{
  "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": 6.264223837447837,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.8653,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.411256699697074,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8939,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.027212861227225,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8222,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 6.094770717952796,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.861,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.446971159716574,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.8116,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.500530798923617,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7778,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.097234677116947,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.775,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 4.3091575269309015,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.7985,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 3.821872023621295,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.7189,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.036132029196842,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.7316,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.168552688043994,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.7169,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.5725797832099504,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.6868,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 2.3271747887261998,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6444,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.853858851417222,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.65,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.3237928410153996,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.6213,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.2891143923566757,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.6337,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.2966845135592888,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.6227,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.1578792077581719,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.6338,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.3234419656554235,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.6246,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.8655919173667078,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.6125,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.8165889585109629,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.569,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.1115583068811594,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5777,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.7628757493964801,
      "learning_rate": 2e-05,
      "loss": 0.6277,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 1.0411028517271135,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5953,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.6804270776946826,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.6098,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7025164734168877,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5633,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.8802100886304541,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5583,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.6163862927237143,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.5541,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.6849997287254687,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5612,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.6298641549528552,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5655,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.7004684552037466,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.5887,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.7253351732744092,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5631,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.7406575180088495,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.6821,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.6795309366762818,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.5633,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.5300947378258845,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4339,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.6948218294926742,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.4979,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.6383161268526382,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.5448,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.5796574691977027,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.4755,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.5205301553697834,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.449,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.6020646847574486,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.517,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.6959773135755725,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4576,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.6088235512117044,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4684,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.6766258484200818,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.4712,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.6249409834640268,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.4912,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.6980550037797275,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.5171,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.4823728113030638,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4755,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.7178649227814365,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.5238,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.47742338433152715,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4681,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.627334443689055,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.4942,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.47781725246848783,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.4346,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.6931187000064182,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.551,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.5185073569252794,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4564,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.6348042347292179,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.5452,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.6054361749046828,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.454,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.5450009949395281,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4683,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.61123856404764,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.4536,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.5906938867185233,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4863,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.6213213252204924,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4862,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5836327926282957,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.5069,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.7548796429456727,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.5135,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.5254706157464364,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4328,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.713641729007418,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.4992,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.5992458896946319,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4924,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.48363149755605384,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.4423,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.6984801529973972,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5937,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.6540653957913342,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.4914,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.5546741099568598,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.384,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.5533344104756105,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.4095,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.6644968396867378,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.3975,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.5747849644459087,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.3921,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.5504373978876771,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.3463,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.6175640089869175,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.3745,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.5572138976873671,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.3469,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.6164348928133362,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.3958,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.5796998598785476,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.3908,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.6116690415189946,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.3618,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.5839247234024287,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.4001,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.5648198606713947,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.3697,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.47424698774743923,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.4042,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.526491138441173,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.3748,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.45913538596337083,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3272,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.5245789145036932,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.3984,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.4564535629632483,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.3488,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.4439029401487793,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.3743,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.45157043394592816,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.3499,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.47118805349317777,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3749,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.4991151007339554,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.3782,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.5070045648686803,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.3816,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.5283211062424125,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.3854,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.49967849508328893,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3318,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.5761824799913082,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.4053,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.457027992926431,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.3331,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.5439180987165674,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.3269,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.5321339617362203,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.4065,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.45872333749508787,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.3782,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.6685263430601783,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.4034,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.4244383332703259,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.3235,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.7804195707932797,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.4883,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.6898289222972395,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.2912,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.6526461198800771,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.2993,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.6109871722734659,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.2829,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.8726587547995472,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.2862,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.5915965883830315,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.3082,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.7018059359740425,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.2775,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.5648020640782025,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.2484,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.5491041130301101,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.2713,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.542398015518937,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.2809,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.5673079398195483,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.2771,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.5560061684468235,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.2785,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.5142315215324735,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.2655,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.508027635053969,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.2876,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.5207943414533546,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.2645,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.4794185965888677,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.2843,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5216131678174822,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.2705,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.4362210510365554,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.2405,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.5484752284700704,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.2874,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.44411698682566847,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.2722,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.4256453593025832,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.2606,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.4840576533102752,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.2627,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.4763344324363398,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.2778,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.4520353656979555,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.2782,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.484359088787161,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.2822,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.5122783646022456,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.2624,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.4497907628730785,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.2518,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.4376125761126391,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.2679,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.45266135210460645,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.2617,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.4286208278697542,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.2857,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.4164821744132408,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.2676,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.42463497579882714,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.2743,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.45143201821249485,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.2827,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.5178904309434295,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.3169,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.6255008822350463,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.2049,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.4663360072666728,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.2082,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.5525690051399497,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.1795,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.7424606332536962,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.2008,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.5745096218393714,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.2018,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.4702614444346243,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.1703,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.6390290266050745,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.2185,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.4628581214262268,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.1862,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.5399100867642865,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.1989,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.574931790699253,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.2267,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.4439955496878499,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.1753,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.5047861912578576,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.2041,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.43821876188470377,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.1844,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.450828883558137,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.1676,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.47294294676149884,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.2021,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.4116079904869393,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.1892,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.44665415136996167,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.185,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.45228880304893426,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.2135,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.4211840314348762,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.1744,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.42731254775297467,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.179,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.4114726369883331,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.1736,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.41713312191238006,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.1883,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.4441935778969085,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.2123,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.41563812394315813,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.1899,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.4362364289458295,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.1847,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.40794185942171385,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.174,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.38071092887816593,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.1667,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.46842208821628495,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.2158,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.40121416255265785,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.1846,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.42931887950953884,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.1956,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.4122318318707383,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.202,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.49075311327954135,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.2082,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.6212796674101548,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.2131,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.5462254440984692,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.1501,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.42302184153770833,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.1489,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.3469584459655291,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.1308,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.43320219632040574,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.1345,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.5578330333033935,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.1396,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.640178109725764,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.1526,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.4415130114149953,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.1288,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.37477185848963257,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.1357,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.4454771530728319,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.1524,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.45252662088706425,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.1346,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.40446738836683316,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.1364,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.4168074072408946,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.1445,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.3896504523888332,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.1362,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.3589228993693732,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.1338,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.4002822744684675,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.1388,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.3722210390601156,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.1252,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.36402491121876285,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.1262,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.3694016510988465,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.1337,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.37537279604979884,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.129,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.33440167755873895,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.1116,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.3757913764404746,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.1354,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.34216379773242095,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.1379,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.33803173864730923,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.1279,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.3654303753984185,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.1413,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.32221649213877057,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.1226,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.3810213863554794,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.1554,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.3391070109677399,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.1255,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.34691646635131895,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.1379,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.3342692915245725,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1385,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.3335059460429094,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1212,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.3607583058461707,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1412,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.4698491102661422,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.1821,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.40763370914111263,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.121,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.33102756087066815,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.1205,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.40115234371354197,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.1134,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.33971191739567935,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.106,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.34076051053215356,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.1162,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.30163529551377666,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.1104,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.30486309486129837,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.1214,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.37329930860455734,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.1189,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.30478576584706973,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.1058,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.32163190636375755,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.1234,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.35511722030237625,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.1149,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.3457462845239258,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.1202,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.3107686645474473,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.1146,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.3369894507843882,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.1197,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.2958490695117549,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.1069,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.32195825943174333,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.1216,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.3252260869537387,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.1204,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.328999581616471,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.1142,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.3131690743268465,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.1234,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.2999294804387676,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.1045,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.3069573676918086,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.1066,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.30918289682126204,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.1123,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.3421838237906003,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.1164,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.32110695428743896,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.1156,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.33797817117104373,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.1174,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.31898372133563707,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.1079,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.31465044834388256,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.1184,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.3182540685603932,
      "learning_rate": 0.0,
      "loss": 0.1176,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 3.4820098135320166e+17,
      "train_loss": 0.32787186798772644,
      "train_runtime": 15227.7577,
      "train_samples_per_second": 1.453,
      "train_steps_per_second": 0.015
    }
  ],
  "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": 3.4820098135320166e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}