{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0045385779122542,
  "eval_steps": 500,
  "global_step": 166,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006051437216338881,
      "grad_norm": 0.23612292110919952,
      "learning_rate": 2e-05,
      "loss": 1.6094,
      "step": 1
    },
    {
      "epoch": 0.012102874432677761,
      "grad_norm": 0.21936853229999542,
      "learning_rate": 4e-05,
      "loss": 1.7958,
      "step": 2
    },
    {
      "epoch": 0.018154311649016642,
      "grad_norm": 0.2684606909751892,
      "learning_rate": 6e-05,
      "loss": 2.3944,
      "step": 3
    },
    {
      "epoch": 0.024205748865355523,
      "grad_norm": 0.2940954267978668,
      "learning_rate": 8e-05,
      "loss": 2.3674,
      "step": 4
    },
    {
      "epoch": 0.030257186081694403,
      "grad_norm": 0.2846325933933258,
      "learning_rate": 0.0001,
      "loss": 2.1885,
      "step": 5
    },
    {
      "epoch": 0.036308623298033284,
      "grad_norm": 0.3098210096359253,
      "learning_rate": 9.999048137490364e-05,
      "loss": 1.8546,
      "step": 6
    },
    {
      "epoch": 0.04236006051437216,
      "grad_norm": 0.33546316623687744,
      "learning_rate": 9.99619291237835e-05,
      "loss": 2.073,
      "step": 7
    },
    {
      "epoch": 0.048411497730711045,
      "grad_norm": 0.3401203453540802,
      "learning_rate": 9.991435411776654e-05,
      "loss": 1.9222,
      "step": 8
    },
    {
      "epoch": 0.05446293494704992,
      "grad_norm": 0.48030051589012146,
      "learning_rate": 9.984777447079862e-05,
      "loss": 2.4656,
      "step": 9
    },
    {
      "epoch": 0.060514372163388806,
      "grad_norm": 0.41418835520744324,
      "learning_rate": 9.976221553274767e-05,
      "loss": 2.5115,
      "step": 10
    },
    {
      "epoch": 0.06656580937972768,
      "grad_norm": 0.5238416790962219,
      "learning_rate": 9.965770987975189e-05,
      "loss": 2.4795,
      "step": 11
    },
    {
      "epoch": 0.07261724659606657,
      "grad_norm": 0.48317399621009827,
      "learning_rate": 9.953429730181653e-05,
      "loss": 2.505,
      "step": 12
    },
    {
      "epoch": 0.07866868381240545,
      "grad_norm": 0.6496583223342896,
      "learning_rate": 9.939202478766407e-05,
      "loss": 2.6298,
      "step": 13
    },
    {
      "epoch": 0.08472012102874432,
      "grad_norm": 0.7371865510940552,
      "learning_rate": 9.923094650684345e-05,
      "loss": 2.5862,
      "step": 14
    },
    {
      "epoch": 0.0907715582450832,
      "grad_norm": 0.7955237030982971,
      "learning_rate": 9.905112378910532e-05,
      "loss": 2.6021,
      "step": 15
    },
    {
      "epoch": 0.09682299546142209,
      "grad_norm": 0.747599720954895,
      "learning_rate": 9.885262510105102e-05,
      "loss": 2.3921,
      "step": 16
    },
    {
      "epoch": 0.10287443267776097,
      "grad_norm": 0.8165279626846313,
      "learning_rate": 9.863552602006435e-05,
      "loss": 2.3954,
      "step": 17
    },
    {
      "epoch": 0.10892586989409984,
      "grad_norm": 0.8175137042999268,
      "learning_rate": 9.839990920553565e-05,
      "loss": 2.5323,
      "step": 18
    },
    {
      "epoch": 0.11497730711043873,
      "grad_norm": 0.8728551864624023,
      "learning_rate": 9.814586436738998e-05,
      "loss": 2.3727,
      "step": 19
    },
    {
      "epoch": 0.12102874432677761,
      "grad_norm": 0.9302012920379639,
      "learning_rate": 9.787348823193013e-05,
      "loss": 2.4606,
      "step": 20
    },
    {
      "epoch": 0.12708018154311648,
      "grad_norm": 0.9512920379638672,
      "learning_rate": 9.75828845050089e-05,
      "loss": 2.201,
      "step": 21
    },
    {
      "epoch": 0.13313161875945537,
      "grad_norm": 1.1705322265625,
      "learning_rate": 9.72741638325434e-05,
      "loss": 2.3474,
      "step": 22
    },
    {
      "epoch": 0.13918305597579425,
      "grad_norm": 1.1705471277236938,
      "learning_rate": 9.694744375838725e-05,
      "loss": 2.3085,
      "step": 23
    },
    {
      "epoch": 0.14523449319213314,
      "grad_norm": 1.2117750644683838,
      "learning_rate": 9.660284867957636e-05,
      "loss": 2.0941,
      "step": 24
    },
    {
      "epoch": 0.15128593040847202,
      "grad_norm": 1.2110291719436646,
      "learning_rate": 9.624050979896533e-05,
      "loss": 2.0143,
      "step": 25
    },
    {
      "epoch": 0.1573373676248109,
      "grad_norm": 1.1926169395446777,
      "learning_rate": 9.586056507527266e-05,
      "loss": 1.9534,
      "step": 26
    },
    {
      "epoch": 0.16338880484114976,
      "grad_norm": 1.1735974550247192,
      "learning_rate": 9.546315917055361e-05,
      "loss": 1.9372,
      "step": 27
    },
    {
      "epoch": 0.16944024205748864,
      "grad_norm": 1.21225905418396,
      "learning_rate": 9.504844339512095e-05,
      "loss": 1.8305,
      "step": 28
    },
    {
      "epoch": 0.17549167927382753,
      "grad_norm": 1.3545117378234863,
      "learning_rate": 9.461657564993418e-05,
      "loss": 1.8428,
      "step": 29
    },
    {
      "epoch": 0.1815431164901664,
      "grad_norm": 1.6355072259902954,
      "learning_rate": 9.416772036647958e-05,
      "loss": 1.8407,
      "step": 30
    },
    {
      "epoch": 0.1875945537065053,
      "grad_norm": 2.1837329864501953,
      "learning_rate": 9.370204844416382e-05,
      "loss": 2.284,
      "step": 31
    },
    {
      "epoch": 0.19364599092284418,
      "grad_norm": 1.917792558670044,
      "learning_rate": 9.321973718524472e-05,
      "loss": 2.086,
      "step": 32
    },
    {
      "epoch": 0.19969742813918306,
      "grad_norm": 2.3043246269226074,
      "learning_rate": 9.272097022732443e-05,
      "loss": 2.1633,
      "step": 33
    },
    {
      "epoch": 0.20574886535552195,
      "grad_norm": 2.235715627670288,
      "learning_rate": 9.220593747343027e-05,
      "loss": 2.3737,
      "step": 34
    },
    {
      "epoch": 0.2118003025718608,
      "grad_norm": 2.3425846099853516,
      "learning_rate": 9.16748350197101e-05,
      "loss": 2.2327,
      "step": 35
    },
    {
      "epoch": 0.2178517397881997,
      "grad_norm": 2.3417069911956787,
      "learning_rate": 9.112786508076972e-05,
      "loss": 2.0814,
      "step": 36
    },
    {
      "epoch": 0.22390317700453857,
      "grad_norm": 3.114316701889038,
      "learning_rate": 9.056523591268064e-05,
      "loss": 2.2132,
      "step": 37
    },
    {
      "epoch": 0.22995461422087746,
      "grad_norm": 3.6260831356048584,
      "learning_rate": 8.998716173368762e-05,
      "loss": 2.1062,
      "step": 38
    },
    {
      "epoch": 0.23600605143721634,
      "grad_norm": 4.580545902252197,
      "learning_rate": 8.939386264264616e-05,
      "loss": 2.199,
      "step": 39
    },
    {
      "epoch": 0.24205748865355523,
      "grad_norm": 5.529226779937744,
      "learning_rate": 8.8785564535221e-05,
      "loss": 2.0428,
      "step": 40
    },
    {
      "epoch": 0.2481089258698941,
      "grad_norm": 8.231687545776367,
      "learning_rate": 8.816249901787737e-05,
      "loss": 2.1102,
      "step": 41
    },
    {
      "epoch": 0.25416036308623297,
      "grad_norm": 1.3886011838912964,
      "learning_rate": 8.752490331969807e-05,
      "loss": 1.3385,
      "step": 42
    },
    {
      "epoch": 0.26021180030257185,
      "grad_norm": 1.6931300163269043,
      "learning_rate": 8.687302020205968e-05,
      "loss": 1.8471,
      "step": 43
    },
    {
      "epoch": 0.26626323751891073,
      "grad_norm": 1.7250523567199707,
      "learning_rate": 8.620709786620231e-05,
      "loss": 1.9711,
      "step": 44
    },
    {
      "epoch": 0.2723146747352496,
      "grad_norm": 1.4766898155212402,
      "learning_rate": 8.552738985872833e-05,
      "loss": 1.9272,
      "step": 45
    },
    {
      "epoch": 0.2783661119515885,
      "grad_norm": 1.3497235774993896,
      "learning_rate": 8.483415497506567e-05,
      "loss": 1.7261,
      "step": 46
    },
    {
      "epoch": 0.2844175491679274,
      "grad_norm": 1.4176321029663086,
      "learning_rate": 8.412765716093272e-05,
      "loss": 1.5276,
      "step": 47
    },
    {
      "epoch": 0.29046898638426627,
      "grad_norm": 1.4237560033798218,
      "learning_rate": 8.340816541184249e-05,
      "loss": 1.3707,
      "step": 48
    },
    {
      "epoch": 0.29652042360060515,
      "grad_norm": 1.1526674032211304,
      "learning_rate": 8.267595367068375e-05,
      "loss": 1.1937,
      "step": 49
    },
    {
      "epoch": 0.30257186081694404,
      "grad_norm": 1.3353413343429565,
      "learning_rate": 8.193130072341873e-05,
      "loss": 1.3607,
      "step": 50
    },
    {
      "epoch": 0.3086232980332829,
      "grad_norm": 1.4287394285202026,
      "learning_rate": 8.117449009293668e-05,
      "loss": 1.5557,
      "step": 51
    },
    {
      "epoch": 0.3146747352496218,
      "grad_norm": 1.5747747421264648,
      "learning_rate": 8.040580993110404e-05,
      "loss": 1.6074,
      "step": 52
    },
    {
      "epoch": 0.3207261724659607,
      "grad_norm": 1.8766899108886719,
      "learning_rate": 7.962555290905197e-05,
      "loss": 1.5941,
      "step": 53
    },
    {
      "epoch": 0.3267776096822995,
      "grad_norm": 1.840211033821106,
      "learning_rate": 7.883401610574336e-05,
      "loss": 1.3652,
      "step": 54
    },
    {
      "epoch": 0.3328290468986384,
      "grad_norm": 2.002328872680664,
      "learning_rate": 7.803150089486144e-05,
      "loss": 1.2883,
      "step": 55
    },
    {
      "epoch": 0.3388804841149773,
      "grad_norm": 2.1058497428894043,
      "learning_rate": 7.721831283006322e-05,
      "loss": 1.079,
      "step": 56
    },
    {
      "epoch": 0.34493192133131617,
      "grad_norm": 1.7998846769332886,
      "learning_rate": 7.639476152864162e-05,
      "loss": 0.9134,
      "step": 57
    },
    {
      "epoch": 0.35098335854765506,
      "grad_norm": 1.864816427230835,
      "learning_rate": 7.556116055364008e-05,
      "loss": 0.82,
      "step": 58
    },
    {
      "epoch": 0.35703479576399394,
      "grad_norm": 1.9277801513671875,
      "learning_rate": 7.4717827294465e-05,
      "loss": 0.7148,
      "step": 59
    },
    {
      "epoch": 0.3630862329803328,
      "grad_norm": 2.2417094707489014,
      "learning_rate": 7.386508284604141e-05,
      "loss": 0.7864,
      "step": 60
    },
    {
      "epoch": 0.3691376701966717,
      "grad_norm": 2.0547397136688232,
      "learning_rate": 7.300325188655761e-05,
      "loss": 0.7012,
      "step": 61
    },
    {
      "epoch": 0.3751891074130106,
      "grad_norm": 2.6025710105895996,
      "learning_rate": 7.21326625538456e-05,
      "loss": 0.6169,
      "step": 62
    },
    {
      "epoch": 0.3812405446293495,
      "grad_norm": 2.259634017944336,
      "learning_rate": 7.125364632044422e-05,
      "loss": 0.6292,
      "step": 63
    },
    {
      "epoch": 0.38729198184568836,
      "grad_norm": 1.6727221012115479,
      "learning_rate": 7.036653786739263e-05,
      "loss": 0.5009,
      "step": 64
    },
    {
      "epoch": 0.39334341906202724,
      "grad_norm": 2.62076997756958,
      "learning_rate": 6.947167495680224e-05,
      "loss": 0.3475,
      "step": 65
    },
    {
      "epoch": 0.39939485627836613,
      "grad_norm": 2.0237042903900146,
      "learning_rate": 6.856939830325534e-05,
      "loss": 0.3402,
      "step": 66
    },
    {
      "epoch": 0.405446293494705,
      "grad_norm": 2.6460840702056885,
      "learning_rate": 6.76600514440799e-05,
      "loss": 0.4223,
      "step": 67
    },
    {
      "epoch": 0.4114977307110439,
      "grad_norm": 1.5343953371047974,
      "learning_rate": 6.674398060854931e-05,
      "loss": 0.3363,
      "step": 68
    },
    {
      "epoch": 0.4175491679273828,
      "grad_norm": 2.540858268737793,
      "learning_rate": 6.582153458605738e-05,
      "loss": 0.521,
      "step": 69
    },
    {
      "epoch": 0.4236006051437216,
      "grad_norm": 4.079755783081055,
      "learning_rate": 6.48930645933185e-05,
      "loss": 0.7838,
      "step": 70
    },
    {
      "epoch": 0.4296520423600605,
      "grad_norm": 2.2050883769989014,
      "learning_rate": 6.395892414064362e-05,
      "loss": 0.3974,
      "step": 71
    },
    {
      "epoch": 0.4357034795763994,
      "grad_norm": 3.4968700408935547,
      "learning_rate": 6.301946889734302e-05,
      "loss": 0.5544,
      "step": 72
    },
    {
      "epoch": 0.44175491679273826,
      "grad_norm": 3.0294439792633057,
      "learning_rate": 6.20750565563069e-05,
      "loss": 0.4241,
      "step": 73
    },
    {
      "epoch": 0.44780635400907715,
      "grad_norm": 2.7060914039611816,
      "learning_rate": 6.112604669781572e-05,
      "loss": 0.3418,
      "step": 74
    },
    {
      "epoch": 0.45385779122541603,
      "grad_norm": 1.8465664386749268,
      "learning_rate": 6.01728006526317e-05,
      "loss": 0.2793,
      "step": 75
    },
    {
      "epoch": 0.4599092284417549,
      "grad_norm": 2.7685420513153076,
      "learning_rate": 5.921568136442397e-05,
      "loss": 0.2668,
      "step": 76
    },
    {
      "epoch": 0.4659606656580938,
      "grad_norm": 3.251429319381714,
      "learning_rate": 5.8255053251579616e-05,
      "loss": 0.2514,
      "step": 77
    },
    {
      "epoch": 0.4720121028744327,
      "grad_norm": 3.9114441871643066,
      "learning_rate": 5.7291282068453166e-05,
      "loss": 0.2437,
      "step": 78
    },
    {
      "epoch": 0.47806354009077157,
      "grad_norm": 2.475526809692383,
      "learning_rate": 5.632473476610748e-05,
      "loss": 0.1536,
      "step": 79
    },
    {
      "epoch": 0.48411497730711045,
      "grad_norm": 4.316705703735352,
      "learning_rate": 5.535577935259888e-05,
      "loss": 0.2268,
      "step": 80
    },
    {
      "epoch": 0.49016641452344933,
      "grad_norm": 7.602482795715332,
      "learning_rate": 5.438478475286003e-05,
      "loss": 0.2635,
      "step": 81
    },
    {
      "epoch": 0.4962178517397882,
      "grad_norm": 6.1896653175354,
      "learning_rate": 5.341212066823355e-05,
      "loss": 0.3162,
      "step": 82
    },
    {
      "epoch": 0.5022692889561271,
      "grad_norm": 3.891818046569824,
      "learning_rate": 5.243815743571009e-05,
      "loss": 0.7189,
      "step": 83
    },
    {
      "epoch": 0.5083207261724659,
      "grad_norm": 7.040829658508301,
      "learning_rate": 5.146326588692438e-05,
      "loss": 1.1875,
      "step": 84
    },
    {
      "epoch": 0.5143721633888049,
      "grad_norm": 9.615294456481934,
      "learning_rate": 5.048781720696291e-05,
      "loss": 1.5413,
      "step": 85
    },
    {
      "epoch": 0.5204236006051437,
      "grad_norm": 8.135547637939453,
      "learning_rate": 4.95121827930371e-05,
      "loss": 1.5035,
      "step": 86
    },
    {
      "epoch": 0.5264750378214826,
      "grad_norm": 5.731124401092529,
      "learning_rate": 4.853673411307564e-05,
      "loss": 1.2925,
      "step": 87
    },
    {
      "epoch": 0.5325264750378215,
      "grad_norm": 5.864335060119629,
      "learning_rate": 4.756184256428992e-05,
      "loss": 0.8643,
      "step": 88
    },
    {
      "epoch": 0.5385779122541604,
      "grad_norm": 6.120126247406006,
      "learning_rate": 4.658787933176646e-05,
      "loss": 0.8852,
      "step": 89
    },
    {
      "epoch": 0.5446293494704992,
      "grad_norm": 3.3239099979400635,
      "learning_rate": 4.561521524713997e-05,
      "loss": 0.6142,
      "step": 90
    },
    {
      "epoch": 0.5506807866868382,
      "grad_norm": 3.291146993637085,
      "learning_rate": 4.4644220647401136e-05,
      "loss": 0.6233,
      "step": 91
    },
    {
      "epoch": 0.556732223903177,
      "grad_norm": 2.9767818450927734,
      "learning_rate": 4.367526523389253e-05,
      "loss": 0.4502,
      "step": 92
    },
    {
      "epoch": 0.5627836611195158,
      "grad_norm": 2.5687639713287354,
      "learning_rate": 4.2708717931546825e-05,
      "loss": 0.4639,
      "step": 93
    },
    {
      "epoch": 0.5688350983358548,
      "grad_norm": 2.8956291675567627,
      "learning_rate": 4.174494674842038e-05,
      "loss": 0.3833,
      "step": 94
    },
    {
      "epoch": 0.5748865355521936,
      "grad_norm": 3.41869854927063,
      "learning_rate": 4.0784318635576055e-05,
      "loss": 0.3458,
      "step": 95
    },
    {
      "epoch": 0.5809379727685325,
      "grad_norm": 2.012956142425537,
      "learning_rate": 3.982719934736832e-05,
      "loss": 0.2622,
      "step": 96
    },
    {
      "epoch": 0.5869894099848714,
      "grad_norm": 1.7718323469161987,
      "learning_rate": 3.887395330218429e-05,
      "loss": 0.1466,
      "step": 97
    },
    {
      "epoch": 0.5930408472012103,
      "grad_norm": 1.5787506103515625,
      "learning_rate": 3.792494344369311e-05,
      "loss": 0.1388,
      "step": 98
    },
    {
      "epoch": 0.5990922844175491,
      "grad_norm": 1.3233414888381958,
      "learning_rate": 3.698053110265699e-05,
      "loss": 0.117,
      "step": 99
    },
    {
      "epoch": 0.6051437216338881,
      "grad_norm": 1.1028014421463013,
      "learning_rate": 3.604107585935638e-05,
      "loss": 0.12,
      "step": 100
    },
    {
      "epoch": 0.6111951588502269,
      "grad_norm": 1.057910442352295,
      "learning_rate": 3.510693540668151e-05,
      "loss": 0.0996,
      "step": 101
    },
    {
      "epoch": 0.6172465960665658,
      "grad_norm": 0.8907783031463623,
      "learning_rate": 3.4178465413942625e-05,
      "loss": 0.0811,
      "step": 102
    },
    {
      "epoch": 0.6232980332829047,
      "grad_norm": 0.7459267377853394,
      "learning_rate": 3.325601939145069e-05,
      "loss": 0.0353,
      "step": 103
    },
    {
      "epoch": 0.6293494704992436,
      "grad_norm": 0.9089831113815308,
      "learning_rate": 3.23399485559201e-05,
      "loss": 0.0709,
      "step": 104
    },
    {
      "epoch": 0.6354009077155824,
      "grad_norm": 0.7159168124198914,
      "learning_rate": 3.143060169674468e-05,
      "loss": 0.0451,
      "step": 105
    },
    {
      "epoch": 0.6414523449319214,
      "grad_norm": 0.5289613008499146,
      "learning_rate": 3.0528325043197785e-05,
      "loss": 0.0278,
      "step": 106
    },
    {
      "epoch": 0.6475037821482602,
      "grad_norm": 0.5118010640144348,
      "learning_rate": 2.963346213260737e-05,
      "loss": 0.0221,
      "step": 107
    },
    {
      "epoch": 0.653555219364599,
      "grad_norm": 0.354955792427063,
      "learning_rate": 2.874635367955579e-05,
      "loss": 0.0104,
      "step": 108
    },
    {
      "epoch": 0.659606656580938,
      "grad_norm": 1.1922907829284668,
      "learning_rate": 2.7867337446154396e-05,
      "loss": 0.0275,
      "step": 109
    },
    {
      "epoch": 0.6656580937972768,
      "grad_norm": 0.9284811615943909,
      "learning_rate": 2.6996748113442394e-05,
      "loss": 0.0236,
      "step": 110
    },
    {
      "epoch": 0.6717095310136157,
      "grad_norm": 0.6479511857032776,
      "learning_rate": 2.613491715395861e-05,
      "loss": 0.0328,
      "step": 111
    },
    {
      "epoch": 0.6777609682299546,
      "grad_norm": 0.3264722526073456,
      "learning_rate": 2.5282172705535013e-05,
      "loss": 0.0126,
      "step": 112
    },
    {
      "epoch": 0.6838124054462935,
      "grad_norm": 0.5275784134864807,
      "learning_rate": 2.4438839446359933e-05,
      "loss": 0.0267,
      "step": 113
    },
    {
      "epoch": 0.6898638426626323,
      "grad_norm": 1.2883987426757812,
      "learning_rate": 2.360523847135838e-05,
      "loss": 0.0482,
      "step": 114
    },
    {
      "epoch": 0.6959152798789713,
      "grad_norm": 0.6634830832481384,
      "learning_rate": 2.2781687169936795e-05,
      "loss": 0.0287,
      "step": 115
    },
    {
      "epoch": 0.7019667170953101,
      "grad_norm": 0.8741732239723206,
      "learning_rate": 2.196849910513858e-05,
      "loss": 0.037,
      "step": 116
    },
    {
      "epoch": 0.708018154311649,
      "grad_norm": 0.7417043447494507,
      "learning_rate": 2.1165983894256647e-05,
      "loss": 0.0244,
      "step": 117
    },
    {
      "epoch": 0.7140695915279879,
      "grad_norm": 1.1073297262191772,
      "learning_rate": 2.037444709094804e-05,
      "loss": 0.0391,
      "step": 118
    },
    {
      "epoch": 0.7201210287443268,
      "grad_norm": 0.9330146908760071,
      "learning_rate": 1.9594190068895968e-05,
      "loss": 0.0312,
      "step": 119
    },
    {
      "epoch": 0.7261724659606656,
      "grad_norm": 0.8450430631637573,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 0.0212,
      "step": 120
    },
    {
      "epoch": 0.7322239031770046,
      "grad_norm": 0.6666033864021301,
      "learning_rate": 1.8068699276581285e-05,
      "loss": 0.0202,
      "step": 121
    },
    {
      "epoch": 0.7382753403933434,
      "grad_norm": 0.8927512168884277,
      "learning_rate": 1.732404632931625e-05,
      "loss": 0.0204,
      "step": 122
    },
    {
      "epoch": 0.7443267776096822,
      "grad_norm": 0.919458270072937,
      "learning_rate": 1.6591834588157523e-05,
      "loss": 0.0185,
      "step": 123
    },
    {
      "epoch": 0.7503782148260212,
      "grad_norm": 1.465542197227478,
      "learning_rate": 1.5872342839067306e-05,
      "loss": 0.5584,
      "step": 124
    },
    {
      "epoch": 0.75642965204236,
      "grad_norm": 1.944527506828308,
      "learning_rate": 1.5165845024934366e-05,
      "loss": 0.7492,
      "step": 125
    },
    {
      "epoch": 0.762481089258699,
      "grad_norm": 2.006526470184326,
      "learning_rate": 1.447261014127167e-05,
      "loss": 0.725,
      "step": 126
    },
    {
      "epoch": 0.7685325264750378,
      "grad_norm": 2.1773271560668945,
      "learning_rate": 1.3792902133797692e-05,
      "loss": 0.6446,
      "step": 127
    },
    {
      "epoch": 0.7745839636913767,
      "grad_norm": 2.4729340076446533,
      "learning_rate": 1.3126979797940336e-05,
      "loss": 0.8293,
      "step": 128
    },
    {
      "epoch": 0.7806354009077155,
      "grad_norm": 2.1801207065582275,
      "learning_rate": 1.2475096680301934e-05,
      "loss": 0.6321,
      "step": 129
    },
    {
      "epoch": 0.7866868381240545,
      "grad_norm": 1.851625680923462,
      "learning_rate": 1.1837500982122645e-05,
      "loss": 0.2963,
      "step": 130
    },
    {
      "epoch": 0.7927382753403933,
      "grad_norm": 2.0671639442443848,
      "learning_rate": 1.1214435464779006e-05,
      "loss": 0.2112,
      "step": 131
    },
    {
      "epoch": 0.7987897125567323,
      "grad_norm": 1.930985450744629,
      "learning_rate": 1.060613735735384e-05,
      "loss": 0.2137,
      "step": 132
    },
    {
      "epoch": 0.8048411497730711,
      "grad_norm": 2.326185464859009,
      "learning_rate": 1.0012838266312396e-05,
      "loss": 0.1678,
      "step": 133
    },
    {
      "epoch": 0.81089258698941,
      "grad_norm": 2.306683301925659,
      "learning_rate": 9.434764087319376e-06,
      "loss": 0.1384,
      "step": 134
    },
    {
      "epoch": 0.8169440242057489,
      "grad_norm": 4.009836673736572,
      "learning_rate": 8.87213491923029e-06,
      "loss": 0.1649,
      "step": 135
    },
    {
      "epoch": 0.8229954614220878,
      "grad_norm": 3.4153988361358643,
      "learning_rate": 8.325164980289896e-06,
      "loss": 0.1384,
      "step": 136
    },
    {
      "epoch": 0.8290468986384266,
      "grad_norm": 4.114608287811279,
      "learning_rate": 7.794062526569734e-06,
      "loss": 0.0969,
      "step": 137
    },
    {
      "epoch": 0.8350983358547656,
      "grad_norm": 1.5541014671325684,
      "learning_rate": 7.2790297726755716e-06,
      "loss": 0.1201,
      "step": 138
    },
    {
      "epoch": 0.8411497730711044,
      "grad_norm": 0.6836862564086914,
      "learning_rate": 6.780262814755284e-06,
      "loss": 0.0259,
      "step": 139
    },
    {
      "epoch": 0.8472012102874432,
      "grad_norm": 0.5702232718467712,
      "learning_rate": 6.29795155583619e-06,
      "loss": 0.0119,
      "step": 140
    },
    {
      "epoch": 0.8532526475037822,
      "grad_norm": 0.27770644426345825,
      "learning_rate": 5.83227963352041e-06,
      "loss": 0.0051,
      "step": 141
    },
    {
      "epoch": 0.859304084720121,
      "grad_norm": 0.44976603984832764,
      "learning_rate": 5.383424350065824e-06,
      "loss": 0.0063,
      "step": 142
    },
    {
      "epoch": 0.8653555219364599,
      "grad_norm": 0.40759414434432983,
      "learning_rate": 4.951556604879048e-06,
      "loss": 0.0091,
      "step": 143
    },
    {
      "epoch": 0.8714069591527988,
      "grad_norm": 0.573310136795044,
      "learning_rate": 4.536840829446387e-06,
      "loss": 0.0135,
      "step": 144
    },
    {
      "epoch": 0.8774583963691377,
      "grad_norm": 0.3286815285682678,
      "learning_rate": 4.139434924727359e-06,
      "loss": 0.0067,
      "step": 145
    },
    {
      "epoch": 0.8835098335854765,
      "grad_norm": 0.12208826839923859,
      "learning_rate": 3.7594902010346768e-06,
      "loss": 0.0034,
      "step": 146
    },
    {
      "epoch": 0.8895612708018155,
      "grad_norm": 0.17780545353889465,
      "learning_rate": 3.397151320423647e-06,
      "loss": 0.0044,
      "step": 147
    },
    {
      "epoch": 0.8956127080181543,
      "grad_norm": 0.4996080696582794,
      "learning_rate": 3.0525562416127472e-06,
      "loss": 0.0241,
      "step": 148
    },
    {
      "epoch": 0.9016641452344932,
      "grad_norm": 0.3310739994049072,
      "learning_rate": 2.725836167456608e-06,
      "loss": 0.0066,
      "step": 149
    },
    {
      "epoch": 0.9077155824508321,
      "grad_norm": 0.7427249550819397,
      "learning_rate": 2.4171154949911067e-06,
      "loss": 0.0224,
      "step": 150
    },
    {
      "epoch": 0.913767019667171,
      "grad_norm": 0.7192853689193726,
      "learning_rate": 2.1265117680698665e-06,
      "loss": 0.0321,
      "step": 151
    },
    {
      "epoch": 0.9198184568835098,
      "grad_norm": 0.653380274772644,
      "learning_rate": 1.8541356326100433e-06,
      "loss": 0.0235,
      "step": 152
    },
    {
      "epoch": 0.9258698940998488,
      "grad_norm": 0.7149347066879272,
      "learning_rate": 1.6000907944643428e-06,
      "loss": 0.0172,
      "step": 153
    },
    {
      "epoch": 0.9319213313161876,
      "grad_norm": 0.7151485681533813,
      "learning_rate": 1.3644739799356764e-06,
      "loss": 0.0246,
      "step": 154
    },
    {
      "epoch": 0.9379727685325264,
      "grad_norm": 0.21587873995304108,
      "learning_rate": 1.1473748989489774e-06,
      "loss": 0.0053,
      "step": 155
    },
    {
      "epoch": 0.9440242057488654,
      "grad_norm": 0.3646269738674164,
      "learning_rate": 9.488762108946892e-07,
      "loss": 0.0073,
      "step": 156
    },
    {
      "epoch": 0.9500756429652042,
      "grad_norm": 0.2228485345840454,
      "learning_rate": 7.690534931565518e-07,
      "loss": 0.0057,
      "step": 157
    },
    {
      "epoch": 0.9561270801815431,
      "grad_norm": 0.5082440376281738,
      "learning_rate": 6.079752123359261e-07,
      "loss": 0.0128,
      "step": 158
    },
    {
      "epoch": 0.962178517397882,
      "grad_norm": 0.9114623069763184,
      "learning_rate": 4.6570269818346224e-07,
      "loss": 0.0233,
      "step": 159
    },
    {
      "epoch": 0.9682299546142209,
      "grad_norm": 0.4167875349521637,
      "learning_rate": 3.4229012024811214e-07,
      "loss": 0.0073,
      "step": 160
    },
    {
      "epoch": 0.9742813918305597,
      "grad_norm": 0.13948678970336914,
      "learning_rate": 2.3778446725233327e-07,
      "loss": 0.0032,
      "step": 161
    },
    {
      "epoch": 0.9803328290468987,
      "grad_norm": 0.34661903977394104,
      "learning_rate": 1.5222552920138856e-07,
      "loss": 0.0053,
      "step": 162
    },
    {
      "epoch": 0.9863842662632375,
      "grad_norm": 0.15658196806907654,
      "learning_rate": 8.564588223346492e-08,
      "loss": 0.0029,
      "step": 163
    },
    {
      "epoch": 0.9924357034795764,
      "grad_norm": 0.40814340114593506,
      "learning_rate": 3.8070876216506066e-08,
      "loss": 0.0051,
      "step": 164
    },
    {
      "epoch": 0.9984871406959153,
      "grad_norm": 0.8116607069969177,
      "learning_rate": 9.518625096366363e-09,
      "loss": 0.1255,
      "step": 165
    },
    {
      "epoch": 0.9984871406959153,
      "eval_loss": 0.06893022358417511,
      "eval_runtime": 7.1896,
      "eval_samples_per_second": 19.473,
      "eval_steps_per_second": 4.868,
      "step": 165
    },
    {
      "epoch": 1.0045385779122542,
      "grad_norm": 1.0615386962890625,
      "learning_rate": 0.0,
      "loss": 0.2848,
      "step": 166
    }
  ],
  "logging_steps": 1,
  "max_steps": 166,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 42,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1627739017537126e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}