Commit
•
6c71261
1
Parent(s):
8bfabe6
End of training
Browse files- README.md +3 -0
- all_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +162 -690
README.md
CHANGED
|
@@ -1,4 +1,7 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
| 2 |
license: mit
|
| 3 |
base_model: facebook/mbart-large-50
|
| 4 |
tags:
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- zh
|
| 4 |
+
- yue
|
| 5 |
license: mit
|
| 6 |
base_model: facebook/mbart-large-50
|
| 7 |
tags:
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 0.23955335338535488,
|
| 4 |
+
"train_runtime": 7304.5803,
|
| 5 |
+
"train_samples": 100042,
|
| 6 |
+
"train_samples_per_second": 54.783,
|
| 7 |
+
"train_steps_per_second": 3.424
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 0.23955335338535488,
|
| 4 |
+
"train_runtime": 7304.5803,
|
| 5 |
+
"train_samples": 100042,
|
| 6 |
+
"train_samples_per_second": 54.783,
|
| 7 |
+
"train_steps_per_second": 3.424
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,856 +1,328 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 0.
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 0.
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 4.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 4.
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate": 4.
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 4.
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate": 4.
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 0.
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"learning_rate":
|
| 146 |
-
"loss": 0.
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch":
|
| 151 |
-
"learning_rate":
|
| 152 |
-
"loss": 0.
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0
|
| 157 |
-
"learning_rate":
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"learning_rate":
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 14000
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch":
|
| 181 |
-
"learning_rate":
|
| 182 |
-
"loss": 0.
|
| 183 |
"step": 14500
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch":
|
| 187 |
-
"learning_rate":
|
| 188 |
-
"loss": 0.
|
| 189 |
"step": 15000
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch":
|
| 193 |
-
"learning_rate":
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 15500
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch":
|
| 199 |
-
"learning_rate":
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 16000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
-
"learning_rate":
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 16500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 17000
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 17500
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch":
|
| 223 |
-
"learning_rate":
|
| 224 |
-
"loss": 0.
|
| 225 |
"step": 18000
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch":
|
| 229 |
-
"learning_rate":
|
| 230 |
-
"loss": 0.
|
| 231 |
"step": 18500
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch":
|
| 235 |
-
"learning_rate":
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 19000
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch":
|
| 241 |
-
"learning_rate":
|
| 242 |
-
"loss": 0.
|
| 243 |
"step": 19500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch":
|
| 247 |
-
"learning_rate":
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 20000
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch":
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 20500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 21000
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
-
"learning_rate":
|
| 266 |
-
"loss": 0.
|
| 267 |
"step": 21500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch":
|
| 271 |
-
"learning_rate":
|
| 272 |
-
"loss": 0.
|
| 273 |
"step": 22000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch":
|
| 277 |
-
"learning_rate":
|
| 278 |
-
"loss": 0.
|
| 279 |
"step": 22500
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch":
|
| 283 |
-
"learning_rate": 4.
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 23000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch":
|
| 289 |
-
"learning_rate":
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 23500
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 24000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch":
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 24500
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 0
|
| 307 |
-
"learning_rate": 4.
|
| 308 |
-
"loss": 0.
|
| 309 |
"step": 25000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch": 0
|
| 313 |
-
"
|
| 314 |
-
"
|
| 315 |
-
"
|
| 316 |
-
|
| 317 |
-
|
| 318 |
-
"
|
| 319 |
-
"learning_rate": 4.0589832717792745e-05,
|
| 320 |
-
"loss": 0.2163,
|
| 321 |
-
"step": 26000
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 0.96,
|
| 325 |
-
"learning_rate": 4.040879136794844e-05,
|
| 326 |
-
"loss": 0.2139,
|
| 327 |
-
"step": 26500
|
| 328 |
-
},
|
| 329 |
-
{
|
| 330 |
-
"epoch": 0.98,
|
| 331 |
-
"learning_rate": 4.0227750018104134e-05,
|
| 332 |
-
"loss": 0.2135,
|
| 333 |
-
"step": 27000
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 1.0,
|
| 337 |
-
"learning_rate": 4.004670866825983e-05,
|
| 338 |
-
"loss": 0.1986,
|
| 339 |
-
"step": 27500
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.01,
|
| 343 |
-
"learning_rate": 3.986566731841553e-05,
|
| 344 |
-
"loss": 0.1597,
|
| 345 |
-
"step": 28000
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 1.03,
|
| 349 |
-
"learning_rate": 3.968462596857123e-05,
|
| 350 |
-
"loss": 0.154,
|
| 351 |
-
"step": 28500
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 1.05,
|
| 355 |
-
"learning_rate": 3.950358461872692e-05,
|
| 356 |
-
"loss": 0.1381,
|
| 357 |
-
"step": 29000
|
| 358 |
-
},
|
| 359 |
-
{
|
| 360 |
-
"epoch": 1.07,
|
| 361 |
-
"learning_rate": 3.932326743428199e-05,
|
| 362 |
-
"loss": 0.1525,
|
| 363 |
-
"step": 29500
|
| 364 |
-
},
|
| 365 |
-
{
|
| 366 |
-
"epoch": 1.09,
|
| 367 |
-
"learning_rate": 3.914222608443769e-05,
|
| 368 |
-
"loss": 0.143,
|
| 369 |
-
"step": 30000
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 1.1,
|
| 373 |
-
"learning_rate": 3.896118473459338e-05,
|
| 374 |
-
"loss": 0.1422,
|
| 375 |
-
"step": 30500
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 1.12,
|
| 379 |
-
"learning_rate": 3.878014338474908e-05,
|
| 380 |
-
"loss": 0.1348,
|
| 381 |
-
"step": 31000
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.14,
|
| 385 |
-
"learning_rate": 3.859910203490477e-05,
|
| 386 |
-
"loss": 0.1421,
|
| 387 |
-
"step": 31500
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 1.16,
|
| 391 |
-
"learning_rate": 3.841806068506047e-05,
|
| 392 |
-
"loss": 0.1438,
|
| 393 |
-
"step": 32000
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 1.18,
|
| 397 |
-
"learning_rate": 3.823701933521617e-05,
|
| 398 |
-
"loss": 0.1375,
|
| 399 |
-
"step": 32500
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 1.19,
|
| 403 |
-
"learning_rate": 3.805597798537186e-05,
|
| 404 |
-
"loss": 0.1418,
|
| 405 |
-
"step": 33000
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 1.21,
|
| 409 |
-
"learning_rate": 3.787493663552756e-05,
|
| 410 |
-
"loss": 0.1336,
|
| 411 |
-
"step": 33500
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 1.23,
|
| 415 |
-
"learning_rate": 3.769425736838294e-05,
|
| 416 |
-
"loss": 0.1339,
|
| 417 |
-
"step": 34000
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 1.25,
|
| 421 |
-
"learning_rate": 3.751321601853864e-05,
|
| 422 |
-
"loss": 0.1383,
|
| 423 |
-
"step": 34500
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 1.27,
|
| 427 |
-
"learning_rate": 3.733217466869433e-05,
|
| 428 |
-
"loss": 0.1298,
|
| 429 |
-
"step": 35000
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 1.29,
|
| 433 |
-
"learning_rate": 3.715113331885003e-05,
|
| 434 |
-
"loss": 0.1314,
|
| 435 |
-
"step": 35500
|
| 436 |
-
},
|
| 437 |
-
{
|
| 438 |
-
"epoch": 1.3,
|
| 439 |
-
"learning_rate": 3.697009196900572e-05,
|
| 440 |
-
"loss": 0.1249,
|
| 441 |
-
"step": 36000
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"epoch": 1.32,
|
| 445 |
-
"learning_rate": 3.678905061916142e-05,
|
| 446 |
-
"loss": 0.1289,
|
| 447 |
-
"step": 36500
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 1.34,
|
| 451 |
-
"learning_rate": 3.660800926931711e-05,
|
| 452 |
-
"loss": 0.1289,
|
| 453 |
-
"step": 37000
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 1.36,
|
| 457 |
-
"learning_rate": 3.642696791947281e-05,
|
| 458 |
-
"loss": 0.1266,
|
| 459 |
-
"step": 37500
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 1.38,
|
| 463 |
-
"learning_rate": 3.62459265696285e-05,
|
| 464 |
-
"loss": 0.1201,
|
| 465 |
-
"step": 38000
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 1.39,
|
| 469 |
-
"learning_rate": 3.606524730248389e-05,
|
| 470 |
-
"loss": 0.1224,
|
| 471 |
-
"step": 38500
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 1.41,
|
| 475 |
-
"learning_rate": 3.5884205952639583e-05,
|
| 476 |
-
"loss": 0.1191,
|
| 477 |
-
"step": 39000
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 1.43,
|
| 481 |
-
"learning_rate": 3.570316460279528e-05,
|
| 482 |
-
"loss": 0.1157,
|
| 483 |
-
"step": 39500
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 1.45,
|
| 487 |
-
"learning_rate": 3.552212325295098e-05,
|
| 488 |
-
"loss": 0.1167,
|
| 489 |
-
"step": 40000
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 1.47,
|
| 493 |
-
"learning_rate": 3.534144398580636e-05,
|
| 494 |
-
"loss": 0.119,
|
| 495 |
-
"step": 40500
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 1.48,
|
| 499 |
-
"learning_rate": 3.5160402635962056e-05,
|
| 500 |
-
"loss": 0.1092,
|
| 501 |
-
"step": 41000
|
| 502 |
-
},
|
| 503 |
-
{
|
| 504 |
-
"epoch": 1.5,
|
| 505 |
-
"learning_rate": 3.4979361286117754e-05,
|
| 506 |
-
"loss": 0.1142,
|
| 507 |
-
"step": 41500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.52,
|
| 511 |
-
"learning_rate": 3.479831993627345e-05,
|
| 512 |
-
"loss": 0.1156,
|
| 513 |
-
"step": 42000
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 1.54,
|
| 517 |
-
"learning_rate": 3.4617278586429144e-05,
|
| 518 |
-
"loss": 0.1122,
|
| 519 |
-
"step": 42500
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 1.56,
|
| 523 |
-
"learning_rate": 3.443623723658484e-05,
|
| 524 |
-
"loss": 0.1104,
|
| 525 |
-
"step": 43000
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 1.58,
|
| 529 |
-
"learning_rate": 3.4255195886740534e-05,
|
| 530 |
-
"loss": 0.1076,
|
| 531 |
-
"step": 43500
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 1.59,
|
| 535 |
-
"learning_rate": 3.407415453689623e-05,
|
| 536 |
-
"loss": 0.1096,
|
| 537 |
-
"step": 44000
|
| 538 |
-
},
|
| 539 |
-
{
|
| 540 |
-
"epoch": 1.61,
|
| 541 |
-
"learning_rate": 3.3893113187051924e-05,
|
| 542 |
-
"loss": 0.1094,
|
| 543 |
-
"step": 44500
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 1.63,
|
| 547 |
-
"learning_rate": 3.371243391990731e-05,
|
| 548 |
-
"loss": 0.105,
|
| 549 |
-
"step": 45000
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 1.65,
|
| 553 |
-
"learning_rate": 3.3531392570063e-05,
|
| 554 |
-
"loss": 0.1059,
|
| 555 |
-
"step": 45500
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 1.67,
|
| 559 |
-
"learning_rate": 3.33503512202187e-05,
|
| 560 |
-
"loss": 0.1012,
|
| 561 |
-
"step": 46000
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 1.68,
|
| 565 |
-
"learning_rate": 3.3169309870374396e-05,
|
| 566 |
-
"loss": 0.0991,
|
| 567 |
-
"step": 46500
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"epoch": 1.7,
|
| 571 |
-
"learning_rate": 3.2988268520530094e-05,
|
| 572 |
-
"loss": 0.0978,
|
| 573 |
-
"step": 47000
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 1.72,
|
| 577 |
-
"learning_rate": 3.2807227170685786e-05,
|
| 578 |
-
"loss": 0.1003,
|
| 579 |
-
"step": 47500
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 1.74,
|
| 583 |
-
"learning_rate": 3.2626185820841484e-05,
|
| 584 |
-
"loss": 0.1,
|
| 585 |
-
"step": 48000
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 1.76,
|
| 589 |
-
"learning_rate": 3.244550655369686e-05,
|
| 590 |
-
"loss": 0.0976,
|
| 591 |
-
"step": 48500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 1.77,
|
| 595 |
-
"learning_rate": 3.226446520385256e-05,
|
| 596 |
-
"loss": 0.0961,
|
| 597 |
-
"step": 49000
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 1.79,
|
| 601 |
-
"learning_rate": 3.208342385400825e-05,
|
| 602 |
-
"loss": 0.1011,
|
| 603 |
-
"step": 49500
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 1.81,
|
| 607 |
-
"learning_rate": 3.190238250416395e-05,
|
| 608 |
-
"loss": 0.0902,
|
| 609 |
-
"step": 50000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 1.83,
|
| 613 |
-
"learning_rate": 3.1721703237019334e-05,
|
| 614 |
-
"loss": 0.0936,
|
| 615 |
-
"step": 50500
|
| 616 |
-
},
|
| 617 |
-
{
|
| 618 |
-
"epoch": 1.85,
|
| 619 |
-
"learning_rate": 3.154066188717503e-05,
|
| 620 |
-
"loss": 0.0903,
|
| 621 |
-
"step": 51000
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 1.86,
|
| 625 |
-
"learning_rate": 3.135962053733073e-05,
|
| 626 |
-
"loss": 0.082,
|
| 627 |
-
"step": 51500
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 1.88,
|
| 631 |
-
"learning_rate": 3.1178941270186115e-05,
|
| 632 |
-
"loss": 0.0841,
|
| 633 |
-
"step": 52000
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 1.9,
|
| 637 |
-
"learning_rate": 3.0997899920341806e-05,
|
| 638 |
-
"loss": 0.0898,
|
| 639 |
-
"step": 52500
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 1.92,
|
| 643 |
-
"learning_rate": 3.0816858570497505e-05,
|
| 644 |
-
"loss": 0.085,
|
| 645 |
-
"step": 53000
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 1.94,
|
| 649 |
-
"learning_rate": 3.06358172206532e-05,
|
| 650 |
-
"loss": 0.0807,
|
| 651 |
-
"step": 53500
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 1.96,
|
| 655 |
-
"learning_rate": 3.0454775870808895e-05,
|
| 656 |
-
"loss": 0.0886,
|
| 657 |
-
"step": 54000
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"epoch": 1.97,
|
| 661 |
-
"learning_rate": 3.0273734520964593e-05,
|
| 662 |
-
"loss": 0.0818,
|
| 663 |
-
"step": 54500
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 1.99,
|
| 667 |
-
"learning_rate": 3.0092693171120284e-05,
|
| 668 |
-
"loss": 0.0839,
|
| 669 |
-
"step": 55000
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 2.01,
|
| 673 |
-
"learning_rate": 2.9911651821275983e-05,
|
| 674 |
-
"loss": 0.0581,
|
| 675 |
-
"step": 55500
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 2.03,
|
| 679 |
-
"learning_rate": 2.9730610471431674e-05,
|
| 680 |
-
"loss": 0.0504,
|
| 681 |
-
"step": 56000
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 2.05,
|
| 685 |
-
"learning_rate": 2.9549569121587373e-05,
|
| 686 |
-
"loss": 0.0486,
|
| 687 |
-
"step": 56500
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 2.06,
|
| 691 |
-
"learning_rate": 2.9368527771743064e-05,
|
| 692 |
-
"loss": 0.0463,
|
| 693 |
-
"step": 57000
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"epoch": 2.08,
|
| 697 |
-
"learning_rate": 2.9187486421898762e-05,
|
| 698 |
-
"loss": 0.0545,
|
| 699 |
-
"step": 57500
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 2.1,
|
| 703 |
-
"learning_rate": 2.900644507205446e-05,
|
| 704 |
-
"loss": 0.0482,
|
| 705 |
-
"step": 58000
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 2.12,
|
| 709 |
-
"learning_rate": 2.8825403722210152e-05,
|
| 710 |
-
"loss": 0.0431,
|
| 711 |
-
"step": 58500
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 2.14,
|
| 715 |
-
"learning_rate": 2.864436237236585e-05,
|
| 716 |
-
"loss": 0.0494,
|
| 717 |
-
"step": 59000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 2.15,
|
| 721 |
-
"learning_rate": 2.8463683105221235e-05,
|
| 722 |
-
"loss": 0.0522,
|
| 723 |
-
"step": 59500
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 4.34,
|
| 727 |
-
"learning_rate": 6.565283510753857e-06,
|
| 728 |
-
"loss": 0.0392,
|
| 729 |
-
"step": 60000
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 4.38,
|
| 733 |
-
"learning_rate": 6.203200811065248e-06,
|
| 734 |
-
"loss": 0.0329,
|
| 735 |
-
"step": 60500
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 4.42,
|
| 739 |
-
"learning_rate": 5.841118111376639e-06,
|
| 740 |
-
"loss": 0.0306,
|
| 741 |
-
"step": 61000
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 4.45,
|
| 745 |
-
"learning_rate": 5.479759577087408e-06,
|
| 746 |
-
"loss": 0.0281,
|
| 747 |
-
"step": 61500
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 4.49,
|
| 751 |
-
"learning_rate": 5.117676877398798e-06,
|
| 752 |
-
"loss": 0.0275,
|
| 753 |
-
"step": 62000
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 4.53,
|
| 757 |
-
"learning_rate": 4.755594177710189e-06,
|
| 758 |
-
"loss": 0.0238,
|
| 759 |
-
"step": 62500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 4.56,
|
| 763 |
-
"learning_rate": 4.394235643420958e-06,
|
| 764 |
-
"loss": 0.0253,
|
| 765 |
-
"step": 63000
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 4.6,
|
| 769 |
-
"learning_rate": 4.032152943732348e-06,
|
| 770 |
-
"loss": 0.0222,
|
| 771 |
-
"step": 63500
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 4.63,
|
| 775 |
-
"learning_rate": 3.67007024404374e-06,
|
| 776 |
-
"loss": 0.0234,
|
| 777 |
-
"step": 64000
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 4.67,
|
| 781 |
-
"learning_rate": 3.307987544355131e-06,
|
| 782 |
-
"loss": 0.0214,
|
| 783 |
-
"step": 64500
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"epoch": 4.71,
|
| 787 |
-
"learning_rate": 2.945904844666522e-06,
|
| 788 |
-
"loss": 0.0217,
|
| 789 |
-
"step": 65000
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 4.74,
|
| 793 |
-
"learning_rate": 2.583822144977913e-06,
|
| 794 |
-
"loss": 0.022,
|
| 795 |
-
"step": 65500
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 4.78,
|
| 799 |
-
"learning_rate": 2.221739445289304e-06,
|
| 800 |
-
"loss": 0.0213,
|
| 801 |
-
"step": 66000
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 4.82,
|
| 805 |
-
"learning_rate": 1.8596567456006953e-06,
|
| 806 |
-
"loss": 0.0216,
|
| 807 |
-
"step": 66500
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 4.85,
|
| 811 |
-
"learning_rate": 1.4975740459120864e-06,
|
| 812 |
-
"loss": 0.0209,
|
| 813 |
-
"step": 67000
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 4.89,
|
| 817 |
-
"learning_rate": 1.1362155116228547e-06,
|
| 818 |
-
"loss": 0.0193,
|
| 819 |
-
"step": 67500
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 4.92,
|
| 823 |
-
"learning_rate": 7.748569773336231e-07,
|
| 824 |
-
"loss": 0.0199,
|
| 825 |
-
"step": 68000
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 4.96,
|
| 829 |
-
"learning_rate": 4.1349844304439136e-07,
|
| 830 |
-
"loss": 0.0183,
|
| 831 |
-
"step": 68500
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 5.0,
|
| 835 |
-
"learning_rate": 5.141574335578247e-08,
|
| 836 |
-
"loss": 0.0212,
|
| 837 |
-
"step": 69000
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 5.0,
|
| 841 |
-
"step": 69045,
|
| 842 |
-
"total_flos": 8.175381920769638e+16,
|
| 843 |
-
"train_loss": 0.003351212132286996,
|
| 844 |
-
"train_runtime": 2726.0601,
|
| 845 |
-
"train_samples_per_second": 405.237,
|
| 846 |
-
"train_steps_per_second": 25.328
|
| 847 |
}
|
| 848 |
],
|
| 849 |
"logging_steps": 500,
|
| 850 |
-
"max_steps":
|
| 851 |
-
"num_train_epochs":
|
| 852 |
"save_steps": 500,
|
| 853 |
-
"total_flos":
|
| 854 |
"trial_name": null,
|
| 855 |
"trial_params": null
|
| 856 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 25012,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.08,
|
| 13 |
+
"learning_rate": 4.9018471133855755e-05,
|
| 14 |
+
"loss": 1.9521,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.16,
|
| 19 |
+
"learning_rate": 4.801895090356629e-05,
|
| 20 |
+
"loss": 0.6834,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.24,
|
| 25 |
+
"learning_rate": 4.701943067327683e-05,
|
| 26 |
+
"loss": 0.55,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.32,
|
| 31 |
+
"learning_rate": 4.6019910442987366e-05,
|
| 32 |
+
"loss": 0.4952,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.4,
|
| 37 |
+
"learning_rate": 4.502039021269791e-05,
|
| 38 |
+
"loss": 0.4475,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.48,
|
| 43 |
+
"learning_rate": 4.402086998240845e-05,
|
| 44 |
+
"loss": 0.4157,
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.56,
|
| 49 |
+
"learning_rate": 4.3021349752118984e-05,
|
| 50 |
+
"loss": 0.3926,
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.64,
|
| 55 |
+
"learning_rate": 4.2021829521829525e-05,
|
| 56 |
+
"loss": 0.3788,
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.72,
|
| 61 |
+
"learning_rate": 4.102230929154006e-05,
|
| 62 |
+
"loss": 0.4176,
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.8,
|
| 67 |
+
"learning_rate": 4.00227890612506e-05,
|
| 68 |
+
"loss": 0.3467,
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.88,
|
| 73 |
+
"learning_rate": 3.902326883096114e-05,
|
| 74 |
+
"loss": 0.3491,
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.96,
|
| 79 |
+
"learning_rate": 3.802374860067168e-05,
|
| 80 |
+
"loss": 0.321,
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"learning_rate": 3.702422837038222e-05,
|
| 86 |
+
"loss": 0.2809,
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 1.12,
|
| 91 |
+
"learning_rate": 3.6024708140092754e-05,
|
| 92 |
+
"loss": 0.2329,
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 1.2,
|
| 97 |
+
"learning_rate": 3.5025187909803295e-05,
|
| 98 |
+
"loss": 0.2385,
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 1.28,
|
| 103 |
+
"learning_rate": 3.402566767951384e-05,
|
| 104 |
+
"loss": 0.2367,
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 1.36,
|
| 109 |
+
"learning_rate": 3.302614744922438e-05,
|
| 110 |
+
"loss": 0.232,
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 1.44,
|
| 115 |
+
"learning_rate": 3.202662721893491e-05,
|
| 116 |
+
"loss": 0.2325,
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 1.52,
|
| 121 |
+
"learning_rate": 3.102710698864545e-05,
|
| 122 |
+
"loss": 0.229,
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 1.6,
|
| 127 |
+
"learning_rate": 3.0027586758355993e-05,
|
| 128 |
+
"loss": 0.2213,
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 1.68,
|
| 133 |
+
"learning_rate": 2.9028066528066527e-05,
|
| 134 |
+
"loss": 0.2153,
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 1.76,
|
| 139 |
+
"learning_rate": 2.802854629777707e-05,
|
| 140 |
+
"loss": 0.2163,
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 1.84,
|
| 145 |
+
"learning_rate": 2.7029026067487607e-05,
|
| 146 |
+
"loss": 0.2152,
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 1.92,
|
| 151 |
+
"learning_rate": 2.602950583719815e-05,
|
| 152 |
+
"loss": 0.2061,
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 2.0,
|
| 157 |
+
"learning_rate": 2.5029985606908683e-05,
|
| 158 |
+
"loss": 0.205,
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 2.08,
|
| 163 |
+
"learning_rate": 2.4030465376619225e-05,
|
| 164 |
+
"loss": 0.1256,
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 2.16,
|
| 169 |
+
"learning_rate": 2.3030945146329763e-05,
|
| 170 |
+
"loss": 0.1247,
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 2.24,
|
| 175 |
+
"learning_rate": 2.20314249160403e-05,
|
| 176 |
+
"loss": 0.1249,
|
| 177 |
"step": 14000
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 2.32,
|
| 181 |
+
"learning_rate": 2.1033903726211422e-05,
|
| 182 |
+
"loss": 0.1263,
|
| 183 |
"step": 14500
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 2.4,
|
| 187 |
+
"learning_rate": 2.0034383495921957e-05,
|
| 188 |
+
"loss": 0.1283,
|
| 189 |
"step": 15000
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 2.48,
|
| 193 |
+
"learning_rate": 1.9034863265632495e-05,
|
| 194 |
+
"loss": 0.1279,
|
| 195 |
"step": 15500
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 2.56,
|
| 199 |
+
"learning_rate": 1.8035343035343037e-05,
|
| 200 |
+
"loss": 0.1247,
|
| 201 |
"step": 16000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 2.64,
|
| 205 |
+
"learning_rate": 1.7035822805053575e-05,
|
| 206 |
+
"loss": 0.1205,
|
| 207 |
"step": 16500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 2.72,
|
| 211 |
+
"learning_rate": 1.6036302574764113e-05,
|
| 212 |
+
"loss": 0.1227,
|
| 213 |
"step": 17000
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 2.8,
|
| 217 |
+
"learning_rate": 1.5036782344474653e-05,
|
| 218 |
+
"loss": 0.1173,
|
| 219 |
"step": 17500
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 2.88,
|
| 223 |
+
"learning_rate": 1.4037262114185193e-05,
|
| 224 |
+
"loss": 0.1211,
|
| 225 |
"step": 18000
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 2.96,
|
| 229 |
+
"learning_rate": 1.3037741883895729e-05,
|
| 230 |
+
"loss": 0.1209,
|
| 231 |
"step": 18500
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 3.04,
|
| 235 |
+
"learning_rate": 1.203822165360627e-05,
|
| 236 |
+
"loss": 0.0903,
|
| 237 |
"step": 19000
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 3.12,
|
| 241 |
+
"learning_rate": 1.1038701423316809e-05,
|
| 242 |
+
"loss": 0.0639,
|
| 243 |
"step": 19500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 3.2,
|
| 247 |
+
"learning_rate": 1.0039181193027347e-05,
|
| 248 |
+
"loss": 0.0616,
|
| 249 |
"step": 20000
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 3.28,
|
| 253 |
+
"learning_rate": 9.041660003198465e-06,
|
| 254 |
+
"loss": 0.0578,
|
| 255 |
"step": 20500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 3.36,
|
| 259 |
+
"learning_rate": 8.042139772909003e-06,
|
| 260 |
+
"loss": 0.0595,
|
| 261 |
"step": 21000
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 3.44,
|
| 265 |
+
"learning_rate": 7.042619542619543e-06,
|
| 266 |
+
"loss": 0.0564,
|
| 267 |
"step": 21500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 3.52,
|
| 271 |
+
"learning_rate": 6.043099312330082e-06,
|
| 272 |
+
"loss": 0.0594,
|
| 273 |
"step": 22000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 3.6,
|
| 277 |
+
"learning_rate": 5.043579082040621e-06,
|
| 278 |
+
"loss": 0.0566,
|
| 279 |
"step": 22500
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 3.68,
|
| 283 |
+
"learning_rate": 4.04405885175116e-06,
|
| 284 |
+
"loss": 0.0582,
|
| 285 |
"step": 23000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 3.76,
|
| 289 |
+
"learning_rate": 3.0445386214616987e-06,
|
| 290 |
+
"loss": 0.0572,
|
| 291 |
"step": 23500
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 3.84,
|
| 295 |
+
"learning_rate": 2.0450183911722376e-06,
|
| 296 |
+
"loss": 0.0561,
|
| 297 |
"step": 24000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 3.92,
|
| 301 |
+
"learning_rate": 1.0454981608827763e-06,
|
| 302 |
+
"loss": 0.0538,
|
| 303 |
"step": 24500
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 4.0,
|
| 307 |
+
"learning_rate": 4.797697105389413e-08,
|
| 308 |
+
"loss": 0.0546,
|
| 309 |
"step": 25000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 4.0,
|
| 313 |
+
"step": 25012,
|
| 314 |
+
"total_flos": 5.323473007568486e+16,
|
| 315 |
+
"train_loss": 0.23955335338535488,
|
| 316 |
+
"train_runtime": 7304.5803,
|
| 317 |
+
"train_samples_per_second": 54.783,
|
| 318 |
+
"train_steps_per_second": 3.424
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 319 |
}
|
| 320 |
],
|
| 321 |
"logging_steps": 500,
|
| 322 |
+
"max_steps": 25012,
|
| 323 |
+
"num_train_epochs": 4,
|
| 324 |
"save_steps": 500,
|
| 325 |
+
"total_flos": 5.323473007568486e+16,
|
| 326 |
"trial_name": null,
|
| 327 |
"trial_params": null
|
| 328 |
}
|