one-align-ntireugc / trainer_state.json
kkfy's picture
Upload folder using huggingface_hub
db95b8a verified
raw
history blame
No virus
63.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 522,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.25e-06,
"loss": 0.2822,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 2.5e-06,
"loss": 0.1688,
"step": 2
},
{
"epoch": 0.02,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.218,
"step": 3
},
{
"epoch": 0.02,
"learning_rate": 5e-06,
"loss": 0.1948,
"step": 4
},
{
"epoch": 0.03,
"learning_rate": 6.25e-06,
"loss": 0.1496,
"step": 5
},
{
"epoch": 0.03,
"learning_rate": 7.500000000000001e-06,
"loss": 0.1308,
"step": 6
},
{
"epoch": 0.04,
"learning_rate": 8.750000000000001e-06,
"loss": 0.1384,
"step": 7
},
{
"epoch": 0.05,
"learning_rate": 1e-05,
"loss": 0.1278,
"step": 8
},
{
"epoch": 0.05,
"learning_rate": 1.125e-05,
"loss": 0.103,
"step": 9
},
{
"epoch": 0.06,
"learning_rate": 1.25e-05,
"loss": 0.1629,
"step": 10
},
{
"epoch": 0.06,
"learning_rate": 1.375e-05,
"loss": 0.1495,
"step": 11
},
{
"epoch": 0.07,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.1234,
"step": 12
},
{
"epoch": 0.07,
"learning_rate": 1.6250000000000002e-05,
"loss": 0.1277,
"step": 13
},
{
"epoch": 0.08,
"learning_rate": 1.7500000000000002e-05,
"loss": 0.134,
"step": 14
},
{
"epoch": 0.09,
"learning_rate": 1.8750000000000002e-05,
"loss": 0.1421,
"step": 15
},
{
"epoch": 0.09,
"learning_rate": 2e-05,
"loss": 0.136,
"step": 16
},
{
"epoch": 0.1,
"learning_rate": 1.9999807262012045e-05,
"loss": 0.1344,
"step": 17
},
{
"epoch": 0.1,
"learning_rate": 1.9999229055477764e-05,
"loss": 0.1124,
"step": 18
},
{
"epoch": 0.11,
"learning_rate": 1.999826540268562e-05,
"loss": 0.1733,
"step": 19
},
{
"epoch": 0.11,
"learning_rate": 1.999691634078213e-05,
"loss": 0.1026,
"step": 20
},
{
"epoch": 0.12,
"learning_rate": 1.999518192177038e-05,
"loss": 0.1288,
"step": 21
},
{
"epoch": 0.13,
"learning_rate": 1.9993062212508054e-05,
"loss": 0.1067,
"step": 22
},
{
"epoch": 0.13,
"learning_rate": 1.9990557294704857e-05,
"loss": 0.102,
"step": 23
},
{
"epoch": 0.14,
"learning_rate": 1.998766726491935e-05,
"loss": 0.0906,
"step": 24
},
{
"epoch": 0.14,
"learning_rate": 1.998439223455524e-05,
"loss": 0.0908,
"step": 25
},
{
"epoch": 0.15,
"learning_rate": 1.9980732329857077e-05,
"loss": 0.0994,
"step": 26
},
{
"epoch": 0.16,
"learning_rate": 1.9976687691905394e-05,
"loss": 0.1055,
"step": 27
},
{
"epoch": 0.16,
"learning_rate": 1.997225847661127e-05,
"loss": 0.101,
"step": 28
},
{
"epoch": 0.17,
"learning_rate": 1.9967444854710313e-05,
"loss": 0.0992,
"step": 29
},
{
"epoch": 0.17,
"learning_rate": 1.996224701175608e-05,
"loss": 0.0992,
"step": 30
},
{
"epoch": 0.18,
"learning_rate": 1.9956665148112936e-05,
"loss": 0.0963,
"step": 31
},
{
"epoch": 0.18,
"learning_rate": 1.995069947894831e-05,
"loss": 0.1659,
"step": 32
},
{
"epoch": 0.19,
"learning_rate": 1.9944350234224415e-05,
"loss": 0.0728,
"step": 33
},
{
"epoch": 0.2,
"learning_rate": 1.9937617658689385e-05,
"loss": 0.1105,
"step": 34
},
{
"epoch": 0.2,
"learning_rate": 1.993050201186783e-05,
"loss": 0.0901,
"step": 35
},
{
"epoch": 0.21,
"learning_rate": 1.9923003568050846e-05,
"loss": 0.0841,
"step": 36
},
{
"epoch": 0.21,
"learning_rate": 1.991512261628542e-05,
"loss": 0.0978,
"step": 37
},
{
"epoch": 0.22,
"learning_rate": 1.9906859460363307e-05,
"loss": 0.0744,
"step": 38
},
{
"epoch": 0.22,
"learning_rate": 1.989821441880933e-05,
"loss": 0.062,
"step": 39
},
{
"epoch": 0.23,
"learning_rate": 1.988918782486906e-05,
"loss": 0.0702,
"step": 40
},
{
"epoch": 0.24,
"learning_rate": 1.987978002649601e-05,
"loss": 0.0832,
"step": 41
},
{
"epoch": 0.24,
"learning_rate": 1.986999138633821e-05,
"loss": 0.1054,
"step": 42
},
{
"epoch": 0.25,
"learning_rate": 1.9859822281724218e-05,
"loss": 0.0867,
"step": 43
},
{
"epoch": 0.25,
"learning_rate": 1.9849273104648592e-05,
"loss": 0.0773,
"step": 44
},
{
"epoch": 0.26,
"learning_rate": 1.9838344261756755e-05,
"loss": 0.0753,
"step": 45
},
{
"epoch": 0.26,
"learning_rate": 1.9827036174329355e-05,
"loss": 0.073,
"step": 46
},
{
"epoch": 0.27,
"learning_rate": 1.981534927826599e-05,
"loss": 0.0787,
"step": 47
},
{
"epoch": 0.28,
"learning_rate": 1.9803284024068428e-05,
"loss": 0.0637,
"step": 48
},
{
"epoch": 0.28,
"learning_rate": 1.979084087682323e-05,
"loss": 0.0597,
"step": 49
},
{
"epoch": 0.29,
"learning_rate": 1.9778020316183834e-05,
"loss": 0.0763,
"step": 50
},
{
"epoch": 0.29,
"learning_rate": 1.9764822836352046e-05,
"loss": 0.0789,
"step": 51
},
{
"epoch": 0.3,
"learning_rate": 1.9751248946059016e-05,
"loss": 0.0657,
"step": 52
},
{
"epoch": 0.3,
"learning_rate": 1.97372991685456e-05,
"loss": 0.0638,
"step": 53
},
{
"epoch": 0.31,
"learning_rate": 1.9722974041542205e-05,
"loss": 0.0477,
"step": 54
},
{
"epoch": 0.32,
"learning_rate": 1.9708274117248067e-05,
"loss": 0.0858,
"step": 55
},
{
"epoch": 0.32,
"learning_rate": 1.969319996230995e-05,
"loss": 0.0748,
"step": 56
},
{
"epoch": 0.33,
"learning_rate": 1.9677752157800316e-05,
"loss": 0.0482,
"step": 57
},
{
"epoch": 0.33,
"learning_rate": 1.966193129919491e-05,
"loss": 0.07,
"step": 58
},
{
"epoch": 0.34,
"learning_rate": 1.9645737996349828e-05,
"loss": 0.0797,
"step": 59
},
{
"epoch": 0.34,
"learning_rate": 1.9629172873477995e-05,
"loss": 0.062,
"step": 60
},
{
"epoch": 0.35,
"learning_rate": 1.9612236569125093e-05,
"loss": 0.0835,
"step": 61
},
{
"epoch": 0.36,
"learning_rate": 1.9594929736144978e-05,
"loss": 0.094,
"step": 62
},
{
"epoch": 0.36,
"learning_rate": 1.957725304167447e-05,
"loss": 0.0967,
"step": 63
},
{
"epoch": 0.37,
"learning_rate": 1.9559207167107684e-05,
"loss": 0.1099,
"step": 64
},
{
"epoch": 0.37,
"learning_rate": 1.9540792808069726e-05,
"loss": 0.0731,
"step": 65
},
{
"epoch": 0.38,
"learning_rate": 1.95220106743899e-05,
"loss": 0.1106,
"step": 66
},
{
"epoch": 0.39,
"learning_rate": 1.950286149007434e-05,
"loss": 0.0963,
"step": 67
},
{
"epoch": 0.39,
"learning_rate": 1.948334599327809e-05,
"loss": 0.0695,
"step": 68
},
{
"epoch": 0.4,
"learning_rate": 1.9463464936276676e-05,
"loss": 0.0863,
"step": 69
},
{
"epoch": 0.4,
"learning_rate": 1.944321908543708e-05,
"loss": 0.1108,
"step": 70
},
{
"epoch": 0.41,
"learning_rate": 1.9422609221188208e-05,
"loss": 0.0636,
"step": 71
},
{
"epoch": 0.41,
"learning_rate": 1.9401636137990814e-05,
"loss": 0.0619,
"step": 72
},
{
"epoch": 0.42,
"learning_rate": 1.938030064430688e-05,
"loss": 0.0784,
"step": 73
},
{
"epoch": 0.43,
"learning_rate": 1.9358603562568417e-05,
"loss": 0.097,
"step": 74
},
{
"epoch": 0.43,
"learning_rate": 1.933654572914581e-05,
"loss": 0.0775,
"step": 75
},
{
"epoch": 0.44,
"learning_rate": 1.931412799431554e-05,
"loss": 0.0618,
"step": 76
},
{
"epoch": 0.44,
"learning_rate": 1.929135122222743e-05,
"loss": 0.0616,
"step": 77
},
{
"epoch": 0.45,
"learning_rate": 1.926821629087133e-05,
"loss": 0.0954,
"step": 78
},
{
"epoch": 0.45,
"learning_rate": 1.924472409204326e-05,
"loss": 0.0902,
"step": 79
},
{
"epoch": 0.46,
"learning_rate": 1.9220875531311044e-05,
"loss": 0.0459,
"step": 80
},
{
"epoch": 0.47,
"learning_rate": 1.9196671527979407e-05,
"loss": 0.0448,
"step": 81
},
{
"epoch": 0.47,
"learning_rate": 1.917211301505453e-05,
"loss": 0.049,
"step": 82
},
{
"epoch": 0.48,
"learning_rate": 1.9147200939208088e-05,
"loss": 0.0417,
"step": 83
},
{
"epoch": 0.48,
"learning_rate": 1.9121936260740752e-05,
"loss": 0.0682,
"step": 84
},
{
"epoch": 0.49,
"learning_rate": 1.9096319953545186e-05,
"loss": 0.0718,
"step": 85
},
{
"epoch": 0.49,
"learning_rate": 1.9070353005068484e-05,
"loss": 0.0888,
"step": 86
},
{
"epoch": 0.5,
"learning_rate": 1.9044036416274133e-05,
"loss": 0.0423,
"step": 87
},
{
"epoch": 0.51,
"learning_rate": 1.901737120160341e-05,
"loss": 0.0517,
"step": 88
},
{
"epoch": 0.51,
"learning_rate": 1.899035838893627e-05,
"loss": 0.0943,
"step": 89
},
{
"epoch": 0.52,
"learning_rate": 1.8962999019551755e-05,
"loss": 0.0598,
"step": 90
},
{
"epoch": 0.52,
"learning_rate": 1.8935294148087822e-05,
"loss": 0.0421,
"step": 91
},
{
"epoch": 0.53,
"learning_rate": 1.8907244842500706e-05,
"loss": 0.0857,
"step": 92
},
{
"epoch": 0.53,
"learning_rate": 1.8878852184023754e-05,
"loss": 0.0567,
"step": 93
},
{
"epoch": 0.54,
"learning_rate": 1.885011726712574e-05,
"loss": 0.0551,
"step": 94
},
{
"epoch": 0.55,
"learning_rate": 1.8821041199468677e-05,
"loss": 0.0335,
"step": 95
},
{
"epoch": 0.55,
"learning_rate": 1.8791625101865117e-05,
"loss": 0.0371,
"step": 96
},
{
"epoch": 0.56,
"learning_rate": 1.876187010823496e-05,
"loss": 0.0431,
"step": 97
},
{
"epoch": 0.56,
"learning_rate": 1.8731777365561723e-05,
"loss": 0.0842,
"step": 98
},
{
"epoch": 0.57,
"learning_rate": 1.870134803384834e-05,
"loss": 0.0509,
"step": 99
},
{
"epoch": 0.57,
"learning_rate": 1.8670583286072443e-05,
"loss": 0.0529,
"step": 100
},
{
"epoch": 0.58,
"learning_rate": 1.8639484308141157e-05,
"loss": 0.0678,
"step": 101
},
{
"epoch": 0.59,
"learning_rate": 1.860805229884536e-05,
"loss": 0.0664,
"step": 102
},
{
"epoch": 0.59,
"learning_rate": 1.8576288469813506e-05,
"loss": 0.0582,
"step": 103
},
{
"epoch": 0.6,
"learning_rate": 1.8544194045464888e-05,
"loss": 0.0479,
"step": 104
},
{
"epoch": 0.6,
"learning_rate": 1.8511770262962462e-05,
"loss": 0.1156,
"step": 105
},
{
"epoch": 0.61,
"learning_rate": 1.8479018372165148e-05,
"loss": 0.082,
"step": 106
},
{
"epoch": 0.61,
"learning_rate": 1.8445939635579658e-05,
"loss": 0.0565,
"step": 107
},
{
"epoch": 0.62,
"learning_rate": 1.8412535328311813e-05,
"loss": 0.0779,
"step": 108
},
{
"epoch": 0.63,
"learning_rate": 1.837880673801741e-05,
"loss": 0.0725,
"step": 109
},
{
"epoch": 0.63,
"learning_rate": 1.8344755164852572e-05,
"loss": 0.0877,
"step": 110
},
{
"epoch": 0.64,
"learning_rate": 1.8310381921423644e-05,
"loss": 0.0567,
"step": 111
},
{
"epoch": 0.64,
"learning_rate": 1.8275688332736577e-05,
"loss": 0.0575,
"step": 112
},
{
"epoch": 0.65,
"learning_rate": 1.8240675736145866e-05,
"loss": 0.0594,
"step": 113
},
{
"epoch": 0.66,
"learning_rate": 1.8205345481303e-05,
"loss": 0.0943,
"step": 114
},
{
"epoch": 0.66,
"learning_rate": 1.816969893010442e-05,
"loss": 0.0793,
"step": 115
},
{
"epoch": 0.67,
"learning_rate": 1.8133737456639043e-05,
"loss": 0.0405,
"step": 116
},
{
"epoch": 0.67,
"learning_rate": 1.8097462447135273e-05,
"loss": 0.0556,
"step": 117
},
{
"epoch": 0.68,
"learning_rate": 1.806087529990758e-05,
"loss": 0.0616,
"step": 118
},
{
"epoch": 0.68,
"learning_rate": 1.802397742530259e-05,
"loss": 0.0953,
"step": 119
},
{
"epoch": 0.69,
"learning_rate": 1.7986770245644728e-05,
"loss": 0.0875,
"step": 120
},
{
"epoch": 0.7,
"learning_rate": 1.7949255195181387e-05,
"loss": 0.0595,
"step": 121
},
{
"epoch": 0.7,
"learning_rate": 1.7911433720027624e-05,
"loss": 0.0854,
"step": 122
},
{
"epoch": 0.71,
"learning_rate": 1.787330727811045e-05,
"loss": 0.0588,
"step": 123
},
{
"epoch": 0.71,
"learning_rate": 1.7834877339112613e-05,
"loss": 0.0625,
"step": 124
},
{
"epoch": 0.72,
"learning_rate": 1.779614538441593e-05,
"loss": 0.0783,
"step": 125
},
{
"epoch": 0.72,
"learning_rate": 1.77571129070442e-05,
"loss": 0.0665,
"step": 126
},
{
"epoch": 0.73,
"learning_rate": 1.771778141160566e-05,
"loss": 0.0409,
"step": 127
},
{
"epoch": 0.74,
"learning_rate": 1.767815241423497e-05,
"loss": 0.0413,
"step": 128
},
{
"epoch": 0.74,
"learning_rate": 1.763822744253477e-05,
"loss": 0.0793,
"step": 129
},
{
"epoch": 0.75,
"learning_rate": 1.75980080355168e-05,
"loss": 0.0511,
"step": 130
},
{
"epoch": 0.75,
"learning_rate": 1.7557495743542586e-05,
"loss": 0.0596,
"step": 131
},
{
"epoch": 0.76,
"learning_rate": 1.7516692128263647e-05,
"loss": 0.0408,
"step": 132
},
{
"epoch": 0.76,
"learning_rate": 1.7475598762561333e-05,
"loss": 0.0682,
"step": 133
},
{
"epoch": 0.77,
"learning_rate": 1.7434217230486162e-05,
"loss": 0.0521,
"step": 134
},
{
"epoch": 0.78,
"learning_rate": 1.739254912719679e-05,
"loss": 0.0379,
"step": 135
},
{
"epoch": 0.78,
"learning_rate": 1.7350596058898482e-05,
"loss": 0.0396,
"step": 136
},
{
"epoch": 0.79,
"learning_rate": 1.730835964278124e-05,
"loss": 0.0865,
"step": 137
},
{
"epoch": 0.79,
"learning_rate": 1.726584150695744e-05,
"loss": 0.0425,
"step": 138
},
{
"epoch": 0.8,
"learning_rate": 1.7223043290399065e-05,
"loss": 0.0389,
"step": 139
},
{
"epoch": 0.8,
"learning_rate": 1.7179966642874543e-05,
"loss": 0.0416,
"step": 140
},
{
"epoch": 0.81,
"learning_rate": 1.7136613224885157e-05,
"loss": 0.0682,
"step": 141
},
{
"epoch": 0.82,
"learning_rate": 1.7092984707601007e-05,
"loss": 0.0344,
"step": 142
},
{
"epoch": 0.82,
"learning_rate": 1.7049082772796635e-05,
"loss": 0.0305,
"step": 143
},
{
"epoch": 0.83,
"learning_rate": 1.7004909112786142e-05,
"loss": 0.0351,
"step": 144
},
{
"epoch": 0.83,
"learning_rate": 1.696046543035801e-05,
"loss": 0.0331,
"step": 145
},
{
"epoch": 0.84,
"learning_rate": 1.6915753438709415e-05,
"loss": 0.0417,
"step": 146
},
{
"epoch": 0.84,
"learning_rate": 1.687077486138023e-05,
"loss": 0.0743,
"step": 147
},
{
"epoch": 0.85,
"learning_rate": 1.6825531432186545e-05,
"loss": 0.0496,
"step": 148
},
{
"epoch": 0.86,
"learning_rate": 1.6780024895153862e-05,
"loss": 0.0403,
"step": 149
},
{
"epoch": 0.86,
"learning_rate": 1.6734257004449863e-05,
"loss": 0.0485,
"step": 150
},
{
"epoch": 0.87,
"learning_rate": 1.6688229524316782e-05,
"loss": 0.0931,
"step": 151
},
{
"epoch": 0.87,
"learning_rate": 1.6641944229003395e-05,
"loss": 0.091,
"step": 152
},
{
"epoch": 0.88,
"learning_rate": 1.6595402902696648e-05,
"loss": 0.0559,
"step": 153
},
{
"epoch": 0.89,
"learning_rate": 1.6548607339452853e-05,
"loss": 0.0597,
"step": 154
},
{
"epoch": 0.89,
"learning_rate": 1.6501559343128553e-05,
"loss": 0.0834,
"step": 155
},
{
"epoch": 0.9,
"learning_rate": 1.645426072731098e-05,
"loss": 0.0368,
"step": 156
},
{
"epoch": 0.9,
"learning_rate": 1.640671331524814e-05,
"loss": 0.1033,
"step": 157
},
{
"epoch": 0.91,
"learning_rate": 1.6358918939778538e-05,
"loss": 0.1323,
"step": 158
},
{
"epoch": 0.91,
"learning_rate": 1.631087944326053e-05,
"loss": 0.0745,
"step": 159
},
{
"epoch": 0.92,
"learning_rate": 1.6262596677501297e-05,
"loss": 0.0638,
"step": 160
},
{
"epoch": 0.93,
"learning_rate": 1.621407250368546e-05,
"loss": 0.1301,
"step": 161
},
{
"epoch": 0.93,
"learning_rate": 1.616530879230335e-05,
"loss": 0.0497,
"step": 162
},
{
"epoch": 0.94,
"learning_rate": 1.611630742307889e-05,
"loss": 0.0638,
"step": 163
},
{
"epoch": 0.94,
"learning_rate": 1.606707028489714e-05,
"loss": 0.0825,
"step": 164
},
{
"epoch": 0.95,
"learning_rate": 1.6017599275731482e-05,
"loss": 0.0518,
"step": 165
},
{
"epoch": 0.95,
"learning_rate": 1.5967896302570485e-05,
"loss": 0.0581,
"step": 166
},
{
"epoch": 0.96,
"learning_rate": 1.5917963281344346e-05,
"loss": 0.0968,
"step": 167
},
{
"epoch": 0.97,
"learning_rate": 1.5867802136851078e-05,
"loss": 0.0735,
"step": 168
},
{
"epoch": 0.97,
"learning_rate": 1.5817414802682292e-05,
"loss": 0.0508,
"step": 169
},
{
"epoch": 0.98,
"learning_rate": 1.5766803221148676e-05,
"loss": 0.0701,
"step": 170
},
{
"epoch": 0.98,
"learning_rate": 1.5715969343205094e-05,
"loss": 0.0579,
"step": 171
},
{
"epoch": 0.99,
"learning_rate": 1.566491512837543e-05,
"loss": 0.0605,
"step": 172
},
{
"epoch": 0.99,
"learning_rate": 1.561364254467701e-05,
"loss": 0.0522,
"step": 173
},
{
"epoch": 1.0,
"learning_rate": 1.5562153568544753e-05,
"loss": 0.061,
"step": 174
},
{
"epoch": 1.01,
"learning_rate": 1.5510450184754996e-05,
"loss": 0.0644,
"step": 175
},
{
"epoch": 1.01,
"learning_rate": 1.5458534386348967e-05,
"loss": 0.0639,
"step": 176
},
{
"epoch": 1.02,
"learning_rate": 1.5406408174555978e-05,
"loss": 0.0509,
"step": 177
},
{
"epoch": 1.02,
"learning_rate": 1.535407355871626e-05,
"loss": 0.077,
"step": 178
},
{
"epoch": 1.03,
"learning_rate": 1.5301532556203524e-05,
"loss": 0.0389,
"step": 179
},
{
"epoch": 1.03,
"learning_rate": 1.5248787192347196e-05,
"loss": 0.0451,
"step": 180
},
{
"epoch": 1.04,
"learning_rate": 1.5195839500354337e-05,
"loss": 0.0536,
"step": 181
},
{
"epoch": 1.05,
"learning_rate": 1.5142691521231267e-05,
"loss": 0.0656,
"step": 182
},
{
"epoch": 1.05,
"learning_rate": 1.5089345303704901e-05,
"loss": 0.0416,
"step": 183
},
{
"epoch": 1.06,
"learning_rate": 1.503580290414376e-05,
"loss": 0.05,
"step": 184
},
{
"epoch": 1.06,
"learning_rate": 1.4982066386478727e-05,
"loss": 0.0606,
"step": 185
},
{
"epoch": 1.07,
"learning_rate": 1.4928137822123451e-05,
"loss": 0.0574,
"step": 186
},
{
"epoch": 1.07,
"learning_rate": 1.4874019289894536e-05,
"loss": 0.0398,
"step": 187
},
{
"epoch": 1.08,
"learning_rate": 1.481971287593138e-05,
"loss": 0.0621,
"step": 188
},
{
"epoch": 1.09,
"learning_rate": 1.4765220673615774e-05,
"loss": 0.0453,
"step": 189
},
{
"epoch": 1.09,
"learning_rate": 1.471054478349121e-05,
"loss": 0.0505,
"step": 190
},
{
"epoch": 1.1,
"learning_rate": 1.4655687313181896e-05,
"loss": 0.0378,
"step": 191
},
{
"epoch": 1.1,
"learning_rate": 1.4600650377311523e-05,
"loss": 0.0464,
"step": 192
},
{
"epoch": 1.11,
"learning_rate": 1.4545436097421745e-05,
"loss": 0.0583,
"step": 193
},
{
"epoch": 1.11,
"learning_rate": 1.4490046601890405e-05,
"loss": 0.0598,
"step": 194
},
{
"epoch": 1.12,
"learning_rate": 1.4434484025849494e-05,
"loss": 0.027,
"step": 195
},
{
"epoch": 1.13,
"learning_rate": 1.4378750511102826e-05,
"loss": 0.0233,
"step": 196
},
{
"epoch": 1.13,
"learning_rate": 1.4322848206043505e-05,
"loss": 0.061,
"step": 197
},
{
"epoch": 1.14,
"learning_rate": 1.4266779265571087e-05,
"loss": 0.0422,
"step": 198
},
{
"epoch": 1.14,
"learning_rate": 1.4210545851008529e-05,
"loss": 0.0427,
"step": 199
},
{
"epoch": 1.15,
"learning_rate": 1.4154150130018867e-05,
"loss": 0.0312,
"step": 200
},
{
"epoch": 1.16,
"learning_rate": 1.4097594276521658e-05,
"loss": 0.053,
"step": 201
},
{
"epoch": 1.16,
"learning_rate": 1.4040880470609186e-05,
"loss": 0.0376,
"step": 202
},
{
"epoch": 1.17,
"learning_rate": 1.3984010898462417e-05,
"loss": 0.0529,
"step": 203
},
{
"epoch": 1.17,
"learning_rate": 1.3926987752266733e-05,
"loss": 0.095,
"step": 204
},
{
"epoch": 1.18,
"learning_rate": 1.3869813230127432e-05,
"loss": 0.0323,
"step": 205
},
{
"epoch": 1.18,
"learning_rate": 1.381248953598498e-05,
"loss": 0.0303,
"step": 206
},
{
"epoch": 1.19,
"learning_rate": 1.3755018879530075e-05,
"loss": 0.0278,
"step": 207
},
{
"epoch": 1.2,
"learning_rate": 1.3697403476118453e-05,
"loss": 0.0428,
"step": 208
},
{
"epoch": 1.2,
"learning_rate": 1.36396455466855e-05,
"loss": 0.0643,
"step": 209
},
{
"epoch": 1.21,
"learning_rate": 1.3581747317660641e-05,
"loss": 0.0391,
"step": 210
},
{
"epoch": 1.21,
"learning_rate": 1.3523711020881504e-05,
"loss": 0.073,
"step": 211
},
{
"epoch": 1.22,
"learning_rate": 1.3465538893507908e-05,
"loss": 0.0438,
"step": 212
},
{
"epoch": 1.22,
"learning_rate": 1.3407233177935608e-05,
"loss": 0.0421,
"step": 213
},
{
"epoch": 1.23,
"learning_rate": 1.3348796121709862e-05,
"loss": 0.0359,
"step": 214
},
{
"epoch": 1.24,
"learning_rate": 1.3290229977438802e-05,
"loss": 0.0508,
"step": 215
},
{
"epoch": 1.24,
"learning_rate": 1.3231537002706594e-05,
"loss": 0.0728,
"step": 216
},
{
"epoch": 1.25,
"learning_rate": 1.31727194599864e-05,
"loss": 0.0435,
"step": 217
},
{
"epoch": 1.25,
"learning_rate": 1.3113779616553189e-05,
"loss": 0.0311,
"step": 218
},
{
"epoch": 1.26,
"learning_rate": 1.3054719744396333e-05,
"loss": 0.0303,
"step": 219
},
{
"epoch": 1.26,
"learning_rate": 1.2995542120132016e-05,
"loss": 0.0245,
"step": 220
},
{
"epoch": 1.27,
"learning_rate": 1.2936249024915487e-05,
"loss": 0.0234,
"step": 221
},
{
"epoch": 1.28,
"learning_rate": 1.2876842744353113e-05,
"loss": 0.055,
"step": 222
},
{
"epoch": 1.28,
"learning_rate": 1.2817325568414299e-05,
"loss": 0.0601,
"step": 223
},
{
"epoch": 1.29,
"learning_rate": 1.2757699791343188e-05,
"loss": 0.0719,
"step": 224
},
{
"epoch": 1.29,
"learning_rate": 1.2697967711570243e-05,
"loss": 0.0415,
"step": 225
},
{
"epoch": 1.3,
"learning_rate": 1.2638131631623641e-05,
"loss": 0.0256,
"step": 226
},
{
"epoch": 1.3,
"learning_rate": 1.2578193858040508e-05,
"loss": 0.0712,
"step": 227
},
{
"epoch": 1.31,
"learning_rate": 1.251815670127802e-05,
"loss": 0.0606,
"step": 228
},
{
"epoch": 1.32,
"learning_rate": 1.2458022475624343e-05,
"loss": 0.0509,
"step": 229
},
{
"epoch": 1.32,
"learning_rate": 1.2397793499109403e-05,
"loss": 0.0687,
"step": 230
},
{
"epoch": 1.33,
"learning_rate": 1.2337472093415554e-05,
"loss": 0.0389,
"step": 231
},
{
"epoch": 1.33,
"learning_rate": 1.2277060583788065e-05,
"loss": 0.0505,
"step": 232
},
{
"epoch": 1.34,
"learning_rate": 1.2216561298945502e-05,
"loss": 0.0419,
"step": 233
},
{
"epoch": 1.34,
"learning_rate": 1.2155976570989949e-05,
"loss": 0.0465,
"step": 234
},
{
"epoch": 1.35,
"learning_rate": 1.2095308735317126e-05,
"loss": 0.0345,
"step": 235
},
{
"epoch": 1.36,
"learning_rate": 1.2034560130526341e-05,
"loss": 0.0346,
"step": 236
},
{
"epoch": 1.36,
"learning_rate": 1.197373309833037e-05,
"loss": 0.0518,
"step": 237
},
{
"epoch": 1.37,
"learning_rate": 1.1912829983465169e-05,
"loss": 0.037,
"step": 238
},
{
"epoch": 1.37,
"learning_rate": 1.1851853133599507e-05,
"loss": 0.0292,
"step": 239
},
{
"epoch": 1.38,
"learning_rate": 1.1790804899244453e-05,
"loss": 0.0217,
"step": 240
},
{
"epoch": 1.39,
"learning_rate": 1.1729687633662776e-05,
"loss": 0.0523,
"step": 241
},
{
"epoch": 1.39,
"learning_rate": 1.1668503692778239e-05,
"loss": 0.0457,
"step": 242
},
{
"epoch": 1.4,
"learning_rate": 1.1607255435084772e-05,
"loss": 0.0521,
"step": 243
},
{
"epoch": 1.4,
"learning_rate": 1.1545945221555571e-05,
"loss": 0.0435,
"step": 244
},
{
"epoch": 1.41,
"learning_rate": 1.1484575415552069e-05,
"loss": 0.0728,
"step": 245
},
{
"epoch": 1.41,
"learning_rate": 1.1423148382732854e-05,
"loss": 0.0434,
"step": 246
},
{
"epoch": 1.42,
"learning_rate": 1.1361666490962468e-05,
"loss": 0.0429,
"step": 247
},
{
"epoch": 1.43,
"learning_rate": 1.1300132110220135e-05,
"loss": 0.0852,
"step": 248
},
{
"epoch": 1.43,
"learning_rate": 1.12385476125084e-05,
"loss": 0.0476,
"step": 249
},
{
"epoch": 1.44,
"learning_rate": 1.1176915371761702e-05,
"loss": 0.0546,
"step": 250
},
{
"epoch": 1.44,
"learning_rate": 1.1115237763754856e-05,
"loss": 0.068,
"step": 251
},
{
"epoch": 1.45,
"learning_rate": 1.1053517166011471e-05,
"loss": 0.0766,
"step": 252
},
{
"epoch": 1.45,
"learning_rate": 1.0991755957712317e-05,
"loss": 0.0683,
"step": 253
},
{
"epoch": 1.46,
"learning_rate": 1.0929956519603595e-05,
"loss": 0.0572,
"step": 254
},
{
"epoch": 1.47,
"learning_rate": 1.0868121233905183e-05,
"loss": 0.029,
"step": 255
},
{
"epoch": 1.47,
"learning_rate": 1.080625248421878e-05,
"loss": 0.0723,
"step": 256
},
{
"epoch": 1.48,
"learning_rate": 1.0744352655436058e-05,
"loss": 0.0753,
"step": 257
},
{
"epoch": 1.48,
"learning_rate": 1.0682424133646712e-05,
"loss": 0.0322,
"step": 258
},
{
"epoch": 1.49,
"learning_rate": 1.0620469306046473e-05,
"loss": 0.0295,
"step": 259
},
{
"epoch": 1.49,
"learning_rate": 1.0558490560845107e-05,
"loss": 0.028,
"step": 260
},
{
"epoch": 1.5,
"learning_rate": 1.0496490287174343e-05,
"loss": 0.0317,
"step": 261
},
{
"epoch": 1.51,
"learning_rate": 1.0434470874995781e-05,
"loss": 0.0646,
"step": 262
},
{
"epoch": 1.51,
"learning_rate": 1.0372434715008763e-05,
"loss": 0.0498,
"step": 263
},
{
"epoch": 1.52,
"learning_rate": 1.0310384198558226e-05,
"loss": 0.0415,
"step": 264
},
{
"epoch": 1.52,
"learning_rate": 1.0248321717542505e-05,
"loss": 0.0381,
"step": 265
},
{
"epoch": 1.53,
"learning_rate": 1.0186249664321139e-05,
"loss": 0.0383,
"step": 266
},
{
"epoch": 1.53,
"learning_rate": 1.012417043162266e-05,
"loss": 0.045,
"step": 267
},
{
"epoch": 1.54,
"learning_rate": 1.0062086412452353e-05,
"loss": 0.0271,
"step": 268
},
{
"epoch": 1.55,
"learning_rate": 1e-05,
"loss": 0.0518,
"step": 269
},
{
"epoch": 1.55,
"learning_rate": 9.93791358754765e-06,
"loss": 0.0501,
"step": 270
},
{
"epoch": 1.56,
"learning_rate": 9.875829568377343e-06,
"loss": 0.0346,
"step": 271
},
{
"epoch": 1.56,
"learning_rate": 9.813750335678866e-06,
"loss": 0.0658,
"step": 272
},
{
"epoch": 1.57,
"learning_rate": 9.751678282457502e-06,
"loss": 0.0356,
"step": 273
},
{
"epoch": 1.57,
"learning_rate": 9.689615801441776e-06,
"loss": 0.0508,
"step": 274
},
{
"epoch": 1.58,
"learning_rate": 9.627565284991237e-06,
"loss": 0.0346,
"step": 275
},
{
"epoch": 1.59,
"learning_rate": 9.56552912500422e-06,
"loss": 0.0546,
"step": 276
},
{
"epoch": 1.59,
"learning_rate": 9.50350971282566e-06,
"loss": 0.0434,
"step": 277
},
{
"epoch": 1.6,
"learning_rate": 9.441509439154896e-06,
"loss": 0.0555,
"step": 278
},
{
"epoch": 1.6,
"learning_rate": 9.37953069395353e-06,
"loss": 0.0463,
"step": 279
},
{
"epoch": 1.61,
"learning_rate": 9.317575866353293e-06,
"loss": 0.0513,
"step": 280
},
{
"epoch": 1.61,
"learning_rate": 9.255647344563943e-06,
"loss": 0.0455,
"step": 281
},
{
"epoch": 1.62,
"learning_rate": 9.193747515781225e-06,
"loss": 0.0433,
"step": 282
},
{
"epoch": 1.63,
"learning_rate": 9.131878766094822e-06,
"loss": 0.0483,
"step": 283
},
{
"epoch": 1.63,
"learning_rate": 9.070043480396404e-06,
"loss": 0.0495,
"step": 284
},
{
"epoch": 1.64,
"learning_rate": 9.008244042287684e-06,
"loss": 0.0322,
"step": 285
},
{
"epoch": 1.64,
"learning_rate": 8.94648283398853e-06,
"loss": 0.0293,
"step": 286
},
{
"epoch": 1.65,
"learning_rate": 8.884762236245146e-06,
"loss": 0.0691,
"step": 287
},
{
"epoch": 1.66,
"learning_rate": 8.8230846282383e-06,
"loss": 0.0315,
"step": 288
},
{
"epoch": 1.66,
"learning_rate": 8.761452387491601e-06,
"loss": 0.062,
"step": 289
},
{
"epoch": 1.67,
"learning_rate": 8.699867889779869e-06,
"loss": 0.0371,
"step": 290
},
{
"epoch": 1.67,
"learning_rate": 8.638333509037537e-06,
"loss": 0.0355,
"step": 291
},
{
"epoch": 1.68,
"learning_rate": 8.576851617267151e-06,
"loss": 0.0345,
"step": 292
},
{
"epoch": 1.68,
"learning_rate": 8.515424584447935e-06,
"loss": 0.0347,
"step": 293
},
{
"epoch": 1.69,
"learning_rate": 8.45405477844443e-06,
"loss": 0.0553,
"step": 294
},
{
"epoch": 1.7,
"learning_rate": 8.392744564915228e-06,
"loss": 0.0448,
"step": 295
},
{
"epoch": 1.7,
"learning_rate": 8.331496307221763e-06,
"loss": 0.0358,
"step": 296
},
{
"epoch": 1.71,
"learning_rate": 8.270312366337227e-06,
"loss": 0.043,
"step": 297
},
{
"epoch": 1.71,
"learning_rate": 8.209195100755552e-06,
"loss": 0.0394,
"step": 298
},
{
"epoch": 1.72,
"learning_rate": 8.148146866400498e-06,
"loss": 0.0373,
"step": 299
},
{
"epoch": 1.72,
"learning_rate": 8.087170016534836e-06,
"loss": 0.0263,
"step": 300
},
{
"epoch": 1.73,
"learning_rate": 8.026266901669635e-06,
"loss": 0.0435,
"step": 301
},
{
"epoch": 1.74,
"learning_rate": 7.965439869473664e-06,
"loss": 0.0352,
"step": 302
},
{
"epoch": 1.74,
"learning_rate": 7.90469126468288e-06,
"loss": 0.0287,
"step": 303
},
{
"epoch": 1.75,
"learning_rate": 7.844023429010051e-06,
"loss": 0.0255,
"step": 304
},
{
"epoch": 1.75,
"learning_rate": 7.7834387010545e-06,
"loss": 0.0532,
"step": 305
},
{
"epoch": 1.76,
"learning_rate": 7.722939416211938e-06,
"loss": 0.0298,
"step": 306
},
{
"epoch": 1.76,
"learning_rate": 7.66252790658445e-06,
"loss": 0.0416,
"step": 307
},
{
"epoch": 1.77,
"learning_rate": 7.6022065008906e-06,
"loss": 0.0505,
"step": 308
},
{
"epoch": 1.78,
"learning_rate": 7.541977524375661e-06,
"loss": 0.0235,
"step": 309
},
{
"epoch": 1.78,
"learning_rate": 7.481843298721982e-06,
"loss": 0.0246,
"step": 310
},
{
"epoch": 1.79,
"learning_rate": 7.421806141959498e-06,
"loss": 0.0587,
"step": 311
},
{
"epoch": 1.79,
"learning_rate": 7.361868368376364e-06,
"loss": 0.0551,
"step": 312
},
{
"epoch": 1.8,
"learning_rate": 7.3020322884297565e-06,
"loss": 0.0473,
"step": 313
},
{
"epoch": 1.8,
"learning_rate": 7.242300208656814e-06,
"loss": 0.0434,
"step": 314
},
{
"epoch": 1.81,
"learning_rate": 7.182674431585703e-06,
"loss": 0.0476,
"step": 315
},
{
"epoch": 1.82,
"learning_rate": 7.123157255646889e-06,
"loss": 0.0361,
"step": 316
},
{
"epoch": 1.82,
"learning_rate": 7.0637509750845175e-06,
"loss": 0.0277,
"step": 317
},
{
"epoch": 1.83,
"learning_rate": 7.004457879867987e-06,
"loss": 0.0484,
"step": 318
},
{
"epoch": 1.83,
"learning_rate": 6.9452802556036705e-06,
"loss": 0.0719,
"step": 319
},
{
"epoch": 1.84,
"learning_rate": 6.886220383446814e-06,
"loss": 0.0343,
"step": 320
},
{
"epoch": 1.84,
"learning_rate": 6.827280540013607e-06,
"loss": 0.0318,
"step": 321
},
{
"epoch": 1.85,
"learning_rate": 6.768462997293412e-06,
"loss": 0.0312,
"step": 322
},
{
"epoch": 1.86,
"learning_rate": 6.709770022561197e-06,
"loss": 0.0354,
"step": 323
},
{
"epoch": 1.86,
"learning_rate": 6.651203878290139e-06,
"loss": 0.0203,
"step": 324
},
{
"epoch": 1.87,
"learning_rate": 6.592766822064396e-06,
"loss": 0.0476,
"step": 325
},
{
"epoch": 1.87,
"learning_rate": 6.534461106492094e-06,
"loss": 0.0392,
"step": 326
},
{
"epoch": 1.88,
"learning_rate": 6.476288979118497e-06,
"loss": 0.0513,
"step": 327
},
{
"epoch": 1.89,
"learning_rate": 6.418252682339362e-06,
"loss": 0.0377,
"step": 328
},
{
"epoch": 1.89,
"learning_rate": 6.360354453314502e-06,
"loss": 0.0247,
"step": 329
},
{
"epoch": 1.9,
"learning_rate": 6.30259652388155e-06,
"loss": 0.0461,
"step": 330
},
{
"epoch": 1.9,
"learning_rate": 6.244981120469928e-06,
"loss": 0.0289,
"step": 331
},
{
"epoch": 1.91,
"learning_rate": 6.187510464015023e-06,
"loss": 0.0363,
"step": 332
},
{
"epoch": 1.91,
"learning_rate": 6.13018676987257e-06,
"loss": 0.0261,
"step": 333
},
{
"epoch": 1.92,
"learning_rate": 6.073012247733267e-06,
"loss": 0.0124,
"step": 334
},
{
"epoch": 1.93,
"learning_rate": 6.015989101537586e-06,
"loss": 0.049,
"step": 335
},
{
"epoch": 1.93,
"learning_rate": 5.959119529390818e-06,
"loss": 0.019,
"step": 336
},
{
"epoch": 1.94,
"learning_rate": 5.902405723478346e-06,
"loss": 0.0537,
"step": 337
},
{
"epoch": 1.94,
"learning_rate": 5.845849869981137e-06,
"loss": 0.0293,
"step": 338
},
{
"epoch": 1.95,
"learning_rate": 5.789454148991477e-06,
"loss": 0.0218,
"step": 339
},
{
"epoch": 1.95,
"learning_rate": 5.733220734428919e-06,
"loss": 0.0476,
"step": 340
},
{
"epoch": 1.96,
"learning_rate": 5.677151793956501e-06,
"loss": 0.0371,
"step": 341
},
{
"epoch": 1.97,
"learning_rate": 5.621249488897176e-06,
"loss": 0.0169,
"step": 342
},
{
"epoch": 1.97,
"learning_rate": 5.5655159741505085e-06,
"loss": 0.022,
"step": 343
},
{
"epoch": 1.98,
"learning_rate": 5.5099533981095945e-06,
"loss": 0.0277,
"step": 344
},
{
"epoch": 1.98,
"learning_rate": 5.454563902578256e-06,
"loss": 0.0329,
"step": 345
},
{
"epoch": 1.99,
"learning_rate": 5.399349622688479e-06,
"loss": 0.0201,
"step": 346
},
{
"epoch": 1.99,
"learning_rate": 5.344312686818106e-06,
"loss": 0.028,
"step": 347
},
{
"epoch": 2.0,
"learning_rate": 5.289455216508792e-06,
"loss": 0.0835,
"step": 348
},
{
"epoch": 2.01,
"learning_rate": 5.234779326384227e-06,
"loss": 0.0333,
"step": 349
},
{
"epoch": 2.01,
"learning_rate": 5.180287124068624e-06,
"loss": 0.0199,
"step": 350
},
{
"epoch": 2.02,
"learning_rate": 5.1259807101054686e-06,
"loss": 0.0414,
"step": 351
},
{
"epoch": 2.02,
"learning_rate": 5.071862177876548e-06,
"loss": 0.0489,
"step": 352
},
{
"epoch": 2.03,
"learning_rate": 5.017933613521273e-06,
"loss": 0.017,
"step": 353
},
{
"epoch": 2.03,
"learning_rate": 4.964197095856237e-06,
"loss": 0.0279,
"step": 354
},
{
"epoch": 2.04,
"learning_rate": 4.9106546962951e-06,
"loss": 0.0298,
"step": 355
},
{
"epoch": 2.05,
"learning_rate": 4.857308478768733e-06,
"loss": 0.0287,
"step": 356
},
{
"epoch": 2.05,
"learning_rate": 4.804160499645667e-06,
"loss": 0.0272,
"step": 357
},
{
"epoch": 2.06,
"learning_rate": 4.7512128076528065e-06,
"loss": 0.0309,
"step": 358
},
{
"epoch": 2.06,
"learning_rate": 4.698467443796479e-06,
"loss": 0.0412,
"step": 359
},
{
"epoch": 2.07,
"learning_rate": 4.645926441283745e-06,
"loss": 0.0247,
"step": 360
},
{
"epoch": 2.07,
"learning_rate": 4.593591825444028e-06,
"loss": 0.0318,
"step": 361
},
{
"epoch": 2.08,
"learning_rate": 4.541465613651034e-06,
"loss": 0.0347,
"step": 362
},
{
"epoch": 2.09,
"learning_rate": 4.489549815245008e-06,
"loss": 0.0384,
"step": 363
},
{
"epoch": 2.09,
"learning_rate": 4.437846431455249e-06,
"loss": 0.0388,
"step": 364
},
{
"epoch": 2.1,
"learning_rate": 4.3863574553229936e-06,
"loss": 0.0434,
"step": 365
},
{
"epoch": 2.1,
"learning_rate": 4.335084871624572e-06,
"loss": 0.0437,
"step": 366
},
{
"epoch": 2.11,
"learning_rate": 4.284030656794908e-06,
"loss": 0.0245,
"step": 367
},
{
"epoch": 2.11,
"learning_rate": 4.2331967788513295e-06,
"loss": 0.0237,
"step": 368
},
{
"epoch": 2.12,
"learning_rate": 4.182585197317709e-06,
"loss": 0.0315,
"step": 369
},
{
"epoch": 2.13,
"learning_rate": 4.132197863148925e-06,
"loss": 0.036,
"step": 370
},
{
"epoch": 2.13,
"learning_rate": 4.082036718655657e-06,
"loss": 0.0301,
"step": 371
},
{
"epoch": 2.14,
"learning_rate": 4.032103697429516e-06,
"loss": 0.044,
"step": 372
},
{
"epoch": 2.14,
"learning_rate": 3.982400724268516e-06,
"loss": 0.0406,
"step": 373
},
{
"epoch": 2.15,
"learning_rate": 3.932929715102863e-06,
"loss": 0.0408,
"step": 374
},
{
"epoch": 2.16,
"learning_rate": 3.883692576921111e-06,
"loss": 0.0158,
"step": 375
},
{
"epoch": 2.16,
"learning_rate": 3.834691207696649e-06,
"loss": 0.0174,
"step": 376
},
{
"epoch": 2.17,
"learning_rate": 3.785927496314543e-06,
"loss": 0.0298,
"step": 377
},
{
"epoch": 2.17,
"learning_rate": 3.7374033224987084e-06,
"loss": 0.026,
"step": 378
},
{
"epoch": 2.18,
"learning_rate": 3.689120556739475e-06,
"loss": 0.0276,
"step": 379
},
{
"epoch": 2.18,
"learning_rate": 3.6410810602214687e-06,
"loss": 0.0461,
"step": 380
},
{
"epoch": 2.19,
"learning_rate": 3.5932866847518654e-06,
"loss": 0.0296,
"step": 381
},
{
"epoch": 2.2,
"learning_rate": 3.545739272689024e-06,
"loss": 0.0433,
"step": 382
},
{
"epoch": 2.2,
"learning_rate": 3.498440656871449e-06,
"loss": 0.017,
"step": 383
},
{
"epoch": 2.21,
"learning_rate": 3.4513926605471504e-06,
"loss": 0.0303,
"step": 384
},
{
"epoch": 2.21,
"learning_rate": 3.404597097303356e-06,
"loss": 0.0378,
"step": 385
},
{
"epoch": 2.22,
"learning_rate": 3.3580557709966066e-06,
"loss": 0.0352,
"step": 386
},
{
"epoch": 2.22,
"learning_rate": 3.3117704756832226e-06,
"loss": 0.0388,
"step": 387
},
{
"epoch": 2.23,
"learning_rate": 3.2657429955501396e-06,
"loss": 0.0247,
"step": 388
},
{
"epoch": 2.24,
"learning_rate": 3.2199751048461414e-06,
"loss": 0.0159,
"step": 389
},
{
"epoch": 2.24,
"learning_rate": 3.174468567813461e-06,
"loss": 0.0131,
"step": 390
},
{
"epoch": 2.25,
"learning_rate": 3.1292251386197727e-06,
"loss": 0.0393,
"step": 391
},
{
"epoch": 2.25,
"learning_rate": 3.084246561290584e-06,
"loss": 0.0113,
"step": 392
},
{
"epoch": 2.26,
"learning_rate": 3.039534569641992e-06,
"loss": 0.0297,
"step": 393
},
{
"epoch": 2.26,
"learning_rate": 2.9950908872138585e-06,
"loss": 0.0223,
"step": 394
},
{
"epoch": 2.27,
"learning_rate": 2.9509172272033683e-06,
"loss": 0.0306,
"step": 395
},
{
"epoch": 2.28,
"learning_rate": 2.9070152923989947e-06,
"loss": 0.0274,
"step": 396
},
{
"epoch": 2.28,
"learning_rate": 2.863386775114848e-06,
"loss": 0.0085,
"step": 397
},
{
"epoch": 2.29,
"learning_rate": 2.8200333571254603e-06,
"loss": 0.0269,
"step": 398
},
{
"epoch": 2.29,
"learning_rate": 2.776956709600941e-06,
"loss": 0.028,
"step": 399
},
{
"epoch": 2.3,
"learning_rate": 2.7341584930425656e-06,
"loss": 0.0118,
"step": 400
},
{
"epoch": 2.3,
"learning_rate": 2.691640357218759e-06,
"loss": 0.0307,
"step": 401
},
{
"epoch": 2.31,
"learning_rate": 2.6494039411015192e-06,
"loss": 0.02,
"step": 402
},
{
"epoch": 2.32,
"learning_rate": 2.6074508728032135e-06,
"loss": 0.0219,
"step": 403
},
{
"epoch": 2.32,
"learning_rate": 2.565782769513837e-06,
"loss": 0.0268,
"step": 404
},
{
"epoch": 2.33,
"learning_rate": 2.5244012374386695e-06,
"loss": 0.0181,
"step": 405
},
{
"epoch": 2.33,
"learning_rate": 2.4833078717363544e-06,
"loss": 0.0239,
"step": 406
},
{
"epoch": 2.34,
"learning_rate": 2.4425042564574186e-06,
"loss": 0.0205,
"step": 407
},
{
"epoch": 2.34,
"learning_rate": 2.4019919644832023e-06,
"loss": 0.0238,
"step": 408
},
{
"epoch": 2.35,
"learning_rate": 2.3617725574652352e-06,
"loss": 0.0221,
"step": 409
},
{
"epoch": 2.36,
"learning_rate": 2.3218475857650347e-06,
"loss": 0.0183,
"step": 410
},
{
"epoch": 2.36,
"learning_rate": 2.2822185883943404e-06,
"loss": 0.0208,
"step": 411
},
{
"epoch": 2.37,
"learning_rate": 2.2428870929558012e-06,
"loss": 0.0366,
"step": 412
},
{
"epoch": 2.37,
"learning_rate": 2.2038546155840733e-06,
"loss": 0.0181,
"step": 413
},
{
"epoch": 2.38,
"learning_rate": 2.1651226608873876e-06,
"loss": 0.0235,
"step": 414
},
{
"epoch": 2.39,
"learning_rate": 2.1266927218895507e-06,
"loss": 0.039,
"step": 415
},
{
"epoch": 2.39,
"learning_rate": 2.0885662799723807e-06,
"loss": 0.0205,
"step": 416
},
{
"epoch": 2.4,
"learning_rate": 2.0507448048186205e-06,
"loss": 0.0209,
"step": 417
},
{
"epoch": 2.4,
"learning_rate": 2.013229754355276e-06,
"loss": 0.0217,
"step": 418
},
{
"epoch": 2.41,
"learning_rate": 1.9760225746974136e-06,
"loss": 0.0346,
"step": 419
},
{
"epoch": 2.41,
"learning_rate": 1.9391247000924227e-06,
"loss": 0.0437,
"step": 420
},
{
"epoch": 2.42,
"learning_rate": 1.9025375528647294e-06,
"loss": 0.0241,
"step": 421
},
{
"epoch": 2.43,
"learning_rate": 1.8662625433609582e-06,
"loss": 0.0282,
"step": 422
},
{
"epoch": 2.43,
"learning_rate": 1.8303010698955803e-06,
"loss": 0.0282,
"step": 423
},
{
"epoch": 2.44,
"learning_rate": 1.7946545186970022e-06,
"loss": 0.0202,
"step": 424
},
{
"epoch": 2.44,
"learning_rate": 1.7593242638541352e-06,
"loss": 0.0394,
"step": 425
},
{
"epoch": 2.45,
"learning_rate": 1.7243116672634264e-06,
"loss": 0.0199,
"step": 426
},
{
"epoch": 2.45,
"learning_rate": 1.6896180785763594e-06,
"loss": 0.0248,
"step": 427
},
{
"epoch": 2.46,
"learning_rate": 1.6552448351474305e-06,
"loss": 0.0064,
"step": 428
},
{
"epoch": 2.47,
"learning_rate": 1.6211932619825932e-06,
"loss": 0.029,
"step": 429
},
{
"epoch": 2.47,
"learning_rate": 1.587464671688187e-06,
"loss": 0.0094,
"step": 430
},
{
"epoch": 2.48,
"learning_rate": 1.5540603644203423e-06,
"loss": 0.0314,
"step": 431
},
{
"epoch": 2.48,
"learning_rate": 1.520981627834851e-06,
"loss": 0.0179,
"step": 432
},
{
"epoch": 2.49,
"learning_rate": 1.4882297370375388e-06,
"loss": 0.0173,
"step": 433
},
{
"epoch": 2.49,
"learning_rate": 1.4558059545351144e-06,
"loss": 0.0153,
"step": 434
},
{
"epoch": 2.5,
"learning_rate": 1.423711530186498e-06,
"loss": 0.0173,
"step": 435
},
{
"epoch": 2.51,
"learning_rate": 1.3919477011546423e-06,
"loss": 0.0516,
"step": 436
},
{
"epoch": 2.51,
"learning_rate": 1.360515691858847e-06,
"loss": 0.0293,
"step": 437
},
{
"epoch": 2.52,
"learning_rate": 1.3294167139275594e-06,
"loss": 0.0132,
"step": 438
},
{
"epoch": 2.52,
"learning_rate": 1.298651966151665e-06,
"loss": 0.0345,
"step": 439
},
{
"epoch": 2.53,
"learning_rate": 1.2682226344382798e-06,
"loss": 0.0168,
"step": 440
},
{
"epoch": 2.53,
"learning_rate": 1.2381298917650408e-06,
"loss": 0.0179,
"step": 441
},
{
"epoch": 2.54,
"learning_rate": 1.2083748981348832e-06,
"loss": 0.0179,
"step": 442
},
{
"epoch": 2.55,
"learning_rate": 1.1789588005313257e-06,
"loss": 0.0188,
"step": 443
},
{
"epoch": 2.55,
"learning_rate": 1.1498827328742623e-06,
"loss": 0.0183,
"step": 444
},
{
"epoch": 2.56,
"learning_rate": 1.121147815976248e-06,
"loss": 0.0221,
"step": 445
},
{
"epoch": 2.56,
"learning_rate": 1.0927551574992967e-06,
"loss": 0.0277,
"step": 446
},
{
"epoch": 2.57,
"learning_rate": 1.0647058519121822e-06,
"loss": 0.0095,
"step": 447
},
{
"epoch": 2.57,
"learning_rate": 1.0370009804482483e-06,
"loss": 0.0457,
"step": 448
},
{
"epoch": 2.58,
"learning_rate": 1.009641611063732e-06,
"loss": 0.0223,
"step": 449
},
{
"epoch": 2.59,
"learning_rate": 9.82628798396592e-07,
"loss": 0.0406,
"step": 450
},
{
"epoch": 2.59,
"learning_rate": 9.559635837258653e-07,
"loss": 0.0112,
"step": 451
},
{
"epoch": 2.6,
"learning_rate": 9.296469949315157e-07,
"loss": 0.0277,
"step": 452
},
{
"epoch": 2.6,
"learning_rate": 9.036800464548157e-07,
"loss": 0.0109,
"step": 453
},
{
"epoch": 2.61,
"learning_rate": 8.780637392592494e-07,
"loss": 0.0244,
"step": 454
},
{
"epoch": 2.61,
"learning_rate": 8.527990607919145e-07,
"loss": 0.0115,
"step": 455
},
{
"epoch": 2.62,
"learning_rate": 8.278869849454718e-07,
"loss": 0.0139,
"step": 456
},
{
"epoch": 2.63,
"learning_rate": 8.033284720205947e-07,
"loss": 0.0254,
"step": 457
},
{
"epoch": 2.63,
"learning_rate": 7.791244686889587e-07,
"loss": 0.0201,
"step": 458
},
{
"epoch": 2.64,
"learning_rate": 7.552759079567418e-07,
"loss": 0.0139,
"step": 459
},
{
"epoch": 2.64,
"learning_rate": 7.317837091286706e-07,
"loss": 0.0268,
"step": 460
},
{
"epoch": 2.65,
"learning_rate": 7.086487777725682e-07,
"loss": 0.0371,
"step": 461
},
{
"epoch": 2.66,
"learning_rate": 6.858720056844614e-07,
"loss": 0.0168,
"step": 462
},
{
"epoch": 2.66,
"learning_rate": 6.634542708541936e-07,
"loss": 0.0146,
"step": 463
},
{
"epoch": 2.67,
"learning_rate": 6.413964374315851e-07,
"loss": 0.0257,
"step": 464
},
{
"epoch": 2.67,
"learning_rate": 6.196993556931242e-07,
"loss": 0.0149,
"step": 465
},
{
"epoch": 2.68,
"learning_rate": 5.983638620091858e-07,
"loss": 0.0067,
"step": 466
},
{
"epoch": 2.68,
"learning_rate": 5.77390778811796e-07,
"loss": 0.0303,
"step": 467
},
{
"epoch": 2.69,
"learning_rate": 5.567809145629244e-07,
"loss": 0.0253,
"step": 468
},
{
"epoch": 2.7,
"learning_rate": 5.365350637233236e-07,
"loss": 0.0198,
"step": 469
},
{
"epoch": 2.7,
"learning_rate": 5.166540067219072e-07,
"loss": 0.0244,
"step": 470
},
{
"epoch": 2.71,
"learning_rate": 4.971385099256609e-07,
"loss": 0.0312,
"step": 471
},
{
"epoch": 2.71,
"learning_rate": 4.779893256100987e-07,
"loss": 0.0287,
"step": 472
},
{
"epoch": 2.72,
"learning_rate": 4.592071919302743e-07,
"loss": 0.0176,
"step": 473
},
{
"epoch": 2.72,
"learning_rate": 4.407928328923194e-07,
"loss": 0.0329,
"step": 474
},
{
"epoch": 2.73,
"learning_rate": 4.227469583255328e-07,
"loss": 0.0308,
"step": 475
},
{
"epoch": 2.74,
"learning_rate": 4.0507026385502747e-07,
"loss": 0.0192,
"step": 476
},
{
"epoch": 2.74,
"learning_rate": 3.877634308749079e-07,
"loss": 0.0165,
"step": 477
},
{
"epoch": 2.75,
"learning_rate": 3.708271265220087e-07,
"loss": 0.0338,
"step": 478
},
{
"epoch": 2.75,
"learning_rate": 3.5426200365017207e-07,
"loss": 0.0318,
"step": 479
},
{
"epoch": 2.76,
"learning_rate": 3.380687008050909e-07,
"loss": 0.0323,
"step": 480
},
{
"epoch": 2.76,
"learning_rate": 3.222478421996877e-07,
"loss": 0.0258,
"step": 481
},
{
"epoch": 2.77,
"learning_rate": 3.0680003769005153e-07,
"loss": 0.0221,
"step": 482
},
{
"epoch": 2.78,
"learning_rate": 2.9172588275193536e-07,
"loss": 0.0214,
"step": 483
},
{
"epoch": 2.78,
"learning_rate": 2.770259584577972e-07,
"loss": 0.0366,
"step": 484
},
{
"epoch": 2.79,
"learning_rate": 2.627008314544044e-07,
"loss": 0.0221,
"step": 485
},
{
"epoch": 2.79,
"learning_rate": 2.4875105394098654e-07,
"loss": 0.0146,
"step": 486
},
{
"epoch": 2.8,
"learning_rate": 2.3517716364795384e-07,
"loss": 0.0264,
"step": 487
},
{
"epoch": 2.8,
"learning_rate": 2.2197968381616807e-07,
"loss": 0.0505,
"step": 488
},
{
"epoch": 2.81,
"learning_rate": 2.091591231767709e-07,
"loss": 0.0171,
"step": 489
},
{
"epoch": 2.82,
"learning_rate": 1.967159759315751e-07,
"loss": 0.0387,
"step": 490
},
{
"epoch": 2.82,
"learning_rate": 1.846507217340121e-07,
"loss": 0.0225,
"step": 491
},
{
"epoch": 2.83,
"learning_rate": 1.7296382567064674e-07,
"loss": 0.0159,
"step": 492
},
{
"epoch": 2.83,
"learning_rate": 1.6165573824324486e-07,
"loss": 0.0341,
"step": 493
},
{
"epoch": 2.84,
"learning_rate": 1.5072689535141072e-07,
"loss": 0.0267,
"step": 494
},
{
"epoch": 2.84,
"learning_rate": 1.401777182757813e-07,
"loss": 0.0194,
"step": 495
},
{
"epoch": 2.85,
"learning_rate": 1.3000861366179064e-07,
"loss": 0.015,
"step": 496
},
{
"epoch": 2.86,
"learning_rate": 1.2021997350399107e-07,
"loss": 0.0113,
"step": 497
},
{
"epoch": 2.86,
"learning_rate": 1.1081217513094211e-07,
"loss": 0.0263,
"step": 498
},
{
"epoch": 2.87,
"learning_rate": 1.0178558119067316e-07,
"loss": 0.0227,
"step": 499
},
{
"epoch": 2.87,
"learning_rate": 9.314053963669245e-08,
"loss": 0.0109,
"step": 500
},
{
"epoch": 2.88,
"learning_rate": 8.487738371458443e-08,
"loss": 0.0164,
"step": 501
},
{
"epoch": 2.89,
"learning_rate": 7.699643194915784e-08,
"loss": 0.023,
"step": 502
},
{
"epoch": 2.89,
"learning_rate": 6.949798813217002e-08,
"loss": 0.013,
"step": 503
},
{
"epoch": 2.9,
"learning_rate": 6.238234131061616e-08,
"loss": 0.0172,
"step": 504
},
{
"epoch": 2.9,
"learning_rate": 5.5649765775586115e-08,
"loss": 0.032,
"step": 505
},
{
"epoch": 2.91,
"learning_rate": 4.930052105169281e-08,
"loss": 0.0223,
"step": 506
},
{
"epoch": 2.91,
"learning_rate": 4.333485188706577e-08,
"loss": 0.0221,
"step": 507
},
{
"epoch": 2.92,
"learning_rate": 3.7752988243919817e-08,
"loss": 0.0179,
"step": 508
},
{
"epoch": 2.93,
"learning_rate": 3.255514528968884e-08,
"loss": 0.0211,
"step": 509
},
{
"epoch": 2.93,
"learning_rate": 2.774152338873126e-08,
"loss": 0.0231,
"step": 510
},
{
"epoch": 2.94,
"learning_rate": 2.3312308094607382e-08,
"loss": 0.0172,
"step": 511
},
{
"epoch": 2.94,
"learning_rate": 1.9267670142926187e-08,
"loss": 0.0188,
"step": 512
},
{
"epoch": 2.95,
"learning_rate": 1.5607765444762834e-08,
"loss": 0.0342,
"step": 513
},
{
"epoch": 2.95,
"learning_rate": 1.2332735080651248e-08,
"loss": 0.0147,
"step": 514
},
{
"epoch": 2.96,
"learning_rate": 9.442705295145127e-09,
"loss": 0.0159,
"step": 515
},
{
"epoch": 2.97,
"learning_rate": 6.937787491947401e-09,
"loss": 0.0444,
"step": 516
},
{
"epoch": 2.97,
"learning_rate": 4.818078229622547e-09,
"loss": 0.0186,
"step": 517
},
{
"epoch": 2.98,
"learning_rate": 3.083659217871793e-09,
"loss": 0.0123,
"step": 518
},
{
"epoch": 2.98,
"learning_rate": 1.7345973143789717e-09,
"loss": 0.0258,
"step": 519
},
{
"epoch": 2.99,
"learning_rate": 7.709445222403578e-10,
"loss": 0.0278,
"step": 520
},
{
"epoch": 2.99,
"learning_rate": 1.9273798795627074e-10,
"loss": 0.0191,
"step": 521
},
{
"epoch": 3.0,
"learning_rate": 0.0,
"loss": 0.0237,
"step": 522
},
{
"epoch": 3.0,
"step": 522,
"total_flos": 6390936502272.0,
"train_loss": 0.050091321954781984,
"train_runtime": 1579.6946,
"train_samples_per_second": 10.55,
"train_steps_per_second": 0.33
}
],
"logging_steps": 1.0,
"max_steps": 522,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 1100,
"total_flos": 6390936502272.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}