{
  "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": 6.2762570503154205,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.7985,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.372603665697634,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8287,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.062767641447401,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.7973,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 5.6613029767923715,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.8092,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.197464170236558,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.743,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.2261820018419862,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7096,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 1.9503775754217643,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.7191,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 3.608471135239456,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.6861,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.065861391567657,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.6984,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.505281158575557,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.7271,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 3.993505256275216,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.6704,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.2461906239411507,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.6602,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 1.8913240182818438,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6359,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.7827255413792524,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.6216,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 1.9373952016631497,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.6064,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 1.5189269144240194,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.6002,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.021532679768449,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.5911,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.2145696535656478,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.5941,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 0.9332576835459129,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.5865,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.9854718835822585,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.5666,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.863534446743324,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.5562,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 0.7855047440591549,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5584,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.8922878440046972,
      "learning_rate": 2e-05,
      "loss": 0.5352,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.8225820187430022,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5461,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.7269612033200328,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.536,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7792255422448295,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5253,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.8066711927499114,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5387,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.6726978183483585,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.5356,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.6050529485762738,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5261,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.7394892576938703,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5216,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.5561243389272098,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.5468,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.6783445230500457,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5023,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.8473261063526972,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.6634,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.5639630772458379,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.5257,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.6018154527114893,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.5002,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.4449257935606156,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.4923,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.6095299328981161,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.4691,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.4365058095200411,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.4843,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.588262834772594,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.4939,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.48798032912159767,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.5024,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.46853548506216736,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4704,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.603637609705948,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4861,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.3574714299256678,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.4771,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.5461637899792279,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.4747,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.5078531190671156,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.4685,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.4407774024703313,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4662,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.41357135702400755,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.4514,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.3863661946111623,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4655,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.41365695519476203,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.4589,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.33719603061370224,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.4518,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.4574097636400911,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.4948,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.3479977065398371,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4424,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.4724172593633243,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.4635,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.4304392477198137,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.4682,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.46503631795249123,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4472,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.41443837221825697,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.46,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.5487340984545669,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4763,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.37023687707825226,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4407,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5042168230774619,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.4467,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.34667620584275416,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.4616,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.5002586337253488,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4464,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.3403749339000905,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.4727,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.5826484487109438,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4781,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.4018797150905205,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.4405,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.6206251584461557,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5285,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.535695839095322,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.4893,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.443131103361226,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.406,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.46068758524257375,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.4289,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.4590090264378599,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.4234,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.48721707665130654,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.4224,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.40333081866206405,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.4188,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.4156220166036803,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.4095,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.5184253817547999,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4158,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.46501725457944976,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.3948,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.5825322765448153,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.4147,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.42008013472004313,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.4127,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.423559095940668,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.3946,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.45231871694743536,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.3899,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.37548541782302813,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.4073,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.525590615002906,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.4136,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.37282049239237197,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3981,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.37111969543001916,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.4065,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.37136294673282466,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.4074,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.3801770139183022,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.4246,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.35416208920443465,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.3873,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.3578966374010058,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3983,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.40555968148299765,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.4086,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.40468319788911045,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.4066,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.37102140930082733,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.4231,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.3273262293104703,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3993,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.370158608571278,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.3965,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.3744419042755021,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.4018,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.38421672659554956,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.4078,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.3275237573172431,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.3947,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.3921279443353858,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.414,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.36365091751766443,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.3995,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.43439625268654203,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.4192,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.5725112171793767,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.5476,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.45398137723690163,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.3616,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.32901250964483086,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.3588,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.4399785786479097,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.3573,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.3846712692482169,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.358,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.4802451180280138,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.3667,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.35241378174627325,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.3548,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.43086405500727976,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.3556,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.38232645998765025,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.3631,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.35847854966536935,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.3604,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.36897397569214757,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.3592,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.37467211844415615,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.3572,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.36003522481494354,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.3739,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.37143532410878527,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.3524,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.4453306490714606,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.3515,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.36315183713364413,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.3513,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.37977185188049756,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.3477,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.4146650557139675,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.3527,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.35575298951162226,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.3688,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.444922147828904,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.3633,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.3309824141226349,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.3641,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.38697481816800394,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.3613,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.3879277473127658,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.3596,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.282250317814845,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.3416,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.3648790777757849,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.3575,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.30443131339417673,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.3536,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.2687608839585301,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.3529,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.3230538781429815,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.3772,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.27733698255984446,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.353,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.2997526920383882,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.352,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.2951024024711004,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.355,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.27955233576724503,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.3566,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.31194797393825047,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.3611,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.3678902368233579,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.4546,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.38827082436770505,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.3237,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.2932322187357908,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.3199,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.32433922673028237,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.3306,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.5054203967391238,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.3054,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.3374827070186633,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.3117,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.342026054996755,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.3144,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.3182120363246575,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.3129,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.3239867250603833,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.3262,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.2905276266280626,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.3168,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.2838957550256554,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.2889,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.30798472723779186,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.3237,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.2985930694931347,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.3177,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.2975687540416297,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.3115,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.2947933952025458,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.3117,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.2792576049515313,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.3138,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.2803003733258494,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.3129,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.2838663067726723,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.3183,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.2738903381476995,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.3252,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.25172458374998063,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.3044,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.33097577225008884,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.3097,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.28889264345089877,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.3048,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.25367866267859174,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.3142,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.2699954399708979,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.3229,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.2942054825468479,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.3113,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.24733670169648372,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.3145,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.28869518200797323,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.3149,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.2639899565601974,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.3158,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.25174531604825895,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.3175,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.25635795561415686,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.314,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.2548776553226238,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.3107,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.24134106593311486,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.3143,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.30077043528093356,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.3754,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.35404762156689296,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.34,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.29244934565178143,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.2857,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.255392258240675,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.2802,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.2647850833197495,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.291,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.3126650485675664,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.2839,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.3426888902495221,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.2959,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.27722583627969555,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.2912,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.27324364837457527,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.289,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.27519141445434037,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.2914,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.2731822522606332,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.2869,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.26057599808611365,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.2745,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.27026903639587385,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.2869,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.24848801604940615,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.2765,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.2890057593880235,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.2871,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.25691024472573787,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.2855,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.2797626452740262,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.2802,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.24802258010907652,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.2903,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.2483600540699912,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.277,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.26469438798640293,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.2827,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.2328899633229245,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.2837,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.22819682740718097,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.2828,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.22811822481106364,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.2797,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.23902550880267787,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.2905,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.22704540975425322,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.2976,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.2158049037694302,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.2781,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.22805428387253524,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.2701,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.24169918977972027,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.2906,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.22568497047547462,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.2936,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.21829647547223588,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.2781,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.22943486342030758,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.2868,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.22215043650428865,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.2774,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.21821517508226881,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.2854,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.3083326056425308,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.3906,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.2643136513123642,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.2703,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.2352915754817002,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.2636,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.24081971629748225,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.2754,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.21054698725431437,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.2767,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.2135593120759992,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.2735,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.2266150866011842,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.2728,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.2274076869554012,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.2798,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.23881124789515334,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.2771,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.22431055048828588,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.2652,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.2309831432591444,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.273,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.21340737179578648,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.2666,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.21899358932824015,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.273,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.21059951568627452,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.2645,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.21918832937530464,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.2699,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.22079490959617132,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.2772,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.21780116231199648,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.2691,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.2149757148808596,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.2665,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.21290811042392016,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.2701,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.20958763548239512,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.2715,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.2169846855969907,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.2787,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.21407022991113825,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.2771,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.22495904618368845,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.2755,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.207246120435824,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.2635,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.2156139701383356,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.2676,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.20729674946451054,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.2703,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.217070764472447,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.2777,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.22237984283499512,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.2712,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.21778692539729155,
      "learning_rate": 0.0,
      "loss": 0.2827,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 8.806509562099139e+17,
      "train_loss": 0.39622728606419905,
      "train_runtime": 26509.0933,
      "train_samples_per_second": 0.834,
      "train_steps_per_second": 0.008
    }
  ],
  "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": 8.806509562099139e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}