{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.850632911392405,
  "eval_steps": 500,
  "global_step": 224,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030379746835443037,
      "grad_norm": 6.728532471421739,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.8452,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.477873118351978,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8315,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.206808643142912,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8029,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 6.086052306045272,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.8264,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.546499515358167,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.7524,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.414904719544462,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7009,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.197296340173741,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.7057,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 4.088091575618223,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.723,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.257395695872859,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.7279,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.042150324881036,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.6732,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.023055913959011,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.6156,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.873155383888412,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.6472,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 2.5689508548129645,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6335,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.8622409566133382,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.6101,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.443395463942045,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.6028,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.0889898616258162,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.5755,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.205476825961952,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.5658,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.5772264724357605,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.6012,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.2291738746310337,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.5798,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.8417914097335203,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.5801,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 1.0142316481342981,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.5435,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.015624724508394,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5351,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.6837493971176926,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 1.0146092701213756,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5367,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.8066720296281878,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.5169,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.9525030113238759,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5544,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.7836793915786231,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5065,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.6366524290355294,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.479,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.8905492057552872,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5347,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5695399428434885,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.4901,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.6920153923154397,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.5243,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.6601304018344886,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5038,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.8505878668391017,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.6836,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.5388830444581212,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.4185,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.7448248641425436,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4716,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.6175198747971177,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.5139,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.5602854106712951,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.4369,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.4627445561883908,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.4306,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.6180378452492742,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.4474,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.5816144477002346,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.4576,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.5220153753051344,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4003,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.5111242506114239,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4415,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.5796719026608445,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.4763,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.5330336629005769,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.432,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.5980369623455133,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.4106,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.5784782615596188,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4446,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.670084251167805,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.457,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.6482612332337225,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4968,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.7326402205670285,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.3984,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.6134162626638894,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.4536,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.4753153166792695,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.417,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 151.60410390215193,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4507,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 1.0014665767200295,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.4176,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.7582407847892562,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.476,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.8822392260186765,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4287,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.7095103794219065,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.4074,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.6560859276484973,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4209,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.8243200720147149,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4379,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.620231233809006,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.4399,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.6477502881876234,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.4291,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.650586796494011,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4048,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.5028356887014183,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.4301,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.6376199277104793,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4539,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.6261709953146879,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.3926,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.6407203370936383,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.479,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.665637280766245,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.4649,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 1.14206727433835,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.3534,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.6997600495305779,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.3923,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.5952645359541701,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.3313,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.806140488022117,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.3629,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.6899775147335421,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.3477,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.5450120267095772,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.3172,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 1.7897329882388493,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.3463,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.8179467499153401,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.3608,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.4796492587712549,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.3082,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.7543593721853677,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.3561,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.6140484015895801,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.3301,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.6010679954164782,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.3384,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.6865792144140253,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.3706,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.6416584067273542,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.3236,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.555203218053069,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3236,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.662835476432764,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.3945,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.4544114720908494,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.2919,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.7245826268865141,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.365,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.5489838948522684,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.3577,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.5683466665508439,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3344,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.6420441475821229,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.2908,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.5239434093064792,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.3387,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.6524063735190525,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.3419,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.4925602547588343,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.2828,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.7420503162847583,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.4265,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.5552173801484004,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.3547,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.47280701768375794,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.3099,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.6245147341456048,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.3449,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.4950226700843929,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.3438,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.5897220711127049,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.3364,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.6775842540155296,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.3481,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.7044476387070703,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.4241,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.8214715816680674,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.277,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.5865615939220649,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.2902,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.8802843082957593,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.2613,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.6342009488469585,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.2744,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.6182591834124829,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.2659,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.6105031181312327,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.2504,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.560452912102368,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.2453,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.6230018134510154,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.2573,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.570933169586084,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.2717,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.48814183205846773,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.2516,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.528570686823949,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.2585,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.508866053599702,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.2605,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.4898484901079688,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.2384,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.48795469934556035,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.2573,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.492272507678816,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.2464,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5872578295549884,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.2581,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.4849611343110372,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.2653,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.4625551501204702,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.2427,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.520939500984656,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.2435,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.44816168069701356,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.2482,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.47687668498551183,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.2557,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.44256538027926445,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.2552,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.5144441583552056,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.2464,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.47675429714679624,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.2652,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.49652445260144373,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.2482,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.44097241910217927,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.2446,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.6109953448970892,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.2548,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.4575929275341041,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.2403,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.43579058324589376,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.2696,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.42502408655640994,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.2503,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.4246848302823355,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.2518,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.4349462810946266,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.2404,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.6145233063273617,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.3034,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.6062670074377573,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.2058,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.44336592617285675,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.1694,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.46411484581929835,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.1792,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.7494053227595222,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.1997,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.5571929322517589,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.1754,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.5161617949994074,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.1884,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.7546014721638269,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.1898,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.49824813309072946,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.1777,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.47011730269110047,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.1683,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.4384470015158467,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.1875,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.49215364697942177,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.2073,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.4629298541259554,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.1695,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.43724424003301365,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.1634,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.47350851243117864,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.1873,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.3842481638676041,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.165,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.45718360134310965,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.1838,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.4102344768195477,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.1631,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.3931320032138753,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.1708,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.4799608445688827,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.1891,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.4006344430587296,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.1619,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.4248964961182278,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.191,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.392546415757788,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.1728,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.4054422881689731,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.1806,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.37729154015685124,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.1662,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.415140587548714,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.1898,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.4578838649770716,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2089,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.37413400094170707,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.1507,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.3846593278745846,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.1823,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.4015034837711947,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.1825,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.39087793553264927,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.1545,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.4147024866318068,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.1766,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.4646552307792971,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.2099,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.49809777526469257,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.197,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.46465876188732097,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.1451,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.40299416898879276,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.1403,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.3564036085676926,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.1243,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.47312411729559195,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.1563,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.4547474032293964,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.1247,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.4864071666983018,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.1383,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.42683784322941737,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.134,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.37008257301690073,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.1175,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.38605523984675955,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.1287,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.38127923602018277,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.1277,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.37681380027938616,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.1353,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.35489658111956907,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.135,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.37377054329517706,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.1346,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.3836705144647881,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.1326,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.3409766109730774,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.1228,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.388042797904289,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.1445,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.3208454759647045,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.1141,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.397736373015668,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.1383,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.35206692341441553,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.1145,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.3735831570296396,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.1248,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.34130452000720274,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.1144,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.34657315328762783,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.1281,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.34838597592165155,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.1264,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.3835482669797013,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.1479,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.3614186996262409,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.1064,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.33810509966804675,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.1386,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.31608971402528485,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.1235,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.32931058651023093,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.1068,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.35449088859176087,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1405,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.3260062307745717,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1273,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.3350452070907351,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1371,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.4469605434294961,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.167,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.3847180304309178,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.1112,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.37479355619260757,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.1086,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.3499116714607568,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.1113,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.3556970878645843,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.1111,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.31753436469541624,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.1178,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.3184408559294196,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.1,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.29859444557341414,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.105,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.33502661524591837,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.1126,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.29826906253575397,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.1026,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.34564702546965725,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.113,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.3316376052058948,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.109,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.3564979904957744,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.1054,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.32245573665876553,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.1131,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.3518253222774433,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.1099,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.3240060671704569,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.1115,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.31741841357972994,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.103,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.34868023901924616,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.1165,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.29797301586297786,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.1045,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.30793643919152075,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.1033,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.3139640415797819,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.1166,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.3090711567568932,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.1031,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.3035735614965039,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.0972,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.3031009491065144,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.1161,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.32915772413278,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.1116,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.29509880968285096,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.1058,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.3159582746177763,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.1075,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.3196048151689102,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.1102,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.2807557993170677,
      "learning_rate": 0.0,
      "loss": 0.1076,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 3.188147913574318e+17,
      "train_loss": 0.30276693496853113,
      "train_runtime": 14303.0706,
      "train_samples_per_second": 1.547,
      "train_steps_per_second": 0.016
    }
  ],
  "logging_steps": 1,
  "max_steps": 224,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.188147913574318e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}