Konrad Wojtasik
commited on
Commit
•
09e9229
1
Parent(s):
eee9989
Update model
Browse files- pytorch_model.bin +1 -1
- trainer_state.json +805 -205
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1100243917
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41e4506ce690c8b65cf380209a694329cf2d1339dc6cfa83d01ee163f28c573c
|
| 3 |
size 1100243917
|
trainer_state.json
CHANGED
|
@@ -2,620 +2,1220 @@
|
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 1.0,
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.01,
|
| 12 |
-
"learning_rate": 9.
|
| 13 |
-
"loss": 7.
|
| 14 |
"step": 100
|
| 15 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
{
|
| 17 |
"epoch": 0.02,
|
| 18 |
"learning_rate": 9.8e-05,
|
| 19 |
-
"loss": 0.
|
| 20 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.03,
|
| 24 |
"learning_rate": 9.7e-05,
|
| 25 |
-
"loss": 0.
|
| 26 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.04,
|
| 30 |
"learning_rate": 9.6e-05,
|
| 31 |
-
"loss": 0.
|
| 32 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.05,
|
| 36 |
"learning_rate": 9.5e-05,
|
| 37 |
-
"loss": 0.
|
| 38 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.06,
|
| 42 |
"learning_rate": 9.4e-05,
|
| 43 |
-
"loss": 0.
|
| 44 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.07,
|
| 48 |
"learning_rate": 9.300000000000001e-05,
|
| 49 |
-
"loss": 0.
|
| 50 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.08,
|
| 54 |
"learning_rate": 9.200000000000001e-05,
|
| 55 |
-
"loss": 0.
|
| 56 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.09,
|
| 60 |
"learning_rate": 9.1e-05,
|
| 61 |
-
"loss": 0.
|
| 62 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.1,
|
| 66 |
"learning_rate": 9e-05,
|
| 67 |
-
"loss": 0.
|
| 68 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.11,
|
| 72 |
"learning_rate": 8.900000000000001e-05,
|
| 73 |
-
"loss": 0.
|
| 74 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.12,
|
| 78 |
"learning_rate": 8.800000000000001e-05,
|
| 79 |
-
"loss": 0.
|
| 80 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.13,
|
| 84 |
"learning_rate": 8.7e-05,
|
| 85 |
-
"loss": 0.
|
| 86 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.14,
|
| 90 |
"learning_rate": 8.6e-05,
|
| 91 |
-
"loss": 0.
|
| 92 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.15,
|
| 96 |
"learning_rate": 8.5e-05,
|
| 97 |
-
"loss": 0.
|
| 98 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.16,
|
| 102 |
"learning_rate": 8.4e-05,
|
| 103 |
-
"loss": 0.
|
| 104 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.17,
|
| 108 |
"learning_rate": 8.3e-05,
|
| 109 |
-
"loss": 0.
|
| 110 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.18,
|
| 114 |
"learning_rate": 8.2e-05,
|
| 115 |
-
"loss": 0.
|
| 116 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.19,
|
| 120 |
"learning_rate": 8.1e-05,
|
| 121 |
-
"loss": 0.
|
| 122 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.2,
|
| 126 |
"learning_rate": 8e-05,
|
| 127 |
-
"loss": 0.
|
| 128 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.21,
|
| 132 |
"learning_rate": 7.900000000000001e-05,
|
| 133 |
-
"loss": 0.
|
| 134 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.22,
|
| 138 |
"learning_rate": 7.800000000000001e-05,
|
| 139 |
-
"loss": 0.
|
| 140 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.23,
|
| 144 |
"learning_rate": 7.7e-05,
|
| 145 |
-
"loss": 0.
|
| 146 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.24,
|
| 150 |
"learning_rate": 7.6e-05,
|
| 151 |
-
"loss": 0.
|
| 152 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.25,
|
| 156 |
"learning_rate": 7.500000000000001e-05,
|
| 157 |
-
"loss": 0.
|
| 158 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.26,
|
| 162 |
"learning_rate": 7.4e-05,
|
| 163 |
-
"loss": 0.
|
| 164 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.27,
|
| 168 |
"learning_rate": 7.3e-05,
|
| 169 |
-
"loss": 0.
|
| 170 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.28,
|
| 174 |
"learning_rate": 7.2e-05,
|
| 175 |
-
"loss": 0.
|
| 176 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.29,
|
| 180 |
"learning_rate": 7.1e-05,
|
| 181 |
-
"loss": 0.
|
| 182 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.3,
|
| 186 |
"learning_rate": 7e-05,
|
| 187 |
-
"loss": 0.
|
| 188 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.31,
|
| 192 |
"learning_rate": 6.9e-05,
|
| 193 |
-
"loss": 0.
|
| 194 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.32,
|
| 198 |
"learning_rate": 6.800000000000001e-05,
|
| 199 |
-
"loss": 0.
|
| 200 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.33,
|
| 204 |
"learning_rate": 6.7e-05,
|
| 205 |
-
"loss": 0.
|
| 206 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.34,
|
| 210 |
"learning_rate": 6.6e-05,
|
| 211 |
-
"loss": 0.
|
| 212 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.35,
|
| 216 |
"learning_rate": 6.500000000000001e-05,
|
| 217 |
-
"loss": 0.
|
| 218 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.36,
|
| 222 |
"learning_rate": 6.400000000000001e-05,
|
| 223 |
-
"loss": 0.
|
| 224 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.37,
|
| 228 |
"learning_rate": 6.3e-05,
|
| 229 |
-
"loss": 0.
|
| 230 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.38,
|
| 234 |
"learning_rate": 6.2e-05,
|
| 235 |
-
"loss": 0.
|
| 236 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.39,
|
| 240 |
"learning_rate": 6.1e-05,
|
| 241 |
-
"loss": 0.
|
| 242 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.4,
|
| 246 |
"learning_rate": 6e-05,
|
| 247 |
-
"loss": 0.
|
| 248 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.41,
|
| 252 |
"learning_rate": 5.9e-05,
|
| 253 |
-
"loss": 0.
|
| 254 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.42,
|
| 258 |
"learning_rate": 5.8e-05,
|
| 259 |
-
"loss": 0.
|
| 260 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.43,
|
| 264 |
"learning_rate": 5.6999999999999996e-05,
|
| 265 |
-
"loss": 0.
|
| 266 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 267 |
},
|
| 268 |
{
|
| 269 |
"epoch": 0.44,
|
| 270 |
"learning_rate": 5.6000000000000006e-05,
|
| 271 |
-
"loss": 0.
|
| 272 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 0.45,
|
| 276 |
"learning_rate": 5.500000000000001e-05,
|
| 277 |
-
"loss": 0.
|
| 278 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 279 |
},
|
| 280 |
{
|
| 281 |
"epoch": 0.46,
|
| 282 |
"learning_rate": 5.4000000000000005e-05,
|
| 283 |
-
"loss": 0.
|
| 284 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.47,
|
| 288 |
"learning_rate": 5.300000000000001e-05,
|
| 289 |
-
"loss": 0.
|
| 290 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.48,
|
| 294 |
"learning_rate": 5.2000000000000004e-05,
|
| 295 |
-
"loss": 0.
|
| 296 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.49,
|
| 300 |
"learning_rate": 5.1000000000000006e-05,
|
| 301 |
-
"loss": 0.
|
| 302 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.5,
|
| 306 |
"learning_rate": 5e-05,
|
| 307 |
-
"loss": 0.
|
| 308 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.51,
|
| 312 |
"learning_rate": 4.9e-05,
|
| 313 |
-
"loss": 0.
|
| 314 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.52,
|
| 318 |
"learning_rate": 4.8e-05,
|
| 319 |
-
"loss": 0.
|
| 320 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.53,
|
| 324 |
"learning_rate": 4.7e-05,
|
| 325 |
-
"loss": 0.
|
| 326 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.54,
|
| 330 |
"learning_rate": 4.600000000000001e-05,
|
| 331 |
-
"loss": 0.
|
| 332 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.55,
|
| 336 |
"learning_rate": 4.5e-05,
|
| 337 |
-
"loss": 0.
|
| 338 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.56,
|
| 342 |
"learning_rate": 4.4000000000000006e-05,
|
| 343 |
-
"loss": 0.
|
| 344 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 0.57,
|
| 348 |
"learning_rate": 4.3e-05,
|
| 349 |
-
"loss": 0.
|
| 350 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 0.58,
|
| 354 |
"learning_rate": 4.2e-05,
|
| 355 |
-
"loss": 0.
|
| 356 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.59,
|
| 360 |
"learning_rate": 4.1e-05,
|
| 361 |
-
"loss": 0.
|
| 362 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.6,
|
| 366 |
"learning_rate": 4e-05,
|
| 367 |
-
"loss": 0.
|
| 368 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.61,
|
| 372 |
"learning_rate": 3.9000000000000006e-05,
|
| 373 |
-
"loss": 0.
|
| 374 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.62,
|
| 378 |
"learning_rate": 3.8e-05,
|
| 379 |
-
"loss": 0.
|
| 380 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.63,
|
| 384 |
"learning_rate": 3.7e-05,
|
| 385 |
-
"loss": 0.
|
| 386 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 0.64,
|
| 390 |
"learning_rate": 3.6e-05,
|
| 391 |
-
"loss": 0.
|
| 392 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 393 |
},
|
| 394 |
{
|
| 395 |
"epoch": 0.65,
|
| 396 |
"learning_rate": 3.5e-05,
|
| 397 |
-
"loss": 0.
|
| 398 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.66,
|
| 402 |
"learning_rate": 3.4000000000000007e-05,
|
| 403 |
-
"loss": 0.
|
| 404 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 405 |
},
|
| 406 |
{
|
| 407 |
"epoch": 0.67,
|
| 408 |
"learning_rate": 3.3e-05,
|
| 409 |
-
"loss": 0.
|
| 410 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.68,
|
| 414 |
"learning_rate": 3.2000000000000005e-05,
|
| 415 |
-
"loss": 0.
|
| 416 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.69,
|
| 420 |
"learning_rate": 3.1e-05,
|
| 421 |
-
"loss": 0.
|
| 422 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.7,
|
| 426 |
"learning_rate": 3e-05,
|
| 427 |
-
"loss": 0.
|
| 428 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.71,
|
| 432 |
"learning_rate": 2.9e-05,
|
| 433 |
-
"loss": 0.
|
| 434 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.72,
|
| 438 |
"learning_rate": 2.8000000000000003e-05,
|
| 439 |
-
"loss": 0.
|
| 440 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.73,
|
| 444 |
"learning_rate": 2.7000000000000002e-05,
|
| 445 |
-
"loss": 0.
|
| 446 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 0.74,
|
| 450 |
"learning_rate": 2.6000000000000002e-05,
|
| 451 |
-
"loss": 0.
|
| 452 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.75,
|
| 456 |
"learning_rate": 2.5e-05,
|
| 457 |
-
"loss": 0.
|
| 458 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.76,
|
| 462 |
"learning_rate": 2.4e-05,
|
| 463 |
-
"loss": 0.
|
| 464 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.77,
|
| 468 |
"learning_rate": 2.3000000000000003e-05,
|
| 469 |
-
"loss": 0.
|
| 470 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.78,
|
| 474 |
"learning_rate": 2.2000000000000003e-05,
|
| 475 |
-
"loss": 0.
|
| 476 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 477 |
},
|
| 478 |
{
|
| 479 |
"epoch": 0.79,
|
| 480 |
"learning_rate": 2.1e-05,
|
| 481 |
-
"loss": 0.
|
| 482 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 0.8,
|
| 486 |
"learning_rate": 2e-05,
|
| 487 |
-
"loss": 0.
|
| 488 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.81,
|
| 492 |
"learning_rate": 1.9e-05,
|
| 493 |
-
"loss": 0.
|
| 494 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.82,
|
| 498 |
"learning_rate": 1.8e-05,
|
| 499 |
-
"loss": 0.
|
| 500 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.83,
|
| 504 |
"learning_rate": 1.7000000000000003e-05,
|
| 505 |
-
"loss": 0.
|
| 506 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.84,
|
| 510 |
"learning_rate": 1.6000000000000003e-05,
|
| 511 |
-
"loss": 0.
|
| 512 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 0.85,
|
| 516 |
"learning_rate": 1.5e-05,
|
| 517 |
-
"loss": 0.
|
| 518 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 0.86,
|
| 522 |
"learning_rate": 1.4000000000000001e-05,
|
| 523 |
-
"loss": 0.
|
| 524 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 525 |
},
|
| 526 |
{
|
| 527 |
"epoch": 0.87,
|
| 528 |
"learning_rate": 1.3000000000000001e-05,
|
| 529 |
-
"loss": 0.
|
| 530 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.88,
|
| 534 |
"learning_rate": 1.2e-05,
|
| 535 |
-
"loss": 0.
|
| 536 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.89,
|
| 540 |
"learning_rate": 1.1000000000000001e-05,
|
| 541 |
-
"loss": 0.
|
| 542 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.9,
|
| 546 |
"learning_rate": 1e-05,
|
| 547 |
-
"loss": 0.
|
| 548 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.91,
|
| 552 |
"learning_rate": 9e-06,
|
| 553 |
-
"loss": 0.
|
| 554 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 0.92,
|
| 558 |
"learning_rate": 8.000000000000001e-06,
|
| 559 |
-
"loss": 0.
|
| 560 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 561 |
},
|
| 562 |
{
|
| 563 |
"epoch": 0.93,
|
| 564 |
"learning_rate": 7.000000000000001e-06,
|
| 565 |
-
"loss": 0.
|
| 566 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 567 |
},
|
| 568 |
{
|
| 569 |
"epoch": 0.94,
|
| 570 |
"learning_rate": 6e-06,
|
| 571 |
-
"loss": 0.
|
| 572 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.95,
|
| 576 |
"learning_rate": 5e-06,
|
| 577 |
-
"loss": 0.
|
| 578 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 0.96,
|
| 582 |
"learning_rate": 4.000000000000001e-06,
|
| 583 |
-
"loss": 0.
|
| 584 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.97,
|
| 588 |
"learning_rate": 3e-06,
|
| 589 |
-
"loss": 0.
|
| 590 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.98,
|
| 594 |
"learning_rate": 2.0000000000000003e-06,
|
| 595 |
-
"loss": 0.
|
| 596 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 597 |
},
|
| 598 |
{
|
| 599 |
"epoch": 0.99,
|
| 600 |
"learning_rate": 1.0000000000000002e-06,
|
| 601 |
-
"loss": 0.
|
| 602 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 603 |
},
|
| 604 |
{
|
| 605 |
"epoch": 1.0,
|
| 606 |
"learning_rate": 0.0,
|
| 607 |
-
"loss": 0.
|
| 608 |
-
"step":
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 1.0,
|
| 612 |
-
"step":
|
| 613 |
"total_flos": 0,
|
| 614 |
-
"train_runtime":
|
| 615 |
-
"train_samples_per_second": 0.
|
| 616 |
}
|
| 617 |
],
|
| 618 |
-
"max_steps":
|
| 619 |
"num_train_epochs": 1,
|
| 620 |
"total_flos": 0,
|
| 621 |
"trial_name": null,
|
|
|
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 1.0,
|
| 5 |
+
"global_step": 20000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 9.95e-05,
|
| 13 |
+
"loss": 7.5958,
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 9.900000000000001e-05,
|
| 19 |
+
"loss": 0.3186,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 9.850000000000001e-05,
|
| 25 |
+
"loss": 0.2868,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
{
|
| 29 |
"epoch": 0.02,
|
| 30 |
"learning_rate": 9.8e-05,
|
| 31 |
+
"loss": 0.2434,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.03,
|
| 36 |
+
"learning_rate": 9.75e-05,
|
| 37 |
+
"loss": 0.1965,
|
| 38 |
+
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.03,
|
| 42 |
"learning_rate": 9.7e-05,
|
| 43 |
+
"loss": 0.1758,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04,
|
| 48 |
+
"learning_rate": 9.65e-05,
|
| 49 |
+
"loss": 0.1722,
|
| 50 |
+
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.04,
|
| 54 |
"learning_rate": 9.6e-05,
|
| 55 |
+
"loss": 0.1605,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.04,
|
| 60 |
+
"learning_rate": 9.55e-05,
|
| 61 |
+
"loss": 0.1588,
|
| 62 |
+
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.05,
|
| 66 |
"learning_rate": 9.5e-05,
|
| 67 |
+
"loss": 0.1493,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.06,
|
| 72 |
+
"learning_rate": 9.449999999999999e-05,
|
| 73 |
+
"loss": 0.1429,
|
| 74 |
+
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.06,
|
| 78 |
"learning_rate": 9.4e-05,
|
| 79 |
+
"loss": 0.1375,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.07,
|
| 84 |
+
"learning_rate": 9.350000000000001e-05,
|
| 85 |
+
"loss": 0.1378,
|
| 86 |
+
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.07,
|
| 90 |
"learning_rate": 9.300000000000001e-05,
|
| 91 |
+
"loss": 0.1372,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.07,
|
| 96 |
+
"learning_rate": 9.250000000000001e-05,
|
| 97 |
+
"loss": 0.1291,
|
| 98 |
+
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.08,
|
| 102 |
"learning_rate": 9.200000000000001e-05,
|
| 103 |
+
"loss": 0.129,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.09,
|
| 108 |
+
"learning_rate": 9.15e-05,
|
| 109 |
+
"loss": 0.1302,
|
| 110 |
+
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.09,
|
| 114 |
"learning_rate": 9.1e-05,
|
| 115 |
+
"loss": 0.1255,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.1,
|
| 120 |
+
"learning_rate": 9.05e-05,
|
| 121 |
+
"loss": 0.1256,
|
| 122 |
+
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.1,
|
| 126 |
"learning_rate": 9e-05,
|
| 127 |
+
"loss": 0.1214,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1,
|
| 132 |
+
"learning_rate": 8.950000000000001e-05,
|
| 133 |
+
"loss": 0.1199,
|
| 134 |
+
"step": 2100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.11,
|
| 138 |
"learning_rate": 8.900000000000001e-05,
|
| 139 |
+
"loss": 0.1193,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.12,
|
| 144 |
+
"learning_rate": 8.850000000000001e-05,
|
| 145 |
+
"loss": 0.1208,
|
| 146 |
+
"step": 2300
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.12,
|
| 150 |
"learning_rate": 8.800000000000001e-05,
|
| 151 |
+
"loss": 0.118,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.12,
|
| 156 |
+
"learning_rate": 8.75e-05,
|
| 157 |
+
"loss": 0.1167,
|
| 158 |
+
"step": 2500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.13,
|
| 162 |
"learning_rate": 8.7e-05,
|
| 163 |
+
"loss": 0.1136,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.14,
|
| 168 |
+
"learning_rate": 8.65e-05,
|
| 169 |
+
"loss": 0.1164,
|
| 170 |
+
"step": 2700
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.14,
|
| 174 |
"learning_rate": 8.6e-05,
|
| 175 |
+
"loss": 0.1124,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.14,
|
| 180 |
+
"learning_rate": 8.55e-05,
|
| 181 |
+
"loss": 0.1138,
|
| 182 |
+
"step": 2900
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.15,
|
| 186 |
"learning_rate": 8.5e-05,
|
| 187 |
+
"loss": 0.114,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.15,
|
| 192 |
+
"learning_rate": 8.450000000000001e-05,
|
| 193 |
+
"loss": 0.1137,
|
| 194 |
+
"step": 3100
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.16,
|
| 198 |
"learning_rate": 8.4e-05,
|
| 199 |
+
"loss": 0.1096,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.17,
|
| 204 |
+
"learning_rate": 8.35e-05,
|
| 205 |
+
"loss": 0.1087,
|
| 206 |
+
"step": 3300
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.17,
|
| 210 |
"learning_rate": 8.3e-05,
|
| 211 |
+
"loss": 0.1079,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.17,
|
| 216 |
+
"learning_rate": 8.25e-05,
|
| 217 |
+
"loss": 0.1085,
|
| 218 |
+
"step": 3500
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.18,
|
| 222 |
"learning_rate": 8.2e-05,
|
| 223 |
+
"loss": 0.112,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.18,
|
| 228 |
+
"learning_rate": 8.15e-05,
|
| 229 |
+
"loss": 0.1071,
|
| 230 |
+
"step": 3700
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.19,
|
| 234 |
"learning_rate": 8.1e-05,
|
| 235 |
+
"loss": 0.1077,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.2,
|
| 240 |
+
"learning_rate": 8.05e-05,
|
| 241 |
+
"loss": 0.1059,
|
| 242 |
+
"step": 3900
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.2,
|
| 246 |
"learning_rate": 8e-05,
|
| 247 |
+
"loss": 0.1105,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2,
|
| 252 |
+
"learning_rate": 7.950000000000001e-05,
|
| 253 |
+
"loss": 0.107,
|
| 254 |
+
"step": 4100
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.21,
|
| 258 |
"learning_rate": 7.900000000000001e-05,
|
| 259 |
+
"loss": 0.1064,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.21,
|
| 264 |
+
"learning_rate": 7.850000000000001e-05,
|
| 265 |
+
"loss": 0.1049,
|
| 266 |
+
"step": 4300
|
| 267 |
},
|
| 268 |
{
|
| 269 |
"epoch": 0.22,
|
| 270 |
"learning_rate": 7.800000000000001e-05,
|
| 271 |
+
"loss": 0.1043,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.23,
|
| 276 |
+
"learning_rate": 7.75e-05,
|
| 277 |
+
"loss": 0.1033,
|
| 278 |
+
"step": 4500
|
| 279 |
},
|
| 280 |
{
|
| 281 |
"epoch": 0.23,
|
| 282 |
"learning_rate": 7.7e-05,
|
| 283 |
+
"loss": 0.103,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.23,
|
| 288 |
+
"learning_rate": 7.65e-05,
|
| 289 |
+
"loss": 0.1033,
|
| 290 |
+
"step": 4700
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.24,
|
| 294 |
"learning_rate": 7.6e-05,
|
| 295 |
+
"loss": 0.1014,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.24,
|
| 300 |
+
"learning_rate": 7.55e-05,
|
| 301 |
+
"loss": 0.1038,
|
| 302 |
+
"step": 4900
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.25,
|
| 306 |
"learning_rate": 7.500000000000001e-05,
|
| 307 |
+
"loss": 0.1011,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.26,
|
| 312 |
+
"learning_rate": 7.450000000000001e-05,
|
| 313 |
+
"loss": 0.1011,
|
| 314 |
+
"step": 5100
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.26,
|
| 318 |
"learning_rate": 7.4e-05,
|
| 319 |
+
"loss": 0.1006,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.27,
|
| 324 |
+
"learning_rate": 7.35e-05,
|
| 325 |
+
"loss": 0.1024,
|
| 326 |
+
"step": 5300
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.27,
|
| 330 |
"learning_rate": 7.3e-05,
|
| 331 |
+
"loss": 0.1003,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.28,
|
| 336 |
+
"learning_rate": 7.25e-05,
|
| 337 |
+
"loss": 0.0997,
|
| 338 |
+
"step": 5500
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.28,
|
| 342 |
"learning_rate": 7.2e-05,
|
| 343 |
+
"loss": 0.1019,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.28,
|
| 348 |
+
"learning_rate": 7.15e-05,
|
| 349 |
+
"loss": 0.1025,
|
| 350 |
+
"step": 5700
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 0.29,
|
| 354 |
"learning_rate": 7.1e-05,
|
| 355 |
+
"loss": 0.0994,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.29,
|
| 360 |
+
"learning_rate": 7.05e-05,
|
| 361 |
+
"loss": 0.0982,
|
| 362 |
+
"step": 5900
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.3,
|
| 366 |
"learning_rate": 7e-05,
|
| 367 |
+
"loss": 0.0998,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.3,
|
| 372 |
+
"learning_rate": 6.95e-05,
|
| 373 |
+
"loss": 0.0966,
|
| 374 |
+
"step": 6100
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.31,
|
| 378 |
"learning_rate": 6.9e-05,
|
| 379 |
+
"loss": 0.096,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.32,
|
| 384 |
+
"learning_rate": 6.850000000000001e-05,
|
| 385 |
+
"loss": 0.0972,
|
| 386 |
+
"step": 6300
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 0.32,
|
| 390 |
"learning_rate": 6.800000000000001e-05,
|
| 391 |
+
"loss": 0.0973,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.33,
|
| 396 |
+
"learning_rate": 6.750000000000001e-05,
|
| 397 |
+
"loss": 0.0996,
|
| 398 |
+
"step": 6500
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.33,
|
| 402 |
"learning_rate": 6.7e-05,
|
| 403 |
+
"loss": 0.0961,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.34,
|
| 408 |
+
"learning_rate": 6.65e-05,
|
| 409 |
+
"loss": 0.0964,
|
| 410 |
+
"step": 6700
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.34,
|
| 414 |
"learning_rate": 6.6e-05,
|
| 415 |
+
"loss": 0.0966,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.34,
|
| 420 |
+
"learning_rate": 6.55e-05,
|
| 421 |
+
"loss": 0.0959,
|
| 422 |
+
"step": 6900
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.35,
|
| 426 |
"learning_rate": 6.500000000000001e-05,
|
| 427 |
+
"loss": 0.0975,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.35,
|
| 432 |
+
"learning_rate": 6.450000000000001e-05,
|
| 433 |
+
"loss": 0.0983,
|
| 434 |
+
"step": 7100
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.36,
|
| 438 |
"learning_rate": 6.400000000000001e-05,
|
| 439 |
+
"loss": 0.0989,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.36,
|
| 444 |
+
"learning_rate": 6.35e-05,
|
| 445 |
+
"loss": 0.095,
|
| 446 |
+
"step": 7300
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 0.37,
|
| 450 |
"learning_rate": 6.3e-05,
|
| 451 |
+
"loss": 0.0943,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.38,
|
| 456 |
+
"learning_rate": 6.25e-05,
|
| 457 |
+
"loss": 0.097,
|
| 458 |
+
"step": 7500
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.38,
|
| 462 |
"learning_rate": 6.2e-05,
|
| 463 |
+
"loss": 0.094,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.39,
|
| 468 |
+
"learning_rate": 6.15e-05,
|
| 469 |
+
"loss": 0.0942,
|
| 470 |
+
"step": 7700
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.39,
|
| 474 |
"learning_rate": 6.1e-05,
|
| 475 |
+
"loss": 0.0966,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.4,
|
| 480 |
+
"learning_rate": 6.05e-05,
|
| 481 |
+
"loss": 0.0934,
|
| 482 |
+
"step": 7900
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 0.4,
|
| 486 |
"learning_rate": 6e-05,
|
| 487 |
+
"loss": 0.0941,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.41,
|
| 492 |
+
"learning_rate": 5.95e-05,
|
| 493 |
+
"loss": 0.0947,
|
| 494 |
+
"step": 8100
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.41,
|
| 498 |
"learning_rate": 5.9e-05,
|
| 499 |
+
"loss": 0.0929,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.41,
|
| 504 |
+
"learning_rate": 5.85e-05,
|
| 505 |
+
"loss": 0.0911,
|
| 506 |
+
"step": 8300
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.42,
|
| 510 |
"learning_rate": 5.8e-05,
|
| 511 |
+
"loss": 0.0962,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.42,
|
| 516 |
+
"learning_rate": 5.7499999999999995e-05,
|
| 517 |
+
"loss": 0.0953,
|
| 518 |
+
"step": 8500
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 0.43,
|
| 522 |
"learning_rate": 5.6999999999999996e-05,
|
| 523 |
+
"loss": 0.0926,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.43,
|
| 528 |
+
"learning_rate": 5.65e-05,
|
| 529 |
+
"loss": 0.0957,
|
| 530 |
+
"step": 8700
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.44,
|
| 534 |
"learning_rate": 5.6000000000000006e-05,
|
| 535 |
+
"loss": 0.0935,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.45,
|
| 540 |
+
"learning_rate": 5.550000000000001e-05,
|
| 541 |
+
"loss": 0.0931,
|
| 542 |
+
"step": 8900
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.45,
|
| 546 |
"learning_rate": 5.500000000000001e-05,
|
| 547 |
+
"loss": 0.0937,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.46,
|
| 552 |
+
"learning_rate": 5.45e-05,
|
| 553 |
+
"loss": 0.0937,
|
| 554 |
+
"step": 9100
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 0.46,
|
| 558 |
"learning_rate": 5.4000000000000005e-05,
|
| 559 |
+
"loss": 0.0929,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.47,
|
| 564 |
+
"learning_rate": 5.3500000000000006e-05,
|
| 565 |
+
"loss": 0.0913,
|
| 566 |
+
"step": 9300
|
| 567 |
},
|
| 568 |
{
|
| 569 |
"epoch": 0.47,
|
| 570 |
"learning_rate": 5.300000000000001e-05,
|
| 571 |
+
"loss": 0.0929,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.47,
|
| 576 |
+
"learning_rate": 5.25e-05,
|
| 577 |
+
"loss": 0.0896,
|
| 578 |
+
"step": 9500
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 0.48,
|
| 582 |
"learning_rate": 5.2000000000000004e-05,
|
| 583 |
+
"loss": 0.0923,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.48,
|
| 588 |
+
"learning_rate": 5.1500000000000005e-05,
|
| 589 |
+
"loss": 0.0909,
|
| 590 |
+
"step": 9700
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.49,
|
| 594 |
"learning_rate": 5.1000000000000006e-05,
|
| 595 |
+
"loss": 0.0925,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.49,
|
| 600 |
+
"learning_rate": 5.05e-05,
|
| 601 |
+
"loss": 0.0934,
|
| 602 |
+
"step": 9900
|
| 603 |
},
|
| 604 |
{
|
| 605 |
"epoch": 0.5,
|
| 606 |
"learning_rate": 5e-05,
|
| 607 |
+
"loss": 0.0915,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.51,
|
| 612 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 613 |
+
"loss": 0.0891,
|
| 614 |
+
"step": 10100
|
| 615 |
},
|
| 616 |
{
|
| 617 |
"epoch": 0.51,
|
| 618 |
"learning_rate": 4.9e-05,
|
| 619 |
+
"loss": 0.0898,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.52,
|
| 624 |
+
"learning_rate": 4.85e-05,
|
| 625 |
+
"loss": 0.0931,
|
| 626 |
+
"step": 10300
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.52,
|
| 630 |
"learning_rate": 4.8e-05,
|
| 631 |
+
"loss": 0.0905,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.53,
|
| 636 |
+
"learning_rate": 4.75e-05,
|
| 637 |
+
"loss": 0.0886,
|
| 638 |
+
"step": 10500
|
| 639 |
},
|
| 640 |
{
|
| 641 |
"epoch": 0.53,
|
| 642 |
"learning_rate": 4.7e-05,
|
| 643 |
+
"loss": 0.0872,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.54,
|
| 648 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 649 |
+
"loss": 0.09,
|
| 650 |
+
"step": 10700
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.54,
|
| 654 |
"learning_rate": 4.600000000000001e-05,
|
| 655 |
+
"loss": 0.0894,
|
| 656 |
+
"step": 10800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.55,
|
| 660 |
+
"learning_rate": 4.55e-05,
|
| 661 |
+
"loss": 0.0878,
|
| 662 |
+
"step": 10900
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 0.55,
|
| 666 |
"learning_rate": 4.5e-05,
|
| 667 |
+
"loss": 0.0903,
|
| 668 |
+
"step": 11000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.56,
|
| 672 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 673 |
+
"loss": 0.0904,
|
| 674 |
+
"step": 11100
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.56,
|
| 678 |
"learning_rate": 4.4000000000000006e-05,
|
| 679 |
+
"loss": 0.0905,
|
| 680 |
+
"step": 11200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.56,
|
| 684 |
+
"learning_rate": 4.35e-05,
|
| 685 |
+
"loss": 0.0877,
|
| 686 |
+
"step": 11300
|
| 687 |
},
|
| 688 |
{
|
| 689 |
"epoch": 0.57,
|
| 690 |
"learning_rate": 4.3e-05,
|
| 691 |
+
"loss": 0.0888,
|
| 692 |
+
"step": 11400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.57,
|
| 696 |
+
"learning_rate": 4.25e-05,
|
| 697 |
+
"loss": 0.0904,
|
| 698 |
+
"step": 11500
|
| 699 |
},
|
| 700 |
{
|
| 701 |
"epoch": 0.58,
|
| 702 |
"learning_rate": 4.2e-05,
|
| 703 |
+
"loss": 0.0903,
|
| 704 |
+
"step": 11600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.58,
|
| 708 |
+
"learning_rate": 4.15e-05,
|
| 709 |
+
"loss": 0.0865,
|
| 710 |
+
"step": 11700
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.59,
|
| 714 |
"learning_rate": 4.1e-05,
|
| 715 |
+
"loss": 0.0895,
|
| 716 |
+
"step": 11800
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.59,
|
| 720 |
+
"learning_rate": 4.05e-05,
|
| 721 |
+
"loss": 0.0888,
|
| 722 |
+
"step": 11900
|
| 723 |
},
|
| 724 |
{
|
| 725 |
"epoch": 0.6,
|
| 726 |
"learning_rate": 4e-05,
|
| 727 |
+
"loss": 0.0884,
|
| 728 |
+
"step": 12000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.6,
|
| 732 |
+
"learning_rate": 3.9500000000000005e-05,
|
| 733 |
+
"loss": 0.0897,
|
| 734 |
+
"step": 12100
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.61,
|
| 738 |
"learning_rate": 3.9000000000000006e-05,
|
| 739 |
+
"loss": 0.0891,
|
| 740 |
+
"step": 12200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.61,
|
| 744 |
+
"learning_rate": 3.85e-05,
|
| 745 |
+
"loss": 0.0904,
|
| 746 |
+
"step": 12300
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 0.62,
|
| 750 |
"learning_rate": 3.8e-05,
|
| 751 |
+
"loss": 0.0864,
|
| 752 |
+
"step": 12400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.62,
|
| 756 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 757 |
+
"loss": 0.0871,
|
| 758 |
+
"step": 12500
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.63,
|
| 762 |
"learning_rate": 3.7e-05,
|
| 763 |
+
"loss": 0.0859,
|
| 764 |
+
"step": 12600
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.64,
|
| 768 |
+
"learning_rate": 3.65e-05,
|
| 769 |
+
"loss": 0.0878,
|
| 770 |
+
"step": 12700
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.64,
|
| 774 |
"learning_rate": 3.6e-05,
|
| 775 |
+
"loss": 0.0862,
|
| 776 |
+
"step": 12800
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.65,
|
| 780 |
+
"learning_rate": 3.55e-05,
|
| 781 |
+
"loss": 0.085,
|
| 782 |
+
"step": 12900
|
| 783 |
},
|
| 784 |
{
|
| 785 |
"epoch": 0.65,
|
| 786 |
"learning_rate": 3.5e-05,
|
| 787 |
+
"loss": 0.0885,
|
| 788 |
+
"step": 13000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.66,
|
| 792 |
+
"learning_rate": 3.45e-05,
|
| 793 |
+
"loss": 0.0889,
|
| 794 |
+
"step": 13100
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.66,
|
| 798 |
"learning_rate": 3.4000000000000007e-05,
|
| 799 |
+
"loss": 0.0876,
|
| 800 |
+
"step": 13200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.67,
|
| 804 |
+
"learning_rate": 3.35e-05,
|
| 805 |
+
"loss": 0.0882,
|
| 806 |
+
"step": 13300
|
| 807 |
},
|
| 808 |
{
|
| 809 |
"epoch": 0.67,
|
| 810 |
"learning_rate": 3.3e-05,
|
| 811 |
+
"loss": 0.0884,
|
| 812 |
+
"step": 13400
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.68,
|
| 816 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 817 |
+
"loss": 0.0848,
|
| 818 |
+
"step": 13500
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 0.68,
|
| 822 |
"learning_rate": 3.2000000000000005e-05,
|
| 823 |
+
"loss": 0.0859,
|
| 824 |
+
"step": 13600
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.69,
|
| 828 |
+
"learning_rate": 3.15e-05,
|
| 829 |
+
"loss": 0.0895,
|
| 830 |
+
"step": 13700
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 0.69,
|
| 834 |
"learning_rate": 3.1e-05,
|
| 835 |
+
"loss": 0.0898,
|
| 836 |
+
"step": 13800
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.69,
|
| 840 |
+
"learning_rate": 3.05e-05,
|
| 841 |
+
"loss": 0.0849,
|
| 842 |
+
"step": 13900
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.7,
|
| 846 |
"learning_rate": 3e-05,
|
| 847 |
+
"loss": 0.0866,
|
| 848 |
+
"step": 14000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.7,
|
| 852 |
+
"learning_rate": 2.95e-05,
|
| 853 |
+
"loss": 0.0839,
|
| 854 |
+
"step": 14100
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.71,
|
| 858 |
"learning_rate": 2.9e-05,
|
| 859 |
+
"loss": 0.0855,
|
| 860 |
+
"step": 14200
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.71,
|
| 864 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 865 |
+
"loss": 0.0848,
|
| 866 |
+
"step": 14300
|
| 867 |
},
|
| 868 |
{
|
| 869 |
"epoch": 0.72,
|
| 870 |
"learning_rate": 2.8000000000000003e-05,
|
| 871 |
+
"loss": 0.0889,
|
| 872 |
+
"step": 14400
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.72,
|
| 876 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 877 |
+
"loss": 0.0861,
|
| 878 |
+
"step": 14500
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.73,
|
| 882 |
"learning_rate": 2.7000000000000002e-05,
|
| 883 |
+
"loss": 0.0838,
|
| 884 |
+
"step": 14600
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.73,
|
| 888 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 889 |
+
"loss": 0.0878,
|
| 890 |
+
"step": 14700
|
| 891 |
},
|
| 892 |
{
|
| 893 |
"epoch": 0.74,
|
| 894 |
"learning_rate": 2.6000000000000002e-05,
|
| 895 |
+
"loss": 0.0866,
|
| 896 |
+
"step": 14800
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.74,
|
| 900 |
+
"learning_rate": 2.5500000000000003e-05,
|
| 901 |
+
"loss": 0.086,
|
| 902 |
+
"step": 14900
|
| 903 |
},
|
| 904 |
{
|
| 905 |
"epoch": 0.75,
|
| 906 |
"learning_rate": 2.5e-05,
|
| 907 |
+
"loss": 0.0841,
|
| 908 |
+
"step": 15000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.76,
|
| 912 |
+
"learning_rate": 2.45e-05,
|
| 913 |
+
"loss": 0.0864,
|
| 914 |
+
"step": 15100
|
| 915 |
},
|
| 916 |
{
|
| 917 |
"epoch": 0.76,
|
| 918 |
"learning_rate": 2.4e-05,
|
| 919 |
+
"loss": 0.0855,
|
| 920 |
+
"step": 15200
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.77,
|
| 924 |
+
"learning_rate": 2.35e-05,
|
| 925 |
+
"loss": 0.0861,
|
| 926 |
+
"step": 15300
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.77,
|
| 930 |
"learning_rate": 2.3000000000000003e-05,
|
| 931 |
+
"loss": 0.0856,
|
| 932 |
+
"step": 15400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.78,
|
| 936 |
+
"learning_rate": 2.25e-05,
|
| 937 |
+
"loss": 0.0872,
|
| 938 |
+
"step": 15500
|
| 939 |
},
|
| 940 |
{
|
| 941 |
"epoch": 0.78,
|
| 942 |
"learning_rate": 2.2000000000000003e-05,
|
| 943 |
+
"loss": 0.0838,
|
| 944 |
+
"step": 15600
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.79,
|
| 948 |
+
"learning_rate": 2.15e-05,
|
| 949 |
+
"loss": 0.0851,
|
| 950 |
+
"step": 15700
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 0.79,
|
| 954 |
"learning_rate": 2.1e-05,
|
| 955 |
+
"loss": 0.0865,
|
| 956 |
+
"step": 15800
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.8,
|
| 960 |
+
"learning_rate": 2.05e-05,
|
| 961 |
+
"loss": 0.0866,
|
| 962 |
+
"step": 15900
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.8,
|
| 966 |
"learning_rate": 2e-05,
|
| 967 |
+
"loss": 0.0868,
|
| 968 |
+
"step": 16000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.81,
|
| 972 |
+
"learning_rate": 1.9500000000000003e-05,
|
| 973 |
+
"loss": 0.087,
|
| 974 |
+
"step": 16100
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 0.81,
|
| 978 |
"learning_rate": 1.9e-05,
|
| 979 |
+
"loss": 0.0844,
|
| 980 |
+
"step": 16200
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.81,
|
| 984 |
+
"learning_rate": 1.85e-05,
|
| 985 |
+
"loss": 0.0835,
|
| 986 |
+
"step": 16300
|
| 987 |
},
|
| 988 |
{
|
| 989 |
"epoch": 0.82,
|
| 990 |
"learning_rate": 1.8e-05,
|
| 991 |
+
"loss": 0.0863,
|
| 992 |
+
"step": 16400
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.82,
|
| 996 |
+
"learning_rate": 1.75e-05,
|
| 997 |
+
"loss": 0.0854,
|
| 998 |
+
"step": 16500
|
| 999 |
},
|
| 1000 |
{
|
| 1001 |
"epoch": 0.83,
|
| 1002 |
"learning_rate": 1.7000000000000003e-05,
|
| 1003 |
+
"loss": 0.0857,
|
| 1004 |
+
"step": 16600
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.83,
|
| 1008 |
+
"learning_rate": 1.65e-05,
|
| 1009 |
+
"loss": 0.0844,
|
| 1010 |
+
"step": 16700
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.84,
|
| 1014 |
"learning_rate": 1.6000000000000003e-05,
|
| 1015 |
+
"loss": 0.0864,
|
| 1016 |
+
"step": 16800
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.84,
|
| 1020 |
+
"learning_rate": 1.55e-05,
|
| 1021 |
+
"loss": 0.0867,
|
| 1022 |
+
"step": 16900
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
"epoch": 0.85,
|
| 1026 |
"learning_rate": 1.5e-05,
|
| 1027 |
+
"loss": 0.0849,
|
| 1028 |
+
"step": 17000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.85,
|
| 1032 |
+
"learning_rate": 1.45e-05,
|
| 1033 |
+
"loss": 0.0815,
|
| 1034 |
+
"step": 17100
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 0.86,
|
| 1038 |
"learning_rate": 1.4000000000000001e-05,
|
| 1039 |
+
"loss": 0.0823,
|
| 1040 |
+
"step": 17200
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.86,
|
| 1044 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 1045 |
+
"loss": 0.0872,
|
| 1046 |
+
"step": 17300
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 0.87,
|
| 1050 |
"learning_rate": 1.3000000000000001e-05,
|
| 1051 |
+
"loss": 0.084,
|
| 1052 |
+
"step": 17400
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.88,
|
| 1056 |
+
"learning_rate": 1.25e-05,
|
| 1057 |
+
"loss": 0.0812,
|
| 1058 |
+
"step": 17500
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
"epoch": 0.88,
|
| 1062 |
"learning_rate": 1.2e-05,
|
| 1063 |
+
"loss": 0.0842,
|
| 1064 |
+
"step": 17600
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.89,
|
| 1068 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 1069 |
+
"loss": 0.087,
|
| 1070 |
+
"step": 17700
|
| 1071 |
},
|
| 1072 |
{
|
| 1073 |
"epoch": 0.89,
|
| 1074 |
"learning_rate": 1.1000000000000001e-05,
|
| 1075 |
+
"loss": 0.0843,
|
| 1076 |
+
"step": 17800
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.9,
|
| 1080 |
+
"learning_rate": 1.05e-05,
|
| 1081 |
+
"loss": 0.0833,
|
| 1082 |
+
"step": 17900
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
"epoch": 0.9,
|
| 1086 |
"learning_rate": 1e-05,
|
| 1087 |
+
"loss": 0.0836,
|
| 1088 |
+
"step": 18000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.91,
|
| 1092 |
+
"learning_rate": 9.5e-06,
|
| 1093 |
+
"loss": 0.0864,
|
| 1094 |
+
"step": 18100
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.91,
|
| 1098 |
"learning_rate": 9e-06,
|
| 1099 |
+
"loss": 0.0847,
|
| 1100 |
+
"step": 18200
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.92,
|
| 1104 |
+
"learning_rate": 8.500000000000002e-06,
|
| 1105 |
+
"loss": 0.084,
|
| 1106 |
+
"step": 18300
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
"epoch": 0.92,
|
| 1110 |
"learning_rate": 8.000000000000001e-06,
|
| 1111 |
+
"loss": 0.085,
|
| 1112 |
+
"step": 18400
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.93,
|
| 1116 |
+
"learning_rate": 7.5e-06,
|
| 1117 |
+
"loss": 0.0847,
|
| 1118 |
+
"step": 18500
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
"epoch": 0.93,
|
| 1122 |
"learning_rate": 7.000000000000001e-06,
|
| 1123 |
+
"loss": 0.0847,
|
| 1124 |
+
"step": 18600
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.94,
|
| 1128 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 1129 |
+
"loss": 0.085,
|
| 1130 |
+
"step": 18700
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 0.94,
|
| 1134 |
"learning_rate": 6e-06,
|
| 1135 |
+
"loss": 0.0836,
|
| 1136 |
+
"step": 18800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.94,
|
| 1140 |
+
"learning_rate": 5.500000000000001e-06,
|
| 1141 |
+
"loss": 0.0838,
|
| 1142 |
+
"step": 18900
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
"epoch": 0.95,
|
| 1146 |
"learning_rate": 5e-06,
|
| 1147 |
+
"loss": 0.0822,
|
| 1148 |
+
"step": 19000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.95,
|
| 1152 |
+
"learning_rate": 4.5e-06,
|
| 1153 |
+
"loss": 0.0821,
|
| 1154 |
+
"step": 19100
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
"epoch": 0.96,
|
| 1158 |
"learning_rate": 4.000000000000001e-06,
|
| 1159 |
+
"loss": 0.0852,
|
| 1160 |
+
"step": 19200
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.96,
|
| 1164 |
+
"learning_rate": 3.5000000000000004e-06,
|
| 1165 |
+
"loss": 0.0829,
|
| 1166 |
+
"step": 19300
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
"epoch": 0.97,
|
| 1170 |
"learning_rate": 3e-06,
|
| 1171 |
+
"loss": 0.085,
|
| 1172 |
+
"step": 19400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.97,
|
| 1176 |
+
"learning_rate": 2.5e-06,
|
| 1177 |
+
"loss": 0.0839,
|
| 1178 |
+
"step": 19500
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.98,
|
| 1182 |
"learning_rate": 2.0000000000000003e-06,
|
| 1183 |
+
"loss": 0.0829,
|
| 1184 |
+
"step": 19600
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.98,
|
| 1188 |
+
"learning_rate": 1.5e-06,
|
| 1189 |
+
"loss": 0.0849,
|
| 1190 |
+
"step": 19700
|
| 1191 |
},
|
| 1192 |
{
|
| 1193 |
"epoch": 0.99,
|
| 1194 |
"learning_rate": 1.0000000000000002e-06,
|
| 1195 |
+
"loss": 0.0842,
|
| 1196 |
+
"step": 19800
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.99,
|
| 1200 |
+
"learning_rate": 5.000000000000001e-07,
|
| 1201 |
+
"loss": 0.0837,
|
| 1202 |
+
"step": 19900
|
| 1203 |
},
|
| 1204 |
{
|
| 1205 |
"epoch": 1.0,
|
| 1206 |
"learning_rate": 0.0,
|
| 1207 |
+
"loss": 0.0797,
|
| 1208 |
+
"step": 20000
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
"epoch": 1.0,
|
| 1212 |
+
"step": 20000,
|
| 1213 |
"total_flos": 0,
|
| 1214 |
+
"train_runtime": 84238.4398,
|
| 1215 |
+
"train_samples_per_second": 0.237
|
| 1216 |
}
|
| 1217 |
],
|
| 1218 |
+
"max_steps": 20000,
|
| 1219 |
"num_train_epochs": 1,
|
| 1220 |
"total_flos": 0,
|
| 1221 |
"trial_name": null,
|