cs_cot_64 / trainer_state.json
brettbbb's picture
End of training
c538287
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"eval_steps": 500,
"global_step": 320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.06,
"learning_rate": 2e-05,
"loss": 2.6289,
"step": 1
},
{
"epoch": 0.12,
"learning_rate": 4e-05,
"loss": 3.0301,
"step": 2
},
{
"epoch": 0.19,
"learning_rate": 6e-05,
"loss": 2.8284,
"step": 3
},
{
"epoch": 0.25,
"learning_rate": 8e-05,
"loss": 2.4221,
"step": 4
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 2.9842,
"step": 5
},
{
"epoch": 0.38,
"learning_rate": 9.968253968253969e-05,
"loss": 2.6203,
"step": 6
},
{
"epoch": 0.44,
"learning_rate": 9.936507936507938e-05,
"loss": 2.2694,
"step": 7
},
{
"epoch": 0.5,
"learning_rate": 9.904761904761905e-05,
"loss": 2.2085,
"step": 8
},
{
"epoch": 0.56,
"learning_rate": 9.873015873015874e-05,
"loss": 2.0651,
"step": 9
},
{
"epoch": 0.62,
"learning_rate": 9.841269841269841e-05,
"loss": 1.9613,
"step": 10
},
{
"epoch": 0.69,
"learning_rate": 9.80952380952381e-05,
"loss": 2.0691,
"step": 11
},
{
"epoch": 0.75,
"learning_rate": 9.777777777777778e-05,
"loss": 1.8524,
"step": 12
},
{
"epoch": 0.81,
"learning_rate": 9.746031746031747e-05,
"loss": 2.1834,
"step": 13
},
{
"epoch": 0.88,
"learning_rate": 9.714285714285715e-05,
"loss": 2.2191,
"step": 14
},
{
"epoch": 0.94,
"learning_rate": 9.682539682539682e-05,
"loss": 1.8665,
"step": 15
},
{
"epoch": 1.0,
"learning_rate": 9.650793650793651e-05,
"loss": 1.9496,
"step": 16
},
{
"epoch": 1.06,
"learning_rate": 9.61904761904762e-05,
"loss": 1.5671,
"step": 17
},
{
"epoch": 1.12,
"learning_rate": 9.587301587301588e-05,
"loss": 1.7273,
"step": 18
},
{
"epoch": 1.19,
"learning_rate": 9.555555555555557e-05,
"loss": 1.696,
"step": 19
},
{
"epoch": 1.25,
"learning_rate": 9.523809523809524e-05,
"loss": 1.5308,
"step": 20
},
{
"epoch": 1.31,
"learning_rate": 9.492063492063493e-05,
"loss": 1.6791,
"step": 21
},
{
"epoch": 1.38,
"learning_rate": 9.46031746031746e-05,
"loss": 1.36,
"step": 22
},
{
"epoch": 1.44,
"learning_rate": 9.428571428571429e-05,
"loss": 1.525,
"step": 23
},
{
"epoch": 1.5,
"learning_rate": 9.396825396825397e-05,
"loss": 1.4626,
"step": 24
},
{
"epoch": 1.56,
"learning_rate": 9.365079365079366e-05,
"loss": 1.5763,
"step": 25
},
{
"epoch": 1.62,
"learning_rate": 9.333333333333334e-05,
"loss": 1.4599,
"step": 26
},
{
"epoch": 1.69,
"learning_rate": 9.301587301587303e-05,
"loss": 1.5587,
"step": 27
},
{
"epoch": 1.75,
"learning_rate": 9.26984126984127e-05,
"loss": 1.4923,
"step": 28
},
{
"epoch": 1.81,
"learning_rate": 9.238095238095239e-05,
"loss": 1.5451,
"step": 29
},
{
"epoch": 1.88,
"learning_rate": 9.206349206349206e-05,
"loss": 1.3339,
"step": 30
},
{
"epoch": 1.94,
"learning_rate": 9.174603174603175e-05,
"loss": 1.3308,
"step": 31
},
{
"epoch": 2.0,
"learning_rate": 9.142857142857143e-05,
"loss": 1.2522,
"step": 32
},
{
"epoch": 2.06,
"learning_rate": 9.111111111111112e-05,
"loss": 1.2509,
"step": 33
},
{
"epoch": 2.12,
"learning_rate": 9.079365079365079e-05,
"loss": 1.0789,
"step": 34
},
{
"epoch": 2.19,
"learning_rate": 9.047619047619048e-05,
"loss": 1.1804,
"step": 35
},
{
"epoch": 2.25,
"learning_rate": 9.015873015873016e-05,
"loss": 1.5871,
"step": 36
},
{
"epoch": 2.31,
"learning_rate": 8.984126984126985e-05,
"loss": 1.3956,
"step": 37
},
{
"epoch": 2.38,
"learning_rate": 8.952380952380953e-05,
"loss": 0.9778,
"step": 38
},
{
"epoch": 2.44,
"learning_rate": 8.920634920634922e-05,
"loss": 1.1076,
"step": 39
},
{
"epoch": 2.5,
"learning_rate": 8.888888888888889e-05,
"loss": 1.0738,
"step": 40
},
{
"epoch": 2.56,
"learning_rate": 8.857142857142857e-05,
"loss": 1.196,
"step": 41
},
{
"epoch": 2.62,
"learning_rate": 8.825396825396825e-05,
"loss": 0.9954,
"step": 42
},
{
"epoch": 2.69,
"learning_rate": 8.793650793650794e-05,
"loss": 1.3212,
"step": 43
},
{
"epoch": 2.75,
"learning_rate": 8.761904761904762e-05,
"loss": 1.0755,
"step": 44
},
{
"epoch": 2.81,
"learning_rate": 8.730158730158731e-05,
"loss": 0.9754,
"step": 45
},
{
"epoch": 2.88,
"learning_rate": 8.6984126984127e-05,
"loss": 1.1576,
"step": 46
},
{
"epoch": 2.94,
"learning_rate": 8.666666666666667e-05,
"loss": 1.1855,
"step": 47
},
{
"epoch": 3.0,
"learning_rate": 8.634920634920635e-05,
"loss": 1.2223,
"step": 48
},
{
"epoch": 3.06,
"learning_rate": 8.603174603174604e-05,
"loss": 0.8029,
"step": 49
},
{
"epoch": 3.12,
"learning_rate": 8.571428571428571e-05,
"loss": 0.738,
"step": 50
},
{
"epoch": 3.19,
"learning_rate": 8.53968253968254e-05,
"loss": 0.72,
"step": 51
},
{
"epoch": 3.25,
"learning_rate": 8.507936507936508e-05,
"loss": 0.6629,
"step": 52
},
{
"epoch": 3.31,
"learning_rate": 8.476190476190477e-05,
"loss": 0.7514,
"step": 53
},
{
"epoch": 3.38,
"learning_rate": 8.444444444444444e-05,
"loss": 0.6566,
"step": 54
},
{
"epoch": 3.44,
"learning_rate": 8.412698412698413e-05,
"loss": 0.6229,
"step": 55
},
{
"epoch": 3.5,
"learning_rate": 8.380952380952382e-05,
"loss": 0.9274,
"step": 56
},
{
"epoch": 3.56,
"learning_rate": 8.34920634920635e-05,
"loss": 0.6264,
"step": 57
},
{
"epoch": 3.62,
"learning_rate": 8.317460317460319e-05,
"loss": 0.6957,
"step": 58
},
{
"epoch": 3.69,
"learning_rate": 8.285714285714287e-05,
"loss": 0.6733,
"step": 59
},
{
"epoch": 3.75,
"learning_rate": 8.253968253968255e-05,
"loss": 0.6615,
"step": 60
},
{
"epoch": 3.81,
"learning_rate": 8.222222222222222e-05,
"loss": 0.6241,
"step": 61
},
{
"epoch": 3.88,
"learning_rate": 8.19047619047619e-05,
"loss": 0.7035,
"step": 62
},
{
"epoch": 3.94,
"learning_rate": 8.158730158730159e-05,
"loss": 0.6843,
"step": 63
},
{
"epoch": 4.0,
"learning_rate": 8.126984126984128e-05,
"loss": 0.7902,
"step": 64
},
{
"epoch": 4.06,
"learning_rate": 8.095238095238096e-05,
"loss": 0.5331,
"step": 65
},
{
"epoch": 4.12,
"learning_rate": 8.063492063492063e-05,
"loss": 0.3752,
"step": 66
},
{
"epoch": 4.19,
"learning_rate": 8.031746031746032e-05,
"loss": 0.2817,
"step": 67
},
{
"epoch": 4.25,
"learning_rate": 8e-05,
"loss": 0.2742,
"step": 68
},
{
"epoch": 4.31,
"learning_rate": 7.968253968253968e-05,
"loss": 0.2818,
"step": 69
},
{
"epoch": 4.38,
"learning_rate": 7.936507936507937e-05,
"loss": 0.2879,
"step": 70
},
{
"epoch": 4.44,
"learning_rate": 7.904761904761905e-05,
"loss": 0.3607,
"step": 71
},
{
"epoch": 4.5,
"learning_rate": 7.873015873015874e-05,
"loss": 0.2886,
"step": 72
},
{
"epoch": 4.56,
"learning_rate": 7.841269841269841e-05,
"loss": 0.2365,
"step": 73
},
{
"epoch": 4.62,
"learning_rate": 7.80952380952381e-05,
"loss": 0.4007,
"step": 74
},
{
"epoch": 4.69,
"learning_rate": 7.777777777777778e-05,
"loss": 0.3153,
"step": 75
},
{
"epoch": 4.75,
"learning_rate": 7.746031746031747e-05,
"loss": 0.3685,
"step": 76
},
{
"epoch": 4.81,
"learning_rate": 7.714285714285715e-05,
"loss": 0.246,
"step": 77
},
{
"epoch": 4.88,
"learning_rate": 7.682539682539684e-05,
"loss": 0.3306,
"step": 78
},
{
"epoch": 4.94,
"learning_rate": 7.650793650793651e-05,
"loss": 0.2472,
"step": 79
},
{
"epoch": 5.0,
"learning_rate": 7.619047619047618e-05,
"loss": 0.2675,
"step": 80
},
{
"epoch": 5.06,
"learning_rate": 7.587301587301587e-05,
"loss": 0.2271,
"step": 81
},
{
"epoch": 5.12,
"learning_rate": 7.555555555555556e-05,
"loss": 0.2076,
"step": 82
},
{
"epoch": 5.19,
"learning_rate": 7.523809523809524e-05,
"loss": 0.1268,
"step": 83
},
{
"epoch": 5.25,
"learning_rate": 7.492063492063493e-05,
"loss": 0.1842,
"step": 84
},
{
"epoch": 5.31,
"learning_rate": 7.460317460317461e-05,
"loss": 0.1736,
"step": 85
},
{
"epoch": 5.38,
"learning_rate": 7.428571428571429e-05,
"loss": 0.1544,
"step": 86
},
{
"epoch": 5.44,
"learning_rate": 7.396825396825397e-05,
"loss": 0.2019,
"step": 87
},
{
"epoch": 5.5,
"learning_rate": 7.365079365079366e-05,
"loss": 0.2301,
"step": 88
},
{
"epoch": 5.56,
"learning_rate": 7.333333333333333e-05,
"loss": 0.1117,
"step": 89
},
{
"epoch": 5.62,
"learning_rate": 7.301587301587302e-05,
"loss": 0.2011,
"step": 90
},
{
"epoch": 5.69,
"learning_rate": 7.26984126984127e-05,
"loss": 0.146,
"step": 91
},
{
"epoch": 5.75,
"learning_rate": 7.238095238095238e-05,
"loss": 0.2428,
"step": 92
},
{
"epoch": 5.81,
"learning_rate": 7.206349206349206e-05,
"loss": 0.1983,
"step": 93
},
{
"epoch": 5.88,
"learning_rate": 7.174603174603175e-05,
"loss": 0.2065,
"step": 94
},
{
"epoch": 5.94,
"learning_rate": 7.142857142857143e-05,
"loss": 0.1183,
"step": 95
},
{
"epoch": 6.0,
"learning_rate": 7.111111111111112e-05,
"loss": 0.1644,
"step": 96
},
{
"epoch": 6.06,
"learning_rate": 7.07936507936508e-05,
"loss": 0.0957,
"step": 97
},
{
"epoch": 6.12,
"learning_rate": 7.047619047619048e-05,
"loss": 0.0854,
"step": 98
},
{
"epoch": 6.19,
"learning_rate": 7.015873015873015e-05,
"loss": 0.1129,
"step": 99
},
{
"epoch": 6.25,
"learning_rate": 6.984126984126984e-05,
"loss": 0.1483,
"step": 100
},
{
"epoch": 6.31,
"learning_rate": 6.952380952380952e-05,
"loss": 0.1008,
"step": 101
},
{
"epoch": 6.38,
"learning_rate": 6.920634920634921e-05,
"loss": 0.103,
"step": 102
},
{
"epoch": 6.44,
"learning_rate": 6.88888888888889e-05,
"loss": 0.0979,
"step": 103
},
{
"epoch": 6.5,
"learning_rate": 6.857142857142858e-05,
"loss": 0.1576,
"step": 104
},
{
"epoch": 6.56,
"learning_rate": 6.825396825396825e-05,
"loss": 0.106,
"step": 105
},
{
"epoch": 6.62,
"learning_rate": 6.793650793650794e-05,
"loss": 0.0994,
"step": 106
},
{
"epoch": 6.69,
"learning_rate": 6.761904761904763e-05,
"loss": 0.0897,
"step": 107
},
{
"epoch": 6.75,
"learning_rate": 6.730158730158731e-05,
"loss": 0.1564,
"step": 108
},
{
"epoch": 6.81,
"learning_rate": 6.698412698412698e-05,
"loss": 0.1115,
"step": 109
},
{
"epoch": 6.88,
"learning_rate": 6.666666666666667e-05,
"loss": 0.1547,
"step": 110
},
{
"epoch": 6.94,
"learning_rate": 6.634920634920636e-05,
"loss": 0.1216,
"step": 111
},
{
"epoch": 7.0,
"learning_rate": 6.603174603174603e-05,
"loss": 0.0885,
"step": 112
},
{
"epoch": 7.06,
"learning_rate": 6.571428571428571e-05,
"loss": 0.093,
"step": 113
},
{
"epoch": 7.12,
"learning_rate": 6.53968253968254e-05,
"loss": 0.1215,
"step": 114
},
{
"epoch": 7.19,
"learning_rate": 6.507936507936509e-05,
"loss": 0.1106,
"step": 115
},
{
"epoch": 7.25,
"learning_rate": 6.476190476190477e-05,
"loss": 0.064,
"step": 116
},
{
"epoch": 7.31,
"learning_rate": 6.444444444444446e-05,
"loss": 0.1215,
"step": 117
},
{
"epoch": 7.38,
"learning_rate": 6.412698412698413e-05,
"loss": 0.1622,
"step": 118
},
{
"epoch": 7.44,
"learning_rate": 6.38095238095238e-05,
"loss": 0.0723,
"step": 119
},
{
"epoch": 7.5,
"learning_rate": 6.349206349206349e-05,
"loss": 0.0807,
"step": 120
},
{
"epoch": 7.56,
"learning_rate": 6.317460317460318e-05,
"loss": 0.06,
"step": 121
},
{
"epoch": 7.62,
"learning_rate": 6.285714285714286e-05,
"loss": 0.1014,
"step": 122
},
{
"epoch": 7.69,
"learning_rate": 6.253968253968255e-05,
"loss": 0.0692,
"step": 123
},
{
"epoch": 7.75,
"learning_rate": 6.222222222222222e-05,
"loss": 0.0913,
"step": 124
},
{
"epoch": 7.81,
"learning_rate": 6.19047619047619e-05,
"loss": 0.0789,
"step": 125
},
{
"epoch": 7.88,
"learning_rate": 6.158730158730159e-05,
"loss": 0.0921,
"step": 126
},
{
"epoch": 7.94,
"learning_rate": 6.126984126984128e-05,
"loss": 0.0957,
"step": 127
},
{
"epoch": 8.0,
"learning_rate": 6.0952380952380964e-05,
"loss": 0.1072,
"step": 128
},
{
"epoch": 8.06,
"learning_rate": 6.0634920634920636e-05,
"loss": 0.0453,
"step": 129
},
{
"epoch": 8.12,
"learning_rate": 6.0317460317460316e-05,
"loss": 0.0571,
"step": 130
},
{
"epoch": 8.19,
"learning_rate": 6e-05,
"loss": 0.0746,
"step": 131
},
{
"epoch": 8.25,
"learning_rate": 5.968253968253969e-05,
"loss": 0.0604,
"step": 132
},
{
"epoch": 8.31,
"learning_rate": 5.936507936507937e-05,
"loss": 0.1931,
"step": 133
},
{
"epoch": 8.38,
"learning_rate": 5.904761904761905e-05,
"loss": 0.0664,
"step": 134
},
{
"epoch": 8.44,
"learning_rate": 5.873015873015873e-05,
"loss": 0.0724,
"step": 135
},
{
"epoch": 8.5,
"learning_rate": 5.841269841269842e-05,
"loss": 0.067,
"step": 136
},
{
"epoch": 8.56,
"learning_rate": 5.8095238095238104e-05,
"loss": 0.063,
"step": 137
},
{
"epoch": 8.62,
"learning_rate": 5.7777777777777776e-05,
"loss": 0.0738,
"step": 138
},
{
"epoch": 8.69,
"learning_rate": 5.7460317460317456e-05,
"loss": 0.0773,
"step": 139
},
{
"epoch": 8.75,
"learning_rate": 5.714285714285714e-05,
"loss": 0.0729,
"step": 140
},
{
"epoch": 8.81,
"learning_rate": 5.682539682539683e-05,
"loss": 0.0813,
"step": 141
},
{
"epoch": 8.88,
"learning_rate": 5.650793650793651e-05,
"loss": 0.0831,
"step": 142
},
{
"epoch": 8.94,
"learning_rate": 5.619047619047619e-05,
"loss": 0.1075,
"step": 143
},
{
"epoch": 9.0,
"learning_rate": 5.587301587301588e-05,
"loss": 0.1084,
"step": 144
},
{
"epoch": 9.06,
"learning_rate": 5.555555555555556e-05,
"loss": 0.0595,
"step": 145
},
{
"epoch": 9.12,
"learning_rate": 5.5238095238095244e-05,
"loss": 0.0705,
"step": 146
},
{
"epoch": 9.19,
"learning_rate": 5.492063492063493e-05,
"loss": 0.0512,
"step": 147
},
{
"epoch": 9.25,
"learning_rate": 5.46031746031746e-05,
"loss": 0.0758,
"step": 148
},
{
"epoch": 9.31,
"learning_rate": 5.428571428571428e-05,
"loss": 0.0625,
"step": 149
},
{
"epoch": 9.38,
"learning_rate": 5.396825396825397e-05,
"loss": 0.0951,
"step": 150
},
{
"epoch": 9.44,
"learning_rate": 5.3650793650793654e-05,
"loss": 0.0493,
"step": 151
},
{
"epoch": 9.5,
"learning_rate": 5.333333333333333e-05,
"loss": 0.0615,
"step": 152
},
{
"epoch": 9.56,
"learning_rate": 5.301587301587302e-05,
"loss": 0.0602,
"step": 153
},
{
"epoch": 9.62,
"learning_rate": 5.2698412698412705e-05,
"loss": 0.0458,
"step": 154
},
{
"epoch": 9.69,
"learning_rate": 5.2380952380952384e-05,
"loss": 0.1141,
"step": 155
},
{
"epoch": 9.75,
"learning_rate": 5.206349206349207e-05,
"loss": 0.0977,
"step": 156
},
{
"epoch": 9.81,
"learning_rate": 5.1746031746031756e-05,
"loss": 0.071,
"step": 157
},
{
"epoch": 9.88,
"learning_rate": 5.142857142857143e-05,
"loss": 0.0554,
"step": 158
},
{
"epoch": 9.94,
"learning_rate": 5.111111111111111e-05,
"loss": 0.0905,
"step": 159
},
{
"epoch": 10.0,
"learning_rate": 5.0793650793650794e-05,
"loss": 0.0736,
"step": 160
},
{
"epoch": 10.06,
"learning_rate": 5.047619047619048e-05,
"loss": 0.0494,
"step": 161
},
{
"epoch": 10.12,
"learning_rate": 5.015873015873016e-05,
"loss": 0.0439,
"step": 162
},
{
"epoch": 10.19,
"learning_rate": 4.9841269841269845e-05,
"loss": 0.1108,
"step": 163
},
{
"epoch": 10.25,
"learning_rate": 4.9523809523809525e-05,
"loss": 0.0346,
"step": 164
},
{
"epoch": 10.31,
"learning_rate": 4.9206349206349204e-05,
"loss": 0.091,
"step": 165
},
{
"epoch": 10.38,
"learning_rate": 4.888888888888889e-05,
"loss": 0.0572,
"step": 166
},
{
"epoch": 10.44,
"learning_rate": 4.8571428571428576e-05,
"loss": 0.0491,
"step": 167
},
{
"epoch": 10.5,
"learning_rate": 4.8253968253968255e-05,
"loss": 0.072,
"step": 168
},
{
"epoch": 10.56,
"learning_rate": 4.793650793650794e-05,
"loss": 0.1037,
"step": 169
},
{
"epoch": 10.62,
"learning_rate": 4.761904761904762e-05,
"loss": 0.0466,
"step": 170
},
{
"epoch": 10.69,
"learning_rate": 4.73015873015873e-05,
"loss": 0.0797,
"step": 171
},
{
"epoch": 10.75,
"learning_rate": 4.6984126984126986e-05,
"loss": 0.0525,
"step": 172
},
{
"epoch": 10.81,
"learning_rate": 4.666666666666667e-05,
"loss": 0.0559,
"step": 173
},
{
"epoch": 10.88,
"learning_rate": 4.634920634920635e-05,
"loss": 0.0699,
"step": 174
},
{
"epoch": 10.94,
"learning_rate": 4.603174603174603e-05,
"loss": 0.0805,
"step": 175
},
{
"epoch": 11.0,
"learning_rate": 4.5714285714285716e-05,
"loss": 0.0833,
"step": 176
},
{
"epoch": 11.06,
"learning_rate": 4.5396825396825395e-05,
"loss": 0.0391,
"step": 177
},
{
"epoch": 11.12,
"learning_rate": 4.507936507936508e-05,
"loss": 0.0553,
"step": 178
},
{
"epoch": 11.19,
"learning_rate": 4.476190476190477e-05,
"loss": 0.0433,
"step": 179
},
{
"epoch": 11.25,
"learning_rate": 4.4444444444444447e-05,
"loss": 0.0669,
"step": 180
},
{
"epoch": 11.31,
"learning_rate": 4.4126984126984126e-05,
"loss": 0.0707,
"step": 181
},
{
"epoch": 11.38,
"learning_rate": 4.380952380952381e-05,
"loss": 0.0642,
"step": 182
},
{
"epoch": 11.44,
"learning_rate": 4.34920634920635e-05,
"loss": 0.068,
"step": 183
},
{
"epoch": 11.5,
"learning_rate": 4.317460317460318e-05,
"loss": 0.0942,
"step": 184
},
{
"epoch": 11.56,
"learning_rate": 4.2857142857142856e-05,
"loss": 0.0399,
"step": 185
},
{
"epoch": 11.62,
"learning_rate": 4.253968253968254e-05,
"loss": 0.0591,
"step": 186
},
{
"epoch": 11.69,
"learning_rate": 4.222222222222222e-05,
"loss": 0.0739,
"step": 187
},
{
"epoch": 11.75,
"learning_rate": 4.190476190476191e-05,
"loss": 0.0574,
"step": 188
},
{
"epoch": 11.81,
"learning_rate": 4.1587301587301594e-05,
"loss": 0.0591,
"step": 189
},
{
"epoch": 11.88,
"learning_rate": 4.126984126984127e-05,
"loss": 0.0675,
"step": 190
},
{
"epoch": 11.94,
"learning_rate": 4.095238095238095e-05,
"loss": 0.0608,
"step": 191
},
{
"epoch": 12.0,
"learning_rate": 4.063492063492064e-05,
"loss": 0.0754,
"step": 192
},
{
"epoch": 12.06,
"learning_rate": 4.031746031746032e-05,
"loss": 0.0478,
"step": 193
},
{
"epoch": 12.12,
"learning_rate": 4e-05,
"loss": 0.0552,
"step": 194
},
{
"epoch": 12.19,
"learning_rate": 3.968253968253968e-05,
"loss": 0.0509,
"step": 195
},
{
"epoch": 12.25,
"learning_rate": 3.936507936507937e-05,
"loss": 0.061,
"step": 196
},
{
"epoch": 12.31,
"learning_rate": 3.904761904761905e-05,
"loss": 0.0435,
"step": 197
},
{
"epoch": 12.38,
"learning_rate": 3.8730158730158734e-05,
"loss": 0.0449,
"step": 198
},
{
"epoch": 12.44,
"learning_rate": 3.841269841269842e-05,
"loss": 0.0844,
"step": 199
},
{
"epoch": 12.5,
"learning_rate": 3.809523809523809e-05,
"loss": 0.0428,
"step": 200
},
{
"epoch": 12.56,
"learning_rate": 3.777777777777778e-05,
"loss": 0.1022,
"step": 201
},
{
"epoch": 12.62,
"learning_rate": 3.7460317460317464e-05,
"loss": 0.0589,
"step": 202
},
{
"epoch": 12.69,
"learning_rate": 3.7142857142857143e-05,
"loss": 0.0439,
"step": 203
},
{
"epoch": 12.75,
"learning_rate": 3.682539682539683e-05,
"loss": 0.071,
"step": 204
},
{
"epoch": 12.81,
"learning_rate": 3.650793650793651e-05,
"loss": 0.0763,
"step": 205
},
{
"epoch": 12.88,
"learning_rate": 3.619047619047619e-05,
"loss": 0.0908,
"step": 206
},
{
"epoch": 12.94,
"learning_rate": 3.5873015873015874e-05,
"loss": 0.0436,
"step": 207
},
{
"epoch": 13.0,
"learning_rate": 3.555555555555556e-05,
"loss": 0.048,
"step": 208
},
{
"epoch": 13.06,
"learning_rate": 3.523809523809524e-05,
"loss": 0.0563,
"step": 209
},
{
"epoch": 13.12,
"learning_rate": 3.492063492063492e-05,
"loss": 0.068,
"step": 210
},
{
"epoch": 13.19,
"learning_rate": 3.4603174603174604e-05,
"loss": 0.1335,
"step": 211
},
{
"epoch": 13.25,
"learning_rate": 3.428571428571429e-05,
"loss": 0.0595,
"step": 212
},
{
"epoch": 13.31,
"learning_rate": 3.396825396825397e-05,
"loss": 0.0648,
"step": 213
},
{
"epoch": 13.38,
"learning_rate": 3.3650793650793656e-05,
"loss": 0.0348,
"step": 214
},
{
"epoch": 13.44,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.0471,
"step": 215
},
{
"epoch": 13.5,
"learning_rate": 3.3015873015873014e-05,
"loss": 0.0431,
"step": 216
},
{
"epoch": 13.56,
"learning_rate": 3.26984126984127e-05,
"loss": 0.0666,
"step": 217
},
{
"epoch": 13.62,
"learning_rate": 3.2380952380952386e-05,
"loss": 0.0657,
"step": 218
},
{
"epoch": 13.69,
"learning_rate": 3.2063492063492065e-05,
"loss": 0.0435,
"step": 219
},
{
"epoch": 13.75,
"learning_rate": 3.1746031746031745e-05,
"loss": 0.0543,
"step": 220
},
{
"epoch": 13.81,
"learning_rate": 3.142857142857143e-05,
"loss": 0.0422,
"step": 221
},
{
"epoch": 13.88,
"learning_rate": 3.111111111111111e-05,
"loss": 0.0636,
"step": 222
},
{
"epoch": 13.94,
"learning_rate": 3.0793650793650796e-05,
"loss": 0.0465,
"step": 223
},
{
"epoch": 14.0,
"learning_rate": 3.0476190476190482e-05,
"loss": 0.067,
"step": 224
},
{
"epoch": 14.06,
"learning_rate": 3.0158730158730158e-05,
"loss": 0.0518,
"step": 225
},
{
"epoch": 14.12,
"learning_rate": 2.9841269841269844e-05,
"loss": 0.0492,
"step": 226
},
{
"epoch": 14.19,
"learning_rate": 2.9523809523809526e-05,
"loss": 0.0403,
"step": 227
},
{
"epoch": 14.25,
"learning_rate": 2.920634920634921e-05,
"loss": 0.0394,
"step": 228
},
{
"epoch": 14.31,
"learning_rate": 2.8888888888888888e-05,
"loss": 0.065,
"step": 229
},
{
"epoch": 14.38,
"learning_rate": 2.857142857142857e-05,
"loss": 0.0469,
"step": 230
},
{
"epoch": 14.44,
"learning_rate": 2.8253968253968253e-05,
"loss": 0.0788,
"step": 231
},
{
"epoch": 14.5,
"learning_rate": 2.793650793650794e-05,
"loss": 0.0501,
"step": 232
},
{
"epoch": 14.56,
"learning_rate": 2.7619047619047622e-05,
"loss": 0.0512,
"step": 233
},
{
"epoch": 14.62,
"learning_rate": 2.73015873015873e-05,
"loss": 0.054,
"step": 234
},
{
"epoch": 14.69,
"learning_rate": 2.6984126984126984e-05,
"loss": 0.0812,
"step": 235
},
{
"epoch": 14.75,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.0383,
"step": 236
},
{
"epoch": 14.81,
"learning_rate": 2.6349206349206353e-05,
"loss": 0.0538,
"step": 237
},
{
"epoch": 14.88,
"learning_rate": 2.6031746031746035e-05,
"loss": 0.0607,
"step": 238
},
{
"epoch": 14.94,
"learning_rate": 2.5714285714285714e-05,
"loss": 0.0492,
"step": 239
},
{
"epoch": 15.0,
"learning_rate": 2.5396825396825397e-05,
"loss": 0.045,
"step": 240
},
{
"epoch": 15.06,
"learning_rate": 2.507936507936508e-05,
"loss": 0.0322,
"step": 241
},
{
"epoch": 15.12,
"learning_rate": 2.4761904761904762e-05,
"loss": 0.0451,
"step": 242
},
{
"epoch": 15.19,
"learning_rate": 2.4444444444444445e-05,
"loss": 0.0858,
"step": 243
},
{
"epoch": 15.25,
"learning_rate": 2.4126984126984128e-05,
"loss": 0.0502,
"step": 244
},
{
"epoch": 15.31,
"learning_rate": 2.380952380952381e-05,
"loss": 0.0717,
"step": 245
},
{
"epoch": 15.38,
"learning_rate": 2.3492063492063493e-05,
"loss": 0.0399,
"step": 246
},
{
"epoch": 15.44,
"learning_rate": 2.3174603174603175e-05,
"loss": 0.0553,
"step": 247
},
{
"epoch": 15.5,
"learning_rate": 2.2857142857142858e-05,
"loss": 0.0447,
"step": 248
},
{
"epoch": 15.56,
"learning_rate": 2.253968253968254e-05,
"loss": 0.0746,
"step": 249
},
{
"epoch": 15.62,
"learning_rate": 2.2222222222222223e-05,
"loss": 0.0587,
"step": 250
},
{
"epoch": 15.69,
"learning_rate": 2.1904761904761906e-05,
"loss": 0.0453,
"step": 251
},
{
"epoch": 15.75,
"learning_rate": 2.158730158730159e-05,
"loss": 0.0513,
"step": 252
},
{
"epoch": 15.81,
"learning_rate": 2.126984126984127e-05,
"loss": 0.038,
"step": 253
},
{
"epoch": 15.88,
"learning_rate": 2.0952380952380954e-05,
"loss": 0.0437,
"step": 254
},
{
"epoch": 15.94,
"learning_rate": 2.0634920634920636e-05,
"loss": 0.0606,
"step": 255
},
{
"epoch": 16.0,
"learning_rate": 2.031746031746032e-05,
"loss": 0.0587,
"step": 256
},
{
"epoch": 16.06,
"learning_rate": 2e-05,
"loss": 0.043,
"step": 257
},
{
"epoch": 16.12,
"learning_rate": 1.9682539682539684e-05,
"loss": 0.0541,
"step": 258
},
{
"epoch": 16.19,
"learning_rate": 1.9365079365079367e-05,
"loss": 0.0575,
"step": 259
},
{
"epoch": 16.25,
"learning_rate": 1.9047619047619046e-05,
"loss": 0.0369,
"step": 260
},
{
"epoch": 16.31,
"learning_rate": 1.8730158730158732e-05,
"loss": 0.0482,
"step": 261
},
{
"epoch": 16.38,
"learning_rate": 1.8412698412698415e-05,
"loss": 0.0321,
"step": 262
},
{
"epoch": 16.44,
"learning_rate": 1.8095238095238094e-05,
"loss": 0.0524,
"step": 263
},
{
"epoch": 16.5,
"learning_rate": 1.777777777777778e-05,
"loss": 0.1042,
"step": 264
},
{
"epoch": 16.56,
"learning_rate": 1.746031746031746e-05,
"loss": 0.0362,
"step": 265
},
{
"epoch": 16.62,
"learning_rate": 1.7142857142857145e-05,
"loss": 0.0389,
"step": 266
},
{
"epoch": 16.69,
"learning_rate": 1.6825396825396828e-05,
"loss": 0.0727,
"step": 267
},
{
"epoch": 16.75,
"learning_rate": 1.6507936507936507e-05,
"loss": 0.0585,
"step": 268
},
{
"epoch": 16.81,
"learning_rate": 1.6190476190476193e-05,
"loss": 0.0475,
"step": 269
},
{
"epoch": 16.88,
"learning_rate": 1.5873015873015872e-05,
"loss": 0.0688,
"step": 270
},
{
"epoch": 16.94,
"learning_rate": 1.5555555555555555e-05,
"loss": 0.0499,
"step": 271
},
{
"epoch": 17.0,
"learning_rate": 1.5238095238095241e-05,
"loss": 0.0531,
"step": 272
},
{
"epoch": 17.06,
"learning_rate": 1.4920634920634922e-05,
"loss": 0.04,
"step": 273
},
{
"epoch": 17.12,
"learning_rate": 1.4603174603174605e-05,
"loss": 0.0433,
"step": 274
},
{
"epoch": 17.19,
"learning_rate": 1.4285714285714285e-05,
"loss": 0.0427,
"step": 275
},
{
"epoch": 17.25,
"learning_rate": 1.396825396825397e-05,
"loss": 0.0632,
"step": 276
},
{
"epoch": 17.31,
"learning_rate": 1.365079365079365e-05,
"loss": 0.0545,
"step": 277
},
{
"epoch": 17.38,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.0397,
"step": 278
},
{
"epoch": 17.44,
"learning_rate": 1.3015873015873018e-05,
"loss": 0.0357,
"step": 279
},
{
"epoch": 17.5,
"learning_rate": 1.2698412698412699e-05,
"loss": 0.0563,
"step": 280
},
{
"epoch": 17.56,
"learning_rate": 1.2380952380952381e-05,
"loss": 0.0444,
"step": 281
},
{
"epoch": 17.62,
"learning_rate": 1.2063492063492064e-05,
"loss": 0.0464,
"step": 282
},
{
"epoch": 17.69,
"learning_rate": 1.1746031746031746e-05,
"loss": 0.0386,
"step": 283
},
{
"epoch": 17.75,
"learning_rate": 1.1428571428571429e-05,
"loss": 0.0692,
"step": 284
},
{
"epoch": 17.81,
"learning_rate": 1.1111111111111112e-05,
"loss": 0.0563,
"step": 285
},
{
"epoch": 17.88,
"learning_rate": 1.0793650793650794e-05,
"loss": 0.0449,
"step": 286
},
{
"epoch": 17.94,
"learning_rate": 1.0476190476190477e-05,
"loss": 0.0712,
"step": 287
},
{
"epoch": 18.0,
"learning_rate": 1.015873015873016e-05,
"loss": 0.0708,
"step": 288
},
{
"epoch": 18.06,
"learning_rate": 9.841269841269842e-06,
"loss": 0.0485,
"step": 289
},
{
"epoch": 18.12,
"learning_rate": 9.523809523809523e-06,
"loss": 0.0491,
"step": 290
},
{
"epoch": 18.19,
"learning_rate": 9.206349206349207e-06,
"loss": 0.055,
"step": 291
},
{
"epoch": 18.25,
"learning_rate": 8.88888888888889e-06,
"loss": 0.0349,
"step": 292
},
{
"epoch": 18.31,
"learning_rate": 8.571428571428573e-06,
"loss": 0.0594,
"step": 293
},
{
"epoch": 18.38,
"learning_rate": 8.253968253968254e-06,
"loss": 0.0444,
"step": 294
},
{
"epoch": 18.44,
"learning_rate": 7.936507936507936e-06,
"loss": 0.0481,
"step": 295
},
{
"epoch": 18.5,
"learning_rate": 7.6190476190476205e-06,
"loss": 0.0435,
"step": 296
},
{
"epoch": 18.56,
"learning_rate": 7.301587301587302e-06,
"loss": 0.0438,
"step": 297
},
{
"epoch": 18.62,
"learning_rate": 6.984126984126985e-06,
"loss": 0.0369,
"step": 298
},
{
"epoch": 18.69,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0425,
"step": 299
},
{
"epoch": 18.75,
"learning_rate": 6.349206349206349e-06,
"loss": 0.0524,
"step": 300
},
{
"epoch": 18.81,
"learning_rate": 6.031746031746032e-06,
"loss": 0.1133,
"step": 301
},
{
"epoch": 18.88,
"learning_rate": 5.7142857142857145e-06,
"loss": 0.044,
"step": 302
},
{
"epoch": 18.94,
"learning_rate": 5.396825396825397e-06,
"loss": 0.0581,
"step": 303
},
{
"epoch": 19.0,
"learning_rate": 5.07936507936508e-06,
"loss": 0.0465,
"step": 304
},
{
"epoch": 19.06,
"learning_rate": 4.7619047619047615e-06,
"loss": 0.0556,
"step": 305
},
{
"epoch": 19.12,
"learning_rate": 4.444444444444445e-06,
"loss": 0.0374,
"step": 306
},
{
"epoch": 19.19,
"learning_rate": 4.126984126984127e-06,
"loss": 0.0518,
"step": 307
},
{
"epoch": 19.25,
"learning_rate": 3.8095238095238102e-06,
"loss": 0.0399,
"step": 308
},
{
"epoch": 19.31,
"learning_rate": 3.4920634920634924e-06,
"loss": 0.0508,
"step": 309
},
{
"epoch": 19.38,
"learning_rate": 3.1746031746031746e-06,
"loss": 0.0476,
"step": 310
},
{
"epoch": 19.44,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.0493,
"step": 311
},
{
"epoch": 19.5,
"learning_rate": 2.53968253968254e-06,
"loss": 0.0612,
"step": 312
},
{
"epoch": 19.56,
"learning_rate": 2.2222222222222225e-06,
"loss": 0.0774,
"step": 313
},
{
"epoch": 19.62,
"learning_rate": 1.9047619047619051e-06,
"loss": 0.046,
"step": 314
},
{
"epoch": 19.69,
"learning_rate": 1.5873015873015873e-06,
"loss": 0.0504,
"step": 315
},
{
"epoch": 19.75,
"learning_rate": 1.26984126984127e-06,
"loss": 0.0413,
"step": 316
},
{
"epoch": 19.81,
"learning_rate": 9.523809523809526e-07,
"loss": 0.0516,
"step": 317
},
{
"epoch": 19.88,
"learning_rate": 6.34920634920635e-07,
"loss": 0.0475,
"step": 318
},
{
"epoch": 19.94,
"learning_rate": 3.174603174603175e-07,
"loss": 0.0394,
"step": 319
},
{
"epoch": 20.0,
"learning_rate": 0.0,
"loss": 0.0384,
"step": 320
},
{
"epoch": 20.0,
"step": 320,
"total_flos": 1.490701523091456e+16,
"train_loss": 0.3566903839586303,
"train_runtime": 520.0586,
"train_samples_per_second": 2.461,
"train_steps_per_second": 0.615
}
],
"logging_steps": 1,
"max_steps": 320,
"num_train_epochs": 20,
"save_steps": 500,
"total_flos": 1.490701523091456e+16,
"trial_name": null,
"trial_params": null
}