Commit
•
6350c45
1
Parent(s):
100fb74
Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +503 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,503 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 191.38755980861245,
|
| 5 |
+
"global_step": 40000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 2.39,
|
| 12 |
+
"learning_rate": 3.125e-06,
|
| 13 |
+
"loss": 0.4016,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 4.78,
|
| 18 |
+
"learning_rate": 6.25e-06,
|
| 19 |
+
"loss": 0.202,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 7.18,
|
| 24 |
+
"learning_rate": 9.375000000000001e-06,
|
| 25 |
+
"loss": 0.1987,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 9.57,
|
| 30 |
+
"learning_rate": 1.25e-05,
|
| 31 |
+
"loss": 0.1964,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 11.96,
|
| 36 |
+
"learning_rate": 1.5625e-05,
|
| 37 |
+
"loss": 0.1579,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 14.35,
|
| 42 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 43 |
+
"loss": 0.1305,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 16.75,
|
| 48 |
+
"learning_rate": 2.1875e-05,
|
| 49 |
+
"loss": 0.1194,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 19.14,
|
| 54 |
+
"learning_rate": 2.5e-05,
|
| 55 |
+
"loss": 0.1094,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 21.53,
|
| 60 |
+
"learning_rate": 2.465277777777778e-05,
|
| 61 |
+
"loss": 0.0994,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 23.92,
|
| 66 |
+
"learning_rate": 2.4305555555555558e-05,
|
| 67 |
+
"loss": 0.0881,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 26.32,
|
| 72 |
+
"learning_rate": 2.3958333333333334e-05,
|
| 73 |
+
"loss": 0.0795,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 28.71,
|
| 78 |
+
"learning_rate": 2.361111111111111e-05,
|
| 79 |
+
"loss": 0.0738,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 31.1,
|
| 84 |
+
"learning_rate": 2.326388888888889e-05,
|
| 85 |
+
"loss": 0.0673,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 33.49,
|
| 90 |
+
"learning_rate": 2.2916666666666667e-05,
|
| 91 |
+
"loss": 0.0621,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 35.89,
|
| 96 |
+
"learning_rate": 2.2569444444444447e-05,
|
| 97 |
+
"loss": 0.058,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 38.28,
|
| 102 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 103 |
+
"loss": 0.0539,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 40.67,
|
| 108 |
+
"learning_rate": 2.1875e-05,
|
| 109 |
+
"loss": 0.0489,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 43.06,
|
| 114 |
+
"learning_rate": 2.152777777777778e-05,
|
| 115 |
+
"loss": 0.0463,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 45.45,
|
| 120 |
+
"learning_rate": 2.1180555555555556e-05,
|
| 121 |
+
"loss": 0.0424,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 47.85,
|
| 126 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 127 |
+
"loss": 0.0402,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 50.24,
|
| 132 |
+
"learning_rate": 2.0486111111111113e-05,
|
| 133 |
+
"loss": 0.0375,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 52.63,
|
| 138 |
+
"learning_rate": 2.013888888888889e-05,
|
| 139 |
+
"loss": 0.0348,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 55.02,
|
| 144 |
+
"learning_rate": 1.9791666666666665e-05,
|
| 145 |
+
"loss": 0.0328,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 57.42,
|
| 150 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 151 |
+
"loss": 0.0306,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 59.81,
|
| 156 |
+
"learning_rate": 1.9097222222222222e-05,
|
| 157 |
+
"loss": 0.0291,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 62.2,
|
| 162 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 163 |
+
"loss": 0.0271,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 64.59,
|
| 168 |
+
"learning_rate": 1.8402777777777778e-05,
|
| 169 |
+
"loss": 0.0261,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 66.99,
|
| 174 |
+
"learning_rate": 1.8055555555555555e-05,
|
| 175 |
+
"loss": 0.0244,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 69.38,
|
| 180 |
+
"learning_rate": 1.7708333333333335e-05,
|
| 181 |
+
"loss": 0.0227,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 71.77,
|
| 186 |
+
"learning_rate": 1.736111111111111e-05,
|
| 187 |
+
"loss": 0.0217,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 74.16,
|
| 192 |
+
"learning_rate": 1.701388888888889e-05,
|
| 193 |
+
"loss": 0.0212,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 76.56,
|
| 198 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 199 |
+
"loss": 0.0199,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 78.95,
|
| 204 |
+
"learning_rate": 1.6319444444444444e-05,
|
| 205 |
+
"loss": 0.019,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 81.34,
|
| 210 |
+
"learning_rate": 1.597222222222222e-05,
|
| 211 |
+
"loss": 0.0183,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 83.73,
|
| 216 |
+
"learning_rate": 1.5625e-05,
|
| 217 |
+
"loss": 0.0169,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 86.12,
|
| 222 |
+
"learning_rate": 1.527777777777778e-05,
|
| 223 |
+
"loss": 0.0166,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 88.52,
|
| 228 |
+
"learning_rate": 1.4930555555555557e-05,
|
| 229 |
+
"loss": 0.016,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 90.91,
|
| 234 |
+
"learning_rate": 1.4583333333333335e-05,
|
| 235 |
+
"loss": 0.015,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 93.3,
|
| 240 |
+
"learning_rate": 1.4236111111111111e-05,
|
| 241 |
+
"loss": 0.0148,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 95.69,
|
| 246 |
+
"learning_rate": 1.388888888888889e-05,
|
| 247 |
+
"loss": 0.0139,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 98.09,
|
| 252 |
+
"learning_rate": 1.3541666666666666e-05,
|
| 253 |
+
"loss": 0.0138,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 100.48,
|
| 258 |
+
"learning_rate": 1.3194444444444446e-05,
|
| 259 |
+
"loss": 0.0137,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 102.87,
|
| 264 |
+
"learning_rate": 1.2847222222222222e-05,
|
| 265 |
+
"loss": 0.0123,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 105.26,
|
| 270 |
+
"learning_rate": 1.25e-05,
|
| 271 |
+
"loss": 0.0122,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 107.66,
|
| 276 |
+
"learning_rate": 1.2152777777777779e-05,
|
| 277 |
+
"loss": 0.0115,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 110.05,
|
| 282 |
+
"learning_rate": 1.1805555555555555e-05,
|
| 283 |
+
"loss": 0.0116,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 112.44,
|
| 288 |
+
"learning_rate": 1.1458333333333333e-05,
|
| 289 |
+
"loss": 0.0109,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 114.83,
|
| 294 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 295 |
+
"loss": 0.0105,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 117.22,
|
| 300 |
+
"learning_rate": 1.076388888888889e-05,
|
| 301 |
+
"loss": 0.0105,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 119.62,
|
| 306 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 307 |
+
"loss": 0.01,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 122.01,
|
| 312 |
+
"learning_rate": 1.0069444444444445e-05,
|
| 313 |
+
"loss": 0.0095,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 124.4,
|
| 318 |
+
"learning_rate": 9.722222222222223e-06,
|
| 319 |
+
"loss": 0.0097,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 126.79,
|
| 324 |
+
"learning_rate": 9.375000000000001e-06,
|
| 325 |
+
"loss": 0.0092,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 129.19,
|
| 330 |
+
"learning_rate": 9.027777777777777e-06,
|
| 331 |
+
"loss": 0.0088,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 131.58,
|
| 336 |
+
"learning_rate": 8.680555555555556e-06,
|
| 337 |
+
"loss": 0.0087,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 133.97,
|
| 342 |
+
"learning_rate": 8.333333333333334e-06,
|
| 343 |
+
"loss": 0.0086,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 136.36,
|
| 348 |
+
"learning_rate": 7.98611111111111e-06,
|
| 349 |
+
"loss": 0.0085,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 138.76,
|
| 354 |
+
"learning_rate": 7.63888888888889e-06,
|
| 355 |
+
"loss": 0.0081,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 141.15,
|
| 360 |
+
"learning_rate": 7.2916666666666674e-06,
|
| 361 |
+
"loss": 0.0081,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 143.54,
|
| 366 |
+
"learning_rate": 6.944444444444445e-06,
|
| 367 |
+
"loss": 0.0077,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 145.93,
|
| 372 |
+
"learning_rate": 6.597222222222223e-06,
|
| 373 |
+
"loss": 0.0075,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 148.33,
|
| 378 |
+
"learning_rate": 6.25e-06,
|
| 379 |
+
"loss": 0.0077,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 150.72,
|
| 384 |
+
"learning_rate": 5.902777777777778e-06,
|
| 385 |
+
"loss": 0.0078,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 153.11,
|
| 390 |
+
"learning_rate": 5.555555555555556e-06,
|
| 391 |
+
"loss": 0.0071,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 155.5,
|
| 396 |
+
"learning_rate": 5.208333333333334e-06,
|
| 397 |
+
"loss": 0.0073,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 157.89,
|
| 402 |
+
"learning_rate": 4.861111111111111e-06,
|
| 403 |
+
"loss": 0.007,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 160.29,
|
| 408 |
+
"learning_rate": 4.513888888888889e-06,
|
| 409 |
+
"loss": 0.0068,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 162.68,
|
| 414 |
+
"learning_rate": 4.166666666666667e-06,
|
| 415 |
+
"loss": 0.0071,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 165.07,
|
| 420 |
+
"learning_rate": 3.819444444444445e-06,
|
| 421 |
+
"loss": 0.0068,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 167.46,
|
| 426 |
+
"learning_rate": 3.4722222222222224e-06,
|
| 427 |
+
"loss": 0.0068,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 169.86,
|
| 432 |
+
"learning_rate": 3.125e-06,
|
| 433 |
+
"loss": 0.0066,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 172.25,
|
| 438 |
+
"learning_rate": 2.777777777777778e-06,
|
| 439 |
+
"loss": 0.0065,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 174.64,
|
| 444 |
+
"learning_rate": 2.4305555555555557e-06,
|
| 445 |
+
"loss": 0.0064,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 177.03,
|
| 450 |
+
"learning_rate": 2.0833333333333334e-06,
|
| 451 |
+
"loss": 0.0066,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 179.43,
|
| 456 |
+
"learning_rate": 1.7361111111111112e-06,
|
| 457 |
+
"loss": 0.0064,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 181.82,
|
| 462 |
+
"learning_rate": 1.388888888888889e-06,
|
| 463 |
+
"loss": 0.0062,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 184.21,
|
| 468 |
+
"learning_rate": 1.0416666666666667e-06,
|
| 469 |
+
"loss": 0.0061,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 186.6,
|
| 474 |
+
"learning_rate": 6.944444444444445e-07,
|
| 475 |
+
"loss": 0.0061,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 189.0,
|
| 480 |
+
"learning_rate": 3.4722222222222224e-07,
|
| 481 |
+
"loss": 0.006,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 191.39,
|
| 486 |
+
"learning_rate": 0.0,
|
| 487 |
+
"loss": 0.0063,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 191.39,
|
| 492 |
+
"step": 40000,
|
| 493 |
+
"total_flos": 2.8134821491393536e+17,
|
| 494 |
+
"train_runtime": 19555.7785,
|
| 495 |
+
"train_samples_per_second": 2.045
|
| 496 |
+
}
|
| 497 |
+
],
|
| 498 |
+
"max_steps": 40000,
|
| 499 |
+
"num_train_epochs": 192,
|
| 500 |
+
"total_flos": 2.8134821491393536e+17,
|
| 501 |
+
"trial_name": null,
|
| 502 |
+
"trial_params": null
|
| 503 |
+
}
|