Upload 19 files
Browse files- scheduler.pt +1 -1
- trainer_state.json +225 -225
- training_args.bin +1 -1
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9014c6af467caecfd470aa6d7788105f86c8acbdb634cf6a0ad3654ec70328ed
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 3.
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 150,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -10,461 +10,461 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.05,
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 0.
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.1,
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 0.
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.15,
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 0.
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 4.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 4.
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate": 4.
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 4.
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate": 4.
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate": 4.
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate": 4.
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate": 4.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate": 4.
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate": 4.
|
| 98 |
-
"loss": 0.
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate": 4.
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate": 4.
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate": 4.
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate": 4.
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"learning_rate": 4.
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 1.
|
| 133 |
-
"learning_rate": 4.
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 1.
|
| 139 |
-
"learning_rate": 4.
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 1.
|
| 145 |
-
"learning_rate": 4.
|
| 146 |
-
"loss": 0.
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 1.
|
| 151 |
-
"learning_rate": 4.
|
| 152 |
-
"loss": 0.
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 1.
|
| 157 |
-
"learning_rate": 4.
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 1.
|
| 163 |
-
"learning_rate": 4.
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 1.
|
| 169 |
-
"learning_rate": 4.
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 1.
|
| 175 |
-
"learning_rate": 4.
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 1.
|
| 181 |
-
"learning_rate":
|
| 182 |
-
"loss": 0.
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 1.
|
| 187 |
-
"learning_rate":
|
| 188 |
-
"loss": 0.
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 1.
|
| 193 |
-
"learning_rate":
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 1.
|
| 199 |
-
"learning_rate": 3.
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 64
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 1.
|
| 205 |
-
"learning_rate": 3.
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 66
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch": 1.
|
| 211 |
-
"learning_rate": 3.
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 68
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 1.
|
| 217 |
-
"learning_rate": 3.
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 70
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 1.
|
| 223 |
-
"learning_rate": 3.
|
| 224 |
-
"loss": 0.
|
| 225 |
"step": 72
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch": 1.
|
| 229 |
-
"learning_rate": 3.
|
| 230 |
-
"loss": 0.
|
| 231 |
"step": 74
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch": 1.
|
| 235 |
-
"learning_rate": 3.
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 76
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch":
|
| 241 |
-
"learning_rate": 3.
|
| 242 |
-
"loss": 0.
|
| 243 |
"step": 78
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch":
|
| 247 |
-
"learning_rate": 3.
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 80
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 2.
|
| 253 |
-
"learning_rate": 3.
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 82
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 2.
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 84
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 2.
|
| 265 |
-
"learning_rate":
|
| 266 |
-
"loss": 0.
|
| 267 |
"step": 86
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch": 2.
|
| 271 |
-
"learning_rate":
|
| 272 |
-
"loss": 0.
|
| 273 |
"step": 88
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch": 2.
|
| 277 |
-
"learning_rate":
|
| 278 |
-
"loss": 0.
|
| 279 |
"step": 90
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch": 2.
|
| 283 |
-
"learning_rate": 2.
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 92
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch": 2.
|
| 289 |
-
"learning_rate": 2.
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 94
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch": 2.
|
| 295 |
-
"learning_rate": 2.
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 96
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 2.
|
| 301 |
-
"learning_rate": 2.
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 98
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 2.
|
| 307 |
-
"learning_rate": 2.
|
| 308 |
-
"loss": 0.
|
| 309 |
"step": 100
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch": 2.
|
| 313 |
-
"learning_rate": 2.
|
| 314 |
-
"loss": 0.
|
| 315 |
"step": 102
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch": 2.
|
| 319 |
-
"learning_rate": 2.
|
| 320 |
-
"loss": 0.
|
| 321 |
"step": 104
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch": 2.
|
| 325 |
-
"learning_rate": 2.
|
| 326 |
-
"loss": 0.
|
| 327 |
"step": 106
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch": 2.
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 108
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 2.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 110
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 2.
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 112
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 2.
|
| 349 |
-
"learning_rate":
|
| 350 |
-
"loss": 0.
|
| 351 |
"step": 114
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch": 2.
|
| 355 |
-
"learning_rate":
|
| 356 |
-
"loss": 0.
|
| 357 |
"step": 116
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch":
|
| 361 |
-
"learning_rate": 1.
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 118
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch":
|
| 367 |
-
"learning_rate": 1.
|
| 368 |
-
"loss": 0.
|
| 369 |
"step": 120
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 3.
|
| 373 |
-
"learning_rate": 1.
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 122
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch": 3.
|
| 379 |
-
"learning_rate": 1.
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 124
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 3.
|
| 385 |
-
"learning_rate": 1.
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 126
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 3.
|
| 391 |
-
"learning_rate": 1.
|
| 392 |
-
"loss": 0.
|
| 393 |
"step": 128
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch": 3.
|
| 397 |
-
"learning_rate": 1.
|
| 398 |
-
"loss": 0.
|
| 399 |
"step": 130
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch": 3.
|
| 403 |
-
"learning_rate": 1.
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 132
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch": 3.
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 134
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch": 3.
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 136
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch": 3.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 138
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 3.
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 140
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 3.
|
| 433 |
-
"learning_rate":
|
| 434 |
-
"loss": 0.
|
| 435 |
"step": 142
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch": 3.
|
| 439 |
-
"learning_rate":
|
| 440 |
-
"loss": 0.
|
| 441 |
"step": 144
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch": 3.
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss": 0.
|
| 447 |
"step": 146
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch": 3.
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 148
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch": 3.
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 150
|
| 460 |
}
|
| 461 |
],
|
| 462 |
"logging_steps": 2,
|
| 463 |
-
"max_steps":
|
| 464 |
"num_input_tokens_seen": 0,
|
| 465 |
"num_train_epochs": 5,
|
| 466 |
"save_steps": 25,
|
| 467 |
-
"total_flos": 2.
|
| 468 |
"train_batch_size": 2,
|
| 469 |
"trial_name": null,
|
| 470 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.8585209003215435,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 150,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.05,
|
| 13 |
+
"learning_rate": 4.9986331433523156e-05,
|
| 14 |
+
"loss": 0.6891,
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.1,
|
| 19 |
+
"learning_rate": 4.994534068046937e-05,
|
| 20 |
+
"loss": 0.5774,
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.15,
|
| 25 |
+
"learning_rate": 4.9877072563625285e-05,
|
| 26 |
+
"loss": 0.5359,
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.21,
|
| 31 |
+
"learning_rate": 4.978160173317438e-05,
|
| 32 |
+
"loss": 0.541,
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.26,
|
| 37 |
+
"learning_rate": 4.965903258506806e-05,
|
| 38 |
+
"loss": 0.5038,
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.31,
|
| 43 |
+
"learning_rate": 4.9509499146870236e-05,
|
| 44 |
+
"loss": 0.5397,
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.36,
|
| 49 |
+
"learning_rate": 4.933316493120015e-05,
|
| 50 |
+
"loss": 0.5097,
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.41,
|
| 55 |
+
"learning_rate": 4.913022275693372e-05,
|
| 56 |
+
"loss": 0.4829,
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.46,
|
| 61 |
+
"learning_rate": 4.8900894538358944e-05,
|
| 62 |
+
"loss": 0.503,
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.51,
|
| 67 |
+
"learning_rate": 4.864543104251587e-05,
|
| 68 |
+
"loss": 0.535,
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.57,
|
| 73 |
+
"learning_rate": 4.8364111614986527e-05,
|
| 74 |
+
"loss": 0.4685,
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.62,
|
| 79 |
+
"learning_rate": 4.805724387443462e-05,
|
| 80 |
+
"loss": 0.5223,
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.67,
|
| 85 |
+
"learning_rate": 4.7725163376229064e-05,
|
| 86 |
+
"loss": 0.5033,
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.72,
|
| 91 |
+
"learning_rate": 4.736823324551909e-05,
|
| 92 |
+
"loss": 0.4991,
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.77,
|
| 97 |
+
"learning_rate": 4.698684378016222e-05,
|
| 98 |
+
"loss": 0.5225,
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.82,
|
| 103 |
+
"learning_rate": 4.6581412023939354e-05,
|
| 104 |
+
"loss": 0.4711,
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.87,
|
| 109 |
+
"learning_rate": 4.6152381310523387e-05,
|
| 110 |
+
"loss": 0.5054,
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.93,
|
| 115 |
+
"learning_rate": 4.5700220778700504e-05,
|
| 116 |
+
"loss": 0.5237,
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.98,
|
| 121 |
+
"learning_rate": 4.522542485937369e-05,
|
| 122 |
+
"loss": 0.4992,
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 1.03,
|
| 127 |
+
"learning_rate": 4.4728512734909844e-05,
|
| 128 |
+
"loss": 0.3735,
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 1.08,
|
| 133 |
+
"learning_rate": 4.421002777142148e-05,
|
| 134 |
+
"loss": 0.2735,
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 1.13,
|
| 139 |
+
"learning_rate": 4.367053692460385e-05,
|
| 140 |
+
"loss": 0.2565,
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 1.18,
|
| 145 |
+
"learning_rate": 4.311063011977723e-05,
|
| 146 |
+
"loss": 0.2926,
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 1.23,
|
| 151 |
+
"learning_rate": 4.2530919606812216e-05,
|
| 152 |
+
"loss": 0.281,
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 1.29,
|
| 157 |
+
"learning_rate": 4.193203929064353e-05,
|
| 158 |
+
"loss": 0.2619,
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 1.34,
|
| 163 |
+
"learning_rate": 4.131464403810422e-05,
|
| 164 |
+
"loss": 0.2456,
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 1.39,
|
| 169 |
+
"learning_rate": 4.067940896183843e-05,
|
| 170 |
+
"loss": 0.2442,
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 1.44,
|
| 175 |
+
"learning_rate": 4.002702868207563e-05,
|
| 176 |
+
"loss": 0.2625,
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 1.49,
|
| 181 |
+
"learning_rate": 3.935821656707359e-05,
|
| 182 |
+
"loss": 0.2616,
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 1.54,
|
| 187 |
+
"learning_rate": 3.867370395306068e-05,
|
| 188 |
+
"loss": 0.2545,
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 1.59,
|
| 193 |
+
"learning_rate": 3.797423934453038e-05,
|
| 194 |
+
"loss": 0.2728,
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 1.65,
|
| 199 |
+
"learning_rate": 3.726058759576271e-05,
|
| 200 |
+
"loss": 0.234,
|
| 201 |
"step": 64
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 1.7,
|
| 205 |
+
"learning_rate": 3.65335290744672e-05,
|
| 206 |
+
"loss": 0.2356,
|
| 207 |
"step": 66
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 1.75,
|
| 211 |
+
"learning_rate": 3.579385880846232e-05,
|
| 212 |
+
"loss": 0.2369,
|
| 213 |
"step": 68
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 1.8,
|
| 217 |
+
"learning_rate": 3.504238561632424e-05,
|
| 218 |
+
"loss": 0.236,
|
| 219 |
"step": 70
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 1.85,
|
| 223 |
+
"learning_rate": 3.427993122295552e-05,
|
| 224 |
+
"loss": 0.2607,
|
| 225 |
"step": 72
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 1.9,
|
| 229 |
+
"learning_rate": 3.350732936104108e-05,
|
| 230 |
+
"loss": 0.2637,
|
| 231 |
"step": 74
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 1.95,
|
| 235 |
+
"learning_rate": 3.272542485937369e-05,
|
| 236 |
+
"loss": 0.2338,
|
| 237 |
"step": 76
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 2.01,
|
| 241 |
+
"learning_rate": 3.1935072719046115e-05,
|
| 242 |
+
"loss": 0.2489,
|
| 243 |
"step": 78
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 2.06,
|
| 247 |
+
"learning_rate": 3.1137137178519985e-05,
|
| 248 |
+
"loss": 0.1337,
|
| 249 |
"step": 80
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 2.11,
|
| 253 |
+
"learning_rate": 3.0332490768593675e-05,
|
| 254 |
+
"loss": 0.1188,
|
| 255 |
"step": 82
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 2.16,
|
| 259 |
+
"learning_rate": 2.952201335830275e-05,
|
| 260 |
+
"loss": 0.1067,
|
| 261 |
"step": 84
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 2.21,
|
| 265 |
+
"learning_rate": 2.870659119279605e-05,
|
| 266 |
+
"loss": 0.1131,
|
| 267 |
"step": 86
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 2.26,
|
| 271 |
+
"learning_rate": 2.788711592423966e-05,
|
| 272 |
+
"loss": 0.1308,
|
| 273 |
"step": 88
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 2.32,
|
| 277 |
+
"learning_rate": 2.7064483636808313e-05,
|
| 278 |
+
"loss": 0.1236,
|
| 279 |
"step": 90
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 2.37,
|
| 283 |
+
"learning_rate": 2.623959386683056e-05,
|
| 284 |
+
"loss": 0.1079,
|
| 285 |
"step": 92
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 2.42,
|
| 289 |
+
"learning_rate": 2.5413348619158967e-05,
|
| 290 |
+
"loss": 0.1102,
|
| 291 |
"step": 94
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 2.47,
|
| 295 |
+
"learning_rate": 2.458665138084104e-05,
|
| 296 |
+
"loss": 0.1218,
|
| 297 |
"step": 96
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 2.52,
|
| 301 |
+
"learning_rate": 2.3760406133169443e-05,
|
| 302 |
+
"loss": 0.1148,
|
| 303 |
"step": 98
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 2.57,
|
| 307 |
+
"learning_rate": 2.2935516363191693e-05,
|
| 308 |
+
"loss": 0.1109,
|
| 309 |
"step": 100
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 2.62,
|
| 313 |
+
"learning_rate": 2.2112884075760347e-05,
|
| 314 |
+
"loss": 0.1088,
|
| 315 |
"step": 102
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 2.68,
|
| 319 |
+
"learning_rate": 2.1293408807203947e-05,
|
| 320 |
+
"loss": 0.1109,
|
| 321 |
"step": 104
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 2.73,
|
| 325 |
+
"learning_rate": 2.047798664169726e-05,
|
| 326 |
+
"loss": 0.0939,
|
| 327 |
"step": 106
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 2.78,
|
| 331 |
+
"learning_rate": 1.9667509231406334e-05,
|
| 332 |
+
"loss": 0.1179,
|
| 333 |
"step": 108
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 2.83,
|
| 337 |
+
"learning_rate": 1.8862862821480025e-05,
|
| 338 |
+
"loss": 0.095,
|
| 339 |
"step": 110
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 2.88,
|
| 343 |
+
"learning_rate": 1.806492728095389e-05,
|
| 344 |
+
"loss": 0.115,
|
| 345 |
"step": 112
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 2.93,
|
| 349 |
+
"learning_rate": 1.7274575140626318e-05,
|
| 350 |
+
"loss": 0.1219,
|
| 351 |
"step": 114
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 2.98,
|
| 355 |
+
"learning_rate": 1.6492670638958924e-05,
|
| 356 |
+
"loss": 0.1153,
|
| 357 |
"step": 116
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 3.04,
|
| 361 |
+
"learning_rate": 1.5720068777044476e-05,
|
| 362 |
+
"loss": 0.0693,
|
| 363 |
"step": 118
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 3.09,
|
| 367 |
+
"learning_rate": 1.495761438367577e-05,
|
| 368 |
+
"loss": 0.0535,
|
| 369 |
"step": 120
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 3.14,
|
| 373 |
+
"learning_rate": 1.4206141191537682e-05,
|
| 374 |
+
"loss": 0.0452,
|
| 375 |
"step": 122
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 3.19,
|
| 379 |
+
"learning_rate": 1.346647092553281e-05,
|
| 380 |
+
"loss": 0.0444,
|
| 381 |
"step": 124
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 3.24,
|
| 385 |
+
"learning_rate": 1.2739412404237306e-05,
|
| 386 |
+
"loss": 0.0364,
|
| 387 |
"step": 126
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 3.29,
|
| 391 |
+
"learning_rate": 1.202576065546963e-05,
|
| 392 |
+
"loss": 0.0591,
|
| 393 |
"step": 128
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 3.34,
|
| 397 |
+
"learning_rate": 1.1326296046939333e-05,
|
| 398 |
+
"loss": 0.0466,
|
| 399 |
"step": 130
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 3.4,
|
| 403 |
+
"learning_rate": 1.064178343292641e-05,
|
| 404 |
+
"loss": 0.0558,
|
| 405 |
"step": 132
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 3.45,
|
| 409 |
+
"learning_rate": 9.972971317924374e-06,
|
| 410 |
+
"loss": 0.0533,
|
| 411 |
"step": 134
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 3.5,
|
| 415 |
+
"learning_rate": 9.320591038161574e-06,
|
| 416 |
+
"loss": 0.0549,
|
| 417 |
"step": 136
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 3.55,
|
| 421 |
+
"learning_rate": 8.685355961895784e-06,
|
| 422 |
+
"loss": 0.0446,
|
| 423 |
"step": 138
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 3.6,
|
| 427 |
+
"learning_rate": 8.067960709356478e-06,
|
| 428 |
+
"loss": 0.0622,
|
| 429 |
"step": 140
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 3.65,
|
| 433 |
+
"learning_rate": 7.469080393187786e-06,
|
| 434 |
+
"loss": 0.0494,
|
| 435 |
"step": 142
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 3.7,
|
| 439 |
+
"learning_rate": 6.889369880222776e-06,
|
| 440 |
+
"loss": 0.0391,
|
| 441 |
"step": 144
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 3.76,
|
| 445 |
+
"learning_rate": 6.329463075396161e-06,
|
| 446 |
+
"loss": 0.0399,
|
| 447 |
"step": 146
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 3.81,
|
| 451 |
+
"learning_rate": 5.78997222857853e-06,
|
| 452 |
+
"loss": 0.0509,
|
| 453 |
"step": 148
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 3.86,
|
| 457 |
+
"learning_rate": 5.271487265090163e-06,
|
| 458 |
+
"loss": 0.0451,
|
| 459 |
"step": 150
|
| 460 |
}
|
| 461 |
],
|
| 462 |
"logging_steps": 2,
|
| 463 |
+
"max_steps": 190,
|
| 464 |
"num_input_tokens_seen": 0,
|
| 465 |
"num_train_epochs": 5,
|
| 466 |
"save_steps": 25,
|
| 467 |
+
"total_flos": 2.0183987688434565e+18,
|
| 468 |
"train_batch_size": 2,
|
| 469 |
"trial_name": null,
|
| 470 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6136
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab83e5c27dd3eaa07b3eceb08fb1dbeb5b64cc39f8d17b2630b3cc537cb322ad
|
| 3 |
size 6136
|