rate": 0.0001,
      "loss": 1.4619,
      "step": 3791
    },
    {
      "epoch": 0.6212319790301442,
      "grad_norm": 0.170904740691185,
      "learning_rate": 0.0001,
      "loss": 1.4633,
      "step": 3792
    },
    {
      "epoch": 0.6213958060288336,
      "grad_norm": 0.16501669585704803,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 3793
    },
    {
      "epoch": 0.6215596330275229,
      "grad_norm": 0.15866239368915558,
      "learning_rate": 0.0001,
      "loss": 1.4144,
      "step": 3794
    },
    {
      "epoch": 0.6217234600262124,
      "grad_norm": 0.16114723682403564,
      "learning_rate": 0.0001,
      "loss": 1.4881,
      "step": 3795
    },
    {
      "epoch": 0.6218872870249017,
      "grad_norm": 0.17188474535942078,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 3796
    },
    {
      "epoch": 0.622051114023591,
      "grad_norm": 0.17615817487239838,
      "learning_rate": 0.0001,
      "loss": 1.3902,
      "step": 3797
    },
    {
      "epoch": 0.6222149410222805,
      "grad_norm": 0.170726016163826,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 3798
    },
    {
      "epoch": 0.6223787680209698,
      "grad_norm": 0.1615934669971466,
      "learning_rate": 0.0001,
      "loss": 1.4716,
      "step": 3799
    },
    {
      "epoch": 0.6225425950196593,
      "grad_norm": 0.16659456491470337,
      "learning_rate": 0.0001,
      "loss": 1.3537,
      "step": 3800
    },
    {
      "epoch": 0.6227064220183486,
      "grad_norm": 0.158842533826828,
      "learning_rate": 0.0001,
      "loss": 1.4343,
      "step": 3801
    },
    {
      "epoch": 0.622870249017038,
      "grad_norm": 0.1644839346408844,
      "learning_rate": 0.0001,
      "loss": 1.3946,
      "step": 3802
    },
    {
      "epoch": 0.6230340760157274,
      "grad_norm": 0.16246142983436584,
      "learning_rate": 0.0001,
      "loss": 1.4206,
      "step": 3803
    },
    {
      "epoch": 0.6231979030144168,
      "grad_norm": 0.1583700180053711,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 3804
    },
    {
      "epoch": 0.6233617300131061,
      "grad_norm": 0.16597220301628113,
      "learning_rate": 0.0001,
      "loss": 1.3397,
      "step": 3805
    },
    {
      "epoch": 0.6235255570117956,
      "grad_norm": 0.15929041802883148,
      "learning_rate": 0.0001,
      "loss": 1.3956,
      "step": 3806
    },
    {
      "epoch": 0.6236893840104849,
      "grad_norm": 0.1733919233083725,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 3807
    },
    {
      "epoch": 0.6238532110091743,
      "grad_norm": 0.16109827160835266,
      "learning_rate": 0.0001,
      "loss": 1.3431,
      "step": 3808
    },
    {
      "epoch": 0.6240170380078637,
      "grad_norm": 0.1615375429391861,
      "learning_rate": 0.0001,
      "loss": 1.3829,
      "step": 3809
    },
    {
      "epoch": 0.6241808650065531,
      "grad_norm": 0.15567992627620697,
      "learning_rate": 0.0001,
      "loss": 1.3601,
      "step": 3810
    },
    {
      "epoch": 0.6243446920052425,
      "grad_norm": 0.16420871019363403,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 3811
    },
    {
      "epoch": 0.6245085190039319,
      "grad_norm": 0.16483160853385925,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 3812
    },
    {
      "epoch": 0.6246723460026212,
      "grad_norm": 0.1639026403427124,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 3813
    },
    {
      "epoch": 0.6248361730013107,
      "grad_norm": 0.16676728427410126,
      "learning_rate": 0.0001,
      "loss": 1.3513,
      "step": 3814
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.16143567860126495,
      "learning_rate": 0.0001,
      "loss": 1.4663,
      "step": 3815
    },
    {
      "epoch": 0.6251638269986893,
      "grad_norm": 0.16601897776126862,
      "learning_rate": 0.0001,
      "loss": 1.3018,
      "step": 3816
    },
    {
      "epoch": 0.6253276539973788,
      "grad_norm": 0.15338711440563202,
      "learning_rate": 0.0001,
      "loss": 1.3065,
      "step": 3817
    },
    {
      "epoch": 0.6254914809960681,
      "grad_norm": 0.1588733047246933,
      "learning_rate": 0.0001,
      "loss": 1.2699,
      "step": 3818
    },
    {
      "epoch": 0.6256553079947575,
      "grad_norm": 0.16771571338176727,
      "learning_rate": 0.0001,
      "loss": 1.3979,
      "step": 3819
    },
    {
      "epoch": 0.6258191349934469,
      "grad_norm": 0.15618999302387238,
      "learning_rate": 0.0001,
      "loss": 1.3313,
      "step": 3820
    },
    {
      "epoch": 0.6259829619921363,
      "grad_norm": 0.15756724774837494,
      "learning_rate": 0.0001,
      "loss": 1.3306,
      "step": 3821
    },
    {
      "epoch": 0.6261467889908257,
      "grad_norm": 0.16654078662395477,
      "learning_rate": 0.0001,
      "loss": 1.3225,
      "step": 3822
    },
    {
      "epoch": 0.6263106159895151,
      "grad_norm": 0.1622321456670761,
      "learning_rate": 0.0001,
      "loss": 1.3443,
      "step": 3823
    },
    {
      "epoch": 0.6264744429882044,
      "grad_norm": 0.16181911528110504,
      "learning_rate": 0.0001,
      "loss": 1.4295,
      "step": 3824
    },
    {
      "epoch": 0.6266382699868939,
      "grad_norm": 0.1633685827255249,
      "learning_rate": 0.0001,
      "loss": 1.4777,
      "step": 3825
    },
    {
      "epoch": 0.6268020969855832,
      "grad_norm": 0.16299821436405182,
      "learning_rate": 0.0001,
      "loss": 1.3843,
      "step": 3826
    },
    {
      "epoch": 0.6269659239842726,
      "grad_norm": 0.15381893515586853,
      "learning_rate": 0.0001,
      "loss": 1.2619,
      "step": 3827
    },
    {
      "epoch": 0.627129750982962,
      "grad_norm": 0.16390787065029144,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 3828
    },
    {
      "epoch": 0.6272935779816514,
      "grad_norm": 0.1640421748161316,
      "learning_rate": 0.0001,
      "loss": 1.3552,
      "step": 3829
    },
    {
      "epoch": 0.6274574049803407,
      "grad_norm": 0.16939347982406616,
      "learning_rate": 0.0001,
      "loss": 1.4536,
      "step": 3830
    },
    {
      "epoch": 0.6276212319790302,
      "grad_norm": 0.16471584141254425,
      "learning_rate": 0.0001,
      "loss": 1.4442,
      "step": 3831
    },
    {
      "epoch": 0.6277850589777195,
      "grad_norm": 0.16148585081100464,
      "learning_rate": 0.0001,
      "loss": 1.3615,
      "step": 3832
    },
    {
      "epoch": 0.627948885976409,
      "grad_norm": 0.16862063109874725,
      "learning_rate": 0.0001,
      "loss": 1.3262,
      "step": 3833
    },
    {
      "epoch": 0.6281127129750983,
      "grad_norm": 0.1644550859928131,
      "learning_rate": 0.0001,
      "loss": 1.4059,
      "step": 3834
    },
    {
      "epoch": 0.6282765399737876,
      "grad_norm": 0.17839914560317993,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 3835
    },
    {
      "epoch": 0.6284403669724771,
      "grad_norm": 0.1639111042022705,
      "learning_rate": 0.0001,
      "loss": 1.3026,
      "step": 3836
    },
    {
      "epoch": 0.6286041939711664,
      "grad_norm": 0.16070345044136047,
      "learning_rate": 0.0001,
      "loss": 1.3561,
      "step": 3837
    },
    {
      "epoch": 0.6287680209698558,
      "grad_norm": 0.1552884429693222,
      "learning_rate": 0.0001,
      "loss": 1.2911,
      "step": 3838
    },
    {
      "epoch": 0.6289318479685452,
      "grad_norm": 0.16018107533454895,
      "learning_rate": 0.0001,
      "loss": 1.2819,
      "step": 3839
    },
    {
      "epoch": 0.6290956749672346,
      "grad_norm": 0.15923333168029785,
      "learning_rate": 0.0001,
      "loss": 1.3705,
      "step": 3840
    },
    {
      "epoch": 0.629259501965924,
      "grad_norm": 0.15980994701385498,
      "learning_rate": 0.0001,
      "loss": 1.3058,
      "step": 3841
    },
    {
      "epoch": 0.6294233289646134,
      "grad_norm": 0.16156591475009918,
      "learning_rate": 0.0001,
      "loss": 1.371,
      "step": 3842
    },
    {
      "epoch": 0.6295871559633027,
      "grad_norm": 0.15801456570625305,
      "learning_rate": 0.0001,
      "loss": 1.4021,
      "step": 3843
    },
    {
      "epoch": 0.6297509829619922,
      "grad_norm": 0.1614512950181961,
      "learning_rate": 0.0001,
      "loss": 1.3493,
      "step": 3844
    },
    {
      "epoch": 0.6299148099606815,
      "grad_norm": 0.16139286756515503,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 3845
    },
    {
      "epoch": 0.6300786369593709,
      "grad_norm": 0.17395786941051483,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 3846
    },
    {
      "epoch": 0.6302424639580603,
      "grad_norm": 0.16714876890182495,
      "learning_rate": 0.0001,
      "loss": 1.3977,
      "step": 3847
    },
    {
      "epoch": 0.6304062909567497,
      "grad_norm": 0.1567513644695282,
      "learning_rate": 0.0001,
      "loss": 1.3009,
      "step": 3848
    },
    {
      "epoch": 0.630570117955439,
      "grad_norm": 0.1593196541070938,
      "learning_rate": 0.0001,
      "loss": 1.4326,
      "step": 3849
    },
    {
      "epoch": 0.6307339449541285,
      "grad_norm": 0.1621357798576355,
      "learning_rate": 0.0001,
      "loss": 1.3699,
      "step": 3850
    },
    {
      "epoch": 0.6308977719528178,
      "grad_norm": 0.1602923721075058,
      "learning_rate": 0.0001,
      "loss": 1.3881,
      "step": 3851
    },
    {
      "epoch": 0.6310615989515073,
      "grad_norm": 0.1674702912569046,
      "learning_rate": 0.0001,
      "loss": 1.3824,
      "step": 3852
    },
    {
      "epoch": 0.6312254259501966,
      "grad_norm": 0.17191509902477264,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 3853
    },
    {
      "epoch": 0.6313892529488859,
      "grad_norm": 0.1665770262479782,
      "learning_rate": 0.0001,
      "loss": 1.3255,
      "step": 3854
    },
    {
      "epoch": 0.6315530799475754,
      "grad_norm": 0.17583486437797546,
      "learning_rate": 0.0001,
      "loss": 1.4296,
      "step": 3855
    },
    {
      "epoch": 0.6317169069462647,
      "grad_norm": 0.1641882061958313,
      "learning_rate": 0.0001,
      "loss": 1.348,
      "step": 3856
    },
    {
      "epoch": 0.6318807339449541,
      "grad_norm": 0.16187115013599396,
      "learning_rate": 0.0001,
      "loss": 1.2989,
      "step": 3857
    },
    {
      "epoch": 0.6320445609436435,
      "grad_norm": 0.16480030119419098,
      "learning_rate": 0.0001,
      "loss": 1.3406,
      "step": 3858
    },
    {
      "epoch": 0.6322083879423329,
      "grad_norm": 0.1626310497522354,
      "learning_rate": 0.0001,
      "loss": 1.3207,
      "step": 3859
    },
    {
      "epoch": 0.6323722149410222,
      "grad_norm": 0.1617509424686432,
      "learning_rate": 0.0001,
      "loss": 1.3039,
      "step": 3860
    },
    {
      "epoch": 0.6325360419397117,
      "grad_norm": 0.16461776196956635,
      "learning_rate": 0.0001,
      "loss": 1.3623,
      "step": 3861
    },
    {
      "epoch": 0.632699868938401,
      "grad_norm": 0.1789625883102417,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 3862
    },
    {
      "epoch": 0.6328636959370905,
      "grad_norm": 0.15942223370075226,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 3863
    },
    {
      "epoch": 0.6330275229357798,
      "grad_norm": 0.1689371019601822,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 3864
    },
    {
      "epoch": 0.6331913499344692,
      "grad_norm": 0.16234508156776428,
      "learning_rate": 0.0001,
      "loss": 1.2872,
      "step": 3865
    },
    {
      "epoch": 0.6333551769331586,
      "grad_norm": 0.1583353877067566,
      "learning_rate": 0.0001,
      "loss": 1.4037,
      "step": 3866
    },
    {
      "epoch": 0.633519003931848,
      "grad_norm": 0.16006843745708466,
      "learning_rate": 0.0001,
      "loss": 1.3476,
      "step": 3867
    },
    {
      "epoch": 0.6336828309305373,
      "grad_norm": 0.16448190808296204,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 3868
    },
    {
      "epoch": 0.6338466579292268,
      "grad_norm": 0.156049445271492,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 3869
    },
    {
      "epoch": 0.6340104849279161,
      "grad_norm": 0.1614864319562912,
      "learning_rate": 0.0001,
      "loss": 1.3723,
      "step": 3870
    },
    {
      "epoch": 0.6341743119266054,
      "grad_norm": 0.1622834950685501,
      "learning_rate": 0.0001,
      "loss": 1.4247,
      "step": 3871
    },
    {
      "epoch": 0.6343381389252949,
      "grad_norm": 0.17389027774333954,
      "learning_rate": 0.0001,
      "loss": 1.405,
      "step": 3872
    },
    {
      "epoch": 0.6345019659239842,
      "grad_norm": 0.16242709755897522,
      "learning_rate": 0.0001,
      "loss": 1.3033,
      "step": 3873
    },
    {
      "epoch": 0.6346657929226737,
      "grad_norm": 0.16618596017360687,
      "learning_rate": 0.0001,
      "loss": 1.3513,
      "step": 3874
    },
    {
      "epoch": 0.634829619921363,
      "grad_norm": 0.1598692536354065,
      "learning_rate": 0.0001,
      "loss": 1.3157,
      "step": 3875
    },
    {
      "epoch": 0.6349934469200524,
      "grad_norm": 0.16502657532691956,
      "learning_rate": 0.0001,
      "loss": 1.4283,
      "step": 3876
    },
    {
      "epoch": 0.6351572739187418,
      "grad_norm": 0.16291506588459015,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 3877
    },
    {
      "epoch": 0.6353211009174312,
      "grad_norm": 0.16224052011966705,
      "learning_rate": 0.0001,
      "loss": 1.3801,
      "step": 3878
    },
    {
      "epoch": 0.6354849279161205,
      "grad_norm": 0.15981119871139526,
      "learning_rate": 0.0001,
      "loss": 1.3948,
      "step": 3879
    },
    {
      "epoch": 0.63564875491481,
      "grad_norm": 0.15993766486644745,
      "learning_rate": 0.0001,
      "loss": 1.3656,
      "step": 3880
    },
    {
      "epoch": 0.6358125819134993,
      "grad_norm": 0.16109542548656464,
      "learning_rate": 0.0001,
      "loss": 1.3227,
      "step": 3881
    },
    {
      "epoch": 0.6359764089121888,
      "grad_norm": 0.1648918092250824,
      "learning_rate": 0.0001,
      "loss": 1.3491,
      "step": 3882
    },
    {
      "epoch": 0.6361402359108781,
      "grad_norm": 0.15971651673316956,
      "learning_rate": 0.0001,
      "loss": 1.2781,
      "step": 3883
    },
    {
      "epoch": 0.6363040629095675,
      "grad_norm": 0.15533030033111572,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 3884
    },
    {
      "epoch": 0.6364678899082569,
      "grad_norm": 0.15848372876644135,
      "learning_rate": 0.0001,
      "loss": 1.3175,
      "step": 3885
    },
    {
      "epoch": 0.6366317169069463,
      "grad_norm": 0.15979355573654175,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 3886
    },
    {
      "epoch": 0.6367955439056356,
      "grad_norm": 0.1698591411113739,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 3887
    },
    {
      "epoch": 0.6369593709043251,
      "grad_norm": 0.1599453091621399,
      "learning_rate": 0.0001,
      "loss": 1.4144,
      "step": 3888
    },
    {
      "epoch": 0.6371231979030144,
      "grad_norm": 0.17061246931552887,
      "learning_rate": 0.0001,
      "loss": 1.3191,
      "step": 3889
    },
    {
      "epoch": 0.6372870249017037,
      "grad_norm": 0.16545043885707855,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 3890
    },
    {
      "epoch": 0.6374508519003932,
      "grad_norm": 0.15710335969924927,
      "learning_rate": 0.0001,
      "loss": 1.3892,
      "step": 3891
    },
    {
      "epoch": 0.6376146788990825,
      "grad_norm": 0.1566462516784668,
      "learning_rate": 0.0001,
      "loss": 1.3394,
      "step": 3892
    },
    {
      "epoch": 0.637778505897772,
      "grad_norm": 0.16730795800685883,
      "learning_rate": 0.0001,
      "loss": 1.4295,
      "step": 3893
    },
    {
      "epoch": 0.6379423328964613,
      "grad_norm": 0.16129820048809052,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 3894
    },
    {
      "epoch": 0.6381061598951507,
      "grad_norm": 0.17170317471027374,
      "learning_rate": 0.0001,
      "loss": 1.396,
      "step": 3895
    },
    {
      "epoch": 0.6382699868938401,
      "grad_norm": 0.1550309807062149,
      "learning_rate": 0.0001,
      "loss": 1.3523,
      "step": 3896
    },
    {
      "epoch": 0.6384338138925295,
      "grad_norm": 0.1701720952987671,
      "learning_rate": 0.0001,
      "loss": 1.5927,
      "step": 3897
    },
    {
      "epoch": 0.6385976408912188,
      "grad_norm": 0.1629335880279541,
      "learning_rate": 0.0001,
      "loss": 1.2829,
      "step": 3898
    },
    {
      "epoch": 0.6387614678899083,
      "grad_norm": 0.15978534519672394,
      "learning_rate": 0.0001,
      "loss": 1.3604,
      "step": 3899
    },
    {
      "epoch": 0.6389252948885976,
      "grad_norm": 0.1594710648059845,
      "learning_rate": 0.0001,
      "loss": 1.3481,
      "step": 3900
    },
    {
      "epoch": 0.6390891218872871,
      "grad_norm": 0.16098937392234802,
      "learning_rate": 0.0001,
      "loss": 1.3562,
      "step": 3901
    },
    {
      "epoch": 0.6392529488859764,
      "grad_norm": 0.1624188870191574,
      "learning_rate": 0.0001,
      "loss": 1.273,
      "step": 3902
    },
    {
      "epoch": 0.6394167758846658,
      "grad_norm": 0.1626436412334442,
      "learning_rate": 0.0001,
      "loss": 1.2807,
      "step": 3903
    },
    {
      "epoch": 0.6395806028833552,
      "grad_norm": 0.16410709917545319,
      "learning_rate": 0.0001,
      "loss": 1.3139,
      "step": 3904
    },
    {
      "epoch": 0.6397444298820446,
      "grad_norm": 0.15975883603096008,
      "learning_rate": 0.0001,
      "loss": 1.3527,
      "step": 3905
    },
    {
      "epoch": 0.6399082568807339,
      "grad_norm": 0.16345946490764618,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 3906
    },
    {
      "epoch": 0.6400720838794234,
      "grad_norm": 0.17001353204250336,
      "learning_rate": 0.0001,
      "loss": 1.3017,
      "step": 3907
    },
    {
      "epoch": 0.6402359108781127,
      "grad_norm": 0.171365424990654,
      "learning_rate": 0.0001,
      "loss": 1.3928,
      "step": 3908
    },
    {
      "epoch": 0.640399737876802,
      "grad_norm": 0.16232171654701233,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 3909
    },
    {
      "epoch": 0.6405635648754915,
      "grad_norm": 0.15717989206314087,
      "learning_rate": 0.0001,
      "loss": 1.4136,
      "step": 3910
    },
    {
      "epoch": 0.6407273918741808,
      "grad_norm": 0.1582614779472351,
      "learning_rate": 0.0001,
      "loss": 1.2839,
      "step": 3911
    },
    {
      "epoch": 0.6408912188728703,
      "grad_norm": 0.1611819863319397,
      "learning_rate": 0.0001,
      "loss": 1.2491,
      "step": 3912
    },
    {
      "epoch": 0.6410550458715596,
      "grad_norm": 0.16319823265075684,
      "learning_rate": 0.0001,
      "loss": 1.3342,
      "step": 3913
    },
    {
      "epoch": 0.641218872870249,
      "grad_norm": 0.1602475941181183,
      "learning_rate": 0.0001,
      "loss": 1.3443,
      "step": 3914
    },
    {
      "epoch": 0.6413826998689384,
      "grad_norm": 0.17944063246250153,
      "learning_rate": 0.0001,
      "loss": 1.3544,
      "step": 3915
    },
    {
      "epoch": 0.6415465268676278,
      "grad_norm": 0.16555175185203552,
      "learning_rate": 0.0001,
      "loss": 1.4313,
      "step": 3916
    },
    {
      "epoch": 0.6417103538663171,
      "grad_norm": 0.16580349206924438,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 3917
    },
    {
      "epoch": 0.6418741808650066,
      "grad_norm": 0.16074681282043457,
      "learning_rate": 0.0001,
      "loss": 1.3185,
      "step": 3918
    },
    {
      "epoch": 0.6420380078636959,
      "grad_norm": 0.16264094412326813,
      "learning_rate": 0.0001,
      "loss": 1.3612,
      "step": 3919
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 0.16087374091148376,
      "learning_rate": 0.0001,
      "loss": 1.3798,
      "step": 3920
    },
    {
      "epoch": 0.6423656618610747,
      "grad_norm": 0.16556943953037262,
      "learning_rate": 0.0001,
      "loss": 1.3644,
      "step": 3921
    },
    {
      "epoch": 0.6425294888597641,
      "grad_norm": 0.1585896611213684,
      "learning_rate": 0.0001,
      "loss": 1.387,
      "step": 3922
    },
    {
      "epoch": 0.6426933158584535,
      "grad_norm": 0.16715915501117706,
      "learning_rate": 0.0001,
      "loss": 1.3843,
      "step": 3923
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 0.16411927342414856,
      "learning_rate": 0.0001,
      "loss": 1.4021,
      "step": 3924
    },
    {
      "epoch": 0.6430209698558322,
      "grad_norm": 0.16385437548160553,
      "learning_rate": 0.0001,
      "loss": 1.3973,
      "step": 3925
    },
    {
      "epoch": 0.6431847968545217,
      "grad_norm": 0.15864966809749603,
      "learning_rate": 0.0001,
      "loss": 1.3647,
      "step": 3926
    },
    {
      "epoch": 0.643348623853211,
      "grad_norm": 0.15976394712924957,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 3927
    },
    {
      "epoch": 0.6435124508519003,
      "grad_norm": 0.1716068834066391,
      "learning_rate": 0.0001,
      "loss": 1.3667,
      "step": 3928
    },
    {
      "epoch": 0.6436762778505898,
      "grad_norm": 0.16125215590000153,
      "learning_rate": 0.0001,
      "loss": 1.3731,
      "step": 3929
    },
    {
      "epoch": 0.6438401048492791,
      "grad_norm": 0.1670350283384323,
      "learning_rate": 0.0001,
      "loss": 1.462,
      "step": 3930
    },
    {
      "epoch": 0.6440039318479686,
      "grad_norm": 0.16712746024131775,
      "learning_rate": 0.0001,
      "loss": 1.4633,
      "step": 3931
    },
    {
      "epoch": 0.6441677588466579,
      "grad_norm": 0.16175366938114166,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 3932
    },
    {
      "epoch": 0.6443315858453473,
      "grad_norm": 0.1613425612449646,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 3933
    },
    {
      "epoch": 0.6444954128440367,
      "grad_norm": 0.1602337807416916,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 3934
    },
    {
      "epoch": 0.6446592398427261,
      "grad_norm": 0.17125588655471802,
      "learning_rate": 0.0001,
      "loss": 1.3675,
      "step": 3935
    },
    {
      "epoch": 0.6448230668414154,
      "grad_norm": 0.15695072710514069,
      "learning_rate": 0.0001,
      "loss": 1.2881,
      "step": 3936
    },
    {
      "epoch": 0.6449868938401049,
      "grad_norm": 0.16820430755615234,
      "learning_rate": 0.0001,
      "loss": 1.4396,
      "step": 3937
    },
    {
      "epoch": 0.6451507208387942,
      "grad_norm": 0.17179587483406067,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 3938
    },
    {
      "epoch": 0.6453145478374837,
      "grad_norm": 0.16124428808689117,
      "learning_rate": 0.0001,
      "loss": 1.2433,
      "step": 3939
    },
    {
      "epoch": 0.645478374836173,
      "grad_norm": 0.16672751307487488,
      "learning_rate": 0.0001,
      "loss": 1.3474,
      "step": 3940
    },
    {
      "epoch": 0.6456422018348624,
      "grad_norm": 0.15707114338874817,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 3941
    },
    {
      "epoch": 0.6458060288335518,
      "grad_norm": 0.1631290465593338,
      "learning_rate": 0.0001,
      "loss": 1.3497,
      "step": 3942
    },
    {
      "epoch": 0.6459698558322412,
      "grad_norm": 0.16066884994506836,
      "learning_rate": 0.0001,
      "loss": 1.3803,
      "step": 3943
    },
    {
      "epoch": 0.6461336828309305,
      "grad_norm": 0.1604137271642685,
      "learning_rate": 0.0001,
      "loss": 1.2821,
      "step": 3944
    },
    {
      "epoch": 0.64629750982962,
      "grad_norm": 0.16717423498630524,
      "learning_rate": 0.0001,
      "loss": 1.4691,
      "step": 3945
    },
    {
      "epoch": 0.6464613368283093,
      "grad_norm": 0.17609278857707977,
      "learning_rate": 0.0001,
      "loss": 1.4201,
      "step": 3946
    },
    {
      "epoch": 0.6466251638269986,
      "grad_norm": 0.16022911667823792,
      "learning_rate": 0.0001,
      "loss": 1.3589,
      "step": 3947
    },
    {
      "epoch": 0.6467889908256881,
      "grad_norm": 0.16498959064483643,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 3948
    },
    {
      "epoch": 0.6469528178243774,
      "grad_norm": 0.15798121690750122,
      "learning_rate": 0.0001,
      "loss": 1.4563,
      "step": 3949
    },
    {
      "epoch": 0.6471166448230669,
      "grad_norm": 0.1617143601179123,
      "learning_rate": 0.0001,
      "loss": 1.3712,
      "step": 3950
    },
    {
      "epoch": 0.6472804718217562,
      "grad_norm": 0.17213821411132812,
      "learning_rate": 0.0001,
      "loss": 1.3291,
      "step": 3951
    },
    {
      "epoch": 0.6474442988204456,
      "grad_norm": 0.16562137007713318,
      "learning_rate": 0.0001,
      "loss": 1.4236,
      "step": 3952
    },
    {
      "epoch": 0.647608125819135,
      "grad_norm": 0.16060605645179749,
      "learning_rate": 0.0001,
      "loss": 1.3896,
      "step": 3953
    },
    {
      "epoch": 0.6477719528178244,
      "grad_norm": 0.15638956427574158,
      "learning_rate": 0.0001,
      "loss": 1.1876,
      "step": 3954
    },
    {
      "epoch": 0.6479357798165137,
      "grad_norm": 0.15569299459457397,
      "learning_rate": 0.0001,
      "loss": 1.3733,
      "step": 3955
    },
    {
      "epoch": 0.6480996068152032,
      "grad_norm": 0.1672324538230896,
      "learning_rate": 0.0001,
      "loss": 1.3887,
      "step": 3956
    },
    {
      "epoch": 0.6482634338138925,
      "grad_norm": 0.15637236833572388,
      "learning_rate": 0.0001,
      "loss": 1.3152,
      "step": 3957
    },
    {
      "epoch": 0.648427260812582,
      "grad_norm": 0.17335741221904755,
      "learning_rate": 0.0001,
      "loss": 1.3814,
      "step": 3958
    },
    {
      "epoch": 0.6485910878112713,
      "grad_norm": 0.1651751846075058,
      "learning_rate": 0.0001,
      "loss": 1.4131,
      "step": 3959
    },
    {
      "epoch": 0.6487549148099607,
      "grad_norm": 0.17387792468070984,
      "learning_rate": 0.0001,
      "loss": 1.3262,
      "step": 3960
    },
    {
      "epoch": 0.6489187418086501,
      "grad_norm": 0.16304419934749603,
      "learning_rate": 0.0001,
      "loss": 1.3845,
      "step": 3961
    },
    {
      "epoch": 0.6490825688073395,
      "grad_norm": 0.16154825687408447,
      "learning_rate": 0.0001,
      "loss": 1.3782,
      "step": 3962
    },
    {
      "epoch": 0.6492463958060288,
      "grad_norm": 0.16061118245124817,
      "learning_rate": 0.0001,
      "loss": 1.3384,
      "step": 3963
    },
    {
      "epoch": 0.6494102228047182,
      "grad_norm": 0.16443929076194763,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 3964
    },
    {
      "epoch": 0.6495740498034076,
      "grad_norm": 0.16823087632656097,
      "learning_rate": 0.0001,
      "loss": 1.3852,
      "step": 3965
    },
    {
      "epoch": 0.6497378768020969,
      "grad_norm": 0.1627075970172882,
      "learning_rate": 0.0001,
      "loss": 1.3595,
      "step": 3966
    },
    {
      "epoch": 0.6499017038007864,
      "grad_norm": 0.15959672629833221,
      "learning_rate": 0.0001,
      "loss": 1.3098,
      "step": 3967
    },
    {
      "epoch": 0.6500655307994757,
      "grad_norm": 0.15889587998390198,
      "learning_rate": 0.0001,
      "loss": 1.3793,
      "step": 3968
    },
    {
      "epoch": 0.6502293577981652,
      "grad_norm": 0.16548511385917664,
      "learning_rate": 0.0001,
      "loss": 1.3999,
      "step": 3969
    },
    {
      "epoch": 0.6503931847968545,
      "grad_norm": 0.16141435503959656,
      "learning_rate": 0.0001,
      "loss": 1.3994,
      "step": 3970
    },
    {
      "epoch": 0.6505570117955439,
      "grad_norm": 0.15976986289024353,
      "learning_rate": 0.0001,
      "loss": 1.3215,
      "step": 3971
    },
    {
      "epoch": 0.6507208387942333,
      "grad_norm": 0.16159716248512268,
      "learning_rate": 0.0001,
      "loss": 1.3747,
      "step": 3972
    },
    {
      "epoch": 0.6508846657929227,
      "grad_norm": 0.1610044240951538,
      "learning_rate": 0.0001,
      "loss": 1.3369,
      "step": 3973
    },
    {
      "epoch": 0.651048492791612,
      "grad_norm": 0.15994048118591309,
      "learning_rate": 0.0001,
      "loss": 1.3853,
      "step": 3974
    },
    {
      "epoch": 0.6512123197903015,
      "grad_norm": 0.15826471149921417,
      "learning_rate": 0.0001,
      "loss": 1.3738,
      "step": 3975
    },
    {
      "epoch": 0.6513761467889908,
      "grad_norm": 0.18372398614883423,
      "learning_rate": 0.0001,
      "loss": 1.2865,
      "step": 3976
    },
    {
      "epoch": 0.6515399737876802,
      "grad_norm": 0.1689513623714447,
      "learning_rate": 0.0001,
      "loss": 1.4488,
      "step": 3977
    },
    {
      "epoch": 0.6517038007863696,
      "grad_norm": 0.1587538868188858,
      "learning_rate": 0.0001,
      "loss": 1.4381,
      "step": 3978
    },
    {
      "epoch": 0.651867627785059,
      "grad_norm": 0.1752086579799652,
      "learning_rate": 0.0001,
      "loss": 1.3974,
      "step": 3979
    },
    {
      "epoch": 0.6520314547837484,
      "grad_norm": 0.1716785579919815,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 3980
    },
    {
      "epoch": 0.6521952817824378,
      "grad_norm": 0.16021600365638733,
      "learning_rate": 0.0001,
      "loss": 1.345,
      "step": 3981
    },
    {
      "epoch": 0.6523591087811271,
      "grad_norm": 0.16661964356899261,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 3982
    },
    {
      "epoch": 0.6525229357798165,
      "grad_norm": 0.15589876472949982,
      "learning_rate": 0.0001,
      "loss": 1.3211,
      "step": 3983
    },
    {
      "epoch": 0.6526867627785059,
      "grad_norm": 0.16784311830997467,
      "learning_rate": 0.0001,
      "loss": 1.4228,
      "step": 3984
    },
    {
      "epoch": 0.6528505897771952,
      "grad_norm": 0.16351035237312317,
      "learning_rate": 0.0001,
      "loss": 1.3672,
      "step": 3985
    },
    {
      "epoch": 0.6530144167758847,
      "grad_norm": 0.1616494059562683,
      "learning_rate": 0.0001,
      "loss": 1.3428,
      "step": 3986
    },
    {
      "epoch": 0.653178243774574,
      "grad_norm": 0.16106314957141876,
      "learning_rate": 0.0001,
      "loss": 1.3064,
      "step": 3987
    },
    {
      "epoch": 0.6533420707732635,
      "grad_norm": 0.16150890290737152,
      "learning_rate": 0.0001,
      "loss": 1.3027,
      "step": 3988
    },
    {
      "epoch": 0.6535058977719528,
      "grad_norm": 0.1618487685918808,
      "learning_rate": 0.0001,
      "loss": 1.4022,
      "step": 3989
    },
    {
      "epoch": 0.6536697247706422,
      "grad_norm": 0.16354046761989594,
      "learning_rate": 0.0001,
      "loss": 1.3146,
      "step": 3990
    },
    {
      "epoch": 0.6538335517693316,
      "grad_norm": 0.16653063893318176,
      "learning_rate": 0.0001,
      "loss": 1.3848,
      "step": 3991
    },
    {
      "epoch": 0.653997378768021,
      "grad_norm": 0.16782571375370026,
      "learning_rate": 0.0001,
      "loss": 1.3112,
      "step": 3992
    },
    {
      "epoch": 0.6541612057667103,
      "grad_norm": 0.17182514071464539,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 3993
    },
    {
      "epoch": 0.6543250327653998,
      "grad_norm": 0.16193589568138123,
      "learning_rate": 0.0001,
      "loss": 1.3806,
      "step": 3994
    },
    {
      "epoch": 0.6544888597640891,
      "grad_norm": 0.18093083798885345,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 3995
    },
    {
      "epoch": 0.6546526867627785,
      "grad_norm": 0.1665964275598526,
      "learning_rate": 0.0001,
      "loss": 1.2995,
      "step": 3996
    },
    {
      "epoch": 0.6548165137614679,
      "grad_norm": 0.1577063351869583,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 3997
    },
    {
      "epoch": 0.6549803407601573,
      "grad_norm": 0.15508121252059937,
      "learning_rate": 0.0001,
      "loss": 1.3837,
      "step": 3998
    },
    {
      "epoch": 0.6551441677588467,
      "grad_norm": 0.1672736406326294,
      "learning_rate": 0.0001,
      "loss": 1.4124,
      "step": 3999
    },
    {
      "epoch": 0.6553079947575361,
      "grad_norm": 0.1600131392478943,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 4000
    },
    {
      "epoch": 0.6554718217562254,
      "grad_norm": 0.16206787526607513,
      "learning_rate": 0.0001,
      "loss": 1.3523,
      "step": 4001
    },
    {
      "epoch": 0.6556356487549148,
      "grad_norm": 0.16217708587646484,
      "learning_rate": 0.0001,
      "loss": 1.3801,
      "step": 4002
    },
    {
      "epoch": 0.6557994757536042,
      "grad_norm": 0.16968068480491638,
      "learning_rate": 0.0001,
      "loss": 1.4119,
      "step": 4003
    },
    {
      "epoch": 0.6559633027522935,
      "grad_norm": 0.17111101746559143,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 4004
    },
    {
      "epoch": 0.656127129750983,
      "grad_norm": 0.16671720147132874,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 4005
    },
    {
      "epoch": 0.6562909567496723,
      "grad_norm": 0.16679027676582336,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 4006
    },
    {
      "epoch": 0.6564547837483617,
      "grad_norm": 0.16370998322963715,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 4007
    },
    {
      "epoch": 0.6566186107470511,
      "grad_norm": 0.17330212891101837,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 4008
    },
    {
      "epoch": 0.6567824377457405,
      "grad_norm": 0.15965527296066284,
      "learning_rate": 0.0001,
      "loss": 1.2782,
      "step": 4009
    },
    {
      "epoch": 0.6569462647444299,
      "grad_norm": 0.16946756839752197,
      "learning_rate": 0.0001,
      "loss": 1.397,
      "step": 4010
    },
    {
      "epoch": 0.6571100917431193,
      "grad_norm": 0.15740448236465454,
      "learning_rate": 0.0001,
      "loss": 1.286,
      "step": 4011
    },
    {
      "epoch": 0.6572739187418086,
      "grad_norm": 0.16475342214107513,
      "learning_rate": 0.0001,
      "loss": 1.3739,
      "step": 4012
    },
    {
      "epoch": 0.6574377457404981,
      "grad_norm": 0.16602981090545654,
      "learning_rate": 0.0001,
      "loss": 1.4443,
      "step": 4013
    },
    {
      "epoch": 0.6576015727391874,
      "grad_norm": 0.16844983398914337,
      "learning_rate": 0.0001,
      "loss": 1.347,
      "step": 4014
    },
    {
      "epoch": 0.6577653997378768,
      "grad_norm": 0.15654601156711578,
      "learning_rate": 0.0001,
      "loss": 1.4498,
      "step": 4015
    },
    {
      "epoch": 0.6579292267365662,
      "grad_norm": 0.16279636323451996,
      "learning_rate": 0.0001,
      "loss": 1.3244,
      "step": 4016
    },
    {
      "epoch": 0.6580930537352556,
      "grad_norm": 0.18444310128688812,
      "learning_rate": 0.0001,
      "loss": 1.384,
      "step": 4017
    },
    {
      "epoch": 0.658256880733945,
      "grad_norm": 0.16854025423526764,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 4018
    },
    {
      "epoch": 0.6584207077326344,
      "grad_norm": 0.17102056741714478,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 4019
    },
    {
      "epoch": 0.6585845347313237,
      "grad_norm": 0.16322697699069977,
      "learning_rate": 0.0001,
      "loss": 1.38,
      "step": 4020
    },
    {
      "epoch": 0.658748361730013,
      "grad_norm": 0.16835050284862518,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 4021
    },
    {
      "epoch": 0.6589121887287025,
      "grad_norm": 0.16769488155841827,
      "learning_rate": 0.0001,
      "loss": 1.387,
      "step": 4022
    },
    {
      "epoch": 0.6590760157273918,
      "grad_norm": 0.16059188544750214,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 4023
    },
    {
      "epoch": 0.6592398427260813,
      "grad_norm": 0.16435754299163818,
      "learning_rate": 0.0001,
      "loss": 1.3389,
      "step": 4024
    },
    {
      "epoch": 0.6594036697247706,
      "grad_norm": 0.16206341981887817,
      "learning_rate": 0.0001,
      "loss": 1.2807,
      "step": 4025
    },
    {
      "epoch": 0.65956749672346,
      "grad_norm": 0.16297852993011475,
      "learning_rate": 0.0001,
      "loss": 1.3254,
      "step": 4026
    },
    {
      "epoch": 0.6597313237221494,
      "grad_norm": 0.18136312067508698,
      "learning_rate": 0.0001,
      "loss": 1.5559,
      "step": 4027
    },
    {
      "epoch": 0.6598951507208388,
      "grad_norm": 0.17416994273662567,
      "learning_rate": 0.0001,
      "loss": 1.3437,
      "step": 4028
    },
    {
      "epoch": 0.6600589777195282,
      "grad_norm": 0.17748531699180603,
      "learning_rate": 0.0001,
      "loss": 1.3701,
      "step": 4029
    },
    {
      "epoch": 0.6602228047182176,
      "grad_norm": 0.16543366014957428,
      "learning_rate": 0.0001,
      "loss": 1.2312,
      "step": 4030
    },
    {
      "epoch": 0.6603866317169069,
      "grad_norm": 0.20185700058937073,
      "learning_rate": 0.0001,
      "loss": 1.253,
      "step": 4031
    },
    {
      "epoch": 0.6605504587155964,
      "grad_norm": 0.1674085408449173,
      "learning_rate": 0.0001,
      "loss": 1.3818,
      "step": 4032
    },
    {
      "epoch": 0.6607142857142857,
      "grad_norm": 0.16126081347465515,
      "learning_rate": 0.0001,
      "loss": 1.3686,
      "step": 4033
    },
    {
      "epoch": 0.6608781127129751,
      "grad_norm": 0.1627933830022812,
      "learning_rate": 0.0001,
      "loss": 1.4011,
      "step": 4034
    },
    {
      "epoch": 0.6610419397116645,
      "grad_norm": 0.16462725400924683,
      "learning_rate": 0.0001,
      "loss": 1.3206,
      "step": 4035
    },
    {
      "epoch": 0.6612057667103539,
      "grad_norm": 0.16428524255752563,
      "learning_rate": 0.0001,
      "loss": 1.2519,
      "step": 4036
    },
    {
      "epoch": 0.6613695937090432,
      "grad_norm": 0.16091440618038177,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 4037
    },
    {
      "epoch": 0.6615334207077327,
      "grad_norm": 0.16606692969799042,
      "learning_rate": 0.0001,
      "loss": 1.4017,
      "step": 4038
    },
    {
      "epoch": 0.661697247706422,
      "grad_norm": 0.1733231246471405,
      "learning_rate": 0.0001,
      "loss": 1.4151,
      "step": 4039
    },
    {
      "epoch": 0.6618610747051114,
      "grad_norm": 0.15839232504367828,
      "learning_rate": 0.0001,
      "loss": 1.3394,
      "step": 4040
    },
    {
      "epoch": 0.6620249017038008,
      "grad_norm": 0.1624530702829361,
      "learning_rate": 0.0001,
      "loss": 1.3773,
      "step": 4041
    },
    {
      "epoch": 0.6621887287024901,
      "grad_norm": 0.17459256947040558,
      "learning_rate": 0.0001,
      "loss": 1.3956,
      "step": 4042
    },
    {
      "epoch": 0.6623525557011796,
      "grad_norm": 0.1621387004852295,
      "learning_rate": 0.0001,
      "loss": 1.2792,
      "step": 4043
    },
    {
      "epoch": 0.6625163826998689,
      "grad_norm": 0.15965670347213745,
      "learning_rate": 0.0001,
      "loss": 1.3381,
      "step": 4044
    },
    {
      "epoch": 0.6626802096985583,
      "grad_norm": 0.16888858377933502,
      "learning_rate": 0.0001,
      "loss": 1.3674,
      "step": 4045
    },
    {
      "epoch": 0.6628440366972477,
      "grad_norm": 0.1607484519481659,
      "learning_rate": 0.0001,
      "loss": 1.2638,
      "step": 4046
    },
    {
      "epoch": 0.6630078636959371,
      "grad_norm": 0.16322408616542816,
      "learning_rate": 0.0001,
      "loss": 1.3275,
      "step": 4047
    },
    {
      "epoch": 0.6631716906946264,
      "grad_norm": 0.16322803497314453,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 4048
    },
    {
      "epoch": 0.6633355176933159,
      "grad_norm": 0.17161528766155243,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 4049
    },
    {
      "epoch": 0.6634993446920052,
      "grad_norm": 0.16516737639904022,
      "learning_rate": 0.0001,
      "loss": 1.3716,
      "step": 4050
    },
    {
      "epoch": 0.6636631716906947,
      "grad_norm": 0.17842997610569,
      "learning_rate": 0.0001,
      "loss": 1.4817,
      "step": 4051
    },
    {
      "epoch": 0.663826998689384,
      "grad_norm": 0.16868473589420319,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 4052
    },
    {
      "epoch": 0.6639908256880734,
      "grad_norm": 0.1596693992614746,
      "learning_rate": 0.0001,
      "loss": 1.3774,
      "step": 4053
    },
    {
      "epoch": 0.6641546526867628,
      "grad_norm": 0.16381525993347168,
      "learning_rate": 0.0001,
      "loss": 1.3509,
      "step": 4054
    },
    {
      "epoch": 0.6643184796854522,
      "grad_norm": 0.16425177454948425,
      "learning_rate": 0.0001,
      "loss": 1.3988,
      "step": 4055
    },
    {
      "epoch": 0.6644823066841415,
      "grad_norm": 0.16821648180484772,
      "learning_rate": 0.0001,
      "loss": 1.3909,
      "step": 4056
    },
    {
      "epoch": 0.6646461336828309,
      "grad_norm": 0.15382671356201172,
      "learning_rate": 0.0001,
      "loss": 1.3058,
      "step": 4057
    },
    {
      "epoch": 0.6648099606815203,
      "grad_norm": 0.16279609501361847,
      "learning_rate": 0.0001,
      "loss": 1.3852,
      "step": 4058
    },
    {
      "epoch": 0.6649737876802096,
      "grad_norm": 0.16746395826339722,
      "learning_rate": 0.0001,
      "loss": 1.4028,
      "step": 4059
    },
    {
      "epoch": 0.6651376146788991,
      "grad_norm": 0.1691022366285324,
      "learning_rate": 0.0001,
      "loss": 1.4462,
      "step": 4060
    },
    {
      "epoch": 0.6653014416775884,
      "grad_norm": 0.16458293795585632,
      "learning_rate": 0.0001,
      "loss": 1.3604,
      "step": 4061
    },
    {
      "epoch": 0.6654652686762779,
      "grad_norm": 0.15962325036525726,
      "learning_rate": 0.0001,
      "loss": 1.3375,
      "step": 4062
    },
    {
      "epoch": 0.6656290956749672,
      "grad_norm": 0.1562601774930954,
      "learning_rate": 0.0001,
      "loss": 1.3282,
      "step": 4063
    },
    {
      "epoch": 0.6657929226736566,
      "grad_norm": 0.1718924194574356,
      "learning_rate": 0.0001,
      "loss": 1.3998,
      "step": 4064
    },
    {
      "epoch": 0.665956749672346,
      "grad_norm": 0.16325020790100098,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 4065
    },
    {
      "epoch": 0.6661205766710354,
      "grad_norm": 0.16888011991977692,
      "learning_rate": 0.0001,
      "loss": 1.376,
      "step": 4066
    },
    {
      "epoch": 0.6662844036697247,
      "grad_norm": 0.1686350405216217,
      "learning_rate": 0.0001,
      "loss": 1.3518,
      "step": 4067
    },
    {
      "epoch": 0.6664482306684142,
      "grad_norm": 0.16263118386268616,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 4068
    },
    {
      "epoch": 0.6666120576671035,
      "grad_norm": 0.16552060842514038,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 4069
    },
    {
      "epoch": 0.666775884665793,
      "grad_norm": 0.1634436994791031,
      "learning_rate": 0.0001,
      "loss": 1.3351,
      "step": 4070
    },
    {
      "epoch": 0.6669397116644823,
      "grad_norm": 0.18848535418510437,
      "learning_rate": 0.0001,
      "loss": 1.5419,
      "step": 4071
    },
    {
      "epoch": 0.6671035386631717,
      "grad_norm": 0.16974669694900513,
      "learning_rate": 0.0001,
      "loss": 1.3312,
      "step": 4072
    },
    {
      "epoch": 0.6672673656618611,
      "grad_norm": 0.16147065162658691,
      "learning_rate": 0.0001,
      "loss": 1.3873,
      "step": 4073
    },
    {
      "epoch": 0.6674311926605505,
      "grad_norm": 0.1583590805530548,
      "learning_rate": 0.0001,
      "loss": 1.3051,
      "step": 4074
    },
    {
      "epoch": 0.6675950196592398,
      "grad_norm": 0.16460584104061127,
      "learning_rate": 0.0001,
      "loss": 1.3499,
      "step": 4075
    },
    {
      "epoch": 0.6677588466579292,
      "grad_norm": 0.16434387862682343,
      "learning_rate": 0.0001,
      "loss": 1.4158,
      "step": 4076
    },
    {
      "epoch": 0.6679226736566186,
      "grad_norm": 0.16466189920902252,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 4077
    },
    {
      "epoch": 0.668086500655308,
      "grad_norm": 0.1679265797138214,
      "learning_rate": 0.0001,
      "loss": 1.3313,
      "step": 4078
    },
    {
      "epoch": 0.6682503276539974,
      "grad_norm": 0.16714312136173248,
      "learning_rate": 0.0001,
      "loss": 1.4217,
      "step": 4079
    },
    {
      "epoch": 0.6684141546526867,
      "grad_norm": 0.16680048406124115,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 4080
    },
    {
      "epoch": 0.6685779816513762,
      "grad_norm": 0.17156705260276794,
      "learning_rate": 0.0001,
      "loss": 1.2919,
      "step": 4081
    },
    {
      "epoch": 0.6687418086500655,
      "grad_norm": 0.1790386140346527,
      "learning_rate": 0.0001,
      "loss": 1.3732,
      "step": 4082
    },
    {
      "epoch": 0.6689056356487549,
      "grad_norm": 0.17980092763900757,
      "learning_rate": 0.0001,
      "loss": 1.3714,
      "step": 4083
    },
    {
      "epoch": 0.6690694626474443,
      "grad_norm": 0.1712007224559784,
      "learning_rate": 0.0001,
      "loss": 1.3117,
      "step": 4084
    },
    {
      "epoch": 0.6692332896461337,
      "grad_norm": 0.1604013592004776,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 4085
    },
    {
      "epoch": 0.669397116644823,
      "grad_norm": 0.16683967411518097,
      "learning_rate": 0.0001,
      "loss": 1.3856,
      "step": 4086
    },
    {
      "epoch": 0.6695609436435125,
      "grad_norm": 0.1638288050889969,
      "learning_rate": 0.0001,
      "loss": 1.2964,
      "step": 4087
    },
    {
      "epoch": 0.6697247706422018,
      "grad_norm": 0.16388815641403198,
      "learning_rate": 0.0001,
      "loss": 1.4537,
      "step": 4088
    },
    {
      "epoch": 0.6698885976408913,
      "grad_norm": 0.17411208152770996,
      "learning_rate": 0.0001,
      "loss": 1.3429,
      "step": 4089
    },
    {
      "epoch": 0.6700524246395806,
      "grad_norm": 0.15659667551517487,
      "learning_rate": 0.0001,
      "loss": 1.253,
      "step": 4090
    },
    {
      "epoch": 0.67021625163827,
      "grad_norm": 0.17040057480335236,
      "learning_rate": 0.0001,
      "loss": 1.3845,
      "step": 4091
    },
    {
      "epoch": 0.6703800786369594,
      "grad_norm": 0.15858396887779236,
      "learning_rate": 0.0001,
      "loss": 1.3727,
      "step": 4092
    },
    {
      "epoch": 0.6705439056356488,
      "grad_norm": 0.16154345870018005,
      "learning_rate": 0.0001,
      "loss": 1.3969,
      "step": 4093
    },
    {
      "epoch": 0.6707077326343381,
      "grad_norm": 0.16580133140087128,
      "learning_rate": 0.0001,
      "loss": 1.3711,
      "step": 4094
    },
    {
      "epoch": 0.6708715596330275,
      "grad_norm": 0.16609235107898712,
      "learning_rate": 0.0001,
      "loss": 1.4016,
      "step": 4095
    },
    {
      "epoch": 0.6710353866317169,
      "grad_norm": 0.17211686074733734,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 4096
    },
    {
      "epoch": 0.6711992136304062,
      "grad_norm": 0.16503937542438507,
      "learning_rate": 0.0001,
      "loss": 1.3386,
      "step": 4097
    },
    {
      "epoch": 0.6713630406290957,
      "grad_norm": 0.16086646914482117,
      "learning_rate": 0.0001,
      "loss": 1.3748,
      "step": 4098
    },
    {
      "epoch": 0.671526867627785,
      "grad_norm": 0.16312721371650696,
      "learning_rate": 0.0001,
      "loss": 1.3683,
      "step": 4099
    },
    {
      "epoch": 0.6716906946264745,
      "grad_norm": 0.16971954703330994,
      "learning_rate": 0.0001,
      "loss": 1.3962,
      "step": 4100
    },
    {
      "epoch": 0.6718545216251638,
      "grad_norm": 0.16363640129566193,
      "learning_rate": 0.0001,
      "loss": 1.426,
      "step": 4101
    },
    {
      "epoch": 0.6720183486238532,
      "grad_norm": 0.17164845764636993,
      "learning_rate": 0.0001,
      "loss": 1.3104,
      "step": 4102
    },
    {
      "epoch": 0.6721821756225426,
      "grad_norm": 0.16861401498317719,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 4103
    },
    {
      "epoch": 0.672346002621232,
      "grad_norm": 0.1651277095079422,
      "learning_rate": 0.0001,
      "loss": 1.3317,
      "step": 4104
    },
    {
      "epoch": 0.6725098296199213,
      "grad_norm": 0.1561911702156067,
      "learning_rate": 0.0001,
      "loss": 1.2606,
      "step": 4105
    },
    {
      "epoch": 0.6726736566186108,
      "grad_norm": 0.1622130572795868,
      "learning_rate": 0.0001,
      "loss": 1.3116,
      "step": 4106
    },
    {
      "epoch": 0.6728374836173001,
      "grad_norm": 0.16305293142795563,
      "learning_rate": 0.0001,
      "loss": 1.4125,
      "step": 4107
    },
    {
      "epoch": 0.6730013106159896,
      "grad_norm": 0.15749794244766235,
      "learning_rate": 0.0001,
      "loss": 1.3235,
      "step": 4108
    },
    {
      "epoch": 0.6731651376146789,
      "grad_norm": 0.17066410183906555,
      "learning_rate": 0.0001,
      "loss": 1.4025,
      "step": 4109
    },
    {
      "epoch": 0.6733289646133683,
      "grad_norm": 0.1636924147605896,
      "learning_rate": 0.0001,
      "loss": 1.3446,
      "step": 4110
    },
    {
      "epoch": 0.6734927916120577,
      "grad_norm": 0.16096246242523193,
      "learning_rate": 0.0001,
      "loss": 1.3308,
      "step": 4111
    },
    {
      "epoch": 0.6736566186107471,
      "grad_norm": 0.16860613226890564,
      "learning_rate": 0.0001,
      "loss": 1.3385,
      "step": 4112
    },
    {
      "epoch": 0.6738204456094364,
      "grad_norm": 0.16553738713264465,
      "learning_rate": 0.0001,
      "loss": 1.3311,
      "step": 4113
    },
    {
      "epoch": 0.6739842726081258,
      "grad_norm": 0.16507862508296967,
      "learning_rate": 0.0001,
      "loss": 1.3846,
      "step": 4114
    },
    {
      "epoch": 0.6741480996068152,
      "grad_norm": 0.16899451613426208,
      "learning_rate": 0.0001,
      "loss": 1.4273,
      "step": 4115
    },
    {
      "epoch": 0.6743119266055045,
      "grad_norm": 0.16434448957443237,
      "learning_rate": 0.0001,
      "loss": 1.3621,
      "step": 4116
    },
    {
      "epoch": 0.674475753604194,
      "grad_norm": 0.16358985006809235,
      "learning_rate": 0.0001,
      "loss": 1.4139,
      "step": 4117
    },
    {
      "epoch": 0.6746395806028833,
      "grad_norm": 0.1636996865272522,
      "learning_rate": 0.0001,
      "loss": 1.4043,
      "step": 4118
    },
    {
      "epoch": 0.6748034076015728,
      "grad_norm": 0.17148207128047943,
      "learning_rate": 0.0001,
      "loss": 1.3847,
      "step": 4119
    },
    {
      "epoch": 0.6749672346002621,
      "grad_norm": 0.17602895200252533,
      "learning_rate": 0.0001,
      "loss": 1.541,
      "step": 4120
    },
    {
      "epoch": 0.6751310615989515,
      "grad_norm": 0.17301493883132935,
      "learning_rate": 0.0001,
      "loss": 1.4085,
      "step": 4121
    },
    {
      "epoch": 0.6752948885976409,
      "grad_norm": 0.17196182906627655,
      "learning_rate": 0.0001,
      "loss": 1.3534,
      "step": 4122
    },
    {
      "epoch": 0.6754587155963303,
      "grad_norm": 0.16266487538814545,
      "learning_rate": 0.0001,
      "loss": 1.4163,
      "step": 4123
    },
    {
      "epoch": 0.6756225425950196,
      "grad_norm": 0.16352510452270508,
      "learning_rate": 0.0001,
      "loss": 1.4122,
      "step": 4124
    },
    {
      "epoch": 0.6757863695937091,
      "grad_norm": 0.16956479847431183,
      "learning_rate": 0.0001,
      "loss": 1.3306,
      "step": 4125
    },
    {
      "epoch": 0.6759501965923984,
      "grad_norm": 0.163934588432312,
      "learning_rate": 0.0001,
      "loss": 1.3302,
      "step": 4126
    },
    {
      "epoch": 0.6761140235910879,
      "grad_norm": 0.16850948333740234,
      "learning_rate": 0.0001,
      "loss": 1.3246,
      "step": 4127
    },
    {
      "epoch": 0.6762778505897772,
      "grad_norm": 0.16585926711559296,
      "learning_rate": 0.0001,
      "loss": 1.3667,
      "step": 4128
    },
    {
      "epoch": 0.6764416775884666,
      "grad_norm": 0.16664493083953857,
      "learning_rate": 0.0001,
      "loss": 1.4063,
      "step": 4129
    },
    {
      "epoch": 0.676605504587156,
      "grad_norm": 0.16946955025196075,
      "learning_rate": 0.0001,
      "loss": 1.2653,
      "step": 4130
    },
    {
      "epoch": 0.6767693315858454,
      "grad_norm": 0.16848912835121155,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 4131
    },
    {
      "epoch": 0.6769331585845347,
      "grad_norm": 0.1713060587644577,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 4132
    },
    {
      "epoch": 0.6770969855832241,
      "grad_norm": 0.16683495044708252,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 4133
    },
    {
      "epoch": 0.6772608125819135,
      "grad_norm": 0.16309288144111633,
      "learning_rate": 0.0001,
      "loss": 1.3431,
      "step": 4134
    },
    {
      "epoch": 0.6774246395806028,
      "grad_norm": 0.1635657548904419,
      "learning_rate": 0.0001,
      "loss": 1.3202,
      "step": 4135
    },
    {
      "epoch": 0.6775884665792923,
      "grad_norm": 0.16322998702526093,
      "learning_rate": 0.0001,
      "loss": 1.2999,
      "step": 4136
    },
    {
      "epoch": 0.6777522935779816,
      "grad_norm": 0.16496938467025757,
      "learning_rate": 0.0001,
      "loss": 1.2936,
      "step": 4137
    },
    {
      "epoch": 0.677916120576671,
      "grad_norm": 0.16103360056877136,
      "learning_rate": 0.0001,
      "loss": 1.3631,
      "step": 4138
    },
    {
      "epoch": 0.6780799475753604,
      "grad_norm": 0.16864684224128723,
      "learning_rate": 0.0001,
      "loss": 1.3482,
      "step": 4139
    },
    {
      "epoch": 0.6782437745740498,
      "grad_norm": 0.15596038103103638,
      "learning_rate": 0.0001,
      "loss": 1.2436,
      "step": 4140
    },
    {
      "epoch": 0.6784076015727392,
      "grad_norm": 0.17296797037124634,
      "learning_rate": 0.0001,
      "loss": 1.3695,
      "step": 4141
    },
    {
      "epoch": 0.6785714285714286,
      "grad_norm": 0.16775070130825043,
      "learning_rate": 0.0001,
      "loss": 1.343,
      "step": 4142
    },
    {
      "epoch": 0.6787352555701179,
      "grad_norm": 0.18037721514701843,
      "learning_rate": 0.0001,
      "loss": 1.2957,
      "step": 4143
    },
    {
      "epoch": 0.6788990825688074,
      "grad_norm": 0.17146030068397522,
      "learning_rate": 0.0001,
      "loss": 1.2818,
      "step": 4144
    },
    {
      "epoch": 0.6790629095674967,
      "grad_norm": 0.16172686219215393,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 4145
    },
    {
      "epoch": 0.6792267365661862,
      "grad_norm": 0.16416361927986145,
      "learning_rate": 0.0001,
      "loss": 1.379,
      "step": 4146
    },
    {
      "epoch": 0.6793905635648755,
      "grad_norm": 0.17327576875686646,
      "learning_rate": 0.0001,
      "loss": 1.2775,
      "step": 4147
    },
    {
      "epoch": 0.6795543905635649,
      "grad_norm": 0.16746093332767487,
      "learning_rate": 0.0001,
      "loss": 1.3038,
      "step": 4148
    },
    {
      "epoch": 0.6797182175622543,
      "grad_norm": 0.15467825531959534,
      "learning_rate": 0.0001,
      "loss": 1.2982,
      "step": 4149
    },
    {
      "epoch": 0.6798820445609436,
      "grad_norm": 0.16711461544036865,
      "learning_rate": 0.0001,
      "loss": 1.3457,
      "step": 4150
    },
    {
      "epoch": 0.680045871559633,
      "grad_norm": 0.1652970314025879,
      "learning_rate": 0.0001,
      "loss": 1.279,
      "step": 4151
    },
    {
      "epoch": 0.6802096985583224,
      "grad_norm": 0.17012912034988403,
      "learning_rate": 0.0001,
      "loss": 1.3892,
      "step": 4152
    },
    {
      "epoch": 0.6803735255570118,
      "grad_norm": 0.17745596170425415,
      "learning_rate": 0.0001,
      "loss": 1.3586,
      "step": 4153
    },
    {
      "epoch": 0.6805373525557011,
      "grad_norm": 0.17162294685840607,
      "learning_rate": 0.0001,
      "loss": 1.3185,
      "step": 4154
    },
    {
      "epoch": 0.6807011795543906,
      "grad_norm": 0.17102985084056854,
      "learning_rate": 0.0001,
      "loss": 1.4097,
      "step": 4155
    },
    {
      "epoch": 0.6808650065530799,
      "grad_norm": 0.1749163419008255,
      "learning_rate": 0.0001,
      "loss": 1.3885,
      "step": 4156
    },
    {
      "epoch": 0.6810288335517694,
      "grad_norm": 0.17142966389656067,
      "learning_rate": 0.0001,
      "loss": 1.3272,
      "step": 4157
    },
    {
      "epoch": 0.6811926605504587,
      "grad_norm": 0.17098331451416016,
      "learning_rate": 0.0001,
      "loss": 1.2956,
      "step": 4158
    },
    {
      "epoch": 0.6813564875491481,
      "grad_norm": 0.17244622111320496,
      "learning_rate": 0.0001,
      "loss": 1.4677,
      "step": 4159
    },
    {
      "epoch": 0.6815203145478375,
      "grad_norm": 0.1675305813550949,
      "learning_rate": 0.0001,
      "loss": 1.4178,
      "step": 4160
    },
    {
      "epoch": 0.6816841415465269,
      "grad_norm": 0.16499821841716766,
      "learning_rate": 0.0001,
      "loss": 1.3003,
      "step": 4161
    },
    {
      "epoch": 0.6818479685452162,
      "grad_norm": 0.16318471729755402,
      "learning_rate": 0.0001,
      "loss": 1.3559,
      "step": 4162
    },
    {
      "epoch": 0.6820117955439057,
      "grad_norm": 0.15947356820106506,
      "learning_rate": 0.0001,
      "loss": 1.281,
      "step": 4163
    },
    {
      "epoch": 0.682175622542595,
      "grad_norm": 0.17172963917255402,
      "learning_rate": 0.0001,
      "loss": 1.4234,
      "step": 4164
    },
    {
      "epoch": 0.6823394495412844,
      "grad_norm": 0.16903598606586456,
      "learning_rate": 0.0001,
      "loss": 1.3624,
      "step": 4165
    },
    {
      "epoch": 0.6825032765399738,
      "grad_norm": 0.1700305938720703,
      "learning_rate": 0.0001,
      "loss": 1.3253,
      "step": 4166
    },
    {
      "epoch": 0.6826671035386632,
      "grad_norm": 0.16980098187923431,
      "learning_rate": 0.0001,
      "loss": 1.3346,
      "step": 4167
    },
    {
      "epoch": 0.6828309305373526,
      "grad_norm": 0.1631162315607071,
      "learning_rate": 0.0001,
      "loss": 1.291,
      "step": 4168
    },
    {
      "epoch": 0.6829947575360419,
      "grad_norm": 0.16334110498428345,
      "learning_rate": 0.0001,
      "loss": 1.3144,
      "step": 4169
    },
    {
      "epoch": 0.6831585845347313,
      "grad_norm": 0.17061570286750793,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 4170
    },
    {
      "epoch": 0.6833224115334207,
      "grad_norm": 0.16457076370716095,
      "learning_rate": 0.0001,
      "loss": 1.3542,
      "step": 4171
    },
    {
      "epoch": 0.6834862385321101,
      "grad_norm": 0.16414494812488556,
      "learning_rate": 0.0001,
      "loss": 1.3955,
      "step": 4172
    },
    {
      "epoch": 0.6836500655307994,
      "grad_norm": 0.1625957190990448,
      "learning_rate": 0.0001,
      "loss": 1.3374,
      "step": 4173
    },
    {
      "epoch": 0.6838138925294889,
      "grad_norm": 0.15731967985630035,
      "learning_rate": 0.0001,
      "loss": 1.2357,
      "step": 4174
    },
    {
      "epoch": 0.6839777195281782,
      "grad_norm": 0.18374653160572052,
      "learning_rate": 0.0001,
      "loss": 1.2505,
      "step": 4175
    },
    {
      "epoch": 0.6841415465268676,
      "grad_norm": 0.17440789937973022,
      "learning_rate": 0.0001,
      "loss": 1.3494,
      "step": 4176
    },
    {
      "epoch": 0.684305373525557,
      "grad_norm": 0.1662435531616211,
      "learning_rate": 0.0001,
      "loss": 1.4063,
      "step": 4177
    },
    {
      "epoch": 0.6844692005242464,
      "grad_norm": 0.17336072027683258,
      "learning_rate": 0.0001,
      "loss": 1.3196,
      "step": 4178
    },
    {
      "epoch": 0.6846330275229358,
      "grad_norm": 0.1659838706254959,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 4179
    },
    {
      "epoch": 0.6847968545216252,
      "grad_norm": 0.1613444983959198,
      "learning_rate": 0.0001,
      "loss": 1.4094,
      "step": 4180
    },
    {
      "epoch": 0.6849606815203145,
      "grad_norm": 0.1620861291885376,
      "learning_rate": 0.0001,
      "loss": 1.3081,
      "step": 4181
    },
    {
      "epoch": 0.685124508519004,
      "grad_norm": 0.16596737504005432,
      "learning_rate": 0.0001,
      "loss": 1.3597,
      "step": 4182
    },
    {
      "epoch": 0.6852883355176933,
      "grad_norm": 0.1689223349094391,
      "learning_rate": 0.0001,
      "loss": 1.3104,
      "step": 4183
    },
    {
      "epoch": 0.6854521625163827,
      "grad_norm": 0.17275042831897736,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 4184
    },
    {
      "epoch": 0.6856159895150721,
      "grad_norm": 0.17155097424983978,
      "learning_rate": 0.0001,
      "loss": 1.4151,
      "step": 4185
    },
    {
      "epoch": 0.6857798165137615,
      "grad_norm": 0.17189624905586243,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 4186
    },
    {
      "epoch": 0.6859436435124509,
      "grad_norm": 0.16743870079517365,
      "learning_rate": 0.0001,
      "loss": 1.3697,
      "step": 4187
    },
    {
      "epoch": 0.6861074705111402,
      "grad_norm": 0.1634405255317688,
      "learning_rate": 0.0001,
      "loss": 1.3456,
      "step": 4188
    },
    {
      "epoch": 0.6862712975098296,
      "grad_norm": 0.173904687166214,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 4189
    },
    {
      "epoch": 0.686435124508519,
      "grad_norm": 0.177724689245224,
      "learning_rate": 0.0001,
      "loss": 1.4026,
      "step": 4190
    },
    {
      "epoch": 0.6865989515072084,
      "grad_norm": 0.17808717489242554,
      "learning_rate": 0.0001,
      "loss": 1.3305,
      "step": 4191
    },
    {
      "epoch": 0.6867627785058977,
      "grad_norm": 0.1729636788368225,
      "learning_rate": 0.0001,
      "loss": 1.3512,
      "step": 4192
    },
    {
      "epoch": 0.6869266055045872,
      "grad_norm": 0.1692892611026764,
      "learning_rate": 0.0001,
      "loss": 1.4471,
      "step": 4193
    },
    {
      "epoch": 0.6870904325032765,
      "grad_norm": 0.1565169095993042,
      "learning_rate": 0.0001,
      "loss": 1.2536,
      "step": 4194
    },
    {
      "epoch": 0.687254259501966,
      "grad_norm": 0.1769404262304306,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 4195
    },
    {
      "epoch": 0.6874180865006553,
      "grad_norm": 0.17000949382781982,
      "learning_rate": 0.0001,
      "loss": 1.3725,
      "step": 4196
    },
    {
      "epoch": 0.6875819134993447,
      "grad_norm": 0.16640405356884003,
      "learning_rate": 0.0001,
      "loss": 1.4294,
      "step": 4197
    },
    {
      "epoch": 0.687745740498034,
      "grad_norm": 0.18058337271213531,
      "learning_rate": 0.0001,
      "loss": 1.4236,
      "step": 4198
    },
    {
      "epoch": 0.6879095674967235,
      "grad_norm": 0.17365390062332153,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 4199
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 0.15987756848335266,
      "learning_rate": 0.0001,
      "loss": 1.3662,
      "step": 4200
    },
    {
      "epoch": 0.6882372214941023,
      "grad_norm": 0.1778026521205902,
      "learning_rate": 0.0001,
      "loss": 1.4196,
      "step": 4201
    },
    {
      "epoch": 0.6884010484927916,
      "grad_norm": 0.1730334460735321,
      "learning_rate": 0.0001,
      "loss": 1.423,
      "step": 4202
    },
    {
      "epoch": 0.688564875491481,
      "grad_norm": 0.16489045321941376,
      "learning_rate": 0.0001,
      "loss": 1.3479,
      "step": 4203
    },
    {
      "epoch": 0.6887287024901704,
      "grad_norm": 0.19238220155239105,
      "learning_rate": 0.0001,
      "loss": 1.3993,
      "step": 4204
    },
    {
      "epoch": 0.6888925294888598,
      "grad_norm": 0.17252066731452942,
      "learning_rate": 0.0001,
      "loss": 1.3339,
      "step": 4205
    },
    {
      "epoch": 0.6890563564875491,
      "grad_norm": 0.1604117751121521,
      "learning_rate": 0.0001,
      "loss": 1.3891,
      "step": 4206
    },
    {
      "epoch": 0.6892201834862385,
      "grad_norm": 0.17421512305736542,
      "learning_rate": 0.0001,
      "loss": 1.3812,
      "step": 4207
    },
    {
      "epoch": 0.6893840104849279,
      "grad_norm": 0.17498624324798584,
      "learning_rate": 0.0001,
      "loss": 1.3356,
      "step": 4208
    },
    {
      "epoch": 0.6895478374836173,
      "grad_norm": 0.16547861695289612,
      "learning_rate": 0.0001,
      "loss": 1.3434,
      "step": 4209
    },
    {
      "epoch": 0.6897116644823067,
      "grad_norm": 0.17036692798137665,
      "learning_rate": 0.0001,
      "loss": 1.3662,
      "step": 4210
    },
    {
      "epoch": 0.689875491480996,
      "grad_norm": 0.16655053198337555,
      "learning_rate": 0.0001,
      "loss": 1.27,
      "step": 4211
    },
    {
      "epoch": 0.6900393184796855,
      "grad_norm": 0.1640458106994629,
      "learning_rate": 0.0001,
      "loss": 1.372,
      "step": 4212
    },
    {
      "epoch": 0.6902031454783748,
      "grad_norm": 0.1681140810251236,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 4213
    },
    {
      "epoch": 0.6903669724770642,
      "grad_norm": 0.17236199975013733,
      "learning_rate": 0.0001,
      "loss": 1.3254,
      "step": 4214
    },
    {
      "epoch": 0.6905307994757536,
      "grad_norm": 0.1587778478860855,
      "learning_rate": 0.0001,
      "loss": 1.4332,
      "step": 4215
    },
    {
      "epoch": 0.690694626474443,
      "grad_norm": 0.1644793003797531,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 4216
    },
    {
      "epoch": 0.6908584534731324,
      "grad_norm": 0.1690499186515808,
      "learning_rate": 0.0001,
      "loss": 1.3784,
      "step": 4217
    },
    {
      "epoch": 0.6910222804718218,
      "grad_norm": 0.164739727973938,
      "learning_rate": 0.0001,
      "loss": 1.2731,
      "step": 4218
    },
    {
      "epoch": 0.6911861074705111,
      "grad_norm": 0.17082802951335907,
      "learning_rate": 0.0001,
      "loss": 1.3073,
      "step": 4219
    },
    {
      "epoch": 0.6913499344692006,
      "grad_norm": 0.17842885851860046,
      "learning_rate": 0.0001,
      "loss": 1.3764,
      "step": 4220
    },
    {
      "epoch": 0.6915137614678899,
      "grad_norm": 0.16970698535442352,
      "learning_rate": 0.0001,
      "loss": 1.3138,
      "step": 4221
    },
    {
      "epoch": 0.6916775884665793,
      "grad_norm": 0.16660892963409424,
      "learning_rate": 0.0001,
      "loss": 1.4329,
      "step": 4222
    },
    {
      "epoch": 0.6918414154652687,
      "grad_norm": 0.15940697491168976,
      "learning_rate": 0.0001,
      "loss": 1.2641,
      "step": 4223
    },
    {
      "epoch": 0.6920052424639581,
      "grad_norm": 0.16959184408187866,
      "learning_rate": 0.0001,
      "loss": 1.3571,
      "step": 4224
    },
    {
      "epoch": 0.6921690694626474,
      "grad_norm": 0.16229264438152313,
      "learning_rate": 0.0001,
      "loss": 1.3546,
      "step": 4225
    },
    {
      "epoch": 0.6923328964613368,
      "grad_norm": 0.16652409732341766,
      "learning_rate": 0.0001,
      "loss": 1.3503,
      "step": 4226
    },
    {
      "epoch": 0.6924967234600262,
      "grad_norm": 0.1688108891248703,
      "learning_rate": 0.0001,
      "loss": 1.293,
      "step": 4227
    },
    {
      "epoch": 0.6926605504587156,
      "grad_norm": 0.16770289838314056,
      "learning_rate": 0.0001,
      "loss": 1.3726,
      "step": 4228
    },
    {
      "epoch": 0.692824377457405,
      "grad_norm": 0.1718888282775879,
      "learning_rate": 0.0001,
      "loss": 1.3558,
      "step": 4229
    },
    {
      "epoch": 0.6929882044560943,
      "grad_norm": 0.16214312613010406,
      "learning_rate": 0.0001,
      "loss": 1.2731,
      "step": 4230
    },
    {
      "epoch": 0.6931520314547838,
      "grad_norm": 0.16644711792469025,
      "learning_rate": 0.0001,
      "loss": 1.3468,
      "step": 4231
    },
    {
      "epoch": 0.6933158584534731,
      "grad_norm": 0.16556279361248016,
      "learning_rate": 0.0001,
      "loss": 1.378,
      "step": 4232
    },
    {
      "epoch": 0.6934796854521625,
      "grad_norm": 0.17100875079631805,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 4233
    },
    {
      "epoch": 0.6936435124508519,
      "grad_norm": 0.17363494634628296,
      "learning_rate": 0.0001,
      "loss": 1.3457,
      "step": 4234
    },
    {
      "epoch": 0.6938073394495413,
      "grad_norm": 0.17130455374717712,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 4235
    },
    {
      "epoch": 0.6939711664482306,
      "grad_norm": 0.1705896556377411,
      "learning_rate": 0.0001,
      "loss": 1.287,
      "step": 4236
    },
    {
      "epoch": 0.6941349934469201,
      "grad_norm": 0.1571621298789978,
      "learning_rate": 0.0001,
      "loss": 1.3456,
      "step": 4237
    },
    {
      "epoch": 0.6942988204456094,
      "grad_norm": 0.1764112263917923,
      "learning_rate": 0.0001,
      "loss": 1.453,
      "step": 4238
    },
    {
      "epoch": 0.6944626474442989,
      "grad_norm": 0.1660788208246231,
      "learning_rate": 0.0001,
      "loss": 1.4296,
      "step": 4239
    },
    {
      "epoch": 0.6946264744429882,
      "grad_norm": 0.16626258194446564,
      "learning_rate": 0.0001,
      "loss": 1.3894,
      "step": 4240
    },
    {
      "epoch": 0.6947903014416776,
      "grad_norm": 0.17215536534786224,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 4241
    },
    {
      "epoch": 0.694954128440367,
      "grad_norm": 0.16856779158115387,
      "learning_rate": 0.0001,
      "loss": 1.3689,
      "step": 4242
    },
    {
      "epoch": 0.6951179554390564,
      "grad_norm": 0.17358136177062988,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 4243
    },
    {
      "epoch": 0.6952817824377457,
      "grad_norm": 0.1659286916255951,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 4244
    },
    {
      "epoch": 0.6954456094364351,
      "grad_norm": 0.1735316812992096,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 4245
    },
    {
      "epoch": 0.6956094364351245,
      "grad_norm": 0.17300505936145782,
      "learning_rate": 0.0001,
      "loss": 1.3221,
      "step": 4246
    },
    {
      "epoch": 0.6957732634338138,
      "grad_norm": 0.16745714843273163,
      "learning_rate": 0.0001,
      "loss": 1.3822,
      "step": 4247
    },
    {
      "epoch": 0.6959370904325033,
      "grad_norm": 0.16998693346977234,
      "learning_rate": 0.0001,
      "loss": 1.3958,
      "step": 4248
    },
    {
      "epoch": 0.6961009174311926,
      "grad_norm": 0.1716630458831787,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 4249
    },
    {
      "epoch": 0.6962647444298821,
      "grad_norm": 0.1797296553850174,
      "learning_rate": 0.0001,
      "loss": 1.5039,
      "step": 4250
    },
    {
      "epoch": 0.6964285714285714,
      "grad_norm": 0.16687101125717163,
      "learning_rate": 0.0001,
      "loss": 1.3517,
      "step": 4251
    },
    {
      "epoch": 0.6965923984272608,
      "grad_norm": 0.17138449847698212,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 4252
    },
    {
      "epoch": 0.6967562254259502,
      "grad_norm": 0.17325103282928467,
      "learning_rate": 0.0001,
      "loss": 1.4163,
      "step": 4253
    },
    {
      "epoch": 0.6969200524246396,
      "grad_norm": 0.16504858434200287,
      "learning_rate": 0.0001,
      "loss": 1.3471,
      "step": 4254
    },
    {
      "epoch": 0.697083879423329,
      "grad_norm": 0.16345500946044922,
      "learning_rate": 0.0001,
      "loss": 1.3154,
      "step": 4255
    },
    {
      "epoch": 0.6972477064220184,
      "grad_norm": 0.16660076379776,
      "learning_rate": 0.0001,
      "loss": 1.3534,
      "step": 4256
    },
    {
      "epoch": 0.6974115334207077,
      "grad_norm": 0.17552854120731354,
      "learning_rate": 0.0001,
      "loss": 1.4098,
      "step": 4257
    },
    {
      "epoch": 0.6975753604193972,
      "grad_norm": 0.16752435266971588,
      "learning_rate": 0.0001,
      "loss": 1.4201,
      "step": 4258
    },
    {
      "epoch": 0.6977391874180865,
      "grad_norm": 0.16812126338481903,
      "learning_rate": 0.0001,
      "loss": 1.4476,
      "step": 4259
    },
    {
      "epoch": 0.6979030144167759,
      "grad_norm": 0.16575215756893158,
      "learning_rate": 0.0001,
      "loss": 1.3317,
      "step": 4260
    },
    {
      "epoch": 0.6980668414154653,
      "grad_norm": 0.16597339510917664,
      "learning_rate": 0.0001,
      "loss": 1.3655,
      "step": 4261
    },
    {
      "epoch": 0.6982306684141546,
      "grad_norm": 0.1577291041612625,
      "learning_rate": 0.0001,
      "loss": 1.3542,
      "step": 4262
    },
    {
      "epoch": 0.698394495412844,
      "grad_norm": 0.16970524191856384,
      "learning_rate": 0.0001,
      "loss": 1.4083,
      "step": 4263
    },
    {
      "epoch": 0.6985583224115334,
      "grad_norm": 0.17514200508594513,
      "learning_rate": 0.0001,
      "loss": 1.3906,
      "step": 4264
    },
    {
      "epoch": 0.6987221494102228,
      "grad_norm": 0.1772897094488144,
      "learning_rate": 0.0001,
      "loss": 1.3874,
      "step": 4265
    },
    {
      "epoch": 0.6988859764089121,
      "grad_norm": 0.16508911550045013,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 4266
    },
    {
      "epoch": 0.6990498034076016,
      "grad_norm": 0.16553671658039093,
      "learning_rate": 0.0001,
      "loss": 1.3953,
      "step": 4267
    },
    {
      "epoch": 0.6992136304062909,
      "grad_norm": 0.16798104345798492,
      "learning_rate": 0.0001,
      "loss": 1.3563,
      "step": 4268
    },
    {
      "epoch": 0.6993774574049804,
      "grad_norm": 0.17048420011997223,
      "learning_rate": 0.0001,
      "loss": 1.3348,
      "step": 4269
    },
    {
      "epoch": 0.6995412844036697,
      "grad_norm": 0.16777676343917847,
      "learning_rate": 0.0001,
      "loss": 1.2811,
      "step": 4270
    },
    {
      "epoch": 0.6997051114023591,
      "grad_norm": 0.17606405913829803,
      "learning_rate": 0.0001,
      "loss": 1.3514,
      "step": 4271
    },
    {
      "epoch": 0.6998689384010485,
      "grad_norm": 0.17037533223628998,
      "learning_rate": 0.0001,
      "loss": 1.3056,
      "step": 4272
    },
    {
      "epoch": 0.7000327653997379,
      "grad_norm": 0.16879317164421082,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 4273
    },
    {
      "epoch": 0.7001965923984272,
      "grad_norm": 0.1698155403137207,
      "learning_rate": 0.0001,
      "loss": 1.2828,
      "step": 4274
    },
    {
      "epoch": 0.7003604193971167,
      "grad_norm": 0.16418543457984924,
      "learning_rate": 0.0001,
      "loss": 1.2075,
      "step": 4275
    },
    {
      "epoch": 0.700524246395806,
      "grad_norm": 0.17373549938201904,
      "learning_rate": 0.0001,
      "loss": 1.3903,
      "step": 4276
    },
    {
      "epoch": 0.7006880733944955,
      "grad_norm": 0.16422128677368164,
      "learning_rate": 0.0001,
      "loss": 1.3819,
      "step": 4277
    },
    {
      "epoch": 0.7008519003931848,
      "grad_norm": 0.18305690586566925,
      "learning_rate": 0.0001,
      "loss": 1.394,
      "step": 4278
    },
    {
      "epoch": 0.7010157273918742,
      "grad_norm": 0.17729805409908295,
      "learning_rate": 0.0001,
      "loss": 1.3663,
      "step": 4279
    },
    {
      "epoch": 0.7011795543905636,
      "grad_norm": 0.170290008187294,
      "learning_rate": 0.0001,
      "loss": 1.2635,
      "step": 4280
    },
    {
      "epoch": 0.7013433813892529,
      "grad_norm": 0.16621486842632294,
      "learning_rate": 0.0001,
      "loss": 1.361,
      "step": 4281
    },
    {
      "epoch": 0.7015072083879423,
      "grad_norm": 0.16062702238559723,
      "learning_rate": 0.0001,
      "loss": 1.3077,
      "step": 4282
    },
    {
      "epoch": 0.7016710353866317,
      "grad_norm": 0.15878190100193024,
      "learning_rate": 0.0001,
      "loss": 1.2787,
      "step": 4283
    },
    {
      "epoch": 0.7018348623853211,
      "grad_norm": 0.17914032936096191,
      "learning_rate": 0.0001,
      "loss": 1.4565,
      "step": 4284
    },
    {
      "epoch": 0.7019986893840104,
      "grad_norm": 0.15783241391181946,
      "learning_rate": 0.0001,
      "loss": 1.3556,
      "step": 4285
    },
    {
      "epoch": 0.7021625163826999,
      "grad_norm": 0.1717151552438736,
      "learning_rate": 0.0001,
      "loss": 1.4313,
      "step": 4286
    },
    {
      "epoch": 0.7023263433813892,
      "grad_norm": 0.1788409799337387,
      "learning_rate": 0.0001,
      "loss": 1.3182,
      "step": 4287
    },
    {
      "epoch": 0.7024901703800787,
      "grad_norm": 0.16937348246574402,
      "learning_rate": 0.0001,
      "loss": 1.4184,
      "step": 4288
    },
    {
      "epoch": 0.702653997378768,
      "grad_norm": 0.16255520284175873,
      "learning_rate": 0.0001,
      "loss": 1.3539,
      "step": 4289
    },
    {
      "epoch": 0.7028178243774574,
      "grad_norm": 0.17916806042194366,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 4290
    },
    {
      "epoch": 0.7029816513761468,
      "grad_norm": 0.16661284863948822,
      "learning_rate": 0.0001,
      "loss": 1.3691,
      "step": 4291
    },
    {
      "epoch": 0.7031454783748362,
      "grad_norm": 0.16784025728702545,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 4292
    },
    {
      "epoch": 0.7033093053735255,
      "grad_norm": 0.15563228726387024,
      "learning_rate": 0.0001,
      "loss": 1.2921,
      "step": 4293
    },
    {
      "epoch": 0.703473132372215,
      "grad_norm": 0.1602732241153717,
      "learning_rate": 0.0001,
      "loss": 1.4087,
      "step": 4294
    },
    {
      "epoch": 0.7036369593709043,
      "grad_norm": 0.1659712791442871,
      "learning_rate": 0.0001,
      "loss": 1.3832,
      "step": 4295
    },
    {
      "epoch": 0.7038007863695938,
      "grad_norm": 0.17299698293209076,
      "learning_rate": 0.0001,
      "loss": 1.416,
      "step": 4296
    },
    {
      "epoch": 0.7039646133682831,
      "grad_norm": 0.16535215079784393,
      "learning_rate": 0.0001,
      "loss": 1.3159,
      "step": 4297
    },
    {
      "epoch": 0.7041284403669725,
      "grad_norm": 0.17219339311122894,
      "learning_rate": 0.0001,
      "loss": 1.3112,
      "step": 4298
    },
    {
      "epoch": 0.7042922673656619,
      "grad_norm": 0.16171295940876007,
      "learning_rate": 0.0001,
      "loss": 1.2771,
      "step": 4299
    },
    {
      "epoch": 0.7044560943643512,
      "grad_norm": 0.17382307350635529,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 4300
    },
    {
      "epoch": 0.7046199213630406,
      "grad_norm": 0.16152793169021606,
      "learning_rate": 0.0001,
      "loss": 1.3046,
      "step": 4301
    },
    {
      "epoch": 0.70478374836173,
      "grad_norm": 0.1658622920513153,
      "learning_rate": 0.0001,
      "loss": 1.3657,
      "step": 4302
    },
    {
      "epoch": 0.7049475753604194,
      "grad_norm": 0.15937988460063934,
      "learning_rate": 0.0001,
      "loss": 1.2752,
      "step": 4303
    },
    {
      "epoch": 0.7051114023591087,
      "grad_norm": 0.17332307994365692,
      "learning_rate": 0.0001,
      "loss": 1.4194,
      "step": 4304
    },
    {
      "epoch": 0.7052752293577982,
      "grad_norm": 0.16415880620479584,
      "learning_rate": 0.0001,
      "loss": 1.2855,
      "step": 4305
    },
    {
      "epoch": 0.7054390563564875,
      "grad_norm": 0.17166461050510406,
      "learning_rate": 0.0001,
      "loss": 1.4031,
      "step": 4306
    },
    {
      "epoch": 0.705602883355177,
      "grad_norm": 0.166514053940773,
      "learning_rate": 0.0001,
      "loss": 1.3597,
      "step": 4307
    },
    {
      "epoch": 0.7057667103538663,
      "grad_norm": 0.17272552847862244,
      "learning_rate": 0.0001,
      "loss": 1.464,
      "step": 4308
    },
    {
      "epoch": 0.7059305373525557,
      "grad_norm": 0.1716524213552475,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 4309
    },
    {
      "epoch": 0.7060943643512451,
      "grad_norm": 0.16924449801445007,
      "learning_rate": 0.0001,
      "loss": 1.42,
      "step": 4310
    },
    {
      "epoch": 0.7062581913499345,
      "grad_norm": 0.17181625962257385,
      "learning_rate": 0.0001,
      "loss": 1.3279,
      "step": 4311
    },
    {
      "epoch": 0.7064220183486238,
      "grad_norm": 0.1606651246547699,
      "learning_rate": 0.0001,
      "loss": 1.3026,
      "step": 4312
    },
    {
      "epoch": 0.7065858453473133,
      "grad_norm": 0.1900206059217453,
      "learning_rate": 0.0001,
      "loss": 1.299,
      "step": 4313
    },
    {
      "epoch": 0.7067496723460026,
      "grad_norm": 0.1689678132534027,
      "learning_rate": 0.0001,
      "loss": 1.3686,
      "step": 4314
    },
    {
      "epoch": 0.706913499344692,
      "grad_norm": 0.16280287504196167,
      "learning_rate": 0.0001,
      "loss": 1.3458,
      "step": 4315
    },
    {
      "epoch": 0.7070773263433814,
      "grad_norm": 0.1832965910434723,
      "learning_rate": 0.0001,
      "loss": 1.3379,
      "step": 4316
    },
    {
      "epoch": 0.7072411533420708,
      "grad_norm": 0.1669464111328125,
      "learning_rate": 0.0001,
      "loss": 1.4013,
      "step": 4317
    },
    {
      "epoch": 0.7074049803407602,
      "grad_norm": 0.1695455014705658,
      "learning_rate": 0.0001,
      "loss": 1.3914,
      "step": 4318
    },
    {
      "epoch": 0.7075688073394495,
      "grad_norm": 0.1595560908317566,
      "learning_rate": 0.0001,
      "loss": 1.3412,
      "step": 4319
    },
    {
      "epoch": 0.7077326343381389,
      "grad_norm": 0.15695446729660034,
      "learning_rate": 0.0001,
      "loss": 1.3539,
      "step": 4320
    },
    {
      "epoch": 0.7078964613368283,
      "grad_norm": 0.16022048890590668,
      "learning_rate": 0.0001,
      "loss": 1.2817,
      "step": 4321
    },
    {
      "epoch": 0.7080602883355177,
      "grad_norm": 0.1693187803030014,
      "learning_rate": 0.0001,
      "loss": 1.3643,
      "step": 4322
    },
    {
      "epoch": 0.708224115334207,
      "grad_norm": 0.16751790046691895,
      "learning_rate": 0.0001,
      "loss": 1.3823,
      "step": 4323
    },
    {
      "epoch": 0.7083879423328965,
      "grad_norm": 0.18763050436973572,
      "learning_rate": 0.0001,
      "loss": 1.4252,
      "step": 4324
    },
    {
      "epoch": 0.7085517693315858,
      "grad_norm": 0.1649247407913208,
      "learning_rate": 0.0001,
      "loss": 1.2603,
      "step": 4325
    },
    {
      "epoch": 0.7087155963302753,
      "grad_norm": 0.1646609604358673,
      "learning_rate": 0.0001,
      "loss": 1.3418,
      "step": 4326
    },
    {
      "epoch": 0.7088794233289646,
      "grad_norm": 0.17025740444660187,
      "learning_rate": 0.0001,
      "loss": 1.4268,
      "step": 4327
    },
    {
      "epoch": 0.709043250327654,
      "grad_norm": 0.16423244774341583,
      "learning_rate": 0.0001,
      "loss": 1.2706,
      "step": 4328
    },
    {
      "epoch": 0.7092070773263434,
      "grad_norm": 0.1665363311767578,
      "learning_rate": 0.0001,
      "loss": 1.3786,
      "step": 4329
    },
    {
      "epoch": 0.7093709043250328,
      "grad_norm": 0.16953538358211517,
      "learning_rate": 0.0001,
      "loss": 1.3638,
      "step": 4330
    },
    {
      "epoch": 0.7095347313237221,
      "grad_norm": 0.1693251132965088,
      "learning_rate": 0.0001,
      "loss": 1.349,
      "step": 4331
    },
    {
      "epoch": 0.7096985583224116,
      "grad_norm": 0.1587654948234558,
      "learning_rate": 0.0001,
      "loss": 1.3366,
      "step": 4332
    },
    {
      "epoch": 0.7098623853211009,
      "grad_norm": 0.16880369186401367,
      "learning_rate": 0.0001,
      "loss": 1.3799,
      "step": 4333
    },
    {
      "epoch": 0.7100262123197904,
      "grad_norm": 0.16801312565803528,
      "learning_rate": 0.0001,
      "loss": 1.3306,
      "step": 4334
    },
    {
      "epoch": 0.7101900393184797,
      "grad_norm": 0.16383902728557587,
      "learning_rate": 0.0001,
      "loss": 1.365,
      "step": 4335
    },
    {
      "epoch": 0.7103538663171691,
      "grad_norm": 0.16761019825935364,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 4336
    },
    {
      "epoch": 0.7105176933158585,
      "grad_norm": 0.1643943339586258,
      "learning_rate": 0.0001,
      "loss": 1.3728,
      "step": 4337
    },
    {
      "epoch": 0.7106815203145478,
      "grad_norm": 0.1814548373222351,
      "learning_rate": 0.0001,
      "loss": 1.3071,
      "step": 4338
    },
    {
      "epoch": 0.7108453473132372,
      "grad_norm": 0.17213401198387146,
      "learning_rate": 0.0001,
      "loss": 1.4067,
      "step": 4339
    },
    {
      "epoch": 0.7110091743119266,
      "grad_norm": 0.17002005875110626,
      "learning_rate": 0.0001,
      "loss": 1.2804,
      "step": 4340
    },
    {
      "epoch": 0.711173001310616,
      "grad_norm": 0.16581054031848907,
      "learning_rate": 0.0001,
      "loss": 1.2996,
      "step": 4341
    },
    {
      "epoch": 0.7113368283093053,
      "grad_norm": 0.1630421131849289,
      "learning_rate": 0.0001,
      "loss": 1.2449,
      "step": 4342
    },
    {
      "epoch": 0.7115006553079948,
      "grad_norm": 0.1667296141386032,
      "learning_rate": 0.0001,
      "loss": 1.3729,
      "step": 4343
    },
    {
      "epoch": 0.7116644823066841,
      "grad_norm": 0.1692282259464264,
      "learning_rate": 0.0001,
      "loss": 1.3216,
      "step": 4344
    },
    {
      "epoch": 0.7118283093053736,
      "grad_norm": 0.1710938811302185,
      "learning_rate": 0.0001,
      "loss": 1.3335,
      "step": 4345
    },
    {
      "epoch": 0.7119921363040629,
      "grad_norm": 0.16871494054794312,
      "learning_rate": 0.0001,
      "loss": 1.2967,
      "step": 4346
    },
    {
      "epoch": 0.7121559633027523,
      "grad_norm": 0.17432034015655518,
      "learning_rate": 0.0001,
      "loss": 1.2692,
      "step": 4347
    },
    {
      "epoch": 0.7123197903014417,
      "grad_norm": 0.17749032378196716,
      "learning_rate": 0.0001,
      "loss": 1.3308,
      "step": 4348
    },
    {
      "epoch": 0.7124836173001311,
      "grad_norm": 0.1664259433746338,
      "learning_rate": 0.0001,
      "loss": 1.4067,
      "step": 4349
    },
    {
      "epoch": 0.7126474442988204,
      "grad_norm": 0.16246512532234192,
      "learning_rate": 0.0001,
      "loss": 1.3105,
      "step": 4350
    },
    {
      "epoch": 0.7128112712975099,
      "grad_norm": 0.17970354855060577,
      "learning_rate": 0.0001,
      "loss": 1.4747,
      "step": 4351
    },
    {
      "epoch": 0.7129750982961992,
      "grad_norm": 0.169361874461174,
      "learning_rate": 0.0001,
      "loss": 1.3797,
      "step": 4352
    },
    {
      "epoch": 0.7131389252948886,
      "grad_norm": 0.17720437049865723,
      "learning_rate": 0.0001,
      "loss": 1.397,
      "step": 4353
    },
    {
      "epoch": 0.713302752293578,
      "grad_norm": 0.17904874682426453,
      "learning_rate": 0.0001,
      "loss": 1.4152,
      "step": 4354
    },
    {
      "epoch": 0.7134665792922673,
      "grad_norm": 0.17096227407455444,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 4355
    },
    {
      "epoch": 0.7136304062909568,
      "grad_norm": 0.1705613136291504,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 4356
    },
    {
      "epoch": 0.7137942332896461,
      "grad_norm": 0.17309275269508362,
      "learning_rate": 0.0001,
      "loss": 1.47,
      "step": 4357
    },
    {
      "epoch": 0.7139580602883355,
      "grad_norm": 0.16412416100502014,
      "learning_rate": 0.0001,
      "loss": 1.3985,
      "step": 4358
    },
    {
      "epoch": 0.7141218872870249,
      "grad_norm": 0.1760832667350769,
      "learning_rate": 0.0001,
      "loss": 1.3333,
      "step": 4359
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.17686213552951813,
      "learning_rate": 0.0001,
      "loss": 1.4,
      "step": 4360
    },
    {
      "epoch": 0.7144495412844036,
      "grad_norm": 0.1632203608751297,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 4361
    },
    {
      "epoch": 0.7146133682830931,
      "grad_norm": 0.17079944908618927,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 4362
    },
    {
      "epoch": 0.7147771952817824,
      "grad_norm": 0.17286638915538788,
      "learning_rate": 0.0001,
      "loss": 1.4315,
      "step": 4363
    },
    {
      "epoch": 0.7149410222804718,
      "grad_norm": 0.16437038779258728,
      "learning_rate": 0.0001,
      "loss": 1.4208,
      "step": 4364
    },
    {
      "epoch": 0.7151048492791612,
      "grad_norm": 0.17580272257328033,
      "learning_rate": 0.0001,
      "loss": 1.4288,
      "step": 4365
    },
    {
      "epoch": 0.7152686762778506,
      "grad_norm": 0.19919437170028687,
      "learning_rate": 0.0001,
      "loss": 1.3209,
      "step": 4366
    },
    {
      "epoch": 0.71543250327654,
      "grad_norm": 0.1715487837791443,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 4367
    },
    {
      "epoch": 0.7155963302752294,
      "grad_norm": 0.1718926876783371,
      "learning_rate": 0.0001,
      "loss": 1.3851,
      "step": 4368
    },
    {
      "epoch": 0.7157601572739187,
      "grad_norm": 0.1706569641828537,
      "learning_rate": 0.0001,
      "loss": 1.3893,
      "step": 4369
    },
    {
      "epoch": 0.7159239842726082,
      "grad_norm": 0.17102095484733582,
      "learning_rate": 0.0001,
      "loss": 1.3689,
      "step": 4370
    },
    {
      "epoch": 0.7160878112712975,
      "grad_norm": 0.17602436244487762,
      "learning_rate": 0.0001,
      "loss": 1.3268,
      "step": 4371
    },
    {
      "epoch": 0.716251638269987,
      "grad_norm": 0.1628226935863495,
      "learning_rate": 0.0001,
      "loss": 1.3969,
      "step": 4372
    },
    {
      "epoch": 0.7164154652686763,
      "grad_norm": 0.16727468371391296,
      "learning_rate": 0.0001,
      "loss": 1.2699,
      "step": 4373
    },
    {
      "epoch": 0.7165792922673656,
      "grad_norm": 0.16377349197864532,
      "learning_rate": 0.0001,
      "loss": 1.2973,
      "step": 4374
    },
    {
      "epoch": 0.716743119266055,
      "grad_norm": 0.17016445100307465,
      "learning_rate": 0.0001,
      "loss": 1.4356,
      "step": 4375
    },
    {
      "epoch": 0.7169069462647444,
      "grad_norm": 0.1677769422531128,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 4376
    },
    {
      "epoch": 0.7170707732634338,
      "grad_norm": 0.16942602396011353,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 4377
    },
    {
      "epoch": 0.7172346002621232,
      "grad_norm": 0.16406883299350739,
      "learning_rate": 0.0001,
      "loss": 1.3244,
      "step": 4378
    },
    {
      "epoch": 0.7173984272608126,
      "grad_norm": 0.16724401712417603,
      "learning_rate": 0.0001,
      "loss": 1.4277,
      "step": 4379
    },
    {
      "epoch": 0.7175622542595019,
      "grad_norm": 0.1734064221382141,
      "learning_rate": 0.0001,
      "loss": 1.4062,
      "step": 4380
    },
    {
      "epoch": 0.7177260812581914,
      "grad_norm": 0.1786728799343109,
      "learning_rate": 0.0001,
      "loss": 1.3184,
      "step": 4381
    },
    {
      "epoch": 0.7178899082568807,
      "grad_norm": 0.16643370687961578,
      "learning_rate": 0.0001,
      "loss": 1.2883,
      "step": 4382
    },
    {
      "epoch": 0.7180537352555701,
      "grad_norm": 0.18417470157146454,
      "learning_rate": 0.0001,
      "loss": 1.4922,
      "step": 4383
    },
    {
      "epoch": 0.7182175622542595,
      "grad_norm": 0.17789915204048157,
      "learning_rate": 0.0001,
      "loss": 1.3758,
      "step": 4384
    },
    {
      "epoch": 0.7183813892529489,
      "grad_norm": 0.17003776133060455,
      "learning_rate": 0.0001,
      "loss": 1.4254,
      "step": 4385
    },
    {
      "epoch": 0.7185452162516383,
      "grad_norm": 0.16624030470848083,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 4386
    },
    {
      "epoch": 0.7187090432503277,
      "grad_norm": 0.17600126564502716,
      "learning_rate": 0.0001,
      "loss": 1.3451,
      "step": 4387
    },
    {
      "epoch": 0.718872870249017,
      "grad_norm": 0.17194193601608276,
      "learning_rate": 0.0001,
      "loss": 1.3398,
      "step": 4388
    },
    {
      "epoch": 0.7190366972477065,
      "grad_norm": 0.17233455181121826,
      "learning_rate": 0.0001,
      "loss": 1.3569,
      "step": 4389
    },
    {
      "epoch": 0.7192005242463958,
      "grad_norm": 0.17187365889549255,
      "learning_rate": 0.0001,
      "loss": 1.3811,
      "step": 4390
    },
    {
      "epoch": 0.7193643512450852,
      "grad_norm": 0.1725103259086609,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 4391
    },
    {
      "epoch": 0.7195281782437746,
      "grad_norm": 0.1772274523973465,
      "learning_rate": 0.0001,
      "loss": 1.3886,
      "step": 4392
    },
    {
      "epoch": 0.7196920052424639,
      "grad_norm": 0.17333552241325378,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 4393
    },
    {
      "epoch": 0.7198558322411533,
      "grad_norm": 0.17012445628643036,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 4394
    },
    {
      "epoch": 0.7200196592398427,
      "grad_norm": 0.17402830719947815,
      "learning_rate": 0.0001,
      "loss": 1.3049,
      "step": 4395
    },
    {
      "epoch": 0.7201834862385321,
      "grad_norm": 0.17229340970516205,
      "learning_rate": 0.0001,
      "loss": 1.2994,
      "step": 4396
    },
    {
      "epoch": 0.7203473132372215,
      "grad_norm": 0.1713295578956604,
      "learning_rate": 0.0001,
      "loss": 1.3953,
      "step": 4397
    },
    {
      "epoch": 0.7205111402359109,
      "grad_norm": 0.17979997396469116,
      "learning_rate": 0.0001,
      "loss": 1.3494,
      "step": 4398
    },
    {
      "epoch": 0.7206749672346002,
      "grad_norm": 0.17026367783546448,
      "learning_rate": 0.0001,
      "loss": 1.2486,
      "step": 4399
    },
    {
      "epoch": 0.7208387942332897,
      "grad_norm": 0.17037415504455566,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 4400
    },
    {
      "epoch": 0.721002621231979,
      "grad_norm": 0.16863392293453217,
      "learning_rate": 0.0001,
      "loss": 1.3998,
      "step": 4401
    },
    {
      "epoch": 0.7211664482306684,
      "grad_norm": 0.17559753358364105,
      "learning_rate": 0.0001,
      "loss": 1.4077,
      "step": 4402
    },
    {
      "epoch": 0.7213302752293578,
      "grad_norm": 0.17751210927963257,
      "learning_rate": 0.0001,
      "loss": 1.3265,
      "step": 4403
    },
    {
      "epoch": 0.7214941022280472,
      "grad_norm": 0.16118118166923523,
      "learning_rate": 0.0001,
      "loss": 1.3172,
      "step": 4404
    },
    {
      "epoch": 0.7216579292267365,
      "grad_norm": 0.1682058870792389,
      "learning_rate": 0.0001,
      "loss": 1.3602,
      "step": 4405
    },
    {
      "epoch": 0.721821756225426,
      "grad_norm": 0.1637183278799057,
      "learning_rate": 0.0001,
      "loss": 1.2673,
      "step": 4406
    },
    {
      "epoch": 0.7219855832241153,
      "grad_norm": 0.17323052883148193,
      "learning_rate": 0.0001,
      "loss": 1.4508,
      "step": 4407
    },
    {
      "epoch": 0.7221494102228048,
      "grad_norm": 0.16809317469596863,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 4408
    },
    {
      "epoch": 0.7223132372214941,
      "grad_norm": 0.1771746128797531,
      "learning_rate": 0.0001,
      "loss": 1.4096,
      "step": 4409
    },
    {
      "epoch": 0.7224770642201835,
      "grad_norm": 0.17301835119724274,
      "learning_rate": 0.0001,
      "loss": 1.3449,
      "step": 4410
    },
    {
      "epoch": 0.7226408912188729,
      "grad_norm": 0.16404952108860016,
      "learning_rate": 0.0001,
      "loss": 1.2814,
      "step": 4411
    },
    {
      "epoch": 0.7228047182175622,
      "grad_norm": 0.1608683168888092,
      "learning_rate": 0.0001,
      "loss": 1.3872,
      "step": 4412
    },
    {
      "epoch": 0.7229685452162516,
      "grad_norm": 0.17007854580879211,
      "learning_rate": 0.0001,
      "loss": 1.2726,
      "step": 4413
    },
    {
      "epoch": 0.723132372214941,
      "grad_norm": 0.16957929730415344,
      "learning_rate": 0.0001,
      "loss": 1.4856,
      "step": 4414
    },
    {
      "epoch": 0.7232961992136304,
      "grad_norm": 0.1675184965133667,
      "learning_rate": 0.0001,
      "loss": 1.3461,
      "step": 4415
    },
    {
      "epoch": 0.7234600262123198,
      "grad_norm": 0.16420261561870575,
      "learning_rate": 0.0001,
      "loss": 1.3665,
      "step": 4416
    },
    {
      "epoch": 0.7236238532110092,
      "grad_norm": 0.1686844378709793,
      "learning_rate": 0.0001,
      "loss": 1.3898,
      "step": 4417
    },
    {
      "epoch": 0.7237876802096985,
      "grad_norm": 0.17875854671001434,
      "learning_rate": 0.0001,
      "loss": 1.3687,
      "step": 4418
    },
    {
      "epoch": 0.723951507208388,
      "grad_norm": 0.1713966280221939,
      "learning_rate": 0.0001,
      "loss": 1.3726,
      "step": 4419
    },
    {
      "epoch": 0.7241153342070773,
      "grad_norm": 0.18301446735858917,
      "learning_rate": 0.0001,
      "loss": 1.4987,
      "step": 4420
    },
    {
      "epoch": 0.7242791612057667,
      "grad_norm": 0.17245231568813324,
      "learning_rate": 0.0001,
      "loss": 1.3224,
      "step": 4421
    },
    {
      "epoch": 0.7244429882044561,
      "grad_norm": 0.17150509357452393,
      "learning_rate": 0.0001,
      "loss": 1.3304,
      "step": 4422
    },
    {
      "epoch": 0.7246068152031455,
      "grad_norm": 0.19079969823360443,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 4423
    },
    {
      "epoch": 0.7247706422018348,
      "grad_norm": 0.16446708142757416,
      "learning_rate": 0.0001,
      "loss": 1.4698,
      "step": 4424
    },
    {
      "epoch": 0.7249344692005243,
      "grad_norm": 0.17468994855880737,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 4425
    },
    {
      "epoch": 0.7250982961992136,
      "grad_norm": 0.17772060632705688,
      "learning_rate": 0.0001,
      "loss": 1.4024,
      "step": 4426
    },
    {
      "epoch": 0.7252621231979031,
      "grad_norm": 0.1682833880186081,
      "learning_rate": 0.0001,
      "loss": 1.4068,
      "step": 4427
    },
    {
      "epoch": 0.7254259501965924,
      "grad_norm": 0.17357516288757324,
      "learning_rate": 0.0001,
      "loss": 1.3813,
      "step": 4428
    },
    {
      "epoch": 0.7255897771952818,
      "grad_norm": 0.16449694335460663,
      "learning_rate": 0.0001,
      "loss": 1.3747,
      "step": 4429
    },
    {
      "epoch": 0.7257536041939712,
      "grad_norm": 0.1750311702489853,
      "learning_rate": 0.0001,
      "loss": 1.3272,
      "step": 4430
    },
    {
      "epoch": 0.7259174311926605,
      "grad_norm": 0.17448624968528748,
      "learning_rate": 0.0001,
      "loss": 1.2608,
      "step": 4431
    },
    {
      "epoch": 0.7260812581913499,
      "grad_norm": 0.16505135595798492,
      "learning_rate": 0.0001,
      "loss": 1.3633,
      "step": 4432
    },
    {
      "epoch": 0.7262450851900393,
      "grad_norm": 0.17433622479438782,
      "learning_rate": 0.0001,
      "loss": 1.3925,
      "step": 4433
    },
    {
      "epoch": 0.7264089121887287,
      "grad_norm": 0.17769980430603027,
      "learning_rate": 0.0001,
      "loss": 1.277,
      "step": 4434
    },
    {
      "epoch": 0.726572739187418,
      "grad_norm": 0.16517360508441925,
      "learning_rate": 0.0001,
      "loss": 1.3242,
      "step": 4435
    },
    {
      "epoch": 0.7267365661861075,
      "grad_norm": 0.17247077822685242,
      "learning_rate": 0.0001,
      "loss": 1.3215,
      "step": 4436
    },
    {
      "epoch": 0.7269003931847968,
      "grad_norm": 0.18046683073043823,
      "learning_rate": 0.0001,
      "loss": 1.5273,
      "step": 4437
    },
    {
      "epoch": 0.7270642201834863,
      "grad_norm": 0.18210768699645996,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 4438
    },
    {
      "epoch": 0.7272280471821756,
      "grad_norm": 0.17731185257434845,
      "learning_rate": 0.0001,
      "loss": 1.3293,
      "step": 4439
    },
    {
      "epoch": 0.727391874180865,
      "grad_norm": 0.1649036705493927,
      "learning_rate": 0.0001,
      "loss": 1.423,
      "step": 4440
    },
    {
      "epoch": 0.7275557011795544,
      "grad_norm": 0.15927360951900482,
      "learning_rate": 0.0001,
      "loss": 1.3203,
      "step": 4441
    },
    {
      "epoch": 0.7277195281782438,
      "grad_norm": 0.158491313457489,
      "learning_rate": 0.0001,
      "loss": 1.2426,
      "step": 4442
    },
    {
      "epoch": 0.7278833551769331,
      "grad_norm": 0.17703555524349213,
      "learning_rate": 0.0001,
      "loss": 1.4563,
      "step": 4443
    },
    {
      "epoch": 0.7280471821756226,
      "grad_norm": 0.16154339909553528,
      "learning_rate": 0.0001,
      "loss": 1.3409,
      "step": 4444
    },
    {
      "epoch": 0.7282110091743119,
      "grad_norm": 0.16433382034301758,
      "learning_rate": 0.0001,
      "loss": 1.352,
      "step": 4445
    },
    {
      "epoch": 0.7283748361730014,
      "grad_norm": 0.17061969637870789,
      "learning_rate": 0.0001,
      "loss": 1.3671,
      "step": 4446
    },
    {
      "epoch": 0.7285386631716907,
      "grad_norm": 0.176802396774292,
      "learning_rate": 0.0001,
      "loss": 1.3848,
      "step": 4447
    },
    {
      "epoch": 0.72870249017038,
      "grad_norm": 0.1687244325876236,
      "learning_rate": 0.0001,
      "loss": 1.499,
      "step": 4448
    },
    {
      "epoch": 0.7288663171690695,
      "grad_norm": 0.20163214206695557,
      "learning_rate": 0.0001,
      "loss": 1.3977,
      "step": 4449
    },
    {
      "epoch": 0.7290301441677588,
      "grad_norm": 0.17658327519893646,
      "learning_rate": 0.0001,
      "loss": 1.3944,
      "step": 4450
    },
    {
      "epoch": 0.7291939711664482,
      "grad_norm": 0.16570279002189636,
      "learning_rate": 0.0001,
      "loss": 1.41,
      "step": 4451
    },
    {
      "epoch": 0.7293577981651376,
      "grad_norm": 0.17041802406311035,
      "learning_rate": 0.0001,
      "loss": 1.2577,
      "step": 4452
    },
    {
      "epoch": 0.729521625163827,
      "grad_norm": 0.17162474989891052,
      "learning_rate": 0.0001,
      "loss": 1.3629,
      "step": 4453
    },
    {
      "epoch": 0.7296854521625163,
      "grad_norm": 0.17934834957122803,
      "learning_rate": 0.0001,
      "loss": 1.3691,
      "step": 4454
    },
    {
      "epoch": 0.7298492791612058,
      "grad_norm": 0.16813834011554718,
      "learning_rate": 0.0001,
      "loss": 1.2778,
      "step": 4455
    },
    {
      "epoch": 0.7300131061598951,
      "grad_norm": 0.16440553963184357,
      "learning_rate": 0.0001,
      "loss": 1.3031,
      "step": 4456
    },
    {
      "epoch": 0.7301769331585846,
      "grad_norm": 0.16565954685211182,
      "learning_rate": 0.0001,
      "loss": 1.3603,
      "step": 4457
    },
    {
      "epoch": 0.7303407601572739,
      "grad_norm": 0.16787134110927582,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 4458
    },
    {
      "epoch": 0.7305045871559633,
      "grad_norm": 0.167593315243721,
      "learning_rate": 0.0001,
      "loss": 1.385,
      "step": 4459
    },
    {
      "epoch": 0.7306684141546527,
      "grad_norm": 0.16889183223247528,
      "learning_rate": 0.0001,
      "loss": 1.3835,
      "step": 4460
    },
    {
      "epoch": 0.7308322411533421,
      "grad_norm": 0.18262960016727448,
      "learning_rate": 0.0001,
      "loss": 1.3845,
      "step": 4461
    },
    {
      "epoch": 0.7309960681520314,
      "grad_norm": 0.18276306986808777,
      "learning_rate": 0.0001,
      "loss": 1.3891,
      "step": 4462
    },
    {
      "epoch": 0.7311598951507209,
      "grad_norm": 0.17806501686573029,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 4463
    },
    {
      "epoch": 0.7313237221494102,
      "grad_norm": 0.16072405874729156,
      "learning_rate": 0.0001,
      "loss": 1.2516,
      "step": 4464
    },
    {
      "epoch": 0.7314875491480997,
      "grad_norm": 0.17128176987171173,
      "learning_rate": 0.0001,
      "loss": 1.3281,
      "step": 4465
    },
    {
      "epoch": 0.731651376146789,
      "grad_norm": 0.1649724394083023,
      "learning_rate": 0.0001,
      "loss": 1.388,
      "step": 4466
    },
    {
      "epoch": 0.7318152031454783,
      "grad_norm": 0.18052518367767334,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 4467
    },
    {
      "epoch": 0.7319790301441678,
      "grad_norm": 0.17245496809482574,
      "learning_rate": 0.0001,
      "loss": 1.4735,
      "step": 4468
    },
    {
      "epoch": 0.7321428571428571,
      "grad_norm": 0.1750812530517578,
      "learning_rate": 0.0001,
      "loss": 1.4041,
      "step": 4469
    },
    {
      "epoch": 0.7323066841415465,
      "grad_norm": 0.1830882579088211,
      "learning_rate": 0.0001,
      "loss": 1.4199,
      "step": 4470
    },
    {
      "epoch": 0.7324705111402359,
      "grad_norm": 0.19728434085845947,
      "learning_rate": 0.0001,
      "loss": 1.4581,
      "step": 4471
    },
    {
      "epoch": 0.7326343381389253,
      "grad_norm": 0.1774546504020691,
      "learning_rate": 0.0001,
      "loss": 1.4087,
      "step": 4472
    },
    {
      "epoch": 0.7327981651376146,
      "grad_norm": 0.1748822182416916,
      "learning_rate": 0.0001,
      "loss": 1.4538,
      "step": 4473
    },
    {
      "epoch": 0.7329619921363041,
      "grad_norm": 0.17239218950271606,
      "learning_rate": 0.0001,
      "loss": 1.423,
      "step": 4474
    },
    {
      "epoch": 0.7331258191349934,
      "grad_norm": 0.17244772613048553,
      "learning_rate": 0.0001,
      "loss": 1.4139,
      "step": 4475
    },
    {
      "epoch": 0.7332896461336829,
      "grad_norm": 0.1814189851284027,
      "learning_rate": 0.0001,
      "loss": 1.4041,
      "step": 4476
    },
    {
      "epoch": 0.7334534731323722,
      "grad_norm": 0.1604740470647812,
      "learning_rate": 0.0001,
      "loss": 1.2883,
      "step": 4477
    },
    {
      "epoch": 0.7336173001310616,
      "grad_norm": 0.17977887392044067,
      "learning_rate": 0.0001,
      "loss": 1.4685,
      "step": 4478
    },
    {
      "epoch": 0.733781127129751,
      "grad_norm": 0.16503721475601196,
      "learning_rate": 0.0001,
      "loss": 1.3266,
      "step": 4479
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 0.173678457736969,
      "learning_rate": 0.0001,
      "loss": 1.3423,
      "step": 4480
    },
    {
      "epoch": 0.7341087811271297,
      "grad_norm": 0.1697419136762619,
      "learning_rate": 0.0001,
      "loss": 1.3521,
      "step": 4481
    },
    {
      "epoch": 0.7342726081258192,
      "grad_norm": 0.22355693578720093,
      "learning_rate": 0.0001,
      "loss": 1.3722,
      "step": 4482
    },
    {
      "epoch": 0.7344364351245085,
      "grad_norm": 0.1668810397386551,
      "learning_rate": 0.0001,
      "loss": 1.3941,
      "step": 4483
    },
    {
      "epoch": 0.734600262123198,
      "grad_norm": 0.1897396743297577,
      "learning_rate": 0.0001,
      "loss": 1.2777,
      "step": 4484
    },
    {
      "epoch": 0.7347640891218873,
      "grad_norm": 0.17345061898231506,
      "learning_rate": 0.0001,
      "loss": 1.372,
      "step": 4485
    },
    {
      "epoch": 0.7349279161205766,
      "grad_norm": 0.17682553827762604,
      "learning_rate": 0.0001,
      "loss": 1.4044,
      "step": 4486
    },
    {
      "epoch": 0.7350917431192661,
      "grad_norm": 0.1718001812696457,
      "learning_rate": 0.0001,
      "loss": 1.4876,
      "step": 4487
    },
    {
      "epoch": 0.7352555701179554,
      "grad_norm": 0.16768071055412292,
      "learning_rate": 0.0001,
      "loss": 1.2692,
      "step": 4488
    },
    {
      "epoch": 0.7354193971166448,
      "grad_norm": 0.17313914000988007,
      "learning_rate": 0.0001,
      "loss": 1.3456,
      "step": 4489
    },
    {
      "epoch": 0.7355832241153342,
      "grad_norm": 0.17473095655441284,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 4490
    },
    {
      "epoch": 0.7357470511140236,
      "grad_norm": 0.174820676445961,
      "learning_rate": 0.0001,
      "loss": 1.4502,
      "step": 4491
    },
    {
      "epoch": 0.7359108781127129,
      "grad_norm": 0.1649303287267685,
      "learning_rate": 0.0001,
      "loss": 1.4171,
      "step": 4492
    },
    {
      "epoch": 0.7360747051114024,
      "grad_norm": 0.1712983399629593,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 4493
    },
    {
      "epoch": 0.7362385321100917,
      "grad_norm": 0.1719125360250473,
      "learning_rate": 0.0001,
      "loss": 1.3773,
      "step": 4494
    },
    {
      "epoch": 0.7364023591087812,
      "grad_norm": 0.17824265360832214,
      "learning_rate": 0.0001,
      "loss": 1.3954,
      "step": 4495
    },
    {
      "epoch": 0.7365661861074705,
      "grad_norm": 0.1768037974834442,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 4496
    },
    {
      "epoch": 0.7367300131061599,
      "grad_norm": 0.17512641847133636,
      "learning_rate": 0.0001,
      "loss": 1.3845,
      "step": 4497
    },
    {
      "epoch": 0.7368938401048493,
      "grad_norm": 0.17265573143959045,
      "learning_rate": 0.0001,
      "loss": 1.4131,
      "step": 4498
    },
    {
      "epoch": 0.7370576671035387,
      "grad_norm": 0.16891828179359436,
      "learning_rate": 0.0001,
      "loss": 1.3207,
      "step": 4499
    },
    {
      "epoch": 0.737221494102228,
      "grad_norm": 0.17671167850494385,
      "learning_rate": 0.0001,
      "loss": 1.2506,
      "step": 4500
    },
    {
      "epoch": 0.7373853211009175,
      "grad_norm": 0.173119455575943,
      "learning_rate": 0.0001,
      "loss": 1.4571,
      "step": 4501
    },
    {
      "epoch": 0.7375491480996068,
      "grad_norm": 0.17420503497123718,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 4502
    },
    {
      "epoch": 0.7377129750982963,
      "grad_norm": 0.16676120460033417,
      "learning_rate": 0.0001,
      "loss": 1.3855,
      "step": 4503
    },
    {
      "epoch": 0.7378768020969856,
      "grad_norm": 0.16112223267555237,
      "learning_rate": 0.0001,
      "loss": 1.3259,
      "step": 4504
    },
    {
      "epoch": 0.7380406290956749,
      "grad_norm": 0.20170298218727112,
      "learning_rate": 0.0001,
      "loss": 1.3995,
      "step": 4505
    },
    {
      "epoch": 0.7382044560943644,
      "grad_norm": 0.17369957268238068,
      "learning_rate": 0.0001,
      "loss": 1.3247,
      "step": 4506
    },
    {
      "epoch": 0.7383682830930537,
      "grad_norm": 0.1617850661277771,
      "learning_rate": 0.0001,
      "loss": 1.3692,
      "step": 4507
    },
    {
      "epoch": 0.7385321100917431,
      "grad_norm": 0.17002613842487335,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 4508
    },
    {
      "epoch": 0.7386959370904325,
      "grad_norm": 0.16953808069229126,
      "learning_rate": 0.0001,
      "loss": 1.2971,
      "step": 4509
    },
    {
      "epoch": 0.7388597640891219,
      "grad_norm": 0.170583575963974,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 4510
    },
    {
      "epoch": 0.7390235910878112,
      "grad_norm": 0.1741146743297577,
      "learning_rate": 0.0001,
      "loss": 1.3274,
      "step": 4511
    },
    {
      "epoch": 0.7391874180865007,
      "grad_norm": 0.16625410318374634,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 4512
    },
    {
      "epoch": 0.73935124508519,
      "grad_norm": 0.17548058927059174,
      "learning_rate": 0.0001,
      "loss": 1.3434,
      "step": 4513
    },
    {
      "epoch": 0.7395150720838795,
      "grad_norm": 0.1691923439502716,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 4514
    },
    {
      "epoch": 0.7396788990825688,
      "grad_norm": 0.17263559997081757,
      "learning_rate": 0.0001,
      "loss": 1.4156,
      "step": 4515
    },
    {
      "epoch": 0.7398427260812582,
      "grad_norm": 0.17312975227832794,
      "learning_rate": 0.0001,
      "loss": 1.3889,
      "step": 4516
    },
    {
      "epoch": 0.7400065530799476,
      "grad_norm": 0.19523757696151733,
      "learning_rate": 0.0001,
      "loss": 1.3691,
      "step": 4517
    },
    {
      "epoch": 0.740170380078637,
      "grad_norm": 0.17865988612174988,
      "learning_rate": 0.0001,
      "loss": 1.3737,
      "step": 4518
    },
    {
      "epoch": 0.7403342070773263,
      "grad_norm": 0.15798531472682953,
      "learning_rate": 0.0001,
      "loss": 1.2558,
      "step": 4519
    },
    {
      "epoch": 0.7404980340760158,
      "grad_norm": 0.17322781682014465,
      "learning_rate": 0.0001,
      "loss": 1.3727,
      "step": 4520
    },
    {
      "epoch": 0.7406618610747051,
      "grad_norm": 0.17203600704669952,
      "learning_rate": 0.0001,
      "loss": 1.2908,
      "step": 4521
    },
    {
      "epoch": 0.7408256880733946,
      "grad_norm": 0.1752198040485382,
      "learning_rate": 0.0001,
      "loss": 1.3417,
      "step": 4522
    },
    {
      "epoch": 0.7409895150720839,
      "grad_norm": 0.1708817332983017,
      "learning_rate": 0.0001,
      "loss": 1.3512,
      "step": 4523
    },
    {
      "epoch": 0.7411533420707732,
      "grad_norm": 0.17993831634521484,
      "learning_rate": 0.0001,
      "loss": 1.309,
      "step": 4524
    },
    {
      "epoch": 0.7413171690694627,
      "grad_norm": 0.18249964714050293,
      "learning_rate": 0.0001,
      "loss": 1.4167,
      "step": 4525
    },
    {
      "epoch": 0.741480996068152,
      "grad_norm": 0.16165684163570404,
      "learning_rate": 0.0001,
      "loss": 1.311,
      "step": 4526
    },
    {
      "epoch": 0.7416448230668414,
      "grad_norm": 0.1680983603000641,
      "learning_rate": 0.0001,
      "loss": 1.3729,
      "step": 4527
    },
    {
      "epoch": 0.7418086500655308,
      "grad_norm": 0.17835089564323425,
      "learning_rate": 0.0001,
      "loss": 1.3121,
      "step": 4528
    },
    {
      "epoch": 0.7419724770642202,
      "grad_norm": 0.17092838883399963,
      "learning_rate": 0.0001,
      "loss": 1.3579,
      "step": 4529
    },
    {
      "epoch": 0.7421363040629095,
      "grad_norm": 0.1699337363243103,
      "learning_rate": 0.0001,
      "loss": 1.3484,
      "step": 4530
    },
    {
      "epoch": 0.742300131061599,
      "grad_norm": 0.16830603778362274,
      "learning_rate": 0.0001,
      "loss": 1.3759,
      "step": 4531
    },
    {
      "epoch": 0.7424639580602883,
      "grad_norm": 0.16639453172683716,
      "learning_rate": 0.0001,
      "loss": 1.299,
      "step": 4532
    },
    {
      "epoch": 0.7426277850589778,
      "grad_norm": 0.1684856414794922,
      "learning_rate": 0.0001,
      "loss": 1.3858,
      "step": 4533
    },
    {
      "epoch": 0.7427916120576671,
      "grad_norm": 0.16670627892017365,
      "learning_rate": 0.0001,
      "loss": 1.1588,
      "step": 4534
    },
    {
      "epoch": 0.7429554390563565,
      "grad_norm": 0.17405672371387482,
      "learning_rate": 0.0001,
      "loss": 1.3159,
      "step": 4535
    },
    {
      "epoch": 0.7431192660550459,
      "grad_norm": 0.17513605952262878,
      "learning_rate": 0.0001,
      "loss": 1.3041,
      "step": 4536
    },
    {
      "epoch": 0.7432830930537353,
      "grad_norm": 0.17316500842571259,
      "learning_rate": 0.0001,
      "loss": 1.331,
      "step": 4537
    },
    {
      "epoch": 0.7434469200524246,
      "grad_norm": 0.1598939597606659,
      "learning_rate": 0.0001,
      "loss": 1.1827,
      "step": 4538
    },
    {
      "epoch": 0.7436107470511141,
      "grad_norm": 0.16552598774433136,
      "learning_rate": 0.0001,
      "loss": 1.3654,
      "step": 4539
    },
    {
      "epoch": 0.7437745740498034,
      "grad_norm": 0.18211854994297028,
      "learning_rate": 0.0001,
      "loss": 1.3536,
      "step": 4540
    },
    {
      "epoch": 0.7439384010484927,
      "grad_norm": 0.16428428888320923,
      "learning_rate": 0.0001,
      "loss": 1.3378,
      "step": 4541
    },
    {
      "epoch": 0.7441022280471822,
      "grad_norm": 0.17244397103786469,
      "learning_rate": 0.0001,
      "loss": 1.3465,
      "step": 4542
    },
    {
      "epoch": 0.7442660550458715,
      "grad_norm": 0.170182466506958,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 4543
    },
    {
      "epoch": 0.744429882044561,
      "grad_norm": 0.17218910157680511,
      "learning_rate": 0.0001,
      "loss": 1.2936,
      "step": 4544
    },
    {
      "epoch": 0.7445937090432503,
      "grad_norm": 0.17540211975574493,
      "learning_rate": 0.0001,
      "loss": 1.3296,
      "step": 4545
    },
    {
      "epoch": 0.7447575360419397,
      "grad_norm": 0.17295660078525543,
      "learning_rate": 0.0001,
      "loss": 1.3654,
      "step": 4546
    },
    {
      "epoch": 0.7449213630406291,
      "grad_norm": 0.17218296229839325,
      "learning_rate": 0.0001,
      "loss": 1.3729,
      "step": 4547
    },
    {
      "epoch": 0.7450851900393185,
      "grad_norm": 0.177415132522583,
      "learning_rate": 0.0001,
      "loss": 1.4381,
      "step": 4548
    },
    {
      "epoch": 0.7452490170380078,
      "grad_norm": 0.1620887666940689,
      "learning_rate": 0.0001,
      "loss": 1.3606,
      "step": 4549
    },
    {
      "epoch": 0.7454128440366973,
      "grad_norm": 0.17106126248836517,
      "learning_rate": 0.0001,
      "loss": 1.4159,
      "step": 4550
    },
    {
      "epoch": 0.7455766710353866,
      "grad_norm": 0.1693723499774933,
      "learning_rate": 0.0001,
      "loss": 1.4509,
      "step": 4551
    },
    {
      "epoch": 0.745740498034076,
      "grad_norm": 0.1746421456336975,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 4552
    },
    {
      "epoch": 0.7459043250327654,
      "grad_norm": 0.16972675919532776,
      "learning_rate": 0.0001,
      "loss": 1.3841,
      "step": 4553
    },
    {
      "epoch": 0.7460681520314548,
      "grad_norm": 0.17959100008010864,
      "learning_rate": 0.0001,
      "loss": 1.3848,
      "step": 4554
    },
    {
      "epoch": 0.7462319790301442,
      "grad_norm": 0.17013074457645416,
      "learning_rate": 0.0001,
      "loss": 1.4254,
      "step": 4555
    },
    {
      "epoch": 0.7463958060288336,
      "grad_norm": 0.17614351212978363,
      "learning_rate": 0.0001,
      "loss": 1.3986,
      "step": 4556
    },
    {
      "epoch": 0.7465596330275229,
      "grad_norm": 0.17692551016807556,
      "learning_rate": 0.0001,
      "loss": 1.1999,
      "step": 4557
    },
    {
      "epoch": 0.7467234600262124,
      "grad_norm": 0.1794242113828659,
      "learning_rate": 0.0001,
      "loss": 1.4125,
      "step": 4558
    },
    {
      "epoch": 0.7468872870249017,
      "grad_norm": 0.16104134917259216,
      "learning_rate": 0.0001,
      "loss": 1.1594,
      "step": 4559
    },
    {
      "epoch": 0.747051114023591,
      "grad_norm": 0.178700253367424,
      "learning_rate": 0.0001,
      "loss": 1.2141,
      "step": 4560
    },
    {
      "epoch": 0.7472149410222805,
      "grad_norm": 0.17339470982551575,
      "learning_rate": 0.0001,
      "loss": 1.3559,
      "step": 4561
    },
    {
      "epoch": 0.7473787680209698,
      "grad_norm": 0.1657378375530243,
      "learning_rate": 0.0001,
      "loss": 1.2382,
      "step": 4562
    },
    {
      "epoch": 0.7475425950196593,
      "grad_norm": 0.17262624204158783,
      "learning_rate": 0.0001,
      "loss": 1.4092,
      "step": 4563
    },
    {
      "epoch": 0.7477064220183486,
      "grad_norm": 0.17458024621009827,
      "learning_rate": 0.0001,
      "loss": 1.3606,
      "step": 4564
    },
    {
      "epoch": 0.747870249017038,
      "grad_norm": 0.17077352106571198,
      "learning_rate": 0.0001,
      "loss": 1.3034,
      "step": 4565
    },
    {
      "epoch": 0.7480340760157274,
      "grad_norm": 0.1814465969800949,
      "learning_rate": 0.0001,
      "loss": 1.3193,
      "step": 4566
    },
    {
      "epoch": 0.7481979030144168,
      "grad_norm": 0.1748528927564621,
      "learning_rate": 0.0001,
      "loss": 1.3499,
      "step": 4567
    },
    {
      "epoch": 0.7483617300131061,
      "grad_norm": 0.1888834685087204,
      "learning_rate": 0.0001,
      "loss": 1.4679,
      "step": 4568
    },
    {
      "epoch": 0.7485255570117956,
      "grad_norm": 0.17341402173042297,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 4569
    },
    {
      "epoch": 0.7486893840104849,
      "grad_norm": 0.17590221762657166,
      "learning_rate": 0.0001,
      "loss": 1.4069,
      "step": 4570
    },
    {
      "epoch": 0.7488532110091743,
      "grad_norm": 0.1687278300523758,
      "learning_rate": 0.0001,
      "loss": 1.3096,
      "step": 4571
    },
    {
      "epoch": 0.7490170380078637,
      "grad_norm": 0.17642107605934143,
      "learning_rate": 0.0001,
      "loss": 1.3581,
      "step": 4572
    },
    {
      "epoch": 0.7491808650065531,
      "grad_norm": 0.17679116129875183,
      "learning_rate": 0.0001,
      "loss": 1.4214,
      "step": 4573
    },
    {
      "epoch": 0.7493446920052425,
      "grad_norm": 0.1786610335111618,
      "learning_rate": 0.0001,
      "loss": 1.3612,
      "step": 4574
    },
    {
      "epoch": 0.7495085190039319,
      "grad_norm": 0.18640802800655365,
      "learning_rate": 0.0001,
      "loss": 1.3651,
      "step": 4575
    },
    {
      "epoch": 0.7496723460026212,
      "grad_norm": 0.16820506751537323,
      "learning_rate": 0.0001,
      "loss": 1.3489,
      "step": 4576
    },
    {
      "epoch": 0.7498361730013107,
      "grad_norm": 0.1578948199748993,
      "learning_rate": 0.0001,
      "loss": 1.1767,
      "step": 4577
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.16509565711021423,
      "learning_rate": 0.0001,
      "loss": 1.2623,
      "step": 4578
    },
    {
      "epoch": 0.7501638269986893,
      "grad_norm": 0.17232255637645721,
      "learning_rate": 0.0001,
      "loss": 1.3479,
      "step": 4579
    },
    {
      "epoch": 0.7503276539973788,
      "grad_norm": 0.17257779836654663,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 4580
    },
    {
      "epoch": 0.7504914809960681,
      "grad_norm": 0.17106996476650238,
      "learning_rate": 0.0001,
      "loss": 1.3731,
      "step": 4581
    },
    {
      "epoch": 0.7506553079947575,
      "grad_norm": 0.1684521734714508,
      "learning_rate": 0.0001,
      "loss": 1.3439,
      "step": 4582
    },
    {
      "epoch": 0.7508191349934469,
      "grad_norm": 0.16743728518486023,
      "learning_rate": 0.0001,
      "loss": 1.3648,
      "step": 4583
    },
    {
      "epoch": 0.7509829619921363,
      "grad_norm": 0.16442245244979858,
      "learning_rate": 0.0001,
      "loss": 1.3208,
      "step": 4584
    },
    {
      "epoch": 0.7511467889908257,
      "grad_norm": 0.17091837525367737,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 4585
    },
    {
      "epoch": 0.7513106159895151,
      "grad_norm": 0.17069628834724426,
      "learning_rate": 0.0001,
      "loss": 1.374,
      "step": 4586
    },
    {
      "epoch": 0.7514744429882044,
      "grad_norm": 0.18210087716579437,
      "learning_rate": 0.0001,
      "loss": 1.3876,
      "step": 4587
    },
    {
      "epoch": 0.7516382699868939,
      "grad_norm": 0.16469861567020416,
      "learning_rate": 0.0001,
      "loss": 1.3846,
      "step": 4588
    },
    {
      "epoch": 0.7518020969855832,
      "grad_norm": 0.15975788235664368,
      "learning_rate": 0.0001,
      "loss": 1.3283,
      "step": 4589
    },
    {
      "epoch": 0.7519659239842726,
      "grad_norm": 0.15957073867321014,
      "learning_rate": 0.0001,
      "loss": 1.2675,
      "step": 4590
    },
    {
      "epoch": 0.752129750982962,
      "grad_norm": 0.1696610152721405,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 4591
    },
    {
      "epoch": 0.7522935779816514,
      "grad_norm": 0.1784416288137436,
      "learning_rate": 0.0001,
      "loss": 1.4175,
      "step": 4592
    },
    {
      "epoch": 0.7524574049803407,
      "grad_norm": 0.17069284617900848,
      "learning_rate": 0.0001,
      "loss": 1.4403,
      "step": 4593
    },
    {
      "epoch": 0.7526212319790302,
      "grad_norm": 0.16937841475009918,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 4594
    },
    {
      "epoch": 0.7527850589777195,
      "grad_norm": 0.16893677413463593,
      "learning_rate": 0.0001,
      "loss": 1.3222,
      "step": 4595
    },
    {
      "epoch": 0.752948885976409,
      "grad_norm": 0.1714247465133667,
      "learning_rate": 0.0001,
      "loss": 1.3604,
      "step": 4596
    },
    {
      "epoch": 0.7531127129750983,
      "grad_norm": 0.17501959204673767,
      "learning_rate": 0.0001,
      "loss": 1.419,
      "step": 4597
    },
    {
      "epoch": 0.7532765399737876,
      "grad_norm": 0.18099793791770935,
      "learning_rate": 0.0001,
      "loss": 1.3528,
      "step": 4598
    },
    {
      "epoch": 0.7534403669724771,
      "grad_norm": 0.17621523141860962,
      "learning_rate": 0.0001,
      "loss": 1.3754,
      "step": 4599
    },
    {
      "epoch": 0.7536041939711664,
      "grad_norm": 0.17122967541217804,
      "learning_rate": 0.0001,
      "loss": 1.33,
      "step": 4600
    },
    {
      "epoch": 0.7537680209698558,
      "grad_norm": 0.17432557046413422,
      "learning_rate": 0.0001,
      "loss": 1.3654,
      "step": 4601
    },
    {
      "epoch": 0.7539318479685452,
      "grad_norm": 0.17846037447452545,
      "learning_rate": 0.0001,
      "loss": 1.3548,
      "step": 4602
    },
    {
      "epoch": 0.7540956749672346,
      "grad_norm": 0.16744425892829895,
      "learning_rate": 0.0001,
      "loss": 1.3812,
      "step": 4603
    },
    {
      "epoch": 0.754259501965924,
      "grad_norm": 0.17605869472026825,
      "learning_rate": 0.0001,
      "loss": 1.296,
      "step": 4604
    },
    {
      "epoch": 0.7544233289646134,
      "grad_norm": 0.17518101632595062,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 4605
    },
    {
      "epoch": 0.7545871559633027,
      "grad_norm": 0.17423945665359497,
      "learning_rate": 0.0001,
      "loss": 1.3383,
      "step": 4606
    },
    {
      "epoch": 0.7547509829619922,
      "grad_norm": 0.1798768937587738,
      "learning_rate": 0.0001,
      "loss": 1.284,
      "step": 4607
    },
    {
      "epoch": 0.7549148099606815,
      "grad_norm": 0.2077450305223465,
      "learning_rate": 0.0001,
      "loss": 1.3854,
      "step": 4608
    },
    {
      "epoch": 0.7550786369593709,
      "grad_norm": 0.17272299528121948,
      "learning_rate": 0.0001,
      "loss": 1.2677,
      "step": 4609
    },
    {
      "epoch": 0.7552424639580603,
      "grad_norm": 0.16390115022659302,
      "learning_rate": 0.0001,
      "loss": 1.2339,
      "step": 4610
    },
    {
      "epoch": 0.7554062909567497,
      "grad_norm": 0.18877704441547394,
      "learning_rate": 0.0001,
      "loss": 1.4658,
      "step": 4611
    },
    {
      "epoch": 0.755570117955439,
      "grad_norm": 0.16815564036369324,
      "learning_rate": 0.0001,
      "loss": 1.3931,
      "step": 4612
    },
    {
      "epoch": 0.7557339449541285,
      "grad_norm": 0.1722032129764557,
      "learning_rate": 0.0001,
      "loss": 1.3402,
      "step": 4613
    },
    {
      "epoch": 0.7558977719528178,
      "grad_norm": 0.16905440390110016,
      "learning_rate": 0.0001,
      "loss": 1.1953,
      "step": 4614
    },
    {
      "epoch": 0.7560615989515073,
      "grad_norm": 0.16947612166404724,
      "learning_rate": 0.0001,
      "loss": 1.431,
      "step": 4615
    },
    {
      "epoch": 0.7562254259501966,
      "grad_norm": 0.17613200843334198,
      "learning_rate": 0.0001,
      "loss": 1.4258,
      "step": 4616
    },
    {
      "epoch": 0.7563892529488859,
      "grad_norm": 0.18882319331169128,
      "learning_rate": 0.0001,
      "loss": 1.435,
      "step": 4617
    },
    {
      "epoch": 0.7565530799475754,
      "grad_norm": 0.1754630208015442,
      "learning_rate": 0.0001,
      "loss": 1.3504,
      "step": 4618
    },
    {
      "epoch": 0.7567169069462647,
      "grad_norm": 0.1718892604112625,
      "learning_rate": 0.0001,
      "loss": 1.3386,
      "step": 4619
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 0.1748095154762268,
      "learning_rate": 0.0001,
      "loss": 1.3803,
      "step": 4620
    },
    {
      "epoch": 0.7570445609436435,
      "grad_norm": 0.16767315566539764,
      "learning_rate": 0.0001,
      "loss": 1.2962,
      "step": 4621
    },
    {
      "epoch": 0.7572083879423329,
      "grad_norm": 0.16319109499454498,
      "learning_rate": 0.0001,
      "loss": 1.2821,
      "step": 4622
    },
    {
      "epoch": 0.7573722149410222,
      "grad_norm": 0.18372532725334167,
      "learning_rate": 0.0001,
      "loss": 1.3722,
      "step": 4623
    },
    {
      "epoch": 0.7575360419397117,
      "grad_norm": 0.17298179864883423,
      "learning_rate": 0.0001,
      "loss": 1.3683,
      "step": 4624
    },
    {
      "epoch": 0.757699868938401,
      "grad_norm": 0.17572076618671417,
      "learning_rate": 0.0001,
      "loss": 1.4674,
      "step": 4625
    },
    {
      "epoch": 0.7578636959370905,
      "grad_norm": 0.16562461853027344,
      "learning_rate": 0.0001,
      "loss": 1.3835,
      "step": 4626
    },
    {
      "epoch": 0.7580275229357798,
      "grad_norm": 0.16611744463443756,
      "learning_rate": 0.0001,
      "loss": 1.322,
      "step": 4627
    },
    {
      "epoch": 0.7581913499344692,
      "grad_norm": 0.17662778496742249,
      "learning_rate": 0.0001,
      "loss": 1.3476,
      "step": 4628
    },
    {
      "epoch": 0.7583551769331586,
      "grad_norm": 0.17163708806037903,
      "learning_rate": 0.0001,
      "loss": 1.4603,
      "step": 4629
    },
    {
      "epoch": 0.758519003931848,
      "grad_norm": 0.16558293998241425,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 4630
    },
    {
      "epoch": 0.7586828309305373,
      "grad_norm": 0.16949675977230072,
      "learning_rate": 0.0001,
      "loss": 1.3659,
      "step": 4631
    },
    {
      "epoch": 0.7588466579292268,
      "grad_norm": 0.17223426699638367,
      "learning_rate": 0.0001,
      "loss": 1.3636,
      "step": 4632
    },
    {
      "epoch": 0.7590104849279161,
      "grad_norm": 0.17935581505298615,
      "learning_rate": 0.0001,
      "loss": 1.359,
      "step": 4633
    },
    {
      "epoch": 0.7591743119266054,
      "grad_norm": 0.17237117886543274,
      "learning_rate": 0.0001,
      "loss": 1.303,
      "step": 4634
    },
    {
      "epoch": 0.7593381389252949,
      "grad_norm": 0.1827143281698227,
      "learning_rate": 0.0001,
      "loss": 1.313,
      "step": 4635
    },
    {
      "epoch": 0.7595019659239842,
      "grad_norm": 0.18358969688415527,
      "learning_rate": 0.0001,
      "loss": 1.4079,
      "step": 4636
    },
    {
      "epoch": 0.7596657929226737,
      "grad_norm": 0.17454276978969574,
      "learning_rate": 0.0001,
      "loss": 1.4011,
      "step": 4637
    },
    {
      "epoch": 0.759829619921363,
      "grad_norm": 0.1772751808166504,
      "learning_rate": 0.0001,
      "loss": 1.3289,
      "step": 4638
    },
    {
      "epoch": 0.7599934469200524,
      "grad_norm": 0.1746198832988739,
      "learning_rate": 0.0001,
      "loss": 1.3289,
      "step": 4639
    },
    {
      "epoch": 0.7601572739187418,
      "grad_norm": 0.17331965267658234,
      "learning_rate": 0.0001,
      "loss": 1.3206,
      "step": 4640
    },
    {
      "epoch": 0.7603211009174312,
      "grad_norm": 0.18279151618480682,
      "learning_rate": 0.0001,
      "loss": 1.4713,
      "step": 4641
    },
    {
      "epoch": 0.7604849279161205,
      "grad_norm": 0.1732569932937622,
      "learning_rate": 0.0001,
      "loss": 1.3796,
      "step": 4642
    },
    {
      "epoch": 0.76064875491481,
      "grad_norm": 0.17745958268642426,
      "learning_rate": 0.0001,
      "loss": 1.4702,
      "step": 4643
    },
    {
      "epoch": 0.7608125819134993,
      "grad_norm": 0.17831313610076904,
      "learning_rate": 0.0001,
      "loss": 1.33,
      "step": 4644
    },
    {
      "epoch": 0.7609764089121888,
      "grad_norm": 0.1686457395553589,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 4645
    },
    {
      "epoch": 0.7611402359108781,
      "grad_norm": 0.16944076120853424,
      "learning_rate": 0.0001,
      "loss": 1.2804,
      "step": 4646
    },
    {
      "epoch": 0.7613040629095675,
      "grad_norm": 0.16661466658115387,
      "learning_rate": 0.0001,
      "loss": 1.37,
      "step": 4647
    },
    {
      "epoch": 0.7614678899082569,
      "grad_norm": 0.16302818059921265,
      "learning_rate": 0.0001,
      "loss": 1.2623,
      "step": 4648
    },
    {
      "epoch": 0.7616317169069463,
      "grad_norm": 0.17939333617687225,
      "learning_rate": 0.0001,
      "loss": 1.281,
      "step": 4649
    },
    {
      "epoch": 0.7617955439056356,
      "grad_norm": 0.1635691374540329,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 4650
    },
    {
      "epoch": 0.7619593709043251,
      "grad_norm": 0.1740325540304184,
      "learning_rate": 0.0001,
      "loss": 1.3164,
      "step": 4651
    },
    {
      "epoch": 0.7621231979030144,
      "grad_norm": 0.17369712889194489,
      "learning_rate": 0.0001,
      "loss": 1.3742,
      "step": 4652
    },
    {
      "epoch": 0.7622870249017037,
      "grad_norm": 0.17375916242599487,
      "learning_rate": 0.0001,
      "loss": 1.3329,
      "step": 4653
    },
    {
      "epoch": 0.7624508519003932,
      "grad_norm": 0.1694774627685547,
      "learning_rate": 0.0001,
      "loss": 1.2704,
      "step": 4654
    },
    {
      "epoch": 0.7626146788990825,
      "grad_norm": 0.17988425493240356,
      "learning_rate": 0.0001,
      "loss": 1.3739,
      "step": 4655
    },
    {
      "epoch": 0.762778505897772,
      "grad_norm": 0.16767707467079163,
      "learning_rate": 0.0001,
      "loss": 1.3177,
      "step": 4656
    },
    {
      "epoch": 0.7629423328964613,
      "grad_norm": 0.17317421734333038,
      "learning_rate": 0.0001,
      "loss": 1.4004,
      "step": 4657
    },
    {
      "epoch": 0.7631061598951507,
      "grad_norm": 0.18006964027881622,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 4658
    },
    {
      "epoch": 0.7632699868938401,
      "grad_norm": 0.17228150367736816,
      "learning_rate": 0.0001,
      "loss": 1.3864,
      "step": 4659
    },
    {
      "epoch": 0.7634338138925295,
      "grad_norm": 0.17707112431526184,
      "learning_rate": 0.0001,
      "loss": 1.4465,
      "step": 4660
    },
    {
      "epoch": 0.7635976408912188,
      "grad_norm": 0.15880069136619568,
      "learning_rate": 0.0001,
      "loss": 1.2886,
      "step": 4661
    },
    {
      "epoch": 0.7637614678899083,
      "grad_norm": 0.16974800825119019,
      "learning_rate": 0.0001,
      "loss": 1.3257,
      "step": 4662
    },
    {
      "epoch": 0.7639252948885976,
      "grad_norm": 0.17272254824638367,
      "learning_rate": 0.0001,
      "loss": 1.4069,
      "step": 4663
    },
    {
      "epoch": 0.7640891218872871,
      "grad_norm": 0.17689014971256256,
      "learning_rate": 0.0001,
      "loss": 1.2122,
      "step": 4664
    },
    {
      "epoch": 0.7642529488859764,
      "grad_norm": 0.18002170324325562,
      "learning_rate": 0.0001,
      "loss": 1.4022,
      "step": 4665
    },
    {
      "epoch": 0.7644167758846658,
      "grad_norm": 0.17458495497703552,
      "learning_rate": 0.0001,
      "loss": 1.3796,
      "step": 4666
    },
    {
      "epoch": 0.7645806028833552,
      "grad_norm": 0.19345900416374207,
      "learning_rate": 0.0001,
      "loss": 1.3605,
      "step": 4667
    },
    {
      "epoch": 0.7647444298820446,
      "grad_norm": 0.17106018960475922,
      "learning_rate": 0.0001,
      "loss": 1.3273,
      "step": 4668
    },
    {
      "epoch": 0.7649082568807339,
      "grad_norm": 0.16871194541454315,
      "learning_rate": 0.0001,
      "loss": 1.3911,
      "step": 4669
    },
    {
      "epoch": 0.7650720838794234,
      "grad_norm": 0.17621001601219177,
      "learning_rate": 0.0001,
      "loss": 1.4174,
      "step": 4670
    },
    {
      "epoch": 0.7652359108781127,
      "grad_norm": 0.16872651875019073,
      "learning_rate": 0.0001,
      "loss": 1.2796,
      "step": 4671
    },
    {
      "epoch": 0.765399737876802,
      "grad_norm": 0.1767764836549759,
      "learning_rate": 0.0001,
      "loss": 1.3776,
      "step": 4672
    },
    {
      "epoch": 0.7655635648754915,
      "grad_norm": 0.17324422299861908,
      "learning_rate": 0.0001,
      "loss": 1.3685,
      "step": 4673
    },
    {
      "epoch": 0.7657273918741808,
      "grad_norm": 0.17163074016571045,
      "learning_rate": 0.0001,
      "loss": 1.37,
      "step": 4674
    },
    {
      "epoch": 0.7658912188728703,
      "grad_norm": 0.18391771614551544,
      "learning_rate": 0.0001,
      "loss": 1.3581,
      "step": 4675
    },
    {
      "epoch": 0.7660550458715596,
      "grad_norm": 0.19873212277889252,
      "learning_rate": 0.0001,
      "loss": 1.4571,
      "step": 4676
    },
    {
      "epoch": 0.766218872870249,
      "grad_norm": 0.1689997911453247,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 4677
    },
    {
      "epoch": 0.7663826998689384,
      "grad_norm": 0.17770257592201233,
      "learning_rate": 0.0001,
      "loss": 1.3843,
      "step": 4678
    },
    {
      "epoch": 0.7665465268676278,
      "grad_norm": 0.1745116412639618,
      "learning_rate": 0.0001,
      "loss": 1.3601,
      "step": 4679
    },
    {
      "epoch": 0.7667103538663171,
      "grad_norm": 0.1646289974451065,
      "learning_rate": 0.0001,
      "loss": 1.3759,
      "step": 4680
    },
    {
      "epoch": 0.7668741808650066,
      "grad_norm": 0.1808895319700241,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 4681
    },
    {
      "epoch": 0.7670380078636959,
      "grad_norm": 0.18229927122592926,
      "learning_rate": 0.0001,
      "loss": 1.4026,
      "step": 4682
    },
    {
      "epoch": 0.7672018348623854,
      "grad_norm": 0.1743389070034027,
      "learning_rate": 0.0001,
      "loss": 1.4163,
      "step": 4683
    },
    {
      "epoch": 0.7673656618610747,
      "grad_norm": 0.16822372376918793,
      "learning_rate": 0.0001,
      "loss": 1.3786,
      "step": 4684
    },
    {
      "epoch": 0.7675294888597641,
      "grad_norm": 0.16796989738941193,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 4685
    },
    {
      "epoch": 0.7676933158584535,
      "grad_norm": 0.17172668874263763,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 4686
    },
    {
      "epoch": 0.7678571428571429,
      "grad_norm": 0.17432977259159088,
      "learning_rate": 0.0001,
      "loss": 1.3485,
      "step": 4687
    },
    {
      "epoch": 0.7680209698558322,
      "grad_norm": 0.166711688041687,
      "learning_rate": 0.0001,
      "loss": 1.3453,
      "step": 4688
    },
    {
      "epoch": 0.7681847968545217,
      "grad_norm": 0.18035967648029327,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 4689
    },
    {
      "epoch": 0.768348623853211,
      "grad_norm": 0.18411961197853088,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 4690
    },
    {
      "epoch": 0.7685124508519003,
      "grad_norm": 0.16988526284694672,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 4691
    },
    {
      "epoch": 0.7686762778505898,
      "grad_norm": 0.18179896473884583,
      "learning_rate": 0.0001,
      "loss": 1.3527,
      "step": 4692
    },
    {
      "epoch": 0.7688401048492791,
      "grad_norm": 0.16363300383090973,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 4693
    },
    {
      "epoch": 0.7690039318479686,
      "grad_norm": 0.16404902935028076,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 4694
    },
    {
      "epoch": 0.7691677588466579,
      "grad_norm": 0.18874123692512512,
      "learning_rate": 0.0001,
      "loss": 1.3767,
      "step": 4695
    },
    {
      "epoch": 0.7693315858453473,
      "grad_norm": 0.17818941175937653,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 4696
    },
    {
      "epoch": 0.7694954128440367,
      "grad_norm": 0.16507025063037872,
      "learning_rate": 0.0001,
      "loss": 1.2606,
      "step": 4697
    },
    {
      "epoch": 0.7696592398427261,
      "grad_norm": 0.18304623663425446,
      "learning_rate": 0.0001,
      "loss": 1.4438,
      "step": 4698
    },
    {
      "epoch": 0.7698230668414154,
      "grad_norm": 0.16944019496440887,
      "learning_rate": 0.0001,
      "loss": 1.2942,
      "step": 4699
    },
    {
      "epoch": 0.7699868938401049,
      "grad_norm": 0.16216625273227692,
      "learning_rate": 0.0001,
      "loss": 1.2665,
      "step": 4700
    },
    {
      "epoch": 0.7701507208387942,
      "grad_norm": 0.18878161907196045,
      "learning_rate": 0.0001,
      "loss": 1.3902,
      "step": 4701
    },
    {
      "epoch": 0.7703145478374837,
      "grad_norm": 0.17536182701587677,
      "learning_rate": 0.0001,
      "loss": 1.3733,
      "step": 4702
    },
    {
      "epoch": 0.770478374836173,
      "grad_norm": 0.16662415862083435,
      "learning_rate": 0.0001,
      "loss": 1.3918,
      "step": 4703
    },
    {
      "epoch": 0.7706422018348624,
      "grad_norm": 0.17186866700649261,
      "learning_rate": 0.0001,
      "loss": 1.2298,
      "step": 4704
    },
    {
      "epoch": 0.7708060288335518,
      "grad_norm": 0.17966806888580322,
      "learning_rate": 0.0001,
      "loss": 1.4442,
      "step": 4705
    },
    {
      "epoch": 0.7709698558322412,
      "grad_norm": 0.16700203716754913,
      "learning_rate": 0.0001,
      "loss": 1.3497,
      "step": 4706
    },
    {
      "epoch": 0.7711336828309305,
      "grad_norm": 0.1739051640033722,
      "learning_rate": 0.0001,
      "loss": 1.428,
      "step": 4707
    },
    {
      "epoch": 0.77129750982962,
      "grad_norm": 0.17421172559261322,
      "learning_rate": 0.0001,
      "loss": 1.3515,
      "step": 4708
    },
    {
      "epoch": 0.7714613368283093,
      "grad_norm": 0.17472556233406067,
      "learning_rate": 0.0001,
      "loss": 1.3768,
      "step": 4709
    },
    {
      "epoch": 0.7716251638269986,
      "grad_norm": 0.17488868534564972,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 4710
    },
    {
      "epoch": 0.7717889908256881,
      "grad_norm": 0.17985011637210846,
      "learning_rate": 0.0001,
      "loss": 1.3022,
      "step": 4711
    },
    {
      "epoch": 0.7719528178243774,
      "grad_norm": 0.18107278645038605,
      "learning_rate": 0.0001,
      "loss": 1.3521,
      "step": 4712
    },
    {
      "epoch": 0.7721166448230669,
      "grad_norm": 0.17074814438819885,
      "learning_rate": 0.0001,
      "loss": 1.3805,
      "step": 4713
    },
    {
      "epoch": 0.7722804718217562,
      "grad_norm": 0.19346500933170319,
      "learning_rate": 0.0001,
      "loss": 1.421,
      "step": 4714
    },
    {
      "epoch": 0.7724442988204456,
      "grad_norm": 0.17188675701618195,
      "learning_rate": 0.0001,
      "loss": 1.3209,
      "step": 4715
    },
    {
      "epoch": 0.772608125819135,
      "grad_norm": 0.16812275350093842,
      "learning_rate": 0.0001,
      "loss": 1.4098,
      "step": 4716
    },
    {
      "epoch": 0.7727719528178244,
      "grad_norm": 0.168903186917305,
      "learning_rate": 0.0001,
      "loss": 1.3293,
      "step": 4717
    },
    {
      "epoch": 0.7729357798165137,
      "grad_norm": 0.18209514021873474,
      "learning_rate": 0.0001,
      "loss": 1.3448,
      "step": 4718
    },
    {
      "epoch": 0.7730996068152032,
      "grad_norm": 0.17888322472572327,
      "learning_rate": 0.0001,
      "loss": 1.3596,
      "step": 4719
    },
    {
      "epoch": 0.7732634338138925,
      "grad_norm": 0.172575905919075,
      "learning_rate": 0.0001,
      "loss": 1.3058,
      "step": 4720
    },
    {
      "epoch": 0.773427260812582,
      "grad_norm": 0.18656539916992188,
      "learning_rate": 0.0001,
      "loss": 1.3205,
      "step": 4721
    },
    {
      "epoch": 0.7735910878112713,
      "grad_norm": 0.17950211465358734,
      "learning_rate": 0.0001,
      "loss": 1.327,
      "step": 4722
    },
    {
      "epoch": 0.7737549148099607,
      "grad_norm": 0.16698184609413147,
      "learning_rate": 0.0001,
      "loss": 1.327,
      "step": 4723
    },
    {
      "epoch": 0.7739187418086501,
      "grad_norm": 0.17316477000713348,
      "learning_rate": 0.0001,
      "loss": 1.3293,
      "step": 4724
    },
    {
      "epoch": 0.7740825688073395,
      "grad_norm": 0.17474308609962463,
      "learning_rate": 0.0001,
      "loss": 1.3306,
      "step": 4725
    },
    {
      "epoch": 0.7742463958060288,
      "grad_norm": 0.16869492828845978,
      "learning_rate": 0.0001,
      "loss": 1.317,
      "step": 4726
    },
    {
      "epoch": 0.7744102228047182,
      "grad_norm": 0.17539486289024353,
      "learning_rate": 0.0001,
      "loss": 1.4067,
      "step": 4727
    },
    {
      "epoch": 0.7745740498034076,
      "grad_norm": 0.16774654388427734,
      "learning_rate": 0.0001,
      "loss": 1.3367,
      "step": 4728
    },
    {
      "epoch": 0.7747378768020969,
      "grad_norm": 0.1939576119184494,
      "learning_rate": 0.0001,
      "loss": 1.4038,
      "step": 4729
    },
    {
      "epoch": 0.7749017038007864,
      "grad_norm": 0.17474165558815002,
      "learning_rate": 0.0001,
      "loss": 1.3945,
      "step": 4730
    },
    {
      "epoch": 0.7750655307994757,
      "grad_norm": 0.17168042063713074,
      "learning_rate": 0.0001,
      "loss": 1.3762,
      "step": 4731
    },
    {
      "epoch": 0.7752293577981652,
      "grad_norm": 0.17143891751766205,
      "learning_rate": 0.0001,
      "loss": 1.2843,
      "step": 4732
    },
    {
      "epoch": 0.7753931847968545,
      "grad_norm": 0.18785282969474792,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 4733
    },
    {
      "epoch": 0.7755570117955439,
      "grad_norm": 0.1673959344625473,
      "learning_rate": 0.0001,
      "loss": 1.3599,
      "step": 4734
    },
    {
      "epoch": 0.7757208387942333,
      "grad_norm": 0.18157874047756195,
      "learning_rate": 0.0001,
      "loss": 1.3332,
      "step": 4735
    },
    {
      "epoch": 0.7758846657929227,
      "grad_norm": 0.18492090702056885,
      "learning_rate": 0.0001,
      "loss": 1.4914,
      "step": 4736
    },
    {
      "epoch": 0.776048492791612,
      "grad_norm": 0.18624916672706604,
      "learning_rate": 0.0001,
      "loss": 1.3569,
      "step": 4737
    },
    {
      "epoch": 0.7762123197903015,
      "grad_norm": 0.21009081602096558,
      "learning_rate": 0.0001,
      "loss": 1.2632,
      "step": 4738
    },
    {
      "epoch": 0.7763761467889908,
      "grad_norm": 0.17300982773303986,
      "learning_rate": 0.0001,
      "loss": 1.2787,
      "step": 4739
    },
    {
      "epoch": 0.7765399737876802,
      "grad_norm": 0.17233355343341827,
      "learning_rate": 0.0001,
      "loss": 1.3414,
      "step": 4740
    },
    {
      "epoch": 0.7767038007863696,
      "grad_norm": 0.18032491207122803,
      "learning_rate": 0.0001,
      "loss": 1.3491,
      "step": 4741
    },
    {
      "epoch": 0.776867627785059,
      "grad_norm": 0.1801784485578537,
      "learning_rate": 0.0001,
      "loss": 1.4072,
      "step": 4742
    },
    {
      "epoch": 0.7770314547837484,
      "grad_norm": 0.16821463406085968,
      "learning_rate": 0.0001,
      "loss": 1.245,
      "step": 4743
    },
    {
      "epoch": 0.7771952817824378,
      "grad_norm": 0.1787845939397812,
      "learning_rate": 0.0001,
      "loss": 1.274,
      "step": 4744
    },
    {
      "epoch": 0.7773591087811271,
      "grad_norm": 0.18495331704616547,
      "learning_rate": 0.0001,
      "loss": 1.3513,
      "step": 4745
    },
    {
      "epoch": 0.7775229357798165,
      "grad_norm": 0.17415142059326172,
      "learning_rate": 0.0001,
      "loss": 1.4036,
      "step": 4746
    },
    {
      "epoch": 0.7776867627785059,
      "grad_norm": 0.18936291337013245,
      "learning_rate": 0.0001,
      "loss": 1.3889,
      "step": 4747
    },
    {
      "epoch": 0.7778505897771952,
      "grad_norm": 0.15811479091644287,
      "learning_rate": 0.0001,
      "loss": 1.2426,
      "step": 4748
    },
    {
      "epoch": 0.7780144167758847,
      "grad_norm": 0.173636794090271,
      "learning_rate": 0.0001,
      "loss": 1.4268,
      "step": 4749
    },
    {
      "epoch": 0.778178243774574,
      "grad_norm": 0.17409482598304749,
      "learning_rate": 0.0001,
      "loss": 1.4227,
      "step": 4750
    },
    {
      "epoch": 0.7783420707732635,
      "grad_norm": 0.1775992512702942,
      "learning_rate": 0.0001,
      "loss": 1.395,
      "step": 4751
    },
    {
      "epoch": 0.7785058977719528,
      "grad_norm": 0.17167504131793976,
      "learning_rate": 0.0001,
      "loss": 1.3066,
      "step": 4752
    },
    {
      "epoch": 0.7786697247706422,
      "grad_norm": 0.17126527428627014,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 4753
    },
    {
      "epoch": 0.7788335517693316,
      "grad_norm": 0.16880494356155396,
      "learning_rate": 0.0001,
      "loss": 1.2764,
      "step": 4754
    },
    {
      "epoch": 0.778997378768021,
      "grad_norm": 0.17541024088859558,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 4755
    },
    {
      "epoch": 0.7791612057667103,
      "grad_norm": 0.17022307217121124,
      "learning_rate": 0.0001,
      "loss": 1.3097,
      "step": 4756
    },
    {
      "epoch": 0.7793250327653998,
      "grad_norm": 0.17549167573451996,
      "learning_rate": 0.0001,
      "loss": 1.3938,
      "step": 4757
    },
    {
      "epoch": 0.7794888597640891,
      "grad_norm": 0.18678288161754608,
      "learning_rate": 0.0001,
      "loss": 1.3981,
      "step": 4758
    },
    {
      "epoch": 0.7796526867627785,
      "grad_norm": 0.1660197377204895,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 4759
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 0.17085197567939758,
      "learning_rate": 0.0001,
      "loss": 1.2899,
      "step": 4760
    },
    {
      "epoch": 0.7799803407601573,
      "grad_norm": 0.18154369294643402,
      "learning_rate": 0.0001,
      "loss": 1.3363,
      "step": 4761
    },
    {
      "epoch": 0.7801441677588467,
      "grad_norm": 0.17409422993659973,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 4762
    },
    {
      "epoch": 0.7803079947575361,
      "grad_norm": 0.17070910334587097,
      "learning_rate": 0.0001,
      "loss": 1.3819,
      "step": 4763
    },
    {
      "epoch": 0.7804718217562254,
      "grad_norm": 0.17701679468154907,
      "learning_rate": 0.0001,
      "loss": 1.475,
      "step": 4764
    },
    {
      "epoch": 0.7806356487549148,
      "grad_norm": 0.17085996270179749,
      "learning_rate": 0.0001,
      "loss": 1.4976,
      "step": 4765
    },
    {
      "epoch": 0.7807994757536042,
      "grad_norm": 0.17314212024211884,
      "learning_rate": 0.0001,
      "loss": 1.4059,
      "step": 4766
    },
    {
      "epoch": 0.7809633027522935,
      "grad_norm": 0.17745880782604218,
      "learning_rate": 0.0001,
      "loss": 1.369,
      "step": 4767
    },
    {
      "epoch": 0.781127129750983,
      "grad_norm": 0.16604623198509216,
      "learning_rate": 0.0001,
      "loss": 1.2952,
      "step": 4768
    },
    {
      "epoch": 0.7812909567496723,
      "grad_norm": 0.1694408804178238,
      "learning_rate": 0.0001,
      "loss": 1.326,
      "step": 4769
    },
    {
      "epoch": 0.7814547837483617,
      "grad_norm": 0.1859327107667923,
      "learning_rate": 0.0001,
      "loss": 1.3463,
      "step": 4770
    },
    {
      "epoch": 0.7816186107470511,
      "grad_norm": 0.18529583513736725,
      "learning_rate": 0.0001,
      "loss": 1.3319,
      "step": 4771
    },
    {
      "epoch": 0.7817824377457405,
      "grad_norm": 0.1723201423883438,
      "learning_rate": 0.0001,
      "loss": 1.3025,
      "step": 4772
    },
    {
      "epoch": 0.7819462647444299,
      "grad_norm": 0.1796126365661621,
      "learning_rate": 0.0001,
      "loss": 1.2989,
      "step": 4773
    },
    {
      "epoch": 0.7821100917431193,
      "grad_norm": 0.17206594347953796,
      "learning_rate": 0.0001,
      "loss": 1.361,
      "step": 4774
    },
    {
      "epoch": 0.7822739187418086,
      "grad_norm": 0.1766183078289032,
      "learning_rate": 0.0001,
      "loss": 1.4331,
      "step": 4775
    },
    {
      "epoch": 0.7824377457404981,
      "grad_norm": 0.17563536763191223,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 4776
    },
    {
      "epoch": 0.7826015727391874,
      "grad_norm": 0.16515570878982544,
      "learning_rate": 0.0001,
      "loss": 1.2637,
      "step": 4777
    },
    {
      "epoch": 0.7827653997378768,
      "grad_norm": 0.17458342015743256,
      "learning_rate": 0.0001,
      "loss": 1.4375,
      "step": 4778
    },
    {
      "epoch": 0.7829292267365662,
      "grad_norm": 0.17039142549037933,
      "learning_rate": 0.0001,
      "loss": 1.3975,
      "step": 4779
    },
    {
      "epoch": 0.7830930537352556,
      "grad_norm": 0.16792498528957367,
      "learning_rate": 0.0001,
      "loss": 1.2758,
      "step": 4780
    },
    {
      "epoch": 0.783256880733945,
      "grad_norm": 0.17026342451572418,
      "learning_rate": 0.0001,
      "loss": 1.3454,
      "step": 4781
    },
    {
      "epoch": 0.7834207077326344,
      "grad_norm": 0.16476711630821228,
      "learning_rate": 0.0001,
      "loss": 1.256,
      "step": 4782
    },
    {
      "epoch": 0.7835845347313237,
      "grad_norm": 0.23502495884895325,
      "learning_rate": 0.0001,
      "loss": 1.3101,
      "step": 4783
    },
    {
      "epoch": 0.783748361730013,
      "grad_norm": 0.17491395771503448,
      "learning_rate": 0.0001,
      "loss": 1.3282,
      "step": 4784
    },
    {
      "epoch": 0.7839121887287025,
      "grad_norm": 0.1718364655971527,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 4785
    },
    {
      "epoch": 0.7840760157273918,
      "grad_norm": 0.16752943396568298,
      "learning_rate": 0.0001,
      "loss": 1.2857,
      "step": 4786
    },
    {
      "epoch": 0.7842398427260813,
      "grad_norm": 0.18420390784740448,
      "learning_rate": 0.0001,
      "loss": 1.4102,
      "step": 4787
    },
    {
      "epoch": 0.7844036697247706,
      "grad_norm": 0.21202488243579865,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 4788
    },
    {
      "epoch": 0.78456749672346,
      "grad_norm": 0.18660052120685577,
      "learning_rate": 0.0001,
      "loss": 1.4038,
      "step": 4789
    },
    {
      "epoch": 0.7847313237221494,
      "grad_norm": 0.17220887541770935,
      "learning_rate": 0.0001,
      "loss": 1.3189,
      "step": 4790
    },
    {
      "epoch": 0.7848951507208388,
      "grad_norm": 0.16536784172058105,
      "learning_rate": 0.0001,
      "loss": 1.3809,
      "step": 4791
    },
    {
      "epoch": 0.7850589777195282,
      "grad_norm": 0.1783563643693924,
      "learning_rate": 0.0001,
      "loss": 1.3168,
      "step": 4792
    },
    {
      "epoch": 0.7852228047182176,
      "grad_norm": 0.17076222598552704,
      "learning_rate": 0.0001,
      "loss": 1.3362,
      "step": 4793
    },
    {
      "epoch": 0.7853866317169069,
      "grad_norm": 0.17756131291389465,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 4794
    },
    {
      "epoch": 0.7855504587155964,
      "grad_norm": 0.17496101558208466,
      "learning_rate": 0.0001,
      "loss": 1.3619,
      "step": 4795
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 0.18000078201293945,
      "learning_rate": 0.0001,
      "loss": 1.3954,
      "step": 4796
    },
    {
      "epoch": 0.7858781127129751,
      "grad_norm": 0.180303156375885,
      "learning_rate": 0.0001,
      "loss": 1.3461,
      "step": 4797
    },
    {
      "epoch": 0.7860419397116645,
      "grad_norm": 0.1790945827960968,
      "learning_rate": 0.0001,
      "loss": 1.3176,
      "step": 4798
    },
    {
      "epoch": 0.7862057667103539,
      "grad_norm": 0.16458457708358765,
      "learning_rate": 0.0001,
      "loss": 1.2735,
      "step": 4799
    },
    {
      "epoch": 0.7863695937090432,
      "grad_norm": 0.17354458570480347,
      "learning_rate": 0.0001,
      "loss": 1.2049,
      "step": 4800
    },
    {
      "epoch": 0.7865334207077327,
      "grad_norm": 0.17188820242881775,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 4801
    },
    {
      "epoch": 0.786697247706422,
      "grad_norm": 0.17958861589431763,
      "learning_rate": 0.0001,
      "loss": 1.3754,
      "step": 4802
    },
    {
      "epoch": 0.7868610747051114,
      "grad_norm": 0.16757327318191528,
      "learning_rate": 0.0001,
      "loss": 1.2675,
      "step": 4803
    },
    {
      "epoch": 0.7870249017038008,
      "grad_norm": 0.18346403539180756,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 4804
    },
    {
      "epoch": 0.7871887287024901,
      "grad_norm": 0.17945042252540588,
      "learning_rate": 0.0001,
      "loss": 1.3577,
      "step": 4805
    },
    {
      "epoch": 0.7873525557011796,
      "grad_norm": 0.17006942629814148,
      "learning_rate": 0.0001,
      "loss": 1.2674,
      "step": 4806
    },
    {
      "epoch": 0.7875163826998689,
      "grad_norm": 0.16752417385578156,
      "learning_rate": 0.0001,
      "loss": 1.3261,
      "step": 4807
    },
    {
      "epoch": 0.7876802096985583,
      "grad_norm": 0.18581709265708923,
      "learning_rate": 0.0001,
      "loss": 1.4012,
      "step": 4808
    },
    {
      "epoch": 0.7878440366972477,
      "grad_norm": 0.1736760139465332,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 4809
    },
    {
      "epoch": 0.7880078636959371,
      "grad_norm": 0.1669355034828186,
      "learning_rate": 0.0001,
      "loss": 1.3641,
      "step": 4810
    },
    {
      "epoch": 0.7881716906946264,
      "grad_norm": 0.17700818181037903,
      "learning_rate": 0.0001,
      "loss": 1.4471,
      "step": 4811
    },
    {
      "epoch": 0.7883355176933159,
      "grad_norm": 0.18157173693180084,
      "learning_rate": 0.0001,
      "loss": 1.3876,
      "step": 4812
    },
    {
      "epoch": 0.7884993446920052,
      "grad_norm": 0.1859486848115921,
      "learning_rate": 0.0001,
      "loss": 1.3934,
      "step": 4813
    },
    {
      "epoch": 0.7886631716906947,
      "grad_norm": 0.1833890974521637,
      "learning_rate": 0.0001,
      "loss": 1.3438,
      "step": 4814
    },
    {
      "epoch": 0.788826998689384,
      "grad_norm": 0.16570231318473816,
      "learning_rate": 0.0001,
      "loss": 1.2634,
      "step": 4815
    },
    {
      "epoch": 0.7889908256880734,
      "grad_norm": 0.1785404086112976,
      "learning_rate": 0.0001,
      "loss": 1.4167,
      "step": 4816
    },
    {
      "epoch": 0.7891546526867628,
      "grad_norm": 0.18139414489269257,
      "learning_rate": 0.0001,
      "loss": 1.3449,
      "step": 4817
    },
    {
      "epoch": 0.7893184796854522,
      "grad_norm": 0.17651605606079102,
      "learning_rate": 0.0001,
      "loss": 1.3897,
      "step": 4818
    },
    {
      "epoch": 0.7894823066841415,
      "grad_norm": 0.1762923300266266,
      "learning_rate": 0.0001,
      "loss": 1.3463,
      "step": 4819
    },
    {
      "epoch": 0.7896461336828309,
      "grad_norm": 0.1790963113307953,
      "learning_rate": 0.0001,
      "loss": 1.4483,
      "step": 4820
    },
    {
      "epoch": 0.7898099606815203,
      "grad_norm": 0.18711449205875397,
      "learning_rate": 0.0001,
      "loss": 1.4137,
      "step": 4821
    },
    {
      "epoch": 0.7899737876802096,
      "grad_norm": 0.18176965415477753,
      "learning_rate": 0.0001,
      "loss": 1.4463,
      "step": 4822
    },
    {
      "epoch": 0.7901376146788991,
      "grad_norm": 0.1763066053390503,
      "learning_rate": 0.0001,
      "loss": 1.3476,
      "step": 4823
    },
    {
      "epoch": 0.7903014416775884,
      "grad_norm": 0.18448151648044586,
      "learning_rate": 0.0001,
      "loss": 1.3168,
      "step": 4824
    },
    {
      "epoch": 0.7904652686762779,
      "grad_norm": 0.1671684980392456,
      "learning_rate": 0.0001,
      "loss": 1.3356,
      "step": 4825
    },
    {
      "epoch": 0.7906290956749672,
      "grad_norm": 0.1894107609987259,
      "learning_rate": 0.0001,
      "loss": 1.4695,
      "step": 4826
    },
    {
      "epoch": 0.7907929226736566,
      "grad_norm": 0.18858326971530914,
      "learning_rate": 0.0001,
      "loss": 1.33,
      "step": 4827
    },
    {
      "epoch": 0.790956749672346,
      "grad_norm": 0.1758764535188675,
      "learning_rate": 0.0001,
      "loss": 1.3285,
      "step": 4828
    },
    {
      "epoch": 0.7911205766710354,
      "grad_norm": 0.16890691220760345,
      "learning_rate": 0.0001,
      "loss": 1.1645,
      "step": 4829
    },
    {
      "epoch": 0.7912844036697247,
      "grad_norm": 0.18104274570941925,
      "learning_rate": 0.0001,
      "loss": 1.3238,
      "step": 4830
    },
    {
      "epoch": 0.7914482306684142,
      "grad_norm": 0.1656343936920166,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 4831
    },
    {
      "epoch": 0.7916120576671035,
      "grad_norm": 0.1717992126941681,
      "learning_rate": 0.0001,
      "loss": 1.3468,
      "step": 4832
    },
    {
      "epoch": 0.791775884665793,
      "grad_norm": 0.1725059598684311,
      "learning_rate": 0.0001,
      "loss": 1.3226,
      "step": 4833
    },
    {
      "epoch": 0.7919397116644823,
      "grad_norm": 0.1595740020275116,
      "learning_rate": 0.0001,
      "loss": 1.3063,
      "step": 4834
    },
    {
      "epoch": 0.7921035386631717,
      "grad_norm": 0.16826258599758148,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 4835
    },
    {
      "epoch": 0.7922673656618611,
      "grad_norm": 0.16860255599021912,
      "learning_rate": 0.0001,
      "loss": 1.363,
      "step": 4836
    },
    {
      "epoch": 0.7924311926605505,
      "grad_norm": 0.1700536459684372,
      "learning_rate": 0.0001,
      "loss": 1.404,
      "step": 4837
    },
    {
      "epoch": 0.7925950196592398,
      "grad_norm": 0.1672907918691635,
      "learning_rate": 0.0001,
      "loss": 1.3485,
      "step": 4838
    },
    {
      "epoch": 0.7927588466579292,
      "grad_norm": 0.1727263629436493,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 4839
    },
    {
      "epoch": 0.7929226736566186,
      "grad_norm": 0.173997163772583,
      "learning_rate": 0.0001,
      "loss": 1.205,
      "step": 4840
    },
    {
      "epoch": 0.793086500655308,
      "grad_norm": 0.16577838361263275,
      "learning_rate": 0.0001,
      "loss": 1.3016,
      "step": 4841
    },
    {
      "epoch": 0.7932503276539974,
      "grad_norm": 0.17114059627056122,
      "learning_rate": 0.0001,
      "loss": 1.3604,
      "step": 4842
    },
    {
      "epoch": 0.7934141546526867,
      "grad_norm": 0.18054157495498657,
      "learning_rate": 0.0001,
      "loss": 1.3824,
      "step": 4843
    },
    {
      "epoch": 0.7935779816513762,
      "grad_norm": 0.18001917004585266,
      "learning_rate": 0.0001,
      "loss": 1.358,
      "step": 4844
    },
    {
      "epoch": 0.7937418086500655,
      "grad_norm": 0.17045319080352783,
      "learning_rate": 0.0001,
      "loss": 1.3113,
      "step": 4845
    },
    {
      "epoch": 0.7939056356487549,
      "grad_norm": 0.16814693808555603,
      "learning_rate": 0.0001,
      "loss": 1.3669,
      "step": 4846
    },
    {
      "epoch": 0.7940694626474443,
      "grad_norm": 0.17314325273036957,
      "learning_rate": 0.0001,
      "loss": 1.3757,
      "step": 4847
    },
    {
      "epoch": 0.7942332896461337,
      "grad_norm": 0.16509471833705902,
      "learning_rate": 0.0001,
      "loss": 1.3379,
      "step": 4848
    },
    {
      "epoch": 0.794397116644823,
      "grad_norm": 0.16110031306743622,
      "learning_rate": 0.0001,
      "loss": 1.2464,
      "step": 4849
    },
    {
      "epoch": 0.7945609436435125,
      "grad_norm": 0.1750853806734085,
      "learning_rate": 0.0001,
      "loss": 1.3774,
      "step": 4850
    },
    {
      "epoch": 0.7947247706422018,
      "grad_norm": 0.17199696600437164,
      "learning_rate": 0.0001,
      "loss": 1.2728,
      "step": 4851
    },
    {
      "epoch": 0.7948885976408913,
      "grad_norm": 0.18581511080265045,
      "learning_rate": 0.0001,
      "loss": 1.2924,
      "step": 4852
    },
    {
      "epoch": 0.7950524246395806,
      "grad_norm": 0.19106508791446686,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 4853
    },
    {
      "epoch": 0.79521625163827,
      "grad_norm": 0.17648066580295563,
      "learning_rate": 0.0001,
      "loss": 1.3985,
      "step": 4854
    },
    {
      "epoch": 0.7953800786369594,
      "grad_norm": 0.18178033828735352,
      "learning_rate": 0.0001,
      "loss": 1.3858,
      "step": 4855
    },
    {
      "epoch": 0.7955439056356488,
      "grad_norm": 0.1724558025598526,
      "learning_rate": 0.0001,
      "loss": 1.3472,
      "step": 4856
    },
    {
      "epoch": 0.7957077326343381,
      "grad_norm": 0.1709398627281189,
      "learning_rate": 0.0001,
      "loss": 1.4323,
      "step": 4857
    },
    {
      "epoch": 0.7958715596330275,
      "grad_norm": 0.17083822190761566,
      "learning_rate": 0.0001,
      "loss": 1.4404,
      "step": 4858
    },
    {
      "epoch": 0.7960353866317169,
      "grad_norm": 0.17098280787467957,
      "learning_rate": 0.0001,
      "loss": 1.4485,
      "step": 4859
    },
    {
      "epoch": 0.7961992136304062,
      "grad_norm": 0.17101651430130005,
      "learning_rate": 0.0001,
      "loss": 1.3796,
      "step": 4860
    },
    {
      "epoch": 0.7963630406290957,
      "grad_norm": 0.1761382520198822,
      "learning_rate": 0.0001,
      "loss": 1.2671,
      "step": 4861
    },
    {
      "epoch": 0.796526867627785,
      "grad_norm": 0.17990611493587494,
      "learning_rate": 0.0001,
      "loss": 1.3713,
      "step": 4862
    },
    {
      "epoch": 0.7966906946264745,
      "grad_norm": 0.1716904640197754,
      "learning_rate": 0.0001,
      "loss": 1.2906,
      "step": 4863
    },
    {
      "epoch": 0.7968545216251638,
      "grad_norm": 0.16771361231803894,
      "learning_rate": 0.0001,
      "loss": 1.4224,
      "step": 4864
    },
    {
      "epoch": 0.7970183486238532,
      "grad_norm": 0.17457976937294006,
      "learning_rate": 0.0001,
      "loss": 1.3273,
      "step": 4865
    },
    {
      "epoch": 0.7971821756225426,
      "grad_norm": 0.18543384969234467,
      "learning_rate": 0.0001,
      "loss": 1.3962,
      "step": 4866
    },
    {
      "epoch": 0.797346002621232,
      "grad_norm": 0.17329217493534088,
      "learning_rate": 0.0001,
      "loss": 1.3993,
      "step": 4867
    },
    {
      "epoch": 0.7975098296199213,
      "grad_norm": 0.16664788126945496,
      "learning_rate": 0.0001,
      "loss": 1.2743,
      "step": 4868
    },
    {
      "epoch": 0.7976736566186108,
      "grad_norm": 0.17702347040176392,
      "learning_rate": 0.0001,
      "loss": 1.2975,
      "step": 4869
    },
    {
      "epoch": 0.7978374836173001,
      "grad_norm": 0.17701056599617004,
      "learning_rate": 0.0001,
      "loss": 1.3687,
      "step": 4870
    },
    {
      "epoch": 0.7980013106159896,
      "grad_norm": 0.1759897917509079,
      "learning_rate": 0.0001,
      "loss": 1.3375,
      "step": 4871
    },
    {
      "epoch": 0.7981651376146789,
      "grad_norm": 0.16859331727027893,
      "learning_rate": 0.0001,
      "loss": 1.3536,
      "step": 4872
    },
    {
      "epoch": 0.7983289646133683,
      "grad_norm": 0.16938696801662445,
      "learning_rate": 0.0001,
      "loss": 1.2935,
      "step": 4873
    },
    {
      "epoch": 0.7984927916120577,
      "grad_norm": 0.17389410734176636,
      "learning_rate": 0.0001,
      "loss": 1.3473,
      "step": 4874
    },
    {
      "epoch": 0.7986566186107471,
      "grad_norm": 0.1809106469154358,
      "learning_rate": 0.0001,
      "loss": 1.4113,
      "step": 4875
    },
    {
      "epoch": 0.7988204456094364,
      "grad_norm": 0.17796121537685394,
      "learning_rate": 0.0001,
      "loss": 1.4306,
      "step": 4876
    },
    {
      "epoch": 0.7989842726081258,
      "grad_norm": 0.1770482212305069,
      "learning_rate": 0.0001,
      "loss": 1.3713,
      "step": 4877
    },
    {
      "epoch": 0.7991480996068152,
      "grad_norm": 0.17415480315685272,
      "learning_rate": 0.0001,
      "loss": 1.3626,
      "step": 4878
    },
    {
      "epoch": 0.7993119266055045,
      "grad_norm": 0.17479057610034943,
      "learning_rate": 0.0001,
      "loss": 1.3237,
      "step": 4879
    },
    {
      "epoch": 0.799475753604194,
      "grad_norm": 0.1850292682647705,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 4880
    },
    {
      "epoch": 0.7996395806028833,
      "grad_norm": 0.17391563951969147,
      "learning_rate": 0.0001,
      "loss": 1.3392,
      "step": 4881
    },
    {
      "epoch": 0.7998034076015728,
      "grad_norm": 0.17821049690246582,
      "learning_rate": 0.0001,
      "loss": 1.3887,
      "step": 4882
    },
    {
      "epoch": 0.7999672346002621,
      "grad_norm": 0.17821647226810455,
      "learning_rate": 0.0001,
      "loss": 1.3523,
      "step": 4883
    },
    {
      "epoch": 0.8001310615989515,
      "grad_norm": 0.17625659704208374,
      "learning_rate": 0.0001,
      "loss": 1.3334,
      "step": 4884
    },
    {
      "epoch": 0.8002948885976409,
      "grad_norm": 0.16764114797115326,
      "learning_rate": 0.0001,
      "loss": 1.3768,
      "step": 4885
    },
    {
      "epoch": 0.8004587155963303,
      "grad_norm": 0.18034672737121582,
      "learning_rate": 0.0001,
      "loss": 1.2759,
      "step": 4886
    },
    {
      "epoch": 0.8006225425950196,
      "grad_norm": 0.1705685257911682,
      "learning_rate": 0.0001,
      "loss": 1.3058,
      "step": 4887
    },
    {
      "epoch": 0.8007863695937091,
      "grad_norm": 0.17191475629806519,
      "learning_rate": 0.0001,
      "loss": 1.3463,
      "step": 4888
    },
    {
      "epoch": 0.8009501965923984,
      "grad_norm": 0.1900031566619873,
      "learning_rate": 0.0001,
      "loss": 1.4265,
      "step": 4889
    },
    {
      "epoch": 0.8011140235910879,
      "grad_norm": 0.17415285110473633,
      "learning_rate": 0.0001,
      "loss": 1.378,
      "step": 4890
    },
    {
      "epoch": 0.8012778505897772,
      "grad_norm": 0.17753766477108002,
      "learning_rate": 0.0001,
      "loss": 1.2399,
      "step": 4891
    },
    {
      "epoch": 0.8014416775884666,
      "grad_norm": 0.17681509256362915,
      "learning_rate": 0.0001,
      "loss": 1.3609,
      "step": 4892
    },
    {
      "epoch": 0.801605504587156,
      "grad_norm": 0.1825808584690094,
      "learning_rate": 0.0001,
      "loss": 1.2802,
      "step": 4893
    },
    {
      "epoch": 0.8017693315858454,
      "grad_norm": 0.17192426323890686,
      "learning_rate": 0.0001,
      "loss": 1.3836,
      "step": 4894
    },
    {
      "epoch": 0.8019331585845347,
      "grad_norm": 0.1767069548368454,
      "learning_rate": 0.0001,
      "loss": 1.3949,
      "step": 4895
    },
    {
      "epoch": 0.8020969855832241,
      "grad_norm": 0.16324369609355927,
      "learning_rate": 0.0001,
      "loss": 1.3851,
      "step": 4896
    },
    {
      "epoch": 0.8022608125819135,
      "grad_norm": 0.17241418361663818,
      "learning_rate": 0.0001,
      "loss": 1.3427,
      "step": 4897
    },
    {
      "epoch": 0.8024246395806028,
      "grad_norm": 0.17447468638420105,
      "learning_rate": 0.0001,
      "loss": 1.3641,
      "step": 4898
    },
    {
      "epoch": 0.8025884665792923,
      "grad_norm": 0.17620202898979187,
      "learning_rate": 0.0001,
      "loss": 1.3126,
      "step": 4899
    },
    {
      "epoch": 0.8027522935779816,
      "grad_norm": 0.17667293548583984,
      "learning_rate": 0.0001,
      "loss": 1.3374,
      "step": 4900
    },
    {
      "epoch": 0.802916120576671,
      "grad_norm": 0.18006594479084015,
      "learning_rate": 0.0001,
      "loss": 1.4024,
      "step": 4901
    },
    {
      "epoch": 0.8030799475753604,
      "grad_norm": 0.1847354918718338,
      "learning_rate": 0.0001,
      "loss": 1.3032,
      "step": 4902
    },
    {
      "epoch": 0.8032437745740498,
      "grad_norm": 0.1724705547094345,
      "learning_rate": 0.0001,
      "loss": 1.2946,
      "step": 4903
    },
    {
      "epoch": 0.8034076015727392,
      "grad_norm": 0.17989040911197662,
      "learning_rate": 0.0001,
      "loss": 1.4116,
      "step": 4904
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 0.17666670680046082,
      "learning_rate": 0.0001,
      "loss": 1.3954,
      "step": 4905
    },
    {
      "epoch": 0.8037352555701179,
      "grad_norm": 0.17601285874843597,
      "learning_rate": 0.0001,
      "loss": 1.3321,
      "step": 4906
    },
    {
      "epoch": 0.8038990825688074,
      "grad_norm": 0.17367888987064362,
      "learning_rate": 0.0001,
      "loss": 1.3784,
      "step": 4907
    },
    {
      "epoch": 0.8040629095674967,
      "grad_norm": 0.16963951289653778,
      "learning_rate": 0.0001,
      "loss": 1.3089,
      "step": 4908
    },
    {
      "epoch": 0.8042267365661862,
      "grad_norm": 0.21585233509540558,
      "learning_rate": 0.0001,
      "loss": 1.4024,
      "step": 4909
    },
    {
      "epoch": 0.8043905635648755,
      "grad_norm": 0.16553205251693726,
      "learning_rate": 0.0001,
      "loss": 1.1917,
      "step": 4910
    },
    {
      "epoch": 0.8045543905635649,
      "grad_norm": 0.1693912148475647,
      "learning_rate": 0.0001,
      "loss": 1.2277,
      "step": 4911
    },
    {
      "epoch": 0.8047182175622543,
      "grad_norm": 0.17484794557094574,
      "learning_rate": 0.0001,
      "loss": 1.3646,
      "step": 4912
    },
    {
      "epoch": 0.8048820445609436,
      "grad_norm": 0.16625414788722992,
      "learning_rate": 0.0001,
      "loss": 1.3583,
      "step": 4913
    },
    {
      "epoch": 0.805045871559633,
      "grad_norm": 0.17727087438106537,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 4914
    },
    {
      "epoch": 0.8052096985583224,
      "grad_norm": 0.1813659816980362,
      "learning_rate": 0.0001,
      "loss": 1.3055,
      "step": 4915
    },
    {
      "epoch": 0.8053735255570118,
      "grad_norm": 0.1734539419412613,
      "learning_rate": 0.0001,
      "loss": 1.3548,
      "step": 4916
    },
    {
      "epoch": 0.8055373525557011,
      "grad_norm": 0.18211424350738525,
      "learning_rate": 0.0001,
      "loss": 1.3459,
      "step": 4917
    },
    {
      "epoch": 0.8057011795543906,
      "grad_norm": 0.17786502838134766,
      "learning_rate": 0.0001,
      "loss": 1.2957,
      "step": 4918
    },
    {
      "epoch": 0.8058650065530799,
      "grad_norm": 0.17373378574848175,
      "learning_rate": 0.0001,
      "loss": 1.4236,
      "step": 4919
    },
    {
      "epoch": 0.8060288335517694,
      "grad_norm": 0.17845012247562408,
      "learning_rate": 0.0001,
      "loss": 1.3949,
      "step": 4920
    },
    {
      "epoch": 0.8061926605504587,
      "grad_norm": 0.1835368424654007,
      "learning_rate": 0.0001,
      "loss": 1.4314,
      "step": 4921
    },
    {
      "epoch": 0.8063564875491481,
      "grad_norm": 0.1768873631954193,
      "learning_rate": 0.0001,
      "loss": 1.3692,
      "step": 4922
    },
    {
      "epoch": 0.8065203145478375,
      "grad_norm": 0.17463894188404083,
      "learning_rate": 0.0001,
      "loss": 1.3435,
      "step": 4923
    },
    {
      "epoch": 0.8066841415465269,
      "grad_norm": 0.17144183814525604,
      "learning_rate": 0.0001,
      "loss": 1.3588,
      "step": 4924
    },
    {
      "epoch": 0.8068479685452162,
      "grad_norm": 0.18525461852550507,
      "learning_rate": 0.0001,
      "loss": 1.4614,
      "step": 4925
    },
    {
      "epoch": 0.8070117955439057,
      "grad_norm": 0.17609146237373352,
      "learning_rate": 0.0001,
      "loss": 1.4364,
      "step": 4926
    },
    {
      "epoch": 0.807175622542595,
      "grad_norm": 0.19646506011486053,
      "learning_rate": 0.0001,
      "loss": 1.4258,
      "step": 4927
    },
    {
      "epoch": 0.8073394495412844,
      "grad_norm": 0.1777554750442505,
      "learning_rate": 0.0001,
      "loss": 1.3279,
      "step": 4928
    },
    {
      "epoch": 0.8075032765399738,
      "grad_norm": 0.17508076131343842,
      "learning_rate": 0.0001,
      "loss": 1.3525,
      "step": 4929
    },
    {
      "epoch": 0.8076671035386632,
      "grad_norm": 0.1801176369190216,
      "learning_rate": 0.0001,
      "loss": 1.384,
      "step": 4930
    },
    {
      "epoch": 0.8078309305373526,
      "grad_norm": 0.18486642837524414,
      "learning_rate": 0.0001,
      "loss": 1.2917,
      "step": 4931
    },
    {
      "epoch": 0.8079947575360419,
      "grad_norm": 0.17145518958568573,
      "learning_rate": 0.0001,
      "loss": 1.388,
      "step": 4932
    },
    {
      "epoch": 0.8081585845347313,
      "grad_norm": 0.17338161170482635,
      "learning_rate": 0.0001,
      "loss": 1.3603,
      "step": 4933
    },
    {
      "epoch": 0.8083224115334207,
      "grad_norm": 0.17350658774375916,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 4934
    },
    {
      "epoch": 0.8084862385321101,
      "grad_norm": 0.1777477264404297,
      "learning_rate": 0.0001,
      "loss": 1.3453,
      "step": 4935
    },
    {
      "epoch": 0.8086500655307994,
      "grad_norm": 0.1766194999217987,
      "learning_rate": 0.0001,
      "loss": 1.4053,
      "step": 4936
    },
    {
      "epoch": 0.8088138925294889,
      "grad_norm": 0.1891629695892334,
      "learning_rate": 0.0001,
      "loss": 1.3852,
      "step": 4937
    },
    {
      "epoch": 0.8089777195281782,
      "grad_norm": 0.17272910475730896,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 4938
    },
    {
      "epoch": 0.8091415465268676,
      "grad_norm": 0.17881807684898376,
      "learning_rate": 0.0001,
      "loss": 1.3373,
      "step": 4939
    },
    {
      "epoch": 0.809305373525557,
      "grad_norm": 0.17694862186908722,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 4940
    },
    {
      "epoch": 0.8094692005242464,
      "grad_norm": 0.18032588064670563,
      "learning_rate": 0.0001,
      "loss": 1.3396,
      "step": 4941
    },
    {
      "epoch": 0.8096330275229358,
      "grad_norm": 0.17039597034454346,
      "learning_rate": 0.0001,
      "loss": 1.3065,
      "step": 4942
    },
    {
      "epoch": 0.8097968545216252,
      "grad_norm": 0.18488623201847076,
      "learning_rate": 0.0001,
      "loss": 1.3409,
      "step": 4943
    },
    {
      "epoch": 0.8099606815203145,
      "grad_norm": 0.1705605685710907,
      "learning_rate": 0.0001,
      "loss": 1.4215,
      "step": 4944
    },
    {
      "epoch": 0.810124508519004,
      "grad_norm": 0.17377397418022156,
      "learning_rate": 0.0001,
      "loss": 1.2623,
      "step": 4945
    },
    {
      "epoch": 0.8102883355176933,
      "grad_norm": 0.180707648396492,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 4946
    },
    {
      "epoch": 0.8104521625163827,
      "grad_norm": 0.16844408214092255,
      "learning_rate": 0.0001,
      "loss": 1.3406,
      "step": 4947
    },
    {
      "epoch": 0.8106159895150721,
      "grad_norm": 0.17441105842590332,
      "learning_rate": 0.0001,
      "loss": 1.3569,
      "step": 4948
    },
    {
      "epoch": 0.8107798165137615,
      "grad_norm": 0.17138029634952545,
      "learning_rate": 0.0001,
      "loss": 1.3658,
      "step": 4949
    },
    {
      "epoch": 0.8109436435124509,
      "grad_norm": 0.18747495114803314,
      "learning_rate": 0.0001,
      "loss": 1.4227,
      "step": 4950
    },
    {
      "epoch": 0.8111074705111402,
      "grad_norm": 0.17756398022174835,
      "learning_rate": 0.0001,
      "loss": 1.3605,
      "step": 4951
    },
    {
      "epoch": 0.8112712975098296,
      "grad_norm": 0.17808200418949127,
      "learning_rate": 0.0001,
      "loss": 1.3336,
      "step": 4952
    },
    {
      "epoch": 0.811435124508519,
      "grad_norm": 0.17718453705310822,
      "learning_rate": 0.0001,
      "loss": 1.3418,
      "step": 4953
    },
    {
      "epoch": 0.8115989515072084,
      "grad_norm": 0.1784604787826538,
      "learning_rate": 0.0001,
      "loss": 1.3504,
      "step": 4954
    },
    {
      "epoch": 0.8117627785058977,
      "grad_norm": 0.18029475212097168,
      "learning_rate": 0.0001,
      "loss": 1.3684,
      "step": 4955
    },
    {
      "epoch": 0.8119266055045872,
      "grad_norm": 0.18526463210582733,
      "learning_rate": 0.0001,
      "loss": 1.3692,
      "step": 4956
    },
    {
      "epoch": 0.8120904325032765,
      "grad_norm": 0.16927622258663177,
      "learning_rate": 0.0001,
      "loss": 1.272,
      "step": 4957
    },
    {
      "epoch": 0.812254259501966,
      "grad_norm": 0.1710970550775528,
      "learning_rate": 0.0001,
      "loss": 1.2787,
      "step": 4958
    },
    {
      "epoch": 0.8124180865006553,
      "grad_norm": 0.17148584127426147,
      "learning_rate": 0.0001,
      "loss": 1.357,
      "step": 4959
    },
    {
      "epoch": 0.8125819134993447,
      "grad_norm": 0.17131638526916504,
      "learning_rate": 0.0001,
      "loss": 1.3984,
      "step": 4960
    },
    {
      "epoch": 0.812745740498034,
      "grad_norm": 0.17786738276481628,
      "learning_rate": 0.0001,
      "loss": 1.1991,
      "step": 4961
    },
    {
      "epoch": 0.8129095674967235,
      "grad_norm": 0.18751287460327148,
      "learning_rate": 0.0001,
      "loss": 1.5372,
      "step": 4962
    },
    {
      "epoch": 0.8130733944954128,
      "grad_norm": 0.17832112312316895,
      "learning_rate": 0.0001,
      "loss": 1.431,
      "step": 4963
    },
    {
      "epoch": 0.8132372214941023,
      "grad_norm": 0.17120219767093658,
      "learning_rate": 0.0001,
      "loss": 1.3281,
      "step": 4964
    },
    {
      "epoch": 0.8134010484927916,
      "grad_norm": 0.17620599269866943,
      "learning_rate": 0.0001,
      "loss": 1.2978,
      "step": 4965
    },
    {
      "epoch": 0.813564875491481,
      "grad_norm": 0.1813267320394516,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 4966
    },
    {
      "epoch": 0.8137287024901704,
      "grad_norm": 0.171678826212883,
      "learning_rate": 0.0001,
      "loss": 1.2477,
      "step": 4967
    },
    {
      "epoch": 0.8138925294888598,
      "grad_norm": 0.17667651176452637,
      "learning_rate": 0.0001,
      "loss": 1.4651,
      "step": 4968
    },
    {
      "epoch": 0.8140563564875491,
      "grad_norm": 0.17171768844127655,
      "learning_rate": 0.0001,
      "loss": 1.3506,
      "step": 4969
    },
    {
      "epoch": 0.8142201834862385,
      "grad_norm": 0.17753875255584717,
      "learning_rate": 0.0001,
      "loss": 1.4012,
      "step": 4970
    },
    {
      "epoch": 0.8143840104849279,
      "grad_norm": 0.20180834829807281,
      "learning_rate": 0.0001,
      "loss": 1.3294,
      "step": 4971
    },
    {
      "epoch": 0.8145478374836173,
      "grad_norm": 0.18532520532608032,
      "learning_rate": 0.0001,
      "loss": 1.2878,
      "step": 4972
    },
    {
      "epoch": 0.8147116644823067,
      "grad_norm": 0.1747933030128479,
      "learning_rate": 0.0001,
      "loss": 1.2854,
      "step": 4973
    },
    {
      "epoch": 0.814875491480996,
      "grad_norm": 0.16978128254413605,
      "learning_rate": 0.0001,
      "loss": 1.3266,
      "step": 4974
    },
    {
      "epoch": 0.8150393184796855,
      "grad_norm": 0.17468197643756866,
      "learning_rate": 0.0001,
      "loss": 1.2654,
      "step": 4975
    },
    {
      "epoch": 0.8152031454783748,
      "grad_norm": 0.17748500406742096,
      "learning_rate": 0.0001,
      "loss": 1.3816,
      "step": 4976
    },
    {
      "epoch": 0.8153669724770642,
      "grad_norm": 0.17890281975269318,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 4977
    },
    {
      "epoch": 0.8155307994757536,
      "grad_norm": 0.17592841386795044,
      "learning_rate": 0.0001,
      "loss": 1.346,
      "step": 4978
    },
    {
      "epoch": 0.815694626474443,
      "grad_norm": 0.17188353836536407,
      "learning_rate": 0.0001,
      "loss": 1.4145,
      "step": 4979
    },
    {
      "epoch": 0.8158584534731324,
      "grad_norm": 0.18267647922039032,
      "learning_rate": 0.0001,
      "loss": 1.3581,
      "step": 4980
    },
    {
      "epoch": 0.8160222804718218,
      "grad_norm": 0.17526879906654358,
      "learning_rate": 0.0001,
      "loss": 1.3786,
      "step": 4981
    },
    {
      "epoch": 0.8161861074705111,
      "grad_norm": 0.1779385358095169,
      "learning_rate": 0.0001,
      "loss": 1.4616,
      "step": 4982
    },
    {
      "epoch": 0.8163499344692006,
      "grad_norm": 0.17143630981445312,
      "learning_rate": 0.0001,
      "loss": 1.3347,
      "step": 4983
    },
    {
      "epoch": 0.8165137614678899,
      "grad_norm": 0.18249867856502533,
      "learning_rate": 0.0001,
      "loss": 1.349,
      "step": 4984
    },
    {
      "epoch": 0.8166775884665793,
      "grad_norm": 0.17263375222682953,
      "learning_rate": 0.0001,
      "loss": 1.3927,
      "step": 4985
    },
    {
      "epoch": 0.8168414154652687,
      "grad_norm": 0.1733926385641098,
      "learning_rate": 0.0001,
      "loss": 1.3159,
      "step": 4986
    },
    {
      "epoch": 0.8170052424639581,
      "grad_norm": 0.1795656979084015,
      "learning_rate": 0.0001,
      "loss": 1.3554,
      "step": 4987
    },
    {
      "epoch": 0.8171690694626474,
      "grad_norm": 0.1723344326019287,
      "learning_rate": 0.0001,
      "loss": 1.3408,
      "step": 4988
    },
    {
      "epoch": 0.8173328964613368,
      "grad_norm": 0.17750835418701172,
      "learning_rate": 0.0001,
      "loss": 1.2836,
      "step": 4989
    },
    {
      "epoch": 0.8174967234600262,
      "grad_norm": 0.1839168816804886,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 4990
    },
    {
      "epoch": 0.8176605504587156,
      "grad_norm": 0.17688944935798645,
      "learning_rate": 0.0001,
      "loss": 1.3931,
      "step": 4991
    },
    {
      "epoch": 0.817824377457405,
      "grad_norm": 0.1660509556531906,
      "learning_rate": 0.0001,
      "loss": 1.2352,
      "step": 4992
    },
    {
      "epoch": 0.8179882044560943,
      "grad_norm": 0.17612671852111816,
      "learning_rate": 0.0001,
      "loss": 1.4094,
      "step": 4993
    },
    {
      "epoch": 0.8181520314547838,
      "grad_norm": 0.18277350068092346,
      "learning_rate": 0.0001,
      "loss": 1.4133,
      "step": 4994
    },
    {
      "epoch": 0.8183158584534731,
      "grad_norm": 0.1775529831647873,
      "learning_rate": 0.0001,
      "loss": 1.3437,
      "step": 4995
    },
    {
      "epoch": 0.8184796854521625,
      "grad_norm": 0.17067743837833405,
      "learning_rate": 0.0001,
      "loss": 1.3805,
      "step": 4996
    },
    {
      "epoch": 0.8186435124508519,
      "grad_norm": 0.17575551569461823,
      "learning_rate": 0.0001,
      "loss": 1.3434,
      "step": 4997
    },
    {
      "epoch": 0.8188073394495413,
      "grad_norm": 0.17226465046405792,
      "learning_rate": 0.0001,
      "loss": 1.4,
      "step": 4998
    },
    {
      "epoch": 0.8189711664482306,
      "grad_norm": 0.17362995445728302,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 4999
    },
    {
      "epoch": 0.8191349934469201,
      "grad_norm": 0.18064559996128082,
      "learning_rate": 0.0001,
      "loss": 1.3572,
      "step": 5000
    },
    {
      "epoch": 0.8192988204456094,
      "grad_norm": 0.1817115694284439,
      "learning_rate": 0.0001,
      "loss": 1.3625,
      "step": 5001
    },
    {
      "epoch": 0.8194626474442989,
      "grad_norm": 0.17571480572223663,
      "learning_rate": 0.0001,
      "loss": 1.3519,
      "step": 5002
    },
    {
      "epoch": 0.8196264744429882,
      "grad_norm": 0.18456673622131348,
      "learning_rate": 0.0001,
      "loss": 1.4171,
      "step": 5003
    },
    {
      "epoch": 0.8197903014416776,
      "grad_norm": 0.17805778980255127,
      "learning_rate": 0.0001,
      "loss": 1.422,
      "step": 5004
    },
    {
      "epoch": 0.819954128440367,
      "grad_norm": 0.17167864739894867,
      "learning_rate": 0.0001,
      "loss": 1.3509,
      "step": 5005
    },
    {
      "epoch": 0.8201179554390564,
      "grad_norm": 0.17471808195114136,
      "learning_rate": 0.0001,
      "loss": 1.4157,
      "step": 5006
    },
    {
      "epoch": 0.8202817824377457,
      "grad_norm": 0.17034219205379486,
      "learning_rate": 0.0001,
      "loss": 1.2764,
      "step": 5007
    },
    {
      "epoch": 0.8204456094364351,
      "grad_norm": 0.1688457876443863,
      "learning_rate": 0.0001,
      "loss": 1.4164,
      "step": 5008
    },
    {
      "epoch": 0.8206094364351245,
      "grad_norm": 0.17141278088092804,
      "learning_rate": 0.0001,
      "loss": 1.4156,
      "step": 5009
    },
    {
      "epoch": 0.8207732634338138,
      "grad_norm": 0.17908915877342224,
      "learning_rate": 0.0001,
      "loss": 1.2765,
      "step": 5010
    },
    {
      "epoch": 0.8209370904325033,
      "grad_norm": 0.17284944653511047,
      "learning_rate": 0.0001,
      "loss": 1.41,
      "step": 5011
    },
    {
      "epoch": 0.8211009174311926,
      "grad_norm": 0.17615573108196259,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 5012
    },
    {
      "epoch": 0.8212647444298821,
      "grad_norm": 0.1708478182554245,
      "learning_rate": 0.0001,
      "loss": 1.2763,
      "step": 5013
    },
    {
      "epoch": 0.8214285714285714,
      "grad_norm": 0.16773661971092224,
      "learning_rate": 0.0001,
      "loss": 1.373,
      "step": 5014
    },
    {
      "epoch": 0.8215923984272608,
      "grad_norm": 0.19208773970603943,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 5015
    },
    {
      "epoch": 0.8217562254259502,
      "grad_norm": 0.17476460337638855,
      "learning_rate": 0.0001,
      "loss": 1.31,
      "step": 5016
    },
    {
      "epoch": 0.8219200524246396,
      "grad_norm": 0.18825668096542358,
      "learning_rate": 0.0001,
      "loss": 1.2831,
      "step": 5017
    },
    {
      "epoch": 0.822083879423329,
      "grad_norm": 0.17342108488082886,
      "learning_rate": 0.0001,
      "loss": 1.3813,
      "step": 5018
    },
    {
      "epoch": 0.8222477064220184,
      "grad_norm": 0.17711041867733002,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 5019
    },
    {
      "epoch": 0.8224115334207077,
      "grad_norm": 0.17207589745521545,
      "learning_rate": 0.0001,
      "loss": 1.3071,
      "step": 5020
    },
    {
      "epoch": 0.8225753604193972,
      "grad_norm": 0.193899005651474,
      "learning_rate": 0.0001,
      "loss": 1.3246,
      "step": 5021
    },
    {
      "epoch": 0.8227391874180865,
      "grad_norm": 0.17438176274299622,
      "learning_rate": 0.0001,
      "loss": 1.3528,
      "step": 5022
    },
    {
      "epoch": 0.8229030144167759,
      "grad_norm": 0.18860043585300446,
      "learning_rate": 0.0001,
      "loss": 1.3095,
      "step": 5023
    },
    {
      "epoch": 0.8230668414154653,
      "grad_norm": 0.19033053517341614,
      "learning_rate": 0.0001,
      "loss": 1.3262,
      "step": 5024
    },
    {
      "epoch": 0.8232306684141546,
      "grad_norm": 0.18021069467067719,
      "learning_rate": 0.0001,
      "loss": 1.4228,
      "step": 5025
    },
    {
      "epoch": 0.823394495412844,
      "grad_norm": 0.17640307545661926,
      "learning_rate": 0.0001,
      "loss": 1.3687,
      "step": 5026
    },
    {
      "epoch": 0.8235583224115334,
      "grad_norm": 0.1667097955942154,
      "learning_rate": 0.0001,
      "loss": 1.2753,
      "step": 5027
    },
    {
      "epoch": 0.8237221494102228,
      "grad_norm": 0.17063254117965698,
      "learning_rate": 0.0001,
      "loss": 1.3627,
      "step": 5028
    },
    {
      "epoch": 0.8238859764089121,
      "grad_norm": 0.18761074542999268,
      "learning_rate": 0.0001,
      "loss": 1.4208,
      "step": 5029
    },
    {
      "epoch": 0.8240498034076016,
      "grad_norm": 0.17776460945606232,
      "learning_rate": 0.0001,
      "loss": 1.4377,
      "step": 5030
    },
    {
      "epoch": 0.8242136304062909,
      "grad_norm": 0.17480088770389557,
      "learning_rate": 0.0001,
      "loss": 1.3771,
      "step": 5031
    },
    {
      "epoch": 0.8243774574049804,
      "grad_norm": 0.1710280030965805,
      "learning_rate": 0.0001,
      "loss": 1.3632,
      "step": 5032
    },
    {
      "epoch": 0.8245412844036697,
      "grad_norm": 0.16868318617343903,
      "learning_rate": 0.0001,
      "loss": 1.3406,
      "step": 5033
    },
    {
      "epoch": 0.8247051114023591,
      "grad_norm": 0.17941389977931976,
      "learning_rate": 0.0001,
      "loss": 1.4049,
      "step": 5034
    },
    {
      "epoch": 0.8248689384010485,
      "grad_norm": 0.1742735505104065,
      "learning_rate": 0.0001,
      "loss": 1.2109,
      "step": 5035
    },
    {
      "epoch": 0.8250327653997379,
      "grad_norm": 0.17821452021598816,
      "learning_rate": 0.0001,
      "loss": 1.3347,
      "step": 5036
    },
    {
      "epoch": 0.8251965923984272,
      "grad_norm": 0.17109298706054688,
      "learning_rate": 0.0001,
      "loss": 1.266,
      "step": 5037
    },
    {
      "epoch": 0.8253604193971167,
      "grad_norm": 0.16762806475162506,
      "learning_rate": 0.0001,
      "loss": 1.2476,
      "step": 5038
    },
    {
      "epoch": 0.825524246395806,
      "grad_norm": 0.17902772128582,
      "learning_rate": 0.0001,
      "loss": 1.3543,
      "step": 5039
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 0.16528072953224182,
      "learning_rate": 0.0001,
      "loss": 1.1893,
      "step": 5040
    },
    {
      "epoch": 0.8258519003931848,
      "grad_norm": 0.17956794798374176,
      "learning_rate": 0.0001,
      "loss": 1.2998,
      "step": 5041
    },
    {
      "epoch": 0.8260157273918742,
      "grad_norm": 0.1723179817199707,
      "learning_rate": 0.0001,
      "loss": 1.2316,
      "step": 5042
    },
    {
      "epoch": 0.8261795543905636,
      "grad_norm": 0.1753060519695282,
      "learning_rate": 0.0001,
      "loss": 1.3445,
      "step": 5043
    },
    {
      "epoch": 0.8263433813892529,
      "grad_norm": 0.17796047031879425,
      "learning_rate": 0.0001,
      "loss": 1.4251,
      "step": 5044
    },
    {
      "epoch": 0.8265072083879423,
      "grad_norm": 0.1772206872701645,
      "learning_rate": 0.0001,
      "loss": 1.3616,
      "step": 5045
    },
    {
      "epoch": 0.8266710353866317,
      "grad_norm": 0.1716567575931549,
      "learning_rate": 0.0001,
      "loss": 1.3404,
      "step": 5046
    },
    {
      "epoch": 0.8268348623853211,
      "grad_norm": 0.1669720858335495,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 5047
    },
    {
      "epoch": 0.8269986893840104,
      "grad_norm": 0.16869580745697021,
      "learning_rate": 0.0001,
      "loss": 1.3596,
      "step": 5048
    },
    {
      "epoch": 0.8271625163826999,
      "grad_norm": 0.17120392620563507,
      "learning_rate": 0.0001,
      "loss": 1.3259,
      "step": 5049
    },
    {
      "epoch": 0.8273263433813892,
      "grad_norm": 0.17857681214809418,
      "learning_rate": 0.0001,
      "loss": 1.2908,
      "step": 5050
    },
    {
      "epoch": 0.8274901703800787,
      "grad_norm": 0.1810142695903778,
      "learning_rate": 0.0001,
      "loss": 1.3436,
      "step": 5051
    },
    {
      "epoch": 0.827653997378768,
      "grad_norm": 0.1732093095779419,
      "learning_rate": 0.0001,
      "loss": 1.2905,
      "step": 5052
    },
    {
      "epoch": 0.8278178243774574,
      "grad_norm": 0.18079446256160736,
      "learning_rate": 0.0001,
      "loss": 1.3467,
      "step": 5053
    },
    {
      "epoch": 0.8279816513761468,
      "grad_norm": 0.18925921618938446,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 5054
    },
    {
      "epoch": 0.8281454783748362,
      "grad_norm": 0.17643697559833527,
      "learning_rate": 0.0001,
      "loss": 1.3521,
      "step": 5055
    },
    {
      "epoch": 0.8283093053735255,
      "grad_norm": 0.1740826517343521,
      "learning_rate": 0.0001,
      "loss": 1.3136,
      "step": 5056
    },
    {
      "epoch": 0.828473132372215,
      "grad_norm": 0.17326517403125763,
      "learning_rate": 0.0001,
      "loss": 1.2897,
      "step": 5057
    },
    {
      "epoch": 0.8286369593709043,
      "grad_norm": 0.17916835844516754,
      "learning_rate": 0.0001,
      "loss": 1.2466,
      "step": 5058
    },
    {
      "epoch": 0.8288007863695938,
      "grad_norm": 0.18838660418987274,
      "learning_rate": 0.0001,
      "loss": 1.3794,
      "step": 5059
    },
    {
      "epoch": 0.8289646133682831,
      "grad_norm": 0.1742069572210312,
      "learning_rate": 0.0001,
      "loss": 1.3167,
      "step": 5060
    },
    {
      "epoch": 0.8291284403669725,
      "grad_norm": 0.1869794875383377,
      "learning_rate": 0.0001,
      "loss": 1.3329,
      "step": 5061
    },
    {
      "epoch": 0.8292922673656619,
      "grad_norm": 0.16451875865459442,
      "learning_rate": 0.0001,
      "loss": 1.267,
      "step": 5062
    },
    {
      "epoch": 0.8294560943643512,
      "grad_norm": 0.17749285697937012,
      "learning_rate": 0.0001,
      "loss": 1.3178,
      "step": 5063
    },
    {
      "epoch": 0.8296199213630406,
      "grad_norm": 0.18058468401432037,
      "learning_rate": 0.0001,
      "loss": 1.3516,
      "step": 5064
    },
    {
      "epoch": 0.82978374836173,
      "grad_norm": 0.18706613779067993,
      "learning_rate": 0.0001,
      "loss": 1.384,
      "step": 5065
    },
    {
      "epoch": 0.8299475753604194,
      "grad_norm": 0.1692553013563156,
      "learning_rate": 0.0001,
      "loss": 1.4011,
      "step": 5066
    },
    {
      "epoch": 0.8301114023591087,
      "grad_norm": 0.17116416990756989,
      "learning_rate": 0.0001,
      "loss": 1.2381,
      "step": 5067
    },
    {
      "epoch": 0.8302752293577982,
      "grad_norm": 0.1741195172071457,
      "learning_rate": 0.0001,
      "loss": 1.3459,
      "step": 5068
    },
    {
      "epoch": 0.8304390563564875,
      "grad_norm": 0.18828125298023224,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 5069
    },
    {
      "epoch": 0.830602883355177,
      "grad_norm": 0.18853743374347687,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 5070
    },
    {
      "epoch": 0.8307667103538663,
      "grad_norm": 0.181294783949852,
      "learning_rate": 0.0001,
      "loss": 1.3265,
      "step": 5071
    },
    {
      "epoch": 0.8309305373525557,
      "grad_norm": 0.17019270360469818,
      "learning_rate": 0.0001,
      "loss": 1.2751,
      "step": 5072
    },
    {
      "epoch": 0.8310943643512451,
      "grad_norm": 0.1848447322845459,
      "learning_rate": 0.0001,
      "loss": 1.3811,
      "step": 5073
    },
    {
      "epoch": 0.8312581913499345,
      "grad_norm": 0.1838390827178955,
      "learning_rate": 0.0001,
      "loss": 1.3665,
      "step": 5074
    },
    {
      "epoch": 0.8314220183486238,
      "grad_norm": 0.18200309574604034,
      "learning_rate": 0.0001,
      "loss": 1.352,
      "step": 5075
    },
    {
      "epoch": 0.8315858453473133,
      "grad_norm": 0.18295200169086456,
      "learning_rate": 0.0001,
      "loss": 1.397,
      "step": 5076
    },
    {
      "epoch": 0.8317496723460026,
      "grad_norm": 0.18281979858875275,
      "learning_rate": 0.0001,
      "loss": 1.3787,
      "step": 5077
    },
    {
      "epoch": 0.831913499344692,
      "grad_norm": 0.17470628023147583,
      "learning_rate": 0.0001,
      "loss": 1.3266,
      "step": 5078
    },
    {
      "epoch": 0.8320773263433814,
      "grad_norm": 0.194620281457901,
      "learning_rate": 0.0001,
      "loss": 1.406,
      "step": 5079
    },
    {
      "epoch": 0.8322411533420708,
      "grad_norm": 0.16786643862724304,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 5080
    },
    {
      "epoch": 0.8324049803407602,
      "grad_norm": 0.17167212069034576,
      "learning_rate": 0.0001,
      "loss": 1.3407,
      "step": 5081
    },
    {
      "epoch": 0.8325688073394495,
      "grad_norm": 0.19224300980567932,
      "learning_rate": 0.0001,
      "loss": 1.3895,
      "step": 5082
    },
    {
      "epoch": 0.8327326343381389,
      "grad_norm": 0.1875658482313156,
      "learning_rate": 0.0001,
      "loss": 1.2991,
      "step": 5083
    },
    {
      "epoch": 0.8328964613368283,
      "grad_norm": 0.20019041001796722,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 5084
    },
    {
      "epoch": 0.8330602883355177,
      "grad_norm": 0.18387652933597565,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 5085
    },
    {
      "epoch": 0.833224115334207,
      "grad_norm": 0.17794917523860931,
      "learning_rate": 0.0001,
      "loss": 1.4114,
      "step": 5086
    },
    {
      "epoch": 0.8333879423328965,
      "grad_norm": 0.1727590560913086,
      "learning_rate": 0.0001,
      "loss": 1.2675,
      "step": 5087
    },
    {
      "epoch": 0.8335517693315858,
      "grad_norm": 0.1754472553730011,
      "learning_rate": 0.0001,
      "loss": 1.2682,
      "step": 5088
    },
    {
      "epoch": 0.8337155963302753,
      "grad_norm": 0.16397172212600708,
      "learning_rate": 0.0001,
      "loss": 1.2973,
      "step": 5089
    },
    {
      "epoch": 0.8338794233289646,
      "grad_norm": 0.1636645346879959,
      "learning_rate": 0.0001,
      "loss": 1.3068,
      "step": 5090
    },
    {
      "epoch": 0.834043250327654,
      "grad_norm": 0.17277099192142487,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 5091
    },
    {
      "epoch": 0.8342070773263434,
      "grad_norm": 0.16925692558288574,
      "learning_rate": 0.0001,
      "loss": 1.259,
      "step": 5092
    },
    {
      "epoch": 0.8343709043250328,
      "grad_norm": 0.178237646818161,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 5093
    },
    {
      "epoch": 0.8345347313237221,
      "grad_norm": 0.17946214973926544,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 5094
    },
    {
      "epoch": 0.8346985583224116,
      "grad_norm": 0.18018649518489838,
      "learning_rate": 0.0001,
      "loss": 1.2913,
      "step": 5095
    },
    {
      "epoch": 0.8348623853211009,
      "grad_norm": 0.1767573058605194,
      "learning_rate": 0.0001,
      "loss": 1.3322,
      "step": 5096
    },
    {
      "epoch": 0.8350262123197904,
      "grad_norm": 0.18091988563537598,
      "learning_rate": 0.0001,
      "loss": 1.3407,
      "step": 5097
    },
    {
      "epoch": 0.8351900393184797,
      "grad_norm": 0.16831232607364655,
      "learning_rate": 0.0001,
      "loss": 1.3528,
      "step": 5098
    },
    {
      "epoch": 0.8353538663171691,
      "grad_norm": 0.17952695488929749,
      "learning_rate": 0.0001,
      "loss": 1.4096,
      "step": 5099
    },
    {
      "epoch": 0.8355176933158585,
      "grad_norm": 0.17874456942081451,
      "learning_rate": 0.0001,
      "loss": 1.4249,
      "step": 5100
    },
    {
      "epoch": 0.8356815203145478,
      "grad_norm": 0.179121196269989,
      "learning_rate": 0.0001,
      "loss": 1.391,
      "step": 5101
    },
    {
      "epoch": 0.8358453473132372,
      "grad_norm": 0.18079829216003418,
      "learning_rate": 0.0001,
      "loss": 1.3975,
      "step": 5102
    },
    {
      "epoch": 0.8360091743119266,
      "grad_norm": 0.19425858557224274,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 5103
    },
    {
      "epoch": 0.836173001310616,
      "grad_norm": 0.16881072521209717,
      "learning_rate": 0.0001,
      "loss": 1.3615,
      "step": 5104
    },
    {
      "epoch": 0.8363368283093053,
      "grad_norm": 0.17783458530902863,
      "learning_rate": 0.0001,
      "loss": 1.323,
      "step": 5105
    },
    {
      "epoch": 0.8365006553079948,
      "grad_norm": 0.16887445747852325,
      "learning_rate": 0.0001,
      "loss": 1.342,
      "step": 5106
    },
    {
      "epoch": 0.8366644823066841,
      "grad_norm": 0.176887646317482,
      "learning_rate": 0.0001,
      "loss": 1.3062,
      "step": 5107
    },
    {
      "epoch": 0.8368283093053736,
      "grad_norm": 0.17291559278964996,
      "learning_rate": 0.0001,
      "loss": 1.3857,
      "step": 5108
    },
    {
      "epoch": 0.8369921363040629,
      "grad_norm": 0.17166778445243835,
      "learning_rate": 0.0001,
      "loss": 1.351,
      "step": 5109
    },
    {
      "epoch": 0.8371559633027523,
      "grad_norm": 0.17867548763751984,
      "learning_rate": 0.0001,
      "loss": 1.3587,
      "step": 5110
    },
    {
      "epoch": 0.8373197903014417,
      "grad_norm": 0.1753069907426834,
      "learning_rate": 0.0001,
      "loss": 1.3469,
      "step": 5111
    },
    {
      "epoch": 0.8374836173001311,
      "grad_norm": 0.1784152239561081,
      "learning_rate": 0.0001,
      "loss": 1.3963,
      "step": 5112
    },
    {
      "epoch": 0.8376474442988204,
      "grad_norm": 0.17233127355575562,
      "learning_rate": 0.0001,
      "loss": 1.2982,
      "step": 5113
    },
    {
      "epoch": 0.8378112712975099,
      "grad_norm": 0.1714698225259781,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 5114
    },
    {
      "epoch": 0.8379750982961992,
      "grad_norm": 0.16956576704978943,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 5115
    },
    {
      "epoch": 0.8381389252948886,
      "grad_norm": 0.17203588783740997,
      "learning_rate": 0.0001,
      "loss": 1.3048,
      "step": 5116
    },
    {
      "epoch": 0.838302752293578,
      "grad_norm": 0.17448082566261292,
      "learning_rate": 0.0001,
      "loss": 1.3177,
      "step": 5117
    },
    {
      "epoch": 0.8384665792922673,
      "grad_norm": 0.1845913827419281,
      "learning_rate": 0.0001,
      "loss": 1.3472,
      "step": 5118
    },
    {
      "epoch": 0.8386304062909568,
      "grad_norm": 0.18491287529468536,
      "learning_rate": 0.0001,
      "loss": 1.3437,
      "step": 5119
    },
    {
      "epoch": 0.8387942332896461,
      "grad_norm": 0.19441203773021698,
      "learning_rate": 0.0001,
      "loss": 1.4263,
      "step": 5120
    },
    {
      "epoch": 0.8389580602883355,
      "grad_norm": 0.19294461607933044,
      "learning_rate": 0.0001,
      "loss": 1.3741,
      "step": 5121
    },
    {
      "epoch": 0.8391218872870249,
      "grad_norm": 0.17930006980895996,
      "learning_rate": 0.0001,
      "loss": 1.3036,
      "step": 5122
    },
    {
      "epoch": 0.8392857142857143,
      "grad_norm": 0.17647704482078552,
      "learning_rate": 0.0001,
      "loss": 1.2757,
      "step": 5123
    },
    {
      "epoch": 0.8394495412844036,
      "grad_norm": 0.188990518450737,
      "learning_rate": 0.0001,
      "loss": 1.3443,
      "step": 5124
    },
    {
      "epoch": 0.8396133682830931,
      "grad_norm": 0.1672034114599228,
      "learning_rate": 0.0001,
      "loss": 1.2526,
      "step": 5125
    },
    {
      "epoch": 0.8397771952817824,
      "grad_norm": 0.18421845138072968,
      "learning_rate": 0.0001,
      "loss": 1.3826,
      "step": 5126
    },
    {
      "epoch": 0.8399410222804718,
      "grad_norm": 0.19119733572006226,
      "learning_rate": 0.0001,
      "loss": 1.3824,
      "step": 5127
    },
    {
      "epoch": 0.8401048492791612,
      "grad_norm": 0.18512849509716034,
      "learning_rate": 0.0001,
      "loss": 1.3852,
      "step": 5128
    },
    {
      "epoch": 0.8402686762778506,
      "grad_norm": 0.17759408056735992,
      "learning_rate": 0.0001,
      "loss": 1.2519,
      "step": 5129
    },
    {
      "epoch": 0.84043250327654,
      "grad_norm": 0.1735716611146927,
      "learning_rate": 0.0001,
      "loss": 1.2237,
      "step": 5130
    },
    {
      "epoch": 0.8405963302752294,
      "grad_norm": 0.18167077004909515,
      "learning_rate": 0.0001,
      "loss": 1.3227,
      "step": 5131
    },
    {
      "epoch": 0.8407601572739187,
      "grad_norm": 0.170152485370636,
      "learning_rate": 0.0001,
      "loss": 1.3014,
      "step": 5132
    },
    {
      "epoch": 0.8409239842726082,
      "grad_norm": 0.18235193192958832,
      "learning_rate": 0.0001,
      "loss": 1.3085,
      "step": 5133
    },
    {
      "epoch": 0.8410878112712975,
      "grad_norm": 0.17464584112167358,
      "learning_rate": 0.0001,
      "loss": 1.2951,
      "step": 5134
    },
    {
      "epoch": 0.841251638269987,
      "grad_norm": 0.18626801669597626,
      "learning_rate": 0.0001,
      "loss": 1.3922,
      "step": 5135
    },
    {
      "epoch": 0.8414154652686763,
      "grad_norm": 0.17668263614177704,
      "learning_rate": 0.0001,
      "loss": 1.3227,
      "step": 5136
    },
    {
      "epoch": 0.8415792922673656,
      "grad_norm": 0.18894167244434357,
      "learning_rate": 0.0001,
      "loss": 1.2889,
      "step": 5137
    },
    {
      "epoch": 0.841743119266055,
      "grad_norm": 0.19015458226203918,
      "learning_rate": 0.0001,
      "loss": 1.3881,
      "step": 5138
    },
    {
      "epoch": 0.8419069462647444,
      "grad_norm": 0.18185681104660034,
      "learning_rate": 0.0001,
      "loss": 1.3187,
      "step": 5139
    },
    {
      "epoch": 0.8420707732634338,
      "grad_norm": 0.1730969250202179,
      "learning_rate": 0.0001,
      "loss": 1.2185,
      "step": 5140
    },
    {
      "epoch": 0.8422346002621232,
      "grad_norm": 0.19161801040172577,
      "learning_rate": 0.0001,
      "loss": 1.3798,
      "step": 5141
    },
    {
      "epoch": 0.8423984272608126,
      "grad_norm": 0.18491581082344055,
      "learning_rate": 0.0001,
      "loss": 1.4355,
      "step": 5142
    },
    {
      "epoch": 0.8425622542595019,
      "grad_norm": 0.17152124643325806,
      "learning_rate": 0.0001,
      "loss": 1.3229,
      "step": 5143
    },
    {
      "epoch": 0.8427260812581914,
      "grad_norm": 0.18488310277462006,
      "learning_rate": 0.0001,
      "loss": 1.358,
      "step": 5144
    },
    {
      "epoch": 0.8428899082568807,
      "grad_norm": 0.1672462373971939,
      "learning_rate": 0.0001,
      "loss": 1.1837,
      "step": 5145
    },
    {
      "epoch": 0.8430537352555701,
      "grad_norm": 0.17160950601100922,
      "learning_rate": 0.0001,
      "loss": 1.3075,
      "step": 5146
    },
    {
      "epoch": 0.8432175622542595,
      "grad_norm": 0.17597170174121857,
      "learning_rate": 0.0001,
      "loss": 1.3138,
      "step": 5147
    },
    {
      "epoch": 0.8433813892529489,
      "grad_norm": 0.17128923535346985,
      "learning_rate": 0.0001,
      "loss": 1.346,
      "step": 5148
    },
    {
      "epoch": 0.8435452162516383,
      "grad_norm": 0.17555350065231323,
      "learning_rate": 0.0001,
      "loss": 1.4228,
      "step": 5149
    },
    {
      "epoch": 0.8437090432503277,
      "grad_norm": 0.17200085520744324,
      "learning_rate": 0.0001,
      "loss": 1.3714,
      "step": 5150
    },
    {
      "epoch": 0.843872870249017,
      "grad_norm": 0.17451301217079163,
      "learning_rate": 0.0001,
      "loss": 1.2549,
      "step": 5151
    },
    {
      "epoch": 0.8440366972477065,
      "grad_norm": 0.17355452477931976,
      "learning_rate": 0.0001,
      "loss": 1.4406,
      "step": 5152
    },
    {
      "epoch": 0.8442005242463958,
      "grad_norm": 0.17234010994434357,
      "learning_rate": 0.0001,
      "loss": 1.2588,
      "step": 5153
    },
    {
      "epoch": 0.8443643512450852,
      "grad_norm": 0.17192833125591278,
      "learning_rate": 0.0001,
      "loss": 1.2417,
      "step": 5154
    },
    {
      "epoch": 0.8445281782437746,
      "grad_norm": 0.17237181961536407,
      "learning_rate": 0.0001,
      "loss": 1.3013,
      "step": 5155
    },
    {
      "epoch": 0.8446920052424639,
      "grad_norm": 0.171688050031662,
      "learning_rate": 0.0001,
      "loss": 1.421,
      "step": 5156
    },
    {
      "epoch": 0.8448558322411533,
      "grad_norm": 0.17444966733455658,
      "learning_rate": 0.0001,
      "loss": 1.3145,
      "step": 5157
    },
    {
      "epoch": 0.8450196592398427,
      "grad_norm": 0.18291769921779633,
      "learning_rate": 0.0001,
      "loss": 1.3473,
      "step": 5158
    },
    {
      "epoch": 0.8451834862385321,
      "grad_norm": 0.18005597591400146,
      "learning_rate": 0.0001,
      "loss": 1.4313,
      "step": 5159
    },
    {
      "epoch": 0.8453473132372215,
      "grad_norm": 0.18092741072177887,
      "learning_rate": 0.0001,
      "loss": 1.3882,
      "step": 5160
    },
    {
      "epoch": 0.8455111402359109,
      "grad_norm": 0.16394339501857758,
      "learning_rate": 0.0001,
      "loss": 1.2995,
      "step": 5161
    },
    {
      "epoch": 0.8456749672346002,
      "grad_norm": 0.1840459257364273,
      "learning_rate": 0.0001,
      "loss": 1.3089,
      "step": 5162
    },
    {
      "epoch": 0.8458387942332897,
      "grad_norm": 0.17325034737586975,
      "learning_rate": 0.0001,
      "loss": 1.3994,
      "step": 5163
    },
    {
      "epoch": 0.846002621231979,
      "grad_norm": 0.17496997117996216,
      "learning_rate": 0.0001,
      "loss": 1.3897,
      "step": 5164
    },
    {
      "epoch": 0.8461664482306684,
      "grad_norm": 0.1986100822687149,
      "learning_rate": 0.0001,
      "loss": 1.3459,
      "step": 5165
    },
    {
      "epoch": 0.8463302752293578,
      "grad_norm": 0.1853477954864502,
      "learning_rate": 0.0001,
      "loss": 1.3181,
      "step": 5166
    },
    {
      "epoch": 0.8464941022280472,
      "grad_norm": 0.17141953110694885,
      "learning_rate": 0.0001,
      "loss": 1.2707,
      "step": 5167
    },
    {
      "epoch": 0.8466579292267365,
      "grad_norm": 0.17618820071220398,
      "learning_rate": 0.0001,
      "loss": 1.4093,
      "step": 5168
    },
    {
      "epoch": 0.846821756225426,
      "grad_norm": 0.18591833114624023,
      "learning_rate": 0.0001,
      "loss": 1.2972,
      "step": 5169
    },
    {
      "epoch": 0.8469855832241153,
      "grad_norm": 0.17777705192565918,
      "learning_rate": 0.0001,
      "loss": 1.3523,
      "step": 5170
    },
    {
      "epoch": 0.8471494102228048,
      "grad_norm": 0.17619717121124268,
      "learning_rate": 0.0001,
      "loss": 1.2813,
      "step": 5171
    },
    {
      "epoch": 0.8473132372214941,
      "grad_norm": 0.18308453261852264,
      "learning_rate": 0.0001,
      "loss": 1.3554,
      "step": 5172
    },
    {
      "epoch": 0.8474770642201835,
      "grad_norm": 0.176536425948143,
      "learning_rate": 0.0001,
      "loss": 1.4688,
      "step": 5173
    },
    {
      "epoch": 0.8476408912188729,
      "grad_norm": 0.17240986227989197,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 5174
    },
    {
      "epoch": 0.8478047182175622,
      "grad_norm": 0.1784319132566452,
      "learning_rate": 0.0001,
      "loss": 1.3503,
      "step": 5175
    },
    {
      "epoch": 0.8479685452162516,
      "grad_norm": 0.176828995347023,
      "learning_rate": 0.0001,
      "loss": 1.3051,
      "step": 5176
    },
    {
      "epoch": 0.848132372214941,
      "grad_norm": 0.1735338270664215,
      "learning_rate": 0.0001,
      "loss": 1.2281,
      "step": 5177
    },
    {
      "epoch": 0.8482961992136304,
      "grad_norm": 0.19006454944610596,
      "learning_rate": 0.0001,
      "loss": 1.3346,
      "step": 5178
    },
    {
      "epoch": 0.8484600262123198,
      "grad_norm": 0.18143180012702942,
      "learning_rate": 0.0001,
      "loss": 1.3321,
      "step": 5179
    },
    {
      "epoch": 0.8486238532110092,
      "grad_norm": 0.18004751205444336,
      "learning_rate": 0.0001,
      "loss": 1.4058,
      "step": 5180
    },
    {
      "epoch": 0.8487876802096985,
      "grad_norm": 0.20001885294914246,
      "learning_rate": 0.0001,
      "loss": 1.3807,
      "step": 5181
    },
    {
      "epoch": 0.848951507208388,
      "grad_norm": 0.17382235825061798,
      "learning_rate": 0.0001,
      "loss": 1.3154,
      "step": 5182
    },
    {
      "epoch": 0.8491153342070773,
      "grad_norm": 0.18060798943042755,
      "learning_rate": 0.0001,
      "loss": 1.3935,
      "step": 5183
    },
    {
      "epoch": 0.8492791612057667,
      "grad_norm": 0.18655627965927124,
      "learning_rate": 0.0001,
      "loss": 1.3513,
      "step": 5184
    },
    {
      "epoch": 0.8494429882044561,
      "grad_norm": 0.17527756094932556,
      "learning_rate": 0.0001,
      "loss": 1.4293,
      "step": 5185
    },
    {
      "epoch": 0.8496068152031455,
      "grad_norm": 0.17442730069160461,
      "learning_rate": 0.0001,
      "loss": 1.3254,
      "step": 5186
    },
    {
      "epoch": 0.8497706422018348,
      "grad_norm": 0.1838906705379486,
      "learning_rate": 0.0001,
      "loss": 1.5177,
      "step": 5187
    },
    {
      "epoch": 0.8499344692005243,
      "grad_norm": 0.1762714684009552,
      "learning_rate": 0.0001,
      "loss": 1.3987,
      "step": 5188
    },
    {
      "epoch": 0.8500982961992136,
      "grad_norm": 0.173417866230011,
      "learning_rate": 0.0001,
      "loss": 1.2694,
      "step": 5189
    },
    {
      "epoch": 0.8502621231979031,
      "grad_norm": 0.17287255823612213,
      "learning_rate": 0.0001,
      "loss": 1.3603,
      "step": 5190
    },
    {
      "epoch": 0.8504259501965924,
      "grad_norm": 0.1779068261384964,
      "learning_rate": 0.0001,
      "loss": 1.2931,
      "step": 5191
    },
    {
      "epoch": 0.8505897771952818,
      "grad_norm": 0.1699581742286682,
      "learning_rate": 0.0001,
      "loss": 1.2979,
      "step": 5192
    },
    {
      "epoch": 0.8507536041939712,
      "grad_norm": 0.18153899908065796,
      "learning_rate": 0.0001,
      "loss": 1.3901,
      "step": 5193
    },
    {
      "epoch": 0.8509174311926605,
      "grad_norm": 0.19150078296661377,
      "learning_rate": 0.0001,
      "loss": 1.2864,
      "step": 5194
    },
    {
      "epoch": 0.8510812581913499,
      "grad_norm": 0.17802895605564117,
      "learning_rate": 0.0001,
      "loss": 1.3821,
      "step": 5195
    },
    {
      "epoch": 0.8512450851900393,
      "grad_norm": 0.18044593930244446,
      "learning_rate": 0.0001,
      "loss": 1.3906,
      "step": 5196
    },
    {
      "epoch": 0.8514089121887287,
      "grad_norm": 0.17506001889705658,
      "learning_rate": 0.0001,
      "loss": 1.4018,
      "step": 5197
    },
    {
      "epoch": 0.851572739187418,
      "grad_norm": 0.16949091851711273,
      "learning_rate": 0.0001,
      "loss": 1.3614,
      "step": 5198
    },
    {
      "epoch": 0.8517365661861075,
      "grad_norm": 0.18253184854984283,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 5199
    },
    {
      "epoch": 0.8519003931847968,
      "grad_norm": 0.17594082653522491,
      "learning_rate": 0.0001,
      "loss": 1.3728,
      "step": 5200
    },
    {
      "epoch": 0.8520642201834863,
      "grad_norm": 0.18575643002986908,
      "learning_rate": 0.0001,
      "loss": 1.3184,
      "step": 5201
    },
    {
      "epoch": 0.8522280471821756,
      "grad_norm": 0.18192222714424133,
      "learning_rate": 0.0001,
      "loss": 1.3458,
      "step": 5202
    },
    {
      "epoch": 0.852391874180865,
      "grad_norm": 0.20428583025932312,
      "learning_rate": 0.0001,
      "loss": 1.3719,
      "step": 5203
    },
    {
      "epoch": 0.8525557011795544,
      "grad_norm": 0.17848841845989227,
      "learning_rate": 0.0001,
      "loss": 1.3066,
      "step": 5204
    },
    {
      "epoch": 0.8527195281782438,
      "grad_norm": 0.1930730789899826,
      "learning_rate": 0.0001,
      "loss": 1.3668,
      "step": 5205
    },
    {
      "epoch": 0.8528833551769331,
      "grad_norm": 0.17888176441192627,
      "learning_rate": 0.0001,
      "loss": 1.3949,
      "step": 5206
    },
    {
      "epoch": 0.8530471821756226,
      "grad_norm": 0.17274677753448486,
      "learning_rate": 0.0001,
      "loss": 1.4035,
      "step": 5207
    },
    {
      "epoch": 0.8532110091743119,
      "grad_norm": 0.18583513796329498,
      "learning_rate": 0.0001,
      "loss": 1.4688,
      "step": 5208
    },
    {
      "epoch": 0.8533748361730014,
      "grad_norm": 0.1740354597568512,
      "learning_rate": 0.0001,
      "loss": 1.3285,
      "step": 5209
    },
    {
      "epoch": 0.8535386631716907,
      "grad_norm": 0.17746101319789886,
      "learning_rate": 0.0001,
      "loss": 1.2999,
      "step": 5210
    },
    {
      "epoch": 0.85370249017038,
      "grad_norm": 0.18583112955093384,
      "learning_rate": 0.0001,
      "loss": 1.4463,
      "step": 5211
    },
    {
      "epoch": 0.8538663171690695,
      "grad_norm": 0.17411556839942932,
      "learning_rate": 0.0001,
      "loss": 1.3619,
      "step": 5212
    },
    {
      "epoch": 0.8540301441677588,
      "grad_norm": 0.17452983558177948,
      "learning_rate": 0.0001,
      "loss": 1.3692,
      "step": 5213
    },
    {
      "epoch": 0.8541939711664482,
      "grad_norm": 0.185703843832016,
      "learning_rate": 0.0001,
      "loss": 1.3585,
      "step": 5214
    },
    {
      "epoch": 0.8543577981651376,
      "grad_norm": 0.17435337603092194,
      "learning_rate": 0.0001,
      "loss": 1.3721,
      "step": 5215
    },
    {
      "epoch": 0.854521625163827,
      "grad_norm": 0.17187871038913727,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 5216
    },
    {
      "epoch": 0.8546854521625163,
      "grad_norm": 0.17898938059806824,
      "learning_rate": 0.0001,
      "loss": 1.3871,
      "step": 5217
    },
    {
      "epoch": 0.8548492791612058,
      "grad_norm": 0.1736583113670349,
      "learning_rate": 0.0001,
      "loss": 1.2725,
      "step": 5218
    },
    {
      "epoch": 0.8550131061598951,
      "grad_norm": 0.17175041139125824,
      "learning_rate": 0.0001,
      "loss": 1.192,
      "step": 5219
    },
    {
      "epoch": 0.8551769331585846,
      "grad_norm": 0.18453139066696167,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 5220
    },
    {
      "epoch": 0.8553407601572739,
      "grad_norm": 0.18592920899391174,
      "learning_rate": 0.0001,
      "loss": 1.3345,
      "step": 5221
    },
    {
      "epoch": 0.8555045871559633,
      "grad_norm": 0.17385314404964447,
      "learning_rate": 0.0001,
      "loss": 1.3842,
      "step": 5222
    },
    {
      "epoch": 0.8556684141546527,
      "grad_norm": 0.18610812723636627,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 5223
    },
    {
      "epoch": 0.8558322411533421,
      "grad_norm": 0.1765221208333969,
      "learning_rate": 0.0001,
      "loss": 1.4208,
      "step": 5224
    },
    {
      "epoch": 0.8559960681520314,
      "grad_norm": 0.17794473469257355,
      "learning_rate": 0.0001,
      "loss": 1.3336,
      "step": 5225
    },
    {
      "epoch": 0.8561598951507209,
      "grad_norm": 0.1816599816083908,
      "learning_rate": 0.0001,
      "loss": 1.3336,
      "step": 5226
    },
    {
      "epoch": 0.8563237221494102,
      "grad_norm": 0.17736047506332397,
      "learning_rate": 0.0001,
      "loss": 1.3781,
      "step": 5227
    },
    {
      "epoch": 0.8564875491480997,
      "grad_norm": 0.16878631711006165,
      "learning_rate": 0.0001,
      "loss": 1.2569,
      "step": 5228
    },
    {
      "epoch": 0.856651376146789,
      "grad_norm": 0.19243919849395752,
      "learning_rate": 0.0001,
      "loss": 1.3993,
      "step": 5229
    },
    {
      "epoch": 0.8568152031454783,
      "grad_norm": 0.16947904229164124,
      "learning_rate": 0.0001,
      "loss": 1.3167,
      "step": 5230
    },
    {
      "epoch": 0.8569790301441678,
      "grad_norm": 0.1764315515756607,
      "learning_rate": 0.0001,
      "loss": 1.3322,
      "step": 5231
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.1888704150915146,
      "learning_rate": 0.0001,
      "loss": 1.3959,
      "step": 5232
    },
    {
      "epoch": 0.8573066841415465,
      "grad_norm": 0.1748475283384323,
      "learning_rate": 0.0001,
      "loss": 1.4176,
      "step": 5233
    },
    {
      "epoch": 0.8574705111402359,
      "grad_norm": 0.16974985599517822,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 5234
    },
    {
      "epoch": 0.8576343381389253,
      "grad_norm": 0.18378472328186035,
      "learning_rate": 0.0001,
      "loss": 1.3792,
      "step": 5235
    },
    {
      "epoch": 0.8577981651376146,
      "grad_norm": 0.16909654438495636,
      "learning_rate": 0.0001,
      "loss": 1.263,
      "step": 5236
    },
    {
      "epoch": 0.8579619921363041,
      "grad_norm": 0.18088436126708984,
      "learning_rate": 0.0001,
      "loss": 1.3182,
      "step": 5237
    },
    {
      "epoch": 0.8581258191349934,
      "grad_norm": 0.17510980367660522,
      "learning_rate": 0.0001,
      "loss": 1.3307,
      "step": 5238
    },
    {
      "epoch": 0.8582896461336829,
      "grad_norm": 0.18791842460632324,
      "learning_rate": 0.0001,
      "loss": 1.3837,
      "step": 5239
    },
    {
      "epoch": 0.8584534731323722,
      "grad_norm": 0.17521338164806366,
      "learning_rate": 0.0001,
      "loss": 1.3075,
      "step": 5240
    },
    {
      "epoch": 0.8586173001310616,
      "grad_norm": 0.19386246800422668,
      "learning_rate": 0.0001,
      "loss": 1.4163,
      "step": 5241
    },
    {
      "epoch": 0.858781127129751,
      "grad_norm": 0.17249156534671783,
      "learning_rate": 0.0001,
      "loss": 1.3063,
      "step": 5242
    },
    {
      "epoch": 0.8589449541284404,
      "grad_norm": 0.18785534799098969,
      "learning_rate": 0.0001,
      "loss": 1.3316,
      "step": 5243
    },
    {
      "epoch": 0.8591087811271297,
      "grad_norm": 0.18893177807331085,
      "learning_rate": 0.0001,
      "loss": 1.3846,
      "step": 5244
    },
    {
      "epoch": 0.8592726081258192,
      "grad_norm": 0.17869427800178528,
      "learning_rate": 0.0001,
      "loss": 1.2961,
      "step": 5245
    },
    {
      "epoch": 0.8594364351245085,
      "grad_norm": 0.21598877012729645,
      "learning_rate": 0.0001,
      "loss": 1.2935,
      "step": 5246
    },
    {
      "epoch": 0.859600262123198,
      "grad_norm": 0.1778215765953064,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 5247
    },
    {
      "epoch": 0.8597640891218873,
      "grad_norm": 0.16789476573467255,
      "learning_rate": 0.0001,
      "loss": 1.2974,
      "step": 5248
    },
    {
      "epoch": 0.8599279161205766,
      "grad_norm": 0.17872144281864166,
      "learning_rate": 0.0001,
      "loss": 1.2539,
      "step": 5249
    },
    {
      "epoch": 0.8600917431192661,
      "grad_norm": 0.18748104572296143,
      "learning_rate": 0.0001,
      "loss": 1.2784,
      "step": 5250
    },
    {
      "epoch": 0.8602555701179554,
      "grad_norm": 0.18300451338291168,
      "learning_rate": 0.0001,
      "loss": 1.3478,
      "step": 5251
    },
    {
      "epoch": 0.8604193971166448,
      "grad_norm": 0.18990354239940643,
      "learning_rate": 0.0001,
      "loss": 1.4135,
      "step": 5252
    },
    {
      "epoch": 0.8605832241153342,
      "grad_norm": 0.17147907614707947,
      "learning_rate": 0.0001,
      "loss": 1.275,
      "step": 5253
    },
    {
      "epoch": 0.8607470511140236,
      "grad_norm": 0.18788281083106995,
      "learning_rate": 0.0001,
      "loss": 1.3391,
      "step": 5254
    },
    {
      "epoch": 0.8609108781127129,
      "grad_norm": 0.18606455624103546,
      "learning_rate": 0.0001,
      "loss": 1.3917,
      "step": 5255
    },
    {
      "epoch": 0.8610747051114024,
      "grad_norm": 0.19244088232517242,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 5256
    },
    {
      "epoch": 0.8612385321100917,
      "grad_norm": 0.18872381746768951,
      "learning_rate": 0.0001,
      "loss": 1.3475,
      "step": 5257
    },
    {
      "epoch": 0.8614023591087812,
      "grad_norm": 0.17333193123340607,
      "learning_rate": 0.0001,
      "loss": 1.3348,
      "step": 5258
    },
    {
      "epoch": 0.8615661861074705,
      "grad_norm": 0.17794454097747803,
      "learning_rate": 0.0001,
      "loss": 1.3551,
      "step": 5259
    },
    {
      "epoch": 0.8617300131061599,
      "grad_norm": 0.18346138298511505,
      "learning_rate": 0.0001,
      "loss": 1.3102,
      "step": 5260
    },
    {
      "epoch": 0.8618938401048493,
      "grad_norm": 0.17766766250133514,
      "learning_rate": 0.0001,
      "loss": 1.4359,
      "step": 5261
    },
    {
      "epoch": 0.8620576671035387,
      "grad_norm": 0.183883398771286,
      "learning_rate": 0.0001,
      "loss": 1.3371,
      "step": 5262
    },
    {
      "epoch": 0.862221494102228,
      "grad_norm": 0.20065608620643616,
      "learning_rate": 0.0001,
      "loss": 1.3173,
      "step": 5263
    },
    {
      "epoch": 0.8623853211009175,
      "grad_norm": 0.17792199552059174,
      "learning_rate": 0.0001,
      "loss": 1.3323,
      "step": 5264
    },
    {
      "epoch": 0.8625491480996068,
      "grad_norm": 0.17582006752490997,
      "learning_rate": 0.0001,
      "loss": 1.4006,
      "step": 5265
    },
    {
      "epoch": 0.8627129750982963,
      "grad_norm": 0.17928682267665863,
      "learning_rate": 0.0001,
      "loss": 1.4465,
      "step": 5266
    },
    {
      "epoch": 0.8628768020969856,
      "grad_norm": 0.17254024744033813,
      "learning_rate": 0.0001,
      "loss": 1.3082,
      "step": 5267
    },
    {
      "epoch": 0.8630406290956749,
      "grad_norm": 0.1759595274925232,
      "learning_rate": 0.0001,
      "loss": 1.4299,
      "step": 5268
    },
    {
      "epoch": 0.8632044560943644,
      "grad_norm": 0.1723039150238037,
      "learning_rate": 0.0001,
      "loss": 1.3328,
      "step": 5269
    },
    {
      "epoch": 0.8633682830930537,
      "grad_norm": 0.21439236402511597,
      "learning_rate": 0.0001,
      "loss": 1.3502,
      "step": 5270
    },
    {
      "epoch": 0.8635321100917431,
      "grad_norm": 0.18551190197467804,
      "learning_rate": 0.0001,
      "loss": 1.3642,
      "step": 5271
    },
    {
      "epoch": 0.8636959370904325,
      "grad_norm": 0.17605985701084137,
      "learning_rate": 0.0001,
      "loss": 1.2923,
      "step": 5272
    },
    {
      "epoch": 0.8638597640891219,
      "grad_norm": 0.1874447911977768,
      "learning_rate": 0.0001,
      "loss": 1.2569,
      "step": 5273
    },
    {
      "epoch": 0.8640235910878112,
      "grad_norm": 0.19631493091583252,
      "learning_rate": 0.0001,
      "loss": 1.3123,
      "step": 5274
    },
    {
      "epoch": 0.8641874180865007,
      "grad_norm": 0.17296673357486725,
      "learning_rate": 0.0001,
      "loss": 1.2433,
      "step": 5275
    },
    {
      "epoch": 0.86435124508519,
      "grad_norm": 0.19328072667121887,
      "learning_rate": 0.0001,
      "loss": 1.4016,
      "step": 5276
    },
    {
      "epoch": 0.8645150720838795,
      "grad_norm": 0.17468872666358948,
      "learning_rate": 0.0001,
      "loss": 1.2599,
      "step": 5277
    },
    {
      "epoch": 0.8646788990825688,
      "grad_norm": 0.1798609495162964,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 5278
    },
    {
      "epoch": 0.8648427260812582,
      "grad_norm": 0.17448067665100098,
      "learning_rate": 0.0001,
      "loss": 1.3648,
      "step": 5279
    },
    {
      "epoch": 0.8650065530799476,
      "grad_norm": 0.17823894321918488,
      "learning_rate": 0.0001,
      "loss": 1.3257,
      "step": 5280
    },
    {
      "epoch": 0.865170380078637,
      "grad_norm": 0.178439199924469,
      "learning_rate": 0.0001,
      "loss": 1.2858,
      "step": 5281
    },
    {
      "epoch": 0.8653342070773263,
      "grad_norm": 0.19522397220134735,
      "learning_rate": 0.0001,
      "loss": 1.3304,
      "step": 5282
    },
    {
      "epoch": 0.8654980340760158,
      "grad_norm": 0.1731029748916626,
      "learning_rate": 0.0001,
      "loss": 1.3157,
      "step": 5283
    },
    {
      "epoch": 0.8656618610747051,
      "grad_norm": 0.17950063943862915,
      "learning_rate": 0.0001,
      "loss": 1.3293,
      "step": 5284
    },
    {
      "epoch": 0.8658256880733946,
      "grad_norm": 0.17220446467399597,
      "learning_rate": 0.0001,
      "loss": 1.3342,
      "step": 5285
    },
    {
      "epoch": 0.8659895150720839,
      "grad_norm": 0.17436499893665314,
      "learning_rate": 0.0001,
      "loss": 1.2889,
      "step": 5286
    },
    {
      "epoch": 0.8661533420707732,
      "grad_norm": 0.17452797293663025,
      "learning_rate": 0.0001,
      "loss": 1.3123,
      "step": 5287
    },
    {
      "epoch": 0.8663171690694627,
      "grad_norm": 0.1837037354707718,
      "learning_rate": 0.0001,
      "loss": 1.2886,
      "step": 5288
    },
    {
      "epoch": 0.866480996068152,
      "grad_norm": 0.17050479352474213,
      "learning_rate": 0.0001,
      "loss": 1.2648,
      "step": 5289
    },
    {
      "epoch": 0.8666448230668414,
      "grad_norm": 0.18768011033535004,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 5290
    },
    {
      "epoch": 0.8668086500655308,
      "grad_norm": 0.19771833717823029,
      "learning_rate": 0.0001,
      "loss": 1.3067,
      "step": 5291
    },
    {
      "epoch": 0.8669724770642202,
      "grad_norm": 0.17603951692581177,
      "learning_rate": 0.0001,
      "loss": 1.2338,
      "step": 5292
    },
    {
      "epoch": 0.8671363040629095,
      "grad_norm": 0.18933294713497162,
      "learning_rate": 0.0001,
      "loss": 1.2936,
      "step": 5293
    },
    {
      "epoch": 0.867300131061599,
      "grad_norm": 0.18534527719020844,
      "learning_rate": 0.0001,
      "loss": 1.3439,
      "step": 5294
    },
    {
      "epoch": 0.8674639580602883,
      "grad_norm": 0.182266965508461,
      "learning_rate": 0.0001,
      "loss": 1.2732,
      "step": 5295
    },
    {
      "epoch": 0.8676277850589778,
      "grad_norm": 0.17921686172485352,
      "learning_rate": 0.0001,
      "loss": 1.2594,
      "step": 5296
    },
    {
      "epoch": 0.8677916120576671,
      "grad_norm": 0.19289398193359375,
      "learning_rate": 0.0001,
      "loss": 1.3334,
      "step": 5297
    },
    {
      "epoch": 0.8679554390563565,
      "grad_norm": 0.17605628073215485,
      "learning_rate": 0.0001,
      "loss": 1.3878,
      "step": 5298
    },
    {
      "epoch": 0.8681192660550459,
      "grad_norm": 0.17505204677581787,
      "learning_rate": 0.0001,
      "loss": 1.3303,
      "step": 5299
    },
    {
      "epoch": 0.8682830930537353,
      "grad_norm": 0.18474829196929932,
      "learning_rate": 0.0001,
      "loss": 1.2956,
      "step": 5300
    },
    {
      "epoch": 0.8684469200524246,
      "grad_norm": 0.1784558743238449,
      "learning_rate": 0.0001,
      "loss": 1.3378,
      "step": 5301
    },
    {
      "epoch": 0.8686107470511141,
      "grad_norm": 0.17568103969097137,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 5302
    },
    {
      "epoch": 0.8687745740498034,
      "grad_norm": 0.17681390047073364,
      "learning_rate": 0.0001,
      "loss": 1.3005,
      "step": 5303
    },
    {
      "epoch": 0.8689384010484927,
      "grad_norm": 0.1903996765613556,
      "learning_rate": 0.0001,
      "loss": 1.2976,
      "step": 5304
    },
    {
      "epoch": 0.8691022280471822,
      "grad_norm": 0.18759188055992126,
      "learning_rate": 0.0001,
      "loss": 1.3124,
      "step": 5305
    },
    {
      "epoch": 0.8692660550458715,
      "grad_norm": 0.17886239290237427,
      "learning_rate": 0.0001,
      "loss": 1.3128,
      "step": 5306
    },
    {
      "epoch": 0.869429882044561,
      "grad_norm": 0.18403109908103943,
      "learning_rate": 0.0001,
      "loss": 1.4051,
      "step": 5307
    },
    {
      "epoch": 0.8695937090432503,
      "grad_norm": 0.18673934042453766,
      "learning_rate": 0.0001,
      "loss": 1.2996,
      "step": 5308
    },
    {
      "epoch": 0.8697575360419397,
      "grad_norm": 0.16594679653644562,
      "learning_rate": 0.0001,
      "loss": 1.3737,
      "step": 5309
    },
    {
      "epoch": 0.8699213630406291,
      "grad_norm": 0.17200787365436554,
      "learning_rate": 0.0001,
      "loss": 1.2944,
      "step": 5310
    },
    {
      "epoch": 0.8700851900393185,
      "grad_norm": 0.183468759059906,
      "learning_rate": 0.0001,
      "loss": 1.3094,
      "step": 5311
    },
    {
      "epoch": 0.8702490170380078,
      "grad_norm": 0.18481458723545074,
      "learning_rate": 0.0001,
      "loss": 1.3186,
      "step": 5312
    },
    {
      "epoch": 0.8704128440366973,
      "grad_norm": 0.170735165476799,
      "learning_rate": 0.0001,
      "loss": 1.2884,
      "step": 5313
    },
    {
      "epoch": 0.8705766710353866,
      "grad_norm": 0.179454505443573,
      "learning_rate": 0.0001,
      "loss": 1.3283,
      "step": 5314
    },
    {
      "epoch": 0.870740498034076,
      "grad_norm": 0.18144890666007996,
      "learning_rate": 0.0001,
      "loss": 1.3317,
      "step": 5315
    },
    {
      "epoch": 0.8709043250327654,
      "grad_norm": 0.1798287332057953,
      "learning_rate": 0.0001,
      "loss": 1.4185,
      "step": 5316
    },
    {
      "epoch": 0.8710681520314548,
      "grad_norm": 0.1731148064136505,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 5317
    },
    {
      "epoch": 0.8712319790301442,
      "grad_norm": 0.17406146228313446,
      "learning_rate": 0.0001,
      "loss": 1.2869,
      "step": 5318
    },
    {
      "epoch": 0.8713958060288336,
      "grad_norm": 0.18011252582073212,
      "learning_rate": 0.0001,
      "loss": 1.3276,
      "step": 5319
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 0.17287445068359375,
      "learning_rate": 0.0001,
      "loss": 1.3334,
      "step": 5320
    },
    {
      "epoch": 0.8717234600262124,
      "grad_norm": 0.1835472285747528,
      "learning_rate": 0.0001,
      "loss": 1.3544,
      "step": 5321
    },
    {
      "epoch": 0.8718872870249017,
      "grad_norm": 0.17287413775920868,
      "learning_rate": 0.0001,
      "loss": 1.2503,
      "step": 5322
    },
    {
      "epoch": 0.872051114023591,
      "grad_norm": 0.17989295721054077,
      "learning_rate": 0.0001,
      "loss": 1.3816,
      "step": 5323
    },
    {
      "epoch": 0.8722149410222805,
      "grad_norm": 0.18197570741176605,
      "learning_rate": 0.0001,
      "loss": 1.2695,
      "step": 5324
    },
    {
      "epoch": 0.8723787680209698,
      "grad_norm": 0.17987528443336487,
      "learning_rate": 0.0001,
      "loss": 1.3643,
      "step": 5325
    },
    {
      "epoch": 0.8725425950196593,
      "grad_norm": 0.17657896876335144,
      "learning_rate": 0.0001,
      "loss": 1.3975,
      "step": 5326
    },
    {
      "epoch": 0.8727064220183486,
      "grad_norm": 0.17598025500774384,
      "learning_rate": 0.0001,
      "loss": 1.313,
      "step": 5327
    },
    {
      "epoch": 0.872870249017038,
      "grad_norm": 0.19132846593856812,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 5328
    },
    {
      "epoch": 0.8730340760157274,
      "grad_norm": 0.18120944499969482,
      "learning_rate": 0.0001,
      "loss": 1.4014,
      "step": 5329
    },
    {
      "epoch": 0.8731979030144168,
      "grad_norm": 0.1897909790277481,
      "learning_rate": 0.0001,
      "loss": 1.3502,
      "step": 5330
    },
    {
      "epoch": 0.8733617300131061,
      "grad_norm": 0.17509235441684723,
      "learning_rate": 0.0001,
      "loss": 1.2396,
      "step": 5331
    },
    {
      "epoch": 0.8735255570117956,
      "grad_norm": 0.18440690636634827,
      "learning_rate": 0.0001,
      "loss": 1.227,
      "step": 5332
    },
    {
      "epoch": 0.8736893840104849,
      "grad_norm": 0.1786998063325882,
      "learning_rate": 0.0001,
      "loss": 1.3167,
      "step": 5333
    },
    {
      "epoch": 0.8738532110091743,
      "grad_norm": 0.18962566554546356,
      "learning_rate": 0.0001,
      "loss": 1.2725,
      "step": 5334
    },
    {
      "epoch": 0.8740170380078637,
      "grad_norm": 0.1809377521276474,
      "learning_rate": 0.0001,
      "loss": 1.2884,
      "step": 5335
    },
    {
      "epoch": 0.8741808650065531,
      "grad_norm": 0.18887494504451752,
      "learning_rate": 0.0001,
      "loss": 1.2778,
      "step": 5336
    },
    {
      "epoch": 0.8743446920052425,
      "grad_norm": 0.1786268651485443,
      "learning_rate": 0.0001,
      "loss": 1.2845,
      "step": 5337
    },
    {
      "epoch": 0.8745085190039319,
      "grad_norm": 0.1796567738056183,
      "learning_rate": 0.0001,
      "loss": 1.3261,
      "step": 5338
    },
    {
      "epoch": 0.8746723460026212,
      "grad_norm": 0.18101395666599274,
      "learning_rate": 0.0001,
      "loss": 1.2606,
      "step": 5339
    },
    {
      "epoch": 0.8748361730013107,
      "grad_norm": 0.17829802632331848,
      "learning_rate": 0.0001,
      "loss": 1.3276,
      "step": 5340
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.17688913643360138,
      "learning_rate": 0.0001,
      "loss": 1.3994,
      "step": 5341
    },
    {
      "epoch": 0.8751638269986893,
      "grad_norm": 0.17775581777095795,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 5342
    },
    {
      "epoch": 0.8753276539973788,
      "grad_norm": 0.17468737065792084,
      "learning_rate": 0.0001,
      "loss": 1.2762,
      "step": 5343
    },
    {
      "epoch": 0.8754914809960681,
      "grad_norm": 0.18077416718006134,
      "learning_rate": 0.0001,
      "loss": 1.2806,
      "step": 5344
    },
    {
      "epoch": 0.8756553079947575,
      "grad_norm": 0.18424390256404877,
      "learning_rate": 0.0001,
      "loss": 1.4067,
      "step": 5345
    },
    {
      "epoch": 0.8758191349934469,
      "grad_norm": 0.17374926805496216,
      "learning_rate": 0.0001,
      "loss": 1.3489,
      "step": 5346
    },
    {
      "epoch": 0.8759829619921363,
      "grad_norm": 0.18515492975711823,
      "learning_rate": 0.0001,
      "loss": 1.3906,
      "step": 5347
    },
    {
      "epoch": 0.8761467889908257,
      "grad_norm": 0.16797317564487457,
      "learning_rate": 0.0001,
      "loss": 1.3543,
      "step": 5348
    },
    {
      "epoch": 0.8763106159895151,
      "grad_norm": 0.18274426460266113,
      "learning_rate": 0.0001,
      "loss": 1.391,
      "step": 5349
    },
    {
      "epoch": 0.8764744429882044,
      "grad_norm": 0.17476879060268402,
      "learning_rate": 0.0001,
      "loss": 1.2908,
      "step": 5350
    },
    {
      "epoch": 0.8766382699868939,
      "grad_norm": 0.1738181859254837,
      "learning_rate": 0.0001,
      "loss": 1.3395,
      "step": 5351
    },
    {
      "epoch": 0.8768020969855832,
      "grad_norm": 0.17405392229557037,
      "learning_rate": 0.0001,
      "loss": 1.2295,
      "step": 5352
    },
    {
      "epoch": 0.8769659239842726,
      "grad_norm": 0.18539553880691528,
      "learning_rate": 0.0001,
      "loss": 1.1686,
      "step": 5353
    },
    {
      "epoch": 0.877129750982962,
      "grad_norm": 0.18506179749965668,
      "learning_rate": 0.0001,
      "loss": 1.3001,
      "step": 5354
    },
    {
      "epoch": 0.8772935779816514,
      "grad_norm": 0.18494874238967896,
      "learning_rate": 0.0001,
      "loss": 1.3217,
      "step": 5355
    },
    {
      "epoch": 0.8774574049803407,
      "grad_norm": 0.18074442446231842,
      "learning_rate": 0.0001,
      "loss": 1.3023,
      "step": 5356
    },
    {
      "epoch": 0.8776212319790302,
      "grad_norm": 0.17627860605716705,
      "learning_rate": 0.0001,
      "loss": 1.3538,
      "step": 5357
    },
    {
      "epoch": 0.8777850589777195,
      "grad_norm": 0.1795046180486679,
      "learning_rate": 0.0001,
      "loss": 1.3711,
      "step": 5358
    },
    {
      "epoch": 0.877948885976409,
      "grad_norm": 0.17138421535491943,
      "learning_rate": 0.0001,
      "loss": 1.2765,
      "step": 5359
    },
    {
      "epoch": 0.8781127129750983,
      "grad_norm": 0.1788465529680252,
      "learning_rate": 0.0001,
      "loss": 1.3342,
      "step": 5360
    },
    {
      "epoch": 0.8782765399737876,
      "grad_norm": 0.181501567363739,
      "learning_rate": 0.0001,
      "loss": 1.3857,
      "step": 5361
    },
    {
      "epoch": 0.8784403669724771,
      "grad_norm": 0.1880597621202469,
      "learning_rate": 0.0001,
      "loss": 1.408,
      "step": 5362
    },
    {
      "epoch": 0.8786041939711664,
      "grad_norm": 0.17477932572364807,
      "learning_rate": 0.0001,
      "loss": 1.3892,
      "step": 5363
    },
    {
      "epoch": 0.8787680209698558,
      "grad_norm": 0.16944046318531036,
      "learning_rate": 0.0001,
      "loss": 1.3376,
      "step": 5364
    },
    {
      "epoch": 0.8789318479685452,
      "grad_norm": 0.1819402426481247,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 5365
    },
    {
      "epoch": 0.8790956749672346,
      "grad_norm": 0.17478035390377045,
      "learning_rate": 0.0001,
      "loss": 1.2986,
      "step": 5366
    },
    {
      "epoch": 0.879259501965924,
      "grad_norm": 0.19004204869270325,
      "learning_rate": 0.0001,
      "loss": 1.3719,
      "step": 5367
    },
    {
      "epoch": 0.8794233289646134,
      "grad_norm": 0.20203834772109985,
      "learning_rate": 0.0001,
      "loss": 1.283,
      "step": 5368
    },
    {
      "epoch": 0.8795871559633027,
      "grad_norm": 0.17680242657661438,
      "learning_rate": 0.0001,
      "loss": 1.2864,
      "step": 5369
    },
    {
      "epoch": 0.8797509829619922,
      "grad_norm": 0.17632009088993073,
      "learning_rate": 0.0001,
      "loss": 1.3609,
      "step": 5370
    },
    {
      "epoch": 0.8799148099606815,
      "grad_norm": 0.18241798877716064,
      "learning_rate": 0.0001,
      "loss": 1.3527,
      "step": 5371
    },
    {
      "epoch": 0.8800786369593709,
      "grad_norm": 0.19368821382522583,
      "learning_rate": 0.0001,
      "loss": 1.4829,
      "step": 5372
    },
    {
      "epoch": 0.8802424639580603,
      "grad_norm": 0.18820993602275848,
      "learning_rate": 0.0001,
      "loss": 1.3889,
      "step": 5373
    },
    {
      "epoch": 0.8804062909567497,
      "grad_norm": 0.17689287662506104,
      "learning_rate": 0.0001,
      "loss": 1.3973,
      "step": 5374
    },
    {
      "epoch": 0.880570117955439,
      "grad_norm": 0.19351699948310852,
      "learning_rate": 0.0001,
      "loss": 1.3948,
      "step": 5375
    },
    {
      "epoch": 0.8807339449541285,
      "grad_norm": 0.18710441887378693,
      "learning_rate": 0.0001,
      "loss": 1.3162,
      "step": 5376
    },
    {
      "epoch": 0.8808977719528178,
      "grad_norm": 0.17671743035316467,
      "learning_rate": 0.0001,
      "loss": 1.3467,
      "step": 5377
    },
    {
      "epoch": 0.8810615989515073,
      "grad_norm": 0.18709151446819305,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 5378
    },
    {
      "epoch": 0.8812254259501966,
      "grad_norm": 0.18858666718006134,
      "learning_rate": 0.0001,
      "loss": 1.2012,
      "step": 5379
    },
    {
      "epoch": 0.8813892529488859,
      "grad_norm": 0.18672174215316772,
      "learning_rate": 0.0001,
      "loss": 1.4015,
      "step": 5380
    },
    {
      "epoch": 0.8815530799475754,
      "grad_norm": 0.18504728376865387,
      "learning_rate": 0.0001,
      "loss": 1.4346,
      "step": 5381
    },
    {
      "epoch": 0.8817169069462647,
      "grad_norm": 0.18620197474956512,
      "learning_rate": 0.0001,
      "loss": 1.3716,
      "step": 5382
    },
    {
      "epoch": 0.8818807339449541,
      "grad_norm": 0.18409128487110138,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 5383
    },
    {
      "epoch": 0.8820445609436435,
      "grad_norm": 0.16872256994247437,
      "learning_rate": 0.0001,
      "loss": 1.2217,
      "step": 5384
    },
    {
      "epoch": 0.8822083879423329,
      "grad_norm": 0.18875502049922943,
      "learning_rate": 0.0001,
      "loss": 1.4191,
      "step": 5385
    },
    {
      "epoch": 0.8823722149410222,
      "grad_norm": 0.17861731350421906,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 5386
    },
    {
      "epoch": 0.8825360419397117,
      "grad_norm": 0.1864994615316391,
      "learning_rate": 0.0001,
      "loss": 1.3689,
      "step": 5387
    },
    {
      "epoch": 0.882699868938401,
      "grad_norm": 0.18392948806285858,
      "learning_rate": 0.0001,
      "loss": 1.3049,
      "step": 5388
    },
    {
      "epoch": 0.8828636959370905,
      "grad_norm": 0.18644309043884277,
      "learning_rate": 0.0001,
      "loss": 1.3541,
      "step": 5389
    },
    {
      "epoch": 0.8830275229357798,
      "grad_norm": 0.18702152371406555,
      "learning_rate": 0.0001,
      "loss": 1.2664,
      "step": 5390
    },
    {
      "epoch": 0.8831913499344692,
      "grad_norm": 0.19107013940811157,
      "learning_rate": 0.0001,
      "loss": 1.3611,
      "step": 5391
    },
    {
      "epoch": 0.8833551769331586,
      "grad_norm": 0.18728190660476685,
      "learning_rate": 0.0001,
      "loss": 1.3986,
      "step": 5392
    },
    {
      "epoch": 0.883519003931848,
      "grad_norm": 0.18897198140621185,
      "learning_rate": 0.0001,
      "loss": 1.4016,
      "step": 5393
    },
    {
      "epoch": 0.8836828309305373,
      "grad_norm": 0.17785128951072693,
      "learning_rate": 0.0001,
      "loss": 1.2778,
      "step": 5394
    },
    {
      "epoch": 0.8838466579292268,
      "grad_norm": 0.19220805168151855,
      "learning_rate": 0.0001,
      "loss": 1.4259,
      "step": 5395
    },
    {
      "epoch": 0.8840104849279161,
      "grad_norm": 0.17895139753818512,
      "learning_rate": 0.0001,
      "loss": 1.3023,
      "step": 5396
    },
    {
      "epoch": 0.8841743119266054,
      "grad_norm": 0.1776207834482193,
      "learning_rate": 0.0001,
      "loss": 1.3609,
      "step": 5397
    },
    {
      "epoch": 0.8843381389252949,
      "grad_norm": 0.19300536811351776,
      "learning_rate": 0.0001,
      "loss": 1.329,
      "step": 5398
    },
    {
      "epoch": 0.8845019659239842,
      "grad_norm": 0.1776760220527649,
      "learning_rate": 0.0001,
      "loss": 1.3407,
      "step": 5399
    },
    {
      "epoch": 0.8846657929226737,
      "grad_norm": 0.17805445194244385,
      "learning_rate": 0.0001,
      "loss": 1.2709,
      "step": 5400
    },
    {
      "epoch": 0.884829619921363,
      "grad_norm": 0.17985928058624268,
      "learning_rate": 0.0001,
      "loss": 1.3624,
      "step": 5401
    },
    {
      "epoch": 0.8849934469200524,
      "grad_norm": 0.17321844398975372,
      "learning_rate": 0.0001,
      "loss": 1.3523,
      "step": 5402
    },
    {
      "epoch": 0.8851572739187418,
      "grad_norm": 0.18245956301689148,
      "learning_rate": 0.0001,
      "loss": 1.2923,
      "step": 5403
    },
    {
      "epoch": 0.8853211009174312,
      "grad_norm": 0.1812293976545334,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 5404
    },
    {
      "epoch": 0.8854849279161205,
      "grad_norm": 0.18352091312408447,
      "learning_rate": 0.0001,
      "loss": 1.2539,
      "step": 5405
    },
    {
      "epoch": 0.88564875491481,
      "grad_norm": 0.1803702563047409,
      "learning_rate": 0.0001,
      "loss": 1.2316,
      "step": 5406
    },
    {
      "epoch": 0.8858125819134993,
      "grad_norm": 0.17004820704460144,
      "learning_rate": 0.0001,
      "loss": 1.2731,
      "step": 5407
    },
    {
      "epoch": 0.8859764089121888,
      "grad_norm": 0.18559899926185608,
      "learning_rate": 0.0001,
      "loss": 1.4185,
      "step": 5408
    },
    {
      "epoch": 0.8861402359108781,
      "grad_norm": 0.18799439072608948,
      "learning_rate": 0.0001,
      "loss": 1.2982,
      "step": 5409
    },
    {
      "epoch": 0.8863040629095675,
      "grad_norm": 0.1779506355524063,
      "learning_rate": 0.0001,
      "loss": 1.2975,
      "step": 5410
    },
    {
      "epoch": 0.8864678899082569,
      "grad_norm": 0.17749615013599396,
      "learning_rate": 0.0001,
      "loss": 1.3418,
      "step": 5411
    },
    {
      "epoch": 0.8866317169069463,
      "grad_norm": 0.19867770373821259,
      "learning_rate": 0.0001,
      "loss": 1.387,
      "step": 5412
    },
    {
      "epoch": 0.8867955439056356,
      "grad_norm": 0.1769511103630066,
      "learning_rate": 0.0001,
      "loss": 1.3289,
      "step": 5413
    },
    {
      "epoch": 0.8869593709043251,
      "grad_norm": 0.1805976778268814,
      "learning_rate": 0.0001,
      "loss": 1.2555,
      "step": 5414
    },
    {
      "epoch": 0.8871231979030144,
      "grad_norm": 0.17110367119312286,
      "learning_rate": 0.0001,
      "loss": 1.2216,
      "step": 5415
    },
    {
      "epoch": 0.8872870249017037,
      "grad_norm": 0.17837350070476532,
      "learning_rate": 0.0001,
      "loss": 1.2625,
      "step": 5416
    },
    {
      "epoch": 0.8874508519003932,
      "grad_norm": 0.18685494363307953,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 5417
    },
    {
      "epoch": 0.8876146788990825,
      "grad_norm": 0.17962898313999176,
      "learning_rate": 0.0001,
      "loss": 1.3241,
      "step": 5418
    },
    {
      "epoch": 0.887778505897772,
      "grad_norm": 0.18287265300750732,
      "learning_rate": 0.0001,
      "loss": 1.3095,
      "step": 5419
    },
    {
      "epoch": 0.8879423328964613,
      "grad_norm": 0.1846320480108261,
      "learning_rate": 0.0001,
      "loss": 1.311,
      "step": 5420
    },
    {
      "epoch": 0.8881061598951507,
      "grad_norm": 0.1861329972743988,
      "learning_rate": 0.0001,
      "loss": 1.3253,
      "step": 5421
    },
    {
      "epoch": 0.8882699868938401,
      "grad_norm": 0.1786942332983017,
      "learning_rate": 0.0001,
      "loss": 1.3764,
      "step": 5422
    },
    {
      "epoch": 0.8884338138925295,
      "grad_norm": 0.18967953324317932,
      "learning_rate": 0.0001,
      "loss": 1.3025,
      "step": 5423
    },
    {
      "epoch": 0.8885976408912188,
      "grad_norm": 0.18027640879154205,
      "learning_rate": 0.0001,
      "loss": 1.2464,
      "step": 5424
    },
    {
      "epoch": 0.8887614678899083,
      "grad_norm": 0.1733361780643463,
      "learning_rate": 0.0001,
      "loss": 1.3587,
      "step": 5425
    },
    {
      "epoch": 0.8889252948885976,
      "grad_norm": 0.18252219259738922,
      "learning_rate": 0.0001,
      "loss": 1.3632,
      "step": 5426
    },
    {
      "epoch": 0.8890891218872871,
      "grad_norm": 0.18323756754398346,
      "learning_rate": 0.0001,
      "loss": 1.384,
      "step": 5427
    },
    {
      "epoch": 0.8892529488859764,
      "grad_norm": 0.1721859723329544,
      "learning_rate": 0.0001,
      "loss": 1.2769,
      "step": 5428
    },
    {
      "epoch": 0.8894167758846658,
      "grad_norm": 0.182548388838768,
      "learning_rate": 0.0001,
      "loss": 1.4046,
      "step": 5429
    },
    {
      "epoch": 0.8895806028833552,
      "grad_norm": 0.1867980659008026,
      "learning_rate": 0.0001,
      "loss": 1.2346,
      "step": 5430
    },
    {
      "epoch": 0.8897444298820446,
      "grad_norm": 0.185779869556427,
      "learning_rate": 0.0001,
      "loss": 1.2958,
      "step": 5431
    },
    {
      "epoch": 0.8899082568807339,
      "grad_norm": 0.1702088862657547,
      "learning_rate": 0.0001,
      "loss": 1.2895,
      "step": 5432
    },
    {
      "epoch": 0.8900720838794234,
      "grad_norm": 0.17086590826511383,
      "learning_rate": 0.0001,
      "loss": 1.2681,
      "step": 5433
    },
    {
      "epoch": 0.8902359108781127,
      "grad_norm": 0.1741095930337906,
      "learning_rate": 0.0001,
      "loss": 1.2605,
      "step": 5434
    },
    {
      "epoch": 0.890399737876802,
      "grad_norm": 0.17572243511676788,
      "learning_rate": 0.0001,
      "loss": 1.3697,
      "step": 5435
    },
    {
      "epoch": 0.8905635648754915,
      "grad_norm": 0.18491172790527344,
      "learning_rate": 0.0001,
      "loss": 1.3569,
      "step": 5436
    },
    {
      "epoch": 0.8907273918741808,
      "grad_norm": 0.18503624200820923,
      "learning_rate": 0.0001,
      "loss": 1.3097,
      "step": 5437
    },
    {
      "epoch": 0.8908912188728703,
      "grad_norm": 0.1807159036397934,
      "learning_rate": 0.0001,
      "loss": 1.2463,
      "step": 5438
    },
    {
      "epoch": 0.8910550458715596,
      "grad_norm": 0.18927858769893646,
      "learning_rate": 0.0001,
      "loss": 1.4167,
      "step": 5439
    },
    {
      "epoch": 0.891218872870249,
      "grad_norm": 0.1765764355659485,
      "learning_rate": 0.0001,
      "loss": 1.3308,
      "step": 5440
    },
    {
      "epoch": 0.8913826998689384,
      "grad_norm": 0.17622530460357666,
      "learning_rate": 0.0001,
      "loss": 1.3809,
      "step": 5441
    },
    {
      "epoch": 0.8915465268676278,
      "grad_norm": 0.17786026000976562,
      "learning_rate": 0.0001,
      "loss": 1.3739,
      "step": 5442
    },
    {
      "epoch": 0.8917103538663171,
      "grad_norm": 0.17872463166713715,
      "learning_rate": 0.0001,
      "loss": 1.3478,
      "step": 5443
    },
    {
      "epoch": 0.8918741808650066,
      "grad_norm": 0.1890820860862732,
      "learning_rate": 0.0001,
      "loss": 1.3288,
      "step": 5444
    },
    {
      "epoch": 0.8920380078636959,
      "grad_norm": 0.17977099120616913,
      "learning_rate": 0.0001,
      "loss": 1.3051,
      "step": 5445
    },
    {
      "epoch": 0.8922018348623854,
      "grad_norm": 0.18826261162757874,
      "learning_rate": 0.0001,
      "loss": 1.3023,
      "step": 5446
    },
    {
      "epoch": 0.8923656618610747,
      "grad_norm": 0.1778324991464615,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 5447
    },
    {
      "epoch": 0.8925294888597641,
      "grad_norm": 0.18759246170520782,
      "learning_rate": 0.0001,
      "loss": 1.3062,
      "step": 5448
    },
    {
      "epoch": 0.8926933158584535,
      "grad_norm": 0.1872626096010208,
      "learning_rate": 0.0001,
      "loss": 1.3052,
      "step": 5449
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.18577246367931366,
      "learning_rate": 0.0001,
      "loss": 1.2896,
      "step": 5450
    },
    {
      "epoch": 0.8930209698558322,
      "grad_norm": 0.18542899191379547,
      "learning_rate": 0.0001,
      "loss": 1.421,
      "step": 5451
    },
    {
      "epoch": 0.8931847968545217,
      "grad_norm": 0.17555011808872223,
      "learning_rate": 0.0001,
      "loss": 1.3576,
      "step": 5452
    },
    {
      "epoch": 0.893348623853211,
      "grad_norm": 0.18386812508106232,
      "learning_rate": 0.0001,
      "loss": 1.3346,
      "step": 5453
    },
    {
      "epoch": 0.8935124508519003,
      "grad_norm": 0.17990629374980927,
      "learning_rate": 0.0001,
      "loss": 1.3806,
      "step": 5454
    },
    {
      "epoch": 0.8936762778505898,
      "grad_norm": 0.19517874717712402,
      "learning_rate": 0.0001,
      "loss": 1.3121,
      "step": 5455
    },
    {
      "epoch": 0.8938401048492791,
      "grad_norm": 0.18796807527542114,
      "learning_rate": 0.0001,
      "loss": 1.379,
      "step": 5456
    },
    {
      "epoch": 0.8940039318479686,
      "grad_norm": 0.17421090602874756,
      "learning_rate": 0.0001,
      "loss": 1.2031,
      "step": 5457
    },
    {
      "epoch": 0.8941677588466579,
      "grad_norm": 0.17354971170425415,
      "learning_rate": 0.0001,
      "loss": 1.3647,
      "step": 5458
    },
    {
      "epoch": 0.8943315858453473,
      "grad_norm": 0.17970004677772522,
      "learning_rate": 0.0001,
      "loss": 1.2905,
      "step": 5459
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 0.1823984831571579,
      "learning_rate": 0.0001,
      "loss": 1.2313,
      "step": 5460
    },
    {
      "epoch": 0.8946592398427261,
      "grad_norm": 0.18057535588741302,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 5461
    },
    {
      "epoch": 0.8948230668414154,
      "grad_norm": 0.17777150869369507,
      "learning_rate": 0.0001,
      "loss": 1.2489,
      "step": 5462
    },
    {
      "epoch": 0.8949868938401049,
      "grad_norm": 0.17718033492565155,
      "learning_rate": 0.0001,
      "loss": 1.3292,
      "step": 5463
    },
    {
      "epoch": 0.8951507208387942,
      "grad_norm": 0.1845141351222992,
      "learning_rate": 0.0001,
      "loss": 1.2468,
      "step": 5464
    },
    {
      "epoch": 0.8953145478374837,
      "grad_norm": 0.1914377361536026,
      "learning_rate": 0.0001,
      "loss": 1.2938,
      "step": 5465
    },
    {
      "epoch": 0.895478374836173,
      "grad_norm": 0.17033803462982178,
      "learning_rate": 0.0001,
      "loss": 1.3058,
      "step": 5466
    },
    {
      "epoch": 0.8956422018348624,
      "grad_norm": 0.1909252107143402,
      "learning_rate": 0.0001,
      "loss": 1.3718,
      "step": 5467
    },
    {
      "epoch": 0.8958060288335518,
      "grad_norm": 0.1829647421836853,
      "learning_rate": 0.0001,
      "loss": 1.3608,
      "step": 5468
    },
    {
      "epoch": 0.8959698558322412,
      "grad_norm": 0.18356238305568695,
      "learning_rate": 0.0001,
      "loss": 1.3072,
      "step": 5469
    },
    {
      "epoch": 0.8961336828309305,
      "grad_norm": 0.18627101182937622,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 5470
    },
    {
      "epoch": 0.89629750982962,
      "grad_norm": 0.1857389211654663,
      "learning_rate": 0.0001,
      "loss": 1.3786,
      "step": 5471
    },
    {
      "epoch": 0.8964613368283093,
      "grad_norm": 0.19548827409744263,
      "learning_rate": 0.0001,
      "loss": 1.3597,
      "step": 5472
    },
    {
      "epoch": 0.8966251638269986,
      "grad_norm": 0.18759004771709442,
      "learning_rate": 0.0001,
      "loss": 1.3324,
      "step": 5473
    },
    {
      "epoch": 0.8967889908256881,
      "grad_norm": 0.18819496035575867,
      "learning_rate": 0.0001,
      "loss": 1.3947,
      "step": 5474
    },
    {
      "epoch": 0.8969528178243774,
      "grad_norm": 0.18456028401851654,
      "learning_rate": 0.0001,
      "loss": 1.2975,
      "step": 5475
    },
    {
      "epoch": 0.8971166448230669,
      "grad_norm": 0.17670711874961853,
      "learning_rate": 0.0001,
      "loss": 1.3351,
      "step": 5476
    },
    {
      "epoch": 0.8972804718217562,
      "grad_norm": 0.1794968843460083,
      "learning_rate": 0.0001,
      "loss": 1.4155,
      "step": 5477
    },
    {
      "epoch": 0.8974442988204456,
      "grad_norm": 0.1709928661584854,
      "learning_rate": 0.0001,
      "loss": 1.349,
      "step": 5478
    },
    {
      "epoch": 0.897608125819135,
      "grad_norm": 0.18419374525547028,
      "learning_rate": 0.0001,
      "loss": 1.2775,
      "step": 5479
    },
    {
      "epoch": 0.8977719528178244,
      "grad_norm": 0.18374952673912048,
      "learning_rate": 0.0001,
      "loss": 1.3438,
      "step": 5480
    },
    {
      "epoch": 0.8979357798165137,
      "grad_norm": 0.18717342615127563,
      "learning_rate": 0.0001,
      "loss": 1.2705,
      "step": 5481
    },
    {
      "epoch": 0.8980996068152032,
      "grad_norm": 0.17642413079738617,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 5482
    },
    {
      "epoch": 0.8982634338138925,
      "grad_norm": 0.1880856305360794,
      "learning_rate": 0.0001,
      "loss": 1.3067,
      "step": 5483
    },
    {
      "epoch": 0.898427260812582,
      "grad_norm": 0.19163841009140015,
      "learning_rate": 0.0001,
      "loss": 1.4431,
      "step": 5484
    },
    {
      "epoch": 0.8985910878112713,
      "grad_norm": 0.18285274505615234,
      "learning_rate": 0.0001,
      "loss": 1.3559,
      "step": 5485
    },
    {
      "epoch": 0.8987549148099607,
      "grad_norm": 0.18400916457176208,
      "learning_rate": 0.0001,
      "loss": 1.3121,
      "step": 5486
    },
    {
      "epoch": 0.8989187418086501,
      "grad_norm": 0.17311643064022064,
      "learning_rate": 0.0001,
      "loss": 1.3109,
      "step": 5487
    },
    {
      "epoch": 0.8990825688073395,
      "grad_norm": 0.17356769740581512,
      "learning_rate": 0.0001,
      "loss": 1.3361,
      "step": 5488
    },
    {
      "epoch": 0.8992463958060288,
      "grad_norm": 0.1789388507604599,
      "learning_rate": 0.0001,
      "loss": 1.335,
      "step": 5489
    },
    {
      "epoch": 0.8994102228047182,
      "grad_norm": 0.16924963891506195,
      "learning_rate": 0.0001,
      "loss": 1.2302,
      "step": 5490
    },
    {
      "epoch": 0.8995740498034076,
      "grad_norm": 0.18904444575309753,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 5491
    },
    {
      "epoch": 0.8997378768020969,
      "grad_norm": 0.18764923512935638,
      "learning_rate": 0.0001,
      "loss": 1.3492,
      "step": 5492
    },
    {
      "epoch": 0.8999017038007864,
      "grad_norm": 0.197743758559227,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 5493
    },
    {
      "epoch": 0.9000655307994757,
      "grad_norm": 0.17702941596508026,
      "learning_rate": 0.0001,
      "loss": 1.326,
      "step": 5494
    },
    {
      "epoch": 0.9002293577981652,
      "grad_norm": 0.18397684395313263,
      "learning_rate": 0.0001,
      "loss": 1.346,
      "step": 5495
    },
    {
      "epoch": 0.9003931847968545,
      "grad_norm": 0.18810424208641052,
      "learning_rate": 0.0001,
      "loss": 1.4189,
      "step": 5496
    },
    {
      "epoch": 0.9005570117955439,
      "grad_norm": 0.1809644252061844,
      "learning_rate": 0.0001,
      "loss": 1.3849,
      "step": 5497
    },
    {
      "epoch": 0.9007208387942333,
      "grad_norm": 0.1914384365081787,
      "learning_rate": 0.0001,
      "loss": 1.3416,
      "step": 5498
    },
    {
      "epoch": 0.9008846657929227,
      "grad_norm": 0.18307986855506897,
      "learning_rate": 0.0001,
      "loss": 1.4055,
      "step": 5499
    },
    {
      "epoch": 0.901048492791612,
      "grad_norm": 0.178245410323143,
      "learning_rate": 0.0001,
      "loss": 1.3568,
      "step": 5500
    },
    {
      "epoch": 0.9012123197903015,
      "grad_norm": 0.18572796881198883,
      "learning_rate": 0.0001,
      "loss": 1.4358,
      "step": 5501
    },
    {
      "epoch": 0.9013761467889908,
      "grad_norm": 0.18852072954177856,
      "learning_rate": 0.0001,
      "loss": 1.3774,
      "step": 5502
    },
    {
      "epoch": 0.9015399737876802,
      "grad_norm": 0.17863669991493225,
      "learning_rate": 0.0001,
      "loss": 1.3083,
      "step": 5503
    },
    {
      "epoch": 0.9017038007863696,
      "grad_norm": 0.2396285980939865,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 5504
    },
    {
      "epoch": 0.901867627785059,
      "grad_norm": 0.18522495031356812,
      "learning_rate": 0.0001,
      "loss": 1.3469,
      "step": 5505
    },
    {
      "epoch": 0.9020314547837484,
      "grad_norm": 0.17870058119297028,
      "learning_rate": 0.0001,
      "loss": 1.3388,
      "step": 5506
    },
    {
      "epoch": 0.9021952817824378,
      "grad_norm": 0.19535976648330688,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 5507
    },
    {
      "epoch": 0.9023591087811271,
      "grad_norm": 0.18308791518211365,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 5508
    },
    {
      "epoch": 0.9025229357798165,
      "grad_norm": 0.17999528348445892,
      "learning_rate": 0.0001,
      "loss": 1.2835,
      "step": 5509
    },
    {
      "epoch": 0.9026867627785059,
      "grad_norm": 0.17585569620132446,
      "learning_rate": 0.0001,
      "loss": 1.2722,
      "step": 5510
    },
    {
      "epoch": 0.9028505897771952,
      "grad_norm": 0.1898203045129776,
      "learning_rate": 0.0001,
      "loss": 1.3776,
      "step": 5511
    },
    {
      "epoch": 0.9030144167758847,
      "grad_norm": 0.2071542590856552,
      "learning_rate": 0.0001,
      "loss": 1.2992,
      "step": 5512
    },
    {
      "epoch": 0.903178243774574,
      "grad_norm": 0.1899733990430832,
      "learning_rate": 0.0001,
      "loss": 1.39,
      "step": 5513
    },
    {
      "epoch": 0.9033420707732635,
      "grad_norm": 0.17570897936820984,
      "learning_rate": 0.0001,
      "loss": 1.2702,
      "step": 5514
    },
    {
      "epoch": 0.9035058977719528,
      "grad_norm": 0.1821557879447937,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 5515
    },
    {
      "epoch": 0.9036697247706422,
      "grad_norm": 0.1974615901708603,
      "learning_rate": 0.0001,
      "loss": 1.4326,
      "step": 5516
    },
    {
      "epoch": 0.9038335517693316,
      "grad_norm": 0.17997132241725922,
      "learning_rate": 0.0001,
      "loss": 1.2899,
      "step": 5517
    },
    {
      "epoch": 0.903997378768021,
      "grad_norm": 0.17659953236579895,
      "learning_rate": 0.0001,
      "loss": 1.3352,
      "step": 5518
    },
    {
      "epoch": 0.9041612057667103,
      "grad_norm": 0.1793428361415863,
      "learning_rate": 0.0001,
      "loss": 1.3059,
      "step": 5519
    },
    {
      "epoch": 0.9043250327653998,
      "grad_norm": 0.1820075362920761,
      "learning_rate": 0.0001,
      "loss": 1.269,
      "step": 5520
    },
    {
      "epoch": 0.9044888597640891,
      "grad_norm": 0.18372398614883423,
      "learning_rate": 0.0001,
      "loss": 1.4036,
      "step": 5521
    },
    {
      "epoch": 0.9046526867627785,
      "grad_norm": 0.17714956402778625,
      "learning_rate": 0.0001,
      "loss": 1.2759,
      "step": 5522
    },
    {
      "epoch": 0.9048165137614679,
      "grad_norm": 0.1976490318775177,
      "learning_rate": 0.0001,
      "loss": 1.3559,
      "step": 5523
    },
    {
      "epoch": 0.9049803407601573,
      "grad_norm": 0.18388499319553375,
      "learning_rate": 0.0001,
      "loss": 1.3095,
      "step": 5524
    },
    {
      "epoch": 0.9051441677588467,
      "grad_norm": 0.18727831542491913,
      "learning_rate": 0.0001,
      "loss": 1.3817,
      "step": 5525
    },
    {
      "epoch": 0.9053079947575361,
      "grad_norm": 0.1917242556810379,
      "learning_rate": 0.0001,
      "loss": 1.4387,
      "step": 5526
    },
    {
      "epoch": 0.9054718217562254,
      "grad_norm": 0.17862054705619812,
      "learning_rate": 0.0001,
      "loss": 1.351,
      "step": 5527
    },
    {
      "epoch": 0.9056356487549148,
      "grad_norm": 0.17677274346351624,
      "learning_rate": 0.0001,
      "loss": 1.358,
      "step": 5528
    },
    {
      "epoch": 0.9057994757536042,
      "grad_norm": 0.18842044472694397,
      "learning_rate": 0.0001,
      "loss": 1.388,
      "step": 5529
    },
    {
      "epoch": 0.9059633027522935,
      "grad_norm": 0.1900693029165268,
      "learning_rate": 0.0001,
      "loss": 1.2972,
      "step": 5530
    },
    {
      "epoch": 0.906127129750983,
      "grad_norm": 0.18693627417087555,
      "learning_rate": 0.0001,
      "loss": 1.3677,
      "step": 5531
    },
    {
      "epoch": 0.9062909567496723,
      "grad_norm": 0.19752132892608643,
      "learning_rate": 0.0001,
      "loss": 1.3783,
      "step": 5532
    },
    {
      "epoch": 0.9064547837483617,
      "grad_norm": 0.18683332204818726,
      "learning_rate": 0.0001,
      "loss": 1.4806,
      "step": 5533
    },
    {
      "epoch": 0.9066186107470511,
      "grad_norm": 0.19169805943965912,
      "learning_rate": 0.0001,
      "loss": 1.4106,
      "step": 5534
    },
    {
      "epoch": 0.9067824377457405,
      "grad_norm": 0.17466896772384644,
      "learning_rate": 0.0001,
      "loss": 1.3373,
      "step": 5535
    },
    {
      "epoch": 0.9069462647444299,
      "grad_norm": 0.19032089412212372,
      "learning_rate": 0.0001,
      "loss": 1.3674,
      "step": 5536
    },
    {
      "epoch": 0.9071100917431193,
      "grad_norm": 0.18241411447525024,
      "learning_rate": 0.0001,
      "loss": 1.3622,
      "step": 5537
    },
    {
      "epoch": 0.9072739187418086,
      "grad_norm": 0.17700916528701782,
      "learning_rate": 0.0001,
      "loss": 1.3011,
      "step": 5538
    },
    {
      "epoch": 0.9074377457404981,
      "grad_norm": 0.18880009651184082,
      "learning_rate": 0.0001,
      "loss": 1.283,
      "step": 5539
    },
    {
      "epoch": 0.9076015727391874,
      "grad_norm": 0.17539219558238983,
      "learning_rate": 0.0001,
      "loss": 1.2578,
      "step": 5540
    },
    {
      "epoch": 0.9077653997378768,
      "grad_norm": 0.19416998326778412,
      "learning_rate": 0.0001,
      "loss": 1.3353,
      "step": 5541
    },
    {
      "epoch": 0.9079292267365662,
      "grad_norm": 0.1835150122642517,
      "learning_rate": 0.0001,
      "loss": 1.3234,
      "step": 5542
    },
    {
      "epoch": 0.9080930537352556,
      "grad_norm": 0.18620987236499786,
      "learning_rate": 0.0001,
      "loss": 1.3868,
      "step": 5543
    },
    {
      "epoch": 0.908256880733945,
      "grad_norm": 0.19511348009109497,
      "learning_rate": 0.0001,
      "loss": 1.3658,
      "step": 5544
    },
    {
      "epoch": 0.9084207077326344,
      "grad_norm": 0.18302258849143982,
      "learning_rate": 0.0001,
      "loss": 1.3007,
      "step": 5545
    },
    {
      "epoch": 0.9085845347313237,
      "grad_norm": 0.1925557255744934,
      "learning_rate": 0.0001,
      "loss": 1.3879,
      "step": 5546
    },
    {
      "epoch": 0.908748361730013,
      "grad_norm": 0.17658527195453644,
      "learning_rate": 0.0001,
      "loss": 1.421,
      "step": 5547
    },
    {
      "epoch": 0.9089121887287025,
      "grad_norm": 0.18989317119121552,
      "learning_rate": 0.0001,
      "loss": 1.4521,
      "step": 5548
    },
    {
      "epoch": 0.9090760157273918,
      "grad_norm": 0.19499076902866364,
      "learning_rate": 0.0001,
      "loss": 1.4461,
      "step": 5549
    },
    {
      "epoch": 0.9092398427260813,
      "grad_norm": 0.1813579499721527,
      "learning_rate": 0.0001,
      "loss": 1.4518,
      "step": 5550
    },
    {
      "epoch": 0.9094036697247706,
      "grad_norm": 0.1850178986787796,
      "learning_rate": 0.0001,
      "loss": 1.3844,
      "step": 5551
    },
    {
      "epoch": 0.90956749672346,
      "grad_norm": 0.19095247983932495,
      "learning_rate": 0.0001,
      "loss": 1.337,
      "step": 5552
    },
    {
      "epoch": 0.9097313237221494,
      "grad_norm": 0.1775941550731659,
      "learning_rate": 0.0001,
      "loss": 1.3506,
      "step": 5553
    },
    {
      "epoch": 0.9098951507208388,
      "grad_norm": 0.1741885393857956,
      "learning_rate": 0.0001,
      "loss": 1.2018,
      "step": 5554
    },
    {
      "epoch": 0.9100589777195282,
      "grad_norm": 0.18913523852825165,
      "learning_rate": 0.0001,
      "loss": 1.3938,
      "step": 5555
    },
    {
      "epoch": 0.9102228047182176,
      "grad_norm": 0.1837390512228012,
      "learning_rate": 0.0001,
      "loss": 1.3659,
      "step": 5556
    },
    {
      "epoch": 0.9103866317169069,
      "grad_norm": 0.18464884161949158,
      "learning_rate": 0.0001,
      "loss": 1.427,
      "step": 5557
    },
    {
      "epoch": 0.9105504587155964,
      "grad_norm": 0.19117604196071625,
      "learning_rate": 0.0001,
      "loss": 1.4283,
      "step": 5558
    },
    {
      "epoch": 0.9107142857142857,
      "grad_norm": 0.18782788515090942,
      "learning_rate": 0.0001,
      "loss": 1.2959,
      "step": 5559
    },
    {
      "epoch": 0.9108781127129751,
      "grad_norm": 0.1858927607536316,
      "learning_rate": 0.0001,
      "loss": 1.4572,
      "step": 5560
    },
    {
      "epoch": 0.9110419397116645,
      "grad_norm": 0.17805631458759308,
      "learning_rate": 0.0001,
      "loss": 1.2987,
      "step": 5561
    },
    {
      "epoch": 0.9112057667103539,
      "grad_norm": 0.18526165187358856,
      "learning_rate": 0.0001,
      "loss": 1.3757,
      "step": 5562
    },
    {
      "epoch": 0.9113695937090432,
      "grad_norm": 0.2013445347547531,
      "learning_rate": 0.0001,
      "loss": 1.3182,
      "step": 5563
    },
    {
      "epoch": 0.9115334207077327,
      "grad_norm": 0.18806280195713043,
      "learning_rate": 0.0001,
      "loss": 1.2999,
      "step": 5564
    },
    {
      "epoch": 0.911697247706422,
      "grad_norm": 0.18651549518108368,
      "learning_rate": 0.0001,
      "loss": 1.34,
      "step": 5565
    },
    {
      "epoch": 0.9118610747051114,
      "grad_norm": 0.1887681782245636,
      "learning_rate": 0.0001,
      "loss": 1.3879,
      "step": 5566
    },
    {
      "epoch": 0.9120249017038008,
      "grad_norm": 0.18071506917476654,
      "learning_rate": 0.0001,
      "loss": 1.2828,
      "step": 5567
    },
    {
      "epoch": 0.9121887287024901,
      "grad_norm": 0.17542321979999542,
      "learning_rate": 0.0001,
      "loss": 1.3752,
      "step": 5568
    },
    {
      "epoch": 0.9123525557011796,
      "grad_norm": 0.18739664554595947,
      "learning_rate": 0.0001,
      "loss": 1.3807,
      "step": 5569
    },
    {
      "epoch": 0.9125163826998689,
      "grad_norm": 0.19167214632034302,
      "learning_rate": 0.0001,
      "loss": 1.3493,
      "step": 5570
    },
    {
      "epoch": 0.9126802096985583,
      "grad_norm": 0.1889871209859848,
      "learning_rate": 0.0001,
      "loss": 1.3016,
      "step": 5571
    },
    {
      "epoch": 0.9128440366972477,
      "grad_norm": 0.1917068511247635,
      "learning_rate": 0.0001,
      "loss": 1.431,
      "step": 5572
    },
    {
      "epoch": 0.9130078636959371,
      "grad_norm": 0.18522842228412628,
      "learning_rate": 0.0001,
      "loss": 1.3875,
      "step": 5573
    },
    {
      "epoch": 0.9131716906946264,
      "grad_norm": 0.183545783162117,
      "learning_rate": 0.0001,
      "loss": 1.3137,
      "step": 5574
    },
    {
      "epoch": 0.9133355176933159,
      "grad_norm": 0.19515222311019897,
      "learning_rate": 0.0001,
      "loss": 1.3738,
      "step": 5575
    },
    {
      "epoch": 0.9134993446920052,
      "grad_norm": 0.18412986397743225,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 5576
    },
    {
      "epoch": 0.9136631716906947,
      "grad_norm": 0.188499316573143,
      "learning_rate": 0.0001,
      "loss": 1.3965,
      "step": 5577
    },
    {
      "epoch": 0.913826998689384,
      "grad_norm": 0.18549323081970215,
      "learning_rate": 0.0001,
      "loss": 1.3595,
      "step": 5578
    },
    {
      "epoch": 0.9139908256880734,
      "grad_norm": 0.1726742386817932,
      "learning_rate": 0.0001,
      "loss": 1.2995,
      "step": 5579
    },
    {
      "epoch": 0.9141546526867628,
      "grad_norm": 0.1769363433122635,
      "learning_rate": 0.0001,
      "loss": 1.3749,
      "step": 5580
    },
    {
      "epoch": 0.9143184796854522,
      "grad_norm": 0.18674598634243011,
      "learning_rate": 0.0001,
      "loss": 1.4469,
      "step": 5581
    },
    {
      "epoch": 0.9144823066841415,
      "grad_norm": 0.17240197956562042,
      "learning_rate": 0.0001,
      "loss": 1.2539,
      "step": 5582
    },
    {
      "epoch": 0.9146461336828309,
      "grad_norm": 0.18198972940444946,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 5583
    },
    {
      "epoch": 0.9148099606815203,
      "grad_norm": 0.18255572021007538,
      "learning_rate": 0.0001,
      "loss": 1.3179,
      "step": 5584
    },
    {
      "epoch": 0.9149737876802096,
      "grad_norm": 0.19002902507781982,
      "learning_rate": 0.0001,
      "loss": 1.3213,
      "step": 5585
    },
    {
      "epoch": 0.9151376146788991,
      "grad_norm": 0.17450515925884247,
      "learning_rate": 0.0001,
      "loss": 1.2418,
      "step": 5586
    },
    {
      "epoch": 0.9153014416775884,
      "grad_norm": 0.17556728422641754,
      "learning_rate": 0.0001,
      "loss": 1.336,
      "step": 5587
    },
    {
      "epoch": 0.9154652686762779,
      "grad_norm": 0.19653211534023285,
      "learning_rate": 0.0001,
      "loss": 1.2731,
      "step": 5588
    },
    {
      "epoch": 0.9156290956749672,
      "grad_norm": 0.17628030478954315,
      "learning_rate": 0.0001,
      "loss": 1.3068,
      "step": 5589
    },
    {
      "epoch": 0.9157929226736566,
      "grad_norm": 0.18424829840660095,
      "learning_rate": 0.0001,
      "loss": 1.3791,
      "step": 5590
    },
    {
      "epoch": 0.915956749672346,
      "grad_norm": 0.1731567531824112,
      "learning_rate": 0.0001,
      "loss": 1.3167,
      "step": 5591
    },
    {
      "epoch": 0.9161205766710354,
      "grad_norm": 0.1757323443889618,
      "learning_rate": 0.0001,
      "loss": 1.2861,
      "step": 5592
    },
    {
      "epoch": 0.9162844036697247,
      "grad_norm": 0.17931708693504333,
      "learning_rate": 0.0001,
      "loss": 1.2736,
      "step": 5593
    },
    {
      "epoch": 0.9164482306684142,
      "grad_norm": 0.17565813660621643,
      "learning_rate": 0.0001,
      "loss": 1.3239,
      "step": 5594
    },
    {
      "epoch": 0.9166120576671035,
      "grad_norm": 0.173841655254364,
      "learning_rate": 0.0001,
      "loss": 1.2206,
      "step": 5595
    },
    {
      "epoch": 0.916775884665793,
      "grad_norm": 0.17230549454689026,
      "learning_rate": 0.0001,
      "loss": 1.2599,
      "step": 5596
    },
    {
      "epoch": 0.9169397116644823,
      "grad_norm": 0.18162426352500916,
      "learning_rate": 0.0001,
      "loss": 1.4069,
      "step": 5597
    },
    {
      "epoch": 0.9171035386631717,
      "grad_norm": 0.1945345103740692,
      "learning_rate": 0.0001,
      "loss": 1.2749,
      "step": 5598
    },
    {
      "epoch": 0.9172673656618611,
      "grad_norm": 0.18051087856292725,
      "learning_rate": 0.0001,
      "loss": 1.3297,
      "step": 5599
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 0.19030945003032684,
      "learning_rate": 0.0001,
      "loss": 1.3708,
      "step": 5600
    },
    {
      "epoch": 0.9175950196592398,
      "grad_norm": 0.17409515380859375,
      "learning_rate": 0.0001,
      "loss": 1.2319,
      "step": 5601
    },
    {
      "epoch": 0.9177588466579292,
      "grad_norm": 0.1841069608926773,
      "learning_rate": 0.0001,
      "loss": 1.3317,
      "step": 5602
    },
    {
      "epoch": 0.9179226736566186,
      "grad_norm": 0.19598743319511414,
      "learning_rate": 0.0001,
      "loss": 1.3511,
      "step": 5603
    },
    {
      "epoch": 0.918086500655308,
      "grad_norm": 0.17570039629936218,
      "learning_rate": 0.0001,
      "loss": 1.2305,
      "step": 5604
    },
    {
      "epoch": 0.9182503276539974,
      "grad_norm": 0.17618511617183685,
      "learning_rate": 0.0001,
      "loss": 1.2782,
      "step": 5605
    },
    {
      "epoch": 0.9184141546526867,
      "grad_norm": 0.1783740520477295,
      "learning_rate": 0.0001,
      "loss": 1.3597,
      "step": 5606
    },
    {
      "epoch": 0.9185779816513762,
      "grad_norm": 0.18285520374774933,
      "learning_rate": 0.0001,
      "loss": 1.3838,
      "step": 5607
    },
    {
      "epoch": 0.9187418086500655,
      "grad_norm": 0.1818184107542038,
      "learning_rate": 0.0001,
      "loss": 1.3886,
      "step": 5608
    },
    {
      "epoch": 0.9189056356487549,
      "grad_norm": 0.1874217391014099,
      "learning_rate": 0.0001,
      "loss": 1.34,
      "step": 5609
    },
    {
      "epoch": 0.9190694626474443,
      "grad_norm": 0.17900240421295166,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 5610
    },
    {
      "epoch": 0.9192332896461337,
      "grad_norm": 0.17738164961338043,
      "learning_rate": 0.0001,
      "loss": 1.2098,
      "step": 5611
    },
    {
      "epoch": 0.919397116644823,
      "grad_norm": 0.18932491540908813,
      "learning_rate": 0.0001,
      "loss": 1.3177,
      "step": 5612
    },
    {
      "epoch": 0.9195609436435125,
      "grad_norm": 0.20610834658145905,
      "learning_rate": 0.0001,
      "loss": 1.4552,
      "step": 5613
    },
    {
      "epoch": 0.9197247706422018,
      "grad_norm": 0.18792036175727844,
      "learning_rate": 0.0001,
      "loss": 1.3535,
      "step": 5614
    },
    {
      "epoch": 0.9198885976408913,
      "grad_norm": 0.18571782112121582,
      "learning_rate": 0.0001,
      "loss": 1.3101,
      "step": 5615
    },
    {
      "epoch": 0.9200524246395806,
      "grad_norm": 0.18876029551029205,
      "learning_rate": 0.0001,
      "loss": 1.3673,
      "step": 5616
    },
    {
      "epoch": 0.92021625163827,
      "grad_norm": 0.18305331468582153,
      "learning_rate": 0.0001,
      "loss": 1.37,
      "step": 5617
    },
    {
      "epoch": 0.9203800786369594,
      "grad_norm": 0.17201289534568787,
      "learning_rate": 0.0001,
      "loss": 1.3344,
      "step": 5618
    },
    {
      "epoch": 0.9205439056356488,
      "grad_norm": 0.1743224561214447,
      "learning_rate": 0.0001,
      "loss": 1.2783,
      "step": 5619
    },
    {
      "epoch": 0.9207077326343381,
      "grad_norm": 0.1850297898054123,
      "learning_rate": 0.0001,
      "loss": 1.355,
      "step": 5620
    },
    {
      "epoch": 0.9208715596330275,
      "grad_norm": 0.1832486391067505,
      "learning_rate": 0.0001,
      "loss": 1.2978,
      "step": 5621
    },
    {
      "epoch": 0.9210353866317169,
      "grad_norm": 0.18498528003692627,
      "learning_rate": 0.0001,
      "loss": 1.2895,
      "step": 5622
    },
    {
      "epoch": 0.9211992136304062,
      "grad_norm": 0.194499209523201,
      "learning_rate": 0.0001,
      "loss": 1.3456,
      "step": 5623
    },
    {
      "epoch": 0.9213630406290957,
      "grad_norm": 0.17405323684215546,
      "learning_rate": 0.0001,
      "loss": 1.2866,
      "step": 5624
    },
    {
      "epoch": 0.921526867627785,
      "grad_norm": 0.18291433155536652,
      "learning_rate": 0.0001,
      "loss": 1.3212,
      "step": 5625
    },
    {
      "epoch": 0.9216906946264745,
      "grad_norm": 0.1918177306652069,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 5626
    },
    {
      "epoch": 0.9218545216251638,
      "grad_norm": 0.1892549693584442,
      "learning_rate": 0.0001,
      "loss": 1.2951,
      "step": 5627
    },
    {
      "epoch": 0.9220183486238532,
      "grad_norm": 0.18444767594337463,
      "learning_rate": 0.0001,
      "loss": 1.3394,
      "step": 5628
    },
    {
      "epoch": 0.9221821756225426,
      "grad_norm": 0.1926834136247635,
      "learning_rate": 0.0001,
      "loss": 1.3528,
      "step": 5629
    },
    {
      "epoch": 0.922346002621232,
      "grad_norm": 0.19153863191604614,
      "learning_rate": 0.0001,
      "loss": 1.3417,
      "step": 5630
    },
    {
      "epoch": 0.9225098296199213,
      "grad_norm": 0.17809922993183136,
      "learning_rate": 0.0001,
      "loss": 1.4236,
      "step": 5631
    },
    {
      "epoch": 0.9226736566186108,
      "grad_norm": 0.18355277180671692,
      "learning_rate": 0.0001,
      "loss": 1.3514,
      "step": 5632
    },
    {
      "epoch": 0.9228374836173001,
      "grad_norm": 0.17926746606826782,
      "learning_rate": 0.0001,
      "loss": 1.2206,
      "step": 5633
    },
    {
      "epoch": 0.9230013106159896,
      "grad_norm": 0.1827128827571869,
      "learning_rate": 0.0001,
      "loss": 1.2658,
      "step": 5634
    },
    {
      "epoch": 0.9231651376146789,
      "grad_norm": 0.18008919060230255,
      "learning_rate": 0.0001,
      "loss": 1.2942,
      "step": 5635
    },
    {
      "epoch": 0.9233289646133683,
      "grad_norm": 0.18894140422344208,
      "learning_rate": 0.0001,
      "loss": 1.3899,
      "step": 5636
    },
    {
      "epoch": 0.9234927916120577,
      "grad_norm": 0.1802639365196228,
      "learning_rate": 0.0001,
      "loss": 1.3848,
      "step": 5637
    },
    {
      "epoch": 0.9236566186107471,
      "grad_norm": 0.1900857537984848,
      "learning_rate": 0.0001,
      "loss": 1.337,
      "step": 5638
    },
    {
      "epoch": 0.9238204456094364,
      "grad_norm": 0.1770707368850708,
      "learning_rate": 0.0001,
      "loss": 1.3389,
      "step": 5639
    },
    {
      "epoch": 0.9239842726081258,
      "grad_norm": 0.17742325365543365,
      "learning_rate": 0.0001,
      "loss": 1.3509,
      "step": 5640
    },
    {
      "epoch": 0.9241480996068152,
      "grad_norm": 0.17320477962493896,
      "learning_rate": 0.0001,
      "loss": 1.4167,
      "step": 5641
    },
    {
      "epoch": 0.9243119266055045,
      "grad_norm": 0.18075187504291534,
      "learning_rate": 0.0001,
      "loss": 1.2675,
      "step": 5642
    },
    {
      "epoch": 0.924475753604194,
      "grad_norm": 0.17540131509304047,
      "learning_rate": 0.0001,
      "loss": 1.3447,
      "step": 5643
    },
    {
      "epoch": 0.9246395806028833,
      "grad_norm": 0.17761822044849396,
      "learning_rate": 0.0001,
      "loss": 1.2998,
      "step": 5644
    },
    {
      "epoch": 0.9248034076015728,
      "grad_norm": 0.1848095953464508,
      "learning_rate": 0.0001,
      "loss": 1.4193,
      "step": 5645
    },
    {
      "epoch": 0.9249672346002621,
      "grad_norm": 0.17699971795082092,
      "learning_rate": 0.0001,
      "loss": 1.1521,
      "step": 5646
    },
    {
      "epoch": 0.9251310615989515,
      "grad_norm": 0.22643209993839264,
      "learning_rate": 0.0001,
      "loss": 1.2832,
      "step": 5647
    },
    {
      "epoch": 0.9252948885976409,
      "grad_norm": 0.18714049458503723,
      "learning_rate": 0.0001,
      "loss": 1.3016,
      "step": 5648
    },
    {
      "epoch": 0.9254587155963303,
      "grad_norm": 0.1961517333984375,
      "learning_rate": 0.0001,
      "loss": 1.3929,
      "step": 5649
    },
    {
      "epoch": 0.9256225425950196,
      "grad_norm": 0.1827218234539032,
      "learning_rate": 0.0001,
      "loss": 1.3385,
      "step": 5650
    },
    {
      "epoch": 0.9257863695937091,
      "grad_norm": 0.2046092003583908,
      "learning_rate": 0.0001,
      "loss": 1.3372,
      "step": 5651
    },
    {
      "epoch": 0.9259501965923984,
      "grad_norm": 0.18520332872867584,
      "learning_rate": 0.0001,
      "loss": 1.2959,
      "step": 5652
    },
    {
      "epoch": 0.9261140235910879,
      "grad_norm": 0.19550631940364838,
      "learning_rate": 0.0001,
      "loss": 1.4174,
      "step": 5653
    },
    {
      "epoch": 0.9262778505897772,
      "grad_norm": 0.17621956765651703,
      "learning_rate": 0.0001,
      "loss": 1.2937,
      "step": 5654
    },
    {
      "epoch": 0.9264416775884666,
      "grad_norm": 0.17850644886493683,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 5655
    },
    {
      "epoch": 0.926605504587156,
      "grad_norm": 0.17772459983825684,
      "learning_rate": 0.0001,
      "loss": 1.4186,
      "step": 5656
    },
    {
      "epoch": 0.9267693315858454,
      "grad_norm": 0.17387913167476654,
      "learning_rate": 0.0001,
      "loss": 1.3846,
      "step": 5657
    },
    {
      "epoch": 0.9269331585845347,
      "grad_norm": 0.19208960235118866,
      "learning_rate": 0.0001,
      "loss": 1.349,
      "step": 5658
    },
    {
      "epoch": 0.9270969855832241,
      "grad_norm": 0.17319418489933014,
      "learning_rate": 0.0001,
      "loss": 1.3727,
      "step": 5659
    },
    {
      "epoch": 0.9272608125819135,
      "grad_norm": 0.1838841438293457,
      "learning_rate": 0.0001,
      "loss": 1.2863,
      "step": 5660
    },
    {
      "epoch": 0.9274246395806028,
      "grad_norm": 0.18982109427452087,
      "learning_rate": 0.0001,
      "loss": 1.2979,
      "step": 5661
    },
    {
      "epoch": 0.9275884665792923,
      "grad_norm": 0.18396282196044922,
      "learning_rate": 0.0001,
      "loss": 1.2852,
      "step": 5662
    },
    {
      "epoch": 0.9277522935779816,
      "grad_norm": 0.1825844943523407,
      "learning_rate": 0.0001,
      "loss": 1.3384,
      "step": 5663
    },
    {
      "epoch": 0.927916120576671,
      "grad_norm": 0.17749926447868347,
      "learning_rate": 0.0001,
      "loss": 1.348,
      "step": 5664
    },
    {
      "epoch": 0.9280799475753604,
      "grad_norm": 0.20148234069347382,
      "learning_rate": 0.0001,
      "loss": 1.3226,
      "step": 5665
    },
    {
      "epoch": 0.9282437745740498,
      "grad_norm": 0.18093286454677582,
      "learning_rate": 0.0001,
      "loss": 1.3482,
      "step": 5666
    },
    {
      "epoch": 0.9284076015727392,
      "grad_norm": 0.17682741582393646,
      "learning_rate": 0.0001,
      "loss": 1.2846,
      "step": 5667
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 0.17843587696552277,
      "learning_rate": 0.0001,
      "loss": 1.4007,
      "step": 5668
    },
    {
      "epoch": 0.9287352555701179,
      "grad_norm": 0.19075974822044373,
      "learning_rate": 0.0001,
      "loss": 1.3421,
      "step": 5669
    },
    {
      "epoch": 0.9288990825688074,
      "grad_norm": 0.19797421991825104,
      "learning_rate": 0.0001,
      "loss": 1.4118,
      "step": 5670
    },
    {
      "epoch": 0.9290629095674967,
      "grad_norm": 0.182387575507164,
      "learning_rate": 0.0001,
      "loss": 1.3863,
      "step": 5671
    },
    {
      "epoch": 0.9292267365661862,
      "grad_norm": 0.19039319455623627,
      "learning_rate": 0.0001,
      "loss": 1.2874,
      "step": 5672
    },
    {
      "epoch": 0.9293905635648755,
      "grad_norm": 0.17803128063678741,
      "learning_rate": 0.0001,
      "loss": 1.2109,
      "step": 5673
    },
    {
      "epoch": 0.9295543905635649,
      "grad_norm": 0.1878783106803894,
      "learning_rate": 0.0001,
      "loss": 1.3498,
      "step": 5674
    },
    {
      "epoch": 0.9297182175622543,
      "grad_norm": 0.16922688484191895,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 5675
    },
    {
      "epoch": 0.9298820445609436,
      "grad_norm": 0.17712166905403137,
      "learning_rate": 0.0001,
      "loss": 1.2719,
      "step": 5676
    },
    {
      "epoch": 0.930045871559633,
      "grad_norm": 0.17922960221767426,
      "learning_rate": 0.0001,
      "loss": 1.2533,
      "step": 5677
    },
    {
      "epoch": 0.9302096985583224,
      "grad_norm": 0.18749691545963287,
      "learning_rate": 0.0001,
      "loss": 1.3431,
      "step": 5678
    },
    {
      "epoch": 0.9303735255570118,
      "grad_norm": 0.19696854054927826,
      "learning_rate": 0.0001,
      "loss": 1.3488,
      "step": 5679
    },
    {
      "epoch": 0.9305373525557011,
      "grad_norm": 0.18480108678340912,
      "learning_rate": 0.0001,
      "loss": 1.3636,
      "step": 5680
    },
    {
      "epoch": 0.9307011795543906,
      "grad_norm": 0.18998616933822632,
      "learning_rate": 0.0001,
      "loss": 1.191,
      "step": 5681
    },
    {
      "epoch": 0.9308650065530799,
      "grad_norm": 0.1942920684814453,
      "learning_rate": 0.0001,
      "loss": 1.4094,
      "step": 5682
    },
    {
      "epoch": 0.9310288335517694,
      "grad_norm": 0.1812295913696289,
      "learning_rate": 0.0001,
      "loss": 1.3573,
      "step": 5683
    },
    {
      "epoch": 0.9311926605504587,
      "grad_norm": 0.18912464380264282,
      "learning_rate": 0.0001,
      "loss": 1.332,
      "step": 5684
    },
    {
      "epoch": 0.9313564875491481,
      "grad_norm": 0.1889471709728241,
      "learning_rate": 0.0001,
      "loss": 1.2108,
      "step": 5685
    },
    {
      "epoch": 0.9315203145478375,
      "grad_norm": 0.1913864016532898,
      "learning_rate": 0.0001,
      "loss": 1.2493,
      "step": 5686
    },
    {
      "epoch": 0.9316841415465269,
      "grad_norm": 0.18113252520561218,
      "learning_rate": 0.0001,
      "loss": 1.3399,
      "step": 5687
    },
    {
      "epoch": 0.9318479685452162,
      "grad_norm": 0.1873491406440735,
      "learning_rate": 0.0001,
      "loss": 1.3583,
      "step": 5688
    },
    {
      "epoch": 0.9320117955439057,
      "grad_norm": 0.1850830763578415,
      "learning_rate": 0.0001,
      "loss": 1.4142,
      "step": 5689
    },
    {
      "epoch": 0.932175622542595,
      "grad_norm": 0.17995549738407135,
      "learning_rate": 0.0001,
      "loss": 1.2749,
      "step": 5690
    },
    {
      "epoch": 0.9323394495412844,
      "grad_norm": 0.18324103951454163,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 5691
    },
    {
      "epoch": 0.9325032765399738,
      "grad_norm": 0.17573407292366028,
      "learning_rate": 0.0001,
      "loss": 1.2855,
      "step": 5692
    },
    {
      "epoch": 0.9326671035386632,
      "grad_norm": 0.18313400447368622,
      "learning_rate": 0.0001,
      "loss": 1.4282,
      "step": 5693
    },
    {
      "epoch": 0.9328309305373526,
      "grad_norm": 0.17332670092582703,
      "learning_rate": 0.0001,
      "loss": 1.347,
      "step": 5694
    },
    {
      "epoch": 0.9329947575360419,
      "grad_norm": 0.19994238018989563,
      "learning_rate": 0.0001,
      "loss": 1.3559,
      "step": 5695
    },
    {
      "epoch": 0.9331585845347313,
      "grad_norm": 0.18467143177986145,
      "learning_rate": 0.0001,
      "loss": 1.3757,
      "step": 5696
    },
    {
      "epoch": 0.9333224115334207,
      "grad_norm": 0.18546327948570251,
      "learning_rate": 0.0001,
      "loss": 1.3956,
      "step": 5697
    },
    {
      "epoch": 0.9334862385321101,
      "grad_norm": 0.19376644492149353,
      "learning_rate": 0.0001,
      "loss": 1.4728,
      "step": 5698
    },
    {
      "epoch": 0.9336500655307994,
      "grad_norm": 0.19520986080169678,
      "learning_rate": 0.0001,
      "loss": 1.3539,
      "step": 5699
    },
    {
      "epoch": 0.9338138925294889,
      "grad_norm": 0.18435563147068024,
      "learning_rate": 0.0001,
      "loss": 1.3475,
      "step": 5700
    },
    {
      "epoch": 0.9339777195281782,
      "grad_norm": 0.18301647901535034,
      "learning_rate": 0.0001,
      "loss": 1.2851,
      "step": 5701
    },
    {
      "epoch": 0.9341415465268676,
      "grad_norm": 0.17487266659736633,
      "learning_rate": 0.0001,
      "loss": 1.1469,
      "step": 5702
    },
    {
      "epoch": 0.934305373525557,
      "grad_norm": 0.19854606688022614,
      "learning_rate": 0.0001,
      "loss": 1.3436,
      "step": 5703
    },
    {
      "epoch": 0.9344692005242464,
      "grad_norm": 0.18084201216697693,
      "learning_rate": 0.0001,
      "loss": 1.2979,
      "step": 5704
    },
    {
      "epoch": 0.9346330275229358,
      "grad_norm": 0.18113991618156433,
      "learning_rate": 0.0001,
      "loss": 1.3557,
      "step": 5705
    },
    {
      "epoch": 0.9347968545216252,
      "grad_norm": 0.20837324857711792,
      "learning_rate": 0.0001,
      "loss": 1.348,
      "step": 5706
    },
    {
      "epoch": 0.9349606815203145,
      "grad_norm": 0.18285943567752838,
      "learning_rate": 0.0001,
      "loss": 1.4048,
      "step": 5707
    },
    {
      "epoch": 0.935124508519004,
      "grad_norm": 0.17657490074634552,
      "learning_rate": 0.0001,
      "loss": 1.2886,
      "step": 5708
    },
    {
      "epoch": 0.9352883355176933,
      "grad_norm": 0.19367755949497223,
      "learning_rate": 0.0001,
      "loss": 1.4244,
      "step": 5709
    },
    {
      "epoch": 0.9354521625163827,
      "grad_norm": 0.18295475840568542,
      "learning_rate": 0.0001,
      "loss": 1.3932,
      "step": 5710
    },
    {
      "epoch": 0.9356159895150721,
      "grad_norm": 0.18270567059516907,
      "learning_rate": 0.0001,
      "loss": 1.2843,
      "step": 5711
    },
    {
      "epoch": 0.9357798165137615,
      "grad_norm": 0.18926969170570374,
      "learning_rate": 0.0001,
      "loss": 1.3673,
      "step": 5712
    },
    {
      "epoch": 0.9359436435124509,
      "grad_norm": 0.18234802782535553,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 5713
    },
    {
      "epoch": 0.9361074705111402,
      "grad_norm": 0.19117791950702667,
      "learning_rate": 0.0001,
      "loss": 1.3867,
      "step": 5714
    },
    {
      "epoch": 0.9362712975098296,
      "grad_norm": 0.21295741200447083,
      "learning_rate": 0.0001,
      "loss": 1.3919,
      "step": 5715
    },
    {
      "epoch": 0.936435124508519,
      "grad_norm": 0.19208292663097382,
      "learning_rate": 0.0001,
      "loss": 1.3343,
      "step": 5716
    },
    {
      "epoch": 0.9365989515072084,
      "grad_norm": 0.17502108216285706,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 5717
    },
    {
      "epoch": 0.9367627785058977,
      "grad_norm": 0.18331573903560638,
      "learning_rate": 0.0001,
      "loss": 1.3108,
      "step": 5718
    },
    {
      "epoch": 0.9369266055045872,
      "grad_norm": 0.18730387091636658,
      "learning_rate": 0.0001,
      "loss": 1.3127,
      "step": 5719
    },
    {
      "epoch": 0.9370904325032765,
      "grad_norm": 0.1836262196302414,
      "learning_rate": 0.0001,
      "loss": 1.3681,
      "step": 5720
    },
    {
      "epoch": 0.937254259501966,
      "grad_norm": 0.19262485206127167,
      "learning_rate": 0.0001,
      "loss": 1.4645,
      "step": 5721
    },
    {
      "epoch": 0.9374180865006553,
      "grad_norm": 0.1855093091726303,
      "learning_rate": 0.0001,
      "loss": 1.2699,
      "step": 5722
    },
    {
      "epoch": 0.9375819134993447,
      "grad_norm": 0.182024747133255,
      "learning_rate": 0.0001,
      "loss": 1.3264,
      "step": 5723
    },
    {
      "epoch": 0.937745740498034,
      "grad_norm": 0.19097791612148285,
      "learning_rate": 0.0001,
      "loss": 1.348,
      "step": 5724
    },
    {
      "epoch": 0.9379095674967235,
      "grad_norm": 0.1877085566520691,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 5725
    },
    {
      "epoch": 0.9380733944954128,
      "grad_norm": 0.17502102255821228,
      "learning_rate": 0.0001,
      "loss": 1.3223,
      "step": 5726
    },
    {
      "epoch": 0.9382372214941023,
      "grad_norm": 0.18380030989646912,
      "learning_rate": 0.0001,
      "loss": 1.3012,
      "step": 5727
    },
    {
      "epoch": 0.9384010484927916,
      "grad_norm": 0.1787099987268448,
      "learning_rate": 0.0001,
      "loss": 1.3239,
      "step": 5728
    },
    {
      "epoch": 0.938564875491481,
      "grad_norm": 0.18960724771022797,
      "learning_rate": 0.0001,
      "loss": 1.3335,
      "step": 5729
    },
    {
      "epoch": 0.9387287024901704,
      "grad_norm": 0.17247061431407928,
      "learning_rate": 0.0001,
      "loss": 1.2674,
      "step": 5730
    },
    {
      "epoch": 0.9388925294888598,
      "grad_norm": 0.17018543183803558,
      "learning_rate": 0.0001,
      "loss": 1.2417,
      "step": 5731
    },
    {
      "epoch": 0.9390563564875491,
      "grad_norm": 0.18222808837890625,
      "learning_rate": 0.0001,
      "loss": 1.4381,
      "step": 5732
    },
    {
      "epoch": 0.9392201834862385,
      "grad_norm": 0.17912523448467255,
      "learning_rate": 0.0001,
      "loss": 1.2941,
      "step": 5733
    },
    {
      "epoch": 0.9393840104849279,
      "grad_norm": 0.18290793895721436,
      "learning_rate": 0.0001,
      "loss": 1.2275,
      "step": 5734
    },
    {
      "epoch": 0.9395478374836173,
      "grad_norm": 0.18493065237998962,
      "learning_rate": 0.0001,
      "loss": 1.2838,
      "step": 5735
    },
    {
      "epoch": 0.9397116644823067,
      "grad_norm": 0.18806838989257812,
      "learning_rate": 0.0001,
      "loss": 1.4094,
      "step": 5736
    },
    {
      "epoch": 0.939875491480996,
      "grad_norm": 0.1996355652809143,
      "learning_rate": 0.0001,
      "loss": 1.3762,
      "step": 5737
    },
    {
      "epoch": 0.9400393184796855,
      "grad_norm": 0.1970035284757614,
      "learning_rate": 0.0001,
      "loss": 1.3511,
      "step": 5738
    },
    {
      "epoch": 0.9402031454783748,
      "grad_norm": 0.18428586423397064,
      "learning_rate": 0.0001,
      "loss": 1.2733,
      "step": 5739
    },
    {
      "epoch": 0.9403669724770642,
      "grad_norm": 0.18839658796787262,
      "learning_rate": 0.0001,
      "loss": 1.3298,
      "step": 5740
    },
    {
      "epoch": 0.9405307994757536,
      "grad_norm": 0.18523912131786346,
      "learning_rate": 0.0001,
      "loss": 1.3615,
      "step": 5741
    },
    {
      "epoch": 0.940694626474443,
      "grad_norm": 0.1806224137544632,
      "learning_rate": 0.0001,
      "loss": 1.3632,
      "step": 5742
    },
    {
      "epoch": 0.9408584534731324,
      "grad_norm": 0.1830713003873825,
      "learning_rate": 0.0001,
      "loss": 1.3389,
      "step": 5743
    },
    {
      "epoch": 0.9410222804718218,
      "grad_norm": 0.18165744841098785,
      "learning_rate": 0.0001,
      "loss": 1.3096,
      "step": 5744
    },
    {
      "epoch": 0.9411861074705111,
      "grad_norm": 0.177118182182312,
      "learning_rate": 0.0001,
      "loss": 1.1331,
      "step": 5745
    },
    {
      "epoch": 0.9413499344692006,
      "grad_norm": 0.19564208388328552,
      "learning_rate": 0.0001,
      "loss": 1.3148,
      "step": 5746
    },
    {
      "epoch": 0.9415137614678899,
      "grad_norm": 0.1854276806116104,
      "learning_rate": 0.0001,
      "loss": 1.3218,
      "step": 5747
    },
    {
      "epoch": 0.9416775884665793,
      "grad_norm": 0.1809789389371872,
      "learning_rate": 0.0001,
      "loss": 1.2505,
      "step": 5748
    },
    {
      "epoch": 0.9418414154652687,
      "grad_norm": 0.19584248960018158,
      "learning_rate": 0.0001,
      "loss": 1.3901,
      "step": 5749
    },
    {
      "epoch": 0.9420052424639581,
      "grad_norm": 0.17755156755447388,
      "learning_rate": 0.0001,
      "loss": 1.2216,
      "step": 5750
    },
    {
      "epoch": 0.9421690694626474,
      "grad_norm": 0.17834238708019257,
      "learning_rate": 0.0001,
      "loss": 1.2458,
      "step": 5751
    },
    {
      "epoch": 0.9423328964613368,
      "grad_norm": 0.18143683671951294,
      "learning_rate": 0.0001,
      "loss": 1.3152,
      "step": 5752
    },
    {
      "epoch": 0.9424967234600262,
      "grad_norm": 0.1819131076335907,
      "learning_rate": 0.0001,
      "loss": 1.3795,
      "step": 5753
    },
    {
      "epoch": 0.9426605504587156,
      "grad_norm": 0.18914608657360077,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 5754
    },
    {
      "epoch": 0.942824377457405,
      "grad_norm": 0.1834976077079773,
      "learning_rate": 0.0001,
      "loss": 1.2358,
      "step": 5755
    },
    {
      "epoch": 0.9429882044560943,
      "grad_norm": 0.18650491535663605,
      "learning_rate": 0.0001,
      "loss": 1.2947,
      "step": 5756
    },
    {
      "epoch": 0.9431520314547838,
      "grad_norm": 0.2002469301223755,
      "learning_rate": 0.0001,
      "loss": 1.3433,
      "step": 5757
    },
    {
      "epoch": 0.9433158584534731,
      "grad_norm": 0.18945293128490448,
      "learning_rate": 0.0001,
      "loss": 1.3782,
      "step": 5758
    },
    {
      "epoch": 0.9434796854521625,
      "grad_norm": 0.19752860069274902,
      "learning_rate": 0.0001,
      "loss": 1.3518,
      "step": 5759
    },
    {
      "epoch": 0.9436435124508519,
      "grad_norm": 0.1834288388490677,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 5760
    },
    {
      "epoch": 0.9438073394495413,
      "grad_norm": 0.18759983777999878,
      "learning_rate": 0.0001,
      "loss": 1.2452,
      "step": 5761
    },
    {
      "epoch": 0.9439711664482306,
      "grad_norm": 0.1938808560371399,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 5762
    },
    {
      "epoch": 0.9441349934469201,
      "grad_norm": 0.19915558397769928,
      "learning_rate": 0.0001,
      "loss": 1.3027,
      "step": 5763
    },
    {
      "epoch": 0.9442988204456094,
      "grad_norm": 0.18421991169452667,
      "learning_rate": 0.0001,
      "loss": 1.3123,
      "step": 5764
    },
    {
      "epoch": 0.9444626474442989,
      "grad_norm": 0.19352765381336212,
      "learning_rate": 0.0001,
      "loss": 1.2481,
      "step": 5765
    },
    {
      "epoch": 0.9446264744429882,
      "grad_norm": 0.18281477689743042,
      "learning_rate": 0.0001,
      "loss": 1.3256,
      "step": 5766
    },
    {
      "epoch": 0.9447903014416776,
      "grad_norm": 0.17925100028514862,
      "learning_rate": 0.0001,
      "loss": 1.2853,
      "step": 5767
    },
    {
      "epoch": 0.944954128440367,
      "grad_norm": 0.17982962727546692,
      "learning_rate": 0.0001,
      "loss": 1.301,
      "step": 5768
    },
    {
      "epoch": 0.9451179554390564,
      "grad_norm": 0.18736720085144043,
      "learning_rate": 0.0001,
      "loss": 1.4213,
      "step": 5769
    },
    {
      "epoch": 0.9452817824377457,
      "grad_norm": 0.19392314553260803,
      "learning_rate": 0.0001,
      "loss": 1.3668,
      "step": 5770
    },
    {
      "epoch": 0.9454456094364351,
      "grad_norm": 0.19086749851703644,
      "learning_rate": 0.0001,
      "loss": 1.3513,
      "step": 5771
    },
    {
      "epoch": 0.9456094364351245,
      "grad_norm": 0.19242917001247406,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 5772
    },
    {
      "epoch": 0.9457732634338138,
      "grad_norm": 0.20872873067855835,
      "learning_rate": 0.0001,
      "loss": 1.3023,
      "step": 5773
    },
    {
      "epoch": 0.9459370904325033,
      "grad_norm": 0.18822963535785675,
      "learning_rate": 0.0001,
      "loss": 1.2929,
      "step": 5774
    },
    {
      "epoch": 0.9461009174311926,
      "grad_norm": 0.18495118618011475,
      "learning_rate": 0.0001,
      "loss": 1.3823,
      "step": 5775
    },
    {
      "epoch": 0.9462647444298821,
      "grad_norm": 0.19052712619304657,
      "learning_rate": 0.0001,
      "loss": 1.3077,
      "step": 5776
    },
    {
      "epoch": 0.9464285714285714,
      "grad_norm": 0.19079087674617767,
      "learning_rate": 0.0001,
      "loss": 1.3407,
      "step": 5777
    },
    {
      "epoch": 0.9465923984272608,
      "grad_norm": 0.18597713112831116,
      "learning_rate": 0.0001,
      "loss": 1.376,
      "step": 5778
    },
    {
      "epoch": 0.9467562254259502,
      "grad_norm": 0.1747102588415146,
      "learning_rate": 0.0001,
      "loss": 1.3623,
      "step": 5779
    },
    {
      "epoch": 0.9469200524246396,
      "grad_norm": 0.1854132115840912,
      "learning_rate": 0.0001,
      "loss": 1.3547,
      "step": 5780
    },
    {
      "epoch": 0.947083879423329,
      "grad_norm": 0.1720830202102661,
      "learning_rate": 0.0001,
      "loss": 1.3353,
      "step": 5781
    },
    {
      "epoch": 0.9472477064220184,
      "grad_norm": 0.18552304804325104,
      "learning_rate": 0.0001,
      "loss": 1.3954,
      "step": 5782
    },
    {
      "epoch": 0.9474115334207077,
      "grad_norm": 0.17588958144187927,
      "learning_rate": 0.0001,
      "loss": 1.2561,
      "step": 5783
    },
    {
      "epoch": 0.9475753604193972,
      "grad_norm": 0.18359510600566864,
      "learning_rate": 0.0001,
      "loss": 1.2954,
      "step": 5784
    },
    {
      "epoch": 0.9477391874180865,
      "grad_norm": 0.18837805092334747,
      "learning_rate": 0.0001,
      "loss": 1.3633,
      "step": 5785
    },
    {
      "epoch": 0.9479030144167759,
      "grad_norm": 0.18080727756023407,
      "learning_rate": 0.0001,
      "loss": 1.3666,
      "step": 5786
    },
    {
      "epoch": 0.9480668414154653,
      "grad_norm": 0.18925975263118744,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 5787
    },
    {
      "epoch": 0.9482306684141546,
      "grad_norm": 0.1935735046863556,
      "learning_rate": 0.0001,
      "loss": 1.4193,
      "step": 5788
    },
    {
      "epoch": 0.948394495412844,
      "grad_norm": 0.17755676805973053,
      "learning_rate": 0.0001,
      "loss": 1.3449,
      "step": 5789
    },
    {
      "epoch": 0.9485583224115334,
      "grad_norm": 0.1931099146604538,
      "learning_rate": 0.0001,
      "loss": 1.3136,
      "step": 5790
    },
    {
      "epoch": 0.9487221494102228,
      "grad_norm": 0.18470478057861328,
      "learning_rate": 0.0001,
      "loss": 1.3918,
      "step": 5791
    },
    {
      "epoch": 0.9488859764089121,
      "grad_norm": 0.178767129778862,
      "learning_rate": 0.0001,
      "loss": 1.35,
      "step": 5792
    },
    {
      "epoch": 0.9490498034076016,
      "grad_norm": 0.1916079968214035,
      "learning_rate": 0.0001,
      "loss": 1.3627,
      "step": 5793
    },
    {
      "epoch": 0.9492136304062909,
      "grad_norm": 0.18581002950668335,
      "learning_rate": 0.0001,
      "loss": 1.343,
      "step": 5794
    },
    {
      "epoch": 0.9493774574049804,
      "grad_norm": 0.19296199083328247,
      "learning_rate": 0.0001,
      "loss": 1.3782,
      "step": 5795
    },
    {
      "epoch": 0.9495412844036697,
      "grad_norm": 0.20208586752414703,
      "learning_rate": 0.0001,
      "loss": 1.3413,
      "step": 5796
    },
    {
      "epoch": 0.9497051114023591,
      "grad_norm": 0.18598854541778564,
      "learning_rate": 0.0001,
      "loss": 1.3219,
      "step": 5797
    },
    {
      "epoch": 0.9498689384010485,
      "grad_norm": 0.18153952062129974,
      "learning_rate": 0.0001,
      "loss": 1.3132,
      "step": 5798
    },
    {
      "epoch": 0.9500327653997379,
      "grad_norm": 0.1903291642665863,
      "learning_rate": 0.0001,
      "loss": 1.2954,
      "step": 5799
    },
    {
      "epoch": 0.9501965923984272,
      "grad_norm": 0.18732579052448273,
      "learning_rate": 0.0001,
      "loss": 1.427,
      "step": 5800
    },
    {
      "epoch": 0.9503604193971167,
      "grad_norm": 0.1886163353919983,
      "learning_rate": 0.0001,
      "loss": 1.2684,
      "step": 5801
    },
    {
      "epoch": 0.950524246395806,
      "grad_norm": 0.1831423044204712,
      "learning_rate": 0.0001,
      "loss": 1.2941,
      "step": 5802
    },
    {
      "epoch": 0.9506880733944955,
      "grad_norm": 0.18574419617652893,
      "learning_rate": 0.0001,
      "loss": 1.3815,
      "step": 5803
    },
    {
      "epoch": 0.9508519003931848,
      "grad_norm": 0.1857692301273346,
      "learning_rate": 0.0001,
      "loss": 1.2963,
      "step": 5804
    },
    {
      "epoch": 0.9510157273918742,
      "grad_norm": 0.1888190656900406,
      "learning_rate": 0.0001,
      "loss": 1.2392,
      "step": 5805
    },
    {
      "epoch": 0.9511795543905636,
      "grad_norm": 0.1847425252199173,
      "learning_rate": 0.0001,
      "loss": 1.3797,
      "step": 5806
    },
    {
      "epoch": 0.9513433813892529,
      "grad_norm": 0.17799273133277893,
      "learning_rate": 0.0001,
      "loss": 1.356,
      "step": 5807
    },
    {
      "epoch": 0.9515072083879423,
      "grad_norm": 0.18338064849376678,
      "learning_rate": 0.0001,
      "loss": 1.2679,
      "step": 5808
    },
    {
      "epoch": 0.9516710353866317,
      "grad_norm": 0.1918811947107315,
      "learning_rate": 0.0001,
      "loss": 1.3222,
      "step": 5809
    },
    {
      "epoch": 0.9518348623853211,
      "grad_norm": 0.19962571561336517,
      "learning_rate": 0.0001,
      "loss": 1.335,
      "step": 5810
    },
    {
      "epoch": 0.9519986893840104,
      "grad_norm": 0.18680068850517273,
      "learning_rate": 0.0001,
      "loss": 1.2963,
      "step": 5811
    },
    {
      "epoch": 0.9521625163826999,
      "grad_norm": 0.1927759349346161,
      "learning_rate": 0.0001,
      "loss": 1.326,
      "step": 5812
    },
    {
      "epoch": 0.9523263433813892,
      "grad_norm": 0.18462929129600525,
      "learning_rate": 0.0001,
      "loss": 1.2859,
      "step": 5813
    },
    {
      "epoch": 0.9524901703800787,
      "grad_norm": 0.17634911835193634,
      "learning_rate": 0.0001,
      "loss": 1.2845,
      "step": 5814
    },
    {
      "epoch": 0.952653997378768,
      "grad_norm": 0.18598423898220062,
      "learning_rate": 0.0001,
      "loss": 1.3303,
      "step": 5815
    },
    {
      "epoch": 0.9528178243774574,
      "grad_norm": 0.19075563549995422,
      "learning_rate": 0.0001,
      "loss": 1.3567,
      "step": 5816
    },
    {
      "epoch": 0.9529816513761468,
      "grad_norm": 0.17935127019882202,
      "learning_rate": 0.0001,
      "loss": 1.2223,
      "step": 5817
    },
    {
      "epoch": 0.9531454783748362,
      "grad_norm": 0.19216184318065643,
      "learning_rate": 0.0001,
      "loss": 1.3731,
      "step": 5818
    },
    {
      "epoch": 0.9533093053735255,
      "grad_norm": 0.17871038615703583,
      "learning_rate": 0.0001,
      "loss": 1.2536,
      "step": 5819
    },
    {
      "epoch": 0.953473132372215,
      "grad_norm": 0.189533069729805,
      "learning_rate": 0.0001,
      "loss": 1.3635,
      "step": 5820
    },
    {
      "epoch": 0.9536369593709043,
      "grad_norm": 0.18069539964199066,
      "learning_rate": 0.0001,
      "loss": 1.3279,
      "step": 5821
    },
    {
      "epoch": 0.9538007863695938,
      "grad_norm": 0.20681527256965637,
      "learning_rate": 0.0001,
      "loss": 1.385,
      "step": 5822
    },
    {
      "epoch": 0.9539646133682831,
      "grad_norm": 0.1815904974937439,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 5823
    },
    {
      "epoch": 0.9541284403669725,
      "grad_norm": 0.1855655014514923,
      "learning_rate": 0.0001,
      "loss": 1.223,
      "step": 5824
    },
    {
      "epoch": 0.9542922673656619,
      "grad_norm": 0.19176451861858368,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 5825
    },
    {
      "epoch": 0.9544560943643512,
      "grad_norm": 0.18913203477859497,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 5826
    },
    {
      "epoch": 0.9546199213630406,
      "grad_norm": 0.1790342926979065,
      "learning_rate": 0.0001,
      "loss": 1.4001,
      "step": 5827
    },
    {
      "epoch": 0.95478374836173,
      "grad_norm": 0.1837921142578125,
      "learning_rate": 0.0001,
      "loss": 1.2997,
      "step": 5828
    },
    {
      "epoch": 0.9549475753604194,
      "grad_norm": 0.1829642355442047,
      "learning_rate": 0.0001,
      "loss": 1.379,
      "step": 5829
    },
    {
      "epoch": 0.9551114023591087,
      "grad_norm": 0.18498149514198303,
      "learning_rate": 0.0001,
      "loss": 1.2507,
      "step": 5830
    },
    {
      "epoch": 0.9552752293577982,
      "grad_norm": 0.18816125392913818,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 5831
    },
    {
      "epoch": 0.9554390563564875,
      "grad_norm": 0.17867796123027802,
      "learning_rate": 0.0001,
      "loss": 1.2983,
      "step": 5832
    },
    {
      "epoch": 0.955602883355177,
      "grad_norm": 0.18918296694755554,
      "learning_rate": 0.0001,
      "loss": 1.3401,
      "step": 5833
    },
    {
      "epoch": 0.9557667103538663,
      "grad_norm": 0.18241223692893982,
      "learning_rate": 0.0001,
      "loss": 1.3581,
      "step": 5834
    },
    {
      "epoch": 0.9559305373525557,
      "grad_norm": 0.182402566075325,
      "learning_rate": 0.0001,
      "loss": 1.1632,
      "step": 5835
    },
    {
      "epoch": 0.9560943643512451,
      "grad_norm": 0.1755887269973755,
      "learning_rate": 0.0001,
      "loss": 1.2251,
      "step": 5836
    },
    {
      "epoch": 0.9562581913499345,
      "grad_norm": 0.19440293312072754,
      "learning_rate": 0.0001,
      "loss": 1.3361,
      "step": 5837
    },
    {
      "epoch": 0.9564220183486238,
      "grad_norm": 0.19407479465007782,
      "learning_rate": 0.0001,
      "loss": 1.314,
      "step": 5838
    },
    {
      "epoch": 0.9565858453473133,
      "grad_norm": 0.18544656038284302,
      "learning_rate": 0.0001,
      "loss": 1.3283,
      "step": 5839
    },
    {
      "epoch": 0.9567496723460026,
      "grad_norm": 0.19127807021141052,
      "learning_rate": 0.0001,
      "loss": 1.3464,
      "step": 5840
    },
    {
      "epoch": 0.956913499344692,
      "grad_norm": 0.17968320846557617,
      "learning_rate": 0.0001,
      "loss": 1.308,
      "step": 5841
    },
    {
      "epoch": 0.9570773263433814,
      "grad_norm": 0.18177743256092072,
      "learning_rate": 0.0001,
      "loss": 1.2475,
      "step": 5842
    },
    {
      "epoch": 0.9572411533420708,
      "grad_norm": 0.19056715071201324,
      "learning_rate": 0.0001,
      "loss": 1.2722,
      "step": 5843
    },
    {
      "epoch": 0.9574049803407602,
      "grad_norm": 0.19105693697929382,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 5844
    },
    {
      "epoch": 0.9575688073394495,
      "grad_norm": 0.17830684781074524,
      "learning_rate": 0.0001,
      "loss": 1.2298,
      "step": 5845
    },
    {
      "epoch": 0.9577326343381389,
      "grad_norm": 0.1905314028263092,
      "learning_rate": 0.0001,
      "loss": 1.3515,
      "step": 5846
    },
    {
      "epoch": 0.9578964613368283,
      "grad_norm": 0.18119774758815765,
      "learning_rate": 0.0001,
      "loss": 1.2693,
      "step": 5847
    },
    {
      "epoch": 0.9580602883355177,
      "grad_norm": 0.18602688610553741,
      "learning_rate": 0.0001,
      "loss": 1.3454,
      "step": 5848
    },
    {
      "epoch": 0.958224115334207,
      "grad_norm": 0.1867159605026245,
      "learning_rate": 0.0001,
      "loss": 1.2723,
      "step": 5849
    },
    {
      "epoch": 0.9583879423328965,
      "grad_norm": 0.18529576063156128,
      "learning_rate": 0.0001,
      "loss": 1.2729,
      "step": 5850
    },
    {
      "epoch": 0.9585517693315858,
      "grad_norm": 0.18321257829666138,
      "learning_rate": 0.0001,
      "loss": 1.3576,
      "step": 5851
    },
    {
      "epoch": 0.9587155963302753,
      "grad_norm": 0.18920964002609253,
      "learning_rate": 0.0001,
      "loss": 1.2519,
      "step": 5852
    },
    {
      "epoch": 0.9588794233289646,
      "grad_norm": 0.19373340904712677,
      "learning_rate": 0.0001,
      "loss": 1.3953,
      "step": 5853
    },
    {
      "epoch": 0.959043250327654,
      "grad_norm": 0.18377049267292023,
      "learning_rate": 0.0001,
      "loss": 1.298,
      "step": 5854
    },
    {
      "epoch": 0.9592070773263434,
      "grad_norm": 0.1894654631614685,
      "learning_rate": 0.0001,
      "loss": 1.2903,
      "step": 5855
    },
    {
      "epoch": 0.9593709043250328,
      "grad_norm": 0.1838942915201187,
      "learning_rate": 0.0001,
      "loss": 1.3325,
      "step": 5856
    },
    {
      "epoch": 0.9595347313237221,
      "grad_norm": 0.1748867779970169,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 5857
    },
    {
      "epoch": 0.9596985583224116,
      "grad_norm": 0.18629443645477295,
      "learning_rate": 0.0001,
      "loss": 1.2464,
      "step": 5858
    },
    {
      "epoch": 0.9598623853211009,
      "grad_norm": 0.19271212816238403,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 5859
    },
    {
      "epoch": 0.9600262123197904,
      "grad_norm": 0.19744807481765747,
      "learning_rate": 0.0001,
      "loss": 1.271,
      "step": 5860
    },
    {
      "epoch": 0.9601900393184797,
      "grad_norm": 0.18668554723262787,
      "learning_rate": 0.0001,
      "loss": 1.4072,
      "step": 5861
    },
    {
      "epoch": 0.9603538663171691,
      "grad_norm": 0.1952923834323883,
      "learning_rate": 0.0001,
      "loss": 1.198,
      "step": 5862
    },
    {
      "epoch": 0.9605176933158585,
      "grad_norm": 0.1911560744047165,
      "learning_rate": 0.0001,
      "loss": 1.2948,
      "step": 5863
    },
    {
      "epoch": 0.9606815203145478,
      "grad_norm": 0.17945560812950134,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 5864
    },
    {
      "epoch": 0.9608453473132372,
      "grad_norm": 0.18953195214271545,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 5865
    },
    {
      "epoch": 0.9610091743119266,
      "grad_norm": 0.1803627461194992,
      "learning_rate": 0.0001,
      "loss": 1.3215,
      "step": 5866
    },
    {
      "epoch": 0.961173001310616,
      "grad_norm": 0.18751271069049835,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 5867
    },
    {
      "epoch": 0.9613368283093053,
      "grad_norm": 0.19175055623054504,
      "learning_rate": 0.0001,
      "loss": 1.2834,
      "step": 5868
    },
    {
      "epoch": 0.9615006553079948,
      "grad_norm": 0.18315236270427704,
      "learning_rate": 0.0001,
      "loss": 1.3287,
      "step": 5869
    },
    {
      "epoch": 0.9616644823066841,
      "grad_norm": 0.18544581532478333,
      "learning_rate": 0.0001,
      "loss": 1.2812,
      "step": 5870
    },
    {
      "epoch": 0.9618283093053736,
      "grad_norm": 0.1934356987476349,
      "learning_rate": 0.0001,
      "loss": 1.2125,
      "step": 5871
    },
    {
      "epoch": 0.9619921363040629,
      "grad_norm": 0.1997280865907669,
      "learning_rate": 0.0001,
      "loss": 1.4137,
      "step": 5872
    },
    {
      "epoch": 0.9621559633027523,
      "grad_norm": 0.18194052577018738,
      "learning_rate": 0.0001,
      "loss": 1.3771,
      "step": 5873
    },
    {
      "epoch": 0.9623197903014417,
      "grad_norm": 0.1796754002571106,
      "learning_rate": 0.0001,
      "loss": 1.2338,
      "step": 5874
    },
    {
      "epoch": 0.9624836173001311,
      "grad_norm": 0.191252201795578,
      "learning_rate": 0.0001,
      "loss": 1.2676,
      "step": 5875
    },
    {
      "epoch": 0.9626474442988204,
      "grad_norm": 0.18382108211517334,
      "learning_rate": 0.0001,
      "loss": 1.4261,
      "step": 5876
    },
    {
      "epoch": 0.9628112712975099,
      "grad_norm": 0.1804894655942917,
      "learning_rate": 0.0001,
      "loss": 1.4318,
      "step": 5877
    },
    {
      "epoch": 0.9629750982961992,
      "grad_norm": 0.1857917159795761,
      "learning_rate": 0.0001,
      "loss": 1.2666,
      "step": 5878
    },
    {
      "epoch": 0.9631389252948886,
      "grad_norm": 0.18951931595802307,
      "learning_rate": 0.0001,
      "loss": 1.4512,
      "step": 5879
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 0.17923428118228912,
      "learning_rate": 0.0001,
      "loss": 1.3444,
      "step": 5880
    },
    {
      "epoch": 0.9634665792922673,
      "grad_norm": 0.18020100891590118,
      "learning_rate": 0.0001,
      "loss": 1.316,
      "step": 5881
    },
    {
      "epoch": 0.9636304062909568,
      "grad_norm": 0.18839894235134125,
      "learning_rate": 0.0001,
      "loss": 1.2238,
      "step": 5882
    },
    {
      "epoch": 0.9637942332896461,
      "grad_norm": 0.1891549676656723,
      "learning_rate": 0.0001,
      "loss": 1.3306,
      "step": 5883
    },
    {
      "epoch": 0.9639580602883355,
      "grad_norm": 0.18917536735534668,
      "learning_rate": 0.0001,
      "loss": 1.4323,
      "step": 5884
    },
    {
      "epoch": 0.9641218872870249,
      "grad_norm": 0.18194511532783508,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 5885
    },
    {
      "epoch": 0.9642857142857143,
      "grad_norm": 0.1808996945619583,
      "learning_rate": 0.0001,
      "loss": 1.3245,
      "step": 5886
    },
    {
      "epoch": 0.9644495412844036,
      "grad_norm": 0.18231524527072906,
      "learning_rate": 0.0001,
      "loss": 1.3376,
      "step": 5887
    },
    {
      "epoch": 0.9646133682830931,
      "grad_norm": 0.17767439782619476,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 5888
    },
    {
      "epoch": 0.9647771952817824,
      "grad_norm": 0.17880533635616302,
      "learning_rate": 0.0001,
      "loss": 1.2887,
      "step": 5889
    },
    {
      "epoch": 0.9649410222804718,
      "grad_norm": 0.18566380441188812,
      "learning_rate": 0.0001,
      "loss": 1.3091,
      "step": 5890
    },
    {
      "epoch": 0.9651048492791612,
      "grad_norm": 0.17258523404598236,
      "learning_rate": 0.0001,
      "loss": 1.3321,
      "step": 5891
    },
    {
      "epoch": 0.9652686762778506,
      "grad_norm": 0.1792377233505249,
      "learning_rate": 0.0001,
      "loss": 1.3606,
      "step": 5892
    },
    {
      "epoch": 0.96543250327654,
      "grad_norm": 0.18176598846912384,
      "learning_rate": 0.0001,
      "loss": 1.326,
      "step": 5893
    },
    {
      "epoch": 0.9655963302752294,
      "grad_norm": 0.19055069983005524,
      "learning_rate": 0.0001,
      "loss": 1.268,
      "step": 5894
    },
    {
      "epoch": 0.9657601572739187,
      "grad_norm": 0.18240439891815186,
      "learning_rate": 0.0001,
      "loss": 1.3203,
      "step": 5895
    },
    {
      "epoch": 0.9659239842726082,
      "grad_norm": 0.18444737792015076,
      "learning_rate": 0.0001,
      "loss": 1.4011,
      "step": 5896
    },
    {
      "epoch": 0.9660878112712975,
      "grad_norm": 0.18626035749912262,
      "learning_rate": 0.0001,
      "loss": 1.398,
      "step": 5897
    },
    {
      "epoch": 0.966251638269987,
      "grad_norm": 0.18684779107570648,
      "learning_rate": 0.0001,
      "loss": 1.3471,
      "step": 5898
    },
    {
      "epoch": 0.9664154652686763,
      "grad_norm": 0.18379704654216766,
      "learning_rate": 0.0001,
      "loss": 1.3872,
      "step": 5899
    },
    {
      "epoch": 0.9665792922673656,
      "grad_norm": 0.18797656893730164,
      "learning_rate": 0.0001,
      "loss": 1.346,
      "step": 5900
    },
    {
      "epoch": 0.966743119266055,
      "grad_norm": 0.1836424171924591,
      "learning_rate": 0.0001,
      "loss": 1.2182,
      "step": 5901
    },
    {
      "epoch": 0.9669069462647444,
      "grad_norm": 0.1992177814245224,
      "learning_rate": 0.0001,
      "loss": 1.4136,
      "step": 5902
    },
    {
      "epoch": 0.9670707732634338,
      "grad_norm": 0.18224342167377472,
      "learning_rate": 0.0001,
      "loss": 1.3435,
      "step": 5903
    },
    {
      "epoch": 0.9672346002621232,
      "grad_norm": 0.18377435207366943,
      "learning_rate": 0.0001,
      "loss": 1.359,
      "step": 5904
    },
    {
      "epoch": 0.9673984272608126,
      "grad_norm": 0.18519841134548187,
      "learning_rate": 0.0001,
      "loss": 1.4432,
      "step": 5905
    },
    {
      "epoch": 0.9675622542595019,
      "grad_norm": 0.18603628873825073,
      "learning_rate": 0.0001,
      "loss": 1.4077,
      "step": 5906
    },
    {
      "epoch": 0.9677260812581914,
      "grad_norm": 0.17655418813228607,
      "learning_rate": 0.0001,
      "loss": 1.2695,
      "step": 5907
    },
    {
      "epoch": 0.9678899082568807,
      "grad_norm": 0.1848190277814865,
      "learning_rate": 0.0001,
      "loss": 1.3176,
      "step": 5908
    },
    {
      "epoch": 0.9680537352555701,
      "grad_norm": 0.18395964801311493,
      "learning_rate": 0.0001,
      "loss": 1.2349,
      "step": 5909
    },
    {
      "epoch": 0.9682175622542595,
      "grad_norm": 0.1768827736377716,
      "learning_rate": 0.0001,
      "loss": 1.2688,
      "step": 5910
    },
    {
      "epoch": 0.9683813892529489,
      "grad_norm": 0.18721528351306915,
      "learning_rate": 0.0001,
      "loss": 1.2938,
      "step": 5911
    },
    {
      "epoch": 0.9685452162516383,
      "grad_norm": 0.18426834046840668,
      "learning_rate": 0.0001,
      "loss": 1.3434,
      "step": 5912
    },
    {
      "epoch": 0.9687090432503277,
      "grad_norm": 0.1923646628856659,
      "learning_rate": 0.0001,
      "loss": 1.271,
      "step": 5913
    },
    {
      "epoch": 0.968872870249017,
      "grad_norm": 0.1931050717830658,
      "learning_rate": 0.0001,
      "loss": 1.3889,
      "step": 5914
    },
    {
      "epoch": 0.9690366972477065,
      "grad_norm": 0.18071426451206207,
      "learning_rate": 0.0001,
      "loss": 1.3406,
      "step": 5915
    },
    {
      "epoch": 0.9692005242463958,
      "grad_norm": 0.1839054375886917,
      "learning_rate": 0.0001,
      "loss": 1.3079,
      "step": 5916
    },
    {
      "epoch": 0.9693643512450852,
      "grad_norm": 0.18530108034610748,
      "learning_rate": 0.0001,
      "loss": 1.3632,
      "step": 5917
    },
    {
      "epoch": 0.9695281782437746,
      "grad_norm": 0.17874929308891296,
      "learning_rate": 0.0001,
      "loss": 1.3263,
      "step": 5918
    },
    {
      "epoch": 0.9696920052424639,
      "grad_norm": 0.1821708083152771,
      "learning_rate": 0.0001,
      "loss": 1.3635,
      "step": 5919
    },
    {
      "epoch": 0.9698558322411533,
      "grad_norm": 0.17341230809688568,
      "learning_rate": 0.0001,
      "loss": 1.277,
      "step": 5920
    },
    {
      "epoch": 0.9700196592398427,
      "grad_norm": 0.19591085612773895,
      "learning_rate": 0.0001,
      "loss": 1.3416,
      "step": 5921
    },
    {
      "epoch": 0.9701834862385321,
      "grad_norm": 0.18834348022937775,
      "learning_rate": 0.0001,
      "loss": 1.3362,
      "step": 5922
    },
    {
      "epoch": 0.9703473132372215,
      "grad_norm": 0.18517015874385834,
      "learning_rate": 0.0001,
      "loss": 1.3682,
      "step": 5923
    },
    {
      "epoch": 0.9705111402359109,
      "grad_norm": 0.21794269979000092,
      "learning_rate": 0.0001,
      "loss": 1.3722,
      "step": 5924
    },
    {
      "epoch": 0.9706749672346002,
      "grad_norm": 0.2120686024427414,
      "learning_rate": 0.0001,
      "loss": 1.2956,
      "step": 5925
    },
    {
      "epoch": 0.9708387942332897,
      "grad_norm": 0.192914679646492,
      "learning_rate": 0.0001,
      "loss": 1.4002,
      "step": 5926
    },
    {
      "epoch": 0.971002621231979,
      "grad_norm": 0.17589807510375977,
      "learning_rate": 0.0001,
      "loss": 1.3901,
      "step": 5927
    },
    {
      "epoch": 0.9711664482306684,
      "grad_norm": 0.1763237565755844,
      "learning_rate": 0.0001,
      "loss": 1.2327,
      "step": 5928
    },
    {
      "epoch": 0.9713302752293578,
      "grad_norm": 0.18063847720623016,
      "learning_rate": 0.0001,
      "loss": 1.3344,
      "step": 5929
    },
    {
      "epoch": 0.9714941022280472,
      "grad_norm": 0.18630437552928925,
      "learning_rate": 0.0001,
      "loss": 1.3314,
      "step": 5930
    },
    {
      "epoch": 0.9716579292267365,
      "grad_norm": 0.18185073137283325,
      "learning_rate": 0.0001,
      "loss": 1.3069,
      "step": 5931
    },
    {
      "epoch": 0.971821756225426,
      "grad_norm": 0.1926363855600357,
      "learning_rate": 0.0001,
      "loss": 1.2905,
      "step": 5932
    },
    {
      "epoch": 0.9719855832241153,
      "grad_norm": 0.18470226228237152,
      "learning_rate": 0.0001,
      "loss": 1.3024,
      "step": 5933
    },
    {
      "epoch": 0.9721494102228048,
      "grad_norm": 0.19182080030441284,
      "learning_rate": 0.0001,
      "loss": 1.2532,
      "step": 5934
    },
    {
      "epoch": 0.9723132372214941,
      "grad_norm": 0.18238483369350433,
      "learning_rate": 0.0001,
      "loss": 1.4054,
      "step": 5935
    },
    {
      "epoch": 0.9724770642201835,
      "grad_norm": 0.18444161117076874,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 5936
    },
    {
      "epoch": 0.9726408912188729,
      "grad_norm": 0.17651879787445068,
      "learning_rate": 0.0001,
      "loss": 1.3288,
      "step": 5937
    },
    {
      "epoch": 0.9728047182175622,
      "grad_norm": 0.17311865091323853,
      "learning_rate": 0.0001,
      "loss": 1.3706,
      "step": 5938
    },
    {
      "epoch": 0.9729685452162516,
      "grad_norm": 0.17520973086357117,
      "learning_rate": 0.0001,
      "loss": 1.2717,
      "step": 5939
    },
    {
      "epoch": 0.973132372214941,
      "grad_norm": 0.1881515234708786,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 5940
    },
    {
      "epoch": 0.9732961992136304,
      "grad_norm": 0.18889689445495605,
      "learning_rate": 0.0001,
      "loss": 1.4189,
      "step": 5941
    },
    {
      "epoch": 0.9734600262123198,
      "grad_norm": 0.18491891026496887,
      "learning_rate": 0.0001,
      "loss": 1.3615,
      "step": 5942
    },
    {
      "epoch": 0.9736238532110092,
      "grad_norm": 0.19322897493839264,
      "learning_rate": 0.0001,
      "loss": 1.4039,
      "step": 5943
    },
    {
      "epoch": 0.9737876802096985,
      "grad_norm": 0.18553033471107483,
      "learning_rate": 0.0001,
      "loss": 1.2781,
      "step": 5944
    },
    {
      "epoch": 0.973951507208388,
      "grad_norm": 0.19603769481182098,
      "learning_rate": 0.0001,
      "loss": 1.2391,
      "step": 5945
    },
    {
      "epoch": 0.9741153342070773,
      "grad_norm": 0.1881067305803299,
      "learning_rate": 0.0001,
      "loss": 1.3746,
      "step": 5946
    },
    {
      "epoch": 0.9742791612057667,
      "grad_norm": 0.19265957176685333,
      "learning_rate": 0.0001,
      "loss": 1.2886,
      "step": 5947
    },
    {
      "epoch": 0.9744429882044561,
      "grad_norm": 0.22347135841846466,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 5948
    },
    {
      "epoch": 0.9746068152031455,
      "grad_norm": 0.1881948709487915,
      "learning_rate": 0.0001,
      "loss": 1.2916,
      "step": 5949
    },
    {
      "epoch": 0.9747706422018348,
      "grad_norm": 0.1789560317993164,
      "learning_rate": 0.0001,
      "loss": 1.3499,
      "step": 5950
    },
    {
      "epoch": 0.9749344692005243,
      "grad_norm": 0.18851500749588013,
      "learning_rate": 0.0001,
      "loss": 1.3897,
      "step": 5951
    },
    {
      "epoch": 0.9750982961992136,
      "grad_norm": 0.19336290657520294,
      "learning_rate": 0.0001,
      "loss": 1.3713,
      "step": 5952
    },
    {
      "epoch": 0.9752621231979031,
      "grad_norm": 0.18682004511356354,
      "learning_rate": 0.0001,
      "loss": 1.3041,
      "step": 5953
    },
    {
      "epoch": 0.9754259501965924,
      "grad_norm": 0.18698811531066895,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 5954
    },
    {
      "epoch": 0.9755897771952818,
      "grad_norm": 0.18383051455020905,
      "learning_rate": 0.0001,
      "loss": 1.3211,
      "step": 5955
    },
    {
      "epoch": 0.9757536041939712,
      "grad_norm": 0.1905040144920349,
      "learning_rate": 0.0001,
      "loss": 1.3116,
      "step": 5956
    },
    {
      "epoch": 0.9759174311926605,
      "grad_norm": 0.19765515625476837,
      "learning_rate": 0.0001,
      "loss": 1.3657,
      "step": 5957
    },
    {
      "epoch": 0.9760812581913499,
      "grad_norm": 0.18778710067272186,
      "learning_rate": 0.0001,
      "loss": 1.4398,
      "step": 5958
    },
    {
      "epoch": 0.9762450851900393,
      "grad_norm": 0.19105587899684906,
      "learning_rate": 0.0001,
      "loss": 1.3967,
      "step": 5959
    },
    {
      "epoch": 0.9764089121887287,
      "grad_norm": 0.182808980345726,
      "learning_rate": 0.0001,
      "loss": 1.3442,
      "step": 5960
    },
    {
      "epoch": 0.976572739187418,
      "grad_norm": 0.18550336360931396,
      "learning_rate": 0.0001,
      "loss": 1.3176,
      "step": 5961
    },
    {
      "epoch": 0.9767365661861075,
      "grad_norm": 0.191632479429245,
      "learning_rate": 0.0001,
      "loss": 1.2793,
      "step": 5962
    },
    {
      "epoch": 0.9769003931847968,
      "grad_norm": 0.1868181824684143,
      "learning_rate": 0.0001,
      "loss": 1.2869,
      "step": 5963
    },
    {
      "epoch": 0.9770642201834863,
      "grad_norm": 0.18379037082195282,
      "learning_rate": 0.0001,
      "loss": 1.4052,
      "step": 5964
    },
    {
      "epoch": 0.9772280471821756,
      "grad_norm": 0.1819850504398346,
      "learning_rate": 0.0001,
      "loss": 1.43,
      "step": 5965
    },
    {
      "epoch": 0.977391874180865,
      "grad_norm": 0.1858784705400467,
      "learning_rate": 0.0001,
      "loss": 1.2428,
      "step": 5966
    },
    {
      "epoch": 0.9775557011795544,
      "grad_norm": 0.1911894977092743,
      "learning_rate": 0.0001,
      "loss": 1.3832,
      "step": 5967
    },
    {
      "epoch": 0.9777195281782438,
      "grad_norm": 0.1944715827703476,
      "learning_rate": 0.0001,
      "loss": 1.3719,
      "step": 5968
    },
    {
      "epoch": 0.9778833551769331,
      "grad_norm": 0.18842533230781555,
      "learning_rate": 0.0001,
      "loss": 1.2722,
      "step": 5969
    },
    {
      "epoch": 0.9780471821756226,
      "grad_norm": 0.18261021375656128,
      "learning_rate": 0.0001,
      "loss": 1.3152,
      "step": 5970
    },
    {
      "epoch": 0.9782110091743119,
      "grad_norm": 0.19497132301330566,
      "learning_rate": 0.0001,
      "loss": 1.2887,
      "step": 5971
    },
    {
      "epoch": 0.9783748361730014,
      "grad_norm": 0.17155678570270538,
      "learning_rate": 0.0001,
      "loss": 1.2325,
      "step": 5972
    },
    {
      "epoch": 0.9785386631716907,
      "grad_norm": 0.1748793125152588,
      "learning_rate": 0.0001,
      "loss": 1.2619,
      "step": 5973
    },
    {
      "epoch": 0.97870249017038,
      "grad_norm": 0.1979513019323349,
      "learning_rate": 0.0001,
      "loss": 1.1758,
      "step": 5974
    },
    {
      "epoch": 0.9788663171690695,
      "grad_norm": 0.17970775067806244,
      "learning_rate": 0.0001,
      "loss": 1.2837,
      "step": 5975
    },
    {
      "epoch": 0.9790301441677588,
      "grad_norm": 0.19129396975040436,
      "learning_rate": 0.0001,
      "loss": 1.3297,
      "step": 5976
    },
    {
      "epoch": 0.9791939711664482,
      "grad_norm": 0.19276297092437744,
      "learning_rate": 0.0001,
      "loss": 1.4137,
      "step": 5977
    },
    {
      "epoch": 0.9793577981651376,
      "grad_norm": 0.1848108172416687,
      "learning_rate": 0.0001,
      "loss": 1.3113,
      "step": 5978
    },
    {
      "epoch": 0.979521625163827,
      "grad_norm": 0.18232189118862152,
      "learning_rate": 0.0001,
      "loss": 1.1992,
      "step": 5979
    },
    {
      "epoch": 0.9796854521625163,
      "grad_norm": 0.19306521117687225,
      "learning_rate": 0.0001,
      "loss": 1.3741,
      "step": 5980
    },
    {
      "epoch": 0.9798492791612058,
      "grad_norm": 0.1795937716960907,
      "learning_rate": 0.0001,
      "loss": 1.2956,
      "step": 5981
    },
    {
      "epoch": 0.9800131061598951,
      "grad_norm": 0.19006799161434174,
      "learning_rate": 0.0001,
      "loss": 1.2805,
      "step": 5982
    },
    {
      "epoch": 0.9801769331585846,
      "grad_norm": 0.18953931331634521,
      "learning_rate": 0.0001,
      "loss": 1.3731,
      "step": 5983
    },
    {
      "epoch": 0.9803407601572739,
      "grad_norm": 0.18368157744407654,
      "learning_rate": 0.0001,
      "loss": 1.3028,
      "step": 5984
    },
    {
      "epoch": 0.9805045871559633,
      "grad_norm": 0.19864521920681,
      "learning_rate": 0.0001,
      "loss": 1.2395,
      "step": 5985
    },
    {
      "epoch": 0.9806684141546527,
      "grad_norm": 0.19959114491939545,
      "learning_rate": 0.0001,
      "loss": 1.3037,
      "step": 5986
    },
    {
      "epoch": 0.9808322411533421,
      "grad_norm": 0.19988304376602173,
      "learning_rate": 0.0001,
      "loss": 1.3176,
      "step": 5987
    },
    {
      "epoch": 0.9809960681520314,
      "grad_norm": 0.174989715218544,
      "learning_rate": 0.0001,
      "loss": 1.2665,
      "step": 5988
    },
    {
      "epoch": 0.9811598951507209,
      "grad_norm": 0.18232625722885132,
      "learning_rate": 0.0001,
      "loss": 1.3556,
      "step": 5989
    },
    {
      "epoch": 0.9813237221494102,
      "grad_norm": 0.18446971476078033,
      "learning_rate": 0.0001,
      "loss": 1.3225,
      "step": 5990
    },
    {
      "epoch": 0.9814875491480997,
      "grad_norm": 0.1953752487897873,
      "learning_rate": 0.0001,
      "loss": 1.3499,
      "step": 5991
    },
    {
      "epoch": 0.981651376146789,
      "grad_norm": 0.1931132823228836,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 5992
    },
    {
      "epoch": 0.9818152031454783,
      "grad_norm": 0.1791372150182724,
      "learning_rate": 0.0001,
      "loss": 1.3628,
      "step": 5993
    },
    {
      "epoch": 0.9819790301441678,
      "grad_norm": 0.17971336841583252,
      "learning_rate": 0.0001,
      "loss": 1.3418,
      "step": 5994
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 0.18503516912460327,
      "learning_rate": 0.0001,
      "loss": 1.3512,
      "step": 5995
    },
    {
      "epoch": 0.9823066841415465,
      "grad_norm": 0.18169750273227692,
      "learning_rate": 0.0001,
      "loss": 1.3372,
      "step": 5996
    },
    {
      "epoch": 0.9824705111402359,
      "grad_norm": 0.20179560780525208,
      "learning_rate": 0.0001,
      "loss": 1.2948,
      "step": 5997
    },
    {
      "epoch": 0.9826343381389253,
      "grad_norm": 0.18052950501441956,
      "learning_rate": 0.0001,
      "loss": 1.4604,
      "step": 5998
    },
    {
      "epoch": 0.9827981651376146,
      "grad_norm": 0.21019285917282104,
      "learning_rate": 0.0001,
      "loss": 1.3246,
      "step": 5999
    },
    {
      "epoch": 0.9829619921363041,
      "grad_norm": 0.18748429417610168,
      "learning_rate": 0.0001,
      "loss": 1.3295,
      "step": 6000
    },
    {
      "epoch": 0.9831258191349934,
      "grad_norm": 0.2004801332950592,
      "learning_rate": 0.0001,
      "loss": 1.3319,
      "step": 6001
    },
    {
      "epoch": 0.9832896461336829,
      "grad_norm": 0.18769106268882751,
      "learning_rate": 0.0001,
      "loss": 1.3593,
      "step": 6002
    },
    {
      "epoch": 0.9834534731323722,
      "grad_norm": 0.19437390565872192,
      "learning_rate": 0.0001,
      "loss": 1.3467,
      "step": 6003
    },
    {
      "epoch": 0.9836173001310616,
      "grad_norm": 0.1852758824825287,
      "learning_rate": 0.0001,
      "loss": 1.2794,
      "step": 6004
    },
    {
      "epoch": 0.983781127129751,
      "grad_norm": 0.18648913502693176,
      "learning_rate": 0.0001,
      "loss": 1.1828,
      "step": 6005
    },
    {
      "epoch": 0.9839449541284404,
      "grad_norm": 0.1961841732263565,
      "learning_rate": 0.0001,
      "loss": 1.4056,
      "step": 6006
    },
    {
      "epoch": 0.9841087811271297,
      "grad_norm": 0.18835961818695068,
      "learning_rate": 0.0001,
      "loss": 1.3706,
      "step": 6007
    },
    {
      "epoch": 0.9842726081258192,
      "grad_norm": 0.1918230950832367,
      "learning_rate": 0.0001,
      "loss": 1.495,
      "step": 6008
    },
    {
      "epoch": 0.9844364351245085,
      "grad_norm": 0.1900627762079239,
      "learning_rate": 0.0001,
      "loss": 1.411,
      "step": 6009
    },
    {
      "epoch": 0.984600262123198,
      "grad_norm": 0.18386685848236084,
      "learning_rate": 0.0001,
      "loss": 1.3212,
      "step": 6010
    },
    {
      "epoch": 0.9847640891218873,
      "grad_norm": 0.18313123285770416,
      "learning_rate": 0.0001,
      "loss": 1.2495,
      "step": 6011
    },
    {
      "epoch": 0.9849279161205766,
      "grad_norm": 0.19014672935009003,
      "learning_rate": 0.0001,
      "loss": 1.3539,
      "step": 6012
    },
    {
      "epoch": 0.9850917431192661,
      "grad_norm": 0.1812426745891571,
      "learning_rate": 0.0001,
      "loss": 1.3374,
      "step": 6013
    },
    {
      "epoch": 0.9852555701179554,
      "grad_norm": 0.19462090730667114,
      "learning_rate": 0.0001,
      "loss": 1.2527,
      "step": 6014
    },
    {
      "epoch": 0.9854193971166448,
      "grad_norm": 0.18502458930015564,
      "learning_rate": 0.0001,
      "loss": 1.2971,
      "step": 6015
    },
    {
      "epoch": 0.9855832241153342,
      "grad_norm": 0.17847904562950134,
      "learning_rate": 0.0001,
      "loss": 1.2246,
      "step": 6016
    },
    {
      "epoch": 0.9857470511140236,
      "grad_norm": 0.19191966950893402,
      "learning_rate": 0.0001,
      "loss": 1.3323,
      "step": 6017
    },
    {
      "epoch": 0.9859108781127129,
      "grad_norm": 0.18734365701675415,
      "learning_rate": 0.0001,
      "loss": 1.3663,
      "step": 6018
    },
    {
      "epoch": 0.9860747051114024,
      "grad_norm": 0.17877177894115448,
      "learning_rate": 0.0001,
      "loss": 1.2805,
      "step": 6019
    },
    {
      "epoch": 0.9862385321100917,
      "grad_norm": 0.19008183479309082,
      "learning_rate": 0.0001,
      "loss": 1.2775,
      "step": 6020
    },
    {
      "epoch": 0.9864023591087812,
      "grad_norm": 0.18758349120616913,
      "learning_rate": 0.0001,
      "loss": 1.3359,
      "step": 6021
    },
    {
      "epoch": 0.9865661861074705,
      "grad_norm": 0.18526344001293182,
      "learning_rate": 0.0001,
      "loss": 1.2935,
      "step": 6022
    },
    {
      "epoch": 0.9867300131061599,
      "grad_norm": 0.1888788789510727,
      "learning_rate": 0.0001,
      "loss": 1.3314,
      "step": 6023
    },
    {
      "epoch": 0.9868938401048493,
      "grad_norm": 0.18535082042217255,
      "learning_rate": 0.0001,
      "loss": 1.2207,
      "step": 6024
    },
    {
      "epoch": 0.9870576671035387,
      "grad_norm": 0.18888629972934723,
      "learning_rate": 0.0001,
      "loss": 1.3682,
      "step": 6025
    },
    {
      "epoch": 0.987221494102228,
      "grad_norm": 0.18269053101539612,
      "learning_rate": 0.0001,
      "loss": 1.3589,
      "step": 6026
    },
    {
      "epoch": 0.9873853211009175,
      "grad_norm": 0.20327366888523102,
      "learning_rate": 0.0001,
      "loss": 1.3906,
      "step": 6027
    },
    {
      "epoch": 0.9875491480996068,
      "grad_norm": 0.18929427862167358,
      "learning_rate": 0.0001,
      "loss": 1.4975,
      "step": 6028
    },
    {
      "epoch": 0.9877129750982963,
      "grad_norm": 0.17421098053455353,
      "learning_rate": 0.0001,
      "loss": 1.3607,
      "step": 6029
    },
    {
      "epoch": 0.9878768020969856,
      "grad_norm": 0.18264512717723846,
      "learning_rate": 0.0001,
      "loss": 1.3453,
      "step": 6030
    },
    {
      "epoch": 0.9880406290956749,
      "grad_norm": 0.19131623208522797,
      "learning_rate": 0.0001,
      "loss": 1.3396,
      "step": 6031
    },
    {
      "epoch": 0.9882044560943644,
      "grad_norm": 0.18627583980560303,
      "learning_rate": 0.0001,
      "loss": 1.4217,
      "step": 6032
    },
    {
      "epoch": 0.9883682830930537,
      "grad_norm": 0.18448859453201294,
      "learning_rate": 0.0001,
      "loss": 1.193,
      "step": 6033
    },
    {
      "epoch": 0.9885321100917431,
      "grad_norm": 0.18443851172924042,
      "learning_rate": 0.0001,
      "loss": 1.3899,
      "step": 6034
    },
    {
      "epoch": 0.9886959370904325,
      "grad_norm": 0.182394877076149,
      "learning_rate": 0.0001,
      "loss": 1.3921,
      "step": 6035
    },
    {
      "epoch": 0.9888597640891219,
      "grad_norm": 0.19248344004154205,
      "learning_rate": 0.0001,
      "loss": 1.3537,
      "step": 6036
    },
    {
      "epoch": 0.9890235910878112,
      "grad_norm": 0.19311220943927765,
      "learning_rate": 0.0001,
      "loss": 1.3715,
      "step": 6037
    },
    {
      "epoch": 0.9891874180865007,
      "grad_norm": 0.19365935027599335,
      "learning_rate": 0.0001,
      "loss": 1.3298,
      "step": 6038
    },
    {
      "epoch": 0.98935124508519,
      "grad_norm": 0.18348781764507294,
      "learning_rate": 0.0001,
      "loss": 1.3287,
      "step": 6039
    },
    {
      "epoch": 0.9895150720838795,
      "grad_norm": 0.19196191430091858,
      "learning_rate": 0.0001,
      "loss": 1.4101,
      "step": 6040
    },
    {
      "epoch": 0.9896788990825688,
      "grad_norm": 0.18652039766311646,
      "learning_rate": 0.0001,
      "loss": 1.2337,
      "step": 6041
    },
    {
      "epoch": 0.9898427260812582,
      "grad_norm": 0.1875268965959549,
      "learning_rate": 0.0001,
      "loss": 1.3603,
      "step": 6042
    },
    {
      "epoch": 0.9900065530799476,
      "grad_norm": 0.19804003834724426,
      "learning_rate": 0.0001,
      "loss": 1.28,
      "step": 6043
    },
    {
      "epoch": 0.990170380078637,
      "grad_norm": 0.1947629451751709,
      "learning_rate": 0.0001,
      "loss": 1.3518,
      "step": 6044
    },
    {
      "epoch": 0.9903342070773263,
      "grad_norm": 0.19843094050884247,
      "learning_rate": 0.0001,
      "loss": 1.3162,
      "step": 6045
    },
    {
      "epoch": 0.9904980340760158,
      "grad_norm": 0.19607967138290405,
      "learning_rate": 0.0001,
      "loss": 1.3984,
      "step": 6046
    },
    {
      "epoch": 0.9906618610747051,
      "grad_norm": 0.18945427238941193,
      "learning_rate": 0.0001,
      "loss": 1.2058,
      "step": 6047
    },
    {
      "epoch": 0.9908256880733946,
      "grad_norm": 0.18138481676578522,
      "learning_rate": 0.0001,
      "loss": 1.3277,
      "step": 6048
    },
    {
      "epoch": 0.9909895150720839,
      "grad_norm": 0.19943605363368988,
      "learning_rate": 0.0001,
      "loss": 1.2954,
      "step": 6049
    },
    {
      "epoch": 0.9911533420707732,
      "grad_norm": 0.18126089870929718,
      "learning_rate": 0.0001,
      "loss": 1.2144,
      "step": 6050
    },
    {
      "epoch": 0.9913171690694627,
      "grad_norm": 0.1830243617296219,
      "learning_rate": 0.0001,
      "loss": 1.3317,
      "step": 6051
    },
    {
      "epoch": 0.991480996068152,
      "grad_norm": 0.19686347246170044,
      "learning_rate": 0.0001,
      "loss": 1.2336,
      "step": 6052
    },
    {
      "epoch": 0.9916448230668414,
      "grad_norm": 0.19046474993228912,
      "learning_rate": 0.0001,
      "loss": 1.4775,
      "step": 6053
    },
    {
      "epoch": 0.9918086500655308,
      "grad_norm": 0.19410233199596405,
      "learning_rate": 0.0001,
      "loss": 1.3374,
      "step": 6054
    },
    {
      "epoch": 0.9919724770642202,
      "grad_norm": 0.19389519095420837,
      "learning_rate": 0.0001,
      "loss": 1.3025,
      "step": 6055
    },
    {
      "epoch": 0.9921363040629095,
      "grad_norm": 0.18872575461864471,
      "learning_rate": 0.0001,
      "loss": 1.3484,
      "step": 6056
    },
    {
      "epoch": 0.992300131061599,
      "grad_norm": 0.19611869752407074,
      "learning_rate": 0.0001,
      "loss": 1.3092,
      "step": 6057
    },
    {
      "epoch": 0.9924639580602883,
      "grad_norm": 0.1795845627784729,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 6058
    },
    {
      "epoch": 0.9926277850589778,
      "grad_norm": 0.19747640192508698,
      "learning_rate": 0.0001,
      "loss": 1.3629,
      "step": 6059
    },
    {
      "epoch": 0.9927916120576671,
      "grad_norm": 0.18717457354068756,
      "learning_rate": 0.0001,
      "loss": 1.2993,
      "step": 6060
    },
    {
      "epoch": 0.9929554390563565,
      "grad_norm": 0.1838531196117401,
      "learning_rate": 0.0001,
      "loss": 1.1976,
      "step": 6061
    },
    {
      "epoch": 0.9931192660550459,
      "grad_norm": 0.1945561170578003,
      "learning_rate": 0.0001,
      "loss": 1.2922,
      "step": 6062
    },
    {
      "epoch": 0.9932830930537353,
      "grad_norm": 0.196538507938385,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 6063
    },
    {
      "epoch": 0.9934469200524246,
      "grad_norm": 0.21206152439117432,
      "learning_rate": 0.0001,
      "loss": 1.3757,
      "step": 6064
    },
    {
      "epoch": 0.9936107470511141,
      "grad_norm": 0.1980433464050293,
      "learning_rate": 0.0001,
      "loss": 1.2657,
      "step": 6065
    },
    {
      "epoch": 0.9937745740498034,
      "grad_norm": 0.1909942924976349,
      "learning_rate": 0.0001,
      "loss": 1.4213,
      "step": 6066
    },
    {
      "epoch": 0.9939384010484927,
      "grad_norm": 0.1874045580625534,
      "learning_rate": 0.0001,
      "loss": 1.3448,
      "step": 6067
    },
    {
      "epoch": 0.9941022280471822,
      "grad_norm": 0.18641644716262817,
      "learning_rate": 0.0001,
      "loss": 1.2755,
      "step": 6068
    },
    {
      "epoch": 0.9942660550458715,
      "grad_norm": 0.19908373057842255,
      "learning_rate": 0.0001,
      "loss": 1.3548,
      "step": 6069
    },
    {
      "epoch": 0.994429882044561,
      "grad_norm": 0.19219906628131866,
      "learning_rate": 0.0001,
      "loss": 1.3461,
      "step": 6070
    },
    {
      "epoch": 0.9945937090432503,
      "grad_norm": 0.19654007256031036,
      "learning_rate": 0.0001,
      "loss": 1.2647,
      "step": 6071
    },
    {
      "epoch": 0.9947575360419397,
      "grad_norm": 0.19563758373260498,
      "learning_rate": 0.0001,
      "loss": 1.312,
      "step": 6072
    },
    {
      "epoch": 0.9949213630406291,
      "grad_norm": 0.17976713180541992,
      "learning_rate": 0.0001,
      "loss": 1.3378,
      "step": 6073
    },
    {
      "epoch": 0.9950851900393185,
      "grad_norm": 0.18331022560596466,
      "learning_rate": 0.0001,
      "loss": 1.2518,
      "step": 6074
    },
    {
      "epoch": 0.9952490170380078,
      "grad_norm": 0.19330789148807526,
      "learning_rate": 0.0001,
      "loss": 1.3325,
      "step": 6075
    },
    {
      "epoch": 0.9954128440366973,
      "grad_norm": 0.191721111536026,
      "learning_rate": 0.0001,
      "loss": 1.3953,
      "step": 6076
    },
    {
      "epoch": 0.9955766710353866,
      "grad_norm": 0.1794525384902954,
      "learning_rate": 0.0001,
      "loss": 1.2865,
      "step": 6077
    },
    {
      "epoch": 0.995740498034076,
      "grad_norm": 0.17714717984199524,
      "learning_rate": 0.0001,
      "loss": 1.1819,
      "step": 6078
    },
    {
      "epoch": 0.9959043250327654,
      "grad_norm": 0.20261555910110474,
      "learning_rate": 0.0001,
      "loss": 1.2955,
      "step": 6079
    },
    {
      "epoch": 0.9960681520314548,
      "grad_norm": 0.19919881224632263,
      "learning_rate": 0.0001,
      "loss": 1.4045,
      "step": 6080
    },
    {
      "epoch": 0.9962319790301442,
      "grad_norm": 0.19124296307563782,
      "learning_rate": 0.0001,
      "loss": 1.3087,
      "step": 6081
    },
    {
      "epoch": 0.9963958060288336,
      "grad_norm": 0.1943296641111374,
      "learning_rate": 0.0001,
      "loss": 1.3593,
      "step": 6082
    },
    {
      "epoch": 0.9965596330275229,
      "grad_norm": 0.18235737085342407,
      "learning_rate": 0.0001,
      "loss": 1.2983,
      "step": 6083
    },
    {
      "epoch": 0.9967234600262124,
      "grad_norm": 0.18115395307540894,
      "learning_rate": 0.0001,
      "loss": 1.3254,
      "step": 6084
    },
    {
      "epoch": 0.9968872870249017,
      "grad_norm": 0.18962350487709045,
      "learning_rate": 0.0001,
      "loss": 1.2407,
      "step": 6085
    },
    {
      "epoch": 0.997051114023591,
      "grad_norm": 0.20859608054161072,
      "learning_rate": 0.0001,
      "loss": 1.4172,
      "step": 6086
    },
    {
      "epoch": 0.9972149410222805,
      "grad_norm": 0.19508366286754608,
      "learning_rate": 0.0001,
      "loss": 1.3301,
      "step": 6087
    },
    {
      "epoch": 0.9973787680209698,
      "grad_norm": 0.1864568591117859,
      "learning_rate": 0.0001,
      "loss": 1.2418,
      "step": 6088
    },
    {
      "epoch": 0.9975425950196593,
      "grad_norm": 0.19224658608436584,
      "learning_rate": 0.0001,
      "loss": 1.3691,
      "step": 6089
    },
    {
      "epoch": 0.9977064220183486,
      "grad_norm": 0.1914091408252716,
      "learning_rate": 0.0001,
      "loss": 1.3697,
      "step": 6090
    },
    {
      "epoch": 0.997870249017038,
      "grad_norm": 0.18093369901180267,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 6091
    },
    {
      "epoch": 0.9980340760157274,
      "grad_norm": 0.19533345103263855,
      "learning_rate": 0.0001,
      "loss": 1.3561,
      "step": 6092
    },
    {
      "epoch": 0.9981979030144168,
      "grad_norm": 0.18287213146686554,
      "learning_rate": 0.0001,
      "loss": 1.2517,
      "step": 6093
    },
    {
      "epoch": 0.9983617300131061,
      "grad_norm": 0.19861860573291779,
      "learning_rate": 0.0001,
      "loss": 1.2925,
      "step": 6094
    },
    {
      "epoch": 0.9985255570117956,
      "grad_norm": 0.20534437894821167,
      "learning_rate": 0.0001,
      "loss": 1.4319,
      "step": 6095
    },
    {
      "epoch": 0.9986893840104849,
      "grad_norm": 0.19749534130096436,
      "learning_rate": 0.0001,
      "loss": 1.3645,
      "step": 6096
    },
    {
      "epoch": 0.9988532110091743,
      "grad_norm": 0.18745307624340057,
      "learning_rate": 0.0001,
      "loss": 1.3746,
      "step": 6097
    },
    {
      "epoch": 0.9990170380078637,
      "grad_norm": 0.19253526628017426,
      "learning_rate": 0.0001,
      "loss": 1.3772,
      "step": 6098
    },
    {
      "epoch": 0.9991808650065531,
      "grad_norm": 0.1929803490638733,
      "learning_rate": 0.0001,
      "loss": 1.3256,
      "step": 6099
    },
    {
      "epoch": 0.9993446920052425,
      "grad_norm": 0.1811429262161255,
      "learning_rate": 0.0001,
      "loss": 1.3277,
      "step": 6100
    },
    {
      "epoch": 0.9995085190039319,
      "grad_norm": 0.19386540353298187,
      "learning_rate": 0.0001,
      "loss": 1.3419,
      "step": 6101
    },
    {
      "epoch": 0.9996723460026212,
      "grad_norm": 0.18842317163944244,
      "learning_rate": 0.0001,
      "loss": 1.3896,
      "step": 6102
    },
    {
      "epoch": 0.9998361730013107,
      "grad_norm": 0.19492201507091522,
      "learning_rate": 0.0001,
      "loss": 1.3737,
      "step": 6103
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.18299008905887604,
      "learning_rate": 0.0001,
      "loss": 1.2906,
      "step": 6104
    },
    {
      "epoch": 0.00010726161106939826,
      "grad_norm": 0.1879759579896927,
      "learning_rate": 0.0001,
      "loss": 1.2783,
      "step": 6105
    },
    {
      "epoch": 0.00021452322213879653,
      "grad_norm": 0.2005559355020523,
      "learning_rate": 0.0001,
      "loss": 1.2682,
      "step": 6106
    },
    {
      "epoch": 0.0003217848332081948,
      "grad_norm": 0.17516635358333588,
      "learning_rate": 0.0001,
      "loss": 1.2164,
      "step": 6107
    },
    {
      "epoch": 0.00042904644427759305,
      "grad_norm": 0.196987122297287,
      "learning_rate": 0.0001,
      "loss": 1.2122,
      "step": 6108
    },
    {
      "epoch": 0.0005363080553469913,
      "grad_norm": 0.19166997075080872,
      "learning_rate": 0.0001,
      "loss": 1.3515,
      "step": 6109
    },
    {
      "epoch": 0.0006435696664163896,
      "grad_norm": 0.19768540561199188,
      "learning_rate": 0.0001,
      "loss": 1.2442,
      "step": 6110
    },
    {
      "epoch": 0.0007508312774857878,
      "grad_norm": 0.19529056549072266,
      "learning_rate": 0.0001,
      "loss": 1.2641,
      "step": 6111
    },
    {
      "epoch": 0.0008580928885551861,
      "grad_norm": 0.20722614228725433,
      "learning_rate": 0.0001,
      "loss": 1.3164,
      "step": 6112
    },
    {
      "epoch": 0.0009653544996245844,
      "grad_norm": 0.20644807815551758,
      "learning_rate": 0.0001,
      "loss": 1.2751,
      "step": 6113
    },
    {
      "epoch": 0.0010726161106939826,
      "grad_norm": 0.19213657081127167,
      "learning_rate": 0.0001,
      "loss": 1.2308,
      "step": 6114
    },
    {
      "epoch": 0.0011798777217633808,
      "grad_norm": 0.20487812161445618,
      "learning_rate": 0.0001,
      "loss": 1.1965,
      "step": 6115
    },
    {
      "epoch": 0.0012871393328327792,
      "grad_norm": 0.19622506201267242,
      "learning_rate": 0.0001,
      "loss": 1.2509,
      "step": 6116
    },
    {
      "epoch": 0.0013944009439021774,
      "grad_norm": 0.2001352608203888,
      "learning_rate": 0.0001,
      "loss": 1.2781,
      "step": 6117
    },
    {
      "epoch": 0.0015016625549715756,
      "grad_norm": 0.19143225252628326,
      "learning_rate": 0.0001,
      "loss": 1.2413,
      "step": 6118
    },
    {
      "epoch": 0.001608924166040974,
      "grad_norm": 0.1964908242225647,
      "learning_rate": 0.0001,
      "loss": 1.3068,
      "step": 6119
    },
    {
      "epoch": 0.0017161857771103722,
      "grad_norm": 0.19128486514091492,
      "learning_rate": 0.0001,
      "loss": 1.2241,
      "step": 6120
    },
    {
      "epoch": 0.0018234473881797704,
      "grad_norm": 0.20104964077472687,
      "learning_rate": 0.0001,
      "loss": 1.3094,
      "step": 6121
    },
    {
      "epoch": 0.0019307089992491688,
      "grad_norm": 0.18986660242080688,
      "learning_rate": 0.0001,
      "loss": 1.2174,
      "step": 6122
    },
    {
      "epoch": 0.002037970610318567,
      "grad_norm": 0.1939975917339325,
      "learning_rate": 0.0001,
      "loss": 1.1973,
      "step": 6123
    },
    {
      "epoch": 0.002145232221387965,
      "grad_norm": 0.19705505669116974,
      "learning_rate": 0.0001,
      "loss": 1.2352,
      "step": 6124
    },
    {
      "epoch": 0.0022524938324573634,
      "grad_norm": 0.19101518392562866,
      "learning_rate": 0.0001,
      "loss": 1.3198,
      "step": 6125
    },
    {
      "epoch": 0.0023597554435267616,
      "grad_norm": 0.19130852818489075,
      "learning_rate": 0.0001,
      "loss": 1.3898,
      "step": 6126
    },
    {
      "epoch": 0.0024670170545961602,
      "grad_norm": 0.18564487993717194,
      "learning_rate": 0.0001,
      "loss": 1.2562,
      "step": 6127
    },
    {
      "epoch": 0.0025742786656655584,
      "grad_norm": 0.19588503241539001,
      "learning_rate": 0.0001,
      "loss": 1.4117,
      "step": 6128
    },
    {
      "epoch": 0.0026815402767349566,
      "grad_norm": 0.20332716405391693,
      "learning_rate": 0.0001,
      "loss": 1.2472,
      "step": 6129
    },
    {
      "epoch": 0.002788801887804355,
      "grad_norm": 0.18440887331962585,
      "learning_rate": 0.0001,
      "loss": 1.2076,
      "step": 6130
    },
    {
      "epoch": 0.002896063498873753,
      "grad_norm": 0.1997266411781311,
      "learning_rate": 0.0001,
      "loss": 1.3262,
      "step": 6131
    },
    {
      "epoch": 0.003003325109943151,
      "grad_norm": 0.1905851513147354,
      "learning_rate": 0.0001,
      "loss": 1.2107,
      "step": 6132
    },
    {
      "epoch": 0.0031105867210125494,
      "grad_norm": 0.20031443238258362,
      "learning_rate": 0.0001,
      "loss": 1.3033,
      "step": 6133
    },
    {
      "epoch": 0.003217848332081948,
      "grad_norm": 0.1892753392457962,
      "learning_rate": 0.0001,
      "loss": 1.2108,
      "step": 6134
    },
    {
      "epoch": 0.0033251099431513462,
      "grad_norm": 0.18927855789661407,
      "learning_rate": 0.0001,
      "loss": 1.2891,
      "step": 6135
    },
    {
      "epoch": 0.0034323715542207444,
      "grad_norm": 0.19662487506866455,
      "learning_rate": 0.0001,
      "loss": 1.2588,
      "step": 6136
    },
    {
      "epoch": 0.0035396331652901426,
      "grad_norm": 0.19372087717056274,
      "learning_rate": 0.0001,
      "loss": 1.2734,
      "step": 6137
    },
    {
      "epoch": 0.003646894776359541,
      "grad_norm": 0.19679009914398193,
      "learning_rate": 0.0001,
      "loss": 1.2012,
      "step": 6138
    },
    {
      "epoch": 0.003754156387428939,
      "grad_norm": 0.19291631877422333,
      "learning_rate": 0.0001,
      "loss": 1.3038,
      "step": 6139
    },
    {
      "epoch": 0.0038614179984983376,
      "grad_norm": 0.2021826058626175,
      "learning_rate": 0.0001,
      "loss": 1.3234,
      "step": 6140
    },
    {
      "epoch": 0.003968679609567736,
      "grad_norm": 0.2045031487941742,
      "learning_rate": 0.0001,
      "loss": 1.1869,
      "step": 6141
    },
    {
      "epoch": 0.004075941220637134,
      "grad_norm": 0.20047610998153687,
      "learning_rate": 0.0001,
      "loss": 1.3646,
      "step": 6142
    },
    {
      "epoch": 0.004183202831706532,
      "grad_norm": 0.18222448229789734,
      "learning_rate": 0.0001,
      "loss": 1.255,
      "step": 6143
    },
    {
      "epoch": 0.00429046444277593,
      "grad_norm": 0.20449604094028473,
      "learning_rate": 0.0001,
      "loss": 1.3394,
      "step": 6144
    },
    {
      "epoch": 0.004397726053845329,
      "grad_norm": 0.20431110262870789,
      "learning_rate": 0.0001,
      "loss": 1.275,
      "step": 6145
    },
    {
      "epoch": 0.004504987664914727,
      "grad_norm": 0.19033387303352356,
      "learning_rate": 0.0001,
      "loss": 1.3594,
      "step": 6146
    },
    {
      "epoch": 0.004612249275984125,
      "grad_norm": 0.19687025249004364,
      "learning_rate": 0.0001,
      "loss": 1.3439,
      "step": 6147
    },
    {
      "epoch": 0.004719510887053523,
      "grad_norm": 0.20687568187713623,
      "learning_rate": 0.0001,
      "loss": 1.3514,
      "step": 6148
    },
    {
      "epoch": 0.004826772498122921,
      "grad_norm": 0.1987178921699524,
      "learning_rate": 0.0001,
      "loss": 1.2414,
      "step": 6149
    },
    {
      "epoch": 0.0049340341091923205,
      "grad_norm": 0.19861745834350586,
      "learning_rate": 0.0001,
      "loss": 1.2684,
      "step": 6150
    },
    {
      "epoch": 0.005041295720261719,
      "grad_norm": 0.19979241490364075,
      "learning_rate": 0.0001,
      "loss": 1.3054,
      "step": 6151
    },
    {
      "epoch": 0.005148557331331117,
      "grad_norm": 0.20027610659599304,
      "learning_rate": 0.0001,
      "loss": 1.3002,
      "step": 6152
    },
    {
      "epoch": 0.005255818942400515,
      "grad_norm": 0.1865113377571106,
      "learning_rate": 0.0001,
      "loss": 1.2808,
      "step": 6153
    },
    {
      "epoch": 0.005363080553469913,
      "grad_norm": 0.19737236201763153,
      "learning_rate": 0.0001,
      "loss": 1.1693,
      "step": 6154
    },
    {
      "epoch": 0.005470342164539311,
      "grad_norm": 0.20031487941741943,
      "learning_rate": 0.0001,
      "loss": 1.3803,
      "step": 6155
    },
    {
      "epoch": 0.00557760377560871,
      "grad_norm": 0.1939406394958496,
      "learning_rate": 0.0001,
      "loss": 1.0922,
      "step": 6156
    },
    {
      "epoch": 0.005684865386678108,
      "grad_norm": 0.19154506921768188,
      "learning_rate": 0.0001,
      "loss": 1.2647,
      "step": 6157
    },
    {
      "epoch": 0.005792126997747506,
      "grad_norm": 0.2070888727903366,
      "learning_rate": 0.0001,
      "loss": 1.2862,
      "step": 6158
    },
    {
      "epoch": 0.005899388608816904,
      "grad_norm": 0.1955205500125885,
      "learning_rate": 0.0001,
      "loss": 1.3392,
      "step": 6159
    },
    {
      "epoch": 0.006006650219886302,
      "grad_norm": 0.1987791657447815,
      "learning_rate": 0.0001,
      "loss": 1.207,
      "step": 6160
    },
    {
      "epoch": 0.006113911830955701,
      "grad_norm": 0.19708003103733063,
      "learning_rate": 0.0001,
      "loss": 1.3417,
      "step": 6161
    },
    {
      "epoch": 0.006221173442025099,
      "grad_norm": 0.1940406858921051,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 6162
    },
    {
      "epoch": 0.006328435053094498,
      "grad_norm": 0.19884800910949707,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 6163
    },
    {
      "epoch": 0.006435696664163896,
      "grad_norm": 0.1900874376296997,
      "learning_rate": 0.0001,
      "loss": 1.2119,
      "step": 6164
    },
    {
      "epoch": 0.006542958275233294,
      "grad_norm": 0.19195151329040527,
      "learning_rate": 0.0001,
      "loss": 1.2312,
      "step": 6165
    },
    {
      "epoch": 0.0066502198863026924,
      "grad_norm": 0.1886688470840454,
      "learning_rate": 0.0001,
      "loss": 1.2737,
      "step": 6166
    },
    {
      "epoch": 0.006757481497372091,
      "grad_norm": 0.19694329798221588,
      "learning_rate": 0.0001,
      "loss": 1.2918,
      "step": 6167
    },
    {
      "epoch": 0.006864743108441489,
      "grad_norm": 0.1942986249923706,
      "learning_rate": 0.0001,
      "loss": 1.3352,
      "step": 6168
    },
    {
      "epoch": 0.006972004719510887,
      "grad_norm": 0.18204884231090546,
      "learning_rate": 0.0001,
      "loss": 1.2012,
      "step": 6169
    },
    {
      "epoch": 0.007079266330580285,
      "grad_norm": 0.20335207879543304,
      "learning_rate": 0.0001,
      "loss": 1.2826,
      "step": 6170
    },
    {
      "epoch": 0.007186527941649683,
      "grad_norm": 0.20980583131313324,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 6171
    },
    {
      "epoch": 0.007293789552719082,
      "grad_norm": 0.21138647198677063,
      "learning_rate": 0.0001,
      "loss": 1.311,
      "step": 6172
    },
    {
      "epoch": 0.00740105116378848,
      "grad_norm": 0.20155590772628784,
      "learning_rate": 0.0001,
      "loss": 1.1985,
      "step": 6173
    },
    {
      "epoch": 0.007508312774857878,
      "grad_norm": 0.2007051259279251,
      "learning_rate": 0.0001,
      "loss": 1.2913,
      "step": 6174
    },
    {
      "epoch": 0.007615574385927276,
      "grad_norm": 0.2117212414741516,
      "learning_rate": 0.0001,
      "loss": 1.3265,
      "step": 6175
    },
    {
      "epoch": 0.007722835996996675,
      "grad_norm": 0.2031354010105133,
      "learning_rate": 0.0001,
      "loss": 1.2471,
      "step": 6176
    },
    {
      "epoch": 0.007830097608066073,
      "grad_norm": 0.22095496952533722,
      "learning_rate": 0.0001,
      "loss": 1.2788,
      "step": 6177
    },
    {
      "epoch": 0.007937359219135472,
      "grad_norm": 0.2060156911611557,
      "learning_rate": 0.0001,
      "loss": 1.3644,
      "step": 6178
    },
    {
      "epoch": 0.008044620830204869,
      "grad_norm": 0.20743562281131744,
      "learning_rate": 0.0001,
      "loss": 1.254,
      "step": 6179
    },
    {
      "epoch": 0.008151882441274268,
      "grad_norm": 0.19149909913539886,
      "learning_rate": 0.0001,
      "loss": 1.2628,
      "step": 6180
    },
    {
      "epoch": 0.008259144052343665,
      "grad_norm": 0.19186222553253174,
      "learning_rate": 0.0001,
      "loss": 1.321,
      "step": 6181
    },
    {
      "epoch": 0.008366405663413064,
      "grad_norm": 0.1932031363248825,
      "learning_rate": 0.0001,
      "loss": 1.2875,
      "step": 6182
    },
    {
      "epoch": 0.008473667274482463,
      "grad_norm": 0.20239092409610748,
      "learning_rate": 0.0001,
      "loss": 1.356,
      "step": 6183
    },
    {
      "epoch": 0.00858092888555186,
      "grad_norm": 0.19798223674297333,
      "learning_rate": 0.0001,
      "loss": 1.2166,
      "step": 6184
    },
    {
      "epoch": 0.00868819049662126,
      "grad_norm": 0.21399036049842834,
      "learning_rate": 0.0001,
      "loss": 1.1884,
      "step": 6185
    },
    {
      "epoch": 0.008795452107690657,
      "grad_norm": 0.20659665763378143,
      "learning_rate": 0.0001,
      "loss": 1.3044,
      "step": 6186
    },
    {
      "epoch": 0.008902713718760056,
      "grad_norm": 0.19146528840065002,
      "learning_rate": 0.0001,
      "loss": 1.2739,
      "step": 6187
    },
    {
      "epoch": 0.009009975329829454,
      "grad_norm": 0.19835175573825836,
      "learning_rate": 0.0001,
      "loss": 1.2993,
      "step": 6188
    },
    {
      "epoch": 0.009117236940898853,
      "grad_norm": 0.2020396739244461,
      "learning_rate": 0.0001,
      "loss": 1.3083,
      "step": 6189
    },
    {
      "epoch": 0.00922449855196825,
      "grad_norm": 0.19576866924762726,
      "learning_rate": 0.0001,
      "loss": 1.2465,
      "step": 6190
    },
    {
      "epoch": 0.009331760163037649,
      "grad_norm": 0.20117655396461487,
      "learning_rate": 0.0001,
      "loss": 1.3114,
      "step": 6191
    },
    {
      "epoch": 0.009439021774107046,
      "grad_norm": 0.1967199593782425,
      "learning_rate": 0.0001,
      "loss": 1.322,
      "step": 6192
    },
    {
      "epoch": 0.009546283385176445,
      "grad_norm": 0.19504979252815247,
      "learning_rate": 0.0001,
      "loss": 1.1885,
      "step": 6193
    },
    {
      "epoch": 0.009653544996245843,
      "grad_norm": 0.1991693526506424,
      "learning_rate": 0.0001,
      "loss": 1.2538,
      "step": 6194
    },
    {
      "epoch": 0.009760806607315242,
      "grad_norm": 0.19300015270709991,
      "learning_rate": 0.0001,
      "loss": 1.2449,
      "step": 6195
    },
    {
      "epoch": 0.009868068218384641,
      "grad_norm": 0.1869754195213318,
      "learning_rate": 0.0001,
      "loss": 1.2431,
      "step": 6196
    },
    {
      "epoch": 0.009975329829454038,
      "grad_norm": 0.20041504502296448,
      "learning_rate": 0.0001,
      "loss": 1.2149,
      "step": 6197
    },
    {
      "epoch": 0.010082591440523437,
      "grad_norm": 0.198547825217247,
      "learning_rate": 0.0001,
      "loss": 1.1651,
      "step": 6198
    },
    {
      "epoch": 0.010189853051592835,
      "grad_norm": 0.20356014370918274,
      "learning_rate": 0.0001,
      "loss": 1.3648,
      "step": 6199
    },
    {
      "epoch": 0.010297114662662234,
      "grad_norm": 0.1990627944469452,
      "learning_rate": 0.0001,
      "loss": 1.2301,
      "step": 6200
    },
    {
      "epoch": 0.010404376273731631,
      "grad_norm": 0.19690994918346405,
      "learning_rate": 0.0001,
      "loss": 1.3121,
      "step": 6201
    },
    {
      "epoch": 0.01051163788480103,
      "grad_norm": 0.19810788333415985,
      "learning_rate": 0.0001,
      "loss": 1.2372,
      "step": 6202
    },
    {
      "epoch": 0.010618899495870427,
      "grad_norm": 0.19301648437976837,
      "learning_rate": 0.0001,
      "loss": 1.2736,
      "step": 6203
    },
    {
      "epoch": 0.010726161106939826,
      "grad_norm": 0.19997185468673706,
      "learning_rate": 0.0001,
      "loss": 1.2759,
      "step": 6204
    },
    {
      "epoch": 0.010833422718009224,
      "grad_norm": 0.20363043248653412,
      "learning_rate": 0.0001,
      "loss": 1.2595,
      "step": 6205
    },
    {
      "epoch": 0.010940684329078623,
      "grad_norm": 0.20270875096321106,
      "learning_rate": 0.0001,
      "loss": 1.2829,
      "step": 6206
    },
    {
      "epoch": 0.01104794594014802,
      "grad_norm": 0.19621172547340393,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 6207
    },
    {
      "epoch": 0.01115520755121742,
      "grad_norm": 0.19223305583000183,
      "learning_rate": 0.0001,
      "loss": 1.2449,
      "step": 6208
    },
    {
      "epoch": 0.011262469162286818,
      "grad_norm": 0.2049228698015213,
      "learning_rate": 0.0001,
      "loss": 1.3263,
      "step": 6209
    },
    {
      "epoch": 0.011369730773356216,
      "grad_norm": 0.1962745487689972,
      "learning_rate": 0.0001,
      "loss": 1.2734,
      "step": 6210
    },
    {
      "epoch": 0.011476992384425615,
      "grad_norm": 0.18800461292266846,
      "learning_rate": 0.0001,
      "loss": 1.2461,
      "step": 6211
    },
    {
      "epoch": 0.011584253995495012,
      "grad_norm": 0.2013290375471115,
      "learning_rate": 0.0001,
      "loss": 1.2791,
      "step": 6212
    },
    {
      "epoch": 0.011691515606564411,
      "grad_norm": 0.21392536163330078,
      "learning_rate": 0.0001,
      "loss": 1.233,
      "step": 6213
    },
    {
      "epoch": 0.011798777217633808,
      "grad_norm": 0.20830099284648895,
      "learning_rate": 0.0001,
      "loss": 1.1247,
      "step": 6214
    },
    {
      "epoch": 0.011906038828703207,
      "grad_norm": 0.2241673320531845,
      "learning_rate": 0.0001,
      "loss": 1.2861,
      "step": 6215
    },
    {
      "epoch": 0.012013300439772605,
      "grad_norm": 0.21151725947856903,
      "learning_rate": 0.0001,
      "loss": 1.3539,
      "step": 6216
    },
    {
      "epoch": 0.012120562050842004,
      "grad_norm": 0.19376541674137115,
      "learning_rate": 0.0001,
      "loss": 1.2559,
      "step": 6217
    },
    {
      "epoch": 0.012227823661911401,
      "grad_norm": 0.21254031360149384,
      "learning_rate": 0.0001,
      "loss": 1.2282,
      "step": 6218
    },
    {
      "epoch": 0.0123350852729808,
      "grad_norm": 0.19501893222332,
      "learning_rate": 0.0001,
      "loss": 1.3133,
      "step": 6219
    },
    {
      "epoch": 0.012442346884050198,
      "grad_norm": 0.19790813326835632,
      "learning_rate": 0.0001,
      "loss": 1.3568,
      "step": 6220
    },
    {
      "epoch": 0.012549608495119597,
      "grad_norm": 0.19493892788887024,
      "learning_rate": 0.0001,
      "loss": 1.392,
      "step": 6221
    },
    {
      "epoch": 0.012656870106188996,
      "grad_norm": 0.21793167293071747,
      "learning_rate": 0.0001,
      "loss": 1.2711,
      "step": 6222
    },
    {
      "epoch": 0.012764131717258393,
      "grad_norm": 0.2099364846944809,
      "learning_rate": 0.0001,
      "loss": 1.3172,
      "step": 6223
    },
    {
      "epoch": 0.012871393328327792,
      "grad_norm": 0.22136101126670837,
      "learning_rate": 0.0001,
      "loss": 1.3167,
      "step": 6224
    },
    {
      "epoch": 0.01297865493939719,
      "grad_norm": 0.19663767516613007,
      "learning_rate": 0.0001,
      "loss": 1.1889,
      "step": 6225
    },
    {
      "epoch": 0.013085916550466588,
      "grad_norm": 0.20759433507919312,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 6226
    },
    {
      "epoch": 0.013193178161535986,
      "grad_norm": 0.19925859570503235,
      "learning_rate": 0.0001,
      "loss": 1.3072,
      "step": 6227
    },
    {
      "epoch": 0.013300439772605385,
      "grad_norm": 0.20088127255439758,
      "learning_rate": 0.0001,
      "loss": 1.1976,
      "step": 6228
    },
    {
      "epoch": 0.013407701383674782,
      "grad_norm": 0.19392117857933044,
      "learning_rate": 0.0001,
      "loss": 1.239,
      "step": 6229
    },
    {
      "epoch": 0.013514962994744181,
      "grad_norm": 0.19073665142059326,
      "learning_rate": 0.0001,
      "loss": 1.2847,
      "step": 6230
    },
    {
      "epoch": 0.013622224605813579,
      "grad_norm": 0.1936269849538803,
      "learning_rate": 0.0001,
      "loss": 1.2031,
      "step": 6231
    },
    {
      "epoch": 0.013729486216882978,
      "grad_norm": 0.1978171169757843,
      "learning_rate": 0.0001,
      "loss": 1.2268,
      "step": 6232
    },
    {
      "epoch": 0.013836747827952375,
      "grad_norm": 0.2142225056886673,
      "learning_rate": 0.0001,
      "loss": 1.3434,
      "step": 6233
    },
    {
      "epoch": 0.013944009439021774,
      "grad_norm": 0.19933582842350006,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 6234
    },
    {
      "epoch": 0.014051271050091173,
      "grad_norm": 0.19321228563785553,
      "learning_rate": 0.0001,
      "loss": 1.2501,
      "step": 6235
    },
    {
      "epoch": 0.01415853266116057,
      "grad_norm": 0.1945497691631317,
      "learning_rate": 0.0001,
      "loss": 1.2837,
      "step": 6236
    },
    {
      "epoch": 0.01426579427222997,
      "grad_norm": 0.20158077776432037,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 6237
    },
    {
      "epoch": 0.014373055883299367,
      "grad_norm": 0.1881740242242813,
      "learning_rate": 0.0001,
      "loss": 1.2724,
      "step": 6238
    },
    {
      "epoch": 0.014480317494368766,
      "grad_norm": 0.19319957494735718,
      "learning_rate": 0.0001,
      "loss": 1.2179,
      "step": 6239
    },
    {
      "epoch": 0.014587579105438163,
      "grad_norm": 0.2189645767211914,
      "learning_rate": 0.0001,
      "loss": 1.3381,
      "step": 6240
    },
    {
      "epoch": 0.014694840716507562,
      "grad_norm": 0.21203406155109406,
      "learning_rate": 0.0001,
      "loss": 1.2689,
      "step": 6241
    },
    {
      "epoch": 0.01480210232757696,
      "grad_norm": 0.20512275397777557,
      "learning_rate": 0.0001,
      "loss": 1.2261,
      "step": 6242
    },
    {
      "epoch": 0.014909363938646359,
      "grad_norm": 0.17996816337108612,
      "learning_rate": 0.0001,
      "loss": 1.3057,
      "step": 6243
    },
    {
      "epoch": 0.015016625549715756,
      "grad_norm": 0.19469992816448212,
      "learning_rate": 0.0001,
      "loss": 1.213,
      "step": 6244
    },
    {
      "epoch": 0.015123887160785155,
      "grad_norm": 0.19165003299713135,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 6245
    },
    {
      "epoch": 0.015231148771854552,
      "grad_norm": 0.19465158879756927,
      "learning_rate": 0.0001,
      "loss": 1.3131,
      "step": 6246
    },
    {
      "epoch": 0.015338410382923951,
      "grad_norm": 0.19475582242012024,
      "learning_rate": 0.0001,
      "loss": 1.2679,
      "step": 6247
    },
    {
      "epoch": 0.01544567199399335,
      "grad_norm": 0.1920860856771469,
      "learning_rate": 0.0001,
      "loss": 1.2816,
      "step": 6248
    },
    {
      "epoch": 0.015552933605062748,
      "grad_norm": 0.20013321936130524,
      "learning_rate": 0.0001,
      "loss": 1.2776,
      "step": 6249
    },
    {
      "epoch": 0.015660195216132145,
      "grad_norm": 0.2062523365020752,
      "learning_rate": 0.0001,
      "loss": 1.2347,
      "step": 6250
    },
    {
      "epoch": 0.015767456827201546,
      "grad_norm": 0.19912408292293549,
      "learning_rate": 0.0001,
      "loss": 1.2733,
      "step": 6251
    },
    {
      "epoch": 0.015874718438270943,
      "grad_norm": 0.21124784648418427,
      "learning_rate": 0.0001,
      "loss": 1.3493,
      "step": 6252
    },
    {
      "epoch": 0.01598198004934034,
      "grad_norm": 0.1909739375114441,
      "learning_rate": 0.0001,
      "loss": 1.32,
      "step": 6253
    },
    {
      "epoch": 0.016089241660409738,
      "grad_norm": 0.2005842626094818,
      "learning_rate": 0.0001,
      "loss": 1.3568,
      "step": 6254
    },
    {
      "epoch": 0.01619650327147914,
      "grad_norm": 0.20688804984092712,
      "learning_rate": 0.0001,
      "loss": 1.2741,
      "step": 6255
    },
    {
      "epoch": 0.016303764882548536,
      "grad_norm": 0.19142182171344757,
      "learning_rate": 0.0001,
      "loss": 1.2574,
      "step": 6256
    },
    {
      "epoch": 0.016411026493617933,
      "grad_norm": 0.20589090883731842,
      "learning_rate": 0.0001,
      "loss": 1.3696,
      "step": 6257
    },
    {
      "epoch": 0.01651828810468733,
      "grad_norm": 0.20754288136959076,
      "learning_rate": 0.0001,
      "loss": 1.3412,
      "step": 6258
    },
    {
      "epoch": 0.01662554971575673,
      "grad_norm": 0.18890541791915894,
      "learning_rate": 0.0001,
      "loss": 1.2083,
      "step": 6259
    },
    {
      "epoch": 0.01673281132682613,
      "grad_norm": 0.20417417585849762,
      "learning_rate": 0.0001,
      "loss": 1.2682,
      "step": 6260
    },
    {
      "epoch": 0.016840072937895526,
      "grad_norm": 0.20791611075401306,
      "learning_rate": 0.0001,
      "loss": 1.2078,
      "step": 6261
    },
    {
      "epoch": 0.016947334548964927,
      "grad_norm": 0.19717548787593842,
      "learning_rate": 0.0001,
      "loss": 1.2152,
      "step": 6262
    },
    {
      "epoch": 0.017054596160034324,
      "grad_norm": 0.19262859225273132,
      "learning_rate": 0.0001,
      "loss": 1.2542,
      "step": 6263
    },
    {
      "epoch": 0.01716185777110372,
      "grad_norm": 0.20325402915477753,
      "learning_rate": 0.0001,
      "loss": 1.352,
      "step": 6264
    },
    {
      "epoch": 0.01726911938217312,
      "grad_norm": 0.1896151304244995,
      "learning_rate": 0.0001,
      "loss": 1.2101,
      "step": 6265
    },
    {
      "epoch": 0.01737638099324252,
      "grad_norm": 0.19118259847164154,
      "learning_rate": 0.0001,
      "loss": 1.3113,
      "step": 6266
    },
    {
      "epoch": 0.017483642604311917,
      "grad_norm": 0.20240841805934906,
      "learning_rate": 0.0001,
      "loss": 1.3,
      "step": 6267
    },
    {
      "epoch": 0.017590904215381314,
      "grad_norm": 0.2213878631591797,
      "learning_rate": 0.0001,
      "loss": 1.1826,
      "step": 6268
    },
    {
      "epoch": 0.017698165826450712,
      "grad_norm": 0.1897164285182953,
      "learning_rate": 0.0001,
      "loss": 1.293,
      "step": 6269
    },
    {
      "epoch": 0.017805427437520113,
      "grad_norm": 0.20309379696846008,
      "learning_rate": 0.0001,
      "loss": 1.3011,
      "step": 6270
    },
    {
      "epoch": 0.01791268904858951,
      "grad_norm": 0.19685137271881104,
      "learning_rate": 0.0001,
      "loss": 1.2993,
      "step": 6271
    },
    {
      "epoch": 0.018019950659658907,
      "grad_norm": 0.21018575131893158,
      "learning_rate": 0.0001,
      "loss": 1.2896,
      "step": 6272
    },
    {
      "epoch": 0.018127212270728308,
      "grad_norm": 0.19919735193252563,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 6273
    },
    {
      "epoch": 0.018234473881797705,
      "grad_norm": 0.18861176073551178,
      "learning_rate": 0.0001,
      "loss": 1.1385,
      "step": 6274
    },
    {
      "epoch": 0.018341735492867103,
      "grad_norm": 0.19616815447807312,
      "learning_rate": 0.0001,
      "loss": 1.2674,
      "step": 6275
    },
    {
      "epoch": 0.0184489971039365,
      "grad_norm": 0.1926494687795639,
      "learning_rate": 0.0001,
      "loss": 1.3235,
      "step": 6276
    },
    {
      "epoch": 0.0185562587150059,
      "grad_norm": 0.19777683913707733,
      "learning_rate": 0.0001,
      "loss": 1.2542,
      "step": 6277
    },
    {
      "epoch": 0.018663520326075298,
      "grad_norm": 0.19123946130275726,
      "learning_rate": 0.0001,
      "loss": 1.3148,
      "step": 6278
    },
    {
      "epoch": 0.018770781937144695,
      "grad_norm": 0.20623134076595306,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 6279
    },
    {
      "epoch": 0.018878043548214093,
      "grad_norm": 0.19514460861682892,
      "learning_rate": 0.0001,
      "loss": 1.2137,
      "step": 6280
    },
    {
      "epoch": 0.018985305159283494,
      "grad_norm": 0.19414058327674866,
      "learning_rate": 0.0001,
      "loss": 1.2006,
      "step": 6281
    },
    {
      "epoch": 0.01909256677035289,
      "grad_norm": 0.19306626915931702,
      "learning_rate": 0.0001,
      "loss": 1.1741,
      "step": 6282
    },
    {
      "epoch": 0.019199828381422288,
      "grad_norm": 0.20607665181159973,
      "learning_rate": 0.0001,
      "loss": 1.2648,
      "step": 6283
    },
    {
      "epoch": 0.019307089992491686,
      "grad_norm": 0.19547469913959503,
      "learning_rate": 0.0001,
      "loss": 1.2538,
      "step": 6284
    },
    {
      "epoch": 0.019414351603561086,
      "grad_norm": 0.1944737285375595,
      "learning_rate": 0.0001,
      "loss": 1.1871,
      "step": 6285
    },
    {
      "epoch": 0.019521613214630484,
      "grad_norm": 0.19390465319156647,
      "learning_rate": 0.0001,
      "loss": 1.2244,
      "step": 6286
    },
    {
      "epoch": 0.01962887482569988,
      "grad_norm": 0.196940615773201,
      "learning_rate": 0.0001,
      "loss": 1.2081,
      "step": 6287
    },
    {
      "epoch": 0.019736136436769282,
      "grad_norm": 0.19910341501235962,
      "learning_rate": 0.0001,
      "loss": 1.2725,
      "step": 6288
    },
    {
      "epoch": 0.01984339804783868,
      "grad_norm": 0.20511825382709503,
      "learning_rate": 0.0001,
      "loss": 1.3086,
      "step": 6289
    },
    {
      "epoch": 0.019950659658908076,
      "grad_norm": 0.19813752174377441,
      "learning_rate": 0.0001,
      "loss": 1.1624,
      "step": 6290
    },
    {
      "epoch": 0.020057921269977474,
      "grad_norm": 0.19808164238929749,
      "learning_rate": 0.0001,
      "loss": 1.2736,
      "step": 6291
    },
    {
      "epoch": 0.020165182881046875,
      "grad_norm": 0.19842636585235596,
      "learning_rate": 0.0001,
      "loss": 1.2446,
      "step": 6292
    },
    {
      "epoch": 0.020272444492116272,
      "grad_norm": 0.2128971666097641,
      "learning_rate": 0.0001,
      "loss": 1.2669,
      "step": 6293
    },
    {
      "epoch": 0.02037970610318567,
      "grad_norm": 0.2051282525062561,
      "learning_rate": 0.0001,
      "loss": 1.2895,
      "step": 6294
    },
    {
      "epoch": 0.020486967714255067,
      "grad_norm": 0.18451206386089325,
      "learning_rate": 0.0001,
      "loss": 1.1907,
      "step": 6295
    },
    {
      "epoch": 0.020594229325324467,
      "grad_norm": 0.20146286487579346,
      "learning_rate": 0.0001,
      "loss": 1.2604,
      "step": 6296
    },
    {
      "epoch": 0.020701490936393865,
      "grad_norm": 0.2125876545906067,
      "learning_rate": 0.0001,
      "loss": 1.2531,
      "step": 6297
    },
    {
      "epoch": 0.020808752547463262,
      "grad_norm": 0.19938498735427856,
      "learning_rate": 0.0001,
      "loss": 1.2685,
      "step": 6298
    },
    {
      "epoch": 0.020916014158532663,
      "grad_norm": 0.21767385303974152,
      "learning_rate": 0.0001,
      "loss": 1.2314,
      "step": 6299
    },
    {
      "epoch": 0.02102327576960206,
      "grad_norm": 0.19855009019374847,
      "learning_rate": 0.0001,
      "loss": 1.1632,
      "step": 6300
    },
    {
      "epoch": 0.021130537380671457,
      "grad_norm": 0.19611524045467377,
      "learning_rate": 0.0001,
      "loss": 1.2296,
      "step": 6301
    },
    {
      "epoch": 0.021237798991740855,
      "grad_norm": 0.2029307335615158,
      "learning_rate": 0.0001,
      "loss": 1.2835,
      "step": 6302
    },
    {
      "epoch": 0.021345060602810256,
      "grad_norm": 0.20609912276268005,
      "learning_rate": 0.0001,
      "loss": 1.3641,
      "step": 6303
    },
    {
      "epoch": 0.021452322213879653,
      "grad_norm": 0.20068465173244476,
      "learning_rate": 0.0001,
      "loss": 1.2475,
      "step": 6304
    },
    {
      "epoch": 0.02155958382494905,
      "grad_norm": 0.19774819910526276,
      "learning_rate": 0.0001,
      "loss": 1.2627,
      "step": 6305
    },
    {
      "epoch": 0.021666845436018448,
      "grad_norm": 0.1990772783756256,
      "learning_rate": 0.0001,
      "loss": 1.2838,
      "step": 6306
    },
    {
      "epoch": 0.02177410704708785,
      "grad_norm": 0.2008899599313736,
      "learning_rate": 0.0001,
      "loss": 1.2348,
      "step": 6307
    },
    {
      "epoch": 0.021881368658157246,
      "grad_norm": 0.20228905975818634,
      "learning_rate": 0.0001,
      "loss": 1.268,
      "step": 6308
    },
    {
      "epoch": 0.021988630269226643,
      "grad_norm": 0.20469117164611816,
      "learning_rate": 0.0001,
      "loss": 1.2558,
      "step": 6309
    },
    {
      "epoch": 0.02209589188029604,
      "grad_norm": 0.19483739137649536,
      "learning_rate": 0.0001,
      "loss": 1.2943,
      "step": 6310
    },
    {
      "epoch": 0.02220315349136544,
      "grad_norm": 0.19796590507030487,
      "learning_rate": 0.0001,
      "loss": 1.2502,
      "step": 6311
    },
    {
      "epoch": 0.02231041510243484,
      "grad_norm": 0.2017429769039154,
      "learning_rate": 0.0001,
      "loss": 1.3257,
      "step": 6312
    },
    {
      "epoch": 0.022417676713504236,
      "grad_norm": 0.201405867934227,
      "learning_rate": 0.0001,
      "loss": 1.3844,
      "step": 6313
    },
    {
      "epoch": 0.022524938324573637,
      "grad_norm": 0.19712120294570923,
      "learning_rate": 0.0001,
      "loss": 1.234,
      "step": 6314
    },
    {
      "epoch": 0.022632199935643034,
      "grad_norm": 0.2255498617887497,
      "learning_rate": 0.0001,
      "loss": 1.2964,
      "step": 6315
    },
    {
      "epoch": 0.02273946154671243,
      "grad_norm": 0.2037588357925415,
      "learning_rate": 0.0001,
      "loss": 1.2688,
      "step": 6316
    },
    {
      "epoch": 0.02284672315778183,
      "grad_norm": 0.20619845390319824,
      "learning_rate": 0.0001,
      "loss": 1.3452,
      "step": 6317
    },
    {
      "epoch": 0.02295398476885123,
      "grad_norm": 0.20532628893852234,
      "learning_rate": 0.0001,
      "loss": 1.2192,
      "step": 6318
    },
    {
      "epoch": 0.023061246379920627,
      "grad_norm": 0.20827773213386536,
      "learning_rate": 0.0001,
      "loss": 1.2871,
      "step": 6319
    },
    {
      "epoch": 0.023168507990990024,
      "grad_norm": 0.19803524017333984,
      "learning_rate": 0.0001,
      "loss": 1.2231,
      "step": 6320
    },
    {
      "epoch": 0.02327576960205942,
      "grad_norm": 0.19437193870544434,
      "learning_rate": 0.0001,
      "loss": 1.3145,
      "step": 6321
    },
    {
      "epoch": 0.023383031213128822,
      "grad_norm": 0.19703665375709534,
      "learning_rate": 0.0001,
      "loss": 1.2862,
      "step": 6322
    },
    {
      "epoch": 0.02349029282419822,
      "grad_norm": 0.2010456621646881,
      "learning_rate": 0.0001,
      "loss": 1.2422,
      "step": 6323
    },
    {
      "epoch": 0.023597554435267617,
      "grad_norm": 0.20004014670848846,
      "learning_rate": 0.0001,
      "loss": 1.2541,
      "step": 6324
    },
    {
      "epoch": 0.023704816046337018,
      "grad_norm": 0.19463399052619934,
      "learning_rate": 0.0001,
      "loss": 1.2903,
      "step": 6325
    },
    {
      "epoch": 0.023812077657406415,
      "grad_norm": 0.19606485962867737,
      "learning_rate": 0.0001,
      "loss": 1.2271,
      "step": 6326
    },
    {
      "epoch": 0.023919339268475812,
      "grad_norm": 0.20070435106754303,
      "learning_rate": 0.0001,
      "loss": 1.2975,
      "step": 6327
    },
    {
      "epoch": 0.02402660087954521,
      "grad_norm": 0.21480582654476166,
      "learning_rate": 0.0001,
      "loss": 1.3114,
      "step": 6328
    },
    {
      "epoch": 0.02413386249061461,
      "grad_norm": 0.20578330755233765,
      "learning_rate": 0.0001,
      "loss": 1.2112,
      "step": 6329
    },
    {
      "epoch": 0.024241124101684008,
      "grad_norm": 0.20106777548789978,
      "learning_rate": 0.0001,
      "loss": 1.2108,
      "step": 6330
    },
    {
      "epoch": 0.024348385712753405,
      "grad_norm": 0.20215415954589844,
      "learning_rate": 0.0001,
      "loss": 1.2603,
      "step": 6331
    },
    {
      "epoch": 0.024455647323822802,
      "grad_norm": 0.20611786842346191,
      "learning_rate": 0.0001,
      "loss": 1.2854,
      "step": 6332
    },
    {
      "epoch": 0.024562908934892203,
      "grad_norm": 0.19982600212097168,
      "learning_rate": 0.0001,
      "loss": 1.2116,
      "step": 6333
    },
    {
      "epoch": 0.0246701705459616,
      "grad_norm": 0.20088942348957062,
      "learning_rate": 0.0001,
      "loss": 1.2361,
      "step": 6334
    },
    {
      "epoch": 0.024777432157030998,
      "grad_norm": 0.19619706273078918,
      "learning_rate": 0.0001,
      "loss": 1.238,
      "step": 6335
    },
    {
      "epoch": 0.024884693768100395,
      "grad_norm": 0.19817385077476501,
      "learning_rate": 0.0001,
      "loss": 1.2719,
      "step": 6336
    },
    {
      "epoch": 0.024991955379169796,
      "grad_norm": 0.19454598426818848,
      "learning_rate": 0.0001,
      "loss": 1.2136,
      "step": 6337
    },
    {
      "epoch": 0.025099216990239193,
      "grad_norm": 0.19306431710720062,
      "learning_rate": 0.0001,
      "loss": 1.2802,
      "step": 6338
    },
    {
      "epoch": 0.02520647860130859,
      "grad_norm": 0.20527878403663635,
      "learning_rate": 0.0001,
      "loss": 1.2667,
      "step": 6339
    },
    {
      "epoch": 0.02531374021237799,
      "grad_norm": 0.20932601392269135,
      "learning_rate": 0.0001,
      "loss": 1.289,
      "step": 6340
    },
    {
      "epoch": 0.02542100182344739,
      "grad_norm": 0.19407962262630463,
      "learning_rate": 0.0001,
      "loss": 1.2503,
      "step": 6341
    },
    {
      "epoch": 0.025528263434516786,
      "grad_norm": 0.20683352649211884,
      "learning_rate": 0.0001,
      "loss": 1.2901,
      "step": 6342
    },
    {
      "epoch": 0.025635525045586183,
      "grad_norm": 0.19821110367774963,
      "learning_rate": 0.0001,
      "loss": 1.1999,
      "step": 6343
    },
    {
      "epoch": 0.025742786656655584,
      "grad_norm": 0.20723757147789001,
      "learning_rate": 0.0001,
      "loss": 1.3146,
      "step": 6344
    },
    {
      "epoch": 0.02585004826772498,
      "grad_norm": 0.18514275550842285,
      "learning_rate": 0.0001,
      "loss": 1.2999,
      "step": 6345
    },
    {
      "epoch": 0.02595730987879438,
      "grad_norm": 0.19716277718544006,
      "learning_rate": 0.0001,
      "loss": 1.2351,
      "step": 6346
    },
    {
      "epoch": 0.026064571489863776,
      "grad_norm": 0.19866107404232025,
      "learning_rate": 0.0001,
      "loss": 1.2525,
      "step": 6347
    },
    {
      "epoch": 0.026171833100933177,
      "grad_norm": 0.2180846929550171,
      "learning_rate": 0.0001,
      "loss": 1.3253,
      "step": 6348
    },
    {
      "epoch": 0.026279094712002574,
      "grad_norm": 0.20713818073272705,
      "learning_rate": 0.0001,
      "loss": 1.2469,
      "step": 6349
    },
    {
      "epoch": 0.02638635632307197,
      "grad_norm": 0.19356299936771393,
      "learning_rate": 0.0001,
      "loss": 1.2129,
      "step": 6350
    },
    {
      "epoch": 0.026493617934141372,
      "grad_norm": 0.19256369769573212,
      "learning_rate": 0.0001,
      "loss": 1.2826,
      "step": 6351
    },
    {
      "epoch": 0.02660087954521077,
      "grad_norm": 0.1986754983663559,
      "learning_rate": 0.0001,
      "loss": 1.2703,
      "step": 6352
    },
    {
      "epoch": 0.026708141156280167,
      "grad_norm": 0.20007269084453583,
      "learning_rate": 0.0001,
      "loss": 1.3533,
      "step": 6353
    },
    {
      "epoch": 0.026815402767349564,
      "grad_norm": 0.20970192551612854,
      "learning_rate": 0.0001,
      "loss": 1.2827,
      "step": 6354
    },
    {
      "epoch": 0.026922664378418965,
      "grad_norm": 0.199740469455719,
      "learning_rate": 0.0001,
      "loss": 1.2645,
      "step": 6355
    },
    {
      "epoch": 0.027029925989488363,
      "grad_norm": 0.19510234892368317,
      "learning_rate": 0.0001,
      "loss": 1.2829,
      "step": 6356
    },
    {
      "epoch": 0.02713718760055776,
      "grad_norm": 0.1940528005361557,
      "learning_rate": 0.0001,
      "loss": 1.2591,
      "step": 6357
    },
    {
      "epoch": 0.027244449211627157,
      "grad_norm": 0.20905984938144684,
      "learning_rate": 0.0001,
      "loss": 1.3089,
      "step": 6358
    },
    {
      "epoch": 0.027351710822696558,
      "grad_norm": 0.20765948295593262,
      "learning_rate": 0.0001,
      "loss": 1.2806,
      "step": 6359
    },
    {
      "epoch": 0.027458972433765955,
      "grad_norm": 0.21586662530899048,
      "learning_rate": 0.0001,
      "loss": 1.1774,
      "step": 6360
    },
    {
      "epoch": 0.027566234044835353,
      "grad_norm": 0.19193173944950104,
      "learning_rate": 0.0001,
      "loss": 1.3213,
      "step": 6361
    },
    {
      "epoch": 0.02767349565590475,
      "grad_norm": 0.20939195156097412,
      "learning_rate": 0.0001,
      "loss": 1.2935,
      "step": 6362
    },
    {
      "epoch": 0.02778075726697415,
      "grad_norm": 0.20618346333503723,
      "learning_rate": 0.0001,
      "loss": 1.3633,
      "step": 6363
    },
    {
      "epoch": 0.027888018878043548,
      "grad_norm": 0.19994992017745972,
      "learning_rate": 0.0001,
      "loss": 1.2877,
      "step": 6364
    },
    {
      "epoch": 0.027995280489112945,
      "grad_norm": 0.19968253374099731,
      "learning_rate": 0.0001,
      "loss": 1.2924,
      "step": 6365
    },
    {
      "epoch": 0.028102542100182346,
      "grad_norm": 0.20533424615859985,
      "learning_rate": 0.0001,
      "loss": 1.286,
      "step": 6366
    },
    {
      "epoch": 0.028209803711251744,
      "grad_norm": 0.2132868468761444,
      "learning_rate": 0.0001,
      "loss": 1.2508,
      "step": 6367
    },
    {
      "epoch": 0.02831706532232114,
      "grad_norm": 0.2000369131565094,
      "learning_rate": 0.0001,
      "loss": 1.2245,
      "step": 6368
    },
    {
      "epoch": 0.028424326933390538,
      "grad_norm": 0.20690478384494781,
      "learning_rate": 0.0001,
      "loss": 1.2668,
      "step": 6369
    },
    {
      "epoch": 0.02853158854445994,
      "grad_norm": 0.20776191353797913,
      "learning_rate": 0.0001,
      "loss": 1.284,
      "step": 6370
    },
    {
      "epoch": 0.028638850155529336,
      "grad_norm": 0.20370714366436005,
      "learning_rate": 0.0001,
      "loss": 1.2681,
      "step": 6371
    },
    {
      "epoch": 0.028746111766598734,
      "grad_norm": 0.2252381443977356,
      "learning_rate": 0.0001,
      "loss": 1.3009,
      "step": 6372
    },
    {
      "epoch": 0.02885337337766813,
      "grad_norm": 0.20387104153633118,
      "learning_rate": 0.0001,
      "loss": 1.3049,
      "step": 6373
    },
    {
      "epoch": 0.028960634988737532,
      "grad_norm": 0.20707732439041138,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 6374
    },
    {
      "epoch": 0.02906789659980693,
      "grad_norm": 0.20007449388504028,
      "learning_rate": 0.0001,
      "loss": 1.2477,
      "step": 6375
    },
    {
      "epoch": 0.029175158210876326,
      "grad_norm": 0.19986362755298615,
      "learning_rate": 0.0001,
      "loss": 1.236,
      "step": 6376
    },
    {
      "epoch": 0.029282419821945727,
      "grad_norm": 0.2017400711774826,
      "learning_rate": 0.0001,
      "loss": 1.2821,
      "step": 6377
    },
    {
      "epoch": 0.029389681433015125,
      "grad_norm": 0.1957884281873703,
      "learning_rate": 0.0001,
      "loss": 1.228,
      "step": 6378
    },
    {
      "epoch": 0.029496943044084522,
      "grad_norm": 0.20155397057533264,
      "learning_rate": 0.0001,
      "loss": 1.243,
      "step": 6379
    },
    {
      "epoch": 0.02960420465515392,
      "grad_norm": 0.20423264801502228,
      "learning_rate": 0.0001,
      "loss": 1.2913,
      "step": 6380
    },
    {
      "epoch": 0.02971146626622332,
      "grad_norm": 0.19138775765895844,
      "learning_rate": 0.0001,
      "loss": 1.2145,
      "step": 6381
    },
    {
      "epoch": 0.029818727877292717,
      "grad_norm": 0.20034658908843994,
      "learning_rate": 0.0001,
      "loss": 1.2321,
      "step": 6382
    },
    {
      "epoch": 0.029925989488362115,
      "grad_norm": 0.1976722925901413,
      "learning_rate": 0.0001,
      "loss": 1.2394,
      "step": 6383
    },
    {
      "epoch": 0.030033251099431512,
      "grad_norm": 0.2082376480102539,
      "learning_rate": 0.0001,
      "loss": 1.2246,
      "step": 6384
    },
    {
      "epoch": 0.030140512710500913,
      "grad_norm": 0.2051459401845932,
      "learning_rate": 0.0001,
      "loss": 1.2039,
      "step": 6385
    },
    {
      "epoch": 0.03024777432157031,
      "grad_norm": 0.20254188776016235,
      "learning_rate": 0.0001,
      "loss": 1.2,
      "step": 6386
    },
    {
      "epoch": 0.030355035932639707,
      "grad_norm": 0.20614615082740784,
      "learning_rate": 0.0001,
      "loss": 1.2768,
      "step": 6387
    },
    {
      "epoch": 0.030462297543709105,
      "grad_norm": 0.2050320953130722,
      "learning_rate": 0.0001,
      "loss": 1.2447,
      "step": 6388
    },
    {
      "epoch": 0.030569559154778506,
      "grad_norm": 0.20430393517017365,
      "learning_rate": 0.0001,
      "loss": 1.3164,
      "step": 6389
    },
    {
      "epoch": 0.030676820765847903,
      "grad_norm": 0.18581223487854004,
      "learning_rate": 0.0001,
      "loss": 1.2359,
      "step": 6390
    },
    {
      "epoch": 0.0307840823769173,
      "grad_norm": 0.20583638548851013,
      "learning_rate": 0.0001,
      "loss": 1.2881,
      "step": 6391
    },
    {
      "epoch": 0.0308913439879867,
      "grad_norm": 0.20096063613891602,
      "learning_rate": 0.0001,
      "loss": 1.1943,
      "step": 6392
    },
    {
      "epoch": 0.0309986055990561,
      "grad_norm": 0.20690976083278656,
      "learning_rate": 0.0001,
      "loss": 1.3319,
      "step": 6393
    },
    {
      "epoch": 0.031105867210125496,
      "grad_norm": 0.19480377435684204,
      "learning_rate": 0.0001,
      "loss": 1.3122,
      "step": 6394
    },
    {
      "epoch": 0.031213128821194893,
      "grad_norm": 0.21461117267608643,
      "learning_rate": 0.0001,
      "loss": 1.3381,
      "step": 6395
    },
    {
      "epoch": 0.03132039043226429,
      "grad_norm": 0.203879252076149,
      "learning_rate": 0.0001,
      "loss": 1.3091,
      "step": 6396
    },
    {
      "epoch": 0.03142765204333369,
      "grad_norm": 0.1975259631872177,
      "learning_rate": 0.0001,
      "loss": 1.1452,
      "step": 6397
    },
    {
      "epoch": 0.03153491365440309,
      "grad_norm": 0.20698966085910797,
      "learning_rate": 0.0001,
      "loss": 1.3731,
      "step": 6398
    },
    {
      "epoch": 0.031642175265472486,
      "grad_norm": 0.19361287355422974,
      "learning_rate": 0.0001,
      "loss": 1.2706,
      "step": 6399
    },
    {
      "epoch": 0.03174943687654189,
      "grad_norm": 0.20004849135875702,
      "learning_rate": 0.0001,
      "loss": 1.3086,
      "step": 6400
    },
    {
      "epoch": 0.03185669848761128,
      "grad_norm": 0.1884019374847412,
      "learning_rate": 0.0001,
      "loss": 1.1845,
      "step": 6401
    },
    {
      "epoch": 0.03196396009868068,
      "grad_norm": 0.1990640014410019,
      "learning_rate": 0.0001,
      "loss": 1.2911,
      "step": 6402
    },
    {
      "epoch": 0.03207122170975008,
      "grad_norm": 0.19775496423244476,
      "learning_rate": 0.0001,
      "loss": 1.1853,
      "step": 6403
    },
    {
      "epoch": 0.032178483320819476,
      "grad_norm": 0.1964321732521057,
      "learning_rate": 0.0001,
      "loss": 1.2915,
      "step": 6404
    },
    {
      "epoch": 0.03228574493188888,
      "grad_norm": 0.1958484798669815,
      "learning_rate": 0.0001,
      "loss": 1.3267,
      "step": 6405
    },
    {
      "epoch": 0.03239300654295828,
      "grad_norm": 0.20707407593727112,
      "learning_rate": 0.0001,
      "loss": 1.179,
      "step": 6406
    },
    {
      "epoch": 0.03250026815402767,
      "grad_norm": 0.2003541886806488,
      "learning_rate": 0.0001,
      "loss": 1.1684,
      "step": 6407
    },
    {
      "epoch": 0.03260752976509707,
      "grad_norm": 0.21204738318920135,
      "learning_rate": 0.0001,
      "loss": 1.2073,
      "step": 6408
    },
    {
      "epoch": 0.03271479137616647,
      "grad_norm": 0.20408320426940918,
      "learning_rate": 0.0001,
      "loss": 1.3042,
      "step": 6409
    },
    {
      "epoch": 0.03282205298723587,
      "grad_norm": 0.19715996086597443,
      "learning_rate": 0.0001,
      "loss": 1.2273,
      "step": 6410
    },
    {
      "epoch": 0.03292931459830527,
      "grad_norm": 0.19771432876586914,
      "learning_rate": 0.0001,
      "loss": 1.2981,
      "step": 6411
    },
    {
      "epoch": 0.03303657620937466,
      "grad_norm": 0.19907860457897186,
      "learning_rate": 0.0001,
      "loss": 1.293,
      "step": 6412
    },
    {
      "epoch": 0.03314383782044406,
      "grad_norm": 0.20292526483535767,
      "learning_rate": 0.0001,
      "loss": 1.1461,
      "step": 6413
    },
    {
      "epoch": 0.03325109943151346,
      "grad_norm": 0.21137464046478271,
      "learning_rate": 0.0001,
      "loss": 1.3343,
      "step": 6414
    },
    {
      "epoch": 0.03335836104258286,
      "grad_norm": 0.20290479063987732,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 6415
    },
    {
      "epoch": 0.03346562265365226,
      "grad_norm": 0.1902763396501541,
      "learning_rate": 0.0001,
      "loss": 1.29,
      "step": 6416
    },
    {
      "epoch": 0.03357288426472166,
      "grad_norm": 0.21123795211315155,
      "learning_rate": 0.0001,
      "loss": 1.318,
      "step": 6417
    },
    {
      "epoch": 0.03368014587579105,
      "grad_norm": 0.20144324004650116,
      "learning_rate": 0.0001,
      "loss": 1.286,
      "step": 6418
    },
    {
      "epoch": 0.03378740748686045,
      "grad_norm": 0.2028965950012207,
      "learning_rate": 0.0001,
      "loss": 1.3045,
      "step": 6419
    },
    {
      "epoch": 0.033894669097929854,
      "grad_norm": 0.2001783847808838,
      "learning_rate": 0.0001,
      "loss": 1.3583,
      "step": 6420
    },
    {
      "epoch": 0.03400193070899925,
      "grad_norm": 0.20556144416332245,
      "learning_rate": 0.0001,
      "loss": 1.3448,
      "step": 6421
    },
    {
      "epoch": 0.03410919232006865,
      "grad_norm": 0.19935737550258636,
      "learning_rate": 0.0001,
      "loss": 1.2362,
      "step": 6422
    },
    {
      "epoch": 0.03421645393113804,
      "grad_norm": 0.19413204491138458,
      "learning_rate": 0.0001,
      "loss": 1.2216,
      "step": 6423
    },
    {
      "epoch": 0.03432371554220744,
      "grad_norm": 0.20469771325588226,
      "learning_rate": 0.0001,
      "loss": 1.3248,
      "step": 6424
    },
    {
      "epoch": 0.034430977153276844,
      "grad_norm": 0.2033664733171463,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 6425
    },
    {
      "epoch": 0.03453823876434624,
      "grad_norm": 0.19189496338367462,
      "learning_rate": 0.0001,
      "loss": 1.2849,
      "step": 6426
    },
    {
      "epoch": 0.03464550037541564,
      "grad_norm": 0.19957731664180756,
      "learning_rate": 0.0001,
      "loss": 1.2663,
      "step": 6427
    },
    {
      "epoch": 0.03475276198648504,
      "grad_norm": 0.20276470482349396,
      "learning_rate": 0.0001,
      "loss": 1.2809,
      "step": 6428
    },
    {
      "epoch": 0.03486002359755443,
      "grad_norm": 0.20199283957481384,
      "learning_rate": 0.0001,
      "loss": 1.2774,
      "step": 6429
    },
    {
      "epoch": 0.034967285208623834,
      "grad_norm": 0.19239205121994019,
      "learning_rate": 0.0001,
      "loss": 1.0714,
      "step": 6430
    },
    {
      "epoch": 0.035074546819693235,
      "grad_norm": 0.19731643795967102,
      "learning_rate": 0.0001,
      "loss": 1.2208,
      "step": 6431
    },
    {
      "epoch": 0.03518180843076263,
      "grad_norm": 0.20365603268146515,
      "learning_rate": 0.0001,
      "loss": 1.3185,
      "step": 6432
    },
    {
      "epoch": 0.03528907004183203,
      "grad_norm": 0.20422250032424927,
      "learning_rate": 0.0001,
      "loss": 1.264,
      "step": 6433
    },
    {
      "epoch": 0.035396331652901424,
      "grad_norm": 0.20938444137573242,
      "learning_rate": 0.0001,
      "loss": 1.198,
      "step": 6434
    },
    {
      "epoch": 0.035503593263970824,
      "grad_norm": 0.20061884820461273,
      "learning_rate": 0.0001,
      "loss": 1.36,
      "step": 6435
    },
    {
      "epoch": 0.035610854875040225,
      "grad_norm": 0.20913757383823395,
      "learning_rate": 0.0001,
      "loss": 1.2326,
      "step": 6436
    },
    {
      "epoch": 0.03571811648610962,
      "grad_norm": 0.19631624221801758,
      "learning_rate": 0.0001,
      "loss": 1.3286,
      "step": 6437
    },
    {
      "epoch": 0.03582537809717902,
      "grad_norm": 0.2063765674829483,
      "learning_rate": 0.0001,
      "loss": 1.3821,
      "step": 6438
    },
    {
      "epoch": 0.03593263970824842,
      "grad_norm": 0.21130160987377167,
      "learning_rate": 0.0001,
      "loss": 1.3419,
      "step": 6439
    },
    {
      "epoch": 0.036039901319317814,
      "grad_norm": 0.20387044548988342,
      "learning_rate": 0.0001,
      "loss": 1.3566,
      "step": 6440
    },
    {
      "epoch": 0.036147162930387215,
      "grad_norm": 0.19696356356143951,
      "learning_rate": 0.0001,
      "loss": 1.2373,
      "step": 6441
    },
    {
      "epoch": 0.036254424541456616,
      "grad_norm": 0.19401152431964874,
      "learning_rate": 0.0001,
      "loss": 1.2916,
      "step": 6442
    },
    {
      "epoch": 0.03636168615252601,
      "grad_norm": 0.20078755915164948,
      "learning_rate": 0.0001,
      "loss": 1.2841,
      "step": 6443
    },
    {
      "epoch": 0.03646894776359541,
      "grad_norm": 0.19673632085323334,
      "learning_rate": 0.0001,
      "loss": 1.2225,
      "step": 6444
    },
    {
      "epoch": 0.036576209374664805,
      "grad_norm": 0.19479338824748993,
      "learning_rate": 0.0001,
      "loss": 1.24,
      "step": 6445
    },
    {
      "epoch": 0.036683470985734205,
      "grad_norm": 0.21313250064849854,
      "learning_rate": 0.0001,
      "loss": 1.3354,
      "step": 6446
    },
    {
      "epoch": 0.036790732596803606,
      "grad_norm": 0.20058707892894745,
      "learning_rate": 0.0001,
      "loss": 1.2904,
      "step": 6447
    },
    {
      "epoch": 0.036897994207873,
      "grad_norm": 0.1917436420917511,
      "learning_rate": 0.0001,
      "loss": 1.2279,
      "step": 6448
    },
    {
      "epoch": 0.0370052558189424,
      "grad_norm": 0.19693288207054138,
      "learning_rate": 0.0001,
      "loss": 1.2425,
      "step": 6449
    },
    {
      "epoch": 0.0371125174300118,
      "grad_norm": 0.22197000682353973,
      "learning_rate": 0.0001,
      "loss": 1.2912,
      "step": 6450
    },
    {
      "epoch": 0.037219779041081195,
      "grad_norm": 0.2122073769569397,
      "learning_rate": 0.0001,
      "loss": 1.0835,
      "step": 6451
    },
    {
      "epoch": 0.037327040652150596,
      "grad_norm": 0.21103954315185547,
      "learning_rate": 0.0001,
      "loss": 1.3095,
      "step": 6452
    },
    {
      "epoch": 0.03743430226321999,
      "grad_norm": 0.1946851909160614,
      "learning_rate": 0.0001,
      "loss": 1.2114,
      "step": 6453
    },
    {
      "epoch": 0.03754156387428939,
      "grad_norm": 0.20815716683864594,
      "learning_rate": 0.0001,
      "loss": 1.2695,
      "step": 6454
    },
    {
      "epoch": 0.03764882548535879,
      "grad_norm": 0.20227554440498352,
      "learning_rate": 0.0001,
      "loss": 1.2246,
      "step": 6455
    },
    {
      "epoch": 0.037756087096428186,
      "grad_norm": 0.2120084911584854,
      "learning_rate": 0.0001,
      "loss": 1.3578,
      "step": 6456
    },
    {
      "epoch": 0.037863348707497586,
      "grad_norm": 0.2157229781150818,
      "learning_rate": 0.0001,
      "loss": 1.2331,
      "step": 6457
    },
    {
      "epoch": 0.03797061031856699,
      "grad_norm": 0.1915653944015503,
      "learning_rate": 0.0001,
      "loss": 1.2071,
      "step": 6458
    },
    {
      "epoch": 0.03807787192963638,
      "grad_norm": 0.21851873397827148,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 6459
    },
    {
      "epoch": 0.03818513354070578,
      "grad_norm": 0.20287621021270752,
      "learning_rate": 0.0001,
      "loss": 1.301,
      "step": 6460
    },
    {
      "epoch": 0.03829239515177518,
      "grad_norm": 0.2054014503955841,
      "learning_rate": 0.0001,
      "loss": 1.2782,
      "step": 6461
    },
    {
      "epoch": 0.038399656762844576,
      "grad_norm": 0.19133557379245758,
      "learning_rate": 0.0001,
      "loss": 1.3443,
      "step": 6462
    },
    {
      "epoch": 0.03850691837391398,
      "grad_norm": 0.21031908690929413,
      "learning_rate": 0.0001,
      "loss": 1.1869,
      "step": 6463
    },
    {
      "epoch": 0.03861417998498337,
      "grad_norm": 0.2163950651884079,
      "learning_rate": 0.0001,
      "loss": 1.3533,
      "step": 6464
    },
    {
      "epoch": 0.03872144159605277,
      "grad_norm": 0.2081238180398941,
      "learning_rate": 0.0001,
      "loss": 1.288,
      "step": 6465
    },
    {
      "epoch": 0.03882870320712217,
      "grad_norm": 0.20519328117370605,
      "learning_rate": 0.0001,
      "loss": 1.2099,
      "step": 6466
    },
    {
      "epoch": 0.03893596481819157,
      "grad_norm": 0.21248853206634521,
      "learning_rate": 0.0001,
      "loss": 1.3172,
      "step": 6467
    },
    {
      "epoch": 0.03904322642926097,
      "grad_norm": 0.19821541011333466,
      "learning_rate": 0.0001,
      "loss": 1.2865,
      "step": 6468
    },
    {
      "epoch": 0.03915048804033037,
      "grad_norm": 0.18799063563346863,
      "learning_rate": 0.0001,
      "loss": 1.2371,
      "step": 6469
    },
    {
      "epoch": 0.03925774965139976,
      "grad_norm": 0.19435277581214905,
      "learning_rate": 0.0001,
      "loss": 1.1383,
      "step": 6470
    },
    {
      "epoch": 0.03936501126246916,
      "grad_norm": 0.20686005055904388,
      "learning_rate": 0.0001,
      "loss": 1.198,
      "step": 6471
    },
    {
      "epoch": 0.039472272873538564,
      "grad_norm": 0.2053653448820114,
      "learning_rate": 0.0001,
      "loss": 1.2699,
      "step": 6472
    },
    {
      "epoch": 0.03957953448460796,
      "grad_norm": 0.20644475519657135,
      "learning_rate": 0.0001,
      "loss": 1.3548,
      "step": 6473
    },
    {
      "epoch": 0.03968679609567736,
      "grad_norm": 0.20310452580451965,
      "learning_rate": 0.0001,
      "loss": 1.1963,
      "step": 6474
    },
    {
      "epoch": 0.03979405770674675,
      "grad_norm": 0.21412622928619385,
      "learning_rate": 0.0001,
      "loss": 1.3258,
      "step": 6475
    },
    {
      "epoch": 0.03990131931781615,
      "grad_norm": 0.21197278797626495,
      "learning_rate": 0.0001,
      "loss": 1.3217,
      "step": 6476
    },
    {
      "epoch": 0.040008580928885554,
      "grad_norm": 0.19916340708732605,
      "learning_rate": 0.0001,
      "loss": 1.2919,
      "step": 6477
    },
    {
      "epoch": 0.04011584253995495,
      "grad_norm": 0.20010752975940704,
      "learning_rate": 0.0001,
      "loss": 1.2787,
      "step": 6478
    },
    {
      "epoch": 0.04022310415102435,
      "grad_norm": 0.20359405875205994,
      "learning_rate": 0.0001,
      "loss": 1.2316,
      "step": 6479
    },
    {
      "epoch": 0.04033036576209375,
      "grad_norm": 0.20223645865917206,
      "learning_rate": 0.0001,
      "loss": 1.2527,
      "step": 6480
    },
    {
      "epoch": 0.04043762737316314,
      "grad_norm": 0.21061848104000092,
      "learning_rate": 0.0001,
      "loss": 1.3181,
      "step": 6481
    },
    {
      "epoch": 0.040544888984232544,
      "grad_norm": 0.2179330438375473,
      "learning_rate": 0.0001,
      "loss": 1.3467,
      "step": 6482
    },
    {
      "epoch": 0.040652150595301945,
      "grad_norm": 0.20951248705387115,
      "learning_rate": 0.0001,
      "loss": 1.2446,
      "step": 6483
    },
    {
      "epoch": 0.04075941220637134,
      "grad_norm": 0.20356866717338562,
      "learning_rate": 0.0001,
      "loss": 1.2631,
      "step": 6484
    },
    {
      "epoch": 0.04086667381744074,
      "grad_norm": 0.20448976755142212,
      "learning_rate": 0.0001,
      "loss": 1.2463,
      "step": 6485
    },
    {
      "epoch": 0.04097393542851013,
      "grad_norm": 0.19141075015068054,
      "learning_rate": 0.0001,
      "loss": 1.1948,
      "step": 6486
    },
    {
      "epoch": 0.041081197039579534,
      "grad_norm": 0.1999606341123581,
      "learning_rate": 0.0001,
      "loss": 1.2702,
      "step": 6487
    },
    {
      "epoch": 0.041188458650648935,
      "grad_norm": 0.19817966222763062,
      "learning_rate": 0.0001,
      "loss": 1.2264,
      "step": 6488
    },
    {
      "epoch": 0.04129572026171833,
      "grad_norm": 0.19871903955936432,
      "learning_rate": 0.0001,
      "loss": 1.2245,
      "step": 6489
    },
    {
      "epoch": 0.04140298187278773,
      "grad_norm": 0.19437013566493988,
      "learning_rate": 0.0001,
      "loss": 1.216,
      "step": 6490
    },
    {
      "epoch": 0.04151024348385713,
      "grad_norm": 0.21647250652313232,
      "learning_rate": 0.0001,
      "loss": 1.3707,
      "step": 6491
    },
    {
      "epoch": 0.041617505094926524,
      "grad_norm": 0.1906319111585617,
      "learning_rate": 0.0001,
      "loss": 1.2329,
      "step": 6492
    },
    {
      "epoch": 0.041724766705995925,
      "grad_norm": 0.21466998755931854,
      "learning_rate": 0.0001,
      "loss": 1.3143,
      "step": 6493
    },
    {
      "epoch": 0.041832028317065326,
      "grad_norm": 0.19945970177650452,
      "learning_rate": 0.0001,
      "loss": 1.385,
      "step": 6494
    },
    {
      "epoch": 0.04193928992813472,
      "grad_norm": 0.20817287266254425,
      "learning_rate": 0.0001,
      "loss": 1.2686,
      "step": 6495
    },
    {
      "epoch": 0.04204655153920412,
      "grad_norm": 0.21649105846881866,
      "learning_rate": 0.0001,
      "loss": 1.3574,
      "step": 6496
    },
    {
      "epoch": 0.042153813150273514,
      "grad_norm": 0.20259574055671692,
      "learning_rate": 0.0001,
      "loss": 1.2486,
      "step": 6497
    },
    {
      "epoch": 0.042261074761342915,
      "grad_norm": 0.20316894352436066,
      "learning_rate": 0.0001,
      "loss": 1.3255,
      "step": 6498
    },
    {
      "epoch": 0.042368336372412316,
      "grad_norm": 0.200150266289711,
      "learning_rate": 0.0001,
      "loss": 1.2808,
      "step": 6499
    },
    {
      "epoch": 0.04247559798348171,
      "grad_norm": 0.2029149979352951,
      "learning_rate": 0.0001,
      "loss": 1.2509,
      "step": 6500
    },
    {
      "epoch": 0.04258285959455111,
      "grad_norm": 0.2070978581905365,
      "learning_rate": 0.0001,
      "loss": 1.2525,
      "step": 6501
    },
    {
      "epoch": 0.04269012120562051,
      "grad_norm": 0.19688844680786133,
      "learning_rate": 0.0001,
      "loss": 1.3372,
      "step": 6502
    },
    {
      "epoch": 0.042797382816689905,
      "grad_norm": 0.2113284319639206,
      "learning_rate": 0.0001,
      "loss": 1.3056,
      "step": 6503
    },
    {
      "epoch": 0.042904644427759306,
      "grad_norm": 0.20480872690677643,
      "learning_rate": 0.0001,
      "loss": 1.3525,
      "step": 6504
    },
    {
      "epoch": 0.04301190603882871,
      "grad_norm": 0.19923479855060577,
      "learning_rate": 0.0001,
      "loss": 1.3102,
      "step": 6505
    },
    {
      "epoch": 0.0431191676498981,
      "grad_norm": 0.20014123618602753,
      "learning_rate": 0.0001,
      "loss": 1.3039,
      "step": 6506
    },
    {
      "epoch": 0.0432264292609675,
      "grad_norm": 0.20379860699176788,
      "learning_rate": 0.0001,
      "loss": 1.3728,
      "step": 6507
    },
    {
      "epoch": 0.043333690872036895,
      "grad_norm": 0.20651304721832275,
      "learning_rate": 0.0001,
      "loss": 1.3018,
      "step": 6508
    },
    {
      "epoch": 0.043440952483106296,
      "grad_norm": 0.2022644728422165,
      "learning_rate": 0.0001,
      "loss": 1.238,
      "step": 6509
    },
    {
      "epoch": 0.0435482140941757,
      "grad_norm": 0.22375372052192688,
      "learning_rate": 0.0001,
      "loss": 1.2566,
      "step": 6510
    },
    {
      "epoch": 0.04365547570524509,
      "grad_norm": 0.1940920501947403,
      "learning_rate": 0.0001,
      "loss": 1.2341,
      "step": 6511
    },
    {
      "epoch": 0.04376273731631449,
      "grad_norm": 0.23418186604976654,
      "learning_rate": 0.0001,
      "loss": 1.3638,
      "step": 6512
    },
    {
      "epoch": 0.04386999892738389,
      "grad_norm": 0.2020508497953415,
      "learning_rate": 0.0001,
      "loss": 1.2712,
      "step": 6513
    },
    {
      "epoch": 0.043977260538453286,
      "grad_norm": 0.19660769402980804,
      "learning_rate": 0.0001,
      "loss": 1.35,
      "step": 6514
    },
    {
      "epoch": 0.04408452214952269,
      "grad_norm": 0.21207016706466675,
      "learning_rate": 0.0001,
      "loss": 1.2509,
      "step": 6515
    },
    {
      "epoch": 0.04419178376059208,
      "grad_norm": 0.2006383091211319,
      "learning_rate": 0.0001,
      "loss": 1.1449,
      "step": 6516
    },
    {
      "epoch": 0.04429904537166148,
      "grad_norm": 0.20800670981407166,
      "learning_rate": 0.0001,
      "loss": 1.2474,
      "step": 6517
    },
    {
      "epoch": 0.04440630698273088,
      "grad_norm": 0.20379985868930817,
      "learning_rate": 0.0001,
      "loss": 1.3134,
      "step": 6518
    },
    {
      "epoch": 0.044513568593800276,
      "grad_norm": 0.21141880750656128,
      "learning_rate": 0.0001,
      "loss": 1.335,
      "step": 6519
    },
    {
      "epoch": 0.04462083020486968,
      "grad_norm": 0.2099279910326004,
      "learning_rate": 0.0001,
      "loss": 1.2452,
      "step": 6520
    },
    {
      "epoch": 0.04472809181593908,
      "grad_norm": 0.19597786664962769,
      "learning_rate": 0.0001,
      "loss": 1.1615,
      "step": 6521
    },
    {
      "epoch": 0.04483535342700847,
      "grad_norm": 0.19475574791431427,
      "learning_rate": 0.0001,
      "loss": 1.2,
      "step": 6522
    },
    {
      "epoch": 0.04494261503807787,
      "grad_norm": 0.20458227396011353,
      "learning_rate": 0.0001,
      "loss": 1.397,
      "step": 6523
    },
    {
      "epoch": 0.04504987664914727,
      "grad_norm": 0.19668734073638916,
      "learning_rate": 0.0001,
      "loss": 1.2597,
      "step": 6524
    },
    {
      "epoch": 0.04515713826021667,
      "grad_norm": 0.2034541815519333,
      "learning_rate": 0.0001,
      "loss": 1.2747,
      "step": 6525
    },
    {
      "epoch": 0.04526439987128607,
      "grad_norm": 0.2091502845287323,
      "learning_rate": 0.0001,
      "loss": 1.4167,
      "step": 6526
    },
    {
      "epoch": 0.04537166148235546,
      "grad_norm": 0.19303783774375916,
      "learning_rate": 0.0001,
      "loss": 1.2942,
      "step": 6527
    },
    {
      "epoch": 0.04547892309342486,
      "grad_norm": 0.1912301927804947,
      "learning_rate": 0.0001,
      "loss": 1.1542,
      "step": 6528
    },
    {
      "epoch": 0.04558618470449426,
      "grad_norm": 0.20510578155517578,
      "learning_rate": 0.0001,
      "loss": 1.3301,
      "step": 6529
    },
    {
      "epoch": 0.04569344631556366,
      "grad_norm": 0.21992599964141846,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 6530
    },
    {
      "epoch": 0.04580070792663306,
      "grad_norm": 0.21039177477359772,
      "learning_rate": 0.0001,
      "loss": 1.2354,
      "step": 6531
    },
    {
      "epoch": 0.04590796953770246,
      "grad_norm": 0.19904334843158722,
      "learning_rate": 0.0001,
      "loss": 1.2931,
      "step": 6532
    },
    {
      "epoch": 0.04601523114877185,
      "grad_norm": 0.20190420746803284,
      "learning_rate": 0.0001,
      "loss": 1.3953,
      "step": 6533
    },
    {
      "epoch": 0.04612249275984125,
      "grad_norm": 0.20714306831359863,
      "learning_rate": 0.0001,
      "loss": 1.283,
      "step": 6534
    },
    {
      "epoch": 0.046229754370910654,
      "grad_norm": 0.2155359983444214,
      "learning_rate": 0.0001,
      "loss": 1.3232,
      "step": 6535
    },
    {
      "epoch": 0.04633701598198005,
      "grad_norm": 0.19413737952709198,
      "learning_rate": 0.0001,
      "loss": 1.2091,
      "step": 6536
    },
    {
      "epoch": 0.04644427759304945,
      "grad_norm": 0.1905590295791626,
      "learning_rate": 0.0001,
      "loss": 1.228,
      "step": 6537
    },
    {
      "epoch": 0.04655153920411884,
      "grad_norm": 0.19309791922569275,
      "learning_rate": 0.0001,
      "loss": 1.2371,
      "step": 6538
    },
    {
      "epoch": 0.046658800815188244,
      "grad_norm": 0.18759897351264954,
      "learning_rate": 0.0001,
      "loss": 1.1914,
      "step": 6539
    },
    {
      "epoch": 0.046766062426257644,
      "grad_norm": 0.2069128155708313,
      "learning_rate": 0.0001,
      "loss": 1.311,
      "step": 6540
    },
    {
      "epoch": 0.04687332403732704,
      "grad_norm": 0.22500579059123993,
      "learning_rate": 0.0001,
      "loss": 1.2129,
      "step": 6541
    },
    {
      "epoch": 0.04698058564839644,
      "grad_norm": 0.2175019085407257,
      "learning_rate": 0.0001,
      "loss": 1.2667,
      "step": 6542
    },
    {
      "epoch": 0.04708784725946584,
      "grad_norm": 0.2078363001346588,
      "learning_rate": 0.0001,
      "loss": 1.2658,
      "step": 6543
    },
    {
      "epoch": 0.047195108870535234,
      "grad_norm": 0.1948874443769455,
      "learning_rate": 0.0001,
      "loss": 1.2573,
      "step": 6544
    },
    {
      "epoch": 0.047302370481604634,
      "grad_norm": 0.19449086487293243,
      "learning_rate": 0.0001,
      "loss": 1.2585,
      "step": 6545
    },
    {
      "epoch": 0.047409632092674035,
      "grad_norm": 0.19584345817565918,
      "learning_rate": 0.0001,
      "loss": 1.2516,
      "step": 6546
    },
    {
      "epoch": 0.04751689370374343,
      "grad_norm": 0.20663590729236603,
      "learning_rate": 0.0001,
      "loss": 1.1528,
      "step": 6547
    },
    {
      "epoch": 0.04762415531481283,
      "grad_norm": 0.21224422752857208,
      "learning_rate": 0.0001,
      "loss": 1.2321,
      "step": 6548
    },
    {
      "epoch": 0.047731416925882224,
      "grad_norm": 0.20319810509681702,
      "learning_rate": 0.0001,
      "loss": 1.2367,
      "step": 6549
    },
    {
      "epoch": 0.047838678536951625,
      "grad_norm": 0.20018699765205383,
      "learning_rate": 0.0001,
      "loss": 1.2699,
      "step": 6550
    },
    {
      "epoch": 0.047945940148021025,
      "grad_norm": 0.20051078498363495,
      "learning_rate": 0.0001,
      "loss": 1.2648,
      "step": 6551
    },
    {
      "epoch": 0.04805320175909042,
      "grad_norm": 0.2107832431793213,
      "learning_rate": 0.0001,
      "loss": 1.3063,
      "step": 6552
    },
    {
      "epoch": 0.04816046337015982,
      "grad_norm": 0.1990395039319992,
      "learning_rate": 0.0001,
      "loss": 1.332,
      "step": 6553
    },
    {
      "epoch": 0.04826772498122922,
      "grad_norm": 0.1995873749256134,
      "learning_rate": 0.0001,
      "loss": 1.3078,
      "step": 6554
    },
    {
      "epoch": 0.048374986592298615,
      "grad_norm": 0.2110949456691742,
      "learning_rate": 0.0001,
      "loss": 1.2253,
      "step": 6555
    },
    {
      "epoch": 0.048482248203368015,
      "grad_norm": 0.1984872967004776,
      "learning_rate": 0.0001,
      "loss": 1.3001,
      "step": 6556
    },
    {
      "epoch": 0.048589509814437416,
      "grad_norm": 0.20823393762111664,
      "learning_rate": 0.0001,
      "loss": 1.2404,
      "step": 6557
    },
    {
      "epoch": 0.04869677142550681,
      "grad_norm": 0.21082617342472076,
      "learning_rate": 0.0001,
      "loss": 1.238,
      "step": 6558
    },
    {
      "epoch": 0.04880403303657621,
      "grad_norm": 0.19909682869911194,
      "learning_rate": 0.0001,
      "loss": 1.1819,
      "step": 6559
    },
    {
      "epoch": 0.048911294647645605,
      "grad_norm": 0.21199747920036316,
      "learning_rate": 0.0001,
      "loss": 1.3182,
      "step": 6560
    },
    {
      "epoch": 0.049018556258715006,
      "grad_norm": 0.20724953711032867,
      "learning_rate": 0.0001,
      "loss": 1.2557,
      "step": 6561
    },
    {
      "epoch": 0.049125817869784406,
      "grad_norm": 0.20480221509933472,
      "learning_rate": 0.0001,
      "loss": 1.2728,
      "step": 6562
    },
    {
      "epoch": 0.0492330794808538,
      "grad_norm": 0.2060379534959793,
      "learning_rate": 0.0001,
      "loss": 1.2428,
      "step": 6563
    },
    {
      "epoch": 0.0493403410919232,
      "grad_norm": 0.20408302545547485,
      "learning_rate": 0.0001,
      "loss": 1.2793,
      "step": 6564
    },
    {
      "epoch": 0.0494476027029926,
      "grad_norm": 0.21220238506793976,
      "learning_rate": 0.0001,
      "loss": 1.2755,
      "step": 6565
    },
    {
      "epoch": 0.049554864314061996,
      "grad_norm": 0.20843391120433807,
      "learning_rate": 0.0001,
      "loss": 1.3156,
      "step": 6566
    },
    {
      "epoch": 0.049662125925131397,
      "grad_norm": 0.20763564109802246,
      "learning_rate": 0.0001,
      "loss": 1.2781,
      "step": 6567
    },
    {
      "epoch": 0.04976938753620079,
      "grad_norm": 0.20406587421894073,
      "learning_rate": 0.0001,
      "loss": 1.2014,
      "step": 6568
    },
    {
      "epoch": 0.04987664914727019,
      "grad_norm": 0.21208855509757996,
      "learning_rate": 0.0001,
      "loss": 1.3178,
      "step": 6569
    },
    {
      "epoch": 0.04998391075833959,
      "grad_norm": 0.2041846215724945,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 6570
    },
    {
      "epoch": 0.050091172369408986,
      "grad_norm": 0.21703234314918518,
      "learning_rate": 0.0001,
      "loss": 1.2081,
      "step": 6571
    },
    {
      "epoch": 0.05019843398047839,
      "grad_norm": 0.22067606449127197,
      "learning_rate": 0.0001,
      "loss": 1.3399,
      "step": 6572
    },
    {
      "epoch": 0.05030569559154779,
      "grad_norm": 0.2056623250246048,
      "learning_rate": 0.0001,
      "loss": 1.2823,
      "step": 6573
    },
    {
      "epoch": 0.05041295720261718,
      "grad_norm": 0.20470760762691498,
      "learning_rate": 0.0001,
      "loss": 1.3128,
      "step": 6574
    },
    {
      "epoch": 0.05052021881368658,
      "grad_norm": 0.2081969678401947,
      "learning_rate": 0.0001,
      "loss": 1.357,
      "step": 6575
    },
    {
      "epoch": 0.05062748042475598,
      "grad_norm": 0.205609530210495,
      "learning_rate": 0.0001,
      "loss": 1.3532,
      "step": 6576
    },
    {
      "epoch": 0.05073474203582538,
      "grad_norm": 0.19944219291210175,
      "learning_rate": 0.0001,
      "loss": 1.2479,
      "step": 6577
    },
    {
      "epoch": 0.05084200364689478,
      "grad_norm": 0.2038002908229828,
      "learning_rate": 0.0001,
      "loss": 1.1656,
      "step": 6578
    },
    {
      "epoch": 0.05094926525796417,
      "grad_norm": 0.20189796388149261,
      "learning_rate": 0.0001,
      "loss": 1.2855,
      "step": 6579
    },
    {
      "epoch": 0.05105652686903357,
      "grad_norm": 0.20430885255336761,
      "learning_rate": 0.0001,
      "loss": 1.3217,
      "step": 6580
    },
    {
      "epoch": 0.05116378848010297,
      "grad_norm": 0.20550139248371124,
      "learning_rate": 0.0001,
      "loss": 1.3602,
      "step": 6581
    },
    {
      "epoch": 0.05127105009117237,
      "grad_norm": 0.20511266589164734,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 6582
    },
    {
      "epoch": 0.05137831170224177,
      "grad_norm": 0.1998504251241684,
      "learning_rate": 0.0001,
      "loss": 1.2796,
      "step": 6583
    },
    {
      "epoch": 0.05148557331331117,
      "grad_norm": 0.20273751020431519,
      "learning_rate": 0.0001,
      "loss": 1.2387,
      "step": 6584
    },
    {
      "epoch": 0.05159283492438056,
      "grad_norm": 0.20327600836753845,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 6585
    },
    {
      "epoch": 0.05170009653544996,
      "grad_norm": 0.20181895792484283,
      "learning_rate": 0.0001,
      "loss": 1.3223,
      "step": 6586
    },
    {
      "epoch": 0.051807358146519364,
      "grad_norm": 0.1981777846813202,
      "learning_rate": 0.0001,
      "loss": 1.2243,
      "step": 6587
    },
    {
      "epoch": 0.05191461975758876,
      "grad_norm": 0.22853274643421173,
      "learning_rate": 0.0001,
      "loss": 1.3904,
      "step": 6588
    },
    {
      "epoch": 0.05202188136865816,
      "grad_norm": 0.21624144911766052,
      "learning_rate": 0.0001,
      "loss": 1.2864,
      "step": 6589
    },
    {
      "epoch": 0.05212914297972755,
      "grad_norm": 0.19542445242404938,
      "learning_rate": 0.0001,
      "loss": 1.2424,
      "step": 6590
    },
    {
      "epoch": 0.05223640459079695,
      "grad_norm": 0.20953844487667084,
      "learning_rate": 0.0001,
      "loss": 1.4043,
      "step": 6591
    },
    {
      "epoch": 0.052343666201866354,
      "grad_norm": 0.21951621770858765,
      "learning_rate": 0.0001,
      "loss": 1.2871,
      "step": 6592
    },
    {
      "epoch": 0.05245092781293575,
      "grad_norm": 0.2042069435119629,
      "learning_rate": 0.0001,
      "loss": 1.3231,
      "step": 6593
    },
    {
      "epoch": 0.05255818942400515,
      "grad_norm": 0.22284971177577972,
      "learning_rate": 0.0001,
      "loss": 1.2741,
      "step": 6594
    },
    {
      "epoch": 0.05266545103507455,
      "grad_norm": 0.21989920735359192,
      "learning_rate": 0.0001,
      "loss": 1.3267,
      "step": 6595
    },
    {
      "epoch": 0.05277271264614394,
      "grad_norm": 0.2040856033563614,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 6596
    },
    {
      "epoch": 0.052879974257213344,
      "grad_norm": 0.22317636013031006,
      "learning_rate": 0.0001,
      "loss": 1.2472,
      "step": 6597
    },
    {
      "epoch": 0.052987235868282745,
      "grad_norm": 0.20041857659816742,
      "learning_rate": 0.0001,
      "loss": 1.2368,
      "step": 6598
    },
    {
      "epoch": 0.05309449747935214,
      "grad_norm": 0.2028709203004837,
      "learning_rate": 0.0001,
      "loss": 1.2174,
      "step": 6599
    },
    {
      "epoch": 0.05320175909042154,
      "grad_norm": 0.2087966352701187,
      "learning_rate": 0.0001,
      "loss": 1.3471,
      "step": 6600
    },
    {
      "epoch": 0.05330902070149093,
      "grad_norm": 0.21927422285079956,
      "learning_rate": 0.0001,
      "loss": 1.4133,
      "step": 6601
    },
    {
      "epoch": 0.053416282312560334,
      "grad_norm": 0.20717257261276245,
      "learning_rate": 0.0001,
      "loss": 1.2842,
      "step": 6602
    },
    {
      "epoch": 0.053523543923629735,
      "grad_norm": 0.2006472498178482,
      "learning_rate": 0.0001,
      "loss": 1.2997,
      "step": 6603
    },
    {
      "epoch": 0.05363080553469913,
      "grad_norm": 0.20398865640163422,
      "learning_rate": 0.0001,
      "loss": 1.3308,
      "step": 6604
    },
    {
      "epoch": 0.05373806714576853,
      "grad_norm": 0.19548675417900085,
      "learning_rate": 0.0001,
      "loss": 1.1397,
      "step": 6605
    },
    {
      "epoch": 0.05384532875683793,
      "grad_norm": 0.19918756186962128,
      "learning_rate": 0.0001,
      "loss": 1.3025,
      "step": 6606
    },
    {
      "epoch": 0.053952590367907324,
      "grad_norm": 0.20980605483055115,
      "learning_rate": 0.0001,
      "loss": 1.228,
      "step": 6607
    },
    {
      "epoch": 0.054059851978976725,
      "grad_norm": 0.211741641163826,
      "learning_rate": 0.0001,
      "loss": 1.2408,
      "step": 6608
    },
    {
      "epoch": 0.054167113590046126,
      "grad_norm": 0.21114656329154968,
      "learning_rate": 0.0001,
      "loss": 1.2186,
      "step": 6609
    },
    {
      "epoch": 0.05427437520111552,
      "grad_norm": 0.19310502707958221,
      "learning_rate": 0.0001,
      "loss": 1.3184,
      "step": 6610
    },
    {
      "epoch": 0.05438163681218492,
      "grad_norm": 0.2192358374595642,
      "learning_rate": 0.0001,
      "loss": 1.2724,
      "step": 6611
    },
    {
      "epoch": 0.054488898423254314,
      "grad_norm": 0.2306223213672638,
      "learning_rate": 0.0001,
      "loss": 1.3516,
      "step": 6612
    },
    {
      "epoch": 0.054596160034323715,
      "grad_norm": 0.21865743398666382,
      "learning_rate": 0.0001,
      "loss": 1.2922,
      "step": 6613
    },
    {
      "epoch": 0.054703421645393116,
      "grad_norm": 0.2173282504081726,
      "learning_rate": 0.0001,
      "loss": 1.2602,
      "step": 6614
    },
    {
      "epoch": 0.05481068325646251,
      "grad_norm": 0.20915508270263672,
      "learning_rate": 0.0001,
      "loss": 1.3285,
      "step": 6615
    },
    {
      "epoch": 0.05491794486753191,
      "grad_norm": 0.20703886449337006,
      "learning_rate": 0.0001,
      "loss": 1.361,
      "step": 6616
    },
    {
      "epoch": 0.05502520647860131,
      "grad_norm": 0.20058168470859528,
      "learning_rate": 0.0001,
      "loss": 1.2419,
      "step": 6617
    },
    {
      "epoch": 0.055132468089670705,
      "grad_norm": 0.20341211557388306,
      "learning_rate": 0.0001,
      "loss": 1.2788,
      "step": 6618
    },
    {
      "epoch": 0.055239729700740106,
      "grad_norm": 0.20657671988010406,
      "learning_rate": 0.0001,
      "loss": 1.3052,
      "step": 6619
    },
    {
      "epoch": 0.0553469913118095,
      "grad_norm": 0.21297645568847656,
      "learning_rate": 0.0001,
      "loss": 1.3492,
      "step": 6620
    },
    {
      "epoch": 0.0554542529228789,
      "grad_norm": 0.20368823409080505,
      "learning_rate": 0.0001,
      "loss": 1.224,
      "step": 6621
    },
    {
      "epoch": 0.0555615145339483,
      "grad_norm": 0.21632198989391327,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 6622
    },
    {
      "epoch": 0.055668776145017695,
      "grad_norm": 0.21849700808525085,
      "learning_rate": 0.0001,
      "loss": 1.3272,
      "step": 6623
    },
    {
      "epoch": 0.055776037756087096,
      "grad_norm": 0.20529122650623322,
      "learning_rate": 0.0001,
      "loss": 1.2384,
      "step": 6624
    },
    {
      "epoch": 0.0558832993671565,
      "grad_norm": 0.19887523353099823,
      "learning_rate": 0.0001,
      "loss": 1.2413,
      "step": 6625
    },
    {
      "epoch": 0.05599056097822589,
      "grad_norm": 0.20999136567115784,
      "learning_rate": 0.0001,
      "loss": 1.372,
      "step": 6626
    },
    {
      "epoch": 0.05609782258929529,
      "grad_norm": 0.20456062257289886,
      "learning_rate": 0.0001,
      "loss": 1.3042,
      "step": 6627
    },
    {
      "epoch": 0.05620508420036469,
      "grad_norm": 0.21462586522102356,
      "learning_rate": 0.0001,
      "loss": 1.2672,
      "step": 6628
    },
    {
      "epoch": 0.056312345811434086,
      "grad_norm": 0.22142267227172852,
      "learning_rate": 0.0001,
      "loss": 1.358,
      "step": 6629
    },
    {
      "epoch": 0.05641960742250349,
      "grad_norm": 0.20921599864959717,
      "learning_rate": 0.0001,
      "loss": 1.3558,
      "step": 6630
    },
    {
      "epoch": 0.05652686903357288,
      "grad_norm": 0.20699502527713776,
      "learning_rate": 0.0001,
      "loss": 1.2634,
      "step": 6631
    },
    {
      "epoch": 0.05663413064464228,
      "grad_norm": 0.20573115348815918,
      "learning_rate": 0.0001,
      "loss": 1.2925,
      "step": 6632
    },
    {
      "epoch": 0.05674139225571168,
      "grad_norm": 0.21576640009880066,
      "learning_rate": 0.0001,
      "loss": 1.258,
      "step": 6633
    },
    {
      "epoch": 0.056848653866781076,
      "grad_norm": 0.20195481181144714,
      "learning_rate": 0.0001,
      "loss": 1.1833,
      "step": 6634
    },
    {
      "epoch": 0.05695591547785048,
      "grad_norm": 0.21021559834480286,
      "learning_rate": 0.0001,
      "loss": 1.3041,
      "step": 6635
    },
    {
      "epoch": 0.05706317708891988,
      "grad_norm": 0.20662611722946167,
      "learning_rate": 0.0001,
      "loss": 1.2939,
      "step": 6636
    },
    {
      "epoch": 0.05717043869998927,
      "grad_norm": 0.20129045844078064,
      "learning_rate": 0.0001,
      "loss": 1.2153,
      "step": 6637
    },
    {
      "epoch": 0.05727770031105867,
      "grad_norm": 0.2026989609003067,
      "learning_rate": 0.0001,
      "loss": 1.2245,
      "step": 6638
    },
    {
      "epoch": 0.057384961922128074,
      "grad_norm": 0.20151586830615997,
      "learning_rate": 0.0001,
      "loss": 1.3046,
      "step": 6639
    },
    {
      "epoch": 0.05749222353319747,
      "grad_norm": 0.20809774100780487,
      "learning_rate": 0.0001,
      "loss": 1.2509,
      "step": 6640
    },
    {
      "epoch": 0.05759948514426687,
      "grad_norm": 0.19076137244701385,
      "learning_rate": 0.0001,
      "loss": 1.1827,
      "step": 6641
    },
    {
      "epoch": 0.05770674675533626,
      "grad_norm": 0.20619656145572662,
      "learning_rate": 0.0001,
      "loss": 1.2782,
      "step": 6642
    },
    {
      "epoch": 0.05781400836640566,
      "grad_norm": 0.22857913374900818,
      "learning_rate": 0.0001,
      "loss": 1.2811,
      "step": 6643
    },
    {
      "epoch": 0.057921269977475064,
      "grad_norm": 0.20274700224399567,
      "learning_rate": 0.0001,
      "loss": 1.3393,
      "step": 6644
    },
    {
      "epoch": 0.05802853158854446,
      "grad_norm": 0.2003968507051468,
      "learning_rate": 0.0001,
      "loss": 1.2409,
      "step": 6645
    },
    {
      "epoch": 0.05813579319961386,
      "grad_norm": 0.20869502425193787,
      "learning_rate": 0.0001,
      "loss": 1.3197,
      "step": 6646
    },
    {
      "epoch": 0.05824305481068326,
      "grad_norm": 0.2053767442703247,
      "learning_rate": 0.0001,
      "loss": 1.244,
      "step": 6647
    },
    {
      "epoch": 0.05835031642175265,
      "grad_norm": 0.20399190485477448,
      "learning_rate": 0.0001,
      "loss": 1.119,
      "step": 6648
    },
    {
      "epoch": 0.058457578032822054,
      "grad_norm": 0.20072078704833984,
      "learning_rate": 0.0001,
      "loss": 1.31,
      "step": 6649
    },
    {
      "epoch": 0.058564839643891455,
      "grad_norm": 0.20318131148815155,
      "learning_rate": 0.0001,
      "loss": 1.2185,
      "step": 6650
    },
    {
      "epoch": 0.05867210125496085,
      "grad_norm": 0.21572232246398926,
      "learning_rate": 0.0001,
      "loss": 1.3662,
      "step": 6651
    },
    {
      "epoch": 0.05877936286603025,
      "grad_norm": 0.2126794457435608,
      "learning_rate": 0.0001,
      "loss": 1.2689,
      "step": 6652
    },
    {
      "epoch": 0.05888662447709964,
      "grad_norm": 0.21189765632152557,
      "learning_rate": 0.0001,
      "loss": 1.2786,
      "step": 6653
    },
    {
      "epoch": 0.058993886088169044,
      "grad_norm": 0.21470671892166138,
      "learning_rate": 0.0001,
      "loss": 1.2515,
      "step": 6654
    },
    {
      "epoch": 0.059101147699238445,
      "grad_norm": 0.20861688256263733,
      "learning_rate": 0.0001,
      "loss": 1.2093,
      "step": 6655
    },
    {
      "epoch": 0.05920840931030784,
      "grad_norm": 0.19579897820949554,
      "learning_rate": 0.0001,
      "loss": 1.1597,
      "step": 6656
    },
    {
      "epoch": 0.05931567092137724,
      "grad_norm": 0.21322114765644073,
      "learning_rate": 0.0001,
      "loss": 1.2227,
      "step": 6657
    },
    {
      "epoch": 0.05942293253244664,
      "grad_norm": 0.21599280834197998,
      "learning_rate": 0.0001,
      "loss": 1.2744,
      "step": 6658
    },
    {
      "epoch": 0.059530194143516034,
      "grad_norm": 0.21012566983699799,
      "learning_rate": 0.0001,
      "loss": 1.2103,
      "step": 6659
    },
    {
      "epoch": 0.059637455754585435,
      "grad_norm": 0.26727479696273804,
      "learning_rate": 0.0001,
      "loss": 1.3159,
      "step": 6660
    },
    {
      "epoch": 0.059744717365654836,
      "grad_norm": 0.19893427193164825,
      "learning_rate": 0.0001,
      "loss": 1.2138,
      "step": 6661
    },
    {
      "epoch": 0.05985197897672423,
      "grad_norm": 0.19592465460300446,
      "learning_rate": 0.0001,
      "loss": 1.2439,
      "step": 6662
    },
    {
      "epoch": 0.05995924058779363,
      "grad_norm": 0.19461405277252197,
      "learning_rate": 0.0001,
      "loss": 1.2061,
      "step": 6663
    },
    {
      "epoch": 0.060066502198863024,
      "grad_norm": 0.21607816219329834,
      "learning_rate": 0.0001,
      "loss": 1.3588,
      "step": 6664
    },
    {
      "epoch": 0.060173763809932425,
      "grad_norm": 0.2045373022556305,
      "learning_rate": 0.0001,
      "loss": 1.274,
      "step": 6665
    },
    {
      "epoch": 0.060281025421001826,
      "grad_norm": 0.2157997041940689,
      "learning_rate": 0.0001,
      "loss": 1.3251,
      "step": 6666
    },
    {
      "epoch": 0.06038828703207122,
      "grad_norm": 0.20460839569568634,
      "learning_rate": 0.0001,
      "loss": 1.2382,
      "step": 6667
    },
    {
      "epoch": 0.06049554864314062,
      "grad_norm": 0.1938919723033905,
      "learning_rate": 0.0001,
      "loss": 1.1601,
      "step": 6668
    },
    {
      "epoch": 0.06060281025421002,
      "grad_norm": 0.20950157940387726,
      "learning_rate": 0.0001,
      "loss": 1.3109,
      "step": 6669
    },
    {
      "epoch": 0.060710071865279415,
      "grad_norm": 0.2048896998167038,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 6670
    },
    {
      "epoch": 0.060817333476348816,
      "grad_norm": 0.1911996454000473,
      "learning_rate": 0.0001,
      "loss": 1.3152,
      "step": 6671
    },
    {
      "epoch": 0.06092459508741821,
      "grad_norm": 0.2072335034608841,
      "learning_rate": 0.0001,
      "loss": 1.3156,
      "step": 6672
    },
    {
      "epoch": 0.06103185669848761,
      "grad_norm": 0.19701531529426575,
      "learning_rate": 0.0001,
      "loss": 1.2002,
      "step": 6673
    },
    {
      "epoch": 0.06113911830955701,
      "grad_norm": 0.21543365716934204,
      "learning_rate": 0.0001,
      "loss": 1.2925,
      "step": 6674
    },
    {
      "epoch": 0.061246379920626405,
      "grad_norm": 0.205561101436615,
      "learning_rate": 0.0001,
      "loss": 1.2472,
      "step": 6675
    },
    {
      "epoch": 0.061353641531695806,
      "grad_norm": 0.19704127311706543,
      "learning_rate": 0.0001,
      "loss": 1.2343,
      "step": 6676
    },
    {
      "epoch": 0.06146090314276521,
      "grad_norm": 0.19591370224952698,
      "learning_rate": 0.0001,
      "loss": 1.2859,
      "step": 6677
    },
    {
      "epoch": 0.0615681647538346,
      "grad_norm": 0.19962163269519806,
      "learning_rate": 0.0001,
      "loss": 1.2519,
      "step": 6678
    },
    {
      "epoch": 0.061675426364904,
      "grad_norm": 0.20113089680671692,
      "learning_rate": 0.0001,
      "loss": 1.2918,
      "step": 6679
    },
    {
      "epoch": 0.0617826879759734,
      "grad_norm": 0.20801059901714325,
      "learning_rate": 0.0001,
      "loss": 1.2785,
      "step": 6680
    },
    {
      "epoch": 0.061889949587042796,
      "grad_norm": 0.21243323385715485,
      "learning_rate": 0.0001,
      "loss": 1.1568,
      "step": 6681
    },
    {
      "epoch": 0.0619972111981122,
      "grad_norm": 0.19983986020088196,
      "learning_rate": 0.0001,
      "loss": 1.3201,
      "step": 6682
    },
    {
      "epoch": 0.06210447280918159,
      "grad_norm": 0.20299598574638367,
      "learning_rate": 0.0001,
      "loss": 1.2767,
      "step": 6683
    },
    {
      "epoch": 0.06221173442025099,
      "grad_norm": 0.23540109395980835,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 6684
    },
    {
      "epoch": 0.06231899603132039,
      "grad_norm": 0.19764159619808197,
      "learning_rate": 0.0001,
      "loss": 1.1779,
      "step": 6685
    },
    {
      "epoch": 0.062426257642389786,
      "grad_norm": 0.20405392348766327,
      "learning_rate": 0.0001,
      "loss": 1.2985,
      "step": 6686
    },
    {
      "epoch": 0.0625335192534592,
      "grad_norm": 0.2204550802707672,
      "learning_rate": 0.0001,
      "loss": 1.2136,
      "step": 6687
    },
    {
      "epoch": 0.06264078086452858,
      "grad_norm": 0.20251953601837158,
      "learning_rate": 0.0001,
      "loss": 1.2552,
      "step": 6688
    },
    {
      "epoch": 0.06274804247559798,
      "grad_norm": 0.2095031589269638,
      "learning_rate": 0.0001,
      "loss": 1.1416,
      "step": 6689
    },
    {
      "epoch": 0.06285530408666738,
      "grad_norm": 0.21932631731033325,
      "learning_rate": 0.0001,
      "loss": 1.2548,
      "step": 6690
    },
    {
      "epoch": 0.06296256569773678,
      "grad_norm": 0.24361564218997955,
      "learning_rate": 0.0001,
      "loss": 1.2738,
      "step": 6691
    },
    {
      "epoch": 0.06306982730880618,
      "grad_norm": 0.22900396585464478,
      "learning_rate": 0.0001,
      "loss": 1.1969,
      "step": 6692
    },
    {
      "epoch": 0.06317708891987557,
      "grad_norm": 0.20887887477874756,
      "learning_rate": 0.0001,
      "loss": 1.2364,
      "step": 6693
    },
    {
      "epoch": 0.06328435053094497,
      "grad_norm": 0.20808562636375427,
      "learning_rate": 0.0001,
      "loss": 1.2056,
      "step": 6694
    },
    {
      "epoch": 0.06339161214201437,
      "grad_norm": 0.2156120240688324,
      "learning_rate": 0.0001,
      "loss": 1.1955,
      "step": 6695
    },
    {
      "epoch": 0.06349887375308377,
      "grad_norm": 0.1944403201341629,
      "learning_rate": 0.0001,
      "loss": 1.2168,
      "step": 6696
    },
    {
      "epoch": 0.06360613536415317,
      "grad_norm": 0.2140844315290451,
      "learning_rate": 0.0001,
      "loss": 1.2957,
      "step": 6697
    },
    {
      "epoch": 0.06371339697522256,
      "grad_norm": 0.20122727751731873,
      "learning_rate": 0.0001,
      "loss": 1.2926,
      "step": 6698
    },
    {
      "epoch": 0.06382065858629196,
      "grad_norm": 0.20170855522155762,
      "learning_rate": 0.0001,
      "loss": 1.3404,
      "step": 6699
    },
    {
      "epoch": 0.06392792019736136,
      "grad_norm": 0.20638249814510345,
      "learning_rate": 0.0001,
      "loss": 1.2433,
      "step": 6700
    },
    {
      "epoch": 0.06403518180843076,
      "grad_norm": 0.19190165400505066,
      "learning_rate": 0.0001,
      "loss": 1.1315,
      "step": 6701
    },
    {
      "epoch": 0.06414244341950016,
      "grad_norm": 0.2097727209329605,
      "learning_rate": 0.0001,
      "loss": 1.2193,
      "step": 6702
    },
    {
      "epoch": 0.06424970503056956,
      "grad_norm": 0.20811833441257477,
      "learning_rate": 0.0001,
      "loss": 1.1988,
      "step": 6703
    },
    {
      "epoch": 0.06435696664163895,
      "grad_norm": 0.20133881270885468,
      "learning_rate": 0.0001,
      "loss": 1.2489,
      "step": 6704
    },
    {
      "epoch": 0.06446422825270835,
      "grad_norm": 0.2196316421031952,
      "learning_rate": 0.0001,
      "loss": 1.2272,
      "step": 6705
    },
    {
      "epoch": 0.06457148986377775,
      "grad_norm": 0.21474294364452362,
      "learning_rate": 0.0001,
      "loss": 1.3547,
      "step": 6706
    },
    {
      "epoch": 0.06467875147484715,
      "grad_norm": 0.1903819739818573,
      "learning_rate": 0.0001,
      "loss": 1.1631,
      "step": 6707
    },
    {
      "epoch": 0.06478601308591656,
      "grad_norm": 0.20770297944545746,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 6708
    },
    {
      "epoch": 0.06489327469698594,
      "grad_norm": 0.214143306016922,
      "learning_rate": 0.0001,
      "loss": 1.3116,
      "step": 6709
    },
    {
      "epoch": 0.06500053630805534,
      "grad_norm": 0.19081000983715057,
      "learning_rate": 0.0001,
      "loss": 1.1461,
      "step": 6710
    },
    {
      "epoch": 0.06510779791912474,
      "grad_norm": 0.21525467932224274,
      "learning_rate": 0.0001,
      "loss": 1.2406,
      "step": 6711
    },
    {
      "epoch": 0.06521505953019414,
      "grad_norm": 0.21915528178215027,
      "learning_rate": 0.0001,
      "loss": 1.2731,
      "step": 6712
    },
    {
      "epoch": 0.06532232114126355,
      "grad_norm": 0.22151325643062592,
      "learning_rate": 0.0001,
      "loss": 1.2932,
      "step": 6713
    },
    {
      "epoch": 0.06542958275233295,
      "grad_norm": 0.21057818830013275,
      "learning_rate": 0.0001,
      "loss": 1.2704,
      "step": 6714
    },
    {
      "epoch": 0.06553684436340233,
      "grad_norm": 0.21205903589725494,
      "learning_rate": 0.0001,
      "loss": 1.3418,
      "step": 6715
    },
    {
      "epoch": 0.06564410597447173,
      "grad_norm": 0.21628545224666595,
      "learning_rate": 0.0001,
      "loss": 1.257,
      "step": 6716
    },
    {
      "epoch": 0.06575136758554113,
      "grad_norm": 0.2200598120689392,
      "learning_rate": 0.0001,
      "loss": 1.3203,
      "step": 6717
    },
    {
      "epoch": 0.06585862919661054,
      "grad_norm": 0.20088616013526917,
      "learning_rate": 0.0001,
      "loss": 1.152,
      "step": 6718
    },
    {
      "epoch": 0.06596589080767994,
      "grad_norm": 0.21555843949317932,
      "learning_rate": 0.0001,
      "loss": 1.2603,
      "step": 6719
    },
    {
      "epoch": 0.06607315241874932,
      "grad_norm": 0.23004518449306488,
      "learning_rate": 0.0001,
      "loss": 1.3072,
      "step": 6720
    },
    {
      "epoch": 0.06618041402981872,
      "grad_norm": 0.21153129637241364,
      "learning_rate": 0.0001,
      "loss": 1.2546,
      "step": 6721
    },
    {
      "epoch": 0.06628767564088812,
      "grad_norm": 0.2141055166721344,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 6722
    },
    {
      "epoch": 0.06639493725195753,
      "grad_norm": 0.2101924568414688,
      "learning_rate": 0.0001,
      "loss": 1.2941,
      "step": 6723
    },
    {
      "epoch": 0.06650219886302693,
      "grad_norm": 0.20355184376239777,
      "learning_rate": 0.0001,
      "loss": 1.2907,
      "step": 6724
    },
    {
      "epoch": 0.06660946047409633,
      "grad_norm": 0.19924819469451904,
      "learning_rate": 0.0001,
      "loss": 1.2495,
      "step": 6725
    },
    {
      "epoch": 0.06671672208516571,
      "grad_norm": 0.21447542309761047,
      "learning_rate": 0.0001,
      "loss": 1.2558,
      "step": 6726
    },
    {
      "epoch": 0.06682398369623511,
      "grad_norm": 0.2238588035106659,
      "learning_rate": 0.0001,
      "loss": 1.3458,
      "step": 6727
    },
    {
      "epoch": 0.06693124530730452,
      "grad_norm": 0.19425322115421295,
      "learning_rate": 0.0001,
      "loss": 1.1464,
      "step": 6728
    },
    {
      "epoch": 0.06703850691837392,
      "grad_norm": 0.2269909828901291,
      "learning_rate": 0.0001,
      "loss": 1.1798,
      "step": 6729
    },
    {
      "epoch": 0.06714576852944332,
      "grad_norm": 0.212100088596344,
      "learning_rate": 0.0001,
      "loss": 1.29,
      "step": 6730
    },
    {
      "epoch": 0.0672530301405127,
      "grad_norm": 0.20281638205051422,
      "learning_rate": 0.0001,
      "loss": 1.2829,
      "step": 6731
    },
    {
      "epoch": 0.0673602917515821,
      "grad_norm": 0.21290093660354614,
      "learning_rate": 0.0001,
      "loss": 1.3644,
      "step": 6732
    },
    {
      "epoch": 0.0674675533626515,
      "grad_norm": 0.2052324116230011,
      "learning_rate": 0.0001,
      "loss": 1.3114,
      "step": 6733
    },
    {
      "epoch": 0.0675748149737209,
      "grad_norm": 0.20896103978157043,
      "learning_rate": 0.0001,
      "loss": 1.3139,
      "step": 6734
    },
    {
      "epoch": 0.06768207658479031,
      "grad_norm": 0.19818106293678284,
      "learning_rate": 0.0001,
      "loss": 1.2967,
      "step": 6735
    },
    {
      "epoch": 0.06778933819585971,
      "grad_norm": 0.2105032354593277,
      "learning_rate": 0.0001,
      "loss": 1.3017,
      "step": 6736
    },
    {
      "epoch": 0.0678965998069291,
      "grad_norm": 0.21559283137321472,
      "learning_rate": 0.0001,
      "loss": 1.2497,
      "step": 6737
    },
    {
      "epoch": 0.0680038614179985,
      "grad_norm": 0.19357289373874664,
      "learning_rate": 0.0001,
      "loss": 1.21,
      "step": 6738
    },
    {
      "epoch": 0.0681111230290679,
      "grad_norm": 0.21511398255825043,
      "learning_rate": 0.0001,
      "loss": 1.2961,
      "step": 6739
    },
    {
      "epoch": 0.0682183846401373,
      "grad_norm": 0.22976596653461456,
      "learning_rate": 0.0001,
      "loss": 1.3438,
      "step": 6740
    },
    {
      "epoch": 0.0683256462512067,
      "grad_norm": 0.20716087520122528,
      "learning_rate": 0.0001,
      "loss": 1.3126,
      "step": 6741
    },
    {
      "epoch": 0.06843290786227609,
      "grad_norm": 0.2320764660835266,
      "learning_rate": 0.0001,
      "loss": 1.2322,
      "step": 6742
    },
    {
      "epoch": 0.06854016947334549,
      "grad_norm": 0.21846400201320648,
      "learning_rate": 0.0001,
      "loss": 1.2882,
      "step": 6743
    },
    {
      "epoch": 0.06864743108441489,
      "grad_norm": 0.1942097544670105,
      "learning_rate": 0.0001,
      "loss": 1.2236,
      "step": 6744
    },
    {
      "epoch": 0.06875469269548429,
      "grad_norm": 0.20684002339839935,
      "learning_rate": 0.0001,
      "loss": 1.2744,
      "step": 6745
    },
    {
      "epoch": 0.06886195430655369,
      "grad_norm": 0.2384331226348877,
      "learning_rate": 0.0001,
      "loss": 1.3322,
      "step": 6746
    },
    {
      "epoch": 0.06896921591762309,
      "grad_norm": 0.21220791339874268,
      "learning_rate": 0.0001,
      "loss": 1.1986,
      "step": 6747
    },
    {
      "epoch": 0.06907647752869248,
      "grad_norm": 0.208564892411232,
      "learning_rate": 0.0001,
      "loss": 1.2436,
      "step": 6748
    },
    {
      "epoch": 0.06918373913976188,
      "grad_norm": 0.2266736924648285,
      "learning_rate": 0.0001,
      "loss": 1.221,
      "step": 6749
    },
    {
      "epoch": 0.06929100075083128,
      "grad_norm": 0.23634812235832214,
      "learning_rate": 0.0001,
      "loss": 1.287,
      "step": 6750
    },
    {
      "epoch": 0.06939826236190068,
      "grad_norm": 0.22221341729164124,
      "learning_rate": 0.0001,
      "loss": 1.2099,
      "step": 6751
    },
    {
      "epoch": 0.06950552397297008,
      "grad_norm": 0.20796647667884827,
      "learning_rate": 0.0001,
      "loss": 1.2711,
      "step": 6752
    },
    {
      "epoch": 0.06961278558403947,
      "grad_norm": 0.21422341465950012,
      "learning_rate": 0.0001,
      "loss": 1.3674,
      "step": 6753
    },
    {
      "epoch": 0.06972004719510887,
      "grad_norm": 0.20421168208122253,
      "learning_rate": 0.0001,
      "loss": 1.3187,
      "step": 6754
    },
    {
      "epoch": 0.06982730880617827,
      "grad_norm": 0.20531372725963593,
      "learning_rate": 0.0001,
      "loss": 1.2757,
      "step": 6755
    },
    {
      "epoch": 0.06993457041724767,
      "grad_norm": 0.20241886377334595,
      "learning_rate": 0.0001,
      "loss": 1.2748,
      "step": 6756
    },
    {
      "epoch": 0.07004183202831707,
      "grad_norm": 0.2164398580789566,
      "learning_rate": 0.0001,
      "loss": 1.2725,
      "step": 6757
    },
    {
      "epoch": 0.07014909363938647,
      "grad_norm": 0.23333653807640076,
      "learning_rate": 0.0001,
      "loss": 1.2238,
      "step": 6758
    },
    {
      "epoch": 0.07025635525045586,
      "grad_norm": 0.1981683075428009,
      "learning_rate": 0.0001,
      "loss": 1.2017,
      "step": 6759
    },
    {
      "epoch": 0.07036361686152526,
      "grad_norm": 0.20356114208698273,
      "learning_rate": 0.0001,
      "loss": 1.2115,
      "step": 6760
    },
    {
      "epoch": 0.07047087847259466,
      "grad_norm": 0.19593428075313568,
      "learning_rate": 0.0001,
      "loss": 1.0863,
      "step": 6761
    },
    {
      "epoch": 0.07057814008366406,
      "grad_norm": 0.2108919620513916,
      "learning_rate": 0.0001,
      "loss": 1.2406,
      "step": 6762
    },
    {
      "epoch": 0.07068540169473346,
      "grad_norm": 0.2030707597732544,
      "learning_rate": 0.0001,
      "loss": 1.1753,
      "step": 6763
    },
    {
      "epoch": 0.07079266330580285,
      "grad_norm": 0.1948305368423462,
      "learning_rate": 0.0001,
      "loss": 1.2376,
      "step": 6764
    },
    {
      "epoch": 0.07089992491687225,
      "grad_norm": 0.20492315292358398,
      "learning_rate": 0.0001,
      "loss": 1.2851,
      "step": 6765
    },
    {
      "epoch": 0.07100718652794165,
      "grad_norm": 0.22504675388336182,
      "learning_rate": 0.0001,
      "loss": 1.3601,
      "step": 6766
    },
    {
      "epoch": 0.07111444813901105,
      "grad_norm": 0.20682071149349213,
      "learning_rate": 0.0001,
      "loss": 1.2815,
      "step": 6767
    },
    {
      "epoch": 0.07122170975008045,
      "grad_norm": 0.22509431838989258,
      "learning_rate": 0.0001,
      "loss": 1.2149,
      "step": 6768
    },
    {
      "epoch": 0.07132897136114985,
      "grad_norm": 0.20415934920310974,
      "learning_rate": 0.0001,
      "loss": 1.2805,
      "step": 6769
    },
    {
      "epoch": 0.07143623297221924,
      "grad_norm": 0.21055597066879272,
      "learning_rate": 0.0001,
      "loss": 1.23,
      "step": 6770
    },
    {
      "epoch": 0.07154349458328864,
      "grad_norm": 0.21216276288032532,
      "learning_rate": 0.0001,
      "loss": 1.3213,
      "step": 6771
    },
    {
      "epoch": 0.07165075619435804,
      "grad_norm": 0.20328737795352936,
      "learning_rate": 0.0001,
      "loss": 1.2269,
      "step": 6772
    },
    {
      "epoch": 0.07175801780542744,
      "grad_norm": 0.20402726531028748,
      "learning_rate": 0.0001,
      "loss": 1.2456,
      "step": 6773
    },
    {
      "epoch": 0.07186527941649684,
      "grad_norm": 0.19800147414207458,
      "learning_rate": 0.0001,
      "loss": 1.2419,
      "step": 6774
    },
    {
      "epoch": 0.07197254102756623,
      "grad_norm": 0.21936191618442535,
      "learning_rate": 0.0001,
      "loss": 1.2879,
      "step": 6775
    },
    {
      "epoch": 0.07207980263863563,
      "grad_norm": 0.20396055281162262,
      "learning_rate": 0.0001,
      "loss": 1.3019,
      "step": 6776
    },
    {
      "epoch": 0.07218706424970503,
      "grad_norm": 0.225092351436615,
      "learning_rate": 0.0001,
      "loss": 1.3082,
      "step": 6777
    },
    {
      "epoch": 0.07229432586077443,
      "grad_norm": 0.2069796472787857,
      "learning_rate": 0.0001,
      "loss": 1.3695,
      "step": 6778
    },
    {
      "epoch": 0.07240158747184383,
      "grad_norm": 0.19848458468914032,
      "learning_rate": 0.0001,
      "loss": 1.2364,
      "step": 6779
    },
    {
      "epoch": 0.07250884908291323,
      "grad_norm": 0.2055046260356903,
      "learning_rate": 0.0001,
      "loss": 1.3482,
      "step": 6780
    },
    {
      "epoch": 0.07261611069398262,
      "grad_norm": 0.21235422790050507,
      "learning_rate": 0.0001,
      "loss": 1.3108,
      "step": 6781
    },
    {
      "epoch": 0.07272337230505202,
      "grad_norm": 0.20798029005527496,
      "learning_rate": 0.0001,
      "loss": 1.2855,
      "step": 6782
    },
    {
      "epoch": 0.07283063391612142,
      "grad_norm": 0.20485571026802063,
      "learning_rate": 0.0001,
      "loss": 1.2357,
      "step": 6783
    },
    {
      "epoch": 0.07293789552719082,
      "grad_norm": 0.2037992924451828,
      "learning_rate": 0.0001,
      "loss": 1.2329,
      "step": 6784
    },
    {
      "epoch": 0.07304515713826022,
      "grad_norm": 0.21481135487556458,
      "learning_rate": 0.0001,
      "loss": 1.2985,
      "step": 6785
    },
    {
      "epoch": 0.07315241874932961,
      "grad_norm": 0.20337450504302979,
      "learning_rate": 0.0001,
      "loss": 1.3242,
      "step": 6786
    },
    {
      "epoch": 0.07325968036039901,
      "grad_norm": 0.2065802961587906,
      "learning_rate": 0.0001,
      "loss": 1.2173,
      "step": 6787
    },
    {
      "epoch": 0.07336694197146841,
      "grad_norm": 0.21670714020729065,
      "learning_rate": 0.0001,
      "loss": 1.2773,
      "step": 6788
    },
    {
      "epoch": 0.07347420358253781,
      "grad_norm": 0.21832212805747986,
      "learning_rate": 0.0001,
      "loss": 1.2443,
      "step": 6789
    },
    {
      "epoch": 0.07358146519360721,
      "grad_norm": 0.21100521087646484,
      "learning_rate": 0.0001,
      "loss": 1.2453,
      "step": 6790
    },
    {
      "epoch": 0.07368872680467661,
      "grad_norm": 0.21534955501556396,
      "learning_rate": 0.0001,
      "loss": 1.1835,
      "step": 6791
    },
    {
      "epoch": 0.073795988415746,
      "grad_norm": 0.20998810231685638,
      "learning_rate": 0.0001,
      "loss": 1.3181,
      "step": 6792
    },
    {
      "epoch": 0.0739032500268154,
      "grad_norm": 0.21616627275943756,
      "learning_rate": 0.0001,
      "loss": 1.2337,
      "step": 6793
    },
    {
      "epoch": 0.0740105116378848,
      "grad_norm": 0.24178902804851532,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 6794
    },
    {
      "epoch": 0.0741177732489542,
      "grad_norm": 0.21081797778606415,
      "learning_rate": 0.0001,
      "loss": 1.1802,
      "step": 6795
    },
    {
      "epoch": 0.0742250348600236,
      "grad_norm": 0.19732706248760223,
      "learning_rate": 0.0001,
      "loss": 1.1702,
      "step": 6796
    },
    {
      "epoch": 0.07433229647109299,
      "grad_norm": 0.2054906040430069,
      "learning_rate": 0.0001,
      "loss": 1.3413,
      "step": 6797
    },
    {
      "epoch": 0.07443955808216239,
      "grad_norm": 0.19626563787460327,
      "learning_rate": 0.0001,
      "loss": 1.2309,
      "step": 6798
    },
    {
      "epoch": 0.07454681969323179,
      "grad_norm": 0.2038876712322235,
      "learning_rate": 0.0001,
      "loss": 1.2899,
      "step": 6799
    },
    {
      "epoch": 0.07465408130430119,
      "grad_norm": 0.208580881357193,
      "learning_rate": 0.0001,
      "loss": 1.3032,
      "step": 6800
    },
    {
      "epoch": 0.0747613429153706,
      "grad_norm": 0.21136066317558289,
      "learning_rate": 0.0001,
      "loss": 1.2532,
      "step": 6801
    },
    {
      "epoch": 0.07486860452643998,
      "grad_norm": 0.20835670828819275,
      "learning_rate": 0.0001,
      "loss": 1.369,
      "step": 6802
    },
    {
      "epoch": 0.07497586613750938,
      "grad_norm": 0.1989949643611908,
      "learning_rate": 0.0001,
      "loss": 1.1788,
      "step": 6803
    },
    {
      "epoch": 0.07508312774857878,
      "grad_norm": 0.21013732254505157,
      "learning_rate": 0.0001,
      "loss": 1.2133,
      "step": 6804
    },
    {
      "epoch": 0.07519038935964818,
      "grad_norm": 0.2012101709842682,
      "learning_rate": 0.0001,
      "loss": 1.2027,
      "step": 6805
    },
    {
      "epoch": 0.07529765097071758,
      "grad_norm": 0.21837663650512695,
      "learning_rate": 0.0001,
      "loss": 1.342,
      "step": 6806
    },
    {
      "epoch": 0.07540491258178698,
      "grad_norm": 0.20693524181842804,
      "learning_rate": 0.0001,
      "loss": 1.3312,
      "step": 6807
    },
    {
      "epoch": 0.07551217419285637,
      "grad_norm": 0.21874520182609558,
      "learning_rate": 0.0001,
      "loss": 1.2893,
      "step": 6808
    },
    {
      "epoch": 0.07561943580392577,
      "grad_norm": 0.21324707567691803,
      "learning_rate": 0.0001,
      "loss": 1.3924,
      "step": 6809
    },
    {
      "epoch": 0.07572669741499517,
      "grad_norm": 0.20759959518909454,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 6810
    },
    {
      "epoch": 0.07583395902606457,
      "grad_norm": 0.20314756035804749,
      "learning_rate": 0.0001,
      "loss": 1.2604,
      "step": 6811
    },
    {
      "epoch": 0.07594122063713397,
      "grad_norm": 0.21288353204727173,
      "learning_rate": 0.0001,
      "loss": 1.3115,
      "step": 6812
    },
    {
      "epoch": 0.07604848224820336,
      "grad_norm": 0.21782530844211578,
      "learning_rate": 0.0001,
      "loss": 1.2149,
      "step": 6813
    },
    {
      "epoch": 0.07615574385927276,
      "grad_norm": 0.21416328847408295,
      "learning_rate": 0.0001,
      "loss": 1.2183,
      "step": 6814
    },
    {
      "epoch": 0.07626300547034216,
      "grad_norm": 0.21807348728179932,
      "learning_rate": 0.0001,
      "loss": 1.3037,
      "step": 6815
    },
    {
      "epoch": 0.07637026708141156,
      "grad_norm": 0.21556240320205688,
      "learning_rate": 0.0001,
      "loss": 1.2368,
      "step": 6816
    },
    {
      "epoch": 0.07647752869248096,
      "grad_norm": 0.20345638692378998,
      "learning_rate": 0.0001,
      "loss": 1.23,
      "step": 6817
    },
    {
      "epoch": 0.07658479030355037,
      "grad_norm": 0.22400127351284027,
      "learning_rate": 0.0001,
      "loss": 1.3392,
      "step": 6818
    },
    {
      "epoch": 0.07669205191461975,
      "grad_norm": 0.20219965279102325,
      "learning_rate": 0.0001,
      "loss": 1.3542,
      "step": 6819
    },
    {
      "epoch": 0.07679931352568915,
      "grad_norm": 0.21518094837665558,
      "learning_rate": 0.0001,
      "loss": 1.2208,
      "step": 6820
    },
    {
      "epoch": 0.07690657513675855,
      "grad_norm": 0.20603600144386292,
      "learning_rate": 0.0001,
      "loss": 1.2645,
      "step": 6821
    },
    {
      "epoch": 0.07701383674782795,
      "grad_norm": 0.20295265316963196,
      "learning_rate": 0.0001,
      "loss": 1.2622,
      "step": 6822
    },
    {
      "epoch": 0.07712109835889736,
      "grad_norm": 0.22197726368904114,
      "learning_rate": 0.0001,
      "loss": 1.2077,
      "step": 6823
    },
    {
      "epoch": 0.07722835996996674,
      "grad_norm": 0.22839680314064026,
      "learning_rate": 0.0001,
      "loss": 1.2985,
      "step": 6824
    },
    {
      "epoch": 0.07733562158103614,
      "grad_norm": 0.21254132688045502,
      "learning_rate": 0.0001,
      "loss": 1.2447,
      "step": 6825
    },
    {
      "epoch": 0.07744288319210554,
      "grad_norm": 0.21479246020317078,
      "learning_rate": 0.0001,
      "loss": 1.2302,
      "step": 6826
    },
    {
      "epoch": 0.07755014480317494,
      "grad_norm": 0.21713565289974213,
      "learning_rate": 0.0001,
      "loss": 1.3524,
      "step": 6827
    },
    {
      "epoch": 0.07765740641424435,
      "grad_norm": 0.21053829789161682,
      "learning_rate": 0.0001,
      "loss": 1.2115,
      "step": 6828
    },
    {
      "epoch": 0.07776466802531375,
      "grad_norm": 0.2057601660490036,
      "learning_rate": 0.0001,
      "loss": 1.2719,
      "step": 6829
    },
    {
      "epoch": 0.07787192963638313,
      "grad_norm": 0.2004348784685135,
      "learning_rate": 0.0001,
      "loss": 1.3481,
      "step": 6830
    },
    {
      "epoch": 0.07797919124745253,
      "grad_norm": 0.20758037269115448,
      "learning_rate": 0.0001,
      "loss": 1.3226,
      "step": 6831
    },
    {
      "epoch": 0.07808645285852193,
      "grad_norm": 0.20946104824543,
      "learning_rate": 0.0001,
      "loss": 1.2881,
      "step": 6832
    },
    {
      "epoch": 0.07819371446959134,
      "grad_norm": 0.20027771592140198,
      "learning_rate": 0.0001,
      "loss": 1.2142,
      "step": 6833
    },
    {
      "epoch": 0.07830097608066074,
      "grad_norm": 0.21104945242404938,
      "learning_rate": 0.0001,
      "loss": 1.2631,
      "step": 6834
    },
    {
      "epoch": 0.07840823769173012,
      "grad_norm": 0.20132197439670563,
      "learning_rate": 0.0001,
      "loss": 1.1826,
      "step": 6835
    },
    {
      "epoch": 0.07851549930279952,
      "grad_norm": 0.20926308631896973,
      "learning_rate": 0.0001,
      "loss": 1.1645,
      "step": 6836
    },
    {
      "epoch": 0.07862276091386892,
      "grad_norm": 0.21005438268184662,
      "learning_rate": 0.0001,
      "loss": 1.3449,
      "step": 6837
    },
    {
      "epoch": 0.07873002252493833,
      "grad_norm": 0.20999878644943237,
      "learning_rate": 0.0001,
      "loss": 1.3134,
      "step": 6838
    },
    {
      "epoch": 0.07883728413600773,
      "grad_norm": 0.21383635699748993,
      "learning_rate": 0.0001,
      "loss": 1.3139,
      "step": 6839
    },
    {
      "epoch": 0.07894454574707713,
      "grad_norm": 0.20932577550411224,
      "learning_rate": 0.0001,
      "loss": 1.306,
      "step": 6840
    },
    {
      "epoch": 0.07905180735814651,
      "grad_norm": 0.21745724976062775,
      "learning_rate": 0.0001,
      "loss": 1.2087,
      "step": 6841
    },
    {
      "epoch": 0.07915906896921591,
      "grad_norm": 0.2153574824333191,
      "learning_rate": 0.0001,
      "loss": 1.262,
      "step": 6842
    },
    {
      "epoch": 0.07926633058028532,
      "grad_norm": 0.20507581532001495,
      "learning_rate": 0.0001,
      "loss": 1.1338,
      "step": 6843
    },
    {
      "epoch": 0.07937359219135472,
      "grad_norm": 0.22418177127838135,
      "learning_rate": 0.0001,
      "loss": 1.2807,
      "step": 6844
    },
    {
      "epoch": 0.07948085380242412,
      "grad_norm": 0.2057168036699295,
      "learning_rate": 0.0001,
      "loss": 1.2664,
      "step": 6845
    },
    {
      "epoch": 0.0795881154134935,
      "grad_norm": 0.19971846044063568,
      "learning_rate": 0.0001,
      "loss": 1.3231,
      "step": 6846
    },
    {
      "epoch": 0.0796953770245629,
      "grad_norm": 0.2116224318742752,
      "learning_rate": 0.0001,
      "loss": 1.3619,
      "step": 6847
    },
    {
      "epoch": 0.0798026386356323,
      "grad_norm": 0.21259769797325134,
      "learning_rate": 0.0001,
      "loss": 1.3332,
      "step": 6848
    },
    {
      "epoch": 0.0799099002467017,
      "grad_norm": 0.192108154296875,
      "learning_rate": 0.0001,
      "loss": 1.2783,
      "step": 6849
    },
    {
      "epoch": 0.08001716185777111,
      "grad_norm": 0.21051721274852753,
      "learning_rate": 0.0001,
      "loss": 1.2618,
      "step": 6850
    },
    {
      "epoch": 0.08012442346884051,
      "grad_norm": 0.2021142989397049,
      "learning_rate": 0.0001,
      "loss": 1.2264,
      "step": 6851
    },
    {
      "epoch": 0.0802316850799099,
      "grad_norm": 0.2169874608516693,
      "learning_rate": 0.0001,
      "loss": 1.2052,
      "step": 6852
    },
    {
      "epoch": 0.0803389466909793,
      "grad_norm": 0.20690666139125824,
      "learning_rate": 0.0001,
      "loss": 1.2348,
      "step": 6853
    },
    {
      "epoch": 0.0804462083020487,
      "grad_norm": 0.21349063515663147,
      "learning_rate": 0.0001,
      "loss": 1.337,
      "step": 6854
    },
    {
      "epoch": 0.0805534699131181,
      "grad_norm": 0.21485643088817596,
      "learning_rate": 0.0001,
      "loss": 1.2542,
      "step": 6855
    },
    {
      "epoch": 0.0806607315241875,
      "grad_norm": 0.20039376616477966,
      "learning_rate": 0.0001,
      "loss": 1.3138,
      "step": 6856
    },
    {
      "epoch": 0.08076799313525689,
      "grad_norm": 0.2093849927186966,
      "learning_rate": 0.0001,
      "loss": 1.2215,
      "step": 6857
    },
    {
      "epoch": 0.08087525474632629,
      "grad_norm": 0.21955248713493347,
      "learning_rate": 0.0001,
      "loss": 1.2532,
      "step": 6858
    },
    {
      "epoch": 0.08098251635739569,
      "grad_norm": 0.21979402005672455,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 6859
    },
    {
      "epoch": 0.08108977796846509,
      "grad_norm": 0.2225266546010971,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 6860
    },
    {
      "epoch": 0.08119703957953449,
      "grad_norm": 0.23164215683937073,
      "learning_rate": 0.0001,
      "loss": 1.2967,
      "step": 6861
    },
    {
      "epoch": 0.08130430119060389,
      "grad_norm": 0.21872524917125702,
      "learning_rate": 0.0001,
      "loss": 1.3053,
      "step": 6862
    },
    {
      "epoch": 0.08141156280167328,
      "grad_norm": 0.21057257056236267,
      "learning_rate": 0.0001,
      "loss": 1.2167,
      "step": 6863
    },
    {
      "epoch": 0.08151882441274268,
      "grad_norm": 0.19789132475852966,
      "learning_rate": 0.0001,
      "loss": 1.2369,
      "step": 6864
    },
    {
      "epoch": 0.08162608602381208,
      "grad_norm": 0.2224435657262802,
      "learning_rate": 0.0001,
      "loss": 1.3044,
      "step": 6865
    },
    {
      "epoch": 0.08173334763488148,
      "grad_norm": 0.21030688285827637,
      "learning_rate": 0.0001,
      "loss": 1.2044,
      "step": 6866
    },
    {
      "epoch": 0.08184060924595088,
      "grad_norm": 0.21891430020332336,
      "learning_rate": 0.0001,
      "loss": 1.2645,
      "step": 6867
    },
    {
      "epoch": 0.08194787085702027,
      "grad_norm": 0.20965689420700073,
      "learning_rate": 0.0001,
      "loss": 1.3522,
      "step": 6868
    },
    {
      "epoch": 0.08205513246808967,
      "grad_norm": 0.2097690999507904,
      "learning_rate": 0.0001,
      "loss": 1.2114,
      "step": 6869
    },
    {
      "epoch": 0.08216239407915907,
      "grad_norm": 0.21419501304626465,
      "learning_rate": 0.0001,
      "loss": 1.3359,
      "step": 6870
    },
    {
      "epoch": 0.08226965569022847,
      "grad_norm": 0.21591618657112122,
      "learning_rate": 0.0001,
      "loss": 1.3319,
      "step": 6871
    },
    {
      "epoch": 0.08237691730129787,
      "grad_norm": 0.2165345400571823,
      "learning_rate": 0.0001,
      "loss": 1.2417,
      "step": 6872
    },
    {
      "epoch": 0.08248417891236727,
      "grad_norm": 0.20818965137004852,
      "learning_rate": 0.0001,
      "loss": 1.2368,
      "step": 6873
    },
    {
      "epoch": 0.08259144052343666,
      "grad_norm": 0.21708816289901733,
      "learning_rate": 0.0001,
      "loss": 1.241,
      "step": 6874
    },
    {
      "epoch": 0.08269870213450606,
      "grad_norm": 0.21922336518764496,
      "learning_rate": 0.0001,
      "loss": 1.2514,
      "step": 6875
    },
    {
      "epoch": 0.08280596374557546,
      "grad_norm": 0.20558515191078186,
      "learning_rate": 0.0001,
      "loss": 1.2538,
      "step": 6876
    },
    {
      "epoch": 0.08291322535664486,
      "grad_norm": 0.2160593718290329,
      "learning_rate": 0.0001,
      "loss": 1.2831,
      "step": 6877
    },
    {
      "epoch": 0.08302048696771426,
      "grad_norm": 0.21629351377487183,
      "learning_rate": 0.0001,
      "loss": 1.2063,
      "step": 6878
    },
    {
      "epoch": 0.08312774857878365,
      "grad_norm": 0.20306211709976196,
      "learning_rate": 0.0001,
      "loss": 1.2874,
      "step": 6879
    },
    {
      "epoch": 0.08323501018985305,
      "grad_norm": 0.2073632925748825,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 6880
    },
    {
      "epoch": 0.08334227180092245,
      "grad_norm": 0.20167683064937592,
      "learning_rate": 0.0001,
      "loss": 1.2818,
      "step": 6881
    },
    {
      "epoch": 0.08344953341199185,
      "grad_norm": 0.20755524933338165,
      "learning_rate": 0.0001,
      "loss": 1.2461,
      "step": 6882
    },
    {
      "epoch": 0.08355679502306125,
      "grad_norm": 0.21193119883537292,
      "learning_rate": 0.0001,
      "loss": 1.1576,
      "step": 6883
    },
    {
      "epoch": 0.08366405663413065,
      "grad_norm": 0.19815170764923096,
      "learning_rate": 0.0001,
      "loss": 1.216,
      "step": 6884
    },
    {
      "epoch": 0.08377131824520004,
      "grad_norm": 0.21456106007099152,
      "learning_rate": 0.0001,
      "loss": 1.3345,
      "step": 6885
    },
    {
      "epoch": 0.08387857985626944,
      "grad_norm": 0.21315552294254303,
      "learning_rate": 0.0001,
      "loss": 1.2656,
      "step": 6886
    },
    {
      "epoch": 0.08398584146733884,
      "grad_norm": 0.20941351354122162,
      "learning_rate": 0.0001,
      "loss": 1.2172,
      "step": 6887
    },
    {
      "epoch": 0.08409310307840824,
      "grad_norm": 0.2098013311624527,
      "learning_rate": 0.0001,
      "loss": 1.2851,
      "step": 6888
    },
    {
      "epoch": 0.08420036468947764,
      "grad_norm": 0.20719179511070251,
      "learning_rate": 0.0001,
      "loss": 1.177,
      "step": 6889
    },
    {
      "epoch": 0.08430762630054703,
      "grad_norm": 0.21445722877979279,
      "learning_rate": 0.0001,
      "loss": 1.2849,
      "step": 6890
    },
    {
      "epoch": 0.08441488791161643,
      "grad_norm": 0.21011963486671448,
      "learning_rate": 0.0001,
      "loss": 1.2728,
      "step": 6891
    },
    {
      "epoch": 0.08452214952268583,
      "grad_norm": 0.1977025866508484,
      "learning_rate": 0.0001,
      "loss": 1.1318,
      "step": 6892
    },
    {
      "epoch": 0.08462941113375523,
      "grad_norm": 0.2062423676252365,
      "learning_rate": 0.0001,
      "loss": 1.299,
      "step": 6893
    },
    {
      "epoch": 0.08473667274482463,
      "grad_norm": 0.22689256072044373,
      "learning_rate": 0.0001,
      "loss": 1.3818,
      "step": 6894
    },
    {
      "epoch": 0.08484393435589403,
      "grad_norm": 0.19806723296642303,
      "learning_rate": 0.0001,
      "loss": 1.2374,
      "step": 6895
    },
    {
      "epoch": 0.08495119596696342,
      "grad_norm": 0.20516157150268555,
      "learning_rate": 0.0001,
      "loss": 1.2077,
      "step": 6896
    },
    {
      "epoch": 0.08505845757803282,
      "grad_norm": 0.2228326052427292,
      "learning_rate": 0.0001,
      "loss": 1.1336,
      "step": 6897
    },
    {
      "epoch": 0.08516571918910222,
      "grad_norm": 0.20380131900310516,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 6898
    },
    {
      "epoch": 0.08527298080017162,
      "grad_norm": 0.21294976770877838,
      "learning_rate": 0.0001,
      "loss": 1.2936,
      "step": 6899
    },
    {
      "epoch": 0.08538024241124102,
      "grad_norm": 0.2226545661687851,
      "learning_rate": 0.0001,
      "loss": 1.303,
      "step": 6900
    },
    {
      "epoch": 0.08548750402231041,
      "grad_norm": 0.20995016396045685,
      "learning_rate": 0.0001,
      "loss": 1.3078,
      "step": 6901
    },
    {
      "epoch": 0.08559476563337981,
      "grad_norm": 0.20867566764354706,
      "learning_rate": 0.0001,
      "loss": 1.321,
      "step": 6902
    },
    {
      "epoch": 0.08570202724444921,
      "grad_norm": 0.21943379938602448,
      "learning_rate": 0.0001,
      "loss": 1.2794,
      "step": 6903
    },
    {
      "epoch": 0.08580928885551861,
      "grad_norm": 0.2064763754606247,
      "learning_rate": 0.0001,
      "loss": 1.3737,
      "step": 6904
    },
    {
      "epoch": 0.08591655046658801,
      "grad_norm": 0.20807822048664093,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 6905
    },
    {
      "epoch": 0.08602381207765741,
      "grad_norm": 0.19703896343708038,
      "learning_rate": 0.0001,
      "loss": 1.1584,
      "step": 6906
    },
    {
      "epoch": 0.0861310736887268,
      "grad_norm": 0.2009844183921814,
      "learning_rate": 0.0001,
      "loss": 1.2616,
      "step": 6907
    },
    {
      "epoch": 0.0862383352997962,
      "grad_norm": 0.20970489084720612,
      "learning_rate": 0.0001,
      "loss": 1.2702,
      "step": 6908
    },
    {
      "epoch": 0.0863455969108656,
      "grad_norm": 0.22011297941207886,
      "learning_rate": 0.0001,
      "loss": 1.3545,
      "step": 6909
    },
    {
      "epoch": 0.086452858521935,
      "grad_norm": 0.20682942867279053,
      "learning_rate": 0.0001,
      "loss": 1.2741,
      "step": 6910
    },
    {
      "epoch": 0.0865601201330044,
      "grad_norm": 0.20747345685958862,
      "learning_rate": 0.0001,
      "loss": 1.2721,
      "step": 6911
    },
    {
      "epoch": 0.08666738174407379,
      "grad_norm": 0.21859921514987946,
      "learning_rate": 0.0001,
      "loss": 1.2434,
      "step": 6912
    },
    {
      "epoch": 0.08677464335514319,
      "grad_norm": 0.2125132530927658,
      "learning_rate": 0.0001,
      "loss": 1.2947,
      "step": 6913
    },
    {
      "epoch": 0.08688190496621259,
      "grad_norm": 0.22133535146713257,
      "learning_rate": 0.0001,
      "loss": 1.3113,
      "step": 6914
    },
    {
      "epoch": 0.08698916657728199,
      "grad_norm": 0.20808494091033936,
      "learning_rate": 0.0001,
      "loss": 1.408,
      "step": 6915
    },
    {
      "epoch": 0.0870964281883514,
      "grad_norm": 0.21456187963485718,
      "learning_rate": 0.0001,
      "loss": 1.2665,
      "step": 6916
    },
    {
      "epoch": 0.08720368979942078,
      "grad_norm": 0.21925339102745056,
      "learning_rate": 0.0001,
      "loss": 1.3725,
      "step": 6917
    },
    {
      "epoch": 0.08731095141049018,
      "grad_norm": 0.21121463179588318,
      "learning_rate": 0.0001,
      "loss": 1.2437,
      "step": 6918
    },
    {
      "epoch": 0.08741821302155958,
      "grad_norm": 0.21079926192760468,
      "learning_rate": 0.0001,
      "loss": 1.2811,
      "step": 6919
    },
    {
      "epoch": 0.08752547463262898,
      "grad_norm": 0.2214285433292389,
      "learning_rate": 0.0001,
      "loss": 1.2986,
      "step": 6920
    },
    {
      "epoch": 0.08763273624369838,
      "grad_norm": 0.20793890953063965,
      "learning_rate": 0.0001,
      "loss": 1.3236,
      "step": 6921
    },
    {
      "epoch": 0.08773999785476778,
      "grad_norm": 0.20421218872070312,
      "learning_rate": 0.0001,
      "loss": 1.2718,
      "step": 6922
    },
    {
      "epoch": 0.08784725946583717,
      "grad_norm": 0.2114454209804535,
      "learning_rate": 0.0001,
      "loss": 1.1902,
      "step": 6923
    },
    {
      "epoch": 0.08795452107690657,
      "grad_norm": 0.21286360919475555,
      "learning_rate": 0.0001,
      "loss": 1.297,
      "step": 6924
    },
    {
      "epoch": 0.08806178268797597,
      "grad_norm": 0.215081125497818,
      "learning_rate": 0.0001,
      "loss": 1.3084,
      "step": 6925
    },
    {
      "epoch": 0.08816904429904537,
      "grad_norm": 0.22301797568798065,
      "learning_rate": 0.0001,
      "loss": 1.3186,
      "step": 6926
    },
    {
      "epoch": 0.08827630591011477,
      "grad_norm": 0.20982462167739868,
      "learning_rate": 0.0001,
      "loss": 1.2508,
      "step": 6927
    },
    {
      "epoch": 0.08838356752118416,
      "grad_norm": 0.20902182161808014,
      "learning_rate": 0.0001,
      "loss": 1.2401,
      "step": 6928
    },
    {
      "epoch": 0.08849082913225356,
      "grad_norm": 0.20747965574264526,
      "learning_rate": 0.0001,
      "loss": 1.2577,
      "step": 6929
    },
    {
      "epoch": 0.08859809074332296,
      "grad_norm": 0.22279231250286102,
      "learning_rate": 0.0001,
      "loss": 1.3009,
      "step": 6930
    },
    {
      "epoch": 0.08870535235439236,
      "grad_norm": 0.2226146012544632,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 6931
    },
    {
      "epoch": 0.08881261396546176,
      "grad_norm": 0.21976789832115173,
      "learning_rate": 0.0001,
      "loss": 1.2121,
      "step": 6932
    },
    {
      "epoch": 0.08891987557653117,
      "grad_norm": 0.2157488912343979,
      "learning_rate": 0.0001,
      "loss": 1.2319,
      "step": 6933
    },
    {
      "epoch": 0.08902713718760055,
      "grad_norm": 0.21364039182662964,
      "learning_rate": 0.0001,
      "loss": 1.2771,
      "step": 6934
    },
    {
      "epoch": 0.08913439879866995,
      "grad_norm": 0.21153193712234497,
      "learning_rate": 0.0001,
      "loss": 1.3318,
      "step": 6935
    },
    {
      "epoch": 0.08924166040973935,
      "grad_norm": 0.22495612502098083,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 6936
    },
    {
      "epoch": 0.08934892202080875,
      "grad_norm": 0.20530955493450165,
      "learning_rate": 0.0001,
      "loss": 1.2864,
      "step": 6937
    },
    {
      "epoch": 0.08945618363187816,
      "grad_norm": 0.2252662032842636,
      "learning_rate": 0.0001,
      "loss": 1.2619,
      "step": 6938
    },
    {
      "epoch": 0.08956344524294754,
      "grad_norm": 0.20738264918327332,
      "learning_rate": 0.0001,
      "loss": 1.172,
      "step": 6939
    },
    {
      "epoch": 0.08967070685401694,
      "grad_norm": 0.20305366814136505,
      "learning_rate": 0.0001,
      "loss": 1.299,
      "step": 6940
    },
    {
      "epoch": 0.08977796846508634,
      "grad_norm": 0.2125425487756729,
      "learning_rate": 0.0001,
      "loss": 1.2722,
      "step": 6941
    },
    {
      "epoch": 0.08988523007615574,
      "grad_norm": 0.20351597666740417,
      "learning_rate": 0.0001,
      "loss": 1.3315,
      "step": 6942
    },
    {
      "epoch": 0.08999249168722515,
      "grad_norm": 0.22243650257587433,
      "learning_rate": 0.0001,
      "loss": 1.2899,
      "step": 6943
    },
    {
      "epoch": 0.09009975329829455,
      "grad_norm": 0.2086757868528366,
      "learning_rate": 0.0001,
      "loss": 1.2611,
      "step": 6944
    },
    {
      "epoch": 0.09020701490936393,
      "grad_norm": 0.22043496370315552,
      "learning_rate": 0.0001,
      "loss": 1.293,
      "step": 6945
    },
    {
      "epoch": 0.09031427652043333,
      "grad_norm": 0.22915753722190857,
      "learning_rate": 0.0001,
      "loss": 1.2614,
      "step": 6946
    },
    {
      "epoch": 0.09042153813150274,
      "grad_norm": 0.22593452036380768,
      "learning_rate": 0.0001,
      "loss": 1.2398,
      "step": 6947
    },
    {
      "epoch": 0.09052879974257214,
      "grad_norm": 0.21235892176628113,
      "learning_rate": 0.0001,
      "loss": 1.3659,
      "step": 6948
    },
    {
      "epoch": 0.09063606135364154,
      "grad_norm": 0.2069162130355835,
      "learning_rate": 0.0001,
      "loss": 1.2356,
      "step": 6949
    },
    {
      "epoch": 0.09074332296471092,
      "grad_norm": 0.20727179944515228,
      "learning_rate": 0.0001,
      "loss": 1.2491,
      "step": 6950
    },
    {
      "epoch": 0.09085058457578032,
      "grad_norm": 0.2098621279001236,
      "learning_rate": 0.0001,
      "loss": 1.211,
      "step": 6951
    },
    {
      "epoch": 0.09095784618684973,
      "grad_norm": 0.22319509088993073,
      "learning_rate": 0.0001,
      "loss": 1.2704,
      "step": 6952
    },
    {
      "epoch": 0.09106510779791913,
      "grad_norm": 0.20520031452178955,
      "learning_rate": 0.0001,
      "loss": 1.1981,
      "step": 6953
    },
    {
      "epoch": 0.09117236940898853,
      "grad_norm": 0.2299032360315323,
      "learning_rate": 0.0001,
      "loss": 1.2392,
      "step": 6954
    },
    {
      "epoch": 0.09127963102005793,
      "grad_norm": 0.22585853934288025,
      "learning_rate": 0.0001,
      "loss": 1.2385,
      "step": 6955
    },
    {
      "epoch": 0.09138689263112731,
      "grad_norm": 0.2098923623561859,
      "learning_rate": 0.0001,
      "loss": 1.3094,
      "step": 6956
    },
    {
      "epoch": 0.09149415424219672,
      "grad_norm": 0.2080068588256836,
      "learning_rate": 0.0001,
      "loss": 1.2262,
      "step": 6957
    },
    {
      "epoch": 0.09160141585326612,
      "grad_norm": 0.2158178985118866,
      "learning_rate": 0.0001,
      "loss": 1.1785,
      "step": 6958
    },
    {
      "epoch": 0.09170867746433552,
      "grad_norm": 0.21630311012268066,
      "learning_rate": 0.0001,
      "loss": 1.3038,
      "step": 6959
    },
    {
      "epoch": 0.09181593907540492,
      "grad_norm": 0.21798375248908997,
      "learning_rate": 0.0001,
      "loss": 1.2956,
      "step": 6960
    },
    {
      "epoch": 0.0919232006864743,
      "grad_norm": 0.21808694303035736,
      "learning_rate": 0.0001,
      "loss": 1.3433,
      "step": 6961
    },
    {
      "epoch": 0.0920304622975437,
      "grad_norm": 0.21572647988796234,
      "learning_rate": 0.0001,
      "loss": 1.3236,
      "step": 6962
    },
    {
      "epoch": 0.0921377239086131,
      "grad_norm": 0.20417632162570953,
      "learning_rate": 0.0001,
      "loss": 1.1987,
      "step": 6963
    },
    {
      "epoch": 0.0922449855196825,
      "grad_norm": 0.2108839750289917,
      "learning_rate": 0.0001,
      "loss": 1.2151,
      "step": 6964
    },
    {
      "epoch": 0.09235224713075191,
      "grad_norm": 0.20451100170612335,
      "learning_rate": 0.0001,
      "loss": 1.2566,
      "step": 6965
    },
    {
      "epoch": 0.09245950874182131,
      "grad_norm": 0.2052125632762909,
      "learning_rate": 0.0001,
      "loss": 1.2361,
      "step": 6966
    },
    {
      "epoch": 0.0925667703528907,
      "grad_norm": 0.21304607391357422,
      "learning_rate": 0.0001,
      "loss": 1.2622,
      "step": 6967
    },
    {
      "epoch": 0.0926740319639601,
      "grad_norm": 0.20991235971450806,
      "learning_rate": 0.0001,
      "loss": 1.2723,
      "step": 6968
    },
    {
      "epoch": 0.0927812935750295,
      "grad_norm": 0.22598204016685486,
      "learning_rate": 0.0001,
      "loss": 1.2631,
      "step": 6969
    },
    {
      "epoch": 0.0928885551860989,
      "grad_norm": 0.20627529919147491,
      "learning_rate": 0.0001,
      "loss": 1.1556,
      "step": 6970
    },
    {
      "epoch": 0.0929958167971683,
      "grad_norm": 0.21242168545722961,
      "learning_rate": 0.0001,
      "loss": 1.2677,
      "step": 6971
    },
    {
      "epoch": 0.09310307840823769,
      "grad_norm": 0.21997950971126556,
      "learning_rate": 0.0001,
      "loss": 1.1871,
      "step": 6972
    },
    {
      "epoch": 0.09321034001930709,
      "grad_norm": 0.2084026336669922,
      "learning_rate": 0.0001,
      "loss": 1.2768,
      "step": 6973
    },
    {
      "epoch": 0.09331760163037649,
      "grad_norm": 0.2115333527326584,
      "learning_rate": 0.0001,
      "loss": 1.3348,
      "step": 6974
    },
    {
      "epoch": 0.09342486324144589,
      "grad_norm": 0.21591533720493317,
      "learning_rate": 0.0001,
      "loss": 1.367,
      "step": 6975
    },
    {
      "epoch": 0.09353212485251529,
      "grad_norm": 0.21019461750984192,
      "learning_rate": 0.0001,
      "loss": 1.1952,
      "step": 6976
    },
    {
      "epoch": 0.09363938646358469,
      "grad_norm": 0.20765277743339539,
      "learning_rate": 0.0001,
      "loss": 1.2187,
      "step": 6977
    },
    {
      "epoch": 0.09374664807465408,
      "grad_norm": 0.20387408137321472,
      "learning_rate": 0.0001,
      "loss": 1.18,
      "step": 6978
    },
    {
      "epoch": 0.09385390968572348,
      "grad_norm": 0.21831758320331573,
      "learning_rate": 0.0001,
      "loss": 1.3092,
      "step": 6979
    },
    {
      "epoch": 0.09396117129679288,
      "grad_norm": 0.22250396013259888,
      "learning_rate": 0.0001,
      "loss": 1.1316,
      "step": 6980
    },
    {
      "epoch": 0.09406843290786228,
      "grad_norm": 0.2055353969335556,
      "learning_rate": 0.0001,
      "loss": 1.2735,
      "step": 6981
    },
    {
      "epoch": 0.09417569451893168,
      "grad_norm": 0.21207940578460693,
      "learning_rate": 0.0001,
      "loss": 1.2763,
      "step": 6982
    },
    {
      "epoch": 0.09428295613000107,
      "grad_norm": 0.20485833287239075,
      "learning_rate": 0.0001,
      "loss": 1.2817,
      "step": 6983
    },
    {
      "epoch": 0.09439021774107047,
      "grad_norm": 0.21069779992103577,
      "learning_rate": 0.0001,
      "loss": 1.188,
      "step": 6984
    },
    {
      "epoch": 0.09449747935213987,
      "grad_norm": 0.2056363821029663,
      "learning_rate": 0.0001,
      "loss": 1.2654,
      "step": 6985
    },
    {
      "epoch": 0.09460474096320927,
      "grad_norm": 0.20570194721221924,
      "learning_rate": 0.0001,
      "loss": 1.232,
      "step": 6986
    },
    {
      "epoch": 0.09471200257427867,
      "grad_norm": 0.21929268538951874,
      "learning_rate": 0.0001,
      "loss": 1.2925,
      "step": 6987
    },
    {
      "epoch": 0.09481926418534807,
      "grad_norm": 0.21579575538635254,
      "learning_rate": 0.0001,
      "loss": 1.2834,
      "step": 6988
    },
    {
      "epoch": 0.09492652579641746,
      "grad_norm": 0.22024521231651306,
      "learning_rate": 0.0001,
      "loss": 1.3003,
      "step": 6989
    },
    {
      "epoch": 0.09503378740748686,
      "grad_norm": 0.21266323328018188,
      "learning_rate": 0.0001,
      "loss": 1.3349,
      "step": 6990
    },
    {
      "epoch": 0.09514104901855626,
      "grad_norm": 0.21919822692871094,
      "learning_rate": 0.0001,
      "loss": 1.3518,
      "step": 6991
    },
    {
      "epoch": 0.09524831062962566,
      "grad_norm": 0.20695239305496216,
      "learning_rate": 0.0001,
      "loss": 1.2733,
      "step": 6992
    },
    {
      "epoch": 0.09535557224069506,
      "grad_norm": 0.2238692045211792,
      "learning_rate": 0.0001,
      "loss": 1.2453,
      "step": 6993
    },
    {
      "epoch": 0.09546283385176445,
      "grad_norm": 0.20849408209323883,
      "learning_rate": 0.0001,
      "loss": 1.2145,
      "step": 6994
    },
    {
      "epoch": 0.09557009546283385,
      "grad_norm": 0.2099524736404419,
      "learning_rate": 0.0001,
      "loss": 1.2112,
      "step": 6995
    },
    {
      "epoch": 0.09567735707390325,
      "grad_norm": 0.19930967688560486,
      "learning_rate": 0.0001,
      "loss": 1.2267,
      "step": 6996
    },
    {
      "epoch": 0.09578461868497265,
      "grad_norm": 0.20941299200057983,
      "learning_rate": 0.0001,
      "loss": 1.2434,
      "step": 6997
    },
    {
      "epoch": 0.09589188029604205,
      "grad_norm": 0.2044067531824112,
      "learning_rate": 0.0001,
      "loss": 1.2763,
      "step": 6998
    },
    {
      "epoch": 0.09599914190711145,
      "grad_norm": 0.21049202978610992,
      "learning_rate": 0.0001,
      "loss": 1.3581,
      "step": 6999
    },
    {
      "epoch": 0.09610640351818084,
      "grad_norm": 0.21614328026771545,
      "learning_rate": 0.0001,
      "loss": 1.2353,
      "step": 7000
    },
    {
      "epoch": 0.09621366512925024,
      "grad_norm": 0.22843138873577118,
      "learning_rate": 0.0001,
      "loss": 1.267,
      "step": 7001
    },
    {
      "epoch": 0.09632092674031964,
      "grad_norm": 0.21480214595794678,
      "learning_rate": 0.0001,
      "loss": 1.2715,
      "step": 7002
    },
    {
      "epoch": 0.09642818835138904,
      "grad_norm": 0.20692597329616547,
      "learning_rate": 0.0001,
      "loss": 1.3299,
      "step": 7003
    },
    {
      "epoch": 0.09653544996245844,
      "grad_norm": 0.2119862586259842,
      "learning_rate": 0.0001,
      "loss": 1.2442,
      "step": 7004
    },
    {
      "epoch": 0.09664271157352783,
      "grad_norm": 0.2307545244693756,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 7005
    },
    {
      "epoch": 0.09674997318459723,
      "grad_norm": 0.20788228511810303,
      "learning_rate": 0.0001,
      "loss": 1.2732,
      "step": 7006
    },
    {
      "epoch": 0.09685723479566663,
      "grad_norm": 0.20758484303951263,
      "learning_rate": 0.0001,
      "loss": 1.2116,
      "step": 7007
    },
    {
      "epoch": 0.09696449640673603,
      "grad_norm": 0.20724865794181824,
      "learning_rate": 0.0001,
      "loss": 1.2598,
      "step": 7008
    },
    {
      "epoch": 0.09707175801780543,
      "grad_norm": 0.21083128452301025,
      "learning_rate": 0.0001,
      "loss": 1.2515,
      "step": 7009
    },
    {
      "epoch": 0.09717901962887483,
      "grad_norm": 0.21372190117835999,
      "learning_rate": 0.0001,
      "loss": 1.2246,
      "step": 7010
    },
    {
      "epoch": 0.09728628123994422,
      "grad_norm": 0.2085382044315338,
      "learning_rate": 0.0001,
      "loss": 1.232,
      "step": 7011
    },
    {
      "epoch": 0.09739354285101362,
      "grad_norm": 0.21481595933437347,
      "learning_rate": 0.0001,
      "loss": 1.2515,
      "step": 7012
    },
    {
      "epoch": 0.09750080446208302,
      "grad_norm": 0.2113979458808899,
      "learning_rate": 0.0001,
      "loss": 1.28,
      "step": 7013
    },
    {
      "epoch": 0.09760806607315242,
      "grad_norm": 0.20700353384017944,
      "learning_rate": 0.0001,
      "loss": 1.2859,
      "step": 7014
    },
    {
      "epoch": 0.09771532768422182,
      "grad_norm": 0.2207186222076416,
      "learning_rate": 0.0001,
      "loss": 1.4057,
      "step": 7015
    },
    {
      "epoch": 0.09782258929529121,
      "grad_norm": 0.22207523882389069,
      "learning_rate": 0.0001,
      "loss": 1.2253,
      "step": 7016
    },
    {
      "epoch": 0.09792985090636061,
      "grad_norm": 0.20715570449829102,
      "learning_rate": 0.0001,
      "loss": 1.2053,
      "step": 7017
    },
    {
      "epoch": 0.09803711251743001,
      "grad_norm": 0.20161506533622742,
      "learning_rate": 0.0001,
      "loss": 1.2897,
      "step": 7018
    },
    {
      "epoch": 0.09814437412849941,
      "grad_norm": 0.2097478210926056,
      "learning_rate": 0.0001,
      "loss": 1.2346,
      "step": 7019
    },
    {
      "epoch": 0.09825163573956881,
      "grad_norm": 0.2279788851737976,
      "learning_rate": 0.0001,
      "loss": 1.2337,
      "step": 7020
    },
    {
      "epoch": 0.0983588973506382,
      "grad_norm": 0.2294238805770874,
      "learning_rate": 0.0001,
      "loss": 1.2654,
      "step": 7021
    },
    {
      "epoch": 0.0984661589617076,
      "grad_norm": 0.21162375807762146,
      "learning_rate": 0.0001,
      "loss": 1.2224,
      "step": 7022
    },
    {
      "epoch": 0.098573420572777,
      "grad_norm": 0.21770255267620087,
      "learning_rate": 0.0001,
      "loss": 1.26,
      "step": 7023
    },
    {
      "epoch": 0.0986806821838464,
      "grad_norm": 0.21938523650169373,
      "learning_rate": 0.0001,
      "loss": 1.3218,
      "step": 7024
    },
    {
      "epoch": 0.0987879437949158,
      "grad_norm": 0.21223947405815125,
      "learning_rate": 0.0001,
      "loss": 1.2414,
      "step": 7025
    },
    {
      "epoch": 0.0988952054059852,
      "grad_norm": 0.22039122879505157,
      "learning_rate": 0.0001,
      "loss": 1.189,
      "step": 7026
    },
    {
      "epoch": 0.09900246701705459,
      "grad_norm": 0.21175086498260498,
      "learning_rate": 0.0001,
      "loss": 1.2518,
      "step": 7027
    },
    {
      "epoch": 0.09910972862812399,
      "grad_norm": 0.22823889553546906,
      "learning_rate": 0.0001,
      "loss": 1.2716,
      "step": 7028
    },
    {
      "epoch": 0.09921699023919339,
      "grad_norm": 0.21126730740070343,
      "learning_rate": 0.0001,
      "loss": 1.2968,
      "step": 7029
    },
    {
      "epoch": 0.09932425185026279,
      "grad_norm": 0.21292753517627716,
      "learning_rate": 0.0001,
      "loss": 1.2564,
      "step": 7030
    },
    {
      "epoch": 0.0994315134613322,
      "grad_norm": 0.2175334393978119,
      "learning_rate": 0.0001,
      "loss": 1.238,
      "step": 7031
    },
    {
      "epoch": 0.09953877507240158,
      "grad_norm": 0.21743641793727875,
      "learning_rate": 0.0001,
      "loss": 1.3281,
      "step": 7032
    },
    {
      "epoch": 0.09964603668347098,
      "grad_norm": 0.22534632682800293,
      "learning_rate": 0.0001,
      "loss": 1.3065,
      "step": 7033
    },
    {
      "epoch": 0.09975329829454038,
      "grad_norm": 0.23282954096794128,
      "learning_rate": 0.0001,
      "loss": 1.3815,
      "step": 7034
    },
    {
      "epoch": 0.09986055990560978,
      "grad_norm": 0.2182275503873825,
      "learning_rate": 0.0001,
      "loss": 1.2939,
      "step": 7035
    },
    {
      "epoch": 0.09996782151667918,
      "grad_norm": 0.1924121230840683,
      "learning_rate": 0.0001,
      "loss": 1.1529,
      "step": 7036
    },
    {
      "epoch": 0.10007508312774858,
      "grad_norm": 0.22431135177612305,
      "learning_rate": 0.0001,
      "loss": 1.2414,
      "step": 7037
    },
    {
      "epoch": 0.10018234473881797,
      "grad_norm": 0.20779673755168915,
      "learning_rate": 0.0001,
      "loss": 1.2053,
      "step": 7038
    },
    {
      "epoch": 0.10028960634988737,
      "grad_norm": 0.19841653108596802,
      "learning_rate": 0.0001,
      "loss": 1.168,
      "step": 7039
    },
    {
      "epoch": 0.10039686796095677,
      "grad_norm": 0.21695095300674438,
      "learning_rate": 0.0001,
      "loss": 1.2326,
      "step": 7040
    },
    {
      "epoch": 0.10050412957202617,
      "grad_norm": 0.20461317896842957,
      "learning_rate": 0.0001,
      "loss": 1.224,
      "step": 7041
    },
    {
      "epoch": 0.10061139118309557,
      "grad_norm": 0.2132393717765808,
      "learning_rate": 0.0001,
      "loss": 1.3075,
      "step": 7042
    },
    {
      "epoch": 0.10071865279416496,
      "grad_norm": 0.21710306406021118,
      "learning_rate": 0.0001,
      "loss": 1.2416,
      "step": 7043
    },
    {
      "epoch": 0.10082591440523436,
      "grad_norm": 0.2126137763261795,
      "learning_rate": 0.0001,
      "loss": 1.2853,
      "step": 7044
    },
    {
      "epoch": 0.10093317601630376,
      "grad_norm": 0.22432582080364227,
      "learning_rate": 0.0001,
      "loss": 1.2896,
      "step": 7045
    },
    {
      "epoch": 0.10104043762737316,
      "grad_norm": 0.2125449925661087,
      "learning_rate": 0.0001,
      "loss": 1.2951,
      "step": 7046
    },
    {
      "epoch": 0.10114769923844256,
      "grad_norm": 0.2172774225473404,
      "learning_rate": 0.0001,
      "loss": 1.3683,
      "step": 7047
    },
    {
      "epoch": 0.10125496084951197,
      "grad_norm": 0.21216677129268646,
      "learning_rate": 0.0001,
      "loss": 1.2748,
      "step": 7048
    },
    {
      "epoch": 0.10136222246058135,
      "grad_norm": 0.22178226709365845,
      "learning_rate": 0.0001,
      "loss": 1.2502,
      "step": 7049
    },
    {
      "epoch": 0.10146948407165075,
      "grad_norm": 0.21759119629859924,
      "learning_rate": 0.0001,
      "loss": 1.2808,
      "step": 7050
    },
    {
      "epoch": 0.10157674568272015,
      "grad_norm": 0.2307625710964203,
      "learning_rate": 0.0001,
      "loss": 1.3241,
      "step": 7051
    },
    {
      "epoch": 0.10168400729378956,
      "grad_norm": 0.2164677232503891,
      "learning_rate": 0.0001,
      "loss": 1.2972,
      "step": 7052
    },
    {
      "epoch": 0.10179126890485896,
      "grad_norm": 0.21986261010169983,
      "learning_rate": 0.0001,
      "loss": 1.2097,
      "step": 7053
    },
    {
      "epoch": 0.10189853051592834,
      "grad_norm": 0.21567079424858093,
      "learning_rate": 0.0001,
      "loss": 1.2648,
      "step": 7054
    },
    {
      "epoch": 0.10200579212699774,
      "grad_norm": 0.20995615422725677,
      "learning_rate": 0.0001,
      "loss": 1.1887,
      "step": 7055
    },
    {
      "epoch": 0.10211305373806714,
      "grad_norm": 0.2046390175819397,
      "learning_rate": 0.0001,
      "loss": 1.1992,
      "step": 7056
    },
    {
      "epoch": 0.10222031534913655,
      "grad_norm": 0.2204498052597046,
      "learning_rate": 0.0001,
      "loss": 1.2642,
      "step": 7057
    },
    {
      "epoch": 0.10232757696020595,
      "grad_norm": 0.21640609204769135,
      "learning_rate": 0.0001,
      "loss": 1.305,
      "step": 7058
    },
    {
      "epoch": 0.10243483857127535,
      "grad_norm": 0.22031033039093018,
      "learning_rate": 0.0001,
      "loss": 1.3667,
      "step": 7059
    },
    {
      "epoch": 0.10254210018234473,
      "grad_norm": 0.2260427623987198,
      "learning_rate": 0.0001,
      "loss": 1.2578,
      "step": 7060
    },
    {
      "epoch": 0.10264936179341413,
      "grad_norm": 0.2118300497531891,
      "learning_rate": 0.0001,
      "loss": 1.35,
      "step": 7061
    },
    {
      "epoch": 0.10275662340448354,
      "grad_norm": 0.20326462388038635,
      "learning_rate": 0.0001,
      "loss": 1.1896,
      "step": 7062
    },
    {
      "epoch": 0.10286388501555294,
      "grad_norm": 0.2104141116142273,
      "learning_rate": 0.0001,
      "loss": 1.3489,
      "step": 7063
    },
    {
      "epoch": 0.10297114662662234,
      "grad_norm": 0.21123327314853668,
      "learning_rate": 0.0001,
      "loss": 1.2898,
      "step": 7064
    },
    {
      "epoch": 0.10307840823769172,
      "grad_norm": 0.21543483436107635,
      "learning_rate": 0.0001,
      "loss": 1.2498,
      "step": 7065
    },
    {
      "epoch": 0.10318566984876112,
      "grad_norm": 0.22240185737609863,
      "learning_rate": 0.0001,
      "loss": 1.2575,
      "step": 7066
    },
    {
      "epoch": 0.10329293145983053,
      "grad_norm": 0.21836207807064056,
      "learning_rate": 0.0001,
      "loss": 1.2965,
      "step": 7067
    },
    {
      "epoch": 0.10340019307089993,
      "grad_norm": 0.20783105492591858,
      "learning_rate": 0.0001,
      "loss": 1.1287,
      "step": 7068
    },
    {
      "epoch": 0.10350745468196933,
      "grad_norm": 0.2161645144224167,
      "learning_rate": 0.0001,
      "loss": 1.1293,
      "step": 7069
    },
    {
      "epoch": 0.10361471629303873,
      "grad_norm": 0.2122649848461151,
      "learning_rate": 0.0001,
      "loss": 1.2725,
      "step": 7070
    },
    {
      "epoch": 0.10372197790410811,
      "grad_norm": 0.2241399586200714,
      "learning_rate": 0.0001,
      "loss": 1.297,
      "step": 7071
    },
    {
      "epoch": 0.10382923951517752,
      "grad_norm": 0.21234259009361267,
      "learning_rate": 0.0001,
      "loss": 1.243,
      "step": 7072
    },
    {
      "epoch": 0.10393650112624692,
      "grad_norm": 0.207000270485878,
      "learning_rate": 0.0001,
      "loss": 1.2083,
      "step": 7073
    },
    {
      "epoch": 0.10404376273731632,
      "grad_norm": 0.1996849775314331,
      "learning_rate": 0.0001,
      "loss": 1.3261,
      "step": 7074
    },
    {
      "epoch": 0.10415102434838572,
      "grad_norm": 0.20029373466968536,
      "learning_rate": 0.0001,
      "loss": 1.2752,
      "step": 7075
    },
    {
      "epoch": 0.1042582859594551,
      "grad_norm": 0.20973782241344452,
      "learning_rate": 0.0001,
      "loss": 1.2563,
      "step": 7076
    },
    {
      "epoch": 0.1043655475705245,
      "grad_norm": 0.20718656480312347,
      "learning_rate": 0.0001,
      "loss": 1.2464,
      "step": 7077
    },
    {
      "epoch": 0.1044728091815939,
      "grad_norm": 0.2292637825012207,
      "learning_rate": 0.0001,
      "loss": 1.2903,
      "step": 7078
    },
    {
      "epoch": 0.10458007079266331,
      "grad_norm": 0.23252132534980774,
      "learning_rate": 0.0001,
      "loss": 1.3726,
      "step": 7079
    },
    {
      "epoch": 0.10468733240373271,
      "grad_norm": 0.21522647142410278,
      "learning_rate": 0.0001,
      "loss": 1.3,
      "step": 7080
    },
    {
      "epoch": 0.10479459401480211,
      "grad_norm": 0.21788857877254486,
      "learning_rate": 0.0001,
      "loss": 1.3225,
      "step": 7081
    },
    {
      "epoch": 0.1049018556258715,
      "grad_norm": 0.21027007699012756,
      "learning_rate": 0.0001,
      "loss": 1.2415,
      "step": 7082
    },
    {
      "epoch": 0.1050091172369409,
      "grad_norm": 0.20676718652248383,
      "learning_rate": 0.0001,
      "loss": 1.1947,
      "step": 7083
    },
    {
      "epoch": 0.1051163788480103,
      "grad_norm": 0.21479834616184235,
      "learning_rate": 0.0001,
      "loss": 1.2154,
      "step": 7084
    },
    {
      "epoch": 0.1052236404590797,
      "grad_norm": 0.20585700869560242,
      "learning_rate": 0.0001,
      "loss": 1.1448,
      "step": 7085
    },
    {
      "epoch": 0.1053309020701491,
      "grad_norm": 0.20947326719760895,
      "learning_rate": 0.0001,
      "loss": 1.2268,
      "step": 7086
    },
    {
      "epoch": 0.10543816368121849,
      "grad_norm": 0.21986708045005798,
      "learning_rate": 0.0001,
      "loss": 1.3234,
      "step": 7087
    },
    {
      "epoch": 0.10554542529228789,
      "grad_norm": 0.21436205506324768,
      "learning_rate": 0.0001,
      "loss": 1.2521,
      "step": 7088
    },
    {
      "epoch": 0.10565268690335729,
      "grad_norm": 0.2344309687614441,
      "learning_rate": 0.0001,
      "loss": 1.2695,
      "step": 7089
    },
    {
      "epoch": 0.10575994851442669,
      "grad_norm": 0.20944668352603912,
      "learning_rate": 0.0001,
      "loss": 1.204,
      "step": 7090
    },
    {
      "epoch": 0.10586721012549609,
      "grad_norm": 0.2015346735715866,
      "learning_rate": 0.0001,
      "loss": 1.2356,
      "step": 7091
    },
    {
      "epoch": 0.10597447173656549,
      "grad_norm": 0.2163650244474411,
      "learning_rate": 0.0001,
      "loss": 1.2317,
      "step": 7092
    },
    {
      "epoch": 0.10608173334763488,
      "grad_norm": 0.23733004927635193,
      "learning_rate": 0.0001,
      "loss": 1.2592,
      "step": 7093
    },
    {
      "epoch": 0.10618899495870428,
      "grad_norm": 0.21341095864772797,
      "learning_rate": 0.0001,
      "loss": 1.1756,
      "step": 7094
    },
    {
      "epoch": 0.10629625656977368,
      "grad_norm": 0.2084835320711136,
      "learning_rate": 0.0001,
      "loss": 1.2024,
      "step": 7095
    },
    {
      "epoch": 0.10640351818084308,
      "grad_norm": 0.21519392728805542,
      "learning_rate": 0.0001,
      "loss": 1.2725,
      "step": 7096
    },
    {
      "epoch": 0.10651077979191248,
      "grad_norm": 0.22414667904376984,
      "learning_rate": 0.0001,
      "loss": 1.3203,
      "step": 7097
    },
    {
      "epoch": 0.10661804140298187,
      "grad_norm": 0.21686074137687683,
      "learning_rate": 0.0001,
      "loss": 1.2813,
      "step": 7098
    },
    {
      "epoch": 0.10672530301405127,
      "grad_norm": 0.22808215022087097,
      "learning_rate": 0.0001,
      "loss": 1.2496,
      "step": 7099
    },
    {
      "epoch": 0.10683256462512067,
      "grad_norm": 0.23279818892478943,
      "learning_rate": 0.0001,
      "loss": 1.342,
      "step": 7100
    },
    {
      "epoch": 0.10693982623619007,
      "grad_norm": 0.2347738891839981,
      "learning_rate": 0.0001,
      "loss": 1.2255,
      "step": 7101
    },
    {
      "epoch": 0.10704708784725947,
      "grad_norm": 0.2102946937084198,
      "learning_rate": 0.0001,
      "loss": 1.3199,
      "step": 7102
    },
    {
      "epoch": 0.10715434945832887,
      "grad_norm": 0.22693125903606415,
      "learning_rate": 0.0001,
      "loss": 1.2527,
      "step": 7103
    },
    {
      "epoch": 0.10726161106939826,
      "grad_norm": 0.21608470380306244,
      "learning_rate": 0.0001,
      "loss": 1.252,
      "step": 7104
    },
    {
      "epoch": 0.10736887268046766,
      "grad_norm": 0.2293478399515152,
      "learning_rate": 0.0001,
      "loss": 1.2147,
      "step": 7105
    },
    {
      "epoch": 0.10747613429153706,
      "grad_norm": 0.22349244356155396,
      "learning_rate": 0.0001,
      "loss": 1.2663,
      "step": 7106
    },
    {
      "epoch": 0.10758339590260646,
      "grad_norm": 0.22750312089920044,
      "learning_rate": 0.0001,
      "loss": 1.2095,
      "step": 7107
    },
    {
      "epoch": 0.10769065751367586,
      "grad_norm": 0.24642394483089447,
      "learning_rate": 0.0001,
      "loss": 1.4152,
      "step": 7108
    },
    {
      "epoch": 0.10779791912474525,
      "grad_norm": 0.2170550376176834,
      "learning_rate": 0.0001,
      "loss": 1.2841,
      "step": 7109
    },
    {
      "epoch": 0.10790518073581465,
      "grad_norm": 0.23386308550834656,
      "learning_rate": 0.0001,
      "loss": 1.1896,
      "step": 7110
    },
    {
      "epoch": 0.10801244234688405,
      "grad_norm": 0.22292710840702057,
      "learning_rate": 0.0001,
      "loss": 1.2326,
      "step": 7111
    },
    {
      "epoch": 0.10811970395795345,
      "grad_norm": 0.21935787796974182,
      "learning_rate": 0.0001,
      "loss": 1.2853,
      "step": 7112
    },
    {
      "epoch": 0.10822696556902285,
      "grad_norm": 0.22762994468212128,
      "learning_rate": 0.0001,
      "loss": 1.3066,
      "step": 7113
    },
    {
      "epoch": 0.10833422718009225,
      "grad_norm": 0.21435119211673737,
      "learning_rate": 0.0001,
      "loss": 1.1279,
      "step": 7114
    },
    {
      "epoch": 0.10844148879116164,
      "grad_norm": 0.215597003698349,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 7115
    },
    {
      "epoch": 0.10854875040223104,
      "grad_norm": 0.2031688690185547,
      "learning_rate": 0.0001,
      "loss": 1.2796,
      "step": 7116
    },
    {
      "epoch": 0.10865601201330044,
      "grad_norm": 0.22235207259655,
      "learning_rate": 0.0001,
      "loss": 1.1971,
      "step": 7117
    },
    {
      "epoch": 0.10876327362436984,
      "grad_norm": 0.2211044728755951,
      "learning_rate": 0.0001,
      "loss": 1.239,
      "step": 7118
    },
    {
      "epoch": 0.10887053523543924,
      "grad_norm": 0.21769189834594727,
      "learning_rate": 0.0001,
      "loss": 1.2703,
      "step": 7119
    },
    {
      "epoch": 0.10897779684650863,
      "grad_norm": 0.22430458664894104,
      "learning_rate": 0.0001,
      "loss": 1.2615,
      "step": 7120
    },
    {
      "epoch": 0.10908505845757803,
      "grad_norm": 0.22895950078964233,
      "learning_rate": 0.0001,
      "loss": 1.3619,
      "step": 7121
    },
    {
      "epoch": 0.10919232006864743,
      "grad_norm": 0.21303527057170868,
      "learning_rate": 0.0001,
      "loss": 1.2922,
      "step": 7122
    },
    {
      "epoch": 0.10929958167971683,
      "grad_norm": 0.2138407826423645,
      "learning_rate": 0.0001,
      "loss": 1.3341,
      "step": 7123
    },
    {
      "epoch": 0.10940684329078623,
      "grad_norm": 0.21904058754444122,
      "learning_rate": 0.0001,
      "loss": 1.2607,
      "step": 7124
    },
    {
      "epoch": 0.10951410490185562,
      "grad_norm": 0.2224254608154297,
      "learning_rate": 0.0001,
      "loss": 1.154,
      "step": 7125
    },
    {
      "epoch": 0.10962136651292502,
      "grad_norm": 0.2259313017129898,
      "learning_rate": 0.0001,
      "loss": 1.3055,
      "step": 7126
    },
    {
      "epoch": 0.10972862812399442,
      "grad_norm": 0.22481203079223633,
      "learning_rate": 0.0001,
      "loss": 1.1201,
      "step": 7127
    },
    {
      "epoch": 0.10983588973506382,
      "grad_norm": 0.21477752923965454,
      "learning_rate": 0.0001,
      "loss": 1.2259,
      "step": 7128
    },
    {
      "epoch": 0.10994315134613322,
      "grad_norm": 0.2083967924118042,
      "learning_rate": 0.0001,
      "loss": 1.2445,
      "step": 7129
    },
    {
      "epoch": 0.11005041295720262,
      "grad_norm": 0.2102813422679901,
      "learning_rate": 0.0001,
      "loss": 1.2882,
      "step": 7130
    },
    {
      "epoch": 0.11015767456827201,
      "grad_norm": 0.22260163724422455,
      "learning_rate": 0.0001,
      "loss": 1.329,
      "step": 7131
    },
    {
      "epoch": 0.11026493617934141,
      "grad_norm": 0.2037307471036911,
      "learning_rate": 0.0001,
      "loss": 1.1754,
      "step": 7132
    },
    {
      "epoch": 0.11037219779041081,
      "grad_norm": 0.20141363143920898,
      "learning_rate": 0.0001,
      "loss": 1.2629,
      "step": 7133
    },
    {
      "epoch": 0.11047945940148021,
      "grad_norm": 0.2182672619819641,
      "learning_rate": 0.0001,
      "loss": 1.242,
      "step": 7134
    },
    {
      "epoch": 0.11058672101254961,
      "grad_norm": 0.2105564922094345,
      "learning_rate": 0.0001,
      "loss": 1.2541,
      "step": 7135
    },
    {
      "epoch": 0.110693982623619,
      "grad_norm": 0.21500222384929657,
      "learning_rate": 0.0001,
      "loss": 1.2779,
      "step": 7136
    },
    {
      "epoch": 0.1108012442346884,
      "grad_norm": 0.2099532186985016,
      "learning_rate": 0.0001,
      "loss": 1.2407,
      "step": 7137
    },
    {
      "epoch": 0.1109085058457578,
      "grad_norm": 0.22054177522659302,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 7138
    },
    {
      "epoch": 0.1110157674568272,
      "grad_norm": 0.21977338194847107,
      "learning_rate": 0.0001,
      "loss": 1.2864,
      "step": 7139
    },
    {
      "epoch": 0.1111230290678966,
      "grad_norm": 0.2073196917772293,
      "learning_rate": 0.0001,
      "loss": 1.2449,
      "step": 7140
    },
    {
      "epoch": 0.111230290678966,
      "grad_norm": 0.2175893783569336,
      "learning_rate": 0.0001,
      "loss": 1.2494,
      "step": 7141
    },
    {
      "epoch": 0.11133755229003539,
      "grad_norm": 0.2133924663066864,
      "learning_rate": 0.0001,
      "loss": 1.2869,
      "step": 7142
    },
    {
      "epoch": 0.11144481390110479,
      "grad_norm": 0.2022683024406433,
      "learning_rate": 0.0001,
      "loss": 1.2551,
      "step": 7143
    },
    {
      "epoch": 0.11155207551217419,
      "grad_norm": 0.2292499840259552,
      "learning_rate": 0.0001,
      "loss": 1.3116,
      "step": 7144
    },
    {
      "epoch": 0.1116593371232436,
      "grad_norm": 0.20208698511123657,
      "learning_rate": 0.0001,
      "loss": 1.2121,
      "step": 7145
    },
    {
      "epoch": 0.111766598734313,
      "grad_norm": 0.22975072264671326,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 7146
    },
    {
      "epoch": 0.11187386034538238,
      "grad_norm": 0.2110445201396942,
      "learning_rate": 0.0001,
      "loss": 1.2721,
      "step": 7147
    },
    {
      "epoch": 0.11198112195645178,
      "grad_norm": 0.21215656399726868,
      "learning_rate": 0.0001,
      "loss": 1.2884,
      "step": 7148
    },
    {
      "epoch": 0.11208838356752118,
      "grad_norm": 0.20980313420295715,
      "learning_rate": 0.0001,
      "loss": 1.17,
      "step": 7149
    },
    {
      "epoch": 0.11219564517859058,
      "grad_norm": 0.2163141369819641,
      "learning_rate": 0.0001,
      "loss": 1.3339,
      "step": 7150
    },
    {
      "epoch": 0.11230290678965998,
      "grad_norm": 0.21023766696453094,
      "learning_rate": 0.0001,
      "loss": 1.3132,
      "step": 7151
    },
    {
      "epoch": 0.11241016840072938,
      "grad_norm": 0.22798435389995575,
      "learning_rate": 0.0001,
      "loss": 1.2691,
      "step": 7152
    },
    {
      "epoch": 0.11251743001179877,
      "grad_norm": 0.2140314280986786,
      "learning_rate": 0.0001,
      "loss": 1.3694,
      "step": 7153
    },
    {
      "epoch": 0.11262469162286817,
      "grad_norm": 0.20435287058353424,
      "learning_rate": 0.0001,
      "loss": 1.3307,
      "step": 7154
    },
    {
      "epoch": 0.11273195323393757,
      "grad_norm": 0.20238034427165985,
      "learning_rate": 0.0001,
      "loss": 1.2526,
      "step": 7155
    },
    {
      "epoch": 0.11283921484500697,
      "grad_norm": 0.20826512575149536,
      "learning_rate": 0.0001,
      "loss": 1.2365,
      "step": 7156
    },
    {
      "epoch": 0.11294647645607638,
      "grad_norm": 0.20192936062812805,
      "learning_rate": 0.0001,
      "loss": 1.2432,
      "step": 7157
    },
    {
      "epoch": 0.11305373806714576,
      "grad_norm": 0.21550212800502777,
      "learning_rate": 0.0001,
      "loss": 1.2639,
      "step": 7158
    },
    {
      "epoch": 0.11316099967821516,
      "grad_norm": 0.22073625028133392,
      "learning_rate": 0.0001,
      "loss": 1.1003,
      "step": 7159
    },
    {
      "epoch": 0.11326826128928456,
      "grad_norm": 0.2261020690202713,
      "learning_rate": 0.0001,
      "loss": 1.3389,
      "step": 7160
    },
    {
      "epoch": 0.11337552290035396,
      "grad_norm": 0.20779472589492798,
      "learning_rate": 0.0001,
      "loss": 1.2378,
      "step": 7161
    },
    {
      "epoch": 0.11348278451142337,
      "grad_norm": 0.21365298330783844,
      "learning_rate": 0.0001,
      "loss": 1.296,
      "step": 7162
    },
    {
      "epoch": 0.11359004612249277,
      "grad_norm": 0.20171774923801422,
      "learning_rate": 0.0001,
      "loss": 1.2655,
      "step": 7163
    },
    {
      "epoch": 0.11369730773356215,
      "grad_norm": 0.22420760989189148,
      "learning_rate": 0.0001,
      "loss": 1.2859,
      "step": 7164
    },
    {
      "epoch": 0.11380456934463155,
      "grad_norm": 0.21239472925662994,
      "learning_rate": 0.0001,
      "loss": 1.2508,
      "step": 7165
    },
    {
      "epoch": 0.11391183095570095,
      "grad_norm": 0.22570884227752686,
      "learning_rate": 0.0001,
      "loss": 1.2483,
      "step": 7166
    },
    {
      "epoch": 0.11401909256677036,
      "grad_norm": 0.21654345095157623,
      "learning_rate": 0.0001,
      "loss": 1.256,
      "step": 7167
    },
    {
      "epoch": 0.11412635417783976,
      "grad_norm": 0.22484280169010162,
      "learning_rate": 0.0001,
      "loss": 1.2875,
      "step": 7168
    },
    {
      "epoch": 0.11423361578890914,
      "grad_norm": 0.23424921929836273,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 7169
    },
    {
      "epoch": 0.11434087739997854,
      "grad_norm": 0.2215433567762375,
      "learning_rate": 0.0001,
      "loss": 1.2676,
      "step": 7170
    },
    {
      "epoch": 0.11444813901104794,
      "grad_norm": 0.2057960331439972,
      "learning_rate": 0.0001,
      "loss": 1.2382,
      "step": 7171
    },
    {
      "epoch": 0.11455540062211735,
      "grad_norm": 0.21322429180145264,
      "learning_rate": 0.0001,
      "loss": 1.3112,
      "step": 7172
    },
    {
      "epoch": 0.11466266223318675,
      "grad_norm": 0.20186923444271088,
      "learning_rate": 0.0001,
      "loss": 1.2593,
      "step": 7173
    },
    {
      "epoch": 0.11476992384425615,
      "grad_norm": 0.21352845430374146,
      "learning_rate": 0.0001,
      "loss": 1.2285,
      "step": 7174
    },
    {
      "epoch": 0.11487718545532553,
      "grad_norm": 0.22605770826339722,
      "learning_rate": 0.0001,
      "loss": 1.1733,
      "step": 7175
    },
    {
      "epoch": 0.11498444706639493,
      "grad_norm": 0.21000690758228302,
      "learning_rate": 0.0001,
      "loss": 1.189,
      "step": 7176
    },
    {
      "epoch": 0.11509170867746434,
      "grad_norm": 0.21398110687732697,
      "learning_rate": 0.0001,
      "loss": 1.3159,
      "step": 7177
    },
    {
      "epoch": 0.11519897028853374,
      "grad_norm": 0.21193927526474,
      "learning_rate": 0.0001,
      "loss": 1.2408,
      "step": 7178
    },
    {
      "epoch": 0.11530623189960314,
      "grad_norm": 0.20161622762680054,
      "learning_rate": 0.0001,
      "loss": 1.2457,
      "step": 7179
    },
    {
      "epoch": 0.11541349351067252,
      "grad_norm": 0.1998174488544464,
      "learning_rate": 0.0001,
      "loss": 1.1218,
      "step": 7180
    },
    {
      "epoch": 0.11552075512174192,
      "grad_norm": 0.21930988132953644,
      "learning_rate": 0.0001,
      "loss": 1.2294,
      "step": 7181
    },
    {
      "epoch": 0.11562801673281133,
      "grad_norm": 0.21981222927570343,
      "learning_rate": 0.0001,
      "loss": 1.2674,
      "step": 7182
    },
    {
      "epoch": 0.11573527834388073,
      "grad_norm": 0.21731676161289215,
      "learning_rate": 0.0001,
      "loss": 1.2506,
      "step": 7183
    },
    {
      "epoch": 0.11584253995495013,
      "grad_norm": 0.21094343066215515,
      "learning_rate": 0.0001,
      "loss": 1.1763,
      "step": 7184
    },
    {
      "epoch": 0.11594980156601953,
      "grad_norm": 0.22937896847724915,
      "learning_rate": 0.0001,
      "loss": 1.2566,
      "step": 7185
    },
    {
      "epoch": 0.11605706317708891,
      "grad_norm": 0.2237701267004013,
      "learning_rate": 0.0001,
      "loss": 1.2291,
      "step": 7186
    },
    {
      "epoch": 0.11616432478815832,
      "grad_norm": 0.20040398836135864,
      "learning_rate": 0.0001,
      "loss": 1.2341,
      "step": 7187
    },
    {
      "epoch": 0.11627158639922772,
      "grad_norm": 0.2242138534784317,
      "learning_rate": 0.0001,
      "loss": 1.2301,
      "step": 7188
    },
    {
      "epoch": 0.11637884801029712,
      "grad_norm": 0.2242637723684311,
      "learning_rate": 0.0001,
      "loss": 1.2852,
      "step": 7189
    },
    {
      "epoch": 0.11648610962136652,
      "grad_norm": 0.22075611352920532,
      "learning_rate": 0.0001,
      "loss": 1.2647,
      "step": 7190
    },
    {
      "epoch": 0.1165933712324359,
      "grad_norm": 0.21295329928398132,
      "learning_rate": 0.0001,
      "loss": 1.1932,
      "step": 7191
    },
    {
      "epoch": 0.1167006328435053,
      "grad_norm": 0.21563847362995148,
      "learning_rate": 0.0001,
      "loss": 1.2833,
      "step": 7192
    },
    {
      "epoch": 0.1168078944545747,
      "grad_norm": 0.22024795413017273,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 7193
    },
    {
      "epoch": 0.11691515606564411,
      "grad_norm": 0.2142612785100937,
      "learning_rate": 0.0001,
      "loss": 1.2556,
      "step": 7194
    },
    {
      "epoch": 0.11702241767671351,
      "grad_norm": 0.20670832693576813,
      "learning_rate": 0.0001,
      "loss": 1.267,
      "step": 7195
    },
    {
      "epoch": 0.11712967928778291,
      "grad_norm": 0.2134236842393875,
      "learning_rate": 0.0001,
      "loss": 1.2742,
      "step": 7196
    },
    {
      "epoch": 0.1172369408988523,
      "grad_norm": 0.2094675451517105,
      "learning_rate": 0.0001,
      "loss": 1.2574,
      "step": 7197
    },
    {
      "epoch": 0.1173442025099217,
      "grad_norm": 0.2077130675315857,
      "learning_rate": 0.0001,
      "loss": 1.3173,
      "step": 7198
    },
    {
      "epoch": 0.1174514641209911,
      "grad_norm": 0.2153225988149643,
      "learning_rate": 0.0001,
      "loss": 1.1605,
      "step": 7199
    },
    {
      "epoch": 0.1175587257320605,
      "grad_norm": 0.22941166162490845,
      "learning_rate": 0.0001,
      "loss": 1.1464,
      "step": 7200
    },
    {
      "epoch": 0.1176659873431299,
      "grad_norm": 0.22736673057079315,
      "learning_rate": 0.0001,
      "loss": 1.2937,
      "step": 7201
    },
    {
      "epoch": 0.11777324895419929,
      "grad_norm": 0.2224375307559967,
      "learning_rate": 0.0001,
      "loss": 1.1627,
      "step": 7202
    },
    {
      "epoch": 0.11788051056526869,
      "grad_norm": 0.21009278297424316,
      "learning_rate": 0.0001,
      "loss": 1.312,
      "step": 7203
    },
    {
      "epoch": 0.11798777217633809,
      "grad_norm": 0.2150605022907257,
      "learning_rate": 0.0001,
      "loss": 1.2941,
      "step": 7204
    },
    {
      "epoch": 0.11809503378740749,
      "grad_norm": 0.20987826585769653,
      "learning_rate": 0.0001,
      "loss": 1.2506,
      "step": 7205
    },
    {
      "epoch": 0.11820229539847689,
      "grad_norm": 0.21911390125751495,
      "learning_rate": 0.0001,
      "loss": 1.2932,
      "step": 7206
    },
    {
      "epoch": 0.11830955700954629,
      "grad_norm": 0.21584749221801758,
      "learning_rate": 0.0001,
      "loss": 1.262,
      "step": 7207
    },
    {
      "epoch": 0.11841681862061568,
      "grad_norm": 0.20588496327400208,
      "learning_rate": 0.0001,
      "loss": 1.2223,
      "step": 7208
    },
    {
      "epoch": 0.11852408023168508,
      "grad_norm": 0.21777977049350739,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 7209
    },
    {
      "epoch": 0.11863134184275448,
      "grad_norm": 0.22093889117240906,
      "learning_rate": 0.0001,
      "loss": 1.1787,
      "step": 7210
    },
    {
      "epoch": 0.11873860345382388,
      "grad_norm": 0.2183019071817398,
      "learning_rate": 0.0001,
      "loss": 1.2324,
      "step": 7211
    },
    {
      "epoch": 0.11884586506489328,
      "grad_norm": 0.2147025167942047,
      "learning_rate": 0.0001,
      "loss": 1.1726,
      "step": 7212
    },
    {
      "epoch": 0.11895312667596267,
      "grad_norm": 0.21692675352096558,
      "learning_rate": 0.0001,
      "loss": 1.2415,
      "step": 7213
    },
    {
      "epoch": 0.11906038828703207,
      "grad_norm": 0.2163931131362915,
      "learning_rate": 0.0001,
      "loss": 1.2536,
      "step": 7214
    },
    {
      "epoch": 0.11916764989810147,
      "grad_norm": 0.22632332146167755,
      "learning_rate": 0.0001,
      "loss": 1.2904,
      "step": 7215
    },
    {
      "epoch": 0.11927491150917087,
      "grad_norm": 0.21333351731300354,
      "learning_rate": 0.0001,
      "loss": 1.2708,
      "step": 7216
    },
    {
      "epoch": 0.11938217312024027,
      "grad_norm": 0.20705516636371613,
      "learning_rate": 0.0001,
      "loss": 1.1783,
      "step": 7217
    },
    {
      "epoch": 0.11948943473130967,
      "grad_norm": 0.2179739773273468,
      "learning_rate": 0.0001,
      "loss": 1.2049,
      "step": 7218
    },
    {
      "epoch": 0.11959669634237906,
      "grad_norm": 0.2132548689842224,
      "learning_rate": 0.0001,
      "loss": 1.2396,
      "step": 7219
    },
    {
      "epoch": 0.11970395795344846,
      "grad_norm": 0.2132216989994049,
      "learning_rate": 0.0001,
      "loss": 1.2509,
      "step": 7220
    },
    {
      "epoch": 0.11981121956451786,
      "grad_norm": 0.2239689975976944,
      "learning_rate": 0.0001,
      "loss": 1.3071,
      "step": 7221
    },
    {
      "epoch": 0.11991848117558726,
      "grad_norm": 0.22168156504631042,
      "learning_rate": 0.0001,
      "loss": 1.2799,
      "step": 7222
    },
    {
      "epoch": 0.12002574278665666,
      "grad_norm": 0.22431494295597076,
      "learning_rate": 0.0001,
      "loss": 1.2659,
      "step": 7223
    },
    {
      "epoch": 0.12013300439772605,
      "grad_norm": 0.2228117138147354,
      "learning_rate": 0.0001,
      "loss": 1.2172,
      "step": 7224
    },
    {
      "epoch": 0.12024026600879545,
      "grad_norm": 0.223526269197464,
      "learning_rate": 0.0001,
      "loss": 1.1863,
      "step": 7225
    },
    {
      "epoch": 0.12034752761986485,
      "grad_norm": 0.21520599722862244,
      "learning_rate": 0.0001,
      "loss": 1.2639,
      "step": 7226
    },
    {
      "epoch": 0.12045478923093425,
      "grad_norm": 0.20318931341171265,
      "learning_rate": 0.0001,
      "loss": 1.2268,
      "step": 7227
    },
    {
      "epoch": 0.12056205084200365,
      "grad_norm": 0.21449239552021027,
      "learning_rate": 0.0001,
      "loss": 1.2553,
      "step": 7228
    },
    {
      "epoch": 0.12066931245307305,
      "grad_norm": 0.21324463188648224,
      "learning_rate": 0.0001,
      "loss": 1.2978,
      "step": 7229
    },
    {
      "epoch": 0.12077657406414244,
      "grad_norm": 0.22229818999767303,
      "learning_rate": 0.0001,
      "loss": 1.3445,
      "step": 7230
    },
    {
      "epoch": 0.12088383567521184,
      "grad_norm": 0.215840682387352,
      "learning_rate": 0.0001,
      "loss": 1.3279,
      "step": 7231
    },
    {
      "epoch": 0.12099109728628124,
      "grad_norm": 0.20373325049877167,
      "learning_rate": 0.0001,
      "loss": 1.2235,
      "step": 7232
    },
    {
      "epoch": 0.12109835889735064,
      "grad_norm": 0.22447803616523743,
      "learning_rate": 0.0001,
      "loss": 1.2287,
      "step": 7233
    },
    {
      "epoch": 0.12120562050842004,
      "grad_norm": 0.214033842086792,
      "learning_rate": 0.0001,
      "loss": 1.2312,
      "step": 7234
    },
    {
      "epoch": 0.12131288211948943,
      "grad_norm": 0.2110273241996765,
      "learning_rate": 0.0001,
      "loss": 1.2929,
      "step": 7235
    },
    {
      "epoch": 0.12142014373055883,
      "grad_norm": 0.2236384153366089,
      "learning_rate": 0.0001,
      "loss": 1.2945,
      "step": 7236
    },
    {
      "epoch": 0.12152740534162823,
      "grad_norm": 0.2077169418334961,
      "learning_rate": 0.0001,
      "loss": 1.1789,
      "step": 7237
    },
    {
      "epoch": 0.12163466695269763,
      "grad_norm": 0.22286462783813477,
      "learning_rate": 0.0001,
      "loss": 1.2962,
      "step": 7238
    },
    {
      "epoch": 0.12174192856376703,
      "grad_norm": 0.2191968858242035,
      "learning_rate": 0.0001,
      "loss": 1.2785,
      "step": 7239
    },
    {
      "epoch": 0.12184919017483642,
      "grad_norm": 0.217106431722641,
      "learning_rate": 0.0001,
      "loss": 1.34,
      "step": 7240
    },
    {
      "epoch": 0.12195645178590582,
      "grad_norm": 0.22531697154045105,
      "learning_rate": 0.0001,
      "loss": 1.3385,
      "step": 7241
    },
    {
      "epoch": 0.12206371339697522,
      "grad_norm": 0.23907579481601715,
      "learning_rate": 0.0001,
      "loss": 1.298,
      "step": 7242
    },
    {
      "epoch": 0.12217097500804462,
      "grad_norm": 0.21290971338748932,
      "learning_rate": 0.0001,
      "loss": 1.3367,
      "step": 7243
    },
    {
      "epoch": 0.12227823661911402,
      "grad_norm": 0.20894838869571686,
      "learning_rate": 0.0001,
      "loss": 1.189,
      "step": 7244
    },
    {
      "epoch": 0.12238549823018342,
      "grad_norm": 0.2085871696472168,
      "learning_rate": 0.0001,
      "loss": 1.223,
      "step": 7245
    },
    {
      "epoch": 0.12249275984125281,
      "grad_norm": 0.21855273842811584,
      "learning_rate": 0.0001,
      "loss": 1.2452,
      "step": 7246
    },
    {
      "epoch": 0.12260002145232221,
      "grad_norm": 0.2078171968460083,
      "learning_rate": 0.0001,
      "loss": 1.2483,
      "step": 7247
    },
    {
      "epoch": 0.12270728306339161,
      "grad_norm": 0.2184855043888092,
      "learning_rate": 0.0001,
      "loss": 1.3146,
      "step": 7248
    },
    {
      "epoch": 0.12281454467446101,
      "grad_norm": 0.20876085758209229,
      "learning_rate": 0.0001,
      "loss": 1.2613,
      "step": 7249
    },
    {
      "epoch": 0.12292180628553041,
      "grad_norm": 0.22276054322719574,
      "learning_rate": 0.0001,
      "loss": 1.222,
      "step": 7250
    },
    {
      "epoch": 0.1230290678965998,
      "grad_norm": 0.22102509438991547,
      "learning_rate": 0.0001,
      "loss": 1.2802,
      "step": 7251
    },
    {
      "epoch": 0.1231363295076692,
      "grad_norm": 0.2186214178800583,
      "learning_rate": 0.0001,
      "loss": 1.2811,
      "step": 7252
    },
    {
      "epoch": 0.1232435911187386,
      "grad_norm": 0.23004202544689178,
      "learning_rate": 0.0001,
      "loss": 1.1902,
      "step": 7253
    },
    {
      "epoch": 0.123350852729808,
      "grad_norm": 0.22349365055561066,
      "learning_rate": 0.0001,
      "loss": 1.2816,
      "step": 7254
    },
    {
      "epoch": 0.1234581143408774,
      "grad_norm": 0.22234463691711426,
      "learning_rate": 0.0001,
      "loss": 1.3227,
      "step": 7255
    },
    {
      "epoch": 0.1235653759519468,
      "grad_norm": 0.21427021920681,
      "learning_rate": 0.0001,
      "loss": 1.222,
      "step": 7256
    },
    {
      "epoch": 0.12367263756301619,
      "grad_norm": 0.2389194816350937,
      "learning_rate": 0.0001,
      "loss": 1.2101,
      "step": 7257
    },
    {
      "epoch": 0.12377989917408559,
      "grad_norm": 0.21765246987342834,
      "learning_rate": 0.0001,
      "loss": 1.2293,
      "step": 7258
    },
    {
      "epoch": 0.12388716078515499,
      "grad_norm": 0.21371832489967346,
      "learning_rate": 0.0001,
      "loss": 1.2808,
      "step": 7259
    },
    {
      "epoch": 0.1239944223962244,
      "grad_norm": 0.21493513882160187,
      "learning_rate": 0.0001,
      "loss": 1.2859,
      "step": 7260
    },
    {
      "epoch": 0.1241016840072938,
      "grad_norm": 0.20932403206825256,
      "learning_rate": 0.0001,
      "loss": 1.2421,
      "step": 7261
    },
    {
      "epoch": 0.12420894561836318,
      "grad_norm": 0.23662546277046204,
      "learning_rate": 0.0001,
      "loss": 1.2124,
      "step": 7262
    },
    {
      "epoch": 0.12431620722943258,
      "grad_norm": 0.21165381371974945,
      "learning_rate": 0.0001,
      "loss": 1.2624,
      "step": 7263
    },
    {
      "epoch": 0.12442346884050198,
      "grad_norm": 0.20406946539878845,
      "learning_rate": 0.0001,
      "loss": 1.2702,
      "step": 7264
    },
    {
      "epoch": 0.12453073045157138,
      "grad_norm": 0.22528354823589325,
      "learning_rate": 0.0001,
      "loss": 1.1848,
      "step": 7265
    },
    {
      "epoch": 0.12463799206264078,
      "grad_norm": 0.21440185606479645,
      "learning_rate": 0.0001,
      "loss": 1.2804,
      "step": 7266
    },
    {
      "epoch": 0.12474525367371019,
      "grad_norm": 0.2349371463060379,
      "learning_rate": 0.0001,
      "loss": 1.2468,
      "step": 7267
    },
    {
      "epoch": 0.12485251528477957,
      "grad_norm": 0.20809872448444366,
      "learning_rate": 0.0001,
      "loss": 1.2218,
      "step": 7268
    },
    {
      "epoch": 0.12495977689584897,
      "grad_norm": 0.2255915254354477,
      "learning_rate": 0.0001,
      "loss": 1.235,
      "step": 7269
    },
    {
      "epoch": 0.1250670385069184,
      "grad_norm": 0.2175072282552719,
      "learning_rate": 0.0001,
      "loss": 1.257,
      "step": 7270
    },
    {
      "epoch": 0.12517430011798777,
      "grad_norm": 0.21009953320026398,
      "learning_rate": 0.0001,
      "loss": 1.2393,
      "step": 7271
    },
    {
      "epoch": 0.12528156172905716,
      "grad_norm": 0.22381006181240082,
      "learning_rate": 0.0001,
      "loss": 1.2488,
      "step": 7272
    },
    {
      "epoch": 0.12538882334012658,
      "grad_norm": 0.24546962976455688,
      "learning_rate": 0.0001,
      "loss": 1.2952,
      "step": 7273
    },
    {
      "epoch": 0.12549608495119596,
      "grad_norm": 0.22820793092250824,
      "learning_rate": 0.0001,
      "loss": 1.2791,
      "step": 7274
    },
    {
      "epoch": 0.12560334656226538,
      "grad_norm": 0.2147456854581833,
      "learning_rate": 0.0001,
      "loss": 1.1952,
      "step": 7275
    },
    {
      "epoch": 0.12571060817333476,
      "grad_norm": 0.2365228533744812,
      "learning_rate": 0.0001,
      "loss": 1.2364,
      "step": 7276
    },
    {
      "epoch": 0.12581786978440415,
      "grad_norm": 0.22065480053424835,
      "learning_rate": 0.0001,
      "loss": 1.3441,
      "step": 7277
    },
    {
      "epoch": 0.12592513139547357,
      "grad_norm": 0.23094935715198517,
      "learning_rate": 0.0001,
      "loss": 1.1865,
      "step": 7278
    },
    {
      "epoch": 0.12603239300654295,
      "grad_norm": 0.2109258472919464,
      "learning_rate": 0.0001,
      "loss": 1.2527,
      "step": 7279
    },
    {
      "epoch": 0.12613965461761237,
      "grad_norm": 0.21417388319969177,
      "learning_rate": 0.0001,
      "loss": 1.269,
      "step": 7280
    },
    {
      "epoch": 0.12624691622868175,
      "grad_norm": 0.22249490022659302,
      "learning_rate": 0.0001,
      "loss": 1.2808,
      "step": 7281
    },
    {
      "epoch": 0.12635417783975114,
      "grad_norm": 0.2130332738161087,
      "learning_rate": 0.0001,
      "loss": 1.2141,
      "step": 7282
    },
    {
      "epoch": 0.12646143945082056,
      "grad_norm": 0.21551837027072906,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 7283
    },
    {
      "epoch": 0.12656870106188994,
      "grad_norm": 0.22709451615810394,
      "learning_rate": 0.0001,
      "loss": 1.2551,
      "step": 7284
    },
    {
      "epoch": 0.12667596267295936,
      "grad_norm": 0.23041512072086334,
      "learning_rate": 0.0001,
      "loss": 1.2229,
      "step": 7285
    },
    {
      "epoch": 0.12678322428402874,
      "grad_norm": 0.21092672646045685,
      "learning_rate": 0.0001,
      "loss": 1.2062,
      "step": 7286
    },
    {
      "epoch": 0.12689048589509813,
      "grad_norm": 0.2081628292798996,
      "learning_rate": 0.0001,
      "loss": 1.2841,
      "step": 7287
    },
    {
      "epoch": 0.12699774750616755,
      "grad_norm": 0.2321757823228836,
      "learning_rate": 0.0001,
      "loss": 1.1799,
      "step": 7288
    },
    {
      "epoch": 0.12710500911723693,
      "grad_norm": 0.21643729507923126,
      "learning_rate": 0.0001,
      "loss": 1.348,
      "step": 7289
    },
    {
      "epoch": 0.12721227072830635,
      "grad_norm": 0.21812884509563446,
      "learning_rate": 0.0001,
      "loss": 1.2889,
      "step": 7290
    },
    {
      "epoch": 0.12731953233937573,
      "grad_norm": 0.21546997129917145,
      "learning_rate": 0.0001,
      "loss": 1.2465,
      "step": 7291
    },
    {
      "epoch": 0.12742679395044512,
      "grad_norm": 0.22816909849643707,
      "learning_rate": 0.0001,
      "loss": 1.2583,
      "step": 7292
    },
    {
      "epoch": 0.12753405556151454,
      "grad_norm": 0.22025728225708008,
      "learning_rate": 0.0001,
      "loss": 1.336,
      "step": 7293
    },
    {
      "epoch": 0.12764131717258392,
      "grad_norm": 0.22823069989681244,
      "learning_rate": 0.0001,
      "loss": 1.2121,
      "step": 7294
    },
    {
      "epoch": 0.12774857878365334,
      "grad_norm": 0.22743895649909973,
      "learning_rate": 0.0001,
      "loss": 1.1922,
      "step": 7295
    },
    {
      "epoch": 0.12785584039472273,
      "grad_norm": 0.2210685759782791,
      "learning_rate": 0.0001,
      "loss": 1.3029,
      "step": 7296
    },
    {
      "epoch": 0.12796310200579214,
      "grad_norm": 0.2207048386335373,
      "learning_rate": 0.0001,
      "loss": 1.2211,
      "step": 7297
    },
    {
      "epoch": 0.12807036361686153,
      "grad_norm": 0.22524914145469666,
      "learning_rate": 0.0001,
      "loss": 1.2253,
      "step": 7298
    },
    {
      "epoch": 0.1281776252279309,
      "grad_norm": 0.22394651174545288,
      "learning_rate": 0.0001,
      "loss": 1.2878,
      "step": 7299
    },
    {
      "epoch": 0.12828488683900033,
      "grad_norm": 0.2094421237707138,
      "learning_rate": 0.0001,
      "loss": 1.1959,
      "step": 7300
    },
    {
      "epoch": 0.12839214845006972,
      "grad_norm": 0.21504253149032593,
      "learning_rate": 0.0001,
      "loss": 1.2495,
      "step": 7301
    },
    {
      "epoch": 0.12849941006113913,
      "grad_norm": 0.20294159650802612,
      "learning_rate": 0.0001,
      "loss": 1.2183,
      "step": 7302
    },
    {
      "epoch": 0.12860667167220852,
      "grad_norm": 0.23378536105155945,
      "learning_rate": 0.0001,
      "loss": 1.3526,
      "step": 7303
    },
    {
      "epoch": 0.1287139332832779,
      "grad_norm": 0.21178215742111206,
      "learning_rate": 0.0001,
      "loss": 1.261,
      "step": 7304
    },
    {
      "epoch": 0.12882119489434732,
      "grad_norm": 0.2220754772424698,
      "learning_rate": 0.0001,
      "loss": 1.2618,
      "step": 7305
    },
    {
      "epoch": 0.1289284565054167,
      "grad_norm": 0.22847160696983337,
      "learning_rate": 0.0001,
      "loss": 1.2721,
      "step": 7306
    },
    {
      "epoch": 0.12903571811648612,
      "grad_norm": 0.22083237767219543,
      "learning_rate": 0.0001,
      "loss": 1.2751,
      "step": 7307
    },
    {
      "epoch": 0.1291429797275555,
      "grad_norm": 0.22392134368419647,
      "learning_rate": 0.0001,
      "loss": 1.2648,
      "step": 7308
    },
    {
      "epoch": 0.1292502413386249,
      "grad_norm": 0.22847767174243927,
      "learning_rate": 0.0001,
      "loss": 1.2145,
      "step": 7309
    },
    {
      "epoch": 0.1293575029496943,
      "grad_norm": 0.22782005369663239,
      "learning_rate": 0.0001,
      "loss": 1.3782,
      "step": 7310
    },
    {
      "epoch": 0.1294647645607637,
      "grad_norm": 0.2091693878173828,
      "learning_rate": 0.0001,
      "loss": 1.2018,
      "step": 7311
    },
    {
      "epoch": 0.1295720261718331,
      "grad_norm": 0.22186699509620667,
      "learning_rate": 0.0001,
      "loss": 1.2565,
      "step": 7312
    },
    {
      "epoch": 0.1296792877829025,
      "grad_norm": 0.21903042495250702,
      "learning_rate": 0.0001,
      "loss": 1.1633,
      "step": 7313
    },
    {
      "epoch": 0.12978654939397188,
      "grad_norm": 0.291425883769989,
      "learning_rate": 0.0001,
      "loss": 1.1905,
      "step": 7314
    },
    {
      "epoch": 0.1298938110050413,
      "grad_norm": 0.22993090748786926,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 7315
    },
    {
      "epoch": 0.13000107261611069,
      "grad_norm": 0.22336658835411072,
      "learning_rate": 0.0001,
      "loss": 1.2398,
      "step": 7316
    },
    {
      "epoch": 0.1301083342271801,
      "grad_norm": 0.22764341533184052,
      "learning_rate": 0.0001,
      "loss": 1.2012,
      "step": 7317
    },
    {
      "epoch": 0.1302155958382495,
      "grad_norm": 0.22279636561870575,
      "learning_rate": 0.0001,
      "loss": 1.3261,
      "step": 7318
    },
    {
      "epoch": 0.1303228574493189,
      "grad_norm": 0.22805555164813995,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 7319
    },
    {
      "epoch": 0.1304301190603883,
      "grad_norm": 0.21445022523403168,
      "learning_rate": 0.0001,
      "loss": 1.2248,
      "step": 7320
    },
    {
      "epoch": 0.13053738067145768,
      "grad_norm": 0.22362449765205383,
      "learning_rate": 0.0001,
      "loss": 1.2042,
      "step": 7321
    },
    {
      "epoch": 0.1306446422825271,
      "grad_norm": 0.23244518041610718,
      "learning_rate": 0.0001,
      "loss": 1.193,
      "step": 7322
    },
    {
      "epoch": 0.13075190389359648,
      "grad_norm": 0.25523293018341064,
      "learning_rate": 0.0001,
      "loss": 1.2659,
      "step": 7323
    },
    {
      "epoch": 0.1308591655046659,
      "grad_norm": 0.22663532197475433,
      "learning_rate": 0.0001,
      "loss": 1.2691,
      "step": 7324
    },
    {
      "epoch": 0.13096642711573528,
      "grad_norm": 0.2108086198568344,
      "learning_rate": 0.0001,
      "loss": 1.1886,
      "step": 7325
    },
    {
      "epoch": 0.13107368872680467,
      "grad_norm": 0.2075616866350174,
      "learning_rate": 0.0001,
      "loss": 1.223,
      "step": 7326
    },
    {
      "epoch": 0.13118095033787408,
      "grad_norm": 0.21860982477664948,
      "learning_rate": 0.0001,
      "loss": 1.2412,
      "step": 7327
    },
    {
      "epoch": 0.13128821194894347,
      "grad_norm": 0.2414669692516327,
      "learning_rate": 0.0001,
      "loss": 1.2626,
      "step": 7328
    },
    {
      "epoch": 0.13139547356001288,
      "grad_norm": 0.24458475410938263,
      "learning_rate": 0.0001,
      "loss": 1.2665,
      "step": 7329
    },
    {
      "epoch": 0.13150273517108227,
      "grad_norm": 0.2134876251220703,
      "learning_rate": 0.0001,
      "loss": 1.2277,
      "step": 7330
    },
    {
      "epoch": 0.13160999678215166,
      "grad_norm": 0.21106500923633575,
      "learning_rate": 0.0001,
      "loss": 1.3195,
      "step": 7331
    },
    {
      "epoch": 0.13171725839322107,
      "grad_norm": 0.21735619008541107,
      "learning_rate": 0.0001,
      "loss": 1.3066,
      "step": 7332
    },
    {
      "epoch": 0.13182452000429046,
      "grad_norm": 0.21203462779521942,
      "learning_rate": 0.0001,
      "loss": 1.3087,
      "step": 7333
    },
    {
      "epoch": 0.13193178161535987,
      "grad_norm": 0.26657184958457947,
      "learning_rate": 0.0001,
      "loss": 1.2156,
      "step": 7334
    },
    {
      "epoch": 0.13203904322642926,
      "grad_norm": 0.2291346937417984,
      "learning_rate": 0.0001,
      "loss": 1.2102,
      "step": 7335
    },
    {
      "epoch": 0.13214630483749865,
      "grad_norm": 0.20637185871601105,
      "learning_rate": 0.0001,
      "loss": 1.3307,
      "step": 7336
    },
    {
      "epoch": 0.13225356644856806,
      "grad_norm": 0.2240259200334549,
      "learning_rate": 0.0001,
      "loss": 1.2623,
      "step": 7337
    },
    {
      "epoch": 0.13236082805963745,
      "grad_norm": 0.2284618765115738,
      "learning_rate": 0.0001,
      "loss": 1.3035,
      "step": 7338
    },
    {
      "epoch": 0.13246808967070686,
      "grad_norm": 0.22282575070858002,
      "learning_rate": 0.0001,
      "loss": 1.3131,
      "step": 7339
    },
    {
      "epoch": 0.13257535128177625,
      "grad_norm": 0.2319929003715515,
      "learning_rate": 0.0001,
      "loss": 1.2044,
      "step": 7340
    },
    {
      "epoch": 0.13268261289284566,
      "grad_norm": 0.22002626955509186,
      "learning_rate": 0.0001,
      "loss": 1.2914,
      "step": 7341
    },
    {
      "epoch": 0.13278987450391505,
      "grad_norm": 0.22511737048625946,
      "learning_rate": 0.0001,
      "loss": 1.2638,
      "step": 7342
    },
    {
      "epoch": 0.13289713611498444,
      "grad_norm": 0.24431124329566956,
      "learning_rate": 0.0001,
      "loss": 1.2019,
      "step": 7343
    },
    {
      "epoch": 0.13300439772605385,
      "grad_norm": 0.2435828149318695,
      "learning_rate": 0.0001,
      "loss": 1.3042,
      "step": 7344
    },
    {
      "epoch": 0.13311165933712324,
      "grad_norm": 0.2162894606590271,
      "learning_rate": 0.0001,
      "loss": 1.2792,
      "step": 7345
    },
    {
      "epoch": 0.13321892094819265,
      "grad_norm": 0.21546246111392975,
      "learning_rate": 0.0001,
      "loss": 1.2612,
      "step": 7346
    },
    {
      "epoch": 0.13332618255926204,
      "grad_norm": 0.20840440690517426,
      "learning_rate": 0.0001,
      "loss": 1.2582,
      "step": 7347
    },
    {
      "epoch": 0.13343344417033143,
      "grad_norm": 0.22037628293037415,
      "learning_rate": 0.0001,
      "loss": 1.1845,
      "step": 7348
    },
    {
      "epoch": 0.13354070578140084,
      "grad_norm": 0.20994170010089874,
      "learning_rate": 0.0001,
      "loss": 1.2232,
      "step": 7349
    },
    {
      "epoch": 0.13364796739247023,
      "grad_norm": 0.214602530002594,
      "learning_rate": 0.0001,
      "loss": 1.3732,
      "step": 7350
    },
    {
      "epoch": 0.13375522900353964,
      "grad_norm": 0.21441945433616638,
      "learning_rate": 0.0001,
      "loss": 1.3116,
      "step": 7351
    },
    {
      "epoch": 0.13386249061460903,
      "grad_norm": 0.22298334538936615,
      "learning_rate": 0.0001,
      "loss": 1.2201,
      "step": 7352
    },
    {
      "epoch": 0.13396975222567842,
      "grad_norm": 0.22880345582962036,
      "learning_rate": 0.0001,
      "loss": 1.3111,
      "step": 7353
    },
    {
      "epoch": 0.13407701383674783,
      "grad_norm": 0.21852509677410126,
      "learning_rate": 0.0001,
      "loss": 1.3102,
      "step": 7354
    },
    {
      "epoch": 0.13418427544781722,
      "grad_norm": 0.21749910712242126,
      "learning_rate": 0.0001,
      "loss": 1.1316,
      "step": 7355
    },
    {
      "epoch": 0.13429153705888663,
      "grad_norm": 0.21988515555858612,
      "learning_rate": 0.0001,
      "loss": 1.2833,
      "step": 7356
    },
    {
      "epoch": 0.13439879866995602,
      "grad_norm": 0.2285822331905365,
      "learning_rate": 0.0001,
      "loss": 1.2295,
      "step": 7357
    },
    {
      "epoch": 0.1345060602810254,
      "grad_norm": 0.21519239246845245,
      "learning_rate": 0.0001,
      "loss": 1.2157,
      "step": 7358
    },
    {
      "epoch": 0.13461332189209482,
      "grad_norm": 0.21863114833831787,
      "learning_rate": 0.0001,
      "loss": 1.207,
      "step": 7359
    },
    {
      "epoch": 0.1347205835031642,
      "grad_norm": 0.2121662199497223,
      "learning_rate": 0.0001,
      "loss": 1.202,
      "step": 7360
    },
    {
      "epoch": 0.13482784511423362,
      "grad_norm": 0.22282774746418,
      "learning_rate": 0.0001,
      "loss": 1.3031,
      "step": 7361
    },
    {
      "epoch": 0.134935106725303,
      "grad_norm": 0.21829171478748322,
      "learning_rate": 0.0001,
      "loss": 1.1849,
      "step": 7362
    },
    {
      "epoch": 0.13504236833637243,
      "grad_norm": 0.21028287708759308,
      "learning_rate": 0.0001,
      "loss": 1.273,
      "step": 7363
    },
    {
      "epoch": 0.1351496299474418,
      "grad_norm": 0.2205718606710434,
      "learning_rate": 0.0001,
      "loss": 1.2439,
      "step": 7364
    },
    {
      "epoch": 0.1352568915585112,
      "grad_norm": 0.21975602209568024,
      "learning_rate": 0.0001,
      "loss": 1.2761,
      "step": 7365
    },
    {
      "epoch": 0.13536415316958061,
      "grad_norm": 0.23095129430294037,
      "learning_rate": 0.0001,
      "loss": 1.3075,
      "step": 7366
    },
    {
      "epoch": 0.13547141478065,
      "grad_norm": 0.2073153257369995,
      "learning_rate": 0.0001,
      "loss": 1.2014,
      "step": 7367
    },
    {
      "epoch": 0.13557867639171942,
      "grad_norm": 0.22174766659736633,
      "learning_rate": 0.0001,
      "loss": 1.2357,
      "step": 7368
    },
    {
      "epoch": 0.1356859380027888,
      "grad_norm": 0.22985218465328217,
      "learning_rate": 0.0001,
      "loss": 1.2962,
      "step": 7369
    },
    {
      "epoch": 0.1357931996138582,
      "grad_norm": 0.21632228791713715,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 7370
    },
    {
      "epoch": 0.1359004612249276,
      "grad_norm": 0.22658906877040863,
      "learning_rate": 0.0001,
      "loss": 1.2185,
      "step": 7371
    },
    {
      "epoch": 0.136007722835997,
      "grad_norm": 0.21176491677761078,
      "learning_rate": 0.0001,
      "loss": 1.3122,
      "step": 7372
    },
    {
      "epoch": 0.1361149844470664,
      "grad_norm": 0.21737191081047058,
      "learning_rate": 0.0001,
      "loss": 1.2785,
      "step": 7373
    },
    {
      "epoch": 0.1362222460581358,
      "grad_norm": 0.22405731678009033,
      "learning_rate": 0.0001,
      "loss": 1.3211,
      "step": 7374
    },
    {
      "epoch": 0.13632950766920518,
      "grad_norm": 0.21005059778690338,
      "learning_rate": 0.0001,
      "loss": 1.2088,
      "step": 7375
    },
    {
      "epoch": 0.1364367692802746,
      "grad_norm": 0.22044815123081207,
      "learning_rate": 0.0001,
      "loss": 1.2935,
      "step": 7376
    },
    {
      "epoch": 0.13654403089134398,
      "grad_norm": 0.22234636545181274,
      "learning_rate": 0.0001,
      "loss": 1.1836,
      "step": 7377
    },
    {
      "epoch": 0.1366512925024134,
      "grad_norm": 0.2288372814655304,
      "learning_rate": 0.0001,
      "loss": 1.2952,
      "step": 7378
    },
    {
      "epoch": 0.13675855411348278,
      "grad_norm": 0.21990464627742767,
      "learning_rate": 0.0001,
      "loss": 1.2285,
      "step": 7379
    },
    {
      "epoch": 0.13686581572455217,
      "grad_norm": 0.2122788280248642,
      "learning_rate": 0.0001,
      "loss": 1.2789,
      "step": 7380
    },
    {
      "epoch": 0.13697307733562158,
      "grad_norm": 0.227971151471138,
      "learning_rate": 0.0001,
      "loss": 1.3407,
      "step": 7381
    },
    {
      "epoch": 0.13708033894669097,
      "grad_norm": 0.20982177555561066,
      "learning_rate": 0.0001,
      "loss": 1.2517,
      "step": 7382
    },
    {
      "epoch": 0.1371876005577604,
      "grad_norm": 0.21715177595615387,
      "learning_rate": 0.0001,
      "loss": 1.2034,
      "step": 7383
    },
    {
      "epoch": 0.13729486216882977,
      "grad_norm": 0.21131263673305511,
      "learning_rate": 0.0001,
      "loss": 1.2674,
      "step": 7384
    },
    {
      "epoch": 0.1374021237798992,
      "grad_norm": 0.23115260899066925,
      "learning_rate": 0.0001,
      "loss": 1.1829,
      "step": 7385
    },
    {
      "epoch": 0.13750938539096857,
      "grad_norm": 0.2236284613609314,
      "learning_rate": 0.0001,
      "loss": 1.2947,
      "step": 7386
    },
    {
      "epoch": 0.13761664700203796,
      "grad_norm": 0.2353920340538025,
      "learning_rate": 0.0001,
      "loss": 1.2847,
      "step": 7387
    },
    {
      "epoch": 0.13772390861310738,
      "grad_norm": 0.2111012190580368,
      "learning_rate": 0.0001,
      "loss": 1.2266,
      "step": 7388
    },
    {
      "epoch": 0.13783117022417676,
      "grad_norm": 0.21885865926742554,
      "learning_rate": 0.0001,
      "loss": 1.2972,
      "step": 7389
    },
    {
      "epoch": 0.13793843183524618,
      "grad_norm": 0.20621004700660706,
      "learning_rate": 0.0001,
      "loss": 1.2863,
      "step": 7390
    },
    {
      "epoch": 0.13804569344631556,
      "grad_norm": 0.21477754414081573,
      "learning_rate": 0.0001,
      "loss": 1.2339,
      "step": 7391
    },
    {
      "epoch": 0.13815295505738495,
      "grad_norm": 0.2243403047323227,
      "learning_rate": 0.0001,
      "loss": 1.2897,
      "step": 7392
    },
    {
      "epoch": 0.13826021666845437,
      "grad_norm": 0.22679166495800018,
      "learning_rate": 0.0001,
      "loss": 1.1602,
      "step": 7393
    },
    {
      "epoch": 0.13836747827952375,
      "grad_norm": 0.23562149703502655,
      "learning_rate": 0.0001,
      "loss": 1.2536,
      "step": 7394
    },
    {
      "epoch": 0.13847473989059317,
      "grad_norm": 0.2169082760810852,
      "learning_rate": 0.0001,
      "loss": 1.266,
      "step": 7395
    },
    {
      "epoch": 0.13858200150166256,
      "grad_norm": 0.20646169781684875,
      "learning_rate": 0.0001,
      "loss": 1.1714,
      "step": 7396
    },
    {
      "epoch": 0.13868926311273194,
      "grad_norm": 0.2207699716091156,
      "learning_rate": 0.0001,
      "loss": 1.2493,
      "step": 7397
    },
    {
      "epoch": 0.13879652472380136,
      "grad_norm": 0.20963113009929657,
      "learning_rate": 0.0001,
      "loss": 1.2827,
      "step": 7398
    },
    {
      "epoch": 0.13890378633487074,
      "grad_norm": 0.21021732687950134,
      "learning_rate": 0.0001,
      "loss": 1.2153,
      "step": 7399
    },
    {
      "epoch": 0.13901104794594016,
      "grad_norm": 0.214704230427742,
      "learning_rate": 0.0001,
      "loss": 1.2037,
      "step": 7400
    },
    {
      "epoch": 0.13911830955700955,
      "grad_norm": 0.22269321978092194,
      "learning_rate": 0.0001,
      "loss": 1.241,
      "step": 7401
    },
    {
      "epoch": 0.13922557116807893,
      "grad_norm": 0.22747443616390228,
      "learning_rate": 0.0001,
      "loss": 1.2705,
      "step": 7402
    },
    {
      "epoch": 0.13933283277914835,
      "grad_norm": 0.2369653731584549,
      "learning_rate": 0.0001,
      "loss": 1.2143,
      "step": 7403
    },
    {
      "epoch": 0.13944009439021773,
      "grad_norm": 0.22418856620788574,
      "learning_rate": 0.0001,
      "loss": 1.3156,
      "step": 7404
    },
    {
      "epoch": 0.13954735600128715,
      "grad_norm": 0.2203410416841507,
      "learning_rate": 0.0001,
      "loss": 1.3041,
      "step": 7405
    },
    {
      "epoch": 0.13965461761235654,
      "grad_norm": 0.2172269970178604,
      "learning_rate": 0.0001,
      "loss": 1.2268,
      "step": 7406
    },
    {
      "epoch": 0.13976187922342592,
      "grad_norm": 0.21040946245193481,
      "learning_rate": 0.0001,
      "loss": 1.2078,
      "step": 7407
    },
    {
      "epoch": 0.13986914083449534,
      "grad_norm": 0.22879821062088013,
      "learning_rate": 0.0001,
      "loss": 1.2531,
      "step": 7408
    },
    {
      "epoch": 0.13997640244556472,
      "grad_norm": 0.2215765118598938,
      "learning_rate": 0.0001,
      "loss": 1.261,
      "step": 7409
    },
    {
      "epoch": 0.14008366405663414,
      "grad_norm": 0.21393489837646484,
      "learning_rate": 0.0001,
      "loss": 1.258,
      "step": 7410
    },
    {
      "epoch": 0.14019092566770353,
      "grad_norm": 0.2115493267774582,
      "learning_rate": 0.0001,
      "loss": 1.2967,
      "step": 7411
    },
    {
      "epoch": 0.14029818727877294,
      "grad_norm": 0.21597613394260406,
      "learning_rate": 0.0001,
      "loss": 1.201,
      "step": 7412
    },
    {
      "epoch": 0.14040544888984233,
      "grad_norm": 0.22114989161491394,
      "learning_rate": 0.0001,
      "loss": 1.3327,
      "step": 7413
    },
    {
      "epoch": 0.14051271050091171,
      "grad_norm": 0.21591979265213013,
      "learning_rate": 0.0001,
      "loss": 1.3424,
      "step": 7414
    },
    {
      "epoch": 0.14061997211198113,
      "grad_norm": 0.22995525598526,
      "learning_rate": 0.0001,
      "loss": 1.2458,
      "step": 7415
    },
    {
      "epoch": 0.14072723372305052,
      "grad_norm": 0.22999811172485352,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 7416
    },
    {
      "epoch": 0.14083449533411993,
      "grad_norm": 0.21793384850025177,
      "learning_rate": 0.0001,
      "loss": 1.3339,
      "step": 7417
    },
    {
      "epoch": 0.14094175694518932,
      "grad_norm": 0.20320507884025574,
      "learning_rate": 0.0001,
      "loss": 1.2437,
      "step": 7418
    },
    {
      "epoch": 0.1410490185562587,
      "grad_norm": 0.20638053119182587,
      "learning_rate": 0.0001,
      "loss": 1.2164,
      "step": 7419
    },
    {
      "epoch": 0.14115628016732812,
      "grad_norm": 0.21222715079784393,
      "learning_rate": 0.0001,
      "loss": 1.2766,
      "step": 7420
    },
    {
      "epoch": 0.1412635417783975,
      "grad_norm": 0.2159426212310791,
      "learning_rate": 0.0001,
      "loss": 1.3181,
      "step": 7421
    },
    {
      "epoch": 0.14137080338946692,
      "grad_norm": 0.203031525015831,
      "learning_rate": 0.0001,
      "loss": 1.1866,
      "step": 7422
    },
    {
      "epoch": 0.1414780650005363,
      "grad_norm": 0.20765811204910278,
      "learning_rate": 0.0001,
      "loss": 1.2966,
      "step": 7423
    },
    {
      "epoch": 0.1415853266116057,
      "grad_norm": 0.21198171377182007,
      "learning_rate": 0.0001,
      "loss": 1.1915,
      "step": 7424
    },
    {
      "epoch": 0.1416925882226751,
      "grad_norm": 0.2049105018377304,
      "learning_rate": 0.0001,
      "loss": 1.1922,
      "step": 7425
    },
    {
      "epoch": 0.1417998498337445,
      "grad_norm": 0.21518218517303467,
      "learning_rate": 0.0001,
      "loss": 1.2252,
      "step": 7426
    },
    {
      "epoch": 0.1419071114448139,
      "grad_norm": 0.24382136762142181,
      "learning_rate": 0.0001,
      "loss": 1.2816,
      "step": 7427
    },
    {
      "epoch": 0.1420143730558833,
      "grad_norm": 0.20301829278469086,
      "learning_rate": 0.0001,
      "loss": 1.1672,
      "step": 7428
    },
    {
      "epoch": 0.14212163466695268,
      "grad_norm": 0.2224283218383789,
      "learning_rate": 0.0001,
      "loss": 1.2425,
      "step": 7429
    },
    {
      "epoch": 0.1422288962780221,
      "grad_norm": 0.2142074555158615,
      "learning_rate": 0.0001,
      "loss": 1.2669,
      "step": 7430
    },
    {
      "epoch": 0.14233615788909149,
      "grad_norm": 0.21854884922504425,
      "learning_rate": 0.0001,
      "loss": 1.2838,
      "step": 7431
    },
    {
      "epoch": 0.1424434195001609,
      "grad_norm": 0.2232360541820526,
      "learning_rate": 0.0001,
      "loss": 1.266,
      "step": 7432
    },
    {
      "epoch": 0.1425506811112303,
      "grad_norm": 0.22172744572162628,
      "learning_rate": 0.0001,
      "loss": 1.3032,
      "step": 7433
    },
    {
      "epoch": 0.1426579427222997,
      "grad_norm": 0.21986456215381622,
      "learning_rate": 0.0001,
      "loss": 1.3563,
      "step": 7434
    },
    {
      "epoch": 0.1427652043333691,
      "grad_norm": 0.2260645180940628,
      "learning_rate": 0.0001,
      "loss": 1.209,
      "step": 7435
    },
    {
      "epoch": 0.14287246594443848,
      "grad_norm": 0.22453837096691132,
      "learning_rate": 0.0001,
      "loss": 1.2098,
      "step": 7436
    },
    {
      "epoch": 0.1429797275555079,
      "grad_norm": 0.2165333330631256,
      "learning_rate": 0.0001,
      "loss": 1.1999,
      "step": 7437
    },
    {
      "epoch": 0.14308698916657728,
      "grad_norm": 0.22826220095157623,
      "learning_rate": 0.0001,
      "loss": 1.2978,
      "step": 7438
    },
    {
      "epoch": 0.1431942507776467,
      "grad_norm": 0.22917716205120087,
      "learning_rate": 0.0001,
      "loss": 1.3192,
      "step": 7439
    },
    {
      "epoch": 0.14330151238871608,
      "grad_norm": 0.22922612726688385,
      "learning_rate": 0.0001,
      "loss": 1.2682,
      "step": 7440
    },
    {
      "epoch": 0.14340877399978547,
      "grad_norm": 0.22417153418064117,
      "learning_rate": 0.0001,
      "loss": 1.2688,
      "step": 7441
    },
    {
      "epoch": 0.14351603561085488,
      "grad_norm": 0.2236444354057312,
      "learning_rate": 0.0001,
      "loss": 1.3236,
      "step": 7442
    },
    {
      "epoch": 0.14362329722192427,
      "grad_norm": 0.22584351897239685,
      "learning_rate": 0.0001,
      "loss": 1.29,
      "step": 7443
    },
    {
      "epoch": 0.14373055883299368,
      "grad_norm": 0.2304764688014984,
      "learning_rate": 0.0001,
      "loss": 1.2109,
      "step": 7444
    },
    {
      "epoch": 0.14383782044406307,
      "grad_norm": 0.21495509147644043,
      "learning_rate": 0.0001,
      "loss": 1.185,
      "step": 7445
    },
    {
      "epoch": 0.14394508205513246,
      "grad_norm": 0.23522454500198364,
      "learning_rate": 0.0001,
      "loss": 1.2033,
      "step": 7446
    },
    {
      "epoch": 0.14405234366620187,
      "grad_norm": 0.21886788308620453,
      "learning_rate": 0.0001,
      "loss": 1.326,
      "step": 7447
    },
    {
      "epoch": 0.14415960527727126,
      "grad_norm": 0.23325961828231812,
      "learning_rate": 0.0001,
      "loss": 1.2252,
      "step": 7448
    },
    {
      "epoch": 0.14426686688834067,
      "grad_norm": 0.2247277796268463,
      "learning_rate": 0.0001,
      "loss": 1.2893,
      "step": 7449
    },
    {
      "epoch": 0.14437412849941006,
      "grad_norm": 0.2275807410478592,
      "learning_rate": 0.0001,
      "loss": 1.2004,
      "step": 7450
    },
    {
      "epoch": 0.14448139011047945,
      "grad_norm": 0.22689233720302582,
      "learning_rate": 0.0001,
      "loss": 1.3286,
      "step": 7451
    },
    {
      "epoch": 0.14458865172154886,
      "grad_norm": 0.2112688422203064,
      "learning_rate": 0.0001,
      "loss": 1.2163,
      "step": 7452
    },
    {
      "epoch": 0.14469591333261825,
      "grad_norm": 0.2171718031167984,
      "learning_rate": 0.0001,
      "loss": 1.182,
      "step": 7453
    },
    {
      "epoch": 0.14480317494368766,
      "grad_norm": 0.2229878306388855,
      "learning_rate": 0.0001,
      "loss": 1.2876,
      "step": 7454
    },
    {
      "epoch": 0.14491043655475705,
      "grad_norm": 0.2277033030986786,
      "learning_rate": 0.0001,
      "loss": 1.2488,
      "step": 7455
    },
    {
      "epoch": 0.14501769816582646,
      "grad_norm": 0.22310587763786316,
      "learning_rate": 0.0001,
      "loss": 1.1467,
      "step": 7456
    },
    {
      "epoch": 0.14512495977689585,
      "grad_norm": 0.21557246148586273,
      "learning_rate": 0.0001,
      "loss": 1.1927,
      "step": 7457
    },
    {
      "epoch": 0.14523222138796524,
      "grad_norm": 0.21065425872802734,
      "learning_rate": 0.0001,
      "loss": 1.2772,
      "step": 7458
    },
    {
      "epoch": 0.14533948299903465,
      "grad_norm": 0.219790518283844,
      "learning_rate": 0.0001,
      "loss": 1.3133,
      "step": 7459
    },
    {
      "epoch": 0.14544674461010404,
      "grad_norm": 0.2130134105682373,
      "learning_rate": 0.0001,
      "loss": 1.1685,
      "step": 7460
    },
    {
      "epoch": 0.14555400622117345,
      "grad_norm": 0.21193543076515198,
      "learning_rate": 0.0001,
      "loss": 1.2458,
      "step": 7461
    },
    {
      "epoch": 0.14566126783224284,
      "grad_norm": 0.22570787370204926,
      "learning_rate": 0.0001,
      "loss": 1.2225,
      "step": 7462
    },
    {
      "epoch": 0.14576852944331223,
      "grad_norm": 0.21434463560581207,
      "learning_rate": 0.0001,
      "loss": 1.3107,
      "step": 7463
    },
    {
      "epoch": 0.14587579105438164,
      "grad_norm": 0.21946077048778534,
      "learning_rate": 0.0001,
      "loss": 1.2405,
      "step": 7464
    },
    {
      "epoch": 0.14598305266545103,
      "grad_norm": 0.22359634935855865,
      "learning_rate": 0.0001,
      "loss": 1.334,
      "step": 7465
    },
    {
      "epoch": 0.14609031427652044,
      "grad_norm": 0.22306999564170837,
      "learning_rate": 0.0001,
      "loss": 1.3185,
      "step": 7466
    },
    {
      "epoch": 0.14619757588758983,
      "grad_norm": 0.23097921907901764,
      "learning_rate": 0.0001,
      "loss": 1.2854,
      "step": 7467
    },
    {
      "epoch": 0.14630483749865922,
      "grad_norm": 0.22629228234291077,
      "learning_rate": 0.0001,
      "loss": 1.2545,
      "step": 7468
    },
    {
      "epoch": 0.14641209910972863,
      "grad_norm": 0.21986086666584015,
      "learning_rate": 0.0001,
      "loss": 1.1522,
      "step": 7469
    },
    {
      "epoch": 0.14651936072079802,
      "grad_norm": 0.22093212604522705,
      "learning_rate": 0.0001,
      "loss": 1.2177,
      "step": 7470
    },
    {
      "epoch": 0.14662662233186743,
      "grad_norm": 0.2293102741241455,
      "learning_rate": 0.0001,
      "loss": 1.2367,
      "step": 7471
    },
    {
      "epoch": 0.14673388394293682,
      "grad_norm": 0.21750828623771667,
      "learning_rate": 0.0001,
      "loss": 1.2238,
      "step": 7472
    },
    {
      "epoch": 0.1468411455540062,
      "grad_norm": 0.21366819739341736,
      "learning_rate": 0.0001,
      "loss": 1.2647,
      "step": 7473
    },
    {
      "epoch": 0.14694840716507562,
      "grad_norm": 0.2057972103357315,
      "learning_rate": 0.0001,
      "loss": 1.1798,
      "step": 7474
    },
    {
      "epoch": 0.147055668776145,
      "grad_norm": 0.2237442284822464,
      "learning_rate": 0.0001,
      "loss": 1.2174,
      "step": 7475
    },
    {
      "epoch": 0.14716293038721442,
      "grad_norm": 0.22452835738658905,
      "learning_rate": 0.0001,
      "loss": 1.2153,
      "step": 7476
    },
    {
      "epoch": 0.1472701919982838,
      "grad_norm": 0.22250615060329437,
      "learning_rate": 0.0001,
      "loss": 1.2775,
      "step": 7477
    },
    {
      "epoch": 0.14737745360935323,
      "grad_norm": 0.2081121951341629,
      "learning_rate": 0.0001,
      "loss": 1.2015,
      "step": 7478
    },
    {
      "epoch": 0.1474847152204226,
      "grad_norm": 0.21687333285808563,
      "learning_rate": 0.0001,
      "loss": 1.2319,
      "step": 7479
    },
    {
      "epoch": 0.147591976831492,
      "grad_norm": 0.21910057961940765,
      "learning_rate": 0.0001,
      "loss": 1.2787,
      "step": 7480
    },
    {
      "epoch": 0.14769923844256141,
      "grad_norm": 0.22107797861099243,
      "learning_rate": 0.0001,
      "loss": 1.3454,
      "step": 7481
    },
    {
      "epoch": 0.1478065000536308,
      "grad_norm": 0.20737852156162262,
      "learning_rate": 0.0001,
      "loss": 1.2692,
      "step": 7482
    },
    {
      "epoch": 0.14791376166470022,
      "grad_norm": 0.20955613255500793,
      "learning_rate": 0.0001,
      "loss": 1.2002,
      "step": 7483
    },
    {
      "epoch": 0.1480210232757696,
      "grad_norm": 0.21972039341926575,
      "learning_rate": 0.0001,
      "loss": 1.2396,
      "step": 7484
    },
    {
      "epoch": 0.148128284886839,
      "grad_norm": 0.22511513531208038,
      "learning_rate": 0.0001,
      "loss": 1.2857,
      "step": 7485
    },
    {
      "epoch": 0.1482355464979084,
      "grad_norm": 0.2465609610080719,
      "learning_rate": 0.0001,
      "loss": 1.2571,
      "step": 7486
    },
    {
      "epoch": 0.1483428081089778,
      "grad_norm": 0.2184850424528122,
      "learning_rate": 0.0001,
      "loss": 1.283,
      "step": 7487
    },
    {
      "epoch": 0.1484500697200472,
      "grad_norm": 0.21637128293514252,
      "learning_rate": 0.0001,
      "loss": 1.2593,
      "step": 7488
    },
    {
      "epoch": 0.1485573313311166,
      "grad_norm": 0.24443204700946808,
      "learning_rate": 0.0001,
      "loss": 1.2989,
      "step": 7489
    },
    {
      "epoch": 0.14866459294218598,
      "grad_norm": 0.21025007963180542,
      "learning_rate": 0.0001,
      "loss": 1.2951,
      "step": 7490
    },
    {
      "epoch": 0.1487718545532554,
      "grad_norm": 0.21977244317531586,
      "learning_rate": 0.0001,
      "loss": 1.3337,
      "step": 7491
    },
    {
      "epoch": 0.14887911616432478,
      "grad_norm": 0.20850811898708344,
      "learning_rate": 0.0001,
      "loss": 1.3121,
      "step": 7492
    },
    {
      "epoch": 0.1489863777753942,
      "grad_norm": 0.20333832502365112,
      "learning_rate": 0.0001,
      "loss": 1.249,
      "step": 7493
    },
    {
      "epoch": 0.14909363938646358,
      "grad_norm": 0.21459797024726868,
      "learning_rate": 0.0001,
      "loss": 1.247,
      "step": 7494
    },
    {
      "epoch": 0.14920090099753297,
      "grad_norm": 0.2220073938369751,
      "learning_rate": 0.0001,
      "loss": 1.2522,
      "step": 7495
    },
    {
      "epoch": 0.14930816260860238,
      "grad_norm": 0.2282019704580307,
      "learning_rate": 0.0001,
      "loss": 1.2878,
      "step": 7496
    },
    {
      "epoch": 0.14941542421967177,
      "grad_norm": 0.22780145704746246,
      "learning_rate": 0.0001,
      "loss": 1.2883,
      "step": 7497
    },
    {
      "epoch": 0.1495226858307412,
      "grad_norm": 0.22712218761444092,
      "learning_rate": 0.0001,
      "loss": 1.3433,
      "step": 7498
    },
    {
      "epoch": 0.14962994744181057,
      "grad_norm": 0.2315959334373474,
      "learning_rate": 0.0001,
      "loss": 1.2455,
      "step": 7499
    },
    {
      "epoch": 0.14973720905287996,
      "grad_norm": 0.21803683042526245,
      "learning_rate": 0.0001,
      "loss": 1.2616,
      "step": 7500
    },
    {
      "epoch": 0.14984447066394938,
      "grad_norm": 0.23665063083171844,
      "learning_rate": 0.0001,
      "loss": 1.2798,
      "step": 7501
    },
    {
      "epoch": 0.14995173227501876,
      "grad_norm": 0.20731309056282043,
      "learning_rate": 0.0001,
      "loss": 1.248,
      "step": 7502
    },
    {
      "epoch": 0.15005899388608818,
      "grad_norm": 0.2218945026397705,
      "learning_rate": 0.0001,
      "loss": 1.3509,
      "step": 7503
    },
    {
      "epoch": 0.15016625549715756,
      "grad_norm": 0.211062490940094,
      "learning_rate": 0.0001,
      "loss": 1.1838,
      "step": 7504
    },
    {
      "epoch": 0.15027351710822698,
      "grad_norm": 0.219604030251503,
      "learning_rate": 0.0001,
      "loss": 1.2193,
      "step": 7505
    },
    {
      "epoch": 0.15038077871929637,
      "grad_norm": 0.24293804168701172,
      "learning_rate": 0.0001,
      "loss": 1.2306,
      "step": 7506
    },
    {
      "epoch": 0.15048804033036575,
      "grad_norm": 0.2143552154302597,
      "learning_rate": 0.0001,
      "loss": 1.1248,
      "step": 7507
    },
    {
      "epoch": 0.15059530194143517,
      "grad_norm": 0.2229606956243515,
      "learning_rate": 0.0001,
      "loss": 1.2263,
      "step": 7508
    },
    {
      "epoch": 0.15070256355250455,
      "grad_norm": 0.23630668222904205,
      "learning_rate": 0.0001,
      "loss": 1.2832,
      "step": 7509
    },
    {
      "epoch": 0.15080982516357397,
      "grad_norm": 0.22654303908348083,
      "learning_rate": 0.0001,
      "loss": 1.2258,
      "step": 7510
    },
    {
      "epoch": 0.15091708677464336,
      "grad_norm": 0.22789053618907928,
      "learning_rate": 0.0001,
      "loss": 1.286,
      "step": 7511
    },
    {
      "epoch": 0.15102434838571274,
      "grad_norm": 0.20878471434116364,
      "learning_rate": 0.0001,
      "loss": 1.1308,
      "step": 7512
    },
    {
      "epoch": 0.15113160999678216,
      "grad_norm": 0.20191791653633118,
      "learning_rate": 0.0001,
      "loss": 1.2661,
      "step": 7513
    },
    {
      "epoch": 0.15123887160785154,
      "grad_norm": 0.23133893311023712,
      "learning_rate": 0.0001,
      "loss": 1.2931,
      "step": 7514
    },
    {
      "epoch": 0.15134613321892096,
      "grad_norm": 0.22101359069347382,
      "learning_rate": 0.0001,
      "loss": 1.1171,
      "step": 7515
    },
    {
      "epoch": 0.15145339482999035,
      "grad_norm": 0.20708045363426208,
      "learning_rate": 0.0001,
      "loss": 1.1571,
      "step": 7516
    },
    {
      "epoch": 0.15156065644105973,
      "grad_norm": 0.23044602572917938,
      "learning_rate": 0.0001,
      "loss": 1.2887,
      "step": 7517
    },
    {
      "epoch": 0.15166791805212915,
      "grad_norm": 0.22491341829299927,
      "learning_rate": 0.0001,
      "loss": 1.3311,
      "step": 7518
    },
    {
      "epoch": 0.15177517966319853,
      "grad_norm": 0.22092397511005402,
      "learning_rate": 0.0001,
      "loss": 1.2245,
      "step": 7519
    },
    {
      "epoch": 0.15188244127426795,
      "grad_norm": 0.2145213633775711,
      "learning_rate": 0.0001,
      "loss": 1.3306,
      "step": 7520
    },
    {
      "epoch": 0.15198970288533734,
      "grad_norm": 0.2133486121892929,
      "learning_rate": 0.0001,
      "loss": 1.2897,
      "step": 7521
    },
    {
      "epoch": 0.15209696449640672,
      "grad_norm": 0.2251446694135666,
      "learning_rate": 0.0001,
      "loss": 1.2461,
      "step": 7522
    },
    {
      "epoch": 0.15220422610747614,
      "grad_norm": 0.2238990068435669,
      "learning_rate": 0.0001,
      "loss": 1.2747,
      "step": 7523
    },
    {
      "epoch": 0.15231148771854552,
      "grad_norm": 0.2231072187423706,
      "learning_rate": 0.0001,
      "loss": 1.2394,
      "step": 7524
    },
    {
      "epoch": 0.15241874932961494,
      "grad_norm": 0.22413042187690735,
      "learning_rate": 0.0001,
      "loss": 1.2658,
      "step": 7525
    },
    {
      "epoch": 0.15252601094068433,
      "grad_norm": 0.226658895611763,
      "learning_rate": 0.0001,
      "loss": 1.3672,
      "step": 7526
    },
    {
      "epoch": 0.15263327255175374,
      "grad_norm": 0.24440810084342957,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 7527
    },
    {
      "epoch": 0.15274053416282313,
      "grad_norm": 0.21468065679073334,
      "learning_rate": 0.0001,
      "loss": 1.2522,
      "step": 7528
    },
    {
      "epoch": 0.15284779577389251,
      "grad_norm": 0.22186505794525146,
      "learning_rate": 0.0001,
      "loss": 1.2172,
      "step": 7529
    },
    {
      "epoch": 0.15295505738496193,
      "grad_norm": 0.21467094123363495,
      "learning_rate": 0.0001,
      "loss": 1.3183,
      "step": 7530
    },
    {
      "epoch": 0.15306231899603132,
      "grad_norm": 0.23268400132656097,
      "learning_rate": 0.0001,
      "loss": 1.259,
      "step": 7531
    },
    {
      "epoch": 0.15316958060710073,
      "grad_norm": 0.22642701864242554,
      "learning_rate": 0.0001,
      "loss": 1.2566,
      "step": 7532
    },
    {
      "epoch": 0.15327684221817012,
      "grad_norm": 0.20371437072753906,
      "learning_rate": 0.0001,
      "loss": 1.2351,
      "step": 7533
    },
    {
      "epoch": 0.1533841038292395,
      "grad_norm": 0.22084926068782806,
      "learning_rate": 0.0001,
      "loss": 1.2906,
      "step": 7534
    },
    {
      "epoch": 0.15349136544030892,
      "grad_norm": 0.23433780670166016,
      "learning_rate": 0.0001,
      "loss": 1.3189,
      "step": 7535
    },
    {
      "epoch": 0.1535986270513783,
      "grad_norm": 0.2045735865831375,
      "learning_rate": 0.0001,
      "loss": 1.2106,
      "step": 7536
    },
    {
      "epoch": 0.15370588866244772,
      "grad_norm": 0.2098902314901352,
      "learning_rate": 0.0001,
      "loss": 1.1726,
      "step": 7537
    },
    {
      "epoch": 0.1538131502735171,
      "grad_norm": 0.2297247350215912,
      "learning_rate": 0.0001,
      "loss": 1.1494,
      "step": 7538
    },
    {
      "epoch": 0.1539204118845865,
      "grad_norm": 0.22339855134487152,
      "learning_rate": 0.0001,
      "loss": 1.2728,
      "step": 7539
    },
    {
      "epoch": 0.1540276734956559,
      "grad_norm": 0.21853481233119965,
      "learning_rate": 0.0001,
      "loss": 1.2396,
      "step": 7540
    },
    {
      "epoch": 0.1541349351067253,
      "grad_norm": 0.22552470862865448,
      "learning_rate": 0.0001,
      "loss": 1.2768,
      "step": 7541
    },
    {
      "epoch": 0.1542421967177947,
      "grad_norm": 0.21458329260349274,
      "learning_rate": 0.0001,
      "loss": 1.2582,
      "step": 7542
    },
    {
      "epoch": 0.1543494583288641,
      "grad_norm": 0.21953248977661133,
      "learning_rate": 0.0001,
      "loss": 1.2314,
      "step": 7543
    },
    {
      "epoch": 0.15445671993993348,
      "grad_norm": 0.24370814859867096,
      "learning_rate": 0.0001,
      "loss": 1.4001,
      "step": 7544
    },
    {
      "epoch": 0.1545639815510029,
      "grad_norm": 0.21648815274238586,
      "learning_rate": 0.0001,
      "loss": 1.1265,
      "step": 7545
    },
    {
      "epoch": 0.15467124316207229,
      "grad_norm": 0.23643086850643158,
      "learning_rate": 0.0001,
      "loss": 1.3593,
      "step": 7546
    },
    {
      "epoch": 0.1547785047731417,
      "grad_norm": 0.2101329267024994,
      "learning_rate": 0.0001,
      "loss": 1.1477,
      "step": 7547
    },
    {
      "epoch": 0.1548857663842111,
      "grad_norm": 0.21877539157867432,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 7548
    },
    {
      "epoch": 0.1549930279952805,
      "grad_norm": 0.2254510372877121,
      "learning_rate": 0.0001,
      "loss": 1.2404,
      "step": 7549
    },
    {
      "epoch": 0.1551002896063499,
      "grad_norm": 0.23150110244750977,
      "learning_rate": 0.0001,
      "loss": 1.2281,
      "step": 7550
    },
    {
      "epoch": 0.15520755121741928,
      "grad_norm": 0.23232468962669373,
      "learning_rate": 0.0001,
      "loss": 1.2097,
      "step": 7551
    },
    {
      "epoch": 0.1553148128284887,
      "grad_norm": 0.21835677325725555,
      "learning_rate": 0.0001,
      "loss": 1.2062,
      "step": 7552
    },
    {
      "epoch": 0.15542207443955808,
      "grad_norm": 0.2238403707742691,
      "learning_rate": 0.0001,
      "loss": 1.2246,
      "step": 7553
    },
    {
      "epoch": 0.1555293360506275,
      "grad_norm": 0.22711136937141418,
      "learning_rate": 0.0001,
      "loss": 1.2194,
      "step": 7554
    },
    {
      "epoch": 0.15563659766169688,
      "grad_norm": 0.23528936505317688,
      "learning_rate": 0.0001,
      "loss": 1.2999,
      "step": 7555
    },
    {
      "epoch": 0.15574385927276627,
      "grad_norm": 0.22930064797401428,
      "learning_rate": 0.0001,
      "loss": 1.2582,
      "step": 7556
    },
    {
      "epoch": 0.15585112088383568,
      "grad_norm": 0.2302795797586441,
      "learning_rate": 0.0001,
      "loss": 1.3157,
      "step": 7557
    },
    {
      "epoch": 0.15595838249490507,
      "grad_norm": 0.22991178929805756,
      "learning_rate": 0.0001,
      "loss": 1.2029,
      "step": 7558
    },
    {
      "epoch": 0.15606564410597448,
      "grad_norm": 0.21146750450134277,
      "learning_rate": 0.0001,
      "loss": 1.234,
      "step": 7559
    },
    {
      "epoch": 0.15617290571704387,
      "grad_norm": 0.22513684630393982,
      "learning_rate": 0.0001,
      "loss": 1.2652,
      "step": 7560
    },
    {
      "epoch": 0.15628016732811326,
      "grad_norm": 0.22004707157611847,
      "learning_rate": 0.0001,
      "loss": 1.2691,
      "step": 7561
    },
    {
      "epoch": 0.15638742893918267,
      "grad_norm": 0.2270342856645584,
      "learning_rate": 0.0001,
      "loss": 1.3053,
      "step": 7562
    },
    {
      "epoch": 0.15649469055025206,
      "grad_norm": 0.23656301200389862,
      "learning_rate": 0.0001,
      "loss": 1.2735,
      "step": 7563
    },
    {
      "epoch": 0.15660195216132147,
      "grad_norm": 0.2136715054512024,
      "learning_rate": 0.0001,
      "loss": 1.2307,
      "step": 7564
    },
    {
      "epoch": 0.15670921377239086,
      "grad_norm": 0.23571692407131195,
      "learning_rate": 0.0001,
      "loss": 1.3545,
      "step": 7565
    },
    {
      "epoch": 0.15681647538346025,
      "grad_norm": 0.2288350760936737,
      "learning_rate": 0.0001,
      "loss": 1.1632,
      "step": 7566
    },
    {
      "epoch": 0.15692373699452966,
      "grad_norm": 0.22191783785820007,
      "learning_rate": 0.0001,
      "loss": 1.2368,
      "step": 7567
    },
    {
      "epoch": 0.15703099860559905,
      "grad_norm": 0.23188930749893188,
      "learning_rate": 0.0001,
      "loss": 1.2409,
      "step": 7568
    },
    {
      "epoch": 0.15713826021666846,
      "grad_norm": 0.24142982065677643,
      "learning_rate": 0.0001,
      "loss": 1.1817,
      "step": 7569
    },
    {
      "epoch": 0.15724552182773785,
      "grad_norm": 0.21717646718025208,
      "learning_rate": 0.0001,
      "loss": 1.263,
      "step": 7570
    },
    {
      "epoch": 0.15735278343880726,
      "grad_norm": 0.2233358919620514,
      "learning_rate": 0.0001,
      "loss": 1.1961,
      "step": 7571
    },
    {
      "epoch": 0.15746004504987665,
      "grad_norm": 0.21601708233356476,
      "learning_rate": 0.0001,
      "loss": 1.2506,
      "step": 7572
    },
    {
      "epoch": 0.15756730666094604,
      "grad_norm": 0.21407008171081543,
      "learning_rate": 0.0001,
      "loss": 1.1944,
      "step": 7573
    },
    {
      "epoch": 0.15767456827201545,
      "grad_norm": 0.21370936930179596,
      "learning_rate": 0.0001,
      "loss": 1.1011,
      "step": 7574
    },
    {
      "epoch": 0.15778182988308484,
      "grad_norm": 0.22503243386745453,
      "learning_rate": 0.0001,
      "loss": 1.277,
      "step": 7575
    },
    {
      "epoch": 0.15788909149415425,
      "grad_norm": 0.2163635641336441,
      "learning_rate": 0.0001,
      "loss": 1.2547,
      "step": 7576
    },
    {
      "epoch": 0.15799635310522364,
      "grad_norm": 0.21999523043632507,
      "learning_rate": 0.0001,
      "loss": 1.2756,
      "step": 7577
    },
    {
      "epoch": 0.15810361471629303,
      "grad_norm": 0.22796304523944855,
      "learning_rate": 0.0001,
      "loss": 1.3543,
      "step": 7578
    },
    {
      "epoch": 0.15821087632736244,
      "grad_norm": 0.23997057974338531,
      "learning_rate": 0.0001,
      "loss": 1.2772,
      "step": 7579
    },
    {
      "epoch": 0.15831813793843183,
      "grad_norm": 0.23606842756271362,
      "learning_rate": 0.0001,
      "loss": 1.2462,
      "step": 7580
    },
    {
      "epoch": 0.15842539954950124,
      "grad_norm": 0.21631214022636414,
      "learning_rate": 0.0001,
      "loss": 1.2715,
      "step": 7581
    },
    {
      "epoch": 0.15853266116057063,
      "grad_norm": 0.21625769138336182,
      "learning_rate": 0.0001,
      "loss": 1.2383,
      "step": 7582
    },
    {
      "epoch": 0.15863992277164002,
      "grad_norm": 0.2182200849056244,
      "learning_rate": 0.0001,
      "loss": 1.3155,
      "step": 7583
    },
    {
      "epoch": 0.15874718438270943,
      "grad_norm": 0.22784502804279327,
      "learning_rate": 0.0001,
      "loss": 1.3082,
      "step": 7584
    },
    {
      "epoch": 0.15885444599377882,
      "grad_norm": 0.21478214859962463,
      "learning_rate": 0.0001,
      "loss": 1.2107,
      "step": 7585
    },
    {
      "epoch": 0.15896170760484823,
      "grad_norm": 0.2231598198413849,
      "learning_rate": 0.0001,
      "loss": 1.2957,
      "step": 7586
    },
    {
      "epoch": 0.15906896921591762,
      "grad_norm": 0.22904092073440552,
      "learning_rate": 0.0001,
      "loss": 1.2281,
      "step": 7587
    },
    {
      "epoch": 0.159176230826987,
      "grad_norm": 0.2183847278356552,
      "learning_rate": 0.0001,
      "loss": 1.2058,
      "step": 7588
    },
    {
      "epoch": 0.15928349243805642,
      "grad_norm": 0.21577322483062744,
      "learning_rate": 0.0001,
      "loss": 1.2369,
      "step": 7589
    },
    {
      "epoch": 0.1593907540491258,
      "grad_norm": 0.217210590839386,
      "learning_rate": 0.0001,
      "loss": 1.2208,
      "step": 7590
    },
    {
      "epoch": 0.15949801566019522,
      "grad_norm": 0.22572651505470276,
      "learning_rate": 0.0001,
      "loss": 1.2951,
      "step": 7591
    },
    {
      "epoch": 0.1596052772712646,
      "grad_norm": 0.23640713095664978,
      "learning_rate": 0.0001,
      "loss": 1.2674,
      "step": 7592
    },
    {
      "epoch": 0.15971253888233403,
      "grad_norm": 0.23740969598293304,
      "learning_rate": 0.0001,
      "loss": 1.2562,
      "step": 7593
    },
    {
      "epoch": 0.1598198004934034,
      "grad_norm": 0.22014178335666656,
      "learning_rate": 0.0001,
      "loss": 1.3078,
      "step": 7594
    },
    {
      "epoch": 0.1599270621044728,
      "grad_norm": 0.2188398838043213,
      "learning_rate": 0.0001,
      "loss": 1.1937,
      "step": 7595
    },
    {
      "epoch": 0.16003432371554221,
      "grad_norm": 0.22050577402114868,
      "learning_rate": 0.0001,
      "loss": 1.2682,
      "step": 7596
    },
    {
      "epoch": 0.1601415853266116,
      "grad_norm": 0.22540873289108276,
      "learning_rate": 0.0001,
      "loss": 1.3162,
      "step": 7597
    },
    {
      "epoch": 0.16024884693768102,
      "grad_norm": 0.2435135394334793,
      "learning_rate": 0.0001,
      "loss": 1.3578,
      "step": 7598
    },
    {
      "epoch": 0.1603561085487504,
      "grad_norm": 0.21223662793636322,
      "learning_rate": 0.0001,
      "loss": 1.2461,
      "step": 7599
    },
    {
      "epoch": 0.1604633701598198,
      "grad_norm": 0.2085997462272644,
      "learning_rate": 0.0001,
      "loss": 1.2336,
      "step": 7600
    },
    {
      "epoch": 0.1605706317708892,
      "grad_norm": 0.22205691039562225,
      "learning_rate": 0.0001,
      "loss": 1.2466,
      "step": 7601
    },
    {
      "epoch": 0.1606778933819586,
      "grad_norm": 0.22269819676876068,
      "learning_rate": 0.0001,
      "loss": 1.2511,
      "step": 7602
    },
    {
      "epoch": 0.160785154993028,
      "grad_norm": 0.22354461252689362,
      "learning_rate": 0.0001,
      "loss": 1.2462,
      "step": 7603
    },
    {
      "epoch": 0.1608924166040974,
      "grad_norm": 0.22216928005218506,
      "learning_rate": 0.0001,
      "loss": 1.2789,
      "step": 7604
    },
    {
      "epoch": 0.16099967821516678,
      "grad_norm": 0.2229452282190323,
      "learning_rate": 0.0001,
      "loss": 1.2322,
      "step": 7605
    },
    {
      "epoch": 0.1611069398262362,
      "grad_norm": 0.23779748380184174,
      "learning_rate": 0.0001,
      "loss": 1.3451,
      "step": 7606
    },
    {
      "epoch": 0.16121420143730558,
      "grad_norm": 0.2268693745136261,
      "learning_rate": 0.0001,
      "loss": 1.1767,
      "step": 7607
    },
    {
      "epoch": 0.161321463048375,
      "grad_norm": 0.21108025312423706,
      "learning_rate": 0.0001,
      "loss": 1.2267,
      "step": 7608
    },
    {
      "epoch": 0.16142872465944438,
      "grad_norm": 0.22819629311561584,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 7609
    },
    {
      "epoch": 0.16153598627051377,
      "grad_norm": 0.2135351449251175,
      "learning_rate": 0.0001,
      "loss": 1.227,
      "step": 7610
    },
    {
      "epoch": 0.16164324788158319,
      "grad_norm": 0.23527421057224274,
      "learning_rate": 0.0001,
      "loss": 1.2841,
      "step": 7611
    },
    {
      "epoch": 0.16175050949265257,
      "grad_norm": 0.20632819831371307,
      "learning_rate": 0.0001,
      "loss": 1.1199,
      "step": 7612
    },
    {
      "epoch": 0.161857771103722,
      "grad_norm": 0.268951416015625,
      "learning_rate": 0.0001,
      "loss": 1.2928,
      "step": 7613
    },
    {
      "epoch": 0.16196503271479137,
      "grad_norm": 0.22156602144241333,
      "learning_rate": 0.0001,
      "loss": 1.2351,
      "step": 7614
    },
    {
      "epoch": 0.16207229432586076,
      "grad_norm": 0.22396184504032135,
      "learning_rate": 0.0001,
      "loss": 1.2861,
      "step": 7615
    },
    {
      "epoch": 0.16217955593693018,
      "grad_norm": 0.22461967170238495,
      "learning_rate": 0.0001,
      "loss": 1.324,
      "step": 7616
    },
    {
      "epoch": 0.16228681754799956,
      "grad_norm": 0.22867797315120697,
      "learning_rate": 0.0001,
      "loss": 1.3593,
      "step": 7617
    },
    {
      "epoch": 0.16239407915906898,
      "grad_norm": 0.21487951278686523,
      "learning_rate": 0.0001,
      "loss": 1.3004,
      "step": 7618
    },
    {
      "epoch": 0.16250134077013836,
      "grad_norm": 0.2235325425863266,
      "learning_rate": 0.0001,
      "loss": 1.3068,
      "step": 7619
    },
    {
      "epoch": 0.16260860238120778,
      "grad_norm": 0.2257593423128128,
      "learning_rate": 0.0001,
      "loss": 1.2679,
      "step": 7620
    },
    {
      "epoch": 0.16271586399227717,
      "grad_norm": 0.22510571777820587,
      "learning_rate": 0.0001,
      "loss": 1.2294,
      "step": 7621
    },
    {
      "epoch": 0.16282312560334655,
      "grad_norm": 0.21474412083625793,
      "learning_rate": 0.0001,
      "loss": 1.216,
      "step": 7622
    },
    {
      "epoch": 0.16293038721441597,
      "grad_norm": 0.23142385482788086,
      "learning_rate": 0.0001,
      "loss": 1.3252,
      "step": 7623
    },
    {
      "epoch": 0.16303764882548535,
      "grad_norm": 0.23083634674549103,
      "learning_rate": 0.0001,
      "loss": 1.3211,
      "step": 7624
    },
    {
      "epoch": 0.16314491043655477,
      "grad_norm": 0.24309717118740082,
      "learning_rate": 0.0001,
      "loss": 1.2757,
      "step": 7625
    },
    {
      "epoch": 0.16325217204762416,
      "grad_norm": 0.23077541589736938,
      "learning_rate": 0.0001,
      "loss": 1.2755,
      "step": 7626
    },
    {
      "epoch": 0.16335943365869354,
      "grad_norm": 0.22269299626350403,
      "learning_rate": 0.0001,
      "loss": 1.259,
      "step": 7627
    },
    {
      "epoch": 0.16346669526976296,
      "grad_norm": 0.2307964265346527,
      "learning_rate": 0.0001,
      "loss": 1.2698,
      "step": 7628
    },
    {
      "epoch": 0.16357395688083234,
      "grad_norm": 0.23125478625297546,
      "learning_rate": 0.0001,
      "loss": 1.3075,
      "step": 7629
    },
    {
      "epoch": 0.16368121849190176,
      "grad_norm": 0.22310656309127808,
      "learning_rate": 0.0001,
      "loss": 1.307,
      "step": 7630
    },
    {
      "epoch": 0.16378848010297115,
      "grad_norm": 0.2270118147134781,
      "learning_rate": 0.0001,
      "loss": 1.2568,
      "step": 7631
    },
    {
      "epoch": 0.16389574171404053,
      "grad_norm": 0.2355596274137497,
      "learning_rate": 0.0001,
      "loss": 1.2568,
      "step": 7632
    },
    {
      "epoch": 0.16400300332510995,
      "grad_norm": 0.2215803861618042,
      "learning_rate": 0.0001,
      "loss": 1.2322,
      "step": 7633
    },
    {
      "epoch": 0.16411026493617933,
      "grad_norm": 0.22608523070812225,
      "learning_rate": 0.0001,
      "loss": 1.2429,
      "step": 7634
    },
    {
      "epoch": 0.16421752654724875,
      "grad_norm": 0.23490555584430695,
      "learning_rate": 0.0001,
      "loss": 1.261,
      "step": 7635
    },
    {
      "epoch": 0.16432478815831814,
      "grad_norm": 0.2253958284854889,
      "learning_rate": 0.0001,
      "loss": 1.2533,
      "step": 7636
    },
    {
      "epoch": 0.16443204976938752,
      "grad_norm": 0.22934164106845856,
      "learning_rate": 0.0001,
      "loss": 1.3624,
      "step": 7637
    },
    {
      "epoch": 0.16453931138045694,
      "grad_norm": 0.21796710789203644,
      "learning_rate": 0.0001,
      "loss": 1.2197,
      "step": 7638
    },
    {
      "epoch": 0.16464657299152632,
      "grad_norm": 0.2226063758134842,
      "learning_rate": 0.0001,
      "loss": 1.1736,
      "step": 7639
    },
    {
      "epoch": 0.16475383460259574,
      "grad_norm": 0.22796228528022766,
      "learning_rate": 0.0001,
      "loss": 1.1927,
      "step": 7640
    },
    {
      "epoch": 0.16486109621366513,
      "grad_norm": 0.23140260577201843,
      "learning_rate": 0.0001,
      "loss": 1.0769,
      "step": 7641
    },
    {
      "epoch": 0.16496835782473454,
      "grad_norm": 0.2259451150894165,
      "learning_rate": 0.0001,
      "loss": 1.1942,
      "step": 7642
    },
    {
      "epoch": 0.16507561943580393,
      "grad_norm": 0.21588292717933655,
      "learning_rate": 0.0001,
      "loss": 1.3021,
      "step": 7643
    },
    {
      "epoch": 0.16518288104687331,
      "grad_norm": 0.23645634949207306,
      "learning_rate": 0.0001,
      "loss": 1.2323,
      "step": 7644
    },
    {
      "epoch": 0.16529014265794273,
      "grad_norm": 0.21638019382953644,
      "learning_rate": 0.0001,
      "loss": 1.1813,
      "step": 7645
    },
    {
      "epoch": 0.16539740426901212,
      "grad_norm": 0.2332446277141571,
      "learning_rate": 0.0001,
      "loss": 1.2678,
      "step": 7646
    },
    {
      "epoch": 0.16550466588008153,
      "grad_norm": 0.22219178080558777,
      "learning_rate": 0.0001,
      "loss": 1.2035,
      "step": 7647
    },
    {
      "epoch": 0.16561192749115092,
      "grad_norm": 0.23351247608661652,
      "learning_rate": 0.0001,
      "loss": 1.2792,
      "step": 7648
    },
    {
      "epoch": 0.1657191891022203,
      "grad_norm": 0.24037957191467285,
      "learning_rate": 0.0001,
      "loss": 1.1987,
      "step": 7649
    },
    {
      "epoch": 0.16582645071328972,
      "grad_norm": 0.2260466367006302,
      "learning_rate": 0.0001,
      "loss": 1.3283,
      "step": 7650
    },
    {
      "epoch": 0.1659337123243591,
      "grad_norm": 0.2302500456571579,
      "learning_rate": 0.0001,
      "loss": 1.1814,
      "step": 7651
    },
    {
      "epoch": 0.16604097393542852,
      "grad_norm": 0.2128310352563858,
      "learning_rate": 0.0001,
      "loss": 1.1371,
      "step": 7652
    },
    {
      "epoch": 0.1661482355464979,
      "grad_norm": 0.2274576723575592,
      "learning_rate": 0.0001,
      "loss": 1.1763,
      "step": 7653
    },
    {
      "epoch": 0.1662554971575673,
      "grad_norm": 0.22134342789649963,
      "learning_rate": 0.0001,
      "loss": 1.2367,
      "step": 7654
    },
    {
      "epoch": 0.1663627587686367,
      "grad_norm": 0.23355591297149658,
      "learning_rate": 0.0001,
      "loss": 1.2006,
      "step": 7655
    },
    {
      "epoch": 0.1664700203797061,
      "grad_norm": 0.2286710888147354,
      "learning_rate": 0.0001,
      "loss": 1.3151,
      "step": 7656
    },
    {
      "epoch": 0.1665772819907755,
      "grad_norm": 0.23423704504966736,
      "learning_rate": 0.0001,
      "loss": 1.2804,
      "step": 7657
    },
    {
      "epoch": 0.1666845436018449,
      "grad_norm": 0.22804737091064453,
      "learning_rate": 0.0001,
      "loss": 1.3028,
      "step": 7658
    },
    {
      "epoch": 0.16679180521291428,
      "grad_norm": 0.22209496796131134,
      "learning_rate": 0.0001,
      "loss": 1.2425,
      "step": 7659
    },
    {
      "epoch": 0.1668990668239837,
      "grad_norm": 0.2206544280052185,
      "learning_rate": 0.0001,
      "loss": 1.265,
      "step": 7660
    },
    {
      "epoch": 0.1670063284350531,
      "grad_norm": 0.22032949328422546,
      "learning_rate": 0.0001,
      "loss": 1.3379,
      "step": 7661
    },
    {
      "epoch": 0.1671135900461225,
      "grad_norm": 0.24015960097312927,
      "learning_rate": 0.0001,
      "loss": 1.2101,
      "step": 7662
    },
    {
      "epoch": 0.1672208516571919,
      "grad_norm": 0.23016409575939178,
      "learning_rate": 0.0001,
      "loss": 1.2134,
      "step": 7663
    },
    {
      "epoch": 0.1673281132682613,
      "grad_norm": 0.22666355967521667,
      "learning_rate": 0.0001,
      "loss": 1.2038,
      "step": 7664
    },
    {
      "epoch": 0.1674353748793307,
      "grad_norm": 0.21397516131401062,
      "learning_rate": 0.0001,
      "loss": 1.261,
      "step": 7665
    },
    {
      "epoch": 0.16754263649040008,
      "grad_norm": 0.21654586493968964,
      "learning_rate": 0.0001,
      "loss": 1.2358,
      "step": 7666
    },
    {
      "epoch": 0.1676498981014695,
      "grad_norm": 0.23629462718963623,
      "learning_rate": 0.0001,
      "loss": 1.3145,
      "step": 7667
    },
    {
      "epoch": 0.16775715971253888,
      "grad_norm": 0.2320757806301117,
      "learning_rate": 0.0001,
      "loss": 1.1649,
      "step": 7668
    },
    {
      "epoch": 0.1678644213236083,
      "grad_norm": 0.22181932628154755,
      "learning_rate": 0.0001,
      "loss": 1.2599,
      "step": 7669
    },
    {
      "epoch": 0.16797168293467768,
      "grad_norm": 0.23449300229549408,
      "learning_rate": 0.0001,
      "loss": 1.2242,
      "step": 7670
    },
    {
      "epoch": 0.16807894454574707,
      "grad_norm": 0.23548480868339539,
      "learning_rate": 0.0001,
      "loss": 1.1014,
      "step": 7671
    },
    {
      "epoch": 0.16818620615681648,
      "grad_norm": 0.21983493864536285,
      "learning_rate": 0.0001,
      "loss": 1.228,
      "step": 7672
    },
    {
      "epoch": 0.16829346776788587,
      "grad_norm": 0.20820774137973785,
      "learning_rate": 0.0001,
      "loss": 1.207,
      "step": 7673
    },
    {
      "epoch": 0.16840072937895528,
      "grad_norm": 0.21062684059143066,
      "learning_rate": 0.0001,
      "loss": 1.2642,
      "step": 7674
    },
    {
      "epoch": 0.16850799099002467,
      "grad_norm": 0.23275230824947357,
      "learning_rate": 0.0001,
      "loss": 1.2717,
      "step": 7675
    },
    {
      "epoch": 0.16861525260109406,
      "grad_norm": 0.23943538963794708,
      "learning_rate": 0.0001,
      "loss": 1.2222,
      "step": 7676
    },
    {
      "epoch": 0.16872251421216347,
      "grad_norm": 0.2233639508485794,
      "learning_rate": 0.0001,
      "loss": 1.2964,
      "step": 7677
    },
    {
      "epoch": 0.16882977582323286,
      "grad_norm": 0.23022346198558807,
      "learning_rate": 0.0001,
      "loss": 1.3391,
      "step": 7678
    },
    {
      "epoch": 0.16893703743430227,
      "grad_norm": 0.21678200364112854,
      "learning_rate": 0.0001,
      "loss": 1.2465,
      "step": 7679
    },
    {
      "epoch": 0.16904429904537166,
      "grad_norm": 0.23331134021282196,
      "learning_rate": 0.0001,
      "loss": 1.2303,
      "step": 7680
    },
    {
      "epoch": 0.16915156065644105,
      "grad_norm": 0.22280165553092957,
      "learning_rate": 0.0001,
      "loss": 1.2799,
      "step": 7681
    },
    {
      "epoch": 0.16925882226751046,
      "grad_norm": 0.23506620526313782,
      "learning_rate": 0.0001,
      "loss": 1.2934,
      "step": 7682
    },
    {
      "epoch": 0.16936608387857985,
      "grad_norm": 0.23615339398384094,
      "learning_rate": 0.0001,
      "loss": 1.245,
      "step": 7683
    },
    {
      "epoch": 0.16947334548964926,
      "grad_norm": 0.22005102038383484,
      "learning_rate": 0.0001,
      "loss": 1.2011,
      "step": 7684
    },
    {
      "epoch": 0.16958060710071865,
      "grad_norm": 0.22503606975078583,
      "learning_rate": 0.0001,
      "loss": 1.2529,
      "step": 7685
    },
    {
      "epoch": 0.16968786871178806,
      "grad_norm": 0.21512462198734283,
      "learning_rate": 0.0001,
      "loss": 1.2334,
      "step": 7686
    },
    {
      "epoch": 0.16979513032285745,
      "grad_norm": 0.22306402027606964,
      "learning_rate": 0.0001,
      "loss": 1.3315,
      "step": 7687
    },
    {
      "epoch": 0.16990239193392684,
      "grad_norm": 0.22312194108963013,
      "learning_rate": 0.0001,
      "loss": 1.2205,
      "step": 7688
    },
    {
      "epoch": 0.17000965354499625,
      "grad_norm": 0.21831843256950378,
      "learning_rate": 0.0001,
      "loss": 1.2914,
      "step": 7689
    },
    {
      "epoch": 0.17011691515606564,
      "grad_norm": 0.22542566061019897,
      "learning_rate": 0.0001,
      "loss": 1.1864,
      "step": 7690
    },
    {
      "epoch": 0.17022417676713505,
      "grad_norm": 0.23327748477458954,
      "learning_rate": 0.0001,
      "loss": 1.17,
      "step": 7691
    },
    {
      "epoch": 0.17033143837820444,
      "grad_norm": 0.2534845471382141,
      "learning_rate": 0.0001,
      "loss": 1.1237,
      "step": 7692
    },
    {
      "epoch": 0.17043869998927383,
      "grad_norm": 0.23076552152633667,
      "learning_rate": 0.0001,
      "loss": 1.3028,
      "step": 7693
    },
    {
      "epoch": 0.17054596160034324,
      "grad_norm": 0.21895885467529297,
      "learning_rate": 0.0001,
      "loss": 1.2823,
      "step": 7694
    },
    {
      "epoch": 0.17065322321141263,
      "grad_norm": 0.23267461359500885,
      "learning_rate": 0.0001,
      "loss": 1.2824,
      "step": 7695
    },
    {
      "epoch": 0.17076048482248204,
      "grad_norm": 0.23205550014972687,
      "learning_rate": 0.0001,
      "loss": 1.3515,
      "step": 7696
    },
    {
      "epoch": 0.17086774643355143,
      "grad_norm": 0.22105486690998077,
      "learning_rate": 0.0001,
      "loss": 1.2871,
      "step": 7697
    },
    {
      "epoch": 0.17097500804462082,
      "grad_norm": 0.20923849940299988,
      "learning_rate": 0.0001,
      "loss": 1.211,
      "step": 7698
    },
    {
      "epoch": 0.17108226965569023,
      "grad_norm": 0.22797001898288727,
      "learning_rate": 0.0001,
      "loss": 1.2457,
      "step": 7699
    },
    {
      "epoch": 0.17118953126675962,
      "grad_norm": 0.2346571832895279,
      "learning_rate": 0.0001,
      "loss": 1.2869,
      "step": 7700
    },
    {
      "epoch": 0.17129679287782903,
      "grad_norm": 0.23907049000263214,
      "learning_rate": 0.0001,
      "loss": 1.257,
      "step": 7701
    },
    {
      "epoch": 0.17140405448889842,
      "grad_norm": 0.24090802669525146,
      "learning_rate": 0.0001,
      "loss": 1.2289,
      "step": 7702
    },
    {
      "epoch": 0.1715113160999678,
      "grad_norm": 0.2302204668521881,
      "learning_rate": 0.0001,
      "loss": 1.1494,
      "step": 7703
    },
    {
      "epoch": 0.17161857771103722,
      "grad_norm": 0.2332334816455841,
      "learning_rate": 0.0001,
      "loss": 1.3515,
      "step": 7704
    },
    {
      "epoch": 0.1717258393221066,
      "grad_norm": 0.21986088156700134,
      "learning_rate": 0.0001,
      "loss": 1.2595,
      "step": 7705
    },
    {
      "epoch": 0.17183310093317603,
      "grad_norm": 0.22719153761863708,
      "learning_rate": 0.0001,
      "loss": 1.2899,
      "step": 7706
    },
    {
      "epoch": 0.1719403625442454,
      "grad_norm": 0.2247094362974167,
      "learning_rate": 0.0001,
      "loss": 1.2536,
      "step": 7707
    },
    {
      "epoch": 0.17204762415531483,
      "grad_norm": 0.21425427496433258,
      "learning_rate": 0.0001,
      "loss": 1.1874,
      "step": 7708
    },
    {
      "epoch": 0.1721548857663842,
      "grad_norm": 0.22860893607139587,
      "learning_rate": 0.0001,
      "loss": 1.2907,
      "step": 7709
    },
    {
      "epoch": 0.1722621473774536,
      "grad_norm": 0.22564291954040527,
      "learning_rate": 0.0001,
      "loss": 1.301,
      "step": 7710
    },
    {
      "epoch": 0.17236940898852302,
      "grad_norm": 0.2266998142004013,
      "learning_rate": 0.0001,
      "loss": 1.1853,
      "step": 7711
    },
    {
      "epoch": 0.1724766705995924,
      "grad_norm": 0.2391148954629898,
      "learning_rate": 0.0001,
      "loss": 1.2376,
      "step": 7712
    },
    {
      "epoch": 0.17258393221066182,
      "grad_norm": 0.22069832682609558,
      "learning_rate": 0.0001,
      "loss": 1.2253,
      "step": 7713
    },
    {
      "epoch": 0.1726911938217312,
      "grad_norm": 0.22416137158870697,
      "learning_rate": 0.0001,
      "loss": 1.1661,
      "step": 7714
    },
    {
      "epoch": 0.1727984554328006,
      "grad_norm": 0.22339795529842377,
      "learning_rate": 0.0001,
      "loss": 1.2584,
      "step": 7715
    },
    {
      "epoch": 0.17290571704387,
      "grad_norm": 0.22878946363925934,
      "learning_rate": 0.0001,
      "loss": 1.2092,
      "step": 7716
    },
    {
      "epoch": 0.1730129786549394,
      "grad_norm": 0.22408783435821533,
      "learning_rate": 0.0001,
      "loss": 1.2997,
      "step": 7717
    },
    {
      "epoch": 0.1731202402660088,
      "grad_norm": 0.22029435634613037,
      "learning_rate": 0.0001,
      "loss": 1.3201,
      "step": 7718
    },
    {
      "epoch": 0.1732275018770782,
      "grad_norm": 0.24014510214328766,
      "learning_rate": 0.0001,
      "loss": 1.2295,
      "step": 7719
    },
    {
      "epoch": 0.17333476348814758,
      "grad_norm": 0.24565564095973969,
      "learning_rate": 0.0001,
      "loss": 1.2202,
      "step": 7720
    },
    {
      "epoch": 0.173442025099217,
      "grad_norm": 0.22635097801685333,
      "learning_rate": 0.0001,
      "loss": 1.2154,
      "step": 7721
    },
    {
      "epoch": 0.17354928671028638,
      "grad_norm": 0.223466157913208,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 7722
    },
    {
      "epoch": 0.1736565483213558,
      "grad_norm": 0.20541924238204956,
      "learning_rate": 0.0001,
      "loss": 1.1038,
      "step": 7723
    },
    {
      "epoch": 0.17376380993242518,
      "grad_norm": 0.22036656737327576,
      "learning_rate": 0.0001,
      "loss": 1.1358,
      "step": 7724
    },
    {
      "epoch": 0.17387107154349457,
      "grad_norm": 0.22111082077026367,
      "learning_rate": 0.0001,
      "loss": 1.2795,
      "step": 7725
    },
    {
      "epoch": 0.17397833315456399,
      "grad_norm": 0.2091655731201172,
      "learning_rate": 0.0001,
      "loss": 1.229,
      "step": 7726
    },
    {
      "epoch": 0.17408559476563337,
      "grad_norm": 0.24337686598300934,
      "learning_rate": 0.0001,
      "loss": 1.0932,
      "step": 7727
    },
    {
      "epoch": 0.1741928563767028,
      "grad_norm": 0.21958142518997192,
      "learning_rate": 0.0001,
      "loss": 1.2334,
      "step": 7728
    },
    {
      "epoch": 0.17430011798777217,
      "grad_norm": 0.23777784407138824,
      "learning_rate": 0.0001,
      "loss": 1.3873,
      "step": 7729
    },
    {
      "epoch": 0.17440737959884156,
      "grad_norm": 0.2083531767129898,
      "learning_rate": 0.0001,
      "loss": 1.1814,
      "step": 7730
    },
    {
      "epoch": 0.17451464120991098,
      "grad_norm": 0.21372458338737488,
      "learning_rate": 0.0001,
      "loss": 1.2323,
      "step": 7731
    },
    {
      "epoch": 0.17462190282098036,
      "grad_norm": 0.2215520143508911,
      "learning_rate": 0.0001,
      "loss": 1.2706,
      "step": 7732
    },
    {
      "epoch": 0.17472916443204978,
      "grad_norm": 0.2172871232032776,
      "learning_rate": 0.0001,
      "loss": 1.2722,
      "step": 7733
    },
    {
      "epoch": 0.17483642604311916,
      "grad_norm": 0.24800164997577667,
      "learning_rate": 0.0001,
      "loss": 1.3195,
      "step": 7734
    },
    {
      "epoch": 0.17494368765418858,
      "grad_norm": 0.22640378773212433,
      "learning_rate": 0.0001,
      "loss": 1.2459,
      "step": 7735
    },
    {
      "epoch": 0.17505094926525797,
      "grad_norm": 0.22893142700195312,
      "learning_rate": 0.0001,
      "loss": 1.2911,
      "step": 7736
    },
    {
      "epoch": 0.17515821087632735,
      "grad_norm": 0.22226963937282562,
      "learning_rate": 0.0001,
      "loss": 1.2785,
      "step": 7737
    },
    {
      "epoch": 0.17526547248739677,
      "grad_norm": 0.22539906203746796,
      "learning_rate": 0.0001,
      "loss": 1.2012,
      "step": 7738
    },
    {
      "epoch": 0.17537273409846615,
      "grad_norm": 0.23124364018440247,
      "learning_rate": 0.0001,
      "loss": 1.2848,
      "step": 7739
    },
    {
      "epoch": 0.17547999570953557,
      "grad_norm": 0.21758203208446503,
      "learning_rate": 0.0001,
      "loss": 1.2502,
      "step": 7740
    },
    {
      "epoch": 0.17558725732060496,
      "grad_norm": 0.21013686060905457,
      "learning_rate": 0.0001,
      "loss": 1.2193,
      "step": 7741
    },
    {
      "epoch": 0.17569451893167434,
      "grad_norm": 0.24068616330623627,
      "learning_rate": 0.0001,
      "loss": 1.3314,
      "step": 7742
    },
    {
      "epoch": 0.17580178054274376,
      "grad_norm": 0.22824403643608093,
      "learning_rate": 0.0001,
      "loss": 1.2985,
      "step": 7743
    },
    {
      "epoch": 0.17590904215381314,
      "grad_norm": 0.23372404277324677,
      "learning_rate": 0.0001,
      "loss": 1.2419,
      "step": 7744
    },
    {
      "epoch": 0.17601630376488256,
      "grad_norm": 0.2248249500989914,
      "learning_rate": 0.0001,
      "loss": 1.1571,
      "step": 7745
    },
    {
      "epoch": 0.17612356537595195,
      "grad_norm": 0.2235163152217865,
      "learning_rate": 0.0001,
      "loss": 1.2809,
      "step": 7746
    },
    {
      "epoch": 0.17623082698702133,
      "grad_norm": 0.22806639969348907,
      "learning_rate": 0.0001,
      "loss": 1.2041,
      "step": 7747
    },
    {
      "epoch": 0.17633808859809075,
      "grad_norm": 0.22235451638698578,
      "learning_rate": 0.0001,
      "loss": 1.2267,
      "step": 7748
    },
    {
      "epoch": 0.17644535020916013,
      "grad_norm": 0.21582143008708954,
      "learning_rate": 0.0001,
      "loss": 1.1919,
      "step": 7749
    },
    {
      "epoch": 0.17655261182022955,
      "grad_norm": 0.23466689884662628,
      "learning_rate": 0.0001,
      "loss": 1.2909,
      "step": 7750
    },
    {
      "epoch": 0.17665987343129894,
      "grad_norm": 0.21865814924240112,
      "learning_rate": 0.0001,
      "loss": 1.1622,
      "step": 7751
    },
    {
      "epoch": 0.17676713504236832,
      "grad_norm": 0.22443735599517822,
      "learning_rate": 0.0001,
      "loss": 1.2863,
      "step": 7752
    },
    {
      "epoch": 0.17687439665343774,
      "grad_norm": 0.24137379229068756,
      "learning_rate": 0.0001,
      "loss": 1.309,
      "step": 7753
    },
    {
      "epoch": 0.17698165826450712,
      "grad_norm": 0.22363045811653137,
      "learning_rate": 0.0001,
      "loss": 1.197,
      "step": 7754
    },
    {
      "epoch": 0.17708891987557654,
      "grad_norm": 0.22222965955734253,
      "learning_rate": 0.0001,
      "loss": 1.2566,
      "step": 7755
    },
    {
      "epoch": 0.17719618148664593,
      "grad_norm": 0.20905068516731262,
      "learning_rate": 0.0001,
      "loss": 1.1759,
      "step": 7756
    },
    {
      "epoch": 0.17730344309771534,
      "grad_norm": 0.21068793535232544,
      "learning_rate": 0.0001,
      "loss": 1.2392,
      "step": 7757
    },
    {
      "epoch": 0.17741070470878473,
      "grad_norm": 0.2385084182024002,
      "learning_rate": 0.0001,
      "loss": 1.275,
      "step": 7758
    },
    {
      "epoch": 0.17751796631985411,
      "grad_norm": 0.21023958921432495,
      "learning_rate": 0.0001,
      "loss": 1.2701,
      "step": 7759
    },
    {
      "epoch": 0.17762522793092353,
      "grad_norm": 0.2493717074394226,
      "learning_rate": 0.0001,
      "loss": 1.3224,
      "step": 7760
    },
    {
      "epoch": 0.17773248954199292,
      "grad_norm": 0.221265971660614,
      "learning_rate": 0.0001,
      "loss": 1.3369,
      "step": 7761
    },
    {
      "epoch": 0.17783975115306233,
      "grad_norm": 0.21403458714485168,
      "learning_rate": 0.0001,
      "loss": 1.2617,
      "step": 7762
    },
    {
      "epoch": 0.17794701276413172,
      "grad_norm": 0.21058616042137146,
      "learning_rate": 0.0001,
      "loss": 1.1923,
      "step": 7763
    },
    {
      "epoch": 0.1780542743752011,
      "grad_norm": 0.22373180091381073,
      "learning_rate": 0.0001,
      "loss": 1.1855,
      "step": 7764
    },
    {
      "epoch": 0.17816153598627052,
      "grad_norm": 0.2153128981590271,
      "learning_rate": 0.0001,
      "loss": 1.3764,
      "step": 7765
    },
    {
      "epoch": 0.1782687975973399,
      "grad_norm": 0.2114923894405365,
      "learning_rate": 0.0001,
      "loss": 1.2131,
      "step": 7766
    },
    {
      "epoch": 0.17837605920840932,
      "grad_norm": 0.219182088971138,
      "learning_rate": 0.0001,
      "loss": 1.2743,
      "step": 7767
    },
    {
      "epoch": 0.1784833208194787,
      "grad_norm": 0.23258303105831146,
      "learning_rate": 0.0001,
      "loss": 1.1736,
      "step": 7768
    },
    {
      "epoch": 0.1785905824305481,
      "grad_norm": 0.2239978015422821,
      "learning_rate": 0.0001,
      "loss": 1.2275,
      "step": 7769
    },
    {
      "epoch": 0.1786978440416175,
      "grad_norm": 0.21706900000572205,
      "learning_rate": 0.0001,
      "loss": 1.2932,
      "step": 7770
    },
    {
      "epoch": 0.1788051056526869,
      "grad_norm": 0.2266720086336136,
      "learning_rate": 0.0001,
      "loss": 1.2721,
      "step": 7771
    },
    {
      "epoch": 0.1789123672637563,
      "grad_norm": 0.21314726769924164,
      "learning_rate": 0.0001,
      "loss": 1.2885,
      "step": 7772
    },
    {
      "epoch": 0.1790196288748257,
      "grad_norm": 0.22989605367183685,
      "learning_rate": 0.0001,
      "loss": 1.2812,
      "step": 7773
    },
    {
      "epoch": 0.17912689048589509,
      "grad_norm": 0.22425755858421326,
      "learning_rate": 0.0001,
      "loss": 1.3092,
      "step": 7774
    },
    {
      "epoch": 0.1792341520969645,
      "grad_norm": 0.22427092492580414,
      "learning_rate": 0.0001,
      "loss": 1.2543,
      "step": 7775
    },
    {
      "epoch": 0.1793414137080339,
      "grad_norm": 0.22082050144672394,
      "learning_rate": 0.0001,
      "loss": 1.2083,
      "step": 7776
    },
    {
      "epoch": 0.1794486753191033,
      "grad_norm": 0.22838197648525238,
      "learning_rate": 0.0001,
      "loss": 1.1866,
      "step": 7777
    },
    {
      "epoch": 0.1795559369301727,
      "grad_norm": 0.2330292910337448,
      "learning_rate": 0.0001,
      "loss": 1.1821,
      "step": 7778
    },
    {
      "epoch": 0.1796631985412421,
      "grad_norm": 0.22854836285114288,
      "learning_rate": 0.0001,
      "loss": 1.3534,
      "step": 7779
    },
    {
      "epoch": 0.1797704601523115,
      "grad_norm": 0.2117679864168167,
      "learning_rate": 0.0001,
      "loss": 1.2083,
      "step": 7780
    },
    {
      "epoch": 0.17987772176338088,
      "grad_norm": 0.222786545753479,
      "learning_rate": 0.0001,
      "loss": 1.275,
      "step": 7781
    },
    {
      "epoch": 0.1799849833744503,
      "grad_norm": 0.21514858305454254,
      "learning_rate": 0.0001,
      "loss": 1.304,
      "step": 7782
    },
    {
      "epoch": 0.18009224498551968,
      "grad_norm": 0.22367464005947113,
      "learning_rate": 0.0001,
      "loss": 1.2262,
      "step": 7783
    },
    {
      "epoch": 0.1801995065965891,
      "grad_norm": 0.22476045787334442,
      "learning_rate": 0.0001,
      "loss": 1.377,
      "step": 7784
    },
    {
      "epoch": 0.18030676820765848,
      "grad_norm": 0.21330448985099792,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 7785
    },
    {
      "epoch": 0.18041402981872787,
      "grad_norm": 0.22808687388896942,
      "learning_rate": 0.0001,
      "loss": 1.1198,
      "step": 7786
    },
    {
      "epoch": 0.18052129142979728,
      "grad_norm": 0.2189113050699234,
      "learning_rate": 0.0001,
      "loss": 1.2868,
      "step": 7787
    },
    {
      "epoch": 0.18062855304086667,
      "grad_norm": 0.22700880467891693,
      "learning_rate": 0.0001,
      "loss": 1.3088,
      "step": 7788
    },
    {
      "epoch": 0.18073581465193608,
      "grad_norm": 0.22711683809757233,
      "learning_rate": 0.0001,
      "loss": 1.2225,
      "step": 7789
    },
    {
      "epoch": 0.18084307626300547,
      "grad_norm": 0.23217712342739105,
      "learning_rate": 0.0001,
      "loss": 1.2774,
      "step": 7790
    },
    {
      "epoch": 0.18095033787407486,
      "grad_norm": 0.22163331508636475,
      "learning_rate": 0.0001,
      "loss": 1.2239,
      "step": 7791
    },
    {
      "epoch": 0.18105759948514427,
      "grad_norm": 0.22029796242713928,
      "learning_rate": 0.0001,
      "loss": 1.192,
      "step": 7792
    },
    {
      "epoch": 0.18116486109621366,
      "grad_norm": 0.22068904340267181,
      "learning_rate": 0.0001,
      "loss": 1.2844,
      "step": 7793
    },
    {
      "epoch": 0.18127212270728307,
      "grad_norm": 0.235752671957016,
      "learning_rate": 0.0001,
      "loss": 1.1593,
      "step": 7794
    },
    {
      "epoch": 0.18137938431835246,
      "grad_norm": 0.22794514894485474,
      "learning_rate": 0.0001,
      "loss": 1.1186,
      "step": 7795
    },
    {
      "epoch": 0.18148664592942185,
      "grad_norm": 0.22049319744110107,
      "learning_rate": 0.0001,
      "loss": 1.249,
      "step": 7796
    },
    {
      "epoch": 0.18159390754049126,
      "grad_norm": 0.2238261103630066,
      "learning_rate": 0.0001,
      "loss": 1.2716,
      "step": 7797
    },
    {
      "epoch": 0.18170116915156065,
      "grad_norm": 0.2248256355524063,
      "learning_rate": 0.0001,
      "loss": 1.1695,
      "step": 7798
    },
    {
      "epoch": 0.18180843076263006,
      "grad_norm": 0.2165648341178894,
      "learning_rate": 0.0001,
      "loss": 1.2921,
      "step": 7799
    },
    {
      "epoch": 0.18191569237369945,
      "grad_norm": 0.2247770130634308,
      "learning_rate": 0.0001,
      "loss": 1.2649,
      "step": 7800
    },
    {
      "epoch": 0.18202295398476886,
      "grad_norm": 0.23297490179538727,
      "learning_rate": 0.0001,
      "loss": 1.3191,
      "step": 7801
    },
    {
      "epoch": 0.18213021559583825,
      "grad_norm": 0.22116808593273163,
      "learning_rate": 0.0001,
      "loss": 1.3244,
      "step": 7802
    },
    {
      "epoch": 0.18223747720690764,
      "grad_norm": 0.23136909306049347,
      "learning_rate": 0.0001,
      "loss": 1.2807,
      "step": 7803
    },
    {
      "epoch": 0.18234473881797705,
      "grad_norm": 0.22406883537769318,
      "learning_rate": 0.0001,
      "loss": 1.1231,
      "step": 7804
    },
    {
      "epoch": 0.18245200042904644,
      "grad_norm": 0.2450983077287674,
      "learning_rate": 0.0001,
      "loss": 1.215,
      "step": 7805
    },
    {
      "epoch": 0.18255926204011586,
      "grad_norm": 0.2517510652542114,
      "learning_rate": 0.0001,
      "loss": 1.0944,
      "step": 7806
    },
    {
      "epoch": 0.18266652365118524,
      "grad_norm": 0.2208961844444275,
      "learning_rate": 0.0001,
      "loss": 1.2456,
      "step": 7807
    },
    {
      "epoch": 0.18277378526225463,
      "grad_norm": 0.21313634514808655,
      "learning_rate": 0.0001,
      "loss": 1.204,
      "step": 7808
    },
    {
      "epoch": 0.18288104687332404,
      "grad_norm": 0.2135390043258667,
      "learning_rate": 0.0001,
      "loss": 1.1779,
      "step": 7809
    },
    {
      "epoch": 0.18298830848439343,
      "grad_norm": 0.221125066280365,
      "learning_rate": 0.0001,
      "loss": 1.2537,
      "step": 7810
    },
    {
      "epoch": 0.18309557009546285,
      "grad_norm": 0.2124611735343933,
      "learning_rate": 0.0001,
      "loss": 1.24,
      "step": 7811
    },
    {
      "epoch": 0.18320283170653223,
      "grad_norm": 0.22528629004955292,
      "learning_rate": 0.0001,
      "loss": 1.2544,
      "step": 7812
    },
    {
      "epoch": 0.18331009331760162,
      "grad_norm": 0.2339491993188858,
      "learning_rate": 0.0001,
      "loss": 1.2411,
      "step": 7813
    },
    {
      "epoch": 0.18341735492867103,
      "grad_norm": 0.22834119200706482,
      "learning_rate": 0.0001,
      "loss": 1.2084,
      "step": 7814
    },
    {
      "epoch": 0.18352461653974042,
      "grad_norm": 0.2105279117822647,
      "learning_rate": 0.0001,
      "loss": 1.0501,
      "step": 7815
    },
    {
      "epoch": 0.18363187815080984,
      "grad_norm": 0.24834668636322021,
      "learning_rate": 0.0001,
      "loss": 1.2413,
      "step": 7816
    },
    {
      "epoch": 0.18373913976187922,
      "grad_norm": 0.2488594502210617,
      "learning_rate": 0.0001,
      "loss": 1.3305,
      "step": 7817
    },
    {
      "epoch": 0.1838464013729486,
      "grad_norm": 0.24925491213798523,
      "learning_rate": 0.0001,
      "loss": 1.2672,
      "step": 7818
    },
    {
      "epoch": 0.18395366298401802,
      "grad_norm": 0.23584920167922974,
      "learning_rate": 0.0001,
      "loss": 1.3123,
      "step": 7819
    },
    {
      "epoch": 0.1840609245950874,
      "grad_norm": 0.2298722118139267,
      "learning_rate": 0.0001,
      "loss": 1.3565,
      "step": 7820
    },
    {
      "epoch": 0.18416818620615683,
      "grad_norm": 0.2185160219669342,
      "learning_rate": 0.0001,
      "loss": 1.2195,
      "step": 7821
    },
    {
      "epoch": 0.1842754478172262,
      "grad_norm": 0.2173519879579544,
      "learning_rate": 0.0001,
      "loss": 1.2623,
      "step": 7822
    },
    {
      "epoch": 0.1843827094282956,
      "grad_norm": 0.2311968356370926,
      "learning_rate": 0.0001,
      "loss": 1.2107,
      "step": 7823
    },
    {
      "epoch": 0.184489971039365,
      "grad_norm": 0.22122426331043243,
      "learning_rate": 0.0001,
      "loss": 1.305,
      "step": 7824
    },
    {
      "epoch": 0.1845972326504344,
      "grad_norm": 0.23766732215881348,
      "learning_rate": 0.0001,
      "loss": 1.3082,
      "step": 7825
    },
    {
      "epoch": 0.18470449426150382,
      "grad_norm": 0.22537390887737274,
      "learning_rate": 0.0001,
      "loss": 1.2751,
      "step": 7826
    },
    {
      "epoch": 0.1848117558725732,
      "grad_norm": 0.2256898730993271,
      "learning_rate": 0.0001,
      "loss": 1.239,
      "step": 7827
    },
    {
      "epoch": 0.18491901748364262,
      "grad_norm": 0.21987487375736237,
      "learning_rate": 0.0001,
      "loss": 1.1657,
      "step": 7828
    },
    {
      "epoch": 0.185026279094712,
      "grad_norm": 0.2293771356344223,
      "learning_rate": 0.0001,
      "loss": 1.2193,
      "step": 7829
    },
    {
      "epoch": 0.1851335407057814,
      "grad_norm": 0.2414952516555786,
      "learning_rate": 0.0001,
      "loss": 1.2682,
      "step": 7830
    },
    {
      "epoch": 0.1852408023168508,
      "grad_norm": 0.24449177086353302,
      "learning_rate": 0.0001,
      "loss": 1.1996,
      "step": 7831
    },
    {
      "epoch": 0.1853480639279202,
      "grad_norm": 0.22866655886173248,
      "learning_rate": 0.0001,
      "loss": 1.2086,
      "step": 7832
    },
    {
      "epoch": 0.1854553255389896,
      "grad_norm": 0.22292166948318481,
      "learning_rate": 0.0001,
      "loss": 1.2798,
      "step": 7833
    },
    {
      "epoch": 0.185562587150059,
      "grad_norm": 0.23589986562728882,
      "learning_rate": 0.0001,
      "loss": 1.368,
      "step": 7834
    },
    {
      "epoch": 0.18566984876112838,
      "grad_norm": 0.23396727442741394,
      "learning_rate": 0.0001,
      "loss": 1.3053,
      "step": 7835
    },
    {
      "epoch": 0.1857771103721978,
      "grad_norm": 0.23094181716442108,
      "learning_rate": 0.0001,
      "loss": 1.2376,
      "step": 7836
    },
    {
      "epoch": 0.18588437198326718,
      "grad_norm": 0.2157677710056305,
      "learning_rate": 0.0001,
      "loss": 1.2019,
      "step": 7837
    },
    {
      "epoch": 0.1859916335943366,
      "grad_norm": 0.2235535979270935,
      "learning_rate": 0.0001,
      "loss": 1.2573,
      "step": 7838
    },
    {
      "epoch": 0.18609889520540598,
      "grad_norm": 0.21997720003128052,
      "learning_rate": 0.0001,
      "loss": 1.233,
      "step": 7839
    },
    {
      "epoch": 0.18620615681647537,
      "grad_norm": 0.2271537482738495,
      "learning_rate": 0.0001,
      "loss": 1.272,
      "step": 7840
    },
    {
      "epoch": 0.18631341842754479,
      "grad_norm": 0.22450898587703705,
      "learning_rate": 0.0001,
      "loss": 1.2526,
      "step": 7841
    },
    {
      "epoch": 0.18642068003861417,
      "grad_norm": 0.22719940543174744,
      "learning_rate": 0.0001,
      "loss": 1.1734,
      "step": 7842
    },
    {
      "epoch": 0.1865279416496836,
      "grad_norm": 0.24073481559753418,
      "learning_rate": 0.0001,
      "loss": 1.2441,
      "step": 7843
    },
    {
      "epoch": 0.18663520326075297,
      "grad_norm": 0.2198556363582611,
      "learning_rate": 0.0001,
      "loss": 1.3283,
      "step": 7844
    },
    {
      "epoch": 0.18674246487182236,
      "grad_norm": 0.2219853401184082,
      "learning_rate": 0.0001,
      "loss": 1.2757,
      "step": 7845
    },
    {
      "epoch": 0.18684972648289178,
      "grad_norm": 0.2262503057718277,
      "learning_rate": 0.0001,
      "loss": 1.2575,
      "step": 7846
    },
    {
      "epoch": 0.18695698809396116,
      "grad_norm": 0.22552503645420074,
      "learning_rate": 0.0001,
      "loss": 1.3351,
      "step": 7847
    },
    {
      "epoch": 0.18706424970503058,
      "grad_norm": 0.230197012424469,
      "learning_rate": 0.0001,
      "loss": 1.2389,
      "step": 7848
    },
    {
      "epoch": 0.18717151131609996,
      "grad_norm": 0.23879577219486237,
      "learning_rate": 0.0001,
      "loss": 1.2957,
      "step": 7849
    },
    {
      "epoch": 0.18727877292716938,
      "grad_norm": 0.2142232358455658,
      "learning_rate": 0.0001,
      "loss": 1.1904,
      "step": 7850
    },
    {
      "epoch": 0.18738603453823877,
      "grad_norm": 0.21534457802772522,
      "learning_rate": 0.0001,
      "loss": 1.2225,
      "step": 7851
    },
    {
      "epoch": 0.18749329614930815,
      "grad_norm": 0.22531671822071075,
      "learning_rate": 0.0001,
      "loss": 1.2873,
      "step": 7852
    },
    {
      "epoch": 0.18760055776037757,
      "grad_norm": 0.23636344075202942,
      "learning_rate": 0.0001,
      "loss": 1.255,
      "step": 7853
    },
    {
      "epoch": 0.18770781937144695,
      "grad_norm": 0.22364389896392822,
      "learning_rate": 0.0001,
      "loss": 1.2329,
      "step": 7854
    },
    {
      "epoch": 0.18781508098251637,
      "grad_norm": 0.2443992644548416,
      "learning_rate": 0.0001,
      "loss": 1.2612,
      "step": 7855
    },
    {
      "epoch": 0.18792234259358576,
      "grad_norm": 0.23311904072761536,
      "learning_rate": 0.0001,
      "loss": 1.2593,
      "step": 7856
    },
    {
      "epoch": 0.18802960420465514,
      "grad_norm": 0.24158258736133575,
      "learning_rate": 0.0001,
      "loss": 1.252,
      "step": 7857
    },
    {
      "epoch": 0.18813686581572456,
      "grad_norm": 0.2131132334470749,
      "learning_rate": 0.0001,
      "loss": 1.2661,
      "step": 7858
    },
    {
      "epoch": 0.18824412742679394,
      "grad_norm": 0.23356202244758606,
      "learning_rate": 0.0001,
      "loss": 1.3338,
      "step": 7859
    },
    {
      "epoch": 0.18835138903786336,
      "grad_norm": 0.2154780775308609,
      "learning_rate": 0.0001,
      "loss": 1.2628,
      "step": 7860
    },
    {
      "epoch": 0.18845865064893275,
      "grad_norm": 0.22282060980796814,
      "learning_rate": 0.0001,
      "loss": 1.2278,
      "step": 7861
    },
    {
      "epoch": 0.18856591226000213,
      "grad_norm": 0.2295328974723816,
      "learning_rate": 0.0001,
      "loss": 1.1976,
      "step": 7862
    },
    {
      "epoch": 0.18867317387107155,
      "grad_norm": 0.22299306094646454,
      "learning_rate": 0.0001,
      "loss": 1.2574,
      "step": 7863
    },
    {
      "epoch": 0.18878043548214093,
      "grad_norm": 0.2253110259771347,
      "learning_rate": 0.0001,
      "loss": 1.2233,
      "step": 7864
    },
    {
      "epoch": 0.18888769709321035,
      "grad_norm": 0.22706934809684753,
      "learning_rate": 0.0001,
      "loss": 1.2584,
      "step": 7865
    },
    {
      "epoch": 0.18899495870427974,
      "grad_norm": 0.206412211060524,
      "learning_rate": 0.0001,
      "loss": 1.253,
      "step": 7866
    },
    {
      "epoch": 0.18910222031534912,
      "grad_norm": 0.21758870780467987,
      "learning_rate": 0.0001,
      "loss": 1.2537,
      "step": 7867
    },
    {
      "epoch": 0.18920948192641854,
      "grad_norm": 0.2209172546863556,
      "learning_rate": 0.0001,
      "loss": 1.2053,
      "step": 7868
    },
    {
      "epoch": 0.18931674353748792,
      "grad_norm": 0.21989937126636505,
      "learning_rate": 0.0001,
      "loss": 1.3203,
      "step": 7869
    },
    {
      "epoch": 0.18942400514855734,
      "grad_norm": 0.2244998812675476,
      "learning_rate": 0.0001,
      "loss": 1.2149,
      "step": 7870
    },
    {
      "epoch": 0.18953126675962673,
      "grad_norm": 0.2264120876789093,
      "learning_rate": 0.0001,
      "loss": 1.1125,
      "step": 7871
    },
    {
      "epoch": 0.18963852837069614,
      "grad_norm": 0.23782432079315186,
      "learning_rate": 0.0001,
      "loss": 1.339,
      "step": 7872
    },
    {
      "epoch": 0.18974578998176553,
      "grad_norm": 0.232926145195961,
      "learning_rate": 0.0001,
      "loss": 1.3511,
      "step": 7873
    },
    {
      "epoch": 0.18985305159283491,
      "grad_norm": 0.23271745443344116,
      "learning_rate": 0.0001,
      "loss": 1.3563,
      "step": 7874
    },
    {
      "epoch": 0.18996031320390433,
      "grad_norm": 0.21500054001808167,
      "learning_rate": 0.0001,
      "loss": 1.3103,
      "step": 7875
    },
    {
      "epoch": 0.19006757481497372,
      "grad_norm": 0.22452636063098907,
      "learning_rate": 0.0001,
      "loss": 1.2244,
      "step": 7876
    },
    {
      "epoch": 0.19017483642604313,
      "grad_norm": 0.2152405083179474,
      "learning_rate": 0.0001,
      "loss": 1.2301,
      "step": 7877
    },
    {
      "epoch": 0.19028209803711252,
      "grad_norm": 0.2276127189397812,
      "learning_rate": 0.0001,
      "loss": 1.1911,
      "step": 7878
    },
    {
      "epoch": 0.1903893596481819,
      "grad_norm": 0.21858616173267365,
      "learning_rate": 0.0001,
      "loss": 1.3006,
      "step": 7879
    },
    {
      "epoch": 0.19049662125925132,
      "grad_norm": 0.25789034366607666,
      "learning_rate": 0.0001,
      "loss": 1.2184,
      "step": 7880
    },
    {
      "epoch": 0.1906038828703207,
      "grad_norm": 0.23831778764724731,
      "learning_rate": 0.0001,
      "loss": 1.2226,
      "step": 7881
    },
    {
      "epoch": 0.19071114448139012,
      "grad_norm": 0.23043355345726013,
      "learning_rate": 0.0001,
      "loss": 1.2891,
      "step": 7882
    },
    {
      "epoch": 0.1908184060924595,
      "grad_norm": 0.23118950426578522,
      "learning_rate": 0.0001,
      "loss": 1.2346,
      "step": 7883
    },
    {
      "epoch": 0.1909256677035289,
      "grad_norm": 0.22668902575969696,
      "learning_rate": 0.0001,
      "loss": 1.2932,
      "step": 7884
    },
    {
      "epoch": 0.1910329293145983,
      "grad_norm": 0.23808178305625916,
      "learning_rate": 0.0001,
      "loss": 1.2973,
      "step": 7885
    },
    {
      "epoch": 0.1911401909256677,
      "grad_norm": 0.25457432866096497,
      "learning_rate": 0.0001,
      "loss": 1.2292,
      "step": 7886
    },
    {
      "epoch": 0.1912474525367371,
      "grad_norm": 0.23315021395683289,
      "learning_rate": 0.0001,
      "loss": 1.3125,
      "step": 7887
    },
    {
      "epoch": 0.1913547141478065,
      "grad_norm": 0.21752473711967468,
      "learning_rate": 0.0001,
      "loss": 1.2169,
      "step": 7888
    },
    {
      "epoch": 0.19146197575887589,
      "grad_norm": 0.23210564255714417,
      "learning_rate": 0.0001,
      "loss": 1.2764,
      "step": 7889
    },
    {
      "epoch": 0.1915692373699453,
      "grad_norm": 0.22222977876663208,
      "learning_rate": 0.0001,
      "loss": 1.2948,
      "step": 7890
    },
    {
      "epoch": 0.1916764989810147,
      "grad_norm": 0.21245259046554565,
      "learning_rate": 0.0001,
      "loss": 1.2767,
      "step": 7891
    },
    {
      "epoch": 0.1917837605920841,
      "grad_norm": 0.2351693958044052,
      "learning_rate": 0.0001,
      "loss": 1.3173,
      "step": 7892
    },
    {
      "epoch": 0.1918910222031535,
      "grad_norm": 0.22220343351364136,
      "learning_rate": 0.0001,
      "loss": 1.2599,
      "step": 7893
    },
    {
      "epoch": 0.1919982838142229,
      "grad_norm": 0.2304762750864029,
      "learning_rate": 0.0001,
      "loss": 1.266,
      "step": 7894
    },
    {
      "epoch": 0.1921055454252923,
      "grad_norm": 0.27136141061782837,
      "learning_rate": 0.0001,
      "loss": 1.2765,
      "step": 7895
    },
    {
      "epoch": 0.19221280703636168,
      "grad_norm": 0.22533777356147766,
      "learning_rate": 0.0001,
      "loss": 1.1808,
      "step": 7896
    },
    {
      "epoch": 0.1923200686474311,
      "grad_norm": 0.23050127923488617,
      "learning_rate": 0.0001,
      "loss": 1.3126,
      "step": 7897
    },
    {
      "epoch": 0.19242733025850048,
      "grad_norm": 0.22712531685829163,
      "learning_rate": 0.0001,
      "loss": 1.2542,
      "step": 7898
    },
    {
      "epoch": 0.1925345918695699,
      "grad_norm": 0.22426433861255646,
      "learning_rate": 0.0001,
      "loss": 1.167,
      "step": 7899
    },
    {
      "epoch": 0.19264185348063928,
      "grad_norm": 0.23001131415367126,
      "learning_rate": 0.0001,
      "loss": 1.2978,
      "step": 7900
    },
    {
      "epoch": 0.19274911509170867,
      "grad_norm": 0.2233201414346695,
      "learning_rate": 0.0001,
      "loss": 1.188,
      "step": 7901
    },
    {
      "epoch": 0.19285637670277808,
      "grad_norm": 0.21792984008789062,
      "learning_rate": 0.0001,
      "loss": 1.1956,
      "step": 7902
    },
    {
      "epoch": 0.19296363831384747,
      "grad_norm": 0.22599945962429047,
      "learning_rate": 0.0001,
      "loss": 1.1733,
      "step": 7903
    },
    {
      "epoch": 0.19307089992491688,
      "grad_norm": 0.21007733047008514,
      "learning_rate": 0.0001,
      "loss": 1.1166,
      "step": 7904
    },
    {
      "epoch": 0.19317816153598627,
      "grad_norm": 0.22591282427310944,
      "learning_rate": 0.0001,
      "loss": 1.213,
      "step": 7905
    },
    {
      "epoch": 0.19328542314705566,
      "grad_norm": 0.2248995006084442,
      "learning_rate": 0.0001,
      "loss": 1.2427,
      "step": 7906
    },
    {
      "epoch": 0.19339268475812507,
      "grad_norm": 0.21801427006721497,
      "learning_rate": 0.0001,
      "loss": 1.1906,
      "step": 7907
    },
    {
      "epoch": 0.19349994636919446,
      "grad_norm": 0.22156430780887604,
      "learning_rate": 0.0001,
      "loss": 1.1461,
      "step": 7908
    },
    {
      "epoch": 0.19360720798026387,
      "grad_norm": 0.22666575014591217,
      "learning_rate": 0.0001,
      "loss": 1.2744,
      "step": 7909
    },
    {
      "epoch": 0.19371446959133326,
      "grad_norm": 0.2354230433702469,
      "learning_rate": 0.0001,
      "loss": 1.2812,
      "step": 7910
    },
    {
      "epoch": 0.19382173120240265,
      "grad_norm": 0.2221653014421463,
      "learning_rate": 0.0001,
      "loss": 1.2654,
      "step": 7911
    },
    {
      "epoch": 0.19392899281347206,
      "grad_norm": 0.22011888027191162,
      "learning_rate": 0.0001,
      "loss": 1.2224,
      "step": 7912
    },
    {
      "epoch": 0.19403625442454145,
      "grad_norm": 0.21937265992164612,
      "learning_rate": 0.0001,
      "loss": 1.2691,
      "step": 7913
    },
    {
      "epoch": 0.19414351603561086,
      "grad_norm": 0.21382980048656464,
      "learning_rate": 0.0001,
      "loss": 1.2009,
      "step": 7914
    },
    {
      "epoch": 0.19425077764668025,
      "grad_norm": 0.2320699542760849,
      "learning_rate": 0.0001,
      "loss": 1.2704,
      "step": 7915
    },
    {
      "epoch": 0.19435803925774967,
      "grad_norm": 0.2436569184064865,
      "learning_rate": 0.0001,
      "loss": 1.2573,
      "step": 7916
    },
    {
      "epoch": 0.19446530086881905,
      "grad_norm": 0.239033043384552,
      "learning_rate": 0.0001,
      "loss": 1.2279,
      "step": 7917
    },
    {
      "epoch": 0.19457256247988844,
      "grad_norm": 0.22809839248657227,
      "learning_rate": 0.0001,
      "loss": 1.301,
      "step": 7918
    },
    {
      "epoch": 0.19467982409095785,
      "grad_norm": 0.24202363193035126,
      "learning_rate": 0.0001,
      "loss": 1.2982,
      "step": 7919
    },
    {
      "epoch": 0.19478708570202724,
      "grad_norm": 0.21196112036705017,
      "learning_rate": 0.0001,
      "loss": 1.2983,
      "step": 7920
    },
    {
      "epoch": 0.19489434731309666,
      "grad_norm": 0.2105206847190857,
      "learning_rate": 0.0001,
      "loss": 1.261,
      "step": 7921
    },
    {
      "epoch": 0.19500160892416604,
      "grad_norm": 0.22684334218502045,
      "learning_rate": 0.0001,
      "loss": 1.1676,
      "step": 7922
    },
    {
      "epoch": 0.19510887053523543,
      "grad_norm": 0.2240392416715622,
      "learning_rate": 0.0001,
      "loss": 1.2681,
      "step": 7923
    },
    {
      "epoch": 0.19521613214630484,
      "grad_norm": 0.21926553547382355,
      "learning_rate": 0.0001,
      "loss": 1.2998,
      "step": 7924
    },
    {
      "epoch": 0.19532339375737423,
      "grad_norm": 0.2333826720714569,
      "learning_rate": 0.0001,
      "loss": 1.2216,
      "step": 7925
    },
    {
      "epoch": 0.19543065536844365,
      "grad_norm": 0.23035673797130585,
      "learning_rate": 0.0001,
      "loss": 1.181,
      "step": 7926
    },
    {
      "epoch": 0.19553791697951303,
      "grad_norm": 0.2221689373254776,
      "learning_rate": 0.0001,
      "loss": 1.2063,
      "step": 7927
    },
    {
      "epoch": 0.19564517859058242,
      "grad_norm": 0.22691252827644348,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 7928
    },
    {
      "epoch": 0.19575244020165183,
      "grad_norm": 0.2261153757572174,
      "learning_rate": 0.0001,
      "loss": 1.3271,
      "step": 7929
    },
    {
      "epoch": 0.19585970181272122,
      "grad_norm": 0.22848249971866608,
      "learning_rate": 0.0001,
      "loss": 1.1831,
      "step": 7930
    },
    {
      "epoch": 0.19596696342379064,
      "grad_norm": 0.24909637868404388,
      "learning_rate": 0.0001,
      "loss": 1.3092,
      "step": 7931
    },
    {
      "epoch": 0.19607422503486002,
      "grad_norm": 0.2159939706325531,
      "learning_rate": 0.0001,
      "loss": 1.1735,
      "step": 7932
    },
    {
      "epoch": 0.1961814866459294,
      "grad_norm": 0.22955133020877838,
      "learning_rate": 0.0001,
      "loss": 1.2378,
      "step": 7933
    },
    {
      "epoch": 0.19628874825699882,
      "grad_norm": 0.21708634495735168,
      "learning_rate": 0.0001,
      "loss": 1.2625,
      "step": 7934
    },
    {
      "epoch": 0.1963960098680682,
      "grad_norm": 0.22340145707130432,
      "learning_rate": 0.0001,
      "loss": 1.2053,
      "step": 7935
    },
    {
      "epoch": 0.19650327147913763,
      "grad_norm": 0.23467983305454254,
      "learning_rate": 0.0001,
      "loss": 1.2175,
      "step": 7936
    },
    {
      "epoch": 0.196610533090207,
      "grad_norm": 0.23760581016540527,
      "learning_rate": 0.0001,
      "loss": 1.2748,
      "step": 7937
    },
    {
      "epoch": 0.1967177947012764,
      "grad_norm": 0.22550414502620697,
      "learning_rate": 0.0001,
      "loss": 1.2372,
      "step": 7938
    },
    {
      "epoch": 0.19682505631234581,
      "grad_norm": 0.2293098270893097,
      "learning_rate": 0.0001,
      "loss": 1.2193,
      "step": 7939
    },
    {
      "epoch": 0.1969323179234152,
      "grad_norm": 0.23134087026119232,
      "learning_rate": 0.0001,
      "loss": 1.2061,
      "step": 7940
    },
    {
      "epoch": 0.19703957953448462,
      "grad_norm": 0.22746427357196808,
      "learning_rate": 0.0001,
      "loss": 1.2589,
      "step": 7941
    },
    {
      "epoch": 0.197146841145554,
      "grad_norm": 0.25889718532562256,
      "learning_rate": 0.0001,
      "loss": 1.2938,
      "step": 7942
    },
    {
      "epoch": 0.19725410275662342,
      "grad_norm": 0.22594071924686432,
      "learning_rate": 0.0001,
      "loss": 1.1834,
      "step": 7943
    },
    {
      "epoch": 0.1973613643676928,
      "grad_norm": 0.226310133934021,
      "learning_rate": 0.0001,
      "loss": 1.1895,
      "step": 7944
    },
    {
      "epoch": 0.1974686259787622,
      "grad_norm": 0.22776469588279724,
      "learning_rate": 0.0001,
      "loss": 1.2629,
      "step": 7945
    },
    {
      "epoch": 0.1975758875898316,
      "grad_norm": 0.23393048346042633,
      "learning_rate": 0.0001,
      "loss": 1.2649,
      "step": 7946
    },
    {
      "epoch": 0.197683149200901,
      "grad_norm": 0.22387638688087463,
      "learning_rate": 0.0001,
      "loss": 1.1644,
      "step": 7947
    },
    {
      "epoch": 0.1977904108119704,
      "grad_norm": 0.23465077579021454,
      "learning_rate": 0.0001,
      "loss": 1.226,
      "step": 7948
    },
    {
      "epoch": 0.1978976724230398,
      "grad_norm": 0.24498413503170013,
      "learning_rate": 0.0001,
      "loss": 1.1854,
      "step": 7949
    },
    {
      "epoch": 0.19800493403410918,
      "grad_norm": 0.23396636545658112,
      "learning_rate": 0.0001,
      "loss": 1.2444,
      "step": 7950
    },
    {
      "epoch": 0.1981121956451786,
      "grad_norm": 0.23443444073200226,
      "learning_rate": 0.0001,
      "loss": 1.2699,
      "step": 7951
    },
    {
      "epoch": 0.19821945725624798,
      "grad_norm": 0.22723770141601562,
      "learning_rate": 0.0001,
      "loss": 1.2757,
      "step": 7952
    },
    {
      "epoch": 0.1983267188673174,
      "grad_norm": 0.2360474020242691,
      "learning_rate": 0.0001,
      "loss": 1.2289,
      "step": 7953
    },
    {
      "epoch": 0.19843398047838678,
      "grad_norm": 0.22715240716934204,
      "learning_rate": 0.0001,
      "loss": 1.3067,
      "step": 7954
    },
    {
      "epoch": 0.19854124208945617,
      "grad_norm": 0.24764809012413025,
      "learning_rate": 0.0001,
      "loss": 1.2119,
      "step": 7955
    },
    {
      "epoch": 0.19864850370052559,
      "grad_norm": 0.2477036714553833,
      "learning_rate": 0.0001,
      "loss": 1.2727,
      "step": 7956
    },
    {
      "epoch": 0.19875576531159497,
      "grad_norm": 0.2198021113872528,
      "learning_rate": 0.0001,
      "loss": 1.2615,
      "step": 7957
    },
    {
      "epoch": 0.1988630269226644,
      "grad_norm": 0.23521418869495392,
      "learning_rate": 0.0001,
      "loss": 1.2555,
      "step": 7958
    },
    {
      "epoch": 0.19897028853373377,
      "grad_norm": 0.22685666382312775,
      "learning_rate": 0.0001,
      "loss": 1.177,
      "step": 7959
    },
    {
      "epoch": 0.19907755014480316,
      "grad_norm": 0.21676260232925415,
      "learning_rate": 0.0001,
      "loss": 1.2003,
      "step": 7960
    },
    {
      "epoch": 0.19918481175587258,
      "grad_norm": 0.240080788731575,
      "learning_rate": 0.0001,
      "loss": 1.2765,
      "step": 7961
    },
    {
      "epoch": 0.19929207336694196,
      "grad_norm": 0.23905925452709198,
      "learning_rate": 0.0001,
      "loss": 1.267,
      "step": 7962
    },
    {
      "epoch": 0.19939933497801138,
      "grad_norm": 0.21543651819229126,
      "learning_rate": 0.0001,
      "loss": 1.2664,
      "step": 7963
    },
    {
      "epoch": 0.19950659658908076,
      "grad_norm": 0.22770127654075623,
      "learning_rate": 0.0001,
      "loss": 1.3503,
      "step": 7964
    },
    {
      "epoch": 0.19961385820015018,
      "grad_norm": 0.221278578042984,
      "learning_rate": 0.0001,
      "loss": 1.3527,
      "step": 7965
    },
    {
      "epoch": 0.19972111981121957,
      "grad_norm": 0.25916585326194763,
      "learning_rate": 0.0001,
      "loss": 1.2684,
      "step": 7966
    },
    {
      "epoch": 0.19982838142228895,
      "grad_norm": 0.2190437614917755,
      "learning_rate": 0.0001,
      "loss": 1.2418,
      "step": 7967
    },
    {
      "epoch": 0.19993564303335837,
      "grad_norm": 0.22505059838294983,
      "learning_rate": 0.0001,
      "loss": 1.1775,
      "step": 7968
    },
    {
      "epoch": 0.20004290464442775,
      "grad_norm": 0.22435835003852844,
      "learning_rate": 0.0001,
      "loss": 1.267,
      "step": 7969
    },
    {
      "epoch": 0.20015016625549717,
      "grad_norm": 0.2132001668214798,
      "learning_rate": 0.0001,
      "loss": 1.158,
      "step": 7970
    },
    {
      "epoch": 0.20025742786656656,
      "grad_norm": 0.22290150821208954,
      "learning_rate": 0.0001,
      "loss": 1.244,
      "step": 7971
    },
    {
      "epoch": 0.20036468947763594,
      "grad_norm": 0.2320772111415863,
      "learning_rate": 0.0001,
      "loss": 1.3217,
      "step": 7972
    },
    {
      "epoch": 0.20047195108870536,
      "grad_norm": 0.22992585599422455,
      "learning_rate": 0.0001,
      "loss": 1.3102,
      "step": 7973
    },
    {
      "epoch": 0.20057921269977474,
      "grad_norm": 0.2275760918855667,
      "learning_rate": 0.0001,
      "loss": 1.2348,
      "step": 7974
    },
    {
      "epoch": 0.20068647431084416,
      "grad_norm": 0.21720941364765167,
      "learning_rate": 0.0001,
      "loss": 1.3209,
      "step": 7975
    },
    {
      "epoch": 0.20079373592191355,
      "grad_norm": 0.22566057741641998,
      "learning_rate": 0.0001,
      "loss": 1.2246,
      "step": 7976
    },
    {
      "epoch": 0.20090099753298293,
      "grad_norm": 0.23235873878002167,
      "learning_rate": 0.0001,
      "loss": 1.1524,
      "step": 7977
    },
    {
      "epoch": 0.20100825914405235,
      "grad_norm": 0.2323780357837677,
      "learning_rate": 0.0001,
      "loss": 1.2542,
      "step": 7978
    },
    {
      "epoch": 0.20111552075512173,
      "grad_norm": 0.22813846170902252,
      "learning_rate": 0.0001,
      "loss": 1.3328,
      "step": 7979
    },
    {
      "epoch": 0.20122278236619115,
      "grad_norm": 0.21298612654209137,
      "learning_rate": 0.0001,
      "loss": 1.2049,
      "step": 7980
    },
    {
      "epoch": 0.20133004397726054,
      "grad_norm": 0.2640489637851715,
      "learning_rate": 0.0001,
      "loss": 1.2266,
      "step": 7981
    },
    {
      "epoch": 0.20143730558832992,
      "grad_norm": 0.23418591916561127,
      "learning_rate": 0.0001,
      "loss": 1.3057,
      "step": 7982
    },
    {
      "epoch": 0.20154456719939934,
      "grad_norm": 0.22754281759262085,
      "learning_rate": 0.0001,
      "loss": 1.316,
      "step": 7983
    },
    {
      "epoch": 0.20165182881046873,
      "grad_norm": 0.23071229457855225,
      "learning_rate": 0.0001,
      "loss": 1.2315,
      "step": 7984
    },
    {
      "epoch": 0.20175909042153814,
      "grad_norm": 0.23398931324481964,
      "learning_rate": 0.0001,
      "loss": 1.302,
      "step": 7985
    },
    {
      "epoch": 0.20186635203260753,
      "grad_norm": 0.22595639526844025,
      "learning_rate": 0.0001,
      "loss": 1.1434,
      "step": 7986
    },
    {
      "epoch": 0.20197361364367694,
      "grad_norm": 0.22858445346355438,
      "learning_rate": 0.0001,
      "loss": 1.2981,
      "step": 7987
    },
    {
      "epoch": 0.20208087525474633,
      "grad_norm": 0.21995809674263,
      "learning_rate": 0.0001,
      "loss": 1.2591,
      "step": 7988
    },
    {
      "epoch": 0.20218813686581572,
      "grad_norm": 0.22877733409404755,
      "learning_rate": 0.0001,
      "loss": 1.2114,
      "step": 7989
    },
    {
      "epoch": 0.20229539847688513,
      "grad_norm": 0.2326286882162094,
      "learning_rate": 0.0001,
      "loss": 1.2235,
      "step": 7990
    },
    {
      "epoch": 0.20240266008795452,
      "grad_norm": 0.22922207415103912,
      "learning_rate": 0.0001,
      "loss": 1.2231,
      "step": 7991
    },
    {
      "epoch": 0.20250992169902393,
      "grad_norm": 0.2235378623008728,
      "learning_rate": 0.0001,
      "loss": 1.2152,
      "step": 7992
    },
    {
      "epoch": 0.20261718331009332,
      "grad_norm": 0.22180213034152985,
      "learning_rate": 0.0001,
      "loss": 1.152,
      "step": 7993
    },
    {
      "epoch": 0.2027244449211627,
      "grad_norm": 0.21246811747550964,
      "learning_rate": 0.0001,
      "loss": 1.2152,
      "step": 7994
    },
    {
      "epoch": 0.20283170653223212,
      "grad_norm": 0.2278171181678772,
      "learning_rate": 0.0001,
      "loss": 1.3045,
      "step": 7995
    },
    {
      "epoch": 0.2029389681433015,
      "grad_norm": 0.2345476597547531,
      "learning_rate": 0.0001,
      "loss": 1.2493,
      "step": 7996
    },
    {
      "epoch": 0.20304622975437092,
      "grad_norm": 0.2180880457162857,
      "learning_rate": 0.0001,
      "loss": 1.2549,
      "step": 7997
    },
    {
      "epoch": 0.2031534913654403,
      "grad_norm": 0.22175633907318115,
      "learning_rate": 0.0001,
      "loss": 1.2775,
      "step": 7998
    },
    {
      "epoch": 0.2032607529765097,
      "grad_norm": 0.24466894567012787,
      "learning_rate": 0.0001,
      "loss": 1.2447,
      "step": 7999
    },
    {
      "epoch": 0.2033680145875791,
      "grad_norm": 0.23182091116905212,
      "learning_rate": 0.0001,
      "loss": 1.3555,
      "step": 8000
    }
  ],
  "logging_steps": 1,
  "max_steps": 9323,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.564072927133696e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}