{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 269,
  "global_step": 1073,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009319664492078285,
      "grad_norm": 23.36980628967285,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 9.2544,
      "step": 1
    },
    {
      "epoch": 0.0009319664492078285,
      "eval_loss": 9.81535816192627,
      "eval_runtime": 126.0986,
      "eval_samples_per_second": 14.33,
      "eval_steps_per_second": 3.584,
      "step": 1
    },
    {
      "epoch": 0.001863932898415657,
      "grad_norm": 18.088346481323242,
      "learning_rate": 4.000000000000001e-06,
      "loss": 9.099,
      "step": 2
    },
    {
      "epoch": 0.0027958993476234857,
      "grad_norm": 18.9443359375,
      "learning_rate": 6e-06,
      "loss": 8.8085,
      "step": 3
    },
    {
      "epoch": 0.003727865796831314,
      "grad_norm": 19.623462677001953,
      "learning_rate": 8.000000000000001e-06,
      "loss": 8.7971,
      "step": 4
    },
    {
      "epoch": 0.004659832246039142,
      "grad_norm": 20.289655685424805,
      "learning_rate": 1e-05,
      "loss": 8.7141,
      "step": 5
    },
    {
      "epoch": 0.005591798695246971,
      "grad_norm": 24.600669860839844,
      "learning_rate": 1.2e-05,
      "loss": 9.3088,
      "step": 6
    },
    {
      "epoch": 0.0065237651444548,
      "grad_norm": 19.402236938476562,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 9.213,
      "step": 7
    },
    {
      "epoch": 0.007455731593662628,
      "grad_norm": 17.378143310546875,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 8.2946,
      "step": 8
    },
    {
      "epoch": 0.008387698042870456,
      "grad_norm": 20.91484832763672,
      "learning_rate": 1.8e-05,
      "loss": 8.6097,
      "step": 9
    },
    {
      "epoch": 0.009319664492078284,
      "grad_norm": 20.42852020263672,
      "learning_rate": 2e-05,
      "loss": 8.1296,
      "step": 10
    },
    {
      "epoch": 0.010251630941286114,
      "grad_norm": 17.04209327697754,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 8.5266,
      "step": 11
    },
    {
      "epoch": 0.011183597390493943,
      "grad_norm": 21.660465240478516,
      "learning_rate": 2.4e-05,
      "loss": 9.4731,
      "step": 12
    },
    {
      "epoch": 0.012115563839701771,
      "grad_norm": 18.74623680114746,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 7.6045,
      "step": 13
    },
    {
      "epoch": 0.0130475302889096,
      "grad_norm": 19.178571701049805,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 7.9309,
      "step": 14
    },
    {
      "epoch": 0.013979496738117428,
      "grad_norm": 19.65428924560547,
      "learning_rate": 3e-05,
      "loss": 8.9349,
      "step": 15
    },
    {
      "epoch": 0.014911463187325256,
      "grad_norm": 19.72906494140625,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 8.135,
      "step": 16
    },
    {
      "epoch": 0.015843429636533086,
      "grad_norm": 17.341209411621094,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 7.803,
      "step": 17
    },
    {
      "epoch": 0.016775396085740912,
      "grad_norm": 16.997432708740234,
      "learning_rate": 3.6e-05,
      "loss": 7.5903,
      "step": 18
    },
    {
      "epoch": 0.017707362534948742,
      "grad_norm": 18.027070999145508,
      "learning_rate": 3.8e-05,
      "loss": 7.8597,
      "step": 19
    },
    {
      "epoch": 0.01863932898415657,
      "grad_norm": 18.275178909301758,
      "learning_rate": 4e-05,
      "loss": 8.2936,
      "step": 20
    },
    {
      "epoch": 0.0195712954333644,
      "grad_norm": 19.839303970336914,
      "learning_rate": 4.2e-05,
      "loss": 8.1574,
      "step": 21
    },
    {
      "epoch": 0.02050326188257223,
      "grad_norm": 19.413339614868164,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 8.2292,
      "step": 22
    },
    {
      "epoch": 0.021435228331780055,
      "grad_norm": 25.84449577331543,
      "learning_rate": 4.600000000000001e-05,
      "loss": 6.6394,
      "step": 23
    },
    {
      "epoch": 0.022367194780987885,
      "grad_norm": 29.13704490661621,
      "learning_rate": 4.8e-05,
      "loss": 6.0639,
      "step": 24
    },
    {
      "epoch": 0.023299161230195712,
      "grad_norm": 106.7900619506836,
      "learning_rate": 5e-05,
      "loss": 6.7721,
      "step": 25
    },
    {
      "epoch": 0.024231127679403542,
      "grad_norm": 17.748001098632812,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 5.3139,
      "step": 26
    },
    {
      "epoch": 0.02516309412861137,
      "grad_norm": 19.872982025146484,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 4.7766,
      "step": 27
    },
    {
      "epoch": 0.0260950605778192,
      "grad_norm": 19.577491760253906,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 4.7364,
      "step": 28
    },
    {
      "epoch": 0.02702702702702703,
      "grad_norm": 20.095836639404297,
      "learning_rate": 5.8e-05,
      "loss": 3.182,
      "step": 29
    },
    {
      "epoch": 0.027958993476234855,
      "grad_norm": 18.048362731933594,
      "learning_rate": 6e-05,
      "loss": 2.9527,
      "step": 30
    },
    {
      "epoch": 0.028890959925442685,
      "grad_norm": 18.98998260498047,
      "learning_rate": 6.2e-05,
      "loss": 2.4878,
      "step": 31
    },
    {
      "epoch": 0.02982292637465051,
      "grad_norm": 17.988750457763672,
      "learning_rate": 6.400000000000001e-05,
      "loss": 2.4282,
      "step": 32
    },
    {
      "epoch": 0.03075489282385834,
      "grad_norm": 15.404643058776855,
      "learning_rate": 6.6e-05,
      "loss": 2.4879,
      "step": 33
    },
    {
      "epoch": 0.03168685927306617,
      "grad_norm": 16.794700622558594,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.8821,
      "step": 34
    },
    {
      "epoch": 0.032618825722273995,
      "grad_norm": 12.591143608093262,
      "learning_rate": 7e-05,
      "loss": 2.0967,
      "step": 35
    },
    {
      "epoch": 0.033550792171481825,
      "grad_norm": 12.896982192993164,
      "learning_rate": 7.2e-05,
      "loss": 1.4124,
      "step": 36
    },
    {
      "epoch": 0.034482758620689655,
      "grad_norm": 13.917445182800293,
      "learning_rate": 7.4e-05,
      "loss": 1.6743,
      "step": 37
    },
    {
      "epoch": 0.035414725069897485,
      "grad_norm": 11.748924255371094,
      "learning_rate": 7.6e-05,
      "loss": 1.7939,
      "step": 38
    },
    {
      "epoch": 0.036346691519105315,
      "grad_norm": 12.96652603149414,
      "learning_rate": 7.800000000000001e-05,
      "loss": 1.1786,
      "step": 39
    },
    {
      "epoch": 0.03727865796831314,
      "grad_norm": 20.441051483154297,
      "learning_rate": 8e-05,
      "loss": 1.1189,
      "step": 40
    },
    {
      "epoch": 0.03821062441752097,
      "grad_norm": 15.231740951538086,
      "learning_rate": 8.2e-05,
      "loss": 1.3477,
      "step": 41
    },
    {
      "epoch": 0.0391425908667288,
      "grad_norm": 11.24478530883789,
      "learning_rate": 8.4e-05,
      "loss": 1.0997,
      "step": 42
    },
    {
      "epoch": 0.04007455731593663,
      "grad_norm": 18.6605224609375,
      "learning_rate": 8.6e-05,
      "loss": 1.2903,
      "step": 43
    },
    {
      "epoch": 0.04100652376514446,
      "grad_norm": 13.082992553710938,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.7943,
      "step": 44
    },
    {
      "epoch": 0.04193849021435228,
      "grad_norm": 12.835314750671387,
      "learning_rate": 9e-05,
      "loss": 1.2972,
      "step": 45
    },
    {
      "epoch": 0.04287045666356011,
      "grad_norm": 12.402058601379395,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.7079,
      "step": 46
    },
    {
      "epoch": 0.04380242311276794,
      "grad_norm": 12.079973220825195,
      "learning_rate": 9.4e-05,
      "loss": 2.0969,
      "step": 47
    },
    {
      "epoch": 0.04473438956197577,
      "grad_norm": 11.028013229370117,
      "learning_rate": 9.6e-05,
      "loss": 1.0664,
      "step": 48
    },
    {
      "epoch": 0.045666356011183594,
      "grad_norm": 24.05512809753418,
      "learning_rate": 9.8e-05,
      "loss": 1.1056,
      "step": 49
    },
    {
      "epoch": 0.046598322460391424,
      "grad_norm": 89.79100036621094,
      "learning_rate": 0.0001,
      "loss": 1.385,
      "step": 50
    },
    {
      "epoch": 0.047530288909599254,
      "grad_norm": 9.370144844055176,
      "learning_rate": 9.999976423021617e-05,
      "loss": 0.3987,
      "step": 51
    },
    {
      "epoch": 0.048462255358807084,
      "grad_norm": 9.368326187133789,
      "learning_rate": 9.999905692308813e-05,
      "loss": 0.5446,
      "step": 52
    },
    {
      "epoch": 0.049394221808014914,
      "grad_norm": 8.342625617980957,
      "learning_rate": 9.999787808528638e-05,
      "loss": 0.5475,
      "step": 53
    },
    {
      "epoch": 0.05032618825722274,
      "grad_norm": 3.533601999282837,
      "learning_rate": 9.999622772792828e-05,
      "loss": 0.2045,
      "step": 54
    },
    {
      "epoch": 0.05125815470643057,
      "grad_norm": 5.094264507293701,
      "learning_rate": 9.9994105866578e-05,
      "loss": 0.4675,
      "step": 55
    },
    {
      "epoch": 0.0521901211556384,
      "grad_norm": 8.645957946777344,
      "learning_rate": 9.999151252124639e-05,
      "loss": 0.2873,
      "step": 56
    },
    {
      "epoch": 0.05312208760484623,
      "grad_norm": 10.678857803344727,
      "learning_rate": 9.998844771639073e-05,
      "loss": 0.182,
      "step": 57
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 9.783062934875488,
      "learning_rate": 9.998491148091457e-05,
      "loss": 0.5509,
      "step": 58
    },
    {
      "epoch": 0.05498602050326188,
      "grad_norm": 4.751331329345703,
      "learning_rate": 9.99809038481674e-05,
      "loss": 0.1962,
      "step": 59
    },
    {
      "epoch": 0.05591798695246971,
      "grad_norm": 7.471037864685059,
      "learning_rate": 9.997642485594436e-05,
      "loss": 0.4997,
      "step": 60
    },
    {
      "epoch": 0.05684995340167754,
      "grad_norm": 3.7392351627349854,
      "learning_rate": 9.99714745464859e-05,
      "loss": 0.2825,
      "step": 61
    },
    {
      "epoch": 0.05778191985088537,
      "grad_norm": 3.4164586067199707,
      "learning_rate": 9.996605296647737e-05,
      "loss": 0.3177,
      "step": 62
    },
    {
      "epoch": 0.05871388630009319,
      "grad_norm": 7.206740856170654,
      "learning_rate": 9.996016016704854e-05,
      "loss": 0.2987,
      "step": 63
    },
    {
      "epoch": 0.05964585274930102,
      "grad_norm": 3.8595774173736572,
      "learning_rate": 9.995379620377318e-05,
      "loss": 0.3315,
      "step": 64
    },
    {
      "epoch": 0.06057781919850885,
      "grad_norm": 2.3879973888397217,
      "learning_rate": 9.99469611366685e-05,
      "loss": 0.2275,
      "step": 65
    },
    {
      "epoch": 0.06150978564771668,
      "grad_norm": 3.690859794616699,
      "learning_rate": 9.993965503019457e-05,
      "loss": 0.25,
      "step": 66
    },
    {
      "epoch": 0.06244175209692451,
      "grad_norm": 5.447560787200928,
      "learning_rate": 9.99318779532538e-05,
      "loss": 0.5862,
      "step": 67
    },
    {
      "epoch": 0.06337371854613234,
      "grad_norm": 3.505098819732666,
      "learning_rate": 9.992362997919016e-05,
      "loss": 0.4662,
      "step": 68
    },
    {
      "epoch": 0.06430568499534017,
      "grad_norm": 5.845356464385986,
      "learning_rate": 9.991491118578856e-05,
      "loss": 0.393,
      "step": 69
    },
    {
      "epoch": 0.06523765144454799,
      "grad_norm": 5.72480583190918,
      "learning_rate": 9.990572165527413e-05,
      "loss": 0.5041,
      "step": 70
    },
    {
      "epoch": 0.06616961789375582,
      "grad_norm": 5.941817283630371,
      "learning_rate": 9.98960614743114e-05,
      "loss": 0.4596,
      "step": 71
    },
    {
      "epoch": 0.06710158434296365,
      "grad_norm": 9.465463638305664,
      "learning_rate": 9.988593073400354e-05,
      "loss": 0.6977,
      "step": 72
    },
    {
      "epoch": 0.06803355079217148,
      "grad_norm": 8.731477737426758,
      "learning_rate": 9.987532952989145e-05,
      "loss": 0.4233,
      "step": 73
    },
    {
      "epoch": 0.06896551724137931,
      "grad_norm": 15.845611572265625,
      "learning_rate": 9.986425796195287e-05,
      "loss": 0.6312,
      "step": 74
    },
    {
      "epoch": 0.06989748369058714,
      "grad_norm": 37.47452926635742,
      "learning_rate": 9.985271613460144e-05,
      "loss": 1.1223,
      "step": 75
    },
    {
      "epoch": 0.07082945013979497,
      "grad_norm": 10.461603164672852,
      "learning_rate": 9.984070415668574e-05,
      "loss": 0.3865,
      "step": 76
    },
    {
      "epoch": 0.0717614165890028,
      "grad_norm": 4.933948993682861,
      "learning_rate": 9.982822214148821e-05,
      "loss": 0.2947,
      "step": 77
    },
    {
      "epoch": 0.07269338303821063,
      "grad_norm": 5.168891906738281,
      "learning_rate": 9.981527020672413e-05,
      "loss": 0.1991,
      "step": 78
    },
    {
      "epoch": 0.07362534948741846,
      "grad_norm": 8.993535995483398,
      "learning_rate": 9.980184847454052e-05,
      "loss": 0.474,
      "step": 79
    },
    {
      "epoch": 0.07455731593662628,
      "grad_norm": 10.690693855285645,
      "learning_rate": 9.978795707151492e-05,
      "loss": 0.5728,
      "step": 80
    },
    {
      "epoch": 0.0754892823858341,
      "grad_norm": 7.1764044761657715,
      "learning_rate": 9.977359612865423e-05,
      "loss": 0.324,
      "step": 81
    },
    {
      "epoch": 0.07642124883504194,
      "grad_norm": 3.059382438659668,
      "learning_rate": 9.975876578139355e-05,
      "loss": 0.2248,
      "step": 82
    },
    {
      "epoch": 0.07735321528424977,
      "grad_norm": 1.5979907512664795,
      "learning_rate": 9.974346616959476e-05,
      "loss": 0.1305,
      "step": 83
    },
    {
      "epoch": 0.0782851817334576,
      "grad_norm": 4.272276878356934,
      "learning_rate": 9.972769743754531e-05,
      "loss": 0.288,
      "step": 84
    },
    {
      "epoch": 0.07921714818266543,
      "grad_norm": 7.77036190032959,
      "learning_rate": 9.971145973395684e-05,
      "loss": 0.3033,
      "step": 85
    },
    {
      "epoch": 0.08014911463187326,
      "grad_norm": 6.204690933227539,
      "learning_rate": 9.969475321196373e-05,
      "loss": 0.347,
      "step": 86
    },
    {
      "epoch": 0.08108108108108109,
      "grad_norm": 8.32636833190918,
      "learning_rate": 9.96775780291217e-05,
      "loss": 0.2946,
      "step": 87
    },
    {
      "epoch": 0.08201304753028892,
      "grad_norm": 4.755763530731201,
      "learning_rate": 9.965993434740633e-05,
      "loss": 0.3311,
      "step": 88
    },
    {
      "epoch": 0.08294501397949673,
      "grad_norm": 1.5085945129394531,
      "learning_rate": 9.96418223332115e-05,
      "loss": 0.1361,
      "step": 89
    },
    {
      "epoch": 0.08387698042870456,
      "grad_norm": 5.299545764923096,
      "learning_rate": 9.962324215734782e-05,
      "loss": 0.3351,
      "step": 90
    },
    {
      "epoch": 0.08480894687791239,
      "grad_norm": 4.837111949920654,
      "learning_rate": 9.960419399504106e-05,
      "loss": 0.3379,
      "step": 91
    },
    {
      "epoch": 0.08574091332712022,
      "grad_norm": 5.275478839874268,
      "learning_rate": 9.958467802593046e-05,
      "loss": 0.2991,
      "step": 92
    },
    {
      "epoch": 0.08667287977632805,
      "grad_norm": 6.431980609893799,
      "learning_rate": 9.956469443406706e-05,
      "loss": 0.441,
      "step": 93
    },
    {
      "epoch": 0.08760484622553588,
      "grad_norm": 5.4865264892578125,
      "learning_rate": 9.954424340791196e-05,
      "loss": 0.3432,
      "step": 94
    },
    {
      "epoch": 0.08853681267474371,
      "grad_norm": 5.279883861541748,
      "learning_rate": 9.952332514033447e-05,
      "loss": 0.3338,
      "step": 95
    },
    {
      "epoch": 0.08946877912395154,
      "grad_norm": 2.8239941596984863,
      "learning_rate": 9.950193982861047e-05,
      "loss": 0.1205,
      "step": 96
    },
    {
      "epoch": 0.09040074557315937,
      "grad_norm": 10.564650535583496,
      "learning_rate": 9.948008767442034e-05,
      "loss": 0.1239,
      "step": 97
    },
    {
      "epoch": 0.09133271202236719,
      "grad_norm": 5.720789909362793,
      "learning_rate": 9.94577688838472e-05,
      "loss": 0.2613,
      "step": 98
    },
    {
      "epoch": 0.09226467847157502,
      "grad_norm": 28.29906463623047,
      "learning_rate": 9.943498366737487e-05,
      "loss": 0.6763,
      "step": 99
    },
    {
      "epoch": 0.09319664492078285,
      "grad_norm": 85.7387924194336,
      "learning_rate": 9.941173223988602e-05,
      "loss": 1.8956,
      "step": 100
    },
    {
      "epoch": 0.09412861136999068,
      "grad_norm": 2.3096601963043213,
      "learning_rate": 9.938801482065998e-05,
      "loss": 0.2865,
      "step": 101
    },
    {
      "epoch": 0.09506057781919851,
      "grad_norm": 3.718655586242676,
      "learning_rate": 9.93638316333708e-05,
      "loss": 0.3133,
      "step": 102
    },
    {
      "epoch": 0.09599254426840634,
      "grad_norm": 3.8024020195007324,
      "learning_rate": 9.933918290608509e-05,
      "loss": 0.183,
      "step": 103
    },
    {
      "epoch": 0.09692451071761417,
      "grad_norm": 2.2328927516937256,
      "learning_rate": 9.93140688712598e-05,
      "loss": 0.2483,
      "step": 104
    },
    {
      "epoch": 0.097856477166822,
      "grad_norm": 3.0202109813690186,
      "learning_rate": 9.928848976574019e-05,
      "loss": 0.2782,
      "step": 105
    },
    {
      "epoch": 0.09878844361602983,
      "grad_norm": 1.373205304145813,
      "learning_rate": 9.926244583075748e-05,
      "loss": 0.0962,
      "step": 106
    },
    {
      "epoch": 0.09972041006523766,
      "grad_norm": 7.907721519470215,
      "learning_rate": 9.923593731192655e-05,
      "loss": 0.3763,
      "step": 107
    },
    {
      "epoch": 0.10065237651444547,
      "grad_norm": 4.7565598487854,
      "learning_rate": 9.920896445924373e-05,
      "loss": 0.2433,
      "step": 108
    },
    {
      "epoch": 0.1015843429636533,
      "grad_norm": 3.6429123878479004,
      "learning_rate": 9.918152752708438e-05,
      "loss": 0.1288,
      "step": 109
    },
    {
      "epoch": 0.10251630941286113,
      "grad_norm": 3.3588905334472656,
      "learning_rate": 9.915362677420044e-05,
      "loss": 0.2031,
      "step": 110
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 2.161036491394043,
      "learning_rate": 9.912526246371814e-05,
      "loss": 0.0812,
      "step": 111
    },
    {
      "epoch": 0.1043802423112768,
      "grad_norm": 2.162220001220703,
      "learning_rate": 9.909643486313533e-05,
      "loss": 0.1037,
      "step": 112
    },
    {
      "epoch": 0.10531220876048462,
      "grad_norm": 3.823596239089966,
      "learning_rate": 9.906714424431913e-05,
      "loss": 0.1591,
      "step": 113
    },
    {
      "epoch": 0.10624417520969245,
      "grad_norm": 3.6801645755767822,
      "learning_rate": 9.903739088350325e-05,
      "loss": 0.147,
      "step": 114
    },
    {
      "epoch": 0.10717614165890028,
      "grad_norm": 6.397568702697754,
      "learning_rate": 9.900717506128541e-05,
      "loss": 0.3958,
      "step": 115
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 4.3636794090271,
      "learning_rate": 9.897649706262473e-05,
      "loss": 0.2412,
      "step": 116
    },
    {
      "epoch": 0.10904007455731593,
      "grad_norm": 2.604064702987671,
      "learning_rate": 9.894535717683902e-05,
      "loss": 0.2321,
      "step": 117
    },
    {
      "epoch": 0.10997204100652376,
      "grad_norm": 3.3178787231445312,
      "learning_rate": 9.891375569760204e-05,
      "loss": 0.2164,
      "step": 118
    },
    {
      "epoch": 0.11090400745573159,
      "grad_norm": 1.7841252088546753,
      "learning_rate": 9.888169292294076e-05,
      "loss": 0.1164,
      "step": 119
    },
    {
      "epoch": 0.11183597390493942,
      "grad_norm": 3.5098044872283936,
      "learning_rate": 9.884916915523251e-05,
      "loss": 0.2846,
      "step": 120
    },
    {
      "epoch": 0.11276794035414725,
      "grad_norm": 2.65773606300354,
      "learning_rate": 9.881618470120215e-05,
      "loss": 0.2409,
      "step": 121
    },
    {
      "epoch": 0.11369990680335508,
      "grad_norm": 5.54008674621582,
      "learning_rate": 9.87827398719192e-05,
      "loss": 0.2004,
      "step": 122
    },
    {
      "epoch": 0.11463187325256291,
      "grad_norm": 5.397515773773193,
      "learning_rate": 9.874883498279484e-05,
      "loss": 0.2435,
      "step": 123
    },
    {
      "epoch": 0.11556383970177074,
      "grad_norm": 7.1401777267456055,
      "learning_rate": 9.871447035357903e-05,
      "loss": 0.2561,
      "step": 124
    },
    {
      "epoch": 0.11649580615097857,
      "grad_norm": 23.130657196044922,
      "learning_rate": 9.867964630835743e-05,
      "loss": 0.6442,
      "step": 125
    },
    {
      "epoch": 0.11742777260018639,
      "grad_norm": 2.3777284622192383,
      "learning_rate": 9.864436317554831e-05,
      "loss": 0.1827,
      "step": 126
    },
    {
      "epoch": 0.11835973904939422,
      "grad_norm": 2.1437063217163086,
      "learning_rate": 9.860862128789953e-05,
      "loss": 0.2189,
      "step": 127
    },
    {
      "epoch": 0.11929170549860205,
      "grad_norm": 3.7361228466033936,
      "learning_rate": 9.857242098248542e-05,
      "loss": 0.3052,
      "step": 128
    },
    {
      "epoch": 0.12022367194780988,
      "grad_norm": 1.8870540857315063,
      "learning_rate": 9.853576260070347e-05,
      "loss": 0.2004,
      "step": 129
    },
    {
      "epoch": 0.1211556383970177,
      "grad_norm": 4.224307537078857,
      "learning_rate": 9.849864648827125e-05,
      "loss": 0.2409,
      "step": 130
    },
    {
      "epoch": 0.12208760484622554,
      "grad_norm": 2.4363510608673096,
      "learning_rate": 9.846107299522304e-05,
      "loss": 0.077,
      "step": 131
    },
    {
      "epoch": 0.12301957129543337,
      "grad_norm": 1.9710980653762817,
      "learning_rate": 9.842304247590667e-05,
      "loss": 0.1215,
      "step": 132
    },
    {
      "epoch": 0.1239515377446412,
      "grad_norm": 1.8237957954406738,
      "learning_rate": 9.838455528897998e-05,
      "loss": 0.1306,
      "step": 133
    },
    {
      "epoch": 0.12488350419384903,
      "grad_norm": 1.345870018005371,
      "learning_rate": 9.834561179740762e-05,
      "loss": 0.0764,
      "step": 134
    },
    {
      "epoch": 0.12581547064305684,
      "grad_norm": 1.505752682685852,
      "learning_rate": 9.830621236845754e-05,
      "loss": 0.1416,
      "step": 135
    },
    {
      "epoch": 0.1267474370922647,
      "grad_norm": 1.9039655923843384,
      "learning_rate": 9.826635737369751e-05,
      "loss": 0.1965,
      "step": 136
    },
    {
      "epoch": 0.1276794035414725,
      "grad_norm": 1.7544796466827393,
      "learning_rate": 9.82260471889917e-05,
      "loss": 0.122,
      "step": 137
    },
    {
      "epoch": 0.12861136999068035,
      "grad_norm": 3.835097312927246,
      "learning_rate": 9.818528219449705e-05,
      "loss": 0.2722,
      "step": 138
    },
    {
      "epoch": 0.12954333643988816,
      "grad_norm": 2.6565046310424805,
      "learning_rate": 9.814406277465968e-05,
      "loss": 0.1866,
      "step": 139
    },
    {
      "epoch": 0.13047530288909598,
      "grad_norm": 1.8572436571121216,
      "learning_rate": 9.810238931821138e-05,
      "loss": 0.1344,
      "step": 140
    },
    {
      "epoch": 0.13140726933830382,
      "grad_norm": 1.8303385972976685,
      "learning_rate": 9.806026221816581e-05,
      "loss": 0.1429,
      "step": 141
    },
    {
      "epoch": 0.13233923578751164,
      "grad_norm": 2.8735477924346924,
      "learning_rate": 9.801768187181486e-05,
      "loss": 0.1621,
      "step": 142
    },
    {
      "epoch": 0.13327120223671948,
      "grad_norm": 2.632936477661133,
      "learning_rate": 9.797464868072488e-05,
      "loss": 0.2389,
      "step": 143
    },
    {
      "epoch": 0.1342031686859273,
      "grad_norm": 4.749107360839844,
      "learning_rate": 9.793116305073292e-05,
      "loss": 0.3371,
      "step": 144
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 3.6311898231506348,
      "learning_rate": 9.788722539194291e-05,
      "loss": 0.2557,
      "step": 145
    },
    {
      "epoch": 0.13606710158434296,
      "grad_norm": 3.9647810459136963,
      "learning_rate": 9.784283611872169e-05,
      "loss": 0.0985,
      "step": 146
    },
    {
      "epoch": 0.1369990680335508,
      "grad_norm": 7.041486740112305,
      "learning_rate": 9.77979956496953e-05,
      "loss": 0.3009,
      "step": 147
    },
    {
      "epoch": 0.13793103448275862,
      "grad_norm": 2.975712299346924,
      "learning_rate": 9.77527044077448e-05,
      "loss": 0.1418,
      "step": 148
    },
    {
      "epoch": 0.13886300093196646,
      "grad_norm": 6.61073637008667,
      "learning_rate": 9.770696282000244e-05,
      "loss": 0.2365,
      "step": 149
    },
    {
      "epoch": 0.13979496738117428,
      "grad_norm": 8.437379837036133,
      "learning_rate": 9.766077131784763e-05,
      "loss": 0.2798,
      "step": 150
    },
    {
      "epoch": 0.1407269338303821,
      "grad_norm": 3.151240587234497,
      "learning_rate": 9.761413033690276e-05,
      "loss": 0.1862,
      "step": 151
    },
    {
      "epoch": 0.14165890027958994,
      "grad_norm": 4.774233818054199,
      "learning_rate": 9.756704031702918e-05,
      "loss": 0.4372,
      "step": 152
    },
    {
      "epoch": 0.14259086672879775,
      "grad_norm": 5.088889122009277,
      "learning_rate": 9.751950170232309e-05,
      "loss": 0.3417,
      "step": 153
    },
    {
      "epoch": 0.1435228331780056,
      "grad_norm": 4.445777893066406,
      "learning_rate": 9.74715149411112e-05,
      "loss": 0.2874,
      "step": 154
    },
    {
      "epoch": 0.14445479962721341,
      "grad_norm": 3.9014739990234375,
      "learning_rate": 9.742308048594665e-05,
      "loss": 0.3016,
      "step": 155
    },
    {
      "epoch": 0.14538676607642126,
      "grad_norm": 2.5364391803741455,
      "learning_rate": 9.73741987936047e-05,
      "loss": 0.1844,
      "step": 156
    },
    {
      "epoch": 0.14631873252562907,
      "grad_norm": 1.6550700664520264,
      "learning_rate": 9.732487032507837e-05,
      "loss": 0.1792,
      "step": 157
    },
    {
      "epoch": 0.14725069897483692,
      "grad_norm": 1.4614055156707764,
      "learning_rate": 9.727509554557417e-05,
      "loss": 0.1203,
      "step": 158
    },
    {
      "epoch": 0.14818266542404473,
      "grad_norm": 1.4385958909988403,
      "learning_rate": 9.722487492450764e-05,
      "loss": 0.1653,
      "step": 159
    },
    {
      "epoch": 0.14911463187325255,
      "grad_norm": 3.6554222106933594,
      "learning_rate": 9.717420893549902e-05,
      "loss": 0.1831,
      "step": 160
    },
    {
      "epoch": 0.1500465983224604,
      "grad_norm": 2.7897613048553467,
      "learning_rate": 9.712309805636863e-05,
      "loss": 0.155,
      "step": 161
    },
    {
      "epoch": 0.1509785647716682,
      "grad_norm": 7.741319179534912,
      "learning_rate": 9.707154276913256e-05,
      "loss": 0.3681,
      "step": 162
    },
    {
      "epoch": 0.15191053122087605,
      "grad_norm": 6.076152801513672,
      "learning_rate": 9.701954355999792e-05,
      "loss": 0.3269,
      "step": 163
    },
    {
      "epoch": 0.15284249767008387,
      "grad_norm": 5.800349712371826,
      "learning_rate": 9.696710091935841e-05,
      "loss": 0.4059,
      "step": 164
    },
    {
      "epoch": 0.15377446411929171,
      "grad_norm": 3.289402484893799,
      "learning_rate": 9.691421534178966e-05,
      "loss": 0.1641,
      "step": 165
    },
    {
      "epoch": 0.15470643056849953,
      "grad_norm": 1.5645993947982788,
      "learning_rate": 9.686088732604449e-05,
      "loss": 0.1461,
      "step": 166
    },
    {
      "epoch": 0.15563839701770738,
      "grad_norm": 1.1692672967910767,
      "learning_rate": 9.68071173750483e-05,
      "loss": 0.0978,
      "step": 167
    },
    {
      "epoch": 0.1565703634669152,
      "grad_norm": 0.9249667525291443,
      "learning_rate": 9.675290599589429e-05,
      "loss": 0.1108,
      "step": 168
    },
    {
      "epoch": 0.157502329916123,
      "grad_norm": 2.5620081424713135,
      "learning_rate": 9.669825369983865e-05,
      "loss": 0.2324,
      "step": 169
    },
    {
      "epoch": 0.15843429636533085,
      "grad_norm": 2.108088254928589,
      "learning_rate": 9.664316100229578e-05,
      "loss": 0.143,
      "step": 170
    },
    {
      "epoch": 0.15936626281453867,
      "grad_norm": 2.3042526245117188,
      "learning_rate": 9.658762842283343e-05,
      "loss": 0.1438,
      "step": 171
    },
    {
      "epoch": 0.1602982292637465,
      "grad_norm": 3.5743720531463623,
      "learning_rate": 9.653165648516776e-05,
      "loss": 0.0584,
      "step": 172
    },
    {
      "epoch": 0.16123019571295433,
      "grad_norm": 5.048624515533447,
      "learning_rate": 9.647524571715843e-05,
      "loss": 0.2915,
      "step": 173
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 2.8932971954345703,
      "learning_rate": 9.641839665080363e-05,
      "loss": 0.1311,
      "step": 174
    },
    {
      "epoch": 0.16309412861137,
      "grad_norm": 3.3463146686553955,
      "learning_rate": 9.636110982223505e-05,
      "loss": 0.1041,
      "step": 175
    },
    {
      "epoch": 0.16402609506057783,
      "grad_norm": 0.9479485154151917,
      "learning_rate": 9.630338577171282e-05,
      "loss": 0.0865,
      "step": 176
    },
    {
      "epoch": 0.16495806150978565,
      "grad_norm": 2.0499370098114014,
      "learning_rate": 9.624522504362039e-05,
      "loss": 0.1282,
      "step": 177
    },
    {
      "epoch": 0.16589002795899346,
      "grad_norm": 2.943985939025879,
      "learning_rate": 9.618662818645948e-05,
      "loss": 0.274,
      "step": 178
    },
    {
      "epoch": 0.1668219944082013,
      "grad_norm": 1.6030714511871338,
      "learning_rate": 9.612759575284483e-05,
      "loss": 0.1708,
      "step": 179
    },
    {
      "epoch": 0.16775396085740912,
      "grad_norm": 1.8890039920806885,
      "learning_rate": 9.606812829949896e-05,
      "loss": 0.2742,
      "step": 180
    },
    {
      "epoch": 0.16868592730661697,
      "grad_norm": 2.0088531970977783,
      "learning_rate": 9.600822638724705e-05,
      "loss": 0.3307,
      "step": 181
    },
    {
      "epoch": 0.16961789375582478,
      "grad_norm": 1.0239542722702026,
      "learning_rate": 9.594789058101153e-05,
      "loss": 0.171,
      "step": 182
    },
    {
      "epoch": 0.17054986020503263,
      "grad_norm": 1.3413492441177368,
      "learning_rate": 9.588712144980681e-05,
      "loss": 0.111,
      "step": 183
    },
    {
      "epoch": 0.17148182665424044,
      "grad_norm": 1.7198429107666016,
      "learning_rate": 9.582591956673387e-05,
      "loss": 0.1318,
      "step": 184
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 2.8803393840789795,
      "learning_rate": 9.576428550897489e-05,
      "loss": 0.1807,
      "step": 185
    },
    {
      "epoch": 0.1733457595526561,
      "grad_norm": 1.4469658136367798,
      "learning_rate": 9.570221985778785e-05,
      "loss": 0.1252,
      "step": 186
    },
    {
      "epoch": 0.17427772600186392,
      "grad_norm": 3.2417330741882324,
      "learning_rate": 9.563972319850092e-05,
      "loss": 0.2001,
      "step": 187
    },
    {
      "epoch": 0.17520969245107176,
      "grad_norm": 1.7459640502929688,
      "learning_rate": 9.557679612050708e-05,
      "loss": 0.1286,
      "step": 188
    },
    {
      "epoch": 0.17614165890027958,
      "grad_norm": 2.3768630027770996,
      "learning_rate": 9.551343921725843e-05,
      "loss": 0.1013,
      "step": 189
    },
    {
      "epoch": 0.17707362534948742,
      "grad_norm": 1.7927972078323364,
      "learning_rate": 9.544965308626074e-05,
      "loss": 0.1465,
      "step": 190
    },
    {
      "epoch": 0.17800559179869524,
      "grad_norm": 1.5043256282806396,
      "learning_rate": 9.538543832906773e-05,
      "loss": 0.0714,
      "step": 191
    },
    {
      "epoch": 0.17893755824790308,
      "grad_norm": 2.0259816646575928,
      "learning_rate": 9.53207955512753e-05,
      "loss": 0.1627,
      "step": 192
    },
    {
      "epoch": 0.1798695246971109,
      "grad_norm": 3.1285605430603027,
      "learning_rate": 9.525572536251607e-05,
      "loss": 0.1725,
      "step": 193
    },
    {
      "epoch": 0.18080149114631874,
      "grad_norm": 2.060331344604492,
      "learning_rate": 9.519022837645337e-05,
      "loss": 0.1713,
      "step": 194
    },
    {
      "epoch": 0.18173345759552656,
      "grad_norm": 2.1062660217285156,
      "learning_rate": 9.512430521077565e-05,
      "loss": 0.123,
      "step": 195
    },
    {
      "epoch": 0.18266542404473438,
      "grad_norm": 2.594912052154541,
      "learning_rate": 9.505795648719048e-05,
      "loss": 0.1772,
      "step": 196
    },
    {
      "epoch": 0.18359739049394222,
      "grad_norm": 5.78408670425415,
      "learning_rate": 9.499118283141887e-05,
      "loss": 0.272,
      "step": 197
    },
    {
      "epoch": 0.18452935694315004,
      "grad_norm": 3.905864715576172,
      "learning_rate": 9.492398487318921e-05,
      "loss": 0.1512,
      "step": 198
    },
    {
      "epoch": 0.18546132339235788,
      "grad_norm": 6.496005058288574,
      "learning_rate": 9.485636324623146e-05,
      "loss": 0.1588,
      "step": 199
    },
    {
      "epoch": 0.1863932898415657,
      "grad_norm": 28.41652488708496,
      "learning_rate": 9.478831858827104e-05,
      "loss": 1.0845,
      "step": 200
    },
    {
      "epoch": 0.18732525629077354,
      "grad_norm": 1.8884375095367432,
      "learning_rate": 9.471985154102292e-05,
      "loss": 0.1585,
      "step": 201
    },
    {
      "epoch": 0.18825722273998136,
      "grad_norm": 3.1585066318511963,
      "learning_rate": 9.465096275018556e-05,
      "loss": 0.2897,
      "step": 202
    },
    {
      "epoch": 0.1891891891891892,
      "grad_norm": 5.4817423820495605,
      "learning_rate": 9.458165286543476e-05,
      "loss": 0.4149,
      "step": 203
    },
    {
      "epoch": 0.19012115563839702,
      "grad_norm": 3.001276969909668,
      "learning_rate": 9.451192254041758e-05,
      "loss": 0.1301,
      "step": 204
    },
    {
      "epoch": 0.19105312208760486,
      "grad_norm": 4.782146453857422,
      "learning_rate": 9.444177243274618e-05,
      "loss": 0.2397,
      "step": 205
    },
    {
      "epoch": 0.19198508853681268,
      "grad_norm": 2.695045232772827,
      "learning_rate": 9.437120320399158e-05,
      "loss": 0.1019,
      "step": 206
    },
    {
      "epoch": 0.1929170549860205,
      "grad_norm": 0.7935613393783569,
      "learning_rate": 9.430021551967744e-05,
      "loss": 0.0681,
      "step": 207
    },
    {
      "epoch": 0.19384902143522834,
      "grad_norm": 1.3984178304672241,
      "learning_rate": 9.422881004927381e-05,
      "loss": 0.1025,
      "step": 208
    },
    {
      "epoch": 0.19478098788443615,
      "grad_norm": 1.7235757112503052,
      "learning_rate": 9.415698746619079e-05,
      "loss": 0.1831,
      "step": 209
    },
    {
      "epoch": 0.195712954333644,
      "grad_norm": 0.9581148028373718,
      "learning_rate": 9.408474844777217e-05,
      "loss": 0.0724,
      "step": 210
    },
    {
      "epoch": 0.1966449207828518,
      "grad_norm": 1.098017692565918,
      "learning_rate": 9.401209367528908e-05,
      "loss": 0.0322,
      "step": 211
    },
    {
      "epoch": 0.19757688723205966,
      "grad_norm": 1.6468275785446167,
      "learning_rate": 9.393902383393347e-05,
      "loss": 0.2123,
      "step": 212
    },
    {
      "epoch": 0.19850885368126747,
      "grad_norm": 2.875089645385742,
      "learning_rate": 9.386553961281179e-05,
      "loss": 0.3043,
      "step": 213
    },
    {
      "epoch": 0.19944082013047532,
      "grad_norm": 1.3007657527923584,
      "learning_rate": 9.379164170493844e-05,
      "loss": 0.0726,
      "step": 214
    },
    {
      "epoch": 0.20037278657968313,
      "grad_norm": 1.3006619215011597,
      "learning_rate": 9.371733080722911e-05,
      "loss": 0.0819,
      "step": 215
    },
    {
      "epoch": 0.20130475302889095,
      "grad_norm": 1.5371683835983276,
      "learning_rate": 9.36426076204944e-05,
      "loss": 0.1184,
      "step": 216
    },
    {
      "epoch": 0.2022367194780988,
      "grad_norm": 2.0229947566986084,
      "learning_rate": 9.35674728494331e-05,
      "loss": 0.1667,
      "step": 217
    },
    {
      "epoch": 0.2031686859273066,
      "grad_norm": 1.4043385982513428,
      "learning_rate": 9.349192720262555e-05,
      "loss": 0.1001,
      "step": 218
    },
    {
      "epoch": 0.20410065237651445,
      "grad_norm": 1.893733263015747,
      "learning_rate": 9.341597139252698e-05,
      "loss": 0.1334,
      "step": 219
    },
    {
      "epoch": 0.20503261882572227,
      "grad_norm": 2.8895277976989746,
      "learning_rate": 9.333960613546079e-05,
      "loss": 0.1791,
      "step": 220
    },
    {
      "epoch": 0.2059645852749301,
      "grad_norm": 6.10243034362793,
      "learning_rate": 9.326283215161178e-05,
      "loss": 0.1588,
      "step": 221
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 1.887194037437439,
      "learning_rate": 9.31856501650194e-05,
      "loss": 0.1906,
      "step": 222
    },
    {
      "epoch": 0.20782851817334577,
      "grad_norm": 3.0247702598571777,
      "learning_rate": 9.310806090357082e-05,
      "loss": 0.2346,
      "step": 223
    },
    {
      "epoch": 0.2087604846225536,
      "grad_norm": 4.188446998596191,
      "learning_rate": 9.303006509899418e-05,
      "loss": 0.1994,
      "step": 224
    },
    {
      "epoch": 0.2096924510717614,
      "grad_norm": 9.785820960998535,
      "learning_rate": 9.29516634868517e-05,
      "loss": 0.2581,
      "step": 225
    },
    {
      "epoch": 0.21062441752096925,
      "grad_norm": 1.3063938617706299,
      "learning_rate": 9.287285680653254e-05,
      "loss": 0.0835,
      "step": 226
    },
    {
      "epoch": 0.21155638397017706,
      "grad_norm": 2.1981046199798584,
      "learning_rate": 9.279364580124614e-05,
      "loss": 0.1189,
      "step": 227
    },
    {
      "epoch": 0.2124883504193849,
      "grad_norm": 3.547961473464966,
      "learning_rate": 9.27140312180149e-05,
      "loss": 0.1961,
      "step": 228
    },
    {
      "epoch": 0.21342031686859272,
      "grad_norm": 6.077332973480225,
      "learning_rate": 9.263401380766739e-05,
      "loss": 0.4945,
      "step": 229
    },
    {
      "epoch": 0.21435228331780057,
      "grad_norm": 4.83152961730957,
      "learning_rate": 9.255359432483105e-05,
      "loss": 0.3202,
      "step": 230
    },
    {
      "epoch": 0.21528424976700838,
      "grad_norm": 4.037288665771484,
      "learning_rate": 9.247277352792533e-05,
      "loss": 0.2901,
      "step": 231
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 1.8266874551773071,
      "learning_rate": 9.239155217915422e-05,
      "loss": 0.0553,
      "step": 232
    },
    {
      "epoch": 0.21714818266542404,
      "grad_norm": 1.9196511507034302,
      "learning_rate": 9.230993104449939e-05,
      "loss": 0.1628,
      "step": 233
    },
    {
      "epoch": 0.21808014911463186,
      "grad_norm": 2.83617901802063,
      "learning_rate": 9.222791089371265e-05,
      "loss": 0.1828,
      "step": 234
    },
    {
      "epoch": 0.2190121155638397,
      "grad_norm": 1.9007632732391357,
      "learning_rate": 9.214549250030899e-05,
      "loss": 0.0901,
      "step": 235
    },
    {
      "epoch": 0.21994408201304752,
      "grad_norm": 5.230377674102783,
      "learning_rate": 9.206267664155907e-05,
      "loss": 0.2777,
      "step": 236
    },
    {
      "epoch": 0.22087604846225536,
      "grad_norm": 2.382657289505005,
      "learning_rate": 9.197946409848194e-05,
      "loss": 0.1518,
      "step": 237
    },
    {
      "epoch": 0.22180801491146318,
      "grad_norm": 3.907731771469116,
      "learning_rate": 9.189585565583778e-05,
      "loss": 0.2514,
      "step": 238
    },
    {
      "epoch": 0.22273998136067102,
      "grad_norm": 6.802032947540283,
      "learning_rate": 9.181185210212034e-05,
      "loss": 0.2172,
      "step": 239
    },
    {
      "epoch": 0.22367194780987884,
      "grad_norm": 2.1859257221221924,
      "learning_rate": 9.172745422954961e-05,
      "loss": 0.1372,
      "step": 240
    },
    {
      "epoch": 0.22460391425908668,
      "grad_norm": 2.6559228897094727,
      "learning_rate": 9.164266283406433e-05,
      "loss": 0.1124,
      "step": 241
    },
    {
      "epoch": 0.2255358807082945,
      "grad_norm": 1.4599460363388062,
      "learning_rate": 9.155747871531444e-05,
      "loss": 0.0548,
      "step": 242
    },
    {
      "epoch": 0.22646784715750232,
      "grad_norm": 2.3699145317077637,
      "learning_rate": 9.14719026766536e-05,
      "loss": 0.1298,
      "step": 243
    },
    {
      "epoch": 0.22739981360671016,
      "grad_norm": 1.8961679935455322,
      "learning_rate": 9.13859355251316e-05,
      "loss": 0.1816,
      "step": 244
    },
    {
      "epoch": 0.22833178005591798,
      "grad_norm": 1.3365075588226318,
      "learning_rate": 9.129957807148666e-05,
      "loss": 0.0788,
      "step": 245
    },
    {
      "epoch": 0.22926374650512582,
      "grad_norm": 2.557935953140259,
      "learning_rate": 9.121283113013795e-05,
      "loss": 0.0993,
      "step": 246
    },
    {
      "epoch": 0.23019571295433364,
      "grad_norm": 2.207437038421631,
      "learning_rate": 9.112569551917774e-05,
      "loss": 0.1826,
      "step": 247
    },
    {
      "epoch": 0.23112767940354148,
      "grad_norm": 1.3832250833511353,
      "learning_rate": 9.103817206036382e-05,
      "loss": 0.0532,
      "step": 248
    },
    {
      "epoch": 0.2320596458527493,
      "grad_norm": 0.8267796039581299,
      "learning_rate": 9.095026157911166e-05,
      "loss": 0.0325,
      "step": 249
    },
    {
      "epoch": 0.23299161230195714,
      "grad_norm": 22.60721206665039,
      "learning_rate": 9.086196490448668e-05,
      "loss": 0.3952,
      "step": 250
    },
    {
      "epoch": 0.23392357875116496,
      "grad_norm": 1.598435640335083,
      "learning_rate": 9.077328286919638e-05,
      "loss": 0.2231,
      "step": 251
    },
    {
      "epoch": 0.23485554520037277,
      "grad_norm": 2.5959858894348145,
      "learning_rate": 9.068421630958254e-05,
      "loss": 0.2467,
      "step": 252
    },
    {
      "epoch": 0.23578751164958062,
      "grad_norm": 1.6885298490524292,
      "learning_rate": 9.059476606561328e-05,
      "loss": 0.1445,
      "step": 253
    },
    {
      "epoch": 0.23671947809878843,
      "grad_norm": 2.016430616378784,
      "learning_rate": 9.050493298087523e-05,
      "loss": 0.2079,
      "step": 254
    },
    {
      "epoch": 0.23765144454799628,
      "grad_norm": 1.5078867673873901,
      "learning_rate": 9.041471790256542e-05,
      "loss": 0.1513,
      "step": 255
    },
    {
      "epoch": 0.2385834109972041,
      "grad_norm": 1.6564357280731201,
      "learning_rate": 9.032412168148345e-05,
      "loss": 0.1056,
      "step": 256
    },
    {
      "epoch": 0.23951537744641194,
      "grad_norm": 1.5224450826644897,
      "learning_rate": 9.02331451720234e-05,
      "loss": 0.1486,
      "step": 257
    },
    {
      "epoch": 0.24044734389561975,
      "grad_norm": 3.1949379444122314,
      "learning_rate": 9.014178923216572e-05,
      "loss": 0.1631,
      "step": 258
    },
    {
      "epoch": 0.2413793103448276,
      "grad_norm": 1.828347086906433,
      "learning_rate": 9.005005472346924e-05,
      "loss": 0.1613,
      "step": 259
    },
    {
      "epoch": 0.2423112767940354,
      "grad_norm": 2.3112988471984863,
      "learning_rate": 8.995794251106295e-05,
      "loss": 0.1969,
      "step": 260
    },
    {
      "epoch": 0.24324324324324326,
      "grad_norm": 2.2977964878082275,
      "learning_rate": 8.986545346363792e-05,
      "loss": 0.1895,
      "step": 261
    },
    {
      "epoch": 0.24417520969245107,
      "grad_norm": 2.2505693435668945,
      "learning_rate": 8.977258845343905e-05,
      "loss": 0.0995,
      "step": 262
    },
    {
      "epoch": 0.2451071761416589,
      "grad_norm": 1.345773458480835,
      "learning_rate": 8.967934835625689e-05,
      "loss": 0.0899,
      "step": 263
    },
    {
      "epoch": 0.24603914259086673,
      "grad_norm": 0.8952029943466187,
      "learning_rate": 8.958573405141932e-05,
      "loss": 0.059,
      "step": 264
    },
    {
      "epoch": 0.24697110904007455,
      "grad_norm": 1.5607550144195557,
      "learning_rate": 8.949174642178332e-05,
      "loss": 0.1651,
      "step": 265
    },
    {
      "epoch": 0.2479030754892824,
      "grad_norm": 0.8190938234329224,
      "learning_rate": 8.939738635372664e-05,
      "loss": 0.0636,
      "step": 266
    },
    {
      "epoch": 0.2488350419384902,
      "grad_norm": 1.721642255783081,
      "learning_rate": 8.930265473713938e-05,
      "loss": 0.0876,
      "step": 267
    },
    {
      "epoch": 0.24976700838769805,
      "grad_norm": 0.5578976273536682,
      "learning_rate": 8.920755246541563e-05,
      "loss": 0.0285,
      "step": 268
    },
    {
      "epoch": 0.2506989748369059,
      "grad_norm": 1.9145925045013428,
      "learning_rate": 8.911208043544513e-05,
      "loss": 0.1488,
      "step": 269
    },
    {
      "epoch": 0.2506989748369059,
      "eval_loss": 0.107957623898983,
      "eval_runtime": 128.7777,
      "eval_samples_per_second": 14.032,
      "eval_steps_per_second": 3.51,
      "step": 269
    },
    {
      "epoch": 0.2516309412861137,
      "grad_norm": 3.0175607204437256,
      "learning_rate": 8.90162395476046e-05,
      "loss": 0.1851,
      "step": 270
    },
    {
      "epoch": 0.25256290773532153,
      "grad_norm": 0.9700628519058228,
      "learning_rate": 8.89200307057495e-05,
      "loss": 0.0514,
      "step": 271
    },
    {
      "epoch": 0.2534948741845294,
      "grad_norm": 1.3989373445510864,
      "learning_rate": 8.882345481720533e-05,
      "loss": 0.0319,
      "step": 272
    },
    {
      "epoch": 0.25442684063373716,
      "grad_norm": 1.5159484148025513,
      "learning_rate": 8.872651279275916e-05,
      "loss": 0.0655,
      "step": 273
    },
    {
      "epoch": 0.255358807082945,
      "grad_norm": 4.176044940948486,
      "learning_rate": 8.862920554665098e-05,
      "loss": 0.3077,
      "step": 274
    },
    {
      "epoch": 0.25629077353215285,
      "grad_norm": 44.685333251953125,
      "learning_rate": 8.853153399656513e-05,
      "loss": 0.4734,
      "step": 275
    },
    {
      "epoch": 0.2572227399813607,
      "grad_norm": 1.0462782382965088,
      "learning_rate": 8.843349906362163e-05,
      "loss": 0.0498,
      "step": 276
    },
    {
      "epoch": 0.2581547064305685,
      "grad_norm": 2.0390098094940186,
      "learning_rate": 8.833510167236747e-05,
      "loss": 0.0617,
      "step": 277
    },
    {
      "epoch": 0.2590866728797763,
      "grad_norm": 2.5294477939605713,
      "learning_rate": 8.823634275076791e-05,
      "loss": 0.1067,
      "step": 278
    },
    {
      "epoch": 0.26001863932898417,
      "grad_norm": 4.861917972564697,
      "learning_rate": 8.813722323019773e-05,
      "loss": 0.2199,
      "step": 279
    },
    {
      "epoch": 0.26095060577819196,
      "grad_norm": 3.0162718296051025,
      "learning_rate": 8.803774404543247e-05,
      "loss": 0.2016,
      "step": 280
    },
    {
      "epoch": 0.2618825722273998,
      "grad_norm": 2.129127264022827,
      "learning_rate": 8.793790613463955e-05,
      "loss": 0.1283,
      "step": 281
    },
    {
      "epoch": 0.26281453867660765,
      "grad_norm": 0.9576233625411987,
      "learning_rate": 8.783771043936948e-05,
      "loss": 0.0752,
      "step": 282
    },
    {
      "epoch": 0.2637465051258155,
      "grad_norm": 2.868595600128174,
      "learning_rate": 8.773715790454695e-05,
      "loss": 0.1673,
      "step": 283
    },
    {
      "epoch": 0.2646784715750233,
      "grad_norm": 1.5952013731002808,
      "learning_rate": 8.763624947846195e-05,
      "loss": 0.114,
      "step": 284
    },
    {
      "epoch": 0.2656104380242311,
      "grad_norm": 2.122833251953125,
      "learning_rate": 8.753498611276079e-05,
      "loss": 0.137,
      "step": 285
    },
    {
      "epoch": 0.26654240447343897,
      "grad_norm": 1.3185533285140991,
      "learning_rate": 8.743336876243712e-05,
      "loss": 0.0877,
      "step": 286
    },
    {
      "epoch": 0.2674743709226468,
      "grad_norm": 1.3917877674102783,
      "learning_rate": 8.7331398385823e-05,
      "loss": 0.1248,
      "step": 287
    },
    {
      "epoch": 0.2684063373718546,
      "grad_norm": 1.9148253202438354,
      "learning_rate": 8.722907594457974e-05,
      "loss": 0.179,
      "step": 288
    },
    {
      "epoch": 0.26933830382106244,
      "grad_norm": 1.8025591373443604,
      "learning_rate": 8.712640240368899e-05,
      "loss": 0.1714,
      "step": 289
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 2.925537109375,
      "learning_rate": 8.702337873144343e-05,
      "loss": 0.2086,
      "step": 290
    },
    {
      "epoch": 0.2712022367194781,
      "grad_norm": 3.2652766704559326,
      "learning_rate": 8.692000589943785e-05,
      "loss": 0.1623,
      "step": 291
    },
    {
      "epoch": 0.2721342031686859,
      "grad_norm": 2.3668527603149414,
      "learning_rate": 8.681628488255986e-05,
      "loss": 0.1309,
      "step": 292
    },
    {
      "epoch": 0.27306616961789376,
      "grad_norm": 3.6899619102478027,
      "learning_rate": 8.671221665898073e-05,
      "loss": 0.2109,
      "step": 293
    },
    {
      "epoch": 0.2739981360671016,
      "grad_norm": 5.232030868530273,
      "learning_rate": 8.660780221014616e-05,
      "loss": 0.1588,
      "step": 294
    },
    {
      "epoch": 0.2749301025163094,
      "grad_norm": 3.900122880935669,
      "learning_rate": 8.650304252076703e-05,
      "loss": 0.1674,
      "step": 295
    },
    {
      "epoch": 0.27586206896551724,
      "grad_norm": 4.121029853820801,
      "learning_rate": 8.639793857881011e-05,
      "loss": 0.1738,
      "step": 296
    },
    {
      "epoch": 0.2767940354147251,
      "grad_norm": 4.410455703735352,
      "learning_rate": 8.629249137548874e-05,
      "loss": 0.1614,
      "step": 297
    },
    {
      "epoch": 0.2777260018639329,
      "grad_norm": 3.216705083847046,
      "learning_rate": 8.618670190525352e-05,
      "loss": 0.147,
      "step": 298
    },
    {
      "epoch": 0.2786579683131407,
      "grad_norm": 3.378570318222046,
      "learning_rate": 8.608057116578283e-05,
      "loss": 0.0914,
      "step": 299
    },
    {
      "epoch": 0.27958993476234856,
      "grad_norm": 7.987005710601807,
      "learning_rate": 8.597410015797358e-05,
      "loss": 0.3069,
      "step": 300
    },
    {
      "epoch": 0.2805219012115564,
      "grad_norm": 0.6978614926338196,
      "learning_rate": 8.586728988593158e-05,
      "loss": 0.0613,
      "step": 301
    },
    {
      "epoch": 0.2814538676607642,
      "grad_norm": 2.7040798664093018,
      "learning_rate": 8.576014135696226e-05,
      "loss": 0.2154,
      "step": 302
    },
    {
      "epoch": 0.28238583410997203,
      "grad_norm": 1.6410375833511353,
      "learning_rate": 8.565265558156101e-05,
      "loss": 0.14,
      "step": 303
    },
    {
      "epoch": 0.2833178005591799,
      "grad_norm": 1.9833245277404785,
      "learning_rate": 8.554483357340379e-05,
      "loss": 0.1751,
      "step": 304
    },
    {
      "epoch": 0.2842497670083877,
      "grad_norm": 1.208056926727295,
      "learning_rate": 8.543667634933742e-05,
      "loss": 0.1013,
      "step": 305
    },
    {
      "epoch": 0.2851817334575955,
      "grad_norm": 1.0126078128814697,
      "learning_rate": 8.532818492937014e-05,
      "loss": 0.0376,
      "step": 306
    },
    {
      "epoch": 0.28611369990680335,
      "grad_norm": 1.7472506761550903,
      "learning_rate": 8.521936033666187e-05,
      "loss": 0.1245,
      "step": 307
    },
    {
      "epoch": 0.2870456663560112,
      "grad_norm": 0.9348568916320801,
      "learning_rate": 8.511020359751466e-05,
      "loss": 0.1106,
      "step": 308
    },
    {
      "epoch": 0.287977632805219,
      "grad_norm": 5.611238479614258,
      "learning_rate": 8.500071574136295e-05,
      "loss": 0.1801,
      "step": 309
    },
    {
      "epoch": 0.28890959925442683,
      "grad_norm": 1.2382830381393433,
      "learning_rate": 8.489089780076386e-05,
      "loss": 0.0415,
      "step": 310
    },
    {
      "epoch": 0.2898415657036347,
      "grad_norm": 0.8619039058685303,
      "learning_rate": 8.478075081138745e-05,
      "loss": 0.0524,
      "step": 311
    },
    {
      "epoch": 0.2907735321528425,
      "grad_norm": 3.3732051849365234,
      "learning_rate": 8.467027581200702e-05,
      "loss": 0.1547,
      "step": 312
    },
    {
      "epoch": 0.2917054986020503,
      "grad_norm": 2.319859027862549,
      "learning_rate": 8.455947384448924e-05,
      "loss": 0.1353,
      "step": 313
    },
    {
      "epoch": 0.29263746505125815,
      "grad_norm": 3.4164938926696777,
      "learning_rate": 8.444834595378434e-05,
      "loss": 0.2439,
      "step": 314
    },
    {
      "epoch": 0.293569431500466,
      "grad_norm": 1.6757917404174805,
      "learning_rate": 8.433689318791626e-05,
      "loss": 0.2057,
      "step": 315
    },
    {
      "epoch": 0.29450139794967384,
      "grad_norm": 2.328280210494995,
      "learning_rate": 8.422511659797279e-05,
      "loss": 0.1578,
      "step": 316
    },
    {
      "epoch": 0.2954333643988816,
      "grad_norm": 2.0828404426574707,
      "learning_rate": 8.411301723809563e-05,
      "loss": 0.12,
      "step": 317
    },
    {
      "epoch": 0.29636533084808947,
      "grad_norm": 1.3504449129104614,
      "learning_rate": 8.400059616547047e-05,
      "loss": 0.0851,
      "step": 318
    },
    {
      "epoch": 0.2972972972972973,
      "grad_norm": 1.7693885564804077,
      "learning_rate": 8.388785444031695e-05,
      "loss": 0.175,
      "step": 319
    },
    {
      "epoch": 0.2982292637465051,
      "grad_norm": 2.22578501701355,
      "learning_rate": 8.377479312587879e-05,
      "loss": 0.1396,
      "step": 320
    },
    {
      "epoch": 0.29916123019571295,
      "grad_norm": 2.099825382232666,
      "learning_rate": 8.366141328841366e-05,
      "loss": 0.148,
      "step": 321
    },
    {
      "epoch": 0.3000931966449208,
      "grad_norm": 2.9881668090820312,
      "learning_rate": 8.354771599718312e-05,
      "loss": 0.197,
      "step": 322
    },
    {
      "epoch": 0.30102516309412863,
      "grad_norm": 4.4058966636657715,
      "learning_rate": 8.343370232444261e-05,
      "loss": 0.1629,
      "step": 323
    },
    {
      "epoch": 0.3019571295433364,
      "grad_norm": 2.186525344848633,
      "learning_rate": 8.331937334543132e-05,
      "loss": 0.062,
      "step": 324
    },
    {
      "epoch": 0.30288909599254427,
      "grad_norm": 20.07304573059082,
      "learning_rate": 8.320473013836196e-05,
      "loss": 0.2938,
      "step": 325
    },
    {
      "epoch": 0.3038210624417521,
      "grad_norm": 0.5368430614471436,
      "learning_rate": 8.308977378441072e-05,
      "loss": 0.046,
      "step": 326
    },
    {
      "epoch": 0.3047530288909599,
      "grad_norm": 1.0717397928237915,
      "learning_rate": 8.297450536770697e-05,
      "loss": 0.0932,
      "step": 327
    },
    {
      "epoch": 0.30568499534016774,
      "grad_norm": 1.006714940071106,
      "learning_rate": 8.28589259753231e-05,
      "loss": 0.0887,
      "step": 328
    },
    {
      "epoch": 0.3066169617893756,
      "grad_norm": 2.4907236099243164,
      "learning_rate": 8.274303669726426e-05,
      "loss": 0.1883,
      "step": 329
    },
    {
      "epoch": 0.30754892823858343,
      "grad_norm": 1.739026665687561,
      "learning_rate": 8.262683862645804e-05,
      "loss": 0.1322,
      "step": 330
    },
    {
      "epoch": 0.3084808946877912,
      "grad_norm": 0.5353955030441284,
      "learning_rate": 8.251033285874421e-05,
      "loss": 0.0189,
      "step": 331
    },
    {
      "epoch": 0.30941286113699906,
      "grad_norm": 1.517392635345459,
      "learning_rate": 8.239352049286435e-05,
      "loss": 0.0856,
      "step": 332
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 1.9752682447433472,
      "learning_rate": 8.22764026304515e-05,
      "loss": 0.1543,
      "step": 333
    },
    {
      "epoch": 0.31127679403541475,
      "grad_norm": 1.9139810800552368,
      "learning_rate": 8.21589803760198e-05,
      "loss": 0.1547,
      "step": 334
    },
    {
      "epoch": 0.31220876048462254,
      "grad_norm": 2.1578001976013184,
      "learning_rate": 8.204125483695403e-05,
      "loss": 0.1611,
      "step": 335
    },
    {
      "epoch": 0.3131407269338304,
      "grad_norm": 1.5856783390045166,
      "learning_rate": 8.192322712349917e-05,
      "loss": 0.0656,
      "step": 336
    },
    {
      "epoch": 0.3140726933830382,
      "grad_norm": 2.2837724685668945,
      "learning_rate": 8.180489834875e-05,
      "loss": 0.1479,
      "step": 337
    },
    {
      "epoch": 0.315004659832246,
      "grad_norm": 2.434117555618286,
      "learning_rate": 8.168626962864045e-05,
      "loss": 0.2402,
      "step": 338
    },
    {
      "epoch": 0.31593662628145386,
      "grad_norm": 1.2002735137939453,
      "learning_rate": 8.156734208193327e-05,
      "loss": 0.0429,
      "step": 339
    },
    {
      "epoch": 0.3168685927306617,
      "grad_norm": 1.260638952255249,
      "learning_rate": 8.144811683020932e-05,
      "loss": 0.0626,
      "step": 340
    },
    {
      "epoch": 0.31780055917986955,
      "grad_norm": 1.155396580696106,
      "learning_rate": 8.132859499785707e-05,
      "loss": 0.0704,
      "step": 341
    },
    {
      "epoch": 0.31873252562907733,
      "grad_norm": 1.6018532514572144,
      "learning_rate": 8.120877771206201e-05,
      "loss": 0.0605,
      "step": 342
    },
    {
      "epoch": 0.3196644920782852,
      "grad_norm": 1.430546522140503,
      "learning_rate": 8.108866610279593e-05,
      "loss": 0.1085,
      "step": 343
    },
    {
      "epoch": 0.320596458527493,
      "grad_norm": 1.5464401245117188,
      "learning_rate": 8.096826130280639e-05,
      "loss": 0.097,
      "step": 344
    },
    {
      "epoch": 0.32152842497670087,
      "grad_norm": 1.4850672483444214,
      "learning_rate": 8.08475644476059e-05,
      "loss": 0.0894,
      "step": 345
    },
    {
      "epoch": 0.32246039142590865,
      "grad_norm": 2.387909412384033,
      "learning_rate": 8.072657667546136e-05,
      "loss": 0.0806,
      "step": 346
    },
    {
      "epoch": 0.3233923578751165,
      "grad_norm": 1.3122079372406006,
      "learning_rate": 8.060529912738315e-05,
      "loss": 0.0819,
      "step": 347
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 1.7305047512054443,
      "learning_rate": 8.048373294711455e-05,
      "loss": 0.0461,
      "step": 348
    },
    {
      "epoch": 0.32525629077353213,
      "grad_norm": 2.7955398559570312,
      "learning_rate": 8.036187928112087e-05,
      "loss": 0.2149,
      "step": 349
    },
    {
      "epoch": 0.32618825722274,
      "grad_norm": 5.184839725494385,
      "learning_rate": 8.023973927857857e-05,
      "loss": 0.1819,
      "step": 350
    },
    {
      "epoch": 0.3271202236719478,
      "grad_norm": 0.8407028317451477,
      "learning_rate": 8.011731409136454e-05,
      "loss": 0.0468,
      "step": 351
    },
    {
      "epoch": 0.32805219012115566,
      "grad_norm": 1.0992075204849243,
      "learning_rate": 7.99946048740452e-05,
      "loss": 0.0529,
      "step": 352
    },
    {
      "epoch": 0.32898415657036345,
      "grad_norm": 1.9874184131622314,
      "learning_rate": 7.987161278386554e-05,
      "loss": 0.1567,
      "step": 353
    },
    {
      "epoch": 0.3299161230195713,
      "grad_norm": 1.0550768375396729,
      "learning_rate": 7.974833898073832e-05,
      "loss": 0.0591,
      "step": 354
    },
    {
      "epoch": 0.33084808946877914,
      "grad_norm": 2.2478857040405273,
      "learning_rate": 7.962478462723307e-05,
      "loss": 0.1689,
      "step": 355
    },
    {
      "epoch": 0.3317800559179869,
      "grad_norm": 0.4727104604244232,
      "learning_rate": 7.950095088856508e-05,
      "loss": 0.0178,
      "step": 356
    },
    {
      "epoch": 0.33271202236719477,
      "grad_norm": 0.5580483078956604,
      "learning_rate": 7.937683893258454e-05,
      "loss": 0.0216,
      "step": 357
    },
    {
      "epoch": 0.3336439888164026,
      "grad_norm": 2.7151288986206055,
      "learning_rate": 7.925244992976538e-05,
      "loss": 0.2424,
      "step": 358
    },
    {
      "epoch": 0.33457595526561046,
      "grad_norm": 1.1772783994674683,
      "learning_rate": 7.912778505319436e-05,
      "loss": 0.0902,
      "step": 359
    },
    {
      "epoch": 0.33550792171481825,
      "grad_norm": 1.3066917657852173,
      "learning_rate": 7.900284547855991e-05,
      "loss": 0.0794,
      "step": 360
    },
    {
      "epoch": 0.3364398881640261,
      "grad_norm": 1.881451964378357,
      "learning_rate": 7.88776323841411e-05,
      "loss": 0.1948,
      "step": 361
    },
    {
      "epoch": 0.33737185461323393,
      "grad_norm": 0.8755617141723633,
      "learning_rate": 7.875214695079647e-05,
      "loss": 0.0362,
      "step": 362
    },
    {
      "epoch": 0.3383038210624418,
      "grad_norm": 1.7075937986373901,
      "learning_rate": 7.862639036195298e-05,
      "loss": 0.0519,
      "step": 363
    },
    {
      "epoch": 0.33923578751164957,
      "grad_norm": 0.9367966651916504,
      "learning_rate": 7.850036380359478e-05,
      "loss": 0.0451,
      "step": 364
    },
    {
      "epoch": 0.3401677539608574,
      "grad_norm": 7.803852558135986,
      "learning_rate": 7.837406846425204e-05,
      "loss": 0.1543,
      "step": 365
    },
    {
      "epoch": 0.34109972041006525,
      "grad_norm": 8.1025972366333,
      "learning_rate": 7.824750553498976e-05,
      "loss": 0.2252,
      "step": 366
    },
    {
      "epoch": 0.34203168685927304,
      "grad_norm": 1.4972351789474487,
      "learning_rate": 7.812067620939652e-05,
      "loss": 0.1005,
      "step": 367
    },
    {
      "epoch": 0.3429636533084809,
      "grad_norm": 3.528541088104248,
      "learning_rate": 7.799358168357323e-05,
      "loss": 0.192,
      "step": 368
    },
    {
      "epoch": 0.34389561975768873,
      "grad_norm": 2.0039918422698975,
      "learning_rate": 7.786622315612183e-05,
      "loss": 0.1503,
      "step": 369
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 1.0715596675872803,
      "learning_rate": 7.773860182813403e-05,
      "loss": 0.041,
      "step": 370
    },
    {
      "epoch": 0.34575955265610436,
      "grad_norm": 2.713217258453369,
      "learning_rate": 7.761071890317994e-05,
      "loss": 0.1438,
      "step": 371
    },
    {
      "epoch": 0.3466915191053122,
      "grad_norm": 1.925072431564331,
      "learning_rate": 7.748257558729677e-05,
      "loss": 0.0374,
      "step": 372
    },
    {
      "epoch": 0.34762348555452005,
      "grad_norm": 2.379302501678467,
      "learning_rate": 7.735417308897736e-05,
      "loss": 0.0951,
      "step": 373
    },
    {
      "epoch": 0.34855545200372784,
      "grad_norm": 0.7751803398132324,
      "learning_rate": 7.72255126191589e-05,
      "loss": 0.0113,
      "step": 374
    },
    {
      "epoch": 0.3494874184529357,
      "grad_norm": 14.013057708740234,
      "learning_rate": 7.709659539121144e-05,
      "loss": 0.2808,
      "step": 375
    },
    {
      "epoch": 0.3504193849021435,
      "grad_norm": 0.7865239977836609,
      "learning_rate": 7.696742262092644e-05,
      "loss": 0.0394,
      "step": 376
    },
    {
      "epoch": 0.35135135135135137,
      "grad_norm": 1.1983258724212646,
      "learning_rate": 7.683799552650534e-05,
      "loss": 0.0412,
      "step": 377
    },
    {
      "epoch": 0.35228331780055916,
      "grad_norm": 0.8813638091087341,
      "learning_rate": 7.67083153285481e-05,
      "loss": 0.0221,
      "step": 378
    },
    {
      "epoch": 0.353215284249767,
      "grad_norm": 1.4917454719543457,
      "learning_rate": 7.657838325004159e-05,
      "loss": 0.068,
      "step": 379
    },
    {
      "epoch": 0.35414725069897485,
      "grad_norm": 3.252878189086914,
      "learning_rate": 7.644820051634812e-05,
      "loss": 0.1152,
      "step": 380
    },
    {
      "epoch": 0.3550792171481827,
      "grad_norm": 1.7754201889038086,
      "learning_rate": 7.63177683551939e-05,
      "loss": 0.0733,
      "step": 381
    },
    {
      "epoch": 0.3560111835973905,
      "grad_norm": 0.9449477791786194,
      "learning_rate": 7.618708799665744e-05,
      "loss": 0.0237,
      "step": 382
    },
    {
      "epoch": 0.3569431500465983,
      "grad_norm": 1.0870510339736938,
      "learning_rate": 7.605616067315793e-05,
      "loss": 0.0547,
      "step": 383
    },
    {
      "epoch": 0.35787511649580617,
      "grad_norm": 1.218623399734497,
      "learning_rate": 7.592498761944362e-05,
      "loss": 0.0661,
      "step": 384
    },
    {
      "epoch": 0.35880708294501396,
      "grad_norm": 1.4543731212615967,
      "learning_rate": 7.579357007258022e-05,
      "loss": 0.1162,
      "step": 385
    },
    {
      "epoch": 0.3597390493942218,
      "grad_norm": 1.6948089599609375,
      "learning_rate": 7.56619092719392e-05,
      "loss": 0.1639,
      "step": 386
    },
    {
      "epoch": 0.36067101584342964,
      "grad_norm": 2.109131336212158,
      "learning_rate": 7.553000645918611e-05,
      "loss": 0.1001,
      "step": 387
    },
    {
      "epoch": 0.3616029822926375,
      "grad_norm": 3.4576566219329834,
      "learning_rate": 7.539786287826885e-05,
      "loss": 0.1819,
      "step": 388
    },
    {
      "epoch": 0.3625349487418453,
      "grad_norm": 2.4241981506347656,
      "learning_rate": 7.526547977540592e-05,
      "loss": 0.0459,
      "step": 389
    },
    {
      "epoch": 0.3634669151910531,
      "grad_norm": 2.08624005317688,
      "learning_rate": 7.51328583990748e-05,
      "loss": 0.1933,
      "step": 390
    },
    {
      "epoch": 0.36439888164026096,
      "grad_norm": 1.3993889093399048,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0555,
      "step": 391
    },
    {
      "epoch": 0.36533084808946875,
      "grad_norm": 0.9735451340675354,
      "learning_rate": 7.486690583114136e-05,
      "loss": 0.0553,
      "step": 392
    },
    {
      "epoch": 0.3662628145386766,
      "grad_norm": 1.6305480003356934,
      "learning_rate": 7.473357714768222e-05,
      "loss": 0.0586,
      "step": 393
    },
    {
      "epoch": 0.36719478098788444,
      "grad_norm": 4.290209770202637,
      "learning_rate": 7.460001520701755e-05,
      "loss": 0.2194,
      "step": 394
    },
    {
      "epoch": 0.3681267474370923,
      "grad_norm": 2.2773101329803467,
      "learning_rate": 7.446622126874218e-05,
      "loss": 0.1027,
      "step": 395
    },
    {
      "epoch": 0.36905871388630007,
      "grad_norm": 0.5855029821395874,
      "learning_rate": 7.433219659463881e-05,
      "loss": 0.0179,
      "step": 396
    },
    {
      "epoch": 0.3699906803355079,
      "grad_norm": 2.2826790809631348,
      "learning_rate": 7.41979424486662e-05,
      "loss": 0.1378,
      "step": 397
    },
    {
      "epoch": 0.37092264678471576,
      "grad_norm": 1.142947793006897,
      "learning_rate": 7.406346009694713e-05,
      "loss": 0.0291,
      "step": 398
    },
    {
      "epoch": 0.3718546132339236,
      "grad_norm": 3.178868532180786,
      "learning_rate": 7.392875080775665e-05,
      "loss": 0.1048,
      "step": 399
    },
    {
      "epoch": 0.3727865796831314,
      "grad_norm": 9.8319091796875,
      "learning_rate": 7.379381585150996e-05,
      "loss": 0.3356,
      "step": 400
    },
    {
      "epoch": 0.37371854613233924,
      "grad_norm": 1.062192440032959,
      "learning_rate": 7.365865650075046e-05,
      "loss": 0.0727,
      "step": 401
    },
    {
      "epoch": 0.3746505125815471,
      "grad_norm": 1.5438944101333618,
      "learning_rate": 7.35232740301378e-05,
      "loss": 0.1193,
      "step": 402
    },
    {
      "epoch": 0.37558247903075487,
      "grad_norm": 1.8247103691101074,
      "learning_rate": 7.338766971643579e-05,
      "loss": 0.1114,
      "step": 403
    },
    {
      "epoch": 0.3765144454799627,
      "grad_norm": 3.635122060775757,
      "learning_rate": 7.325184483850042e-05,
      "loss": 0.1463,
      "step": 404
    },
    {
      "epoch": 0.37744641192917056,
      "grad_norm": 0.604207456111908,
      "learning_rate": 7.311580067726783e-05,
      "loss": 0.0194,
      "step": 405
    },
    {
      "epoch": 0.3783783783783784,
      "grad_norm": 1.354001760482788,
      "learning_rate": 7.297953851574206e-05,
      "loss": 0.0441,
      "step": 406
    },
    {
      "epoch": 0.3793103448275862,
      "grad_norm": 1.9031493663787842,
      "learning_rate": 7.284305963898314e-05,
      "loss": 0.1449,
      "step": 407
    },
    {
      "epoch": 0.38024231127679403,
      "grad_norm": 1.7530853748321533,
      "learning_rate": 7.27063653340949e-05,
      "loss": 0.131,
      "step": 408
    },
    {
      "epoch": 0.3811742777260019,
      "grad_norm": 1.8988293409347534,
      "learning_rate": 7.25694568902128e-05,
      "loss": 0.1702,
      "step": 409
    },
    {
      "epoch": 0.3821062441752097,
      "grad_norm": 1.3123613595962524,
      "learning_rate": 7.243233559849179e-05,
      "loss": 0.0723,
      "step": 410
    },
    {
      "epoch": 0.3830382106244175,
      "grad_norm": 1.0989536046981812,
      "learning_rate": 7.229500275209418e-05,
      "loss": 0.0745,
      "step": 411
    },
    {
      "epoch": 0.38397017707362535,
      "grad_norm": 1.5770695209503174,
      "learning_rate": 7.215745964617737e-05,
      "loss": 0.0333,
      "step": 412
    },
    {
      "epoch": 0.3849021435228332,
      "grad_norm": 1.0897724628448486,
      "learning_rate": 7.201970757788172e-05,
      "loss": 0.038,
      "step": 413
    },
    {
      "epoch": 0.385834109972041,
      "grad_norm": 0.8779085874557495,
      "learning_rate": 7.188174784631824e-05,
      "loss": 0.0531,
      "step": 414
    },
    {
      "epoch": 0.38676607642124883,
      "grad_norm": 1.3717111349105835,
      "learning_rate": 7.174358175255636e-05,
      "loss": 0.1259,
      "step": 415
    },
    {
      "epoch": 0.38769804287045667,
      "grad_norm": 2.16155743598938,
      "learning_rate": 7.160521059961169e-05,
      "loss": 0.0633,
      "step": 416
    },
    {
      "epoch": 0.3886300093196645,
      "grad_norm": 2.235919952392578,
      "learning_rate": 7.146663569243371e-05,
      "loss": 0.1122,
      "step": 417
    },
    {
      "epoch": 0.3895619757688723,
      "grad_norm": 1.4645483493804932,
      "learning_rate": 7.132785833789344e-05,
      "loss": 0.0639,
      "step": 418
    },
    {
      "epoch": 0.39049394221808015,
      "grad_norm": 1.0879848003387451,
      "learning_rate": 7.118887984477116e-05,
      "loss": 0.0231,
      "step": 419
    },
    {
      "epoch": 0.391425908667288,
      "grad_norm": 1.1921757459640503,
      "learning_rate": 7.104970152374405e-05,
      "loss": 0.041,
      "step": 420
    },
    {
      "epoch": 0.3923578751164958,
      "grad_norm": 0.8043162822723389,
      "learning_rate": 7.091032468737381e-05,
      "loss": 0.0211,
      "step": 421
    },
    {
      "epoch": 0.3932898415657036,
      "grad_norm": 1.0248724222183228,
      "learning_rate": 7.077075065009433e-05,
      "loss": 0.0221,
      "step": 422
    },
    {
      "epoch": 0.39422180801491147,
      "grad_norm": 1.9075223207473755,
      "learning_rate": 7.063098072819919e-05,
      "loss": 0.1292,
      "step": 423
    },
    {
      "epoch": 0.3951537744641193,
      "grad_norm": 2.7208824157714844,
      "learning_rate": 7.049101623982937e-05,
      "loss": 0.0734,
      "step": 424
    },
    {
      "epoch": 0.3960857409133271,
      "grad_norm": 101.83853149414062,
      "learning_rate": 7.03508585049608e-05,
      "loss": 1.1532,
      "step": 425
    },
    {
      "epoch": 0.39701770736253494,
      "grad_norm": 0.779060423374176,
      "learning_rate": 7.021050884539177e-05,
      "loss": 0.0463,
      "step": 426
    },
    {
      "epoch": 0.3979496738117428,
      "grad_norm": 1.509827971458435,
      "learning_rate": 7.006996858473069e-05,
      "loss": 0.0837,
      "step": 427
    },
    {
      "epoch": 0.39888164026095063,
      "grad_norm": 1.621672511100769,
      "learning_rate": 6.99292390483834e-05,
      "loss": 0.0827,
      "step": 428
    },
    {
      "epoch": 0.3998136067101584,
      "grad_norm": 0.5434191823005676,
      "learning_rate": 6.97883215635408e-05,
      "loss": 0.0147,
      "step": 429
    },
    {
      "epoch": 0.40074557315936626,
      "grad_norm": 1.5291364192962646,
      "learning_rate": 6.964721745916631e-05,
      "loss": 0.0406,
      "step": 430
    },
    {
      "epoch": 0.4016775396085741,
      "grad_norm": 0.9197437167167664,
      "learning_rate": 6.950592806598327e-05,
      "loss": 0.0326,
      "step": 431
    },
    {
      "epoch": 0.4026095060577819,
      "grad_norm": 0.8559898138046265,
      "learning_rate": 6.936445471646249e-05,
      "loss": 0.0433,
      "step": 432
    },
    {
      "epoch": 0.40354147250698974,
      "grad_norm": 1.2630029916763306,
      "learning_rate": 6.922279874480959e-05,
      "loss": 0.0317,
      "step": 433
    },
    {
      "epoch": 0.4044734389561976,
      "grad_norm": 1.6737178564071655,
      "learning_rate": 6.908096148695251e-05,
      "loss": 0.0879,
      "step": 434
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 1.7928637266159058,
      "learning_rate": 6.89389442805288e-05,
      "loss": 0.1492,
      "step": 435
    },
    {
      "epoch": 0.4063373718546132,
      "grad_norm": 2.602510452270508,
      "learning_rate": 6.879674846487314e-05,
      "loss": 0.1599,
      "step": 436
    },
    {
      "epoch": 0.40726933830382106,
      "grad_norm": 1.5879359245300293,
      "learning_rate": 6.865437538100457e-05,
      "loss": 0.1823,
      "step": 437
    },
    {
      "epoch": 0.4082013047530289,
      "grad_norm": 0.5362139940261841,
      "learning_rate": 6.851182637161396e-05,
      "loss": 0.0136,
      "step": 438
    },
    {
      "epoch": 0.4091332712022367,
      "grad_norm": 1.3532445430755615,
      "learning_rate": 6.836910278105124e-05,
      "loss": 0.0812,
      "step": 439
    },
    {
      "epoch": 0.41006523765144454,
      "grad_norm": 1.3916680812835693,
      "learning_rate": 6.822620595531286e-05,
      "loss": 0.033,
      "step": 440
    },
    {
      "epoch": 0.4109972041006524,
      "grad_norm": 1.6064777374267578,
      "learning_rate": 6.808313724202894e-05,
      "loss": 0.1176,
      "step": 441
    },
    {
      "epoch": 0.4119291705498602,
      "grad_norm": 1.1576989889144897,
      "learning_rate": 6.793989799045066e-05,
      "loss": 0.0485,
      "step": 442
    },
    {
      "epoch": 0.412861136999068,
      "grad_norm": 1.1500450372695923,
      "learning_rate": 6.779648955143754e-05,
      "loss": 0.0453,
      "step": 443
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 1.5548428297042847,
      "learning_rate": 6.765291327744463e-05,
      "loss": 0.0538,
      "step": 444
    },
    {
      "epoch": 0.4147250698974837,
      "grad_norm": 1.0851874351501465,
      "learning_rate": 6.750917052250982e-05,
      "loss": 0.0226,
      "step": 445
    },
    {
      "epoch": 0.41565703634669154,
      "grad_norm": 0.5055509209632874,
      "learning_rate": 6.736526264224101e-05,
      "loss": 0.0136,
      "step": 446
    },
    {
      "epoch": 0.41658900279589933,
      "grad_norm": 3.6784214973449707,
      "learning_rate": 6.722119099380345e-05,
      "loss": 0.216,
      "step": 447
    },
    {
      "epoch": 0.4175209692451072,
      "grad_norm": 2.3906421661376953,
      "learning_rate": 6.707695693590676e-05,
      "loss": 0.1041,
      "step": 448
    },
    {
      "epoch": 0.418452935694315,
      "grad_norm": 2.9953436851501465,
      "learning_rate": 6.693256182879225e-05,
      "loss": 0.1169,
      "step": 449
    },
    {
      "epoch": 0.4193849021435228,
      "grad_norm": 8.506790161132812,
      "learning_rate": 6.678800703422003e-05,
      "loss": 0.2528,
      "step": 450
    },
    {
      "epoch": 0.42031686859273065,
      "grad_norm": 1.6189558506011963,
      "learning_rate": 6.664329391545626e-05,
      "loss": 0.0498,
      "step": 451
    },
    {
      "epoch": 0.4212488350419385,
      "grad_norm": 0.789339005947113,
      "learning_rate": 6.64984238372601e-05,
      "loss": 0.0302,
      "step": 452
    },
    {
      "epoch": 0.42218080149114634,
      "grad_norm": 2.2830944061279297,
      "learning_rate": 6.635339816587109e-05,
      "loss": 0.0388,
      "step": 453
    },
    {
      "epoch": 0.42311276794035413,
      "grad_norm": 1.3299132585525513,
      "learning_rate": 6.620821826899606e-05,
      "loss": 0.0892,
      "step": 454
    },
    {
      "epoch": 0.424044734389562,
      "grad_norm": 1.234998345375061,
      "learning_rate": 6.606288551579629e-05,
      "loss": 0.0744,
      "step": 455
    },
    {
      "epoch": 0.4249767008387698,
      "grad_norm": 1.1684775352478027,
      "learning_rate": 6.59174012768747e-05,
      "loss": 0.0461,
      "step": 456
    },
    {
      "epoch": 0.42590866728797766,
      "grad_norm": 0.12061901390552521,
      "learning_rate": 6.577176692426279e-05,
      "loss": 0.0041,
      "step": 457
    },
    {
      "epoch": 0.42684063373718545,
      "grad_norm": 1.531051754951477,
      "learning_rate": 6.562598383140772e-05,
      "loss": 0.093,
      "step": 458
    },
    {
      "epoch": 0.4277726001863933,
      "grad_norm": 0.8624515533447266,
      "learning_rate": 6.548005337315944e-05,
      "loss": 0.0424,
      "step": 459
    },
    {
      "epoch": 0.42870456663560114,
      "grad_norm": 0.7969793081283569,
      "learning_rate": 6.533397692575766e-05,
      "loss": 0.0549,
      "step": 460
    },
    {
      "epoch": 0.4296365330848089,
      "grad_norm": 0.7530978918075562,
      "learning_rate": 6.518775586681887e-05,
      "loss": 0.0508,
      "step": 461
    },
    {
      "epoch": 0.43056849953401677,
      "grad_norm": 0.8213361501693726,
      "learning_rate": 6.504139157532337e-05,
      "loss": 0.0552,
      "step": 462
    },
    {
      "epoch": 0.4315004659832246,
      "grad_norm": 1.5611070394515991,
      "learning_rate": 6.489488543160224e-05,
      "loss": 0.0992,
      "step": 463
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 0.9949514269828796,
      "learning_rate": 6.47482388173244e-05,
      "loss": 0.0819,
      "step": 464
    },
    {
      "epoch": 0.43336439888164024,
      "grad_norm": 1.4731316566467285,
      "learning_rate": 6.460145311548341e-05,
      "loss": 0.0341,
      "step": 465
    },
    {
      "epoch": 0.4342963653308481,
      "grad_norm": 1.581841230392456,
      "learning_rate": 6.445452971038464e-05,
      "loss": 0.0797,
      "step": 466
    },
    {
      "epoch": 0.43522833178005593,
      "grad_norm": 0.46201473474502563,
      "learning_rate": 6.430746998763204e-05,
      "loss": 0.0174,
      "step": 467
    },
    {
      "epoch": 0.4361602982292637,
      "grad_norm": 0.8862415552139282,
      "learning_rate": 6.416027533411519e-05,
      "loss": 0.0413,
      "step": 468
    },
    {
      "epoch": 0.43709226467847156,
      "grad_norm": 1.4375804662704468,
      "learning_rate": 6.401294713799618e-05,
      "loss": 0.0863,
      "step": 469
    },
    {
      "epoch": 0.4380242311276794,
      "grad_norm": 2.5881032943725586,
      "learning_rate": 6.386548678869644e-05,
      "loss": 0.0682,
      "step": 470
    },
    {
      "epoch": 0.43895619757688725,
      "grad_norm": 1.366281509399414,
      "learning_rate": 6.37178956768838e-05,
      "loss": 0.0385,
      "step": 471
    },
    {
      "epoch": 0.43988816402609504,
      "grad_norm": 2.332908868789673,
      "learning_rate": 6.35701751944592e-05,
      "loss": 0.0655,
      "step": 472
    },
    {
      "epoch": 0.4408201304753029,
      "grad_norm": 0.8943026661872864,
      "learning_rate": 6.342232673454371e-05,
      "loss": 0.0217,
      "step": 473
    },
    {
      "epoch": 0.44175209692451073,
      "grad_norm": 0.4061681628227234,
      "learning_rate": 6.32743516914653e-05,
      "loss": 0.0063,
      "step": 474
    },
    {
      "epoch": 0.4426840633737186,
      "grad_norm": 2.605149984359741,
      "learning_rate": 6.312625146074575e-05,
      "loss": 0.0873,
      "step": 475
    },
    {
      "epoch": 0.44361602982292636,
      "grad_norm": 0.26587679982185364,
      "learning_rate": 6.297802743908741e-05,
      "loss": 0.0062,
      "step": 476
    },
    {
      "epoch": 0.4445479962721342,
      "grad_norm": 1.5406039953231812,
      "learning_rate": 6.282968102436009e-05,
      "loss": 0.0739,
      "step": 477
    },
    {
      "epoch": 0.44547996272134205,
      "grad_norm": 1.777808666229248,
      "learning_rate": 6.268121361558792e-05,
      "loss": 0.0769,
      "step": 478
    },
    {
      "epoch": 0.44641192917054984,
      "grad_norm": 0.9425985217094421,
      "learning_rate": 6.253262661293604e-05,
      "loss": 0.0383,
      "step": 479
    },
    {
      "epoch": 0.4473438956197577,
      "grad_norm": 1.4720977544784546,
      "learning_rate": 6.238392141769743e-05,
      "loss": 0.0769,
      "step": 480
    },
    {
      "epoch": 0.4482758620689655,
      "grad_norm": 0.7975260019302368,
      "learning_rate": 6.22350994322798e-05,
      "loss": 0.0518,
      "step": 481
    },
    {
      "epoch": 0.44920782851817337,
      "grad_norm": 1.155370831489563,
      "learning_rate": 6.208616206019224e-05,
      "loss": 0.0326,
      "step": 482
    },
    {
      "epoch": 0.45013979496738116,
      "grad_norm": 1.9402527809143066,
      "learning_rate": 6.193711070603202e-05,
      "loss": 0.1378,
      "step": 483
    },
    {
      "epoch": 0.451071761416589,
      "grad_norm": 0.9330565929412842,
      "learning_rate": 6.178794677547137e-05,
      "loss": 0.0259,
      "step": 484
    },
    {
      "epoch": 0.45200372786579684,
      "grad_norm": 6.712258338928223,
      "learning_rate": 6.163867167524419e-05,
      "loss": 0.0925,
      "step": 485
    },
    {
      "epoch": 0.45293569431500463,
      "grad_norm": 0.7303646206855774,
      "learning_rate": 6.14892868131328e-05,
      "loss": 0.0296,
      "step": 486
    },
    {
      "epoch": 0.4538676607642125,
      "grad_norm": 1.3826953172683716,
      "learning_rate": 6.133979359795468e-05,
      "loss": 0.1587,
      "step": 487
    },
    {
      "epoch": 0.4547996272134203,
      "grad_norm": 1.1449270248413086,
      "learning_rate": 6.119019343954914e-05,
      "loss": 0.0431,
      "step": 488
    },
    {
      "epoch": 0.45573159366262816,
      "grad_norm": 1.846737027168274,
      "learning_rate": 6.104048774876406e-05,
      "loss": 0.0455,
      "step": 489
    },
    {
      "epoch": 0.45666356011183595,
      "grad_norm": 1.589819073677063,
      "learning_rate": 6.0890677937442574e-05,
      "loss": 0.066,
      "step": 490
    },
    {
      "epoch": 0.4575955265610438,
      "grad_norm": 2.8043324947357178,
      "learning_rate": 6.074076541840977e-05,
      "loss": 0.1297,
      "step": 491
    },
    {
      "epoch": 0.45852749301025164,
      "grad_norm": 1.2121779918670654,
      "learning_rate": 6.059075160545933e-05,
      "loss": 0.0915,
      "step": 492
    },
    {
      "epoch": 0.4594594594594595,
      "grad_norm": 1.4770679473876953,
      "learning_rate": 6.0440637913340224e-05,
      "loss": 0.0892,
      "step": 493
    },
    {
      "epoch": 0.4603914259086673,
      "grad_norm": 2.293391227722168,
      "learning_rate": 6.029042575774334e-05,
      "loss": 0.0915,
      "step": 494
    },
    {
      "epoch": 0.4613233923578751,
      "grad_norm": 2.9511337280273438,
      "learning_rate": 6.0140116555288204e-05,
      "loss": 0.0979,
      "step": 495
    },
    {
      "epoch": 0.46225535880708296,
      "grad_norm": 1.325278639793396,
      "learning_rate": 5.998971172350953e-05,
      "loss": 0.0623,
      "step": 496
    },
    {
      "epoch": 0.46318732525629075,
      "grad_norm": 1.2071905136108398,
      "learning_rate": 5.9839212680843925e-05,
      "loss": 0.0436,
      "step": 497
    },
    {
      "epoch": 0.4641192917054986,
      "grad_norm": 1.2752379179000854,
      "learning_rate": 5.968862084661643e-05,
      "loss": 0.0375,
      "step": 498
    },
    {
      "epoch": 0.46505125815470644,
      "grad_norm": 2.3906185626983643,
      "learning_rate": 5.953793764102722e-05,
      "loss": 0.0396,
      "step": 499
    },
    {
      "epoch": 0.4659832246039143,
      "grad_norm": 2.577181339263916,
      "learning_rate": 5.938716448513818e-05,
      "loss": 0.1317,
      "step": 500
    },
    {
      "epoch": 0.46691519105312207,
      "grad_norm": 0.3229764699935913,
      "learning_rate": 5.923630280085948e-05,
      "loss": 0.0078,
      "step": 501
    },
    {
      "epoch": 0.4678471575023299,
      "grad_norm": 1.5205317735671997,
      "learning_rate": 5.908535401093618e-05,
      "loss": 0.0789,
      "step": 502
    },
    {
      "epoch": 0.46877912395153776,
      "grad_norm": 0.8096385598182678,
      "learning_rate": 5.8934319538934824e-05,
      "loss": 0.0274,
      "step": 503
    },
    {
      "epoch": 0.46971109040074555,
      "grad_norm": 1.7681595087051392,
      "learning_rate": 5.878320080923001e-05,
      "loss": 0.0511,
      "step": 504
    },
    {
      "epoch": 0.4706430568499534,
      "grad_norm": 2.489217519760132,
      "learning_rate": 5.863199924699095e-05,
      "loss": 0.176,
      "step": 505
    },
    {
      "epoch": 0.47157502329916123,
      "grad_norm": 1.064134120941162,
      "learning_rate": 5.848071627816803e-05,
      "loss": 0.0863,
      "step": 506
    },
    {
      "epoch": 0.4725069897483691,
      "grad_norm": 1.0623358488082886,
      "learning_rate": 5.8329353329479366e-05,
      "loss": 0.0437,
      "step": 507
    },
    {
      "epoch": 0.47343895619757687,
      "grad_norm": 0.8172505497932434,
      "learning_rate": 5.8177911828397336e-05,
      "loss": 0.0331,
      "step": 508
    },
    {
      "epoch": 0.4743709226467847,
      "grad_norm": 0.7944416999816895,
      "learning_rate": 5.802639320313514e-05,
      "loss": 0.0435,
      "step": 509
    },
    {
      "epoch": 0.47530288909599255,
      "grad_norm": 0.8773410320281982,
      "learning_rate": 5.787479888263333e-05,
      "loss": 0.0378,
      "step": 510
    },
    {
      "epoch": 0.4762348555452004,
      "grad_norm": 0.8762723207473755,
      "learning_rate": 5.77231302965463e-05,
      "loss": 0.0607,
      "step": 511
    },
    {
      "epoch": 0.4771668219944082,
      "grad_norm": 0.24209994077682495,
      "learning_rate": 5.757138887522884e-05,
      "loss": 0.0059,
      "step": 512
    },
    {
      "epoch": 0.47809878844361603,
      "grad_norm": 4.610219478607178,
      "learning_rate": 5.7419576049722646e-05,
      "loss": 0.0349,
      "step": 513
    },
    {
      "epoch": 0.4790307548928239,
      "grad_norm": 0.8906161785125732,
      "learning_rate": 5.7267693251742794e-05,
      "loss": 0.0262,
      "step": 514
    },
    {
      "epoch": 0.47996272134203166,
      "grad_norm": 1.6799854040145874,
      "learning_rate": 5.7115741913664264e-05,
      "loss": 0.107,
      "step": 515
    },
    {
      "epoch": 0.4808946877912395,
      "grad_norm": 2.382206678390503,
      "learning_rate": 5.696372346850841e-05,
      "loss": 0.0567,
      "step": 516
    },
    {
      "epoch": 0.48182665424044735,
      "grad_norm": 1.7199441194534302,
      "learning_rate": 5.68116393499295e-05,
      "loss": 0.0622,
      "step": 517
    },
    {
      "epoch": 0.4827586206896552,
      "grad_norm": 0.9822327494621277,
      "learning_rate": 5.6659490992201094e-05,
      "loss": 0.025,
      "step": 518
    },
    {
      "epoch": 0.483690587138863,
      "grad_norm": 1.5693819522857666,
      "learning_rate": 5.650727983020262e-05,
      "loss": 0.2062,
      "step": 519
    },
    {
      "epoch": 0.4846225535880708,
      "grad_norm": 1.9711823463439941,
      "learning_rate": 5.635500729940577e-05,
      "loss": 0.0682,
      "step": 520
    },
    {
      "epoch": 0.48555452003727867,
      "grad_norm": 1.031491994857788,
      "learning_rate": 5.620267483586105e-05,
      "loss": 0.0203,
      "step": 521
    },
    {
      "epoch": 0.4864864864864865,
      "grad_norm": 1.9086527824401855,
      "learning_rate": 5.6050283876184116e-05,
      "loss": 0.0701,
      "step": 522
    },
    {
      "epoch": 0.4874184529356943,
      "grad_norm": 0.8771370053291321,
      "learning_rate": 5.5897835857542317e-05,
      "loss": 0.0241,
      "step": 523
    },
    {
      "epoch": 0.48835041938490215,
      "grad_norm": 2.242940664291382,
      "learning_rate": 5.574533221764109e-05,
      "loss": 0.0484,
      "step": 524
    },
    {
      "epoch": 0.48928238583411,
      "grad_norm": 23.281734466552734,
      "learning_rate": 5.5592774394710465e-05,
      "loss": 0.23,
      "step": 525
    },
    {
      "epoch": 0.4902143522833178,
      "grad_norm": 1.172943115234375,
      "learning_rate": 5.5440163827491454e-05,
      "loss": 0.0466,
      "step": 526
    },
    {
      "epoch": 0.4911463187325256,
      "grad_norm": 0.3696756958961487,
      "learning_rate": 5.5287501955222444e-05,
      "loss": 0.011,
      "step": 527
    },
    {
      "epoch": 0.49207828518173347,
      "grad_norm": 1.8396764993667603,
      "learning_rate": 5.513479021762572e-05,
      "loss": 0.1616,
      "step": 528
    },
    {
      "epoch": 0.4930102516309413,
      "grad_norm": 1.1849040985107422,
      "learning_rate": 5.4982030054893785e-05,
      "loss": 0.0695,
      "step": 529
    },
    {
      "epoch": 0.4939422180801491,
      "grad_norm": 1.8092740774154663,
      "learning_rate": 5.482922290767589e-05,
      "loss": 0.0366,
      "step": 530
    },
    {
      "epoch": 0.49487418452935694,
      "grad_norm": 0.546469509601593,
      "learning_rate": 5.467637021706438e-05,
      "loss": 0.0171,
      "step": 531
    },
    {
      "epoch": 0.4958061509785648,
      "grad_norm": 1.3984041213989258,
      "learning_rate": 5.452347342458104e-05,
      "loss": 0.0303,
      "step": 532
    },
    {
      "epoch": 0.4967381174277726,
      "grad_norm": 0.39972129464149475,
      "learning_rate": 5.437053397216364e-05,
      "loss": 0.0101,
      "step": 533
    },
    {
      "epoch": 0.4976700838769804,
      "grad_norm": 0.5799165964126587,
      "learning_rate": 5.4217553302152237e-05,
      "loss": 0.0173,
      "step": 534
    },
    {
      "epoch": 0.49860205032618826,
      "grad_norm": 1.4319212436676025,
      "learning_rate": 5.406453285727564e-05,
      "loss": 0.0405,
      "step": 535
    },
    {
      "epoch": 0.4995340167753961,
      "grad_norm": 0.3638681173324585,
      "learning_rate": 5.39114740806377e-05,
      "loss": 0.0115,
      "step": 536
    },
    {
      "epoch": 0.5004659832246039,
      "grad_norm": 0.8622692227363586,
      "learning_rate": 5.3758378415703825e-05,
      "loss": 0.0348,
      "step": 537
    },
    {
      "epoch": 0.5013979496738118,
      "grad_norm": 0.7463175654411316,
      "learning_rate": 5.360524730628728e-05,
      "loss": 0.0238,
      "step": 538
    },
    {
      "epoch": 0.5013979496738118,
      "eval_loss": 0.0520755872130394,
      "eval_runtime": 126.0622,
      "eval_samples_per_second": 14.334,
      "eval_steps_per_second": 3.586,
      "step": 538
    },
    {
      "epoch": 0.5023299161230196,
      "grad_norm": 0.33791813254356384,
      "learning_rate": 5.345208219653561e-05,
      "loss": 0.0073,
      "step": 539
    },
    {
      "epoch": 0.5032618825722274,
      "grad_norm": 0.7025349140167236,
      "learning_rate": 5.3298884530917014e-05,
      "loss": 0.0179,
      "step": 540
    },
    {
      "epoch": 0.5041938490214353,
      "grad_norm": 6.270991325378418,
      "learning_rate": 5.3145655754206714e-05,
      "loss": 0.1438,
      "step": 541
    },
    {
      "epoch": 0.5051258154706431,
      "grad_norm": 0.582237184047699,
      "learning_rate": 5.2992397311473316e-05,
      "loss": 0.0119,
      "step": 542
    },
    {
      "epoch": 0.5060577819198508,
      "grad_norm": 0.25732892751693726,
      "learning_rate": 5.2839110648065214e-05,
      "loss": 0.0065,
      "step": 543
    },
    {
      "epoch": 0.5069897483690587,
      "grad_norm": 4.037124156951904,
      "learning_rate": 5.2685797209596976e-05,
      "loss": 0.0951,
      "step": 544
    },
    {
      "epoch": 0.5079217148182665,
      "grad_norm": 1.8424044847488403,
      "learning_rate": 5.2532458441935636e-05,
      "loss": 0.0928,
      "step": 545
    },
    {
      "epoch": 0.5088536812674743,
      "grad_norm": 2.209555149078369,
      "learning_rate": 5.2379095791187124e-05,
      "loss": 0.1429,
      "step": 546
    },
    {
      "epoch": 0.5097856477166822,
      "grad_norm": 0.6223617792129517,
      "learning_rate": 5.222571070368257e-05,
      "loss": 0.0104,
      "step": 547
    },
    {
      "epoch": 0.51071761416589,
      "grad_norm": 1.578021764755249,
      "learning_rate": 5.2072304625964785e-05,
      "loss": 0.0166,
      "step": 548
    },
    {
      "epoch": 0.5116495806150979,
      "grad_norm": 7.880814552307129,
      "learning_rate": 5.191887900477443e-05,
      "loss": 0.1465,
      "step": 549
    },
    {
      "epoch": 0.5125815470643057,
      "grad_norm": 13.242683410644531,
      "learning_rate": 5.176543528703657e-05,
      "loss": 0.3455,
      "step": 550
    },
    {
      "epoch": 0.5135135135135135,
      "grad_norm": 0.05634134262800217,
      "learning_rate": 5.161197491984684e-05,
      "loss": 0.002,
      "step": 551
    },
    {
      "epoch": 0.5144454799627214,
      "grad_norm": 0.34363609552383423,
      "learning_rate": 5.1458499350458e-05,
      "loss": 0.0064,
      "step": 552
    },
    {
      "epoch": 0.5153774464119292,
      "grad_norm": 0.25330761075019836,
      "learning_rate": 5.130501002626609e-05,
      "loss": 0.0046,
      "step": 553
    },
    {
      "epoch": 0.516309412861137,
      "grad_norm": 1.2197694778442383,
      "learning_rate": 5.11515083947969e-05,
      "loss": 0.0342,
      "step": 554
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 3.949093818664551,
      "learning_rate": 5.099799590369231e-05,
      "loss": 0.0867,
      "step": 555
    },
    {
      "epoch": 0.5181733457595527,
      "grad_norm": 0.16666558384895325,
      "learning_rate": 5.084447400069655e-05,
      "loss": 0.0043,
      "step": 556
    },
    {
      "epoch": 0.5191053122087604,
      "grad_norm": 1.1512700319290161,
      "learning_rate": 5.069094413364272e-05,
      "loss": 0.0468,
      "step": 557
    },
    {
      "epoch": 0.5200372786579683,
      "grad_norm": 0.9841926097869873,
      "learning_rate": 5.053740775043892e-05,
      "loss": 0.0752,
      "step": 558
    },
    {
      "epoch": 0.5209692451071761,
      "grad_norm": 1.8312132358551025,
      "learning_rate": 5.038386629905475e-05,
      "loss": 0.0422,
      "step": 559
    },
    {
      "epoch": 0.5219012115563839,
      "grad_norm": 2.9533884525299072,
      "learning_rate": 5.023032122750759e-05,
      "loss": 0.1588,
      "step": 560
    },
    {
      "epoch": 0.5228331780055918,
      "grad_norm": 0.0878317654132843,
      "learning_rate": 5.007677398384901e-05,
      "loss": 0.002,
      "step": 561
    },
    {
      "epoch": 0.5237651444547996,
      "grad_norm": 0.921676516532898,
      "learning_rate": 4.9923226016151e-05,
      "loss": 0.0214,
      "step": 562
    },
    {
      "epoch": 0.5246971109040075,
      "grad_norm": 5.561086177825928,
      "learning_rate": 4.976967877249242e-05,
      "loss": 0.1318,
      "step": 563
    },
    {
      "epoch": 0.5256290773532153,
      "grad_norm": 0.7021181583404541,
      "learning_rate": 4.961613370094527e-05,
      "loss": 0.0727,
      "step": 564
    },
    {
      "epoch": 0.5265610438024231,
      "grad_norm": 1.6535977125167847,
      "learning_rate": 4.9462592249561095e-05,
      "loss": 0.0668,
      "step": 565
    },
    {
      "epoch": 0.527493010251631,
      "grad_norm": 1.1922616958618164,
      "learning_rate": 4.9309055866357295e-05,
      "loss": 0.0224,
      "step": 566
    },
    {
      "epoch": 0.5284249767008388,
      "grad_norm": 0.9673576354980469,
      "learning_rate": 4.915552599930345e-05,
      "loss": 0.0291,
      "step": 567
    },
    {
      "epoch": 0.5293569431500466,
      "grad_norm": 1.8743202686309814,
      "learning_rate": 4.900200409630771e-05,
      "loss": 0.0289,
      "step": 568
    },
    {
      "epoch": 0.5302889095992545,
      "grad_norm": 1.7686008214950562,
      "learning_rate": 4.884849160520311e-05,
      "loss": 0.0814,
      "step": 569
    },
    {
      "epoch": 0.5312208760484622,
      "grad_norm": 0.39702948927879333,
      "learning_rate": 4.869498997373393e-05,
      "loss": 0.0115,
      "step": 570
    },
    {
      "epoch": 0.53215284249767,
      "grad_norm": 0.8344688415527344,
      "learning_rate": 4.8541500649542014e-05,
      "loss": 0.0138,
      "step": 571
    },
    {
      "epoch": 0.5330848089468779,
      "grad_norm": 1.8823225498199463,
      "learning_rate": 4.838802508015316e-05,
      "loss": 0.0605,
      "step": 572
    },
    {
      "epoch": 0.5340167753960857,
      "grad_norm": 1.7391656637191772,
      "learning_rate": 4.823456471296344e-05,
      "loss": 0.0357,
      "step": 573
    },
    {
      "epoch": 0.5349487418452936,
      "grad_norm": 2.7030084133148193,
      "learning_rate": 4.8081120995225574e-05,
      "loss": 0.0893,
      "step": 574
    },
    {
      "epoch": 0.5358807082945014,
      "grad_norm": 5.969890594482422,
      "learning_rate": 4.792769537403523e-05,
      "loss": 0.347,
      "step": 575
    },
    {
      "epoch": 0.5368126747437092,
      "grad_norm": 0.7905582785606384,
      "learning_rate": 4.777428929631743e-05,
      "loss": 0.0267,
      "step": 576
    },
    {
      "epoch": 0.5377446411929171,
      "grad_norm": 1.111538290977478,
      "learning_rate": 4.762090420881289e-05,
      "loss": 0.0523,
      "step": 577
    },
    {
      "epoch": 0.5386766076421249,
      "grad_norm": 1.5851444005966187,
      "learning_rate": 4.746754155806437e-05,
      "loss": 0.1028,
      "step": 578
    },
    {
      "epoch": 0.5396085740913327,
      "grad_norm": 1.3687413930892944,
      "learning_rate": 4.731420279040303e-05,
      "loss": 0.0795,
      "step": 579
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 2.08109712600708,
      "learning_rate": 4.716088935193479e-05,
      "loss": 0.0377,
      "step": 580
    },
    {
      "epoch": 0.5414725069897484,
      "grad_norm": 1.7671325206756592,
      "learning_rate": 4.7007602688526695e-05,
      "loss": 0.0431,
      "step": 581
    },
    {
      "epoch": 0.5424044734389561,
      "grad_norm": 2.354390859603882,
      "learning_rate": 4.6854344245793305e-05,
      "loss": 0.1034,
      "step": 582
    },
    {
      "epoch": 0.543336439888164,
      "grad_norm": 1.8093174695968628,
      "learning_rate": 4.6701115469083e-05,
      "loss": 0.1356,
      "step": 583
    },
    {
      "epoch": 0.5442684063373718,
      "grad_norm": 3.422104597091675,
      "learning_rate": 4.65479178034644e-05,
      "loss": 0.228,
      "step": 584
    },
    {
      "epoch": 0.5452003727865797,
      "grad_norm": 0.0745796486735344,
      "learning_rate": 4.6394752693712734e-05,
      "loss": 0.0021,
      "step": 585
    },
    {
      "epoch": 0.5461323392357875,
      "grad_norm": 0.6440649032592773,
      "learning_rate": 4.624162158429619e-05,
      "loss": 0.0187,
      "step": 586
    },
    {
      "epoch": 0.5470643056849953,
      "grad_norm": 1.015874981880188,
      "learning_rate": 4.608852591936231e-05,
      "loss": 0.0256,
      "step": 587
    },
    {
      "epoch": 0.5479962721342032,
      "grad_norm": 0.12743471562862396,
      "learning_rate": 4.593546714272438e-05,
      "loss": 0.0025,
      "step": 588
    },
    {
      "epoch": 0.548928238583411,
      "grad_norm": 1.2057770490646362,
      "learning_rate": 4.578244669784777e-05,
      "loss": 0.0317,
      "step": 589
    },
    {
      "epoch": 0.5498602050326188,
      "grad_norm": 2.5640909671783447,
      "learning_rate": 4.562946602783636e-05,
      "loss": 0.1007,
      "step": 590
    },
    {
      "epoch": 0.5507921714818267,
      "grad_norm": 1.9403150081634521,
      "learning_rate": 4.547652657541897e-05,
      "loss": 0.0554,
      "step": 591
    },
    {
      "epoch": 0.5517241379310345,
      "grad_norm": 1.5157524347305298,
      "learning_rate": 4.532362978293564e-05,
      "loss": 0.0366,
      "step": 592
    },
    {
      "epoch": 0.5526561043802423,
      "grad_norm": 1.6860204935073853,
      "learning_rate": 4.517077709232411e-05,
      "loss": 0.0322,
      "step": 593
    },
    {
      "epoch": 0.5535880708294502,
      "grad_norm": 0.4395802617073059,
      "learning_rate": 4.501796994510622e-05,
      "loss": 0.0131,
      "step": 594
    },
    {
      "epoch": 0.554520037278658,
      "grad_norm": 1.106860637664795,
      "learning_rate": 4.4865209782374304e-05,
      "loss": 0.04,
      "step": 595
    },
    {
      "epoch": 0.5554520037278659,
      "grad_norm": 0.5238240361213684,
      "learning_rate": 4.471249804477758e-05,
      "loss": 0.0125,
      "step": 596
    },
    {
      "epoch": 0.5563839701770736,
      "grad_norm": 3.6945576667785645,
      "learning_rate": 4.4559836172508565e-05,
      "loss": 0.1404,
      "step": 597
    },
    {
      "epoch": 0.5573159366262814,
      "grad_norm": 1.1280156373977661,
      "learning_rate": 4.4407225605289547e-05,
      "loss": 0.0245,
      "step": 598
    },
    {
      "epoch": 0.5582479030754893,
      "grad_norm": 1.4675111770629883,
      "learning_rate": 4.4254667782358924e-05,
      "loss": 0.0228,
      "step": 599
    },
    {
      "epoch": 0.5591798695246971,
      "grad_norm": 2.4232566356658936,
      "learning_rate": 4.410216414245771e-05,
      "loss": 0.0157,
      "step": 600
    },
    {
      "epoch": 0.5601118359739049,
      "grad_norm": 0.8545390367507935,
      "learning_rate": 4.394971612381591e-05,
      "loss": 0.029,
      "step": 601
    },
    {
      "epoch": 0.5610438024231128,
      "grad_norm": 0.5777714848518372,
      "learning_rate": 4.379732516413897e-05,
      "loss": 0.0127,
      "step": 602
    },
    {
      "epoch": 0.5619757688723206,
      "grad_norm": 0.8869187831878662,
      "learning_rate": 4.364499270059424e-05,
      "loss": 0.0227,
      "step": 603
    },
    {
      "epoch": 0.5629077353215284,
      "grad_norm": 0.9142225980758667,
      "learning_rate": 4.34927201697974e-05,
      "loss": 0.0552,
      "step": 604
    },
    {
      "epoch": 0.5638397017707363,
      "grad_norm": 2.659519910812378,
      "learning_rate": 4.334050900779893e-05,
      "loss": 0.1076,
      "step": 605
    },
    {
      "epoch": 0.5647716682199441,
      "grad_norm": 0.21417352557182312,
      "learning_rate": 4.3188360650070525e-05,
      "loss": 0.0068,
      "step": 606
    },
    {
      "epoch": 0.5657036346691519,
      "grad_norm": 0.3624404966831207,
      "learning_rate": 4.303627653149159e-05,
      "loss": 0.0087,
      "step": 607
    },
    {
      "epoch": 0.5666356011183598,
      "grad_norm": 1.3973774909973145,
      "learning_rate": 4.288425808633575e-05,
      "loss": 0.0599,
      "step": 608
    },
    {
      "epoch": 0.5675675675675675,
      "grad_norm": 1.1630985736846924,
      "learning_rate": 4.2732306748257224e-05,
      "loss": 0.0857,
      "step": 609
    },
    {
      "epoch": 0.5684995340167754,
      "grad_norm": 0.263249933719635,
      "learning_rate": 4.258042395027738e-05,
      "loss": 0.0084,
      "step": 610
    },
    {
      "epoch": 0.5694315004659832,
      "grad_norm": 2.8128879070281982,
      "learning_rate": 4.2428611124771184e-05,
      "loss": 0.1316,
      "step": 611
    },
    {
      "epoch": 0.570363466915191,
      "grad_norm": 1.825242042541504,
      "learning_rate": 4.2276869703453724e-05,
      "loss": 0.102,
      "step": 612
    },
    {
      "epoch": 0.5712954333643989,
      "grad_norm": 0.9243812561035156,
      "learning_rate": 4.21252011173667e-05,
      "loss": 0.0403,
      "step": 613
    },
    {
      "epoch": 0.5722273998136067,
      "grad_norm": 1.2102307081222534,
      "learning_rate": 4.1973606796864884e-05,
      "loss": 0.0637,
      "step": 614
    },
    {
      "epoch": 0.5731593662628145,
      "grad_norm": 1.0278326272964478,
      "learning_rate": 4.182208817160269e-05,
      "loss": 0.0872,
      "step": 615
    },
    {
      "epoch": 0.5740913327120224,
      "grad_norm": 0.973726212978363,
      "learning_rate": 4.167064667052065e-05,
      "loss": 0.0462,
      "step": 616
    },
    {
      "epoch": 0.5750232991612302,
      "grad_norm": 0.3728512227535248,
      "learning_rate": 4.151928372183198e-05,
      "loss": 0.0064,
      "step": 617
    },
    {
      "epoch": 0.575955265610438,
      "grad_norm": 2.0337462425231934,
      "learning_rate": 4.136800075300906e-05,
      "loss": 0.067,
      "step": 618
    },
    {
      "epoch": 0.5768872320596459,
      "grad_norm": 2.5226025581359863,
      "learning_rate": 4.1216799190770006e-05,
      "loss": 0.1414,
      "step": 619
    },
    {
      "epoch": 0.5778191985088537,
      "grad_norm": 0.20914335548877716,
      "learning_rate": 4.1065680461065194e-05,
      "loss": 0.0033,
      "step": 620
    },
    {
      "epoch": 0.5787511649580616,
      "grad_norm": 2.150099754333496,
      "learning_rate": 4.0914645989063845e-05,
      "loss": 0.0502,
      "step": 621
    },
    {
      "epoch": 0.5796831314072693,
      "grad_norm": 3.2852182388305664,
      "learning_rate": 4.076369719914055e-05,
      "loss": 0.0877,
      "step": 622
    },
    {
      "epoch": 0.5806150978564771,
      "grad_norm": 2.739978790283203,
      "learning_rate": 4.061283551486185e-05,
      "loss": 0.2081,
      "step": 623
    },
    {
      "epoch": 0.581547064305685,
      "grad_norm": 3.39498233795166,
      "learning_rate": 4.0462062358972796e-05,
      "loss": 0.1141,
      "step": 624
    },
    {
      "epoch": 0.5824790307548928,
      "grad_norm": 3.0169754028320312,
      "learning_rate": 4.03113791533836e-05,
      "loss": 0.0791,
      "step": 625
    },
    {
      "epoch": 0.5834109972041006,
      "grad_norm": 0.4348333477973938,
      "learning_rate": 4.016078731915608e-05,
      "loss": 0.0113,
      "step": 626
    },
    {
      "epoch": 0.5843429636533085,
      "grad_norm": 0.17418645322322845,
      "learning_rate": 4.001028827649046e-05,
      "loss": 0.0049,
      "step": 627
    },
    {
      "epoch": 0.5852749301025163,
      "grad_norm": 2.0803823471069336,
      "learning_rate": 3.9859883444711794e-05,
      "loss": 0.0276,
      "step": 628
    },
    {
      "epoch": 0.5862068965517241,
      "grad_norm": 0.5942701697349548,
      "learning_rate": 3.970957424225666e-05,
      "loss": 0.0176,
      "step": 629
    },
    {
      "epoch": 0.587138863000932,
      "grad_norm": 1.0880169868469238,
      "learning_rate": 3.955936208665979e-05,
      "loss": 0.0319,
      "step": 630
    },
    {
      "epoch": 0.5880708294501398,
      "grad_norm": 1.0171687602996826,
      "learning_rate": 3.940924839454066e-05,
      "loss": 0.0592,
      "step": 631
    },
    {
      "epoch": 0.5890027958993477,
      "grad_norm": 0.8191186189651489,
      "learning_rate": 3.925923458159023e-05,
      "loss": 0.0173,
      "step": 632
    },
    {
      "epoch": 0.5899347623485555,
      "grad_norm": 0.34623149037361145,
      "learning_rate": 3.9109322062557424e-05,
      "loss": 0.0099,
      "step": 633
    },
    {
      "epoch": 0.5908667287977633,
      "grad_norm": 0.3688358962535858,
      "learning_rate": 3.895951225123595e-05,
      "loss": 0.0094,
      "step": 634
    },
    {
      "epoch": 0.5917986952469712,
      "grad_norm": 1.6616538763046265,
      "learning_rate": 3.880980656045087e-05,
      "loss": 0.0241,
      "step": 635
    },
    {
      "epoch": 0.5927306616961789,
      "grad_norm": 0.9035270810127258,
      "learning_rate": 3.866020640204533e-05,
      "loss": 0.0169,
      "step": 636
    },
    {
      "epoch": 0.5936626281453867,
      "grad_norm": 0.6977134346961975,
      "learning_rate": 3.851071318686721e-05,
      "loss": 0.0251,
      "step": 637
    },
    {
      "epoch": 0.5945945945945946,
      "grad_norm": 1.0212866067886353,
      "learning_rate": 3.8361328324755825e-05,
      "loss": 0.0717,
      "step": 638
    },
    {
      "epoch": 0.5955265610438024,
      "grad_norm": 1.4821285009384155,
      "learning_rate": 3.821205322452863e-05,
      "loss": 0.0705,
      "step": 639
    },
    {
      "epoch": 0.5964585274930102,
      "grad_norm": 0.7820432186126709,
      "learning_rate": 3.8062889293967976e-05,
      "loss": 0.0187,
      "step": 640
    },
    {
      "epoch": 0.5973904939422181,
      "grad_norm": 1.4119162559509277,
      "learning_rate": 3.791383793980776e-05,
      "loss": 0.0756,
      "step": 641
    },
    {
      "epoch": 0.5983224603914259,
      "grad_norm": 0.652569591999054,
      "learning_rate": 3.7764900567720196e-05,
      "loss": 0.0153,
      "step": 642
    },
    {
      "epoch": 0.5992544268406338,
      "grad_norm": 0.5894368290901184,
      "learning_rate": 3.761607858230257e-05,
      "loss": 0.0099,
      "step": 643
    },
    {
      "epoch": 0.6001863932898416,
      "grad_norm": 1.44651198387146,
      "learning_rate": 3.746737338706397e-05,
      "loss": 0.0758,
      "step": 644
    },
    {
      "epoch": 0.6011183597390494,
      "grad_norm": 1.0577771663665771,
      "learning_rate": 3.7318786384412074e-05,
      "loss": 0.0427,
      "step": 645
    },
    {
      "epoch": 0.6020503261882573,
      "grad_norm": 1.540482997894287,
      "learning_rate": 3.71703189756399e-05,
      "loss": 0.0339,
      "step": 646
    },
    {
      "epoch": 0.6029822926374651,
      "grad_norm": 2.182873249053955,
      "learning_rate": 3.70219725609126e-05,
      "loss": 0.075,
      "step": 647
    },
    {
      "epoch": 0.6039142590866728,
      "grad_norm": 1.5290659666061401,
      "learning_rate": 3.6873748539254246e-05,
      "loss": 0.0569,
      "step": 648
    },
    {
      "epoch": 0.6048462255358807,
      "grad_norm": 3.3621015548706055,
      "learning_rate": 3.6725648308534696e-05,
      "loss": 0.1976,
      "step": 649
    },
    {
      "epoch": 0.6057781919850885,
      "grad_norm": 0.9337851405143738,
      "learning_rate": 3.65776732654563e-05,
      "loss": 0.054,
      "step": 650
    },
    {
      "epoch": 0.6067101584342963,
      "grad_norm": 0.25393131375312805,
      "learning_rate": 3.6429824805540815e-05,
      "loss": 0.0052,
      "step": 651
    },
    {
      "epoch": 0.6076421248835042,
      "grad_norm": 1.53010892868042,
      "learning_rate": 3.628210432311621e-05,
      "loss": 0.0433,
      "step": 652
    },
    {
      "epoch": 0.608574091332712,
      "grad_norm": 0.31782206892967224,
      "learning_rate": 3.6134513211303556e-05,
      "loss": 0.0039,
      "step": 653
    },
    {
      "epoch": 0.6095060577819198,
      "grad_norm": 0.7145633101463318,
      "learning_rate": 3.598705286200382e-05,
      "loss": 0.0182,
      "step": 654
    },
    {
      "epoch": 0.6104380242311277,
      "grad_norm": 0.5081955790519714,
      "learning_rate": 3.58397246658848e-05,
      "loss": 0.0138,
      "step": 655
    },
    {
      "epoch": 0.6113699906803355,
      "grad_norm": 1.1797877550125122,
      "learning_rate": 3.5692530012367955e-05,
      "loss": 0.0096,
      "step": 656
    },
    {
      "epoch": 0.6123019571295434,
      "grad_norm": 0.0682310089468956,
      "learning_rate": 3.554547028961537e-05,
      "loss": 0.0023,
      "step": 657
    },
    {
      "epoch": 0.6132339235787512,
      "grad_norm": 1.760521650314331,
      "learning_rate": 3.5398546884516604e-05,
      "loss": 0.0529,
      "step": 658
    },
    {
      "epoch": 0.614165890027959,
      "grad_norm": 0.818846583366394,
      "learning_rate": 3.5251761182675625e-05,
      "loss": 0.0297,
      "step": 659
    },
    {
      "epoch": 0.6150978564771669,
      "grad_norm": 0.3822430372238159,
      "learning_rate": 3.510511456839777e-05,
      "loss": 0.0111,
      "step": 660
    },
    {
      "epoch": 0.6160298229263746,
      "grad_norm": 1.5274235010147095,
      "learning_rate": 3.495860842467664e-05,
      "loss": 0.0688,
      "step": 661
    },
    {
      "epoch": 0.6169617893755824,
      "grad_norm": 2.4907660484313965,
      "learning_rate": 3.481224413318114e-05,
      "loss": 0.0601,
      "step": 662
    },
    {
      "epoch": 0.6178937558247903,
      "grad_norm": 1.6190844774246216,
      "learning_rate": 3.4666023074242356e-05,
      "loss": 0.126,
      "step": 663
    },
    {
      "epoch": 0.6188257222739981,
      "grad_norm": 1.7249010801315308,
      "learning_rate": 3.4519946626840573e-05,
      "loss": 0.0391,
      "step": 664
    },
    {
      "epoch": 0.6197576887232059,
      "grad_norm": 1.1270893812179565,
      "learning_rate": 3.4374016168592296e-05,
      "loss": 0.0391,
      "step": 665
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 2.65583872795105,
      "learning_rate": 3.422823307573722e-05,
      "loss": 0.0979,
      "step": 666
    },
    {
      "epoch": 0.6216216216216216,
      "grad_norm": 0.3144218921661377,
      "learning_rate": 3.40825987231253e-05,
      "loss": 0.0056,
      "step": 667
    },
    {
      "epoch": 0.6225535880708295,
      "grad_norm": 0.4311087727546692,
      "learning_rate": 3.393711448420372e-05,
      "loss": 0.0116,
      "step": 668
    },
    {
      "epoch": 0.6234855545200373,
      "grad_norm": 1.4498568773269653,
      "learning_rate": 3.379178173100396e-05,
      "loss": 0.0887,
      "step": 669
    },
    {
      "epoch": 0.6244175209692451,
      "grad_norm": 0.8447300791740417,
      "learning_rate": 3.364660183412892e-05,
      "loss": 0.0198,
      "step": 670
    },
    {
      "epoch": 0.625349487418453,
      "grad_norm": 0.34523147344589233,
      "learning_rate": 3.3501576162739904e-05,
      "loss": 0.0081,
      "step": 671
    },
    {
      "epoch": 0.6262814538676608,
      "grad_norm": 0.3423440456390381,
      "learning_rate": 3.335670608454376e-05,
      "loss": 0.0096,
      "step": 672
    },
    {
      "epoch": 0.6272134203168686,
      "grad_norm": 0.6186404228210449,
      "learning_rate": 3.321199296577998e-05,
      "loss": 0.0128,
      "step": 673
    },
    {
      "epoch": 0.6281453867660765,
      "grad_norm": 1.3062037229537964,
      "learning_rate": 3.3067438171207766e-05,
      "loss": 0.0293,
      "step": 674
    },
    {
      "epoch": 0.6290773532152842,
      "grad_norm": 8.898473739624023,
      "learning_rate": 3.292304306409325e-05,
      "loss": 0.2124,
      "step": 675
    },
    {
      "epoch": 0.630009319664492,
      "grad_norm": 2.829618453979492,
      "learning_rate": 3.2778809006196564e-05,
      "loss": 0.0664,
      "step": 676
    },
    {
      "epoch": 0.6309412861136999,
      "grad_norm": 1.4280197620391846,
      "learning_rate": 3.263473735775899e-05,
      "loss": 0.0601,
      "step": 677
    },
    {
      "epoch": 0.6318732525629077,
      "grad_norm": 1.7249200344085693,
      "learning_rate": 3.249082947749019e-05,
      "loss": 0.0946,
      "step": 678
    },
    {
      "epoch": 0.6328052190121156,
      "grad_norm": 0.622667670249939,
      "learning_rate": 3.234708672255538e-05,
      "loss": 0.0068,
      "step": 679
    },
    {
      "epoch": 0.6337371854613234,
      "grad_norm": 2.2010300159454346,
      "learning_rate": 3.220351044856247e-05,
      "loss": 0.0679,
      "step": 680
    },
    {
      "epoch": 0.6346691519105312,
      "grad_norm": 1.1353260278701782,
      "learning_rate": 3.206010200954935e-05,
      "loss": 0.0427,
      "step": 681
    },
    {
      "epoch": 0.6356011183597391,
      "grad_norm": 0.8460433483123779,
      "learning_rate": 3.191686275797107e-05,
      "loss": 0.012,
      "step": 682
    },
    {
      "epoch": 0.6365330848089469,
      "grad_norm": 1.015140414237976,
      "learning_rate": 3.177379404468715e-05,
      "loss": 0.0281,
      "step": 683
    },
    {
      "epoch": 0.6374650512581547,
      "grad_norm": 1.5313421487808228,
      "learning_rate": 3.163089721894876e-05,
      "loss": 0.0582,
      "step": 684
    },
    {
      "epoch": 0.6383970177073626,
      "grad_norm": 0.6894136071205139,
      "learning_rate": 3.148817362838606e-05,
      "loss": 0.0102,
      "step": 685
    },
    {
      "epoch": 0.6393289841565704,
      "grad_norm": 0.2898493707180023,
      "learning_rate": 3.134562461899545e-05,
      "loss": 0.0054,
      "step": 686
    },
    {
      "epoch": 0.6402609506057781,
      "grad_norm": 0.8155750036239624,
      "learning_rate": 3.120325153512687e-05,
      "loss": 0.0197,
      "step": 687
    },
    {
      "epoch": 0.641192917054986,
      "grad_norm": 0.26222237944602966,
      "learning_rate": 3.10610557194712e-05,
      "loss": 0.0075,
      "step": 688
    },
    {
      "epoch": 0.6421248835041938,
      "grad_norm": 0.5798500776290894,
      "learning_rate": 3.091903851304751e-05,
      "loss": 0.013,
      "step": 689
    },
    {
      "epoch": 0.6430568499534017,
      "grad_norm": 1.631866693496704,
      "learning_rate": 3.077720125519042e-05,
      "loss": 0.0798,
      "step": 690
    },
    {
      "epoch": 0.6439888164026095,
      "grad_norm": 0.4146805703639984,
      "learning_rate": 3.063554528353752e-05,
      "loss": 0.0091,
      "step": 691
    },
    {
      "epoch": 0.6449207828518173,
      "grad_norm": 1.7114466428756714,
      "learning_rate": 3.0494071934016737e-05,
      "loss": 0.0486,
      "step": 692
    },
    {
      "epoch": 0.6458527493010252,
      "grad_norm": 1.099481225013733,
      "learning_rate": 3.03527825408337e-05,
      "loss": 0.0241,
      "step": 693
    },
    {
      "epoch": 0.646784715750233,
      "grad_norm": 1.2806260585784912,
      "learning_rate": 3.0211678436459212e-05,
      "loss": 0.0481,
      "step": 694
    },
    {
      "epoch": 0.6477166821994408,
      "grad_norm": 0.46307051181793213,
      "learning_rate": 3.0070760951616618e-05,
      "loss": 0.011,
      "step": 695
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 2.406785488128662,
      "learning_rate": 2.993003141526933e-05,
      "loss": 0.0789,
      "step": 696
    },
    {
      "epoch": 0.6495806150978565,
      "grad_norm": 1.8010510206222534,
      "learning_rate": 2.978949115460824e-05,
      "loss": 0.0789,
      "step": 697
    },
    {
      "epoch": 0.6505125815470643,
      "grad_norm": 0.39275264739990234,
      "learning_rate": 2.9649141495039223e-05,
      "loss": 0.0066,
      "step": 698
    },
    {
      "epoch": 0.6514445479962722,
      "grad_norm": 2.9994301795959473,
      "learning_rate": 2.950898376017064e-05,
      "loss": 0.056,
      "step": 699
    },
    {
      "epoch": 0.65237651444548,
      "grad_norm": 6.136080741882324,
      "learning_rate": 2.9369019271800823e-05,
      "loss": 0.1862,
      "step": 700
    },
    {
      "epoch": 0.6533084808946877,
      "grad_norm": 0.296806275844574,
      "learning_rate": 2.9229249349905684e-05,
      "loss": 0.0072,
      "step": 701
    },
    {
      "epoch": 0.6542404473438956,
      "grad_norm": 0.6903855204582214,
      "learning_rate": 2.908967531262618e-05,
      "loss": 0.0312,
      "step": 702
    },
    {
      "epoch": 0.6551724137931034,
      "grad_norm": 1.400633692741394,
      "learning_rate": 2.895029847625595e-05,
      "loss": 0.0153,
      "step": 703
    },
    {
      "epoch": 0.6561043802423113,
      "grad_norm": 0.370590478181839,
      "learning_rate": 2.8811120155228844e-05,
      "loss": 0.0085,
      "step": 704
    },
    {
      "epoch": 0.6570363466915191,
      "grad_norm": 0.47119617462158203,
      "learning_rate": 2.8672141662106578e-05,
      "loss": 0.0088,
      "step": 705
    },
    {
      "epoch": 0.6579683131407269,
      "grad_norm": 0.07778705656528473,
      "learning_rate": 2.853336430756631e-05,
      "loss": 0.0024,
      "step": 706
    },
    {
      "epoch": 0.6589002795899348,
      "grad_norm": 1.7751870155334473,
      "learning_rate": 2.8394789400388328e-05,
      "loss": 0.0729,
      "step": 707
    },
    {
      "epoch": 0.6598322460391426,
      "grad_norm": 1.4420006275177002,
      "learning_rate": 2.8256418247443662e-05,
      "loss": 0.0563,
      "step": 708
    },
    {
      "epoch": 0.6607642124883504,
      "grad_norm": 0.22819720208644867,
      "learning_rate": 2.811825215368179e-05,
      "loss": 0.0068,
      "step": 709
    },
    {
      "epoch": 0.6616961789375583,
      "grad_norm": 0.3278886377811432,
      "learning_rate": 2.798029242211828e-05,
      "loss": 0.0067,
      "step": 710
    },
    {
      "epoch": 0.6626281453867661,
      "grad_norm": 0.25643807649612427,
      "learning_rate": 2.7842540353822632e-05,
      "loss": 0.005,
      "step": 711
    },
    {
      "epoch": 0.6635601118359739,
      "grad_norm": 0.47565677762031555,
      "learning_rate": 2.770499724790584e-05,
      "loss": 0.012,
      "step": 712
    },
    {
      "epoch": 0.6644920782851818,
      "grad_norm": 0.05990108847618103,
      "learning_rate": 2.7567664401508225e-05,
      "loss": 0.0012,
      "step": 713
    },
    {
      "epoch": 0.6654240447343895,
      "grad_norm": 0.6037838459014893,
      "learning_rate": 2.7430543109787222e-05,
      "loss": 0.0086,
      "step": 714
    },
    {
      "epoch": 0.6663560111835974,
      "grad_norm": 2.216022491455078,
      "learning_rate": 2.729363466590511e-05,
      "loss": 0.0832,
      "step": 715
    },
    {
      "epoch": 0.6672879776328052,
      "grad_norm": 1.0071760416030884,
      "learning_rate": 2.7156940361016864e-05,
      "loss": 0.0313,
      "step": 716
    },
    {
      "epoch": 0.668219944082013,
      "grad_norm": 0.5068988800048828,
      "learning_rate": 2.702046148425795e-05,
      "loss": 0.0127,
      "step": 717
    },
    {
      "epoch": 0.6691519105312209,
      "grad_norm": 0.16356121003627777,
      "learning_rate": 2.688419932273219e-05,
      "loss": 0.0029,
      "step": 718
    },
    {
      "epoch": 0.6700838769804287,
      "grad_norm": 0.5813694596290588,
      "learning_rate": 2.6748155161499567e-05,
      "loss": 0.0086,
      "step": 719
    },
    {
      "epoch": 0.6710158434296365,
      "grad_norm": 0.29364094138145447,
      "learning_rate": 2.6612330283564223e-05,
      "loss": 0.0052,
      "step": 720
    },
    {
      "epoch": 0.6719478098788444,
      "grad_norm": 1.4780179262161255,
      "learning_rate": 2.6476725969862227e-05,
      "loss": 0.0357,
      "step": 721
    },
    {
      "epoch": 0.6728797763280522,
      "grad_norm": 0.11548914760351181,
      "learning_rate": 2.634134349924956e-05,
      "loss": 0.002,
      "step": 722
    },
    {
      "epoch": 0.67381174277726,
      "grad_norm": 0.39383208751678467,
      "learning_rate": 2.6206184148490066e-05,
      "loss": 0.0091,
      "step": 723
    },
    {
      "epoch": 0.6747437092264679,
      "grad_norm": 0.3109658658504486,
      "learning_rate": 2.6071249192243363e-05,
      "loss": 0.0018,
      "step": 724
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 3.5623598098754883,
      "learning_rate": 2.5936539903052892e-05,
      "loss": 0.1177,
      "step": 725
    },
    {
      "epoch": 0.6766076421248836,
      "grad_norm": 0.1712770015001297,
      "learning_rate": 2.580205755133384e-05,
      "loss": 0.0013,
      "step": 726
    },
    {
      "epoch": 0.6775396085740913,
      "grad_norm": 0.7612726092338562,
      "learning_rate": 2.5667803405361213e-05,
      "loss": 0.0122,
      "step": 727
    },
    {
      "epoch": 0.6784715750232991,
      "grad_norm": 0.7174275517463684,
      "learning_rate": 2.5533778731257824e-05,
      "loss": 0.0259,
      "step": 728
    },
    {
      "epoch": 0.679403541472507,
      "grad_norm": 1.2396039962768555,
      "learning_rate": 2.5399984792982457e-05,
      "loss": 0.0468,
      "step": 729
    },
    {
      "epoch": 0.6803355079217148,
      "grad_norm": 0.49372637271881104,
      "learning_rate": 2.5266422852317795e-05,
      "loss": 0.0081,
      "step": 730
    },
    {
      "epoch": 0.6812674743709226,
      "grad_norm": 0.16633592545986176,
      "learning_rate": 2.513309416885865e-05,
      "loss": 0.0042,
      "step": 731
    },
    {
      "epoch": 0.6821994408201305,
      "grad_norm": 1.141757607460022,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.0391,
      "step": 732
    },
    {
      "epoch": 0.6831314072693383,
      "grad_norm": 0.2698551118373871,
      "learning_rate": 2.4867141600925214e-05,
      "loss": 0.005,
      "step": 733
    },
    {
      "epoch": 0.6840633737185461,
      "grad_norm": 0.20871886610984802,
      "learning_rate": 2.4734520224594093e-05,
      "loss": 0.0061,
      "step": 734
    },
    {
      "epoch": 0.684995340167754,
      "grad_norm": 0.3639092445373535,
      "learning_rate": 2.4602137121731195e-05,
      "loss": 0.0122,
      "step": 735
    },
    {
      "epoch": 0.6859273066169618,
      "grad_norm": 1.5544925928115845,
      "learning_rate": 2.44699935408139e-05,
      "loss": 0.0348,
      "step": 736
    },
    {
      "epoch": 0.6868592730661697,
      "grad_norm": 0.7567540407180786,
      "learning_rate": 2.433809072806081e-05,
      "loss": 0.0063,
      "step": 737
    },
    {
      "epoch": 0.6877912395153775,
      "grad_norm": 1.6622443199157715,
      "learning_rate": 2.4206429927419794e-05,
      "loss": 0.0182,
      "step": 738
    },
    {
      "epoch": 0.6887232059645852,
      "grad_norm": 0.32770344614982605,
      "learning_rate": 2.40750123805564e-05,
      "loss": 0.0039,
      "step": 739
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 1.0293552875518799,
      "learning_rate": 2.3943839326842092e-05,
      "loss": 0.0309,
      "step": 740
    },
    {
      "epoch": 0.6905871388630009,
      "grad_norm": 1.511672854423523,
      "learning_rate": 2.3812912003342568e-05,
      "loss": 0.0387,
      "step": 741
    },
    {
      "epoch": 0.6915191053122087,
      "grad_norm": 1.2512294054031372,
      "learning_rate": 2.3682231644806108e-05,
      "loss": 0.0276,
      "step": 742
    },
    {
      "epoch": 0.6924510717614166,
      "grad_norm": 0.2696451246738434,
      "learning_rate": 2.3551799483651894e-05,
      "loss": 0.0027,
      "step": 743
    },
    {
      "epoch": 0.6933830382106244,
      "grad_norm": 0.7463300228118896,
      "learning_rate": 2.342161674995843e-05,
      "loss": 0.0067,
      "step": 744
    },
    {
      "epoch": 0.6943150046598322,
      "grad_norm": 0.102654367685318,
      "learning_rate": 2.3291684671451903e-05,
      "loss": 0.0025,
      "step": 745
    },
    {
      "epoch": 0.6952469711090401,
      "grad_norm": 0.2291325330734253,
      "learning_rate": 2.3162004473494657e-05,
      "loss": 0.004,
      "step": 746
    },
    {
      "epoch": 0.6961789375582479,
      "grad_norm": 2.3645198345184326,
      "learning_rate": 2.3032577379073577e-05,
      "loss": 0.0629,
      "step": 747
    },
    {
      "epoch": 0.6971109040074557,
      "grad_norm": 0.990889847278595,
      "learning_rate": 2.2903404608788582e-05,
      "loss": 0.024,
      "step": 748
    },
    {
      "epoch": 0.6980428704566636,
      "grad_norm": 2.9361162185668945,
      "learning_rate": 2.2774487380841115e-05,
      "loss": 0.0389,
      "step": 749
    },
    {
      "epoch": 0.6989748369058714,
      "grad_norm": 1.2799005508422852,
      "learning_rate": 2.2645826911022655e-05,
      "loss": 0.0631,
      "step": 750
    },
    {
      "epoch": 0.6999068033550793,
      "grad_norm": 0.12018714100122452,
      "learning_rate": 2.2517424412703252e-05,
      "loss": 0.0021,
      "step": 751
    },
    {
      "epoch": 0.700838769804287,
      "grad_norm": 0.8238220810890198,
      "learning_rate": 2.2389281096820075e-05,
      "loss": 0.0259,
      "step": 752
    },
    {
      "epoch": 0.7017707362534948,
      "grad_norm": 0.07809086889028549,
      "learning_rate": 2.2261398171865978e-05,
      "loss": 0.0017,
      "step": 753
    },
    {
      "epoch": 0.7027027027027027,
      "grad_norm": 2.065542221069336,
      "learning_rate": 2.2133776843878186e-05,
      "loss": 0.0705,
      "step": 754
    },
    {
      "epoch": 0.7036346691519105,
      "grad_norm": 0.10505253821611404,
      "learning_rate": 2.2006418316426775e-05,
      "loss": 0.0023,
      "step": 755
    },
    {
      "epoch": 0.7045666356011183,
      "grad_norm": 0.19952963292598724,
      "learning_rate": 2.187932379060348e-05,
      "loss": 0.0045,
      "step": 756
    },
    {
      "epoch": 0.7054986020503262,
      "grad_norm": 0.06244141235947609,
      "learning_rate": 2.1752494465010242e-05,
      "loss": 0.0016,
      "step": 757
    },
    {
      "epoch": 0.706430568499534,
      "grad_norm": 1.953023910522461,
      "learning_rate": 2.1625931535747964e-05,
      "loss": 0.0261,
      "step": 758
    },
    {
      "epoch": 0.7073625349487418,
      "grad_norm": 0.7066685557365417,
      "learning_rate": 2.1499636196405228e-05,
      "loss": 0.0232,
      "step": 759
    },
    {
      "epoch": 0.7082945013979497,
      "grad_norm": 0.7350996136665344,
      "learning_rate": 2.1373609638047032e-05,
      "loss": 0.0185,
      "step": 760
    },
    {
      "epoch": 0.7092264678471575,
      "grad_norm": 0.9684416651725769,
      "learning_rate": 2.1247853049203543e-05,
      "loss": 0.0263,
      "step": 761
    },
    {
      "epoch": 0.7101584342963654,
      "grad_norm": 1.195703148841858,
      "learning_rate": 2.112236761585892e-05,
      "loss": 0.0458,
      "step": 762
    },
    {
      "epoch": 0.7110904007455732,
      "grad_norm": 1.525630235671997,
      "learning_rate": 2.09971545214401e-05,
      "loss": 0.0274,
      "step": 763
    },
    {
      "epoch": 0.712022367194781,
      "grad_norm": 0.7897143363952637,
      "learning_rate": 2.087221494680563e-05,
      "loss": 0.0145,
      "step": 764
    },
    {
      "epoch": 0.7129543336439889,
      "grad_norm": 0.4849226772785187,
      "learning_rate": 2.074755007023461e-05,
      "loss": 0.0095,
      "step": 765
    },
    {
      "epoch": 0.7138863000931966,
      "grad_norm": 2.5897316932678223,
      "learning_rate": 2.0623161067415463e-05,
      "loss": 0.0353,
      "step": 766
    },
    {
      "epoch": 0.7148182665424044,
      "grad_norm": 0.49661892652511597,
      "learning_rate": 2.049904911143492e-05,
      "loss": 0.014,
      "step": 767
    },
    {
      "epoch": 0.7157502329916123,
      "grad_norm": 0.16137608885765076,
      "learning_rate": 2.0375215372766944e-05,
      "loss": 0.0023,
      "step": 768
    },
    {
      "epoch": 0.7166821994408201,
      "grad_norm": 1.290913462638855,
      "learning_rate": 2.025166101926168e-05,
      "loss": 0.034,
      "step": 769
    },
    {
      "epoch": 0.7176141658900279,
      "grad_norm": 0.3757019639015198,
      "learning_rate": 2.012838721613447e-05,
      "loss": 0.0048,
      "step": 770
    },
    {
      "epoch": 0.7185461323392358,
      "grad_norm": 0.8790585398674011,
      "learning_rate": 2.0005395125954812e-05,
      "loss": 0.0211,
      "step": 771
    },
    {
      "epoch": 0.7194780987884436,
      "grad_norm": 0.06951785832643509,
      "learning_rate": 1.9882685908635462e-05,
      "loss": 0.001,
      "step": 772
    },
    {
      "epoch": 0.7204100652376515,
      "grad_norm": 0.6821100115776062,
      "learning_rate": 1.9760260721421426e-05,
      "loss": 0.0127,
      "step": 773
    },
    {
      "epoch": 0.7213420316868593,
      "grad_norm": 1.6195906400680542,
      "learning_rate": 1.963812071887913e-05,
      "loss": 0.0168,
      "step": 774
    },
    {
      "epoch": 0.7222739981360671,
      "grad_norm": 6.112545013427734,
      "learning_rate": 1.951626705288544e-05,
      "loss": 0.1486,
      "step": 775
    },
    {
      "epoch": 0.723205964585275,
      "grad_norm": 0.20122942328453064,
      "learning_rate": 1.9394700872616855e-05,
      "loss": 0.0041,
      "step": 776
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.0628390684723854,
      "learning_rate": 1.9273423324538658e-05,
      "loss": 0.0012,
      "step": 777
    },
    {
      "epoch": 0.7250698974836906,
      "grad_norm": 0.9709482192993164,
      "learning_rate": 1.9152435552394104e-05,
      "loss": 0.0271,
      "step": 778
    },
    {
      "epoch": 0.7260018639328985,
      "grad_norm": 1.464906096458435,
      "learning_rate": 1.903173869719362e-05,
      "loss": 0.0446,
      "step": 779
    },
    {
      "epoch": 0.7269338303821062,
      "grad_norm": 4.5541791915893555,
      "learning_rate": 1.891133389720407e-05,
      "loss": 0.0495,
      "step": 780
    },
    {
      "epoch": 0.727865796831314,
      "grad_norm": 0.476789265871048,
      "learning_rate": 1.8791222287937982e-05,
      "loss": 0.0031,
      "step": 781
    },
    {
      "epoch": 0.7287977632805219,
      "grad_norm": 0.5186454653739929,
      "learning_rate": 1.8671405002142918e-05,
      "loss": 0.0167,
      "step": 782
    },
    {
      "epoch": 0.7297297297297297,
      "grad_norm": 2.046664237976074,
      "learning_rate": 1.855188316979068e-05,
      "loss": 0.0679,
      "step": 783
    },
    {
      "epoch": 0.7306616961789375,
      "grad_norm": 0.5090547204017639,
      "learning_rate": 1.8432657918066732e-05,
      "loss": 0.0127,
      "step": 784
    },
    {
      "epoch": 0.7315936626281454,
      "grad_norm": 0.25921696424484253,
      "learning_rate": 1.831373037135955e-05,
      "loss": 0.0038,
      "step": 785
    },
    {
      "epoch": 0.7325256290773532,
      "grad_norm": 1.4198919534683228,
      "learning_rate": 1.8195101651250017e-05,
      "loss": 0.0456,
      "step": 786
    },
    {
      "epoch": 0.7334575955265611,
      "grad_norm": 0.12709754705429077,
      "learning_rate": 1.807677287650083e-05,
      "loss": 0.0022,
      "step": 787
    },
    {
      "epoch": 0.7343895619757689,
      "grad_norm": 0.19620268046855927,
      "learning_rate": 1.7958745163045986e-05,
      "loss": 0.0049,
      "step": 788
    },
    {
      "epoch": 0.7353215284249767,
      "grad_norm": 1.1757110357284546,
      "learning_rate": 1.7841019623980217e-05,
      "loss": 0.061,
      "step": 789
    },
    {
      "epoch": 0.7362534948741846,
      "grad_norm": 0.24634391069412231,
      "learning_rate": 1.77235973695485e-05,
      "loss": 0.0045,
      "step": 790
    },
    {
      "epoch": 0.7371854613233924,
      "grad_norm": 1.143908143043518,
      "learning_rate": 1.760647950713566e-05,
      "loss": 0.0276,
      "step": 791
    },
    {
      "epoch": 0.7381174277726001,
      "grad_norm": 0.6017893552780151,
      "learning_rate": 1.74896671412558e-05,
      "loss": 0.0078,
      "step": 792
    },
    {
      "epoch": 0.739049394221808,
      "grad_norm": 0.18127556145191193,
      "learning_rate": 1.737316137354197e-05,
      "loss": 0.0029,
      "step": 793
    },
    {
      "epoch": 0.7399813606710158,
      "grad_norm": 1.0178278684616089,
      "learning_rate": 1.725696330273575e-05,
      "loss": 0.0244,
      "step": 794
    },
    {
      "epoch": 0.7409133271202236,
      "grad_norm": 0.9616022706031799,
      "learning_rate": 1.7141074024676912e-05,
      "loss": 0.0233,
      "step": 795
    },
    {
      "epoch": 0.7418452935694315,
      "grad_norm": 1.721096396446228,
      "learning_rate": 1.7025494632293048e-05,
      "loss": 0.055,
      "step": 796
    },
    {
      "epoch": 0.7427772600186393,
      "grad_norm": 1.3252190351486206,
      "learning_rate": 1.6910226215589303e-05,
      "loss": 0.024,
      "step": 797
    },
    {
      "epoch": 0.7437092264678472,
      "grad_norm": 0.27513352036476135,
      "learning_rate": 1.6795269861638042e-05,
      "loss": 0.0041,
      "step": 798
    },
    {
      "epoch": 0.744641192917055,
      "grad_norm": 0.22185392677783966,
      "learning_rate": 1.6680626654568686e-05,
      "loss": 0.0031,
      "step": 799
    },
    {
      "epoch": 0.7455731593662628,
      "grad_norm": 4.715059757232666,
      "learning_rate": 1.656629767555739e-05,
      "loss": 0.1378,
      "step": 800
    },
    {
      "epoch": 0.7465051258154707,
      "grad_norm": 0.09731733053922653,
      "learning_rate": 1.645228400281689e-05,
      "loss": 0.0021,
      "step": 801
    },
    {
      "epoch": 0.7474370922646785,
      "grad_norm": 0.505806565284729,
      "learning_rate": 1.6338586711586356e-05,
      "loss": 0.0118,
      "step": 802
    },
    {
      "epoch": 0.7483690587138863,
      "grad_norm": 0.15126951038837433,
      "learning_rate": 1.6225206874121218e-05,
      "loss": 0.0027,
      "step": 803
    },
    {
      "epoch": 0.7493010251630942,
      "grad_norm": 1.0030357837677002,
      "learning_rate": 1.6112145559683055e-05,
      "loss": 0.0262,
      "step": 804
    },
    {
      "epoch": 0.750232991612302,
      "grad_norm": 2.0372867584228516,
      "learning_rate": 1.5999403834529548e-05,
      "loss": 0.0985,
      "step": 805
    },
    {
      "epoch": 0.7511649580615097,
      "grad_norm": 0.2379598617553711,
      "learning_rate": 1.5886982761904377e-05,
      "loss": 0.0056,
      "step": 806
    },
    {
      "epoch": 0.7520969245107176,
      "grad_norm": 0.8561238050460815,
      "learning_rate": 1.5774883402027208e-05,
      "loss": 0.0183,
      "step": 807
    },
    {
      "epoch": 0.7520969245107176,
      "eval_loss": 0.029690111055970192,
      "eval_runtime": 127.9702,
      "eval_samples_per_second": 14.12,
      "eval_steps_per_second": 3.532,
      "step": 807
    },
    {
      "epoch": 0.7530288909599254,
      "grad_norm": 0.1286248415708542,
      "learning_rate": 1.5663106812083743e-05,
      "loss": 0.0035,
      "step": 808
    },
    {
      "epoch": 0.7539608574091333,
      "grad_norm": 0.29819682240486145,
      "learning_rate": 1.555165404621567e-05,
      "loss": 0.0043,
      "step": 809
    },
    {
      "epoch": 0.7548928238583411,
      "grad_norm": 1.4682197570800781,
      "learning_rate": 1.5440526155510763e-05,
      "loss": 0.0349,
      "step": 810
    },
    {
      "epoch": 0.7558247903075489,
      "grad_norm": 2.5306453704833984,
      "learning_rate": 1.532972418799298e-05,
      "loss": 0.1089,
      "step": 811
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 0.7194310426712036,
      "learning_rate": 1.5219249188612556e-05,
      "loss": 0.0092,
      "step": 812
    },
    {
      "epoch": 0.7576887232059646,
      "grad_norm": 0.7676751613616943,
      "learning_rate": 1.5109102199236152e-05,
      "loss": 0.0195,
      "step": 813
    },
    {
      "epoch": 0.7586206896551724,
      "grad_norm": 0.3930986225605011,
      "learning_rate": 1.4999284258637053e-05,
      "loss": 0.0055,
      "step": 814
    },
    {
      "epoch": 0.7595526561043803,
      "grad_norm": 1.3155858516693115,
      "learning_rate": 1.488979640248534e-05,
      "loss": 0.0408,
      "step": 815
    },
    {
      "epoch": 0.7604846225535881,
      "grad_norm": 0.8961852788925171,
      "learning_rate": 1.4780639663338125e-05,
      "loss": 0.014,
      "step": 816
    },
    {
      "epoch": 0.7614165890027959,
      "grad_norm": 0.16851025819778442,
      "learning_rate": 1.4671815070629868e-05,
      "loss": 0.0031,
      "step": 817
    },
    {
      "epoch": 0.7623485554520038,
      "grad_norm": 0.2377876490354538,
      "learning_rate": 1.4563323650662586e-05,
      "loss": 0.0036,
      "step": 818
    },
    {
      "epoch": 0.7632805219012115,
      "grad_norm": 0.26792046427726746,
      "learning_rate": 1.4455166426596223e-05,
      "loss": 0.0053,
      "step": 819
    },
    {
      "epoch": 0.7642124883504194,
      "grad_norm": 0.1533597856760025,
      "learning_rate": 1.434734441843899e-05,
      "loss": 0.0032,
      "step": 820
    },
    {
      "epoch": 0.7651444547996272,
      "grad_norm": 0.13230851292610168,
      "learning_rate": 1.4239858643037751e-05,
      "loss": 0.0025,
      "step": 821
    },
    {
      "epoch": 0.766076421248835,
      "grad_norm": 0.42541301250457764,
      "learning_rate": 1.4132710114068426e-05,
      "loss": 0.0078,
      "step": 822
    },
    {
      "epoch": 0.7670083876980429,
      "grad_norm": 1.640199899673462,
      "learning_rate": 1.4025899842026442e-05,
      "loss": 0.0358,
      "step": 823
    },
    {
      "epoch": 0.7679403541472507,
      "grad_norm": 3.3744680881500244,
      "learning_rate": 1.3919428834217163e-05,
      "loss": 0.066,
      "step": 824
    },
    {
      "epoch": 0.7688723205964585,
      "grad_norm": 4.397242069244385,
      "learning_rate": 1.3813298094746491e-05,
      "loss": 0.1446,
      "step": 825
    },
    {
      "epoch": 0.7698042870456664,
      "grad_norm": 1.2552555799484253,
      "learning_rate": 1.3707508624511262e-05,
      "loss": 0.0247,
      "step": 826
    },
    {
      "epoch": 0.7707362534948742,
      "grad_norm": 0.9916906356811523,
      "learning_rate": 1.36020614211899e-05,
      "loss": 0.033,
      "step": 827
    },
    {
      "epoch": 0.771668219944082,
      "grad_norm": 1.0547466278076172,
      "learning_rate": 1.349695747923298e-05,
      "loss": 0.0706,
      "step": 828
    },
    {
      "epoch": 0.7726001863932899,
      "grad_norm": 0.8943658471107483,
      "learning_rate": 1.339219778985385e-05,
      "loss": 0.0239,
      "step": 829
    },
    {
      "epoch": 0.7735321528424977,
      "grad_norm": 1.5261614322662354,
      "learning_rate": 1.3287783341019278e-05,
      "loss": 0.0195,
      "step": 830
    },
    {
      "epoch": 0.7744641192917054,
      "grad_norm": 0.11581642925739288,
      "learning_rate": 1.3183715117440142e-05,
      "loss": 0.0019,
      "step": 831
    },
    {
      "epoch": 0.7753960857409133,
      "grad_norm": 0.190779909491539,
      "learning_rate": 1.3079994100562159e-05,
      "loss": 0.0051,
      "step": 832
    },
    {
      "epoch": 0.7763280521901211,
      "grad_norm": 0.7171970009803772,
      "learning_rate": 1.2976621268556571e-05,
      "loss": 0.0258,
      "step": 833
    },
    {
      "epoch": 0.777260018639329,
      "grad_norm": 1.5177792310714722,
      "learning_rate": 1.2873597596311027e-05,
      "loss": 0.0589,
      "step": 834
    },
    {
      "epoch": 0.7781919850885368,
      "grad_norm": 0.1316295564174652,
      "learning_rate": 1.2770924055420258e-05,
      "loss": 0.0026,
      "step": 835
    },
    {
      "epoch": 0.7791239515377446,
      "grad_norm": 0.5268474221229553,
      "learning_rate": 1.2668601614177017e-05,
      "loss": 0.0101,
      "step": 836
    },
    {
      "epoch": 0.7800559179869525,
      "grad_norm": 0.8330385684967041,
      "learning_rate": 1.2566631237562893e-05,
      "loss": 0.0351,
      "step": 837
    },
    {
      "epoch": 0.7809878844361603,
      "grad_norm": 0.4103114902973175,
      "learning_rate": 1.246501388723923e-05,
      "loss": 0.009,
      "step": 838
    },
    {
      "epoch": 0.7819198508853681,
      "grad_norm": 0.7705252766609192,
      "learning_rate": 1.2363750521538064e-05,
      "loss": 0.0211,
      "step": 839
    },
    {
      "epoch": 0.782851817334576,
      "grad_norm": 1.3898861408233643,
      "learning_rate": 1.2262842095453065e-05,
      "loss": 0.0678,
      "step": 840
    },
    {
      "epoch": 0.7837837837837838,
      "grad_norm": 2.382798910140991,
      "learning_rate": 1.2162289560630524e-05,
      "loss": 0.0201,
      "step": 841
    },
    {
      "epoch": 0.7847157502329916,
      "grad_norm": 0.8590038418769836,
      "learning_rate": 1.2062093865360458e-05,
      "loss": 0.0196,
      "step": 842
    },
    {
      "epoch": 0.7856477166821995,
      "grad_norm": 0.4806217849254608,
      "learning_rate": 1.1962255954567537e-05,
      "loss": 0.0081,
      "step": 843
    },
    {
      "epoch": 0.7865796831314072,
      "grad_norm": 0.6536293625831604,
      "learning_rate": 1.1862776769802276e-05,
      "loss": 0.01,
      "step": 844
    },
    {
      "epoch": 0.7875116495806151,
      "grad_norm": 1.8112365007400513,
      "learning_rate": 1.1763657249232107e-05,
      "loss": 0.0299,
      "step": 845
    },
    {
      "epoch": 0.7884436160298229,
      "grad_norm": 0.0851447582244873,
      "learning_rate": 1.1664898327632551e-05,
      "loss": 0.0015,
      "step": 846
    },
    {
      "epoch": 0.7893755824790307,
      "grad_norm": 3.3141989707946777,
      "learning_rate": 1.1566500936378388e-05,
      "loss": 0.0935,
      "step": 847
    },
    {
      "epoch": 0.7903075489282386,
      "grad_norm": 0.655958354473114,
      "learning_rate": 1.146846600343488e-05,
      "loss": 0.0092,
      "step": 848
    },
    {
      "epoch": 0.7912395153774464,
      "grad_norm": 2.1537182331085205,
      "learning_rate": 1.1370794453349037e-05,
      "loss": 0.0292,
      "step": 849
    },
    {
      "epoch": 0.7921714818266542,
      "grad_norm": 1.5979948043823242,
      "learning_rate": 1.1273487207240847e-05,
      "loss": 0.0682,
      "step": 850
    },
    {
      "epoch": 0.7931034482758621,
      "grad_norm": 0.32921919226646423,
      "learning_rate": 1.1176545182794674e-05,
      "loss": 0.0051,
      "step": 851
    },
    {
      "epoch": 0.7940354147250699,
      "grad_norm": 0.3104652762413025,
      "learning_rate": 1.1079969294250514e-05,
      "loss": 0.0053,
      "step": 852
    },
    {
      "epoch": 0.7949673811742777,
      "grad_norm": 0.23145949840545654,
      "learning_rate": 1.0983760452395413e-05,
      "loss": 0.0033,
      "step": 853
    },
    {
      "epoch": 0.7958993476234856,
      "grad_norm": 0.44747766852378845,
      "learning_rate": 1.0887919564554894e-05,
      "loss": 0.0089,
      "step": 854
    },
    {
      "epoch": 0.7968313140726934,
      "grad_norm": 2.918243169784546,
      "learning_rate": 1.079244753458437e-05,
      "loss": 0.0767,
      "step": 855
    },
    {
      "epoch": 0.7977632805219013,
      "grad_norm": 0.6843606233596802,
      "learning_rate": 1.0697345262860636e-05,
      "loss": 0.0144,
      "step": 856
    },
    {
      "epoch": 0.798695246971109,
      "grad_norm": 0.1016877144575119,
      "learning_rate": 1.0602613646273374e-05,
      "loss": 0.0033,
      "step": 857
    },
    {
      "epoch": 0.7996272134203168,
      "grad_norm": 0.7905264496803284,
      "learning_rate": 1.0508253578216693e-05,
      "loss": 0.0268,
      "step": 858
    },
    {
      "epoch": 0.8005591798695247,
      "grad_norm": 1.3822872638702393,
      "learning_rate": 1.0414265948580692e-05,
      "loss": 0.0685,
      "step": 859
    },
    {
      "epoch": 0.8014911463187325,
      "grad_norm": 0.5625484585762024,
      "learning_rate": 1.032065164374313e-05,
      "loss": 0.0123,
      "step": 860
    },
    {
      "epoch": 0.8024231127679403,
      "grad_norm": 0.05632966756820679,
      "learning_rate": 1.0227411546560962e-05,
      "loss": 0.001,
      "step": 861
    },
    {
      "epoch": 0.8033550792171482,
      "grad_norm": 0.7446116805076599,
      "learning_rate": 1.01345465363621e-05,
      "loss": 0.0202,
      "step": 862
    },
    {
      "epoch": 0.804287045666356,
      "grad_norm": 0.09283380210399628,
      "learning_rate": 1.0042057488937067e-05,
      "loss": 0.0016,
      "step": 863
    },
    {
      "epoch": 0.8052190121155638,
      "grad_norm": 0.27069157361984253,
      "learning_rate": 9.949945276530781e-06,
      "loss": 0.0053,
      "step": 864
    },
    {
      "epoch": 0.8061509785647717,
      "grad_norm": 0.8382067680358887,
      "learning_rate": 9.858210767834292e-06,
      "loss": 0.0226,
      "step": 865
    },
    {
      "epoch": 0.8070829450139795,
      "grad_norm": 1.0016965866088867,
      "learning_rate": 9.766854827976617e-06,
      "loss": 0.041,
      "step": 866
    },
    {
      "epoch": 0.8080149114631874,
      "grad_norm": 1.245935320854187,
      "learning_rate": 9.675878318516546e-06,
      "loss": 0.0404,
      "step": 867
    },
    {
      "epoch": 0.8089468779123952,
      "grad_norm": 0.5361421704292297,
      "learning_rate": 9.58528209743459e-06,
      "loss": 0.0076,
      "step": 868
    },
    {
      "epoch": 0.809878844361603,
      "grad_norm": 0.09631752967834473,
      "learning_rate": 9.495067019124792e-06,
      "loss": 0.0028,
      "step": 869
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.6793972253799438,
      "learning_rate": 9.405233934386726e-06,
      "loss": 0.0057,
      "step": 870
    },
    {
      "epoch": 0.8117427772600186,
      "grad_norm": 3.4450275897979736,
      "learning_rate": 9.315783690417479e-06,
      "loss": 0.053,
      "step": 871
    },
    {
      "epoch": 0.8126747437092264,
      "grad_norm": 1.280646800994873,
      "learning_rate": 9.226717130803636e-06,
      "loss": 0.042,
      "step": 872
    },
    {
      "epoch": 0.8136067101584343,
      "grad_norm": 1.587612509727478,
      "learning_rate": 9.138035095513336e-06,
      "loss": 0.0671,
      "step": 873
    },
    {
      "epoch": 0.8145386766076421,
      "grad_norm": 1.265060305595398,
      "learning_rate": 9.049738420888349e-06,
      "loss": 0.014,
      "step": 874
    },
    {
      "epoch": 0.8154706430568499,
      "grad_norm": 1.8084266185760498,
      "learning_rate": 8.961827939636196e-06,
      "loss": 0.0891,
      "step": 875
    },
    {
      "epoch": 0.8164026095060578,
      "grad_norm": 0.39547011256217957,
      "learning_rate": 8.87430448082227e-06,
      "loss": 0.0022,
      "step": 876
    },
    {
      "epoch": 0.8173345759552656,
      "grad_norm": 0.3031284213066101,
      "learning_rate": 8.787168869862066e-06,
      "loss": 0.0066,
      "step": 877
    },
    {
      "epoch": 0.8182665424044734,
      "grad_norm": 0.15621918439865112,
      "learning_rate": 8.700421928513352e-06,
      "loss": 0.0032,
      "step": 878
    },
    {
      "epoch": 0.8191985088536813,
      "grad_norm": 0.22168925404548645,
      "learning_rate": 8.614064474868421e-06,
      "loss": 0.0042,
      "step": 879
    },
    {
      "epoch": 0.8201304753028891,
      "grad_norm": 1.5389249324798584,
      "learning_rate": 8.528097323346407e-06,
      "loss": 0.0427,
      "step": 880
    },
    {
      "epoch": 0.821062441752097,
      "grad_norm": 1.5170341730117798,
      "learning_rate": 8.442521284685573e-06,
      "loss": 0.0335,
      "step": 881
    },
    {
      "epoch": 0.8219944082013048,
      "grad_norm": 0.08587615191936493,
      "learning_rate": 8.357337165935675e-06,
      "loss": 0.0023,
      "step": 882
    },
    {
      "epoch": 0.8229263746505125,
      "grad_norm": 0.17892524600028992,
      "learning_rate": 8.27254577045039e-06,
      "loss": 0.0027,
      "step": 883
    },
    {
      "epoch": 0.8238583410997204,
      "grad_norm": 0.17656829953193665,
      "learning_rate": 8.188147897879667e-06,
      "loss": 0.0046,
      "step": 884
    },
    {
      "epoch": 0.8247903075489282,
      "grad_norm": 0.5518335700035095,
      "learning_rate": 8.104144344162229e-06,
      "loss": 0.0124,
      "step": 885
    },
    {
      "epoch": 0.825722273998136,
      "grad_norm": 0.30466359853744507,
      "learning_rate": 8.02053590151805e-06,
      "loss": 0.0077,
      "step": 886
    },
    {
      "epoch": 0.8266542404473439,
      "grad_norm": 0.7674888372421265,
      "learning_rate": 7.937323358440935e-06,
      "loss": 0.0248,
      "step": 887
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.055480506271123886,
      "learning_rate": 7.854507499691005e-06,
      "loss": 0.0011,
      "step": 888
    },
    {
      "epoch": 0.8285181733457595,
      "grad_norm": 0.22888953983783722,
      "learning_rate": 7.772089106287344e-06,
      "loss": 0.0029,
      "step": 889
    },
    {
      "epoch": 0.8294501397949674,
      "grad_norm": 0.0735834464430809,
      "learning_rate": 7.690068955500624e-06,
      "loss": 0.0015,
      "step": 890
    },
    {
      "epoch": 0.8303821062441752,
      "grad_norm": 1.0018421411514282,
      "learning_rate": 7.608447820845771e-06,
      "loss": 0.0309,
      "step": 891
    },
    {
      "epoch": 0.8313140726933831,
      "grad_norm": 1.2352806329727173,
      "learning_rate": 7.527226472074678e-06,
      "loss": 0.0362,
      "step": 892
    },
    {
      "epoch": 0.8322460391425909,
      "grad_norm": 0.9229680299758911,
      "learning_rate": 7.446405675168938e-06,
      "loss": 0.0236,
      "step": 893
    },
    {
      "epoch": 0.8331780055917987,
      "grad_norm": 0.12361752986907959,
      "learning_rate": 7.365986192332625e-06,
      "loss": 0.0015,
      "step": 894
    },
    {
      "epoch": 0.8341099720410066,
      "grad_norm": 0.18358244001865387,
      "learning_rate": 7.285968781985092e-06,
      "loss": 0.0041,
      "step": 895
    },
    {
      "epoch": 0.8350419384902144,
      "grad_norm": 1.7125695943832397,
      "learning_rate": 7.206354198753862e-06,
      "loss": 0.0712,
      "step": 896
    },
    {
      "epoch": 0.8359739049394221,
      "grad_norm": 2.4467711448669434,
      "learning_rate": 7.127143193467445e-06,
      "loss": 0.0988,
      "step": 897
    },
    {
      "epoch": 0.83690587138863,
      "grad_norm": 2.2507357597351074,
      "learning_rate": 7.048336513148307e-06,
      "loss": 0.0513,
      "step": 898
    },
    {
      "epoch": 0.8378378378378378,
      "grad_norm": 2.246809720993042,
      "learning_rate": 6.969934901005809e-06,
      "loss": 0.0265,
      "step": 899
    },
    {
      "epoch": 0.8387698042870456,
      "grad_norm": 1.8378955125808716,
      "learning_rate": 6.8919390964291895e-06,
      "loss": 0.1061,
      "step": 900
    },
    {
      "epoch": 0.8397017707362535,
      "grad_norm": 0.47382405400276184,
      "learning_rate": 6.814349834980621e-06,
      "loss": 0.011,
      "step": 901
    },
    {
      "epoch": 0.8406337371854613,
      "grad_norm": 1.1276845932006836,
      "learning_rate": 6.7371678483882264e-06,
      "loss": 0.025,
      "step": 902
    },
    {
      "epoch": 0.8415657036346692,
      "grad_norm": 0.5048341155052185,
      "learning_rate": 6.660393864539222e-06,
      "loss": 0.0053,
      "step": 903
    },
    {
      "epoch": 0.842497670083877,
      "grad_norm": 0.9290698766708374,
      "learning_rate": 6.58402860747302e-06,
      "loss": 0.0501,
      "step": 904
    },
    {
      "epoch": 0.8434296365330848,
      "grad_norm": 1.0195822715759277,
      "learning_rate": 6.508072797374454e-06,
      "loss": 0.0182,
      "step": 905
    },
    {
      "epoch": 0.8443616029822927,
      "grad_norm": 0.4818008244037628,
      "learning_rate": 6.4325271505669025e-06,
      "loss": 0.0074,
      "step": 906
    },
    {
      "epoch": 0.8452935694315005,
      "grad_norm": 0.051283132284879684,
      "learning_rate": 6.3573923795056e-06,
      "loss": 0.0011,
      "step": 907
    },
    {
      "epoch": 0.8462255358807083,
      "grad_norm": 1.3800734281539917,
      "learning_rate": 6.282669192770896e-06,
      "loss": 0.0442,
      "step": 908
    },
    {
      "epoch": 0.8471575023299162,
      "grad_norm": 1.0327776670455933,
      "learning_rate": 6.208358295061572e-06,
      "loss": 0.037,
      "step": 909
    },
    {
      "epoch": 0.848089468779124,
      "grad_norm": 0.1635047346353531,
      "learning_rate": 6.134460387188206e-06,
      "loss": 0.003,
      "step": 910
    },
    {
      "epoch": 0.8490214352283317,
      "grad_norm": 0.17982259392738342,
      "learning_rate": 6.060976166066546e-06,
      "loss": 0.0032,
      "step": 911
    },
    {
      "epoch": 0.8499534016775396,
      "grad_norm": 0.2328854650259018,
      "learning_rate": 5.9879063247109405e-06,
      "loss": 0.0049,
      "step": 912
    },
    {
      "epoch": 0.8508853681267474,
      "grad_norm": 1.2415688037872314,
      "learning_rate": 5.915251552227829e-06,
      "loss": 0.0307,
      "step": 913
    },
    {
      "epoch": 0.8518173345759553,
      "grad_norm": 0.3968764841556549,
      "learning_rate": 5.8430125338092115e-06,
      "loss": 0.0087,
      "step": 914
    },
    {
      "epoch": 0.8527493010251631,
      "grad_norm": 1.2752315998077393,
      "learning_rate": 5.771189950726191e-06,
      "loss": 0.0312,
      "step": 915
    },
    {
      "epoch": 0.8536812674743709,
      "grad_norm": 0.7192822098731995,
      "learning_rate": 5.699784480322568e-06,
      "loss": 0.0108,
      "step": 916
    },
    {
      "epoch": 0.8546132339235788,
      "grad_norm": 0.34972473978996277,
      "learning_rate": 5.628796796008434e-06,
      "loss": 0.0029,
      "step": 917
    },
    {
      "epoch": 0.8555452003727866,
      "grad_norm": 0.3486124277114868,
      "learning_rate": 5.558227567253832e-06,
      "loss": 0.0051,
      "step": 918
    },
    {
      "epoch": 0.8564771668219944,
      "grad_norm": 0.3966856300830841,
      "learning_rate": 5.488077459582425e-06,
      "loss": 0.0074,
      "step": 919
    },
    {
      "epoch": 0.8574091332712023,
      "grad_norm": 0.9853683114051819,
      "learning_rate": 5.418347134565249e-06,
      "loss": 0.0151,
      "step": 920
    },
    {
      "epoch": 0.8583410997204101,
      "grad_norm": 0.6727381348609924,
      "learning_rate": 5.349037249814442e-06,
      "loss": 0.0114,
      "step": 921
    },
    {
      "epoch": 0.8592730661696178,
      "grad_norm": 1.942173719406128,
      "learning_rate": 5.28014845897708e-06,
      "loss": 0.0432,
      "step": 922
    },
    {
      "epoch": 0.8602050326188257,
      "grad_norm": 0.726831316947937,
      "learning_rate": 5.211681411728969e-06,
      "loss": 0.0217,
      "step": 923
    },
    {
      "epoch": 0.8611369990680335,
      "grad_norm": 1.3569886684417725,
      "learning_rate": 5.143636753768549e-06,
      "loss": 0.0124,
      "step": 924
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 1.4848052263259888,
      "learning_rate": 5.076015126810784e-06,
      "loss": 0.0913,
      "step": 925
    },
    {
      "epoch": 0.8630009319664492,
      "grad_norm": 0.08492793142795563,
      "learning_rate": 5.008817168581137e-06,
      "loss": 0.0021,
      "step": 926
    },
    {
      "epoch": 0.863932898415657,
      "grad_norm": 0.3844028115272522,
      "learning_rate": 4.94204351280953e-06,
      "loss": 0.0082,
      "step": 927
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 0.3236229717731476,
      "learning_rate": 4.8756947892243725e-06,
      "loss": 0.0026,
      "step": 928
    },
    {
      "epoch": 0.8657968313140727,
      "grad_norm": 0.13389159739017487,
      "learning_rate": 4.809771623546627e-06,
      "loss": 0.0024,
      "step": 929
    },
    {
      "epoch": 0.8667287977632805,
      "grad_norm": 0.4836329519748688,
      "learning_rate": 4.744274637483936e-06,
      "loss": 0.0089,
      "step": 930
    },
    {
      "epoch": 0.8676607642124884,
      "grad_norm": 0.7628198862075806,
      "learning_rate": 4.6792044487247e-06,
      "loss": 0.0091,
      "step": 931
    },
    {
      "epoch": 0.8685927306616962,
      "grad_norm": 0.9907669425010681,
      "learning_rate": 4.614561670932288e-06,
      "loss": 0.0322,
      "step": 932
    },
    {
      "epoch": 0.869524697110904,
      "grad_norm": 0.8017989993095398,
      "learning_rate": 4.550346913739256e-06,
      "loss": 0.0208,
      "step": 933
    },
    {
      "epoch": 0.8704566635601119,
      "grad_norm": 0.680810272693634,
      "learning_rate": 4.486560782741578e-06,
      "loss": 0.0122,
      "step": 934
    },
    {
      "epoch": 0.8713886300093197,
      "grad_norm": 2.3900279998779297,
      "learning_rate": 4.423203879492943e-06,
      "loss": 0.0444,
      "step": 935
    },
    {
      "epoch": 0.8723205964585274,
      "grad_norm": 0.30623510479927063,
      "learning_rate": 4.360276801499086e-06,
      "loss": 0.006,
      "step": 936
    },
    {
      "epoch": 0.8732525629077353,
      "grad_norm": 0.19743725657463074,
      "learning_rate": 4.29778014221216e-06,
      "loss": 0.004,
      "step": 937
    },
    {
      "epoch": 0.8741845293569431,
      "grad_norm": 0.06322675198316574,
      "learning_rate": 4.2357144910251e-06,
      "loss": 0.0012,
      "step": 938
    },
    {
      "epoch": 0.875116495806151,
      "grad_norm": 0.14598406851291656,
      "learning_rate": 4.174080433266136e-06,
      "loss": 0.0035,
      "step": 939
    },
    {
      "epoch": 0.8760484622553588,
      "grad_norm": 0.20462970435619354,
      "learning_rate": 4.112878550193194e-06,
      "loss": 0.0028,
      "step": 940
    },
    {
      "epoch": 0.8769804287045666,
      "grad_norm": 1.6550384759902954,
      "learning_rate": 4.05210941898847e-06,
      "loss": 0.0479,
      "step": 941
    },
    {
      "epoch": 0.8779123951537745,
      "grad_norm": 0.5954486131668091,
      "learning_rate": 3.991773612752952e-06,
      "loss": 0.0106,
      "step": 942
    },
    {
      "epoch": 0.8788443616029823,
      "grad_norm": 0.2953099310398102,
      "learning_rate": 3.9318717005010495e-06,
      "loss": 0.0049,
      "step": 943
    },
    {
      "epoch": 0.8797763280521901,
      "grad_norm": 0.3062129020690918,
      "learning_rate": 3.872404247155193e-06,
      "loss": 0.0024,
      "step": 944
    },
    {
      "epoch": 0.880708294501398,
      "grad_norm": 0.44278305768966675,
      "learning_rate": 3.8133718135405285e-06,
      "loss": 0.0084,
      "step": 945
    },
    {
      "epoch": 0.8816402609506058,
      "grad_norm": 0.15503761172294617,
      "learning_rate": 3.754774956379614e-06,
      "loss": 0.0025,
      "step": 946
    },
    {
      "epoch": 0.8825722273998136,
      "grad_norm": 0.15528859198093414,
      "learning_rate": 3.696614228287187e-06,
      "loss": 0.0022,
      "step": 947
    },
    {
      "epoch": 0.8835041938490215,
      "grad_norm": 0.3577142655849457,
      "learning_rate": 3.6388901777649477e-06,
      "loss": 0.006,
      "step": 948
    },
    {
      "epoch": 0.8844361602982292,
      "grad_norm": 1.3068500757217407,
      "learning_rate": 3.581603349196372e-06,
      "loss": 0.0105,
      "step": 949
    },
    {
      "epoch": 0.8853681267474371,
      "grad_norm": 1.0203229188919067,
      "learning_rate": 3.5247542828415747e-06,
      "loss": 0.0584,
      "step": 950
    },
    {
      "epoch": 0.8863000931966449,
      "grad_norm": 1.07167387008667,
      "learning_rate": 3.468343514832251e-06,
      "loss": 0.0127,
      "step": 951
    },
    {
      "epoch": 0.8872320596458527,
      "grad_norm": 0.17283280193805695,
      "learning_rate": 3.4123715771665786e-06,
      "loss": 0.0023,
      "step": 952
    },
    {
      "epoch": 0.8881640260950606,
      "grad_norm": 0.06770434975624084,
      "learning_rate": 3.356838997704226e-06,
      "loss": 0.0018,
      "step": 953
    },
    {
      "epoch": 0.8890959925442684,
      "grad_norm": 0.20179446041584015,
      "learning_rate": 3.3017463001613626e-06,
      "loss": 0.0036,
      "step": 954
    },
    {
      "epoch": 0.8900279589934762,
      "grad_norm": 1.307664394378662,
      "learning_rate": 3.2470940041057106e-06,
      "loss": 0.0415,
      "step": 955
    },
    {
      "epoch": 0.8909599254426841,
      "grad_norm": 0.3763882517814636,
      "learning_rate": 3.1928826249516987e-06,
      "loss": 0.008,
      "step": 956
    },
    {
      "epoch": 0.8918918918918919,
      "grad_norm": 0.4271828532218933,
      "learning_rate": 3.139112673955513e-06,
      "loss": 0.0088,
      "step": 957
    },
    {
      "epoch": 0.8928238583410997,
      "grad_norm": 0.5596781969070435,
      "learning_rate": 3.08578465821035e-06,
      "loss": 0.0064,
      "step": 958
    },
    {
      "epoch": 0.8937558247903076,
      "grad_norm": 0.9955370426177979,
      "learning_rate": 3.0328990806415934e-06,
      "loss": 0.0293,
      "step": 959
    },
    {
      "epoch": 0.8946877912395154,
      "grad_norm": 0.6744177341461182,
      "learning_rate": 2.9804564400020994e-06,
      "loss": 0.0115,
      "step": 960
    },
    {
      "epoch": 0.8956197576887233,
      "grad_norm": 1.0342707633972168,
      "learning_rate": 2.9284572308674596e-06,
      "loss": 0.0688,
      "step": 961
    },
    {
      "epoch": 0.896551724137931,
      "grad_norm": 0.36630138754844666,
      "learning_rate": 2.8769019436313715e-06,
      "loss": 0.0069,
      "step": 962
    },
    {
      "epoch": 0.8974836905871388,
      "grad_norm": 0.9983659982681274,
      "learning_rate": 2.825791064500993e-06,
      "loss": 0.032,
      "step": 963
    },
    {
      "epoch": 0.8984156570363467,
      "grad_norm": 0.27666184306144714,
      "learning_rate": 2.7751250754923573e-06,
      "loss": 0.0053,
      "step": 964
    },
    {
      "epoch": 0.8993476234855545,
      "grad_norm": 0.03350253403186798,
      "learning_rate": 2.7249044544258363e-06,
      "loss": 0.0009,
      "step": 965
    },
    {
      "epoch": 0.9002795899347623,
      "grad_norm": 0.501646876335144,
      "learning_rate": 2.675129674921639e-06,
      "loss": 0.0064,
      "step": 966
    },
    {
      "epoch": 0.9012115563839702,
      "grad_norm": 0.07862615585327148,
      "learning_rate": 2.625801206395312e-06,
      "loss": 0.0015,
      "step": 967
    },
    {
      "epoch": 0.902143522833178,
      "grad_norm": 0.849259078502655,
      "learning_rate": 2.576919514053355e-06,
      "loss": 0.0202,
      "step": 968
    },
    {
      "epoch": 0.9030754892823858,
      "grad_norm": 0.0796375498175621,
      "learning_rate": 2.528485058888813e-06,
      "loss": 0.0017,
      "step": 969
    },
    {
      "epoch": 0.9040074557315937,
      "grad_norm": 1.963661789894104,
      "learning_rate": 2.48049829767692e-06,
      "loss": 0.0798,
      "step": 970
    },
    {
      "epoch": 0.9049394221808015,
      "grad_norm": 0.7925097346305847,
      "learning_rate": 2.4329596829708144e-06,
      "loss": 0.0099,
      "step": 971
    },
    {
      "epoch": 0.9058713886300093,
      "grad_norm": 0.11906258761882782,
      "learning_rate": 2.385869663097251e-06,
      "loss": 0.002,
      "step": 972
    },
    {
      "epoch": 0.9068033550792172,
      "grad_norm": 1.8423750400543213,
      "learning_rate": 2.3392286821523724e-06,
      "loss": 0.0678,
      "step": 973
    },
    {
      "epoch": 0.907735321528425,
      "grad_norm": 0.6542656421661377,
      "learning_rate": 2.2930371799975594e-06,
      "loss": 0.0111,
      "step": 974
    },
    {
      "epoch": 0.9086672879776329,
      "grad_norm": 1.2632553577423096,
      "learning_rate": 2.2472955922552163e-06,
      "loss": 0.0662,
      "step": 975
    },
    {
      "epoch": 0.9095992544268406,
      "grad_norm": 0.9758872985839844,
      "learning_rate": 2.202004350304715e-06,
      "loss": 0.0104,
      "step": 976
    },
    {
      "epoch": 0.9105312208760484,
      "grad_norm": 0.615419864654541,
      "learning_rate": 2.157163881278312e-06,
      "loss": 0.0131,
      "step": 977
    },
    {
      "epoch": 0.9114631873252563,
      "grad_norm": 0.07674716413021088,
      "learning_rate": 2.112774608057111e-06,
      "loss": 0.001,
      "step": 978
    },
    {
      "epoch": 0.9123951537744641,
      "grad_norm": 0.2658063471317291,
      "learning_rate": 2.068836949267089e-06,
      "loss": 0.0017,
      "step": 979
    },
    {
      "epoch": 0.9133271202236719,
      "grad_norm": 1.0758992433547974,
      "learning_rate": 2.0253513192751373e-06,
      "loss": 0.0118,
      "step": 980
    },
    {
      "epoch": 0.9142590866728798,
      "grad_norm": 0.5867937803268433,
      "learning_rate": 1.982318128185151e-06,
      "loss": 0.0053,
      "step": 981
    },
    {
      "epoch": 0.9151910531220876,
      "grad_norm": 0.7973155379295349,
      "learning_rate": 1.9397377818341944e-06,
      "loss": 0.0145,
      "step": 982
    },
    {
      "epoch": 0.9161230195712954,
      "grad_norm": 0.16484799981117249,
      "learning_rate": 1.8976106817886196e-06,
      "loss": 0.0033,
      "step": 983
    },
    {
      "epoch": 0.9170549860205033,
      "grad_norm": 0.21485017240047455,
      "learning_rate": 1.855937225340315e-06,
      "loss": 0.0038,
      "step": 984
    },
    {
      "epoch": 0.9179869524697111,
      "grad_norm": 0.0540773943066597,
      "learning_rate": 1.8147178055029579e-06,
      "loss": 0.0012,
      "step": 985
    },
    {
      "epoch": 0.918918918918919,
      "grad_norm": 0.10457666218280792,
      "learning_rate": 1.7739528110083004e-06,
      "loss": 0.0014,
      "step": 986
    },
    {
      "epoch": 0.9198508853681268,
      "grad_norm": 0.07089216262102127,
      "learning_rate": 1.7336426263024897e-06,
      "loss": 0.0012,
      "step": 987
    },
    {
      "epoch": 0.9207828518173345,
      "grad_norm": 0.041863564401865005,
      "learning_rate": 1.6937876315424705e-06,
      "loss": 0.0008,
      "step": 988
    },
    {
      "epoch": 0.9217148182665424,
      "grad_norm": 0.1343851238489151,
      "learning_rate": 1.6543882025923886e-06,
      "loss": 0.0024,
      "step": 989
    },
    {
      "epoch": 0.9226467847157502,
      "grad_norm": 0.04936739057302475,
      "learning_rate": 1.6154447110200255e-06,
      "loss": 0.0008,
      "step": 990
    },
    {
      "epoch": 0.923578751164958,
      "grad_norm": 0.6000242233276367,
      "learning_rate": 1.5769575240933422e-06,
      "loss": 0.0095,
      "step": 991
    },
    {
      "epoch": 0.9245107176141659,
      "grad_norm": 0.7469452619552612,
      "learning_rate": 1.5389270047769578e-06,
      "loss": 0.0163,
      "step": 992
    },
    {
      "epoch": 0.9254426840633737,
      "grad_norm": 0.09905730932950974,
      "learning_rate": 1.5013535117287647e-06,
      "loss": 0.0014,
      "step": 993
    },
    {
      "epoch": 0.9263746505125815,
      "grad_norm": 0.9133400917053223,
      "learning_rate": 1.4642373992965364e-06,
      "loss": 0.0117,
      "step": 994
    },
    {
      "epoch": 0.9273066169617894,
      "grad_norm": 1.5601509809494019,
      "learning_rate": 1.4275790175145908e-06,
      "loss": 0.0671,
      "step": 995
    },
    {
      "epoch": 0.9282385834109972,
      "grad_norm": 1.1153135299682617,
      "learning_rate": 1.3913787121004719e-06,
      "loss": 0.0184,
      "step": 996
    },
    {
      "epoch": 0.9291705498602051,
      "grad_norm": 1.7211925983428955,
      "learning_rate": 1.3556368244517114e-06,
      "loss": 0.0227,
      "step": 997
    },
    {
      "epoch": 0.9301025163094129,
      "grad_norm": 0.985527753829956,
      "learning_rate": 1.3203536916425841e-06,
      "loss": 0.0273,
      "step": 998
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 1.213148832321167,
      "learning_rate": 1.2855296464209688e-06,
      "loss": 0.0441,
      "step": 999
    },
    {
      "epoch": 0.9319664492078286,
      "grad_norm": 1.5918529033660889,
      "learning_rate": 1.2511650172051637e-06,
      "loss": 0.0944,
      "step": 1000
    },
    {
      "epoch": 0.9328984156570364,
      "grad_norm": 0.06392987817525864,
      "learning_rate": 1.217260128080816e-06,
      "loss": 0.0012,
      "step": 1001
    },
    {
      "epoch": 0.9338303821062441,
      "grad_norm": 1.1391026973724365,
      "learning_rate": 1.1838152987978578e-06,
      "loss": 0.0308,
      "step": 1002
    },
    {
      "epoch": 0.934762348555452,
      "grad_norm": 0.05295366048812866,
      "learning_rate": 1.1508308447674975e-06,
      "loss": 0.0008,
      "step": 1003
    },
    {
      "epoch": 0.9356943150046598,
      "grad_norm": 1.0572638511657715,
      "learning_rate": 1.1183070770592441e-06,
      "loss": 0.0262,
      "step": 1004
    },
    {
      "epoch": 0.9366262814538676,
      "grad_norm": 0.37860023975372314,
      "learning_rate": 1.0862443023979651e-06,
      "loss": 0.0069,
      "step": 1005
    },
    {
      "epoch": 0.9375582479030755,
      "grad_norm": 0.15000490844249725,
      "learning_rate": 1.0546428231609895e-06,
      "loss": 0.0035,
      "step": 1006
    },
    {
      "epoch": 0.9384902143522833,
      "grad_norm": 0.05710078775882721,
      "learning_rate": 1.0235029373752758e-06,
      "loss": 0.001,
      "step": 1007
    },
    {
      "epoch": 0.9394221808014911,
      "grad_norm": 0.5820727348327637,
      "learning_rate": 9.928249387145982e-07,
      "loss": 0.0101,
      "step": 1008
    },
    {
      "epoch": 0.940354147250699,
      "grad_norm": 0.19106589257717133,
      "learning_rate": 9.6260911649676e-07,
      "loss": 0.0041,
      "step": 1009
    },
    {
      "epoch": 0.9412861136999068,
      "grad_norm": 0.3644753396511078,
      "learning_rate": 9.32855755680867e-07,
      "loss": 0.0098,
      "step": 1010
    },
    {
      "epoch": 0.9422180801491147,
      "grad_norm": 0.2526366114616394,
      "learning_rate": 9.035651368646648e-07,
      "loss": 0.0055,
      "step": 1011
    },
    {
      "epoch": 0.9431500465983225,
      "grad_norm": 0.6924790143966675,
      "learning_rate": 8.747375362818666e-07,
      "loss": 0.0188,
      "step": 1012
    },
    {
      "epoch": 0.9440820130475303,
      "grad_norm": 1.0197707414627075,
      "learning_rate": 8.463732257995571e-07,
      "loss": 0.0265,
      "step": 1013
    },
    {
      "epoch": 0.9450139794967382,
      "grad_norm": 0.5194602608680725,
      "learning_rate": 8.184724729156379e-07,
      "loss": 0.0094,
      "step": 1014
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.5754898190498352,
      "learning_rate": 7.910355407562742e-07,
      "loss": 0.0105,
      "step": 1015
    },
    {
      "epoch": 0.9468779123951537,
      "grad_norm": 0.28730660676956177,
      "learning_rate": 7.640626880734581e-07,
      "loss": 0.0034,
      "step": 1016
    },
    {
      "epoch": 0.9478098788443616,
      "grad_norm": 1.0178496837615967,
      "learning_rate": 7.375541692425326e-07,
      "loss": 0.0199,
      "step": 1017
    },
    {
      "epoch": 0.9487418452935694,
      "grad_norm": 0.2701060473918915,
      "learning_rate": 7.115102342598101e-07,
      "loss": 0.003,
      "step": 1018
    },
    {
      "epoch": 0.9496738117427772,
      "grad_norm": 0.5945216417312622,
      "learning_rate": 6.859311287402081e-07,
      "loss": 0.0105,
      "step": 1019
    },
    {
      "epoch": 0.9506057781919851,
      "grad_norm": 0.7214066982269287,
      "learning_rate": 6.608170939149283e-07,
      "loss": 0.0132,
      "step": 1020
    },
    {
      "epoch": 0.9515377446411929,
      "grad_norm": 0.11991623789072037,
      "learning_rate": 6.361683666291973e-07,
      "loss": 0.0031,
      "step": 1021
    },
    {
      "epoch": 0.9524697110904008,
      "grad_norm": 0.24525563418865204,
      "learning_rate": 6.119851793400189e-07,
      "loss": 0.0026,
      "step": 1022
    },
    {
      "epoch": 0.9534016775396086,
      "grad_norm": 1.5672634840011597,
      "learning_rate": 5.882677601139919e-07,
      "loss": 0.0275,
      "step": 1023
    },
    {
      "epoch": 0.9543336439888164,
      "grad_norm": 0.9412722587585449,
      "learning_rate": 5.650163326251345e-07,
      "loss": 0.0178,
      "step": 1024
    },
    {
      "epoch": 0.9552656104380243,
      "grad_norm": 0.39089635014533997,
      "learning_rate": 5.422311161528193e-07,
      "loss": 0.0224,
      "step": 1025
    },
    {
      "epoch": 0.9561975768872321,
      "grad_norm": 0.03957606479525566,
      "learning_rate": 5.199123255796634e-07,
      "loss": 0.0009,
      "step": 1026
    },
    {
      "epoch": 0.9571295433364398,
      "grad_norm": 1.1148029565811157,
      "learning_rate": 4.980601713895305e-07,
      "loss": 0.023,
      "step": 1027
    },
    {
      "epoch": 0.9580615097856477,
      "grad_norm": 0.12970255315303802,
      "learning_rate": 4.7667485966552685e-07,
      "loss": 0.0021,
      "step": 1028
    },
    {
      "epoch": 0.9589934762348555,
      "grad_norm": 0.9582773447036743,
      "learning_rate": 4.55756592088058e-07,
      "loss": 0.0134,
      "step": 1029
    },
    {
      "epoch": 0.9599254426840633,
      "grad_norm": 0.3440712094306946,
      "learning_rate": 4.353055659329419e-07,
      "loss": 0.0089,
      "step": 1030
    },
    {
      "epoch": 0.9608574091332712,
      "grad_norm": 0.0910550057888031,
      "learning_rate": 4.153219740695435e-07,
      "loss": 0.0022,
      "step": 1031
    },
    {
      "epoch": 0.961789375582479,
      "grad_norm": 0.5724707841873169,
      "learning_rate": 3.958060049589485e-07,
      "loss": 0.008,
      "step": 1032
    },
    {
      "epoch": 0.9627213420316869,
      "grad_norm": 0.7248005867004395,
      "learning_rate": 3.767578426521923e-07,
      "loss": 0.0209,
      "step": 1033
    },
    {
      "epoch": 0.9636533084808947,
      "grad_norm": 1.4421683549880981,
      "learning_rate": 3.5817766678850615e-07,
      "loss": 0.0059,
      "step": 1034
    },
    {
      "epoch": 0.9645852749301025,
      "grad_norm": 0.555047869682312,
      "learning_rate": 3.4006565259366295e-07,
      "loss": 0.0095,
      "step": 1035
    },
    {
      "epoch": 0.9655172413793104,
      "grad_norm": 0.6058024764060974,
      "learning_rate": 3.224219708782894e-07,
      "loss": 0.0113,
      "step": 1036
    },
    {
      "epoch": 0.9664492078285182,
      "grad_norm": 0.9548462629318237,
      "learning_rate": 3.052467880362675e-07,
      "loss": 0.0218,
      "step": 1037
    },
    {
      "epoch": 0.967381174277726,
      "grad_norm": 0.2721622586250305,
      "learning_rate": 2.88540266043158e-07,
      "loss": 0.0058,
      "step": 1038
    },
    {
      "epoch": 0.9683131407269339,
      "grad_norm": 0.2145017385482788,
      "learning_rate": 2.723025624546849e-07,
      "loss": 0.0034,
      "step": 1039
    },
    {
      "epoch": 0.9692451071761417,
      "grad_norm": 0.4106197655200958,
      "learning_rate": 2.5653383040524227e-07,
      "loss": 0.0055,
      "step": 1040
    },
    {
      "epoch": 0.9701770736253494,
      "grad_norm": 0.16109812259674072,
      "learning_rate": 2.412342186064565e-07,
      "loss": 0.0024,
      "step": 1041
    },
    {
      "epoch": 0.9711090400745573,
      "grad_norm": 1.1901835203170776,
      "learning_rate": 2.2640387134577058e-07,
      "loss": 0.0369,
      "step": 1042
    },
    {
      "epoch": 0.9720410065237651,
      "grad_norm": 0.8408852815628052,
      "learning_rate": 2.1204292848509555e-07,
      "loss": 0.019,
      "step": 1043
    },
    {
      "epoch": 0.972972972972973,
      "grad_norm": 0.14452603459358215,
      "learning_rate": 1.98151525459489e-07,
      "loss": 0.0023,
      "step": 1044
    },
    {
      "epoch": 0.9739049394221808,
      "grad_norm": 0.34769976139068604,
      "learning_rate": 1.8472979327587292e-07,
      "loss": 0.0066,
      "step": 1045
    },
    {
      "epoch": 0.9748369058713886,
      "grad_norm": 0.8973527550697327,
      "learning_rate": 1.717778585118013e-07,
      "loss": 0.0171,
      "step": 1046
    },
    {
      "epoch": 0.9757688723205965,
      "grad_norm": 0.06965801864862442,
      "learning_rate": 1.5929584331427216e-07,
      "loss": 0.0011,
      "step": 1047
    },
    {
      "epoch": 0.9767008387698043,
      "grad_norm": 0.5106223225593567,
      "learning_rate": 1.4728386539856754e-07,
      "loss": 0.0086,
      "step": 1048
    },
    {
      "epoch": 0.9776328052190121,
      "grad_norm": 1.0167614221572876,
      "learning_rate": 1.3574203804713747e-07,
      "loss": 0.0193,
      "step": 1049
    },
    {
      "epoch": 0.97856477166822,
      "grad_norm": 1.6071141958236694,
      "learning_rate": 1.246704701085566e-07,
      "loss": 0.0923,
      "step": 1050
    },
    {
      "epoch": 0.9794967381174278,
      "grad_norm": 0.06256900727748871,
      "learning_rate": 1.1406926599646373e-07,
      "loss": 0.001,
      "step": 1051
    },
    {
      "epoch": 0.9804287045666356,
      "grad_norm": 0.6707242727279663,
      "learning_rate": 1.0393852568860719e-07,
      "loss": 0.0157,
      "step": 1052
    },
    {
      "epoch": 0.9813606710158435,
      "grad_norm": 0.9473837614059448,
      "learning_rate": 9.427834472588992e-08,
      "loss": 0.0473,
      "step": 1053
    },
    {
      "epoch": 0.9822926374650512,
      "grad_norm": 0.40152788162231445,
      "learning_rate": 8.508881421145365e-08,
      "loss": 0.0091,
      "step": 1054
    },
    {
      "epoch": 0.983224603914259,
      "grad_norm": 0.11309801042079926,
      "learning_rate": 7.637002080985168e-08,
      "loss": 0.002,
      "step": 1055
    },
    {
      "epoch": 0.9841565703634669,
      "grad_norm": 0.1226629987359047,
      "learning_rate": 6.812204674619959e-08,
      "loss": 0.0025,
      "step": 1056
    },
    {
      "epoch": 0.9850885368126747,
      "grad_norm": 0.6275689601898193,
      "learning_rate": 6.034496980542037e-08,
      "loss": 0.0149,
      "step": 1057
    },
    {
      "epoch": 0.9860205032618826,
      "grad_norm": 0.6120526194572449,
      "learning_rate": 5.303886333151153e-08,
      "loss": 0.0088,
      "step": 1058
    },
    {
      "epoch": 0.9869524697110904,
      "grad_norm": 0.8378371000289917,
      "learning_rate": 4.620379622682358e-08,
      "loss": 0.0166,
      "step": 1059
    },
    {
      "epoch": 0.9878844361602982,
      "grad_norm": 0.06605886667966843,
      "learning_rate": 3.9839832951465986e-08,
      "loss": 0.0012,
      "step": 1060
    },
    {
      "epoch": 0.9888164026095061,
      "grad_norm": 0.9707759618759155,
      "learning_rate": 3.394703352263551e-08,
      "loss": 0.0434,
      "step": 1061
    },
    {
      "epoch": 0.9897483690587139,
      "grad_norm": 0.12164362519979477,
      "learning_rate": 2.8525453514099965e-08,
      "loss": 0.0018,
      "step": 1062
    },
    {
      "epoch": 0.9906803355079217,
      "grad_norm": 0.4129875898361206,
      "learning_rate": 2.3575144055643094e-08,
      "loss": 0.0093,
      "step": 1063
    },
    {
      "epoch": 0.9916123019571296,
      "grad_norm": 0.6688860654830933,
      "learning_rate": 1.9096151832609375e-08,
      "loss": 0.0118,
      "step": 1064
    },
    {
      "epoch": 0.9925442684063374,
      "grad_norm": 0.1448785960674286,
      "learning_rate": 1.5088519085437734e-08,
      "loss": 0.0021,
      "step": 1065
    },
    {
      "epoch": 0.9934762348555451,
      "grad_norm": 0.17577403783798218,
      "learning_rate": 1.1552283609272962e-08,
      "loss": 0.0036,
      "step": 1066
    },
    {
      "epoch": 0.994408201304753,
      "grad_norm": 0.0788675993680954,
      "learning_rate": 8.487478753615997e-09,
      "loss": 0.0016,
      "step": 1067
    },
    {
      "epoch": 0.9953401677539608,
      "grad_norm": 0.3536447286605835,
      "learning_rate": 5.894133422001957e-09,
      "loss": 0.004,
      "step": 1068
    },
    {
      "epoch": 0.9962721342031687,
      "grad_norm": 1.099785327911377,
      "learning_rate": 3.772272071722593e-09,
      "loss": 0.016,
      "step": 1069
    },
    {
      "epoch": 0.9972041006523765,
      "grad_norm": 1.2353144884109497,
      "learning_rate": 2.1219147136264382e-09,
      "loss": 0.024,
      "step": 1070
    },
    {
      "epoch": 0.9981360671015843,
      "grad_norm": 0.8307344317436218,
      "learning_rate": 9.43076911874563e-10,
      "loss": 0.0139,
      "step": 1071
    },
    {
      "epoch": 0.9990680335507922,
      "grad_norm": 2.4767701625823975,
      "learning_rate": 2.3576978384065584e-10,
      "loss": 0.0363,
      "step": 1072
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.866561770439148,
      "learning_rate": 0.0,
      "loss": 0.0964,
      "step": 1073
    }
  ],
  "logging_steps": 1,
  "max_steps": 1073,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 269,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5471679295717376e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}