Upload folder using huggingface_hub
Browse files- adapter_config.json +2 -2
- adapter_model.safetensors +1 -1
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +810 -204
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -17,9 +17,9 @@
|
|
| 17 |
"revision": null,
|
| 18 |
"target_modules": [
|
| 19 |
"dense",
|
| 20 |
-
"query_key_value",
|
| 21 |
"dense_h_to_4h",
|
| 22 |
-
"dense_4h_to_h"
|
|
|
|
| 23 |
],
|
| 24 |
"task_type": "CAUSAL_LM"
|
| 25 |
}
|
|
|
|
| 17 |
"revision": null,
|
| 18 |
"target_modules": [
|
| 19 |
"dense",
|
|
|
|
| 20 |
"dense_h_to_4h",
|
| 21 |
+
"dense_4h_to_h",
|
| 22 |
+
"query_key_value"
|
| 23 |
],
|
| 24 |
"task_type": "CAUSAL_LM"
|
| 25 |
}
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 134235712
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87ed2acf6125adcd0b24b33bc85700484416c3f26f621297c7c4b99078d32734
|
| 3 |
size 134235712
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ce2abea04ce84564218c3cf846e1bf599b1376d44e961f38757081a2ec3ee3c
|
| 3 |
+
size 268515002
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5df93484ff14822d285a99b747b77d524936bc9c829bc86f91863db23667392
|
| 3 |
size 14244
|
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 sha256:9b5b7f865fff447a834d063766ffbff2b06cf8776b1ce383609115b7efb4a180
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,619 +1,1225 @@
|
|
| 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": 0.03,
|
| 13 |
-
"learning_rate":
|
| 14 |
"loss": 2.7224,
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.06,
|
| 19 |
-
"learning_rate":
|
| 20 |
-
"loss": 2.
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.09,
|
| 25 |
-
"learning_rate":
|
| 26 |
-
"loss": 2.
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.12,
|
| 31 |
-
"learning_rate": 1.
|
| 32 |
-
"loss": 2.
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.15,
|
| 37 |
-
"learning_rate": 1.
|
| 38 |
-
"loss": 1.
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.18,
|
| 43 |
-
"learning_rate": 1.
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.21,
|
| 49 |
-
"learning_rate": 1.
|
| 50 |
-
"loss": 2.
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.24,
|
| 55 |
-
"learning_rate": 1.
|
| 56 |
-
"loss": 3.
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.27,
|
| 61 |
-
"learning_rate": 1.
|
| 62 |
-
"loss": 2.
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.3,
|
| 67 |
-
"learning_rate": 1.
|
| 68 |
-
"loss": 1.
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.33,
|
| 73 |
-
"learning_rate": 1.
|
| 74 |
-
"loss": 1.
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.36,
|
| 79 |
-
"learning_rate": 1.
|
| 80 |
-
"loss": 1.
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.39,
|
| 85 |
-
"learning_rate": 1.
|
| 86 |
-
"loss": 1.
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.42,
|
| 91 |
-
"learning_rate": 1.
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.45,
|
| 97 |
-
"learning_rate": 1.
|
| 98 |
-
"loss": 2.
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.48,
|
| 103 |
-
"learning_rate": 1.
|
| 104 |
-
"loss": 2.
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.51,
|
| 109 |
-
"learning_rate": 1.
|
| 110 |
-
"loss": 2.
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"epoch": 0.54,
|
| 115 |
-
"learning_rate": 1.
|
| 116 |
-
"loss": 2.
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 0.57,
|
| 121 |
-
"learning_rate": 1.
|
| 122 |
-
"loss": 1.
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 0.6,
|
| 127 |
-
"learning_rate": 1.
|
| 128 |
-
"loss": 1.
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.63,
|
| 133 |
-
"learning_rate": 1.
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.66,
|
| 139 |
-
"learning_rate": 1.
|
| 140 |
-
"loss": 2.
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
"epoch": 0.69,
|
| 145 |
-
"learning_rate": 1.
|
| 146 |
-
"loss": 2.
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
"epoch": 0.72,
|
| 151 |
-
"learning_rate": 1.
|
| 152 |
-
"loss": 2.
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.75,
|
| 157 |
-
"learning_rate": 1.
|
| 158 |
-
"loss": 2.
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.78,
|
| 163 |
-
"learning_rate": 1.
|
| 164 |
-
"loss": 1.
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.81,
|
| 169 |
-
"learning_rate": 1.
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.84,
|
| 175 |
-
"learning_rate": 1.
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.86,
|
| 181 |
-
"learning_rate": 1.
|
| 182 |
-
"loss": 1.
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
"epoch": 0.89,
|
| 187 |
-
"learning_rate": 1.
|
| 188 |
-
"loss": 1.
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.92,
|
| 193 |
-
"learning_rate": 1.
|
| 194 |
-
"loss": 1.
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 0.95,
|
| 199 |
-
"learning_rate": 1.
|
| 200 |
-
"loss": 2.
|
| 201 |
"step": 64
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 0.98,
|
| 205 |
-
"learning_rate": 1.
|
| 206 |
-
"loss": 1.
|
| 207 |
"step": 66
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 1.01,
|
| 211 |
-
"learning_rate": 1.
|
| 212 |
-
"loss": 2.
|
| 213 |
"step": 68
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 1.04,
|
| 217 |
-
"learning_rate": 1.
|
| 218 |
-
"loss": 2.
|
| 219 |
"step": 70
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 1.07,
|
| 223 |
-
"learning_rate": 1.
|
| 224 |
-
"loss": 1.
|
| 225 |
"step": 72
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 1.1,
|
| 229 |
-
"learning_rate": 1.
|
| 230 |
-
"loss": 1.
|
| 231 |
"step": 74
|
| 232 |
},
|
| 233 |
{
|
| 234 |
"epoch": 1.13,
|
| 235 |
-
"learning_rate": 1.
|
| 236 |
-
"loss": 1.
|
| 237 |
"step": 76
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 1.16,
|
| 241 |
-
"learning_rate": 1.
|
| 242 |
-
"loss": 1.
|
| 243 |
"step": 78
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 1.19,
|
| 247 |
-
"learning_rate": 1.
|
| 248 |
-
"loss": 1.
|
| 249 |
"step": 80
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 1.22,
|
| 253 |
-
"learning_rate": 1.
|
| 254 |
-
"loss": 2.
|
| 255 |
"step": 82
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.25,
|
| 259 |
-
"learning_rate": 1.
|
| 260 |
-
"loss": 2.
|
| 261 |
"step": 84
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 1.28,
|
| 265 |
-
"learning_rate": 1.
|
| 266 |
-
"loss": 2.
|
| 267 |
"step": 86
|
| 268 |
},
|
| 269 |
{
|
| 270 |
"epoch": 1.31,
|
| 271 |
-
"learning_rate": 1.
|
| 272 |
-
"loss": 1.
|
| 273 |
"step": 88
|
| 274 |
},
|
| 275 |
{
|
| 276 |
"epoch": 1.34,
|
| 277 |
-
"learning_rate": 1.
|
| 278 |
-
"loss": 1.
|
| 279 |
"step": 90
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 1.37,
|
| 283 |
-
"learning_rate": 1.
|
| 284 |
-
"loss": 1.
|
| 285 |
"step": 92
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 1.4,
|
| 289 |
-
"learning_rate": 1.
|
| 290 |
-
"loss": 1.
|
| 291 |
"step": 94
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 1.43,
|
| 295 |
-
"learning_rate": 1.
|
| 296 |
-
"loss": 1.
|
| 297 |
"step": 96
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.46,
|
| 301 |
-
"learning_rate": 1.
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 98
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 1.49,
|
| 307 |
-
"learning_rate": 1.
|
| 308 |
-
"loss": 1.
|
| 309 |
"step": 100
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 1.52,
|
| 313 |
-
"learning_rate": 1.
|
| 314 |
-
"loss": 2.
|
| 315 |
"step": 102
|
| 316 |
},
|
| 317 |
{
|
| 318 |
"epoch": 1.55,
|
| 319 |
-
"learning_rate":
|
| 320 |
-
"loss": 1.
|
| 321 |
"step": 104
|
| 322 |
},
|
| 323 |
{
|
| 324 |
"epoch": 1.58,
|
| 325 |
-
"learning_rate":
|
| 326 |
-
"loss": 1.
|
| 327 |
"step": 106
|
| 328 |
},
|
| 329 |
{
|
| 330 |
"epoch": 1.61,
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss": 1.
|
| 333 |
"step": 108
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 1.64,
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 1.
|
| 339 |
"step": 110
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.67,
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 112
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 1.7,
|
| 349 |
-
"learning_rate":
|
| 350 |
-
"loss": 2.
|
| 351 |
"step": 114
|
| 352 |
},
|
| 353 |
{
|
| 354 |
"epoch": 1.73,
|
| 355 |
-
"learning_rate":
|
| 356 |
-
"loss": 2.
|
| 357 |
"step": 116
|
| 358 |
},
|
| 359 |
{
|
| 360 |
"epoch": 1.76,
|
| 361 |
-
"learning_rate":
|
| 362 |
-
"loss": 2.
|
| 363 |
"step": 118
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 1.79,
|
| 367 |
-
"learning_rate":
|
| 368 |
-
"loss": 1.
|
| 369 |
"step": 120
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 1.82,
|
| 373 |
-
"learning_rate":
|
| 374 |
-
"loss": 1.
|
| 375 |
"step": 122
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 1.85,
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss": 1.
|
| 381 |
"step": 124
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.88,
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 126
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 1.91,
|
| 391 |
-
"learning_rate":
|
| 392 |
-
"loss": 1.
|
| 393 |
"step": 128
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 1.94,
|
| 397 |
-
"learning_rate":
|
| 398 |
-
"loss": 2.
|
| 399 |
"step": 130
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 1.97,
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss": 1.
|
| 405 |
"step": 132
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 2.0,
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss": 1.
|
| 411 |
"step": 134
|
| 412 |
},
|
| 413 |
{
|
| 414 |
"epoch": 2.03,
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss": 2.
|
| 417 |
"step": 136
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 2.06,
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 138
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 2.09,
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 140
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 2.12,
|
| 433 |
-
"learning_rate":
|
| 434 |
-
"loss": 1.
|
| 435 |
"step": 142
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 2.15,
|
| 439 |
-
"learning_rate":
|
| 440 |
-
"loss": 1.
|
| 441 |
"step": 144
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 2.18,
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss": 1.
|
| 447 |
"step": 146
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 2.21,
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 1.
|
| 453 |
"step": 148
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 2.24,
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss": 1.
|
| 459 |
"step": 150
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 2.27,
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 2.
|
| 465 |
"step": 152
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 2.3,
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 2.
|
| 471 |
"step": 154
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 2.33,
|
| 475 |
-
"learning_rate":
|
| 476 |
-
"loss": 1.
|
| 477 |
"step": 156
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 2.36,
|
| 481 |
-
"learning_rate":
|
| 482 |
-
"loss": 1.
|
| 483 |
"step": 158
|
| 484 |
},
|
| 485 |
{
|
| 486 |
"epoch": 2.39,
|
| 487 |
-
"learning_rate":
|
| 488 |
-
"loss": 1.
|
| 489 |
"step": 160
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 2.42,
|
| 493 |
-
"learning_rate": 1.
|
| 494 |
-
"loss": 1.
|
| 495 |
"step": 162
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 2.45,
|
| 499 |
-
"learning_rate": 1.
|
| 500 |
-
"loss": 1.
|
| 501 |
"step": 164
|
| 502 |
},
|
| 503 |
{
|
| 504 |
"epoch": 2.48,
|
| 505 |
-
"learning_rate": 1.
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 166
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 2.51,
|
| 511 |
-
"learning_rate": 1.
|
| 512 |
-
"loss": 2.
|
| 513 |
"step": 168
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 2.53,
|
| 517 |
-
"learning_rate": 1.
|
| 518 |
-
"loss": 1.
|
| 519 |
"step": 170
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 2.56,
|
| 523 |
-
"learning_rate": 1.
|
| 524 |
-
"loss": 1.
|
| 525 |
"step": 172
|
| 526 |
},
|
| 527 |
{
|
| 528 |
"epoch": 2.59,
|
| 529 |
-
"learning_rate":
|
| 530 |
-
"loss": 1.
|
| 531 |
"step": 174
|
| 532 |
},
|
| 533 |
{
|
| 534 |
"epoch": 2.62,
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss": 1.
|
| 537 |
"step": 176
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 2.65,
|
| 541 |
-
"learning_rate":
|
| 542 |
-
"loss": 1.
|
| 543 |
"step": 178
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 2.68,
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss": 1.
|
| 549 |
"step": 180
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 2.71,
|
| 553 |
-
"learning_rate":
|
| 554 |
-
"loss":
|
| 555 |
"step": 182
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 2.74,
|
| 559 |
-
"learning_rate":
|
| 560 |
-
"loss": 2.
|
| 561 |
"step": 184
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 2.77,
|
| 565 |
-
"learning_rate":
|
| 566 |
-
"loss": 1.
|
| 567 |
"step": 186
|
| 568 |
},
|
| 569 |
{
|
| 570 |
"epoch": 2.8,
|
| 571 |
-
"learning_rate":
|
| 572 |
-
"loss": 1.
|
| 573 |
"step": 188
|
| 574 |
},
|
| 575 |
{
|
| 576 |
"epoch": 2.83,
|
| 577 |
-
"learning_rate": 1.
|
| 578 |
-
"loss": 1.
|
| 579 |
"step": 190
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 2.86,
|
| 583 |
-
"learning_rate": 1.
|
| 584 |
-
"loss": 1.
|
| 585 |
"step": 192
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 2.89,
|
| 589 |
-
"learning_rate":
|
| 590 |
-
"loss": 1.
|
| 591 |
"step": 194
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 2.92,
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss":
|
| 597 |
"step": 196
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 2.95,
|
| 601 |
-
"learning_rate": 1.
|
| 602 |
-
"loss": 1.
|
| 603 |
"step": 198
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 2.98,
|
| 607 |
-
"learning_rate": 1.
|
| 608 |
-
"loss": 1.
|
| 609 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 610 |
}
|
| 611 |
],
|
| 612 |
"logging_steps": 2,
|
| 613 |
-
"max_steps":
|
| 614 |
-
"num_train_epochs":
|
| 615 |
"save_steps": 500,
|
| 616 |
-
"total_flos":
|
| 617 |
"trial_name": null,
|
| 618 |
"trial_params": null
|
| 619 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.994408201304753,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 402,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.03,
|
| 13 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 14 |
"loss": 2.7224,
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.06,
|
| 19 |
+
"learning_rate": 6.153846153846155e-06,
|
| 20 |
+
"loss": 2.023,
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.09,
|
| 25 |
+
"learning_rate": 9.230769230769232e-06,
|
| 26 |
+
"loss": 2.011,
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.12,
|
| 31 |
+
"learning_rate": 1.230769230769231e-05,
|
| 32 |
+
"loss": 2.0488,
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.15,
|
| 37 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 38 |
+
"loss": 1.806,
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.18,
|
| 43 |
+
"learning_rate": 1.8461538461538465e-05,
|
| 44 |
+
"loss": 2.3016,
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.21,
|
| 49 |
+
"learning_rate": 1.9999673886943734e-05,
|
| 50 |
+
"loss": 2.5816,
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.24,
|
| 55 |
+
"learning_rate": 1.9997065110111884e-05,
|
| 56 |
+
"loss": 3.187,
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.27,
|
| 61 |
+
"learning_rate": 1.9991848237042037e-05,
|
| 62 |
+
"loss": 2.1655,
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.3,
|
| 67 |
+
"learning_rate": 1.998402462874433e-05,
|
| 68 |
+
"loss": 1.9262,
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.33,
|
| 73 |
+
"learning_rate": 1.9973596326290136e-05,
|
| 74 |
+
"loss": 1.9535,
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.36,
|
| 79 |
+
"learning_rate": 1.9960566050279568e-05,
|
| 80 |
+
"loss": 1.9173,
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.39,
|
| 85 |
+
"learning_rate": 1.994493720013169e-05,
|
| 86 |
+
"loss": 1.9742,
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.42,
|
| 91 |
+
"learning_rate": 1.9926713853197696e-05,
|
| 92 |
+
"loss": 2.3098,
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.45,
|
| 97 |
+
"learning_rate": 1.9905900763697152e-05,
|
| 98 |
+
"loss": 2.6274,
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.48,
|
| 103 |
+
"learning_rate": 1.9882503361477707e-05,
|
| 104 |
+
"loss": 2.8624,
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.51,
|
| 109 |
+
"learning_rate": 1.9856527750598493e-05,
|
| 110 |
+
"loss": 2.5503,
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"epoch": 0.54,
|
| 115 |
+
"learning_rate": 1.9827980707737704e-05,
|
| 116 |
+
"loss": 2.179,
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 0.57,
|
| 121 |
+
"learning_rate": 1.979686968042461e-05,
|
| 122 |
+
"loss": 1.8671,
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 0.6,
|
| 127 |
+
"learning_rate": 1.976320278509663e-05,
|
| 128 |
+
"loss": 1.9073,
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.63,
|
| 133 |
+
"learning_rate": 1.9726988804981847e-05,
|
| 134 |
+
"loss": 1.7712,
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.66,
|
| 139 |
+
"learning_rate": 1.9688237187807594e-05,
|
| 140 |
+
"loss": 2.0556,
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
"epoch": 0.69,
|
| 145 |
+
"learning_rate": 1.9646958043335678e-05,
|
| 146 |
+
"loss": 2.3458,
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
"epoch": 0.72,
|
| 151 |
+
"learning_rate": 1.9603162140724863e-05,
|
| 152 |
+
"loss": 2.4738,
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.75,
|
| 157 |
+
"learning_rate": 1.9556860905721363e-05,
|
| 158 |
+
"loss": 2.671,
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.78,
|
| 163 |
+
"learning_rate": 1.950806641767802e-05,
|
| 164 |
+
"loss": 1.9429,
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.81,
|
| 169 |
+
"learning_rate": 1.9456791406402964e-05,
|
| 170 |
+
"loss": 1.8356,
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.84,
|
| 175 |
+
"learning_rate": 1.940304924883858e-05,
|
| 176 |
+
"loss": 1.8648,
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.86,
|
| 181 |
+
"learning_rate": 1.934685396557165e-05,
|
| 182 |
+
"loss": 1.8207,
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
"epoch": 0.89,
|
| 187 |
+
"learning_rate": 1.9288220217175583e-05,
|
| 188 |
+
"loss": 1.7519,
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.92,
|
| 193 |
+
"learning_rate": 1.9227163300385662e-05,
|
| 194 |
+
"loss": 1.9448,
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 0.95,
|
| 199 |
+
"learning_rate": 1.9163699144108343e-05,
|
| 200 |
+
"loss": 2.1599,
|
| 201 |
"step": 64
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 0.98,
|
| 205 |
+
"learning_rate": 1.9097844305265625e-05,
|
| 206 |
+
"loss": 1.7499,
|
| 207 |
"step": 66
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 1.01,
|
| 211 |
+
"learning_rate": 1.9029615964475572e-05,
|
| 212 |
+
"loss": 2.3244,
|
| 213 |
"step": 68
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 1.04,
|
| 217 |
+
"learning_rate": 1.8959031921570136e-05,
|
| 218 |
+
"loss": 2.0381,
|
| 219 |
"step": 70
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 1.07,
|
| 223 |
+
"learning_rate": 1.8886110590951417e-05,
|
| 224 |
+
"loss": 1.7528,
|
| 225 |
"step": 72
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 1.1,
|
| 229 |
+
"learning_rate": 1.88108709967876e-05,
|
| 230 |
+
"loss": 1.9127,
|
| 231 |
"step": 74
|
| 232 |
},
|
| 233 |
{
|
| 234 |
"epoch": 1.13,
|
| 235 |
+
"learning_rate": 1.873333276804983e-05,
|
| 236 |
+
"loss": 1.654,
|
| 237 |
"step": 76
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 1.16,
|
| 241 |
+
"learning_rate": 1.865351613339125e-05,
|
| 242 |
+
"loss": 1.642,
|
| 243 |
"step": 78
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 1.19,
|
| 247 |
+
"learning_rate": 1.8571441915869663e-05,
|
| 248 |
+
"loss": 1.9048,
|
| 249 |
"step": 80
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 1.22,
|
| 253 |
+
"learning_rate": 1.848713152751506e-05,
|
| 254 |
+
"loss": 2.04,
|
| 255 |
"step": 82
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.25,
|
| 259 |
+
"learning_rate": 1.8400606963743517e-05,
|
| 260 |
+
"loss": 2.168,
|
| 261 |
"step": 84
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 1.28,
|
| 265 |
+
"learning_rate": 1.8311890797618918e-05,
|
| 266 |
+
"loss": 2.0551,
|
| 267 |
"step": 86
|
| 268 |
},
|
| 269 |
{
|
| 270 |
"epoch": 1.31,
|
| 271 |
+
"learning_rate": 1.822100617396391e-05,
|
| 272 |
+
"loss": 1.8066,
|
| 273 |
"step": 88
|
| 274 |
},
|
| 275 |
{
|
| 276 |
"epoch": 1.34,
|
| 277 |
+
"learning_rate": 1.8127976803321793e-05,
|
| 278 |
+
"loss": 1.7716,
|
| 279 |
"step": 90
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 1.37,
|
| 283 |
+
"learning_rate": 1.8032826955770723e-05,
|
| 284 |
+
"loss": 1.7329,
|
| 285 |
"step": 92
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 1.4,
|
| 289 |
+
"learning_rate": 1.7935581454592005e-05,
|
| 290 |
+
"loss": 1.6597,
|
| 291 |
"step": 94
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 1.43,
|
| 295 |
+
"learning_rate": 1.7836265669794032e-05,
|
| 296 |
+
"loss": 1.7824,
|
| 297 |
"step": 96
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.46,
|
| 301 |
+
"learning_rate": 1.7734905511493614e-05,
|
| 302 |
+
"loss": 1.9783,
|
| 303 |
"step": 98
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 1.49,
|
| 307 |
+
"learning_rate": 1.763152742315637e-05,
|
| 308 |
+
"loss": 1.8963,
|
| 309 |
"step": 100
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 1.52,
|
| 313 |
+
"learning_rate": 1.7526158374697997e-05,
|
| 314 |
+
"loss": 2.5619,
|
| 315 |
"step": 102
|
| 316 |
},
|
| 317 |
{
|
| 318 |
"epoch": 1.55,
|
| 319 |
+
"learning_rate": 1.7418825855448208e-05,
|
| 320 |
+
"loss": 1.6946,
|
| 321 |
"step": 104
|
| 322 |
},
|
| 323 |
{
|
| 324 |
"epoch": 1.58,
|
| 325 |
+
"learning_rate": 1.7309557866979113e-05,
|
| 326 |
+
"loss": 1.7272,
|
| 327 |
"step": 106
|
| 328 |
},
|
| 329 |
{
|
| 330 |
"epoch": 1.61,
|
| 331 |
+
"learning_rate": 1.7198382915800034e-05,
|
| 332 |
+
"loss": 1.6596,
|
| 333 |
"step": 108
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 1.64,
|
| 337 |
+
"learning_rate": 1.7085330005920516e-05,
|
| 338 |
+
"loss": 1.5296,
|
| 339 |
"step": 110
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.67,
|
| 343 |
+
"learning_rate": 1.6970428631283602e-05,
|
| 344 |
+
"loss": 1.7269,
|
| 345 |
"step": 112
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 1.7,
|
| 349 |
+
"learning_rate": 1.6853708768071265e-05,
|
| 350 |
+
"loss": 2.0672,
|
| 351 |
"step": 114
|
| 352 |
},
|
| 353 |
{
|
| 354 |
"epoch": 1.73,
|
| 355 |
+
"learning_rate": 1.6735200866884037e-05,
|
| 356 |
+
"loss": 2.3257,
|
| 357 |
"step": 116
|
| 358 |
},
|
| 359 |
{
|
| 360 |
"epoch": 1.76,
|
| 361 |
+
"learning_rate": 1.6614935844796863e-05,
|
| 362 |
+
"loss": 2.3116,
|
| 363 |
"step": 118
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 1.79,
|
| 367 |
+
"learning_rate": 1.649294507729327e-05,
|
| 368 |
+
"loss": 1.6772,
|
| 369 |
"step": 120
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 1.82,
|
| 373 |
+
"learning_rate": 1.6369260390079933e-05,
|
| 374 |
+
"loss": 1.7853,
|
| 375 |
"step": 122
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 1.85,
|
| 379 |
+
"learning_rate": 1.6243914050783783e-05,
|
| 380 |
+
"loss": 1.6112,
|
| 381 |
"step": 124
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.88,
|
| 385 |
+
"learning_rate": 1.6116938760533843e-05,
|
| 386 |
+
"loss": 1.5548,
|
| 387 |
"step": 126
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 1.91,
|
| 391 |
+
"learning_rate": 1.5988367645429938e-05,
|
| 392 |
+
"loss": 1.8593,
|
| 393 |
"step": 128
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 1.94,
|
| 397 |
+
"learning_rate": 1.585823424790056e-05,
|
| 398 |
+
"loss": 2.0341,
|
| 399 |
"step": 130
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 1.97,
|
| 403 |
+
"learning_rate": 1.5726572517952122e-05,
|
| 404 |
+
"loss": 1.5791,
|
| 405 |
"step": 132
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 2.0,
|
| 409 |
+
"learning_rate": 1.559341680431185e-05,
|
| 410 |
+
"loss": 1.7591,
|
| 411 |
"step": 134
|
| 412 |
},
|
| 413 |
{
|
| 414 |
"epoch": 2.03,
|
| 415 |
+
"learning_rate": 1.545880184546669e-05,
|
| 416 |
+
"loss": 2.2547,
|
| 417 |
"step": 136
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 2.06,
|
| 421 |
+
"learning_rate": 1.532276276060051e-05,
|
| 422 |
+
"loss": 1.8692,
|
| 423 |
"step": 138
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 2.09,
|
| 427 |
+
"learning_rate": 1.518533504043199e-05,
|
| 428 |
+
"loss": 1.7114,
|
| 429 |
"step": 140
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 2.12,
|
| 433 |
+
"learning_rate": 1.5046554537955587e-05,
|
| 434 |
+
"loss": 1.6847,
|
| 435 |
"step": 142
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 2.15,
|
| 439 |
+
"learning_rate": 1.4906457459087977e-05,
|
| 440 |
+
"loss": 1.6882,
|
| 441 |
"step": 144
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 2.18,
|
| 445 |
+
"learning_rate": 1.4765080353222447e-05,
|
| 446 |
+
"loss": 1.6619,
|
| 447 |
"step": 146
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 2.21,
|
| 451 |
+
"learning_rate": 1.462246010369364e-05,
|
| 452 |
+
"loss": 1.7873,
|
| 453 |
"step": 148
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 2.24,
|
| 457 |
+
"learning_rate": 1.4478633918155216e-05,
|
| 458 |
+
"loss": 1.6883,
|
| 459 |
"step": 150
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 2.27,
|
| 463 |
+
"learning_rate": 1.4333639318872891e-05,
|
| 464 |
+
"loss": 2.4291,
|
| 465 |
"step": 152
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 2.3,
|
| 469 |
+
"learning_rate": 1.4187514132935393e-05,
|
| 470 |
+
"loss": 2.2567,
|
| 471 |
"step": 154
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 2.33,
|
| 475 |
+
"learning_rate": 1.4040296482385893e-05,
|
| 476 |
+
"loss": 1.6691,
|
| 477 |
"step": 156
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 2.36,
|
| 481 |
+
"learning_rate": 1.3892024774276496e-05,
|
| 482 |
+
"loss": 1.7297,
|
| 483 |
"step": 158
|
| 484 |
},
|
| 485 |
{
|
| 486 |
"epoch": 2.39,
|
| 487 |
+
"learning_rate": 1.3742737690648362e-05,
|
| 488 |
+
"loss": 1.5559,
|
| 489 |
"step": 160
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 2.42,
|
| 493 |
+
"learning_rate": 1.3592474178440116e-05,
|
| 494 |
+
"loss": 1.6777,
|
| 495 |
"step": 162
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 2.45,
|
| 499 |
+
"learning_rate": 1.34412734393271e-05,
|
| 500 |
+
"loss": 1.8319,
|
| 501 |
"step": 164
|
| 502 |
},
|
| 503 |
{
|
| 504 |
"epoch": 2.48,
|
| 505 |
+
"learning_rate": 1.3289174919494228e-05,
|
| 506 |
+
"loss": 1.8486,
|
| 507 |
"step": 166
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 2.51,
|
| 511 |
+
"learning_rate": 1.3136218299344993e-05,
|
| 512 |
+
"loss": 2.2382,
|
| 513 |
"step": 168
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 2.53,
|
| 517 |
+
"learning_rate": 1.2982443483149423e-05,
|
| 518 |
+
"loss": 1.6871,
|
| 519 |
"step": 170
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 2.56,
|
| 523 |
+
"learning_rate": 1.2827890588633589e-05,
|
| 524 |
+
"loss": 1.8091,
|
| 525 |
"step": 172
|
| 526 |
},
|
| 527 |
{
|
| 528 |
"epoch": 2.59,
|
| 529 |
+
"learning_rate": 1.267259993651345e-05,
|
| 530 |
+
"loss": 1.7956,
|
| 531 |
"step": 174
|
| 532 |
},
|
| 533 |
{
|
| 534 |
"epoch": 2.62,
|
| 535 |
+
"learning_rate": 1.2516612039975745e-05,
|
| 536 |
+
"loss": 1.5036,
|
| 537 |
"step": 176
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 2.65,
|
| 541 |
+
"learning_rate": 1.2359967594108643e-05,
|
| 542 |
+
"loss": 1.5951,
|
| 543 |
"step": 178
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 2.68,
|
| 547 |
+
"learning_rate": 1.2202707465284973e-05,
|
| 548 |
+
"loss": 1.8153,
|
| 549 |
"step": 180
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 2.71,
|
| 553 |
+
"learning_rate": 1.2044872680500743e-05,
|
| 554 |
+
"loss": 1.9109,
|
| 555 |
"step": 182
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 2.74,
|
| 559 |
+
"learning_rate": 1.188650441667177e-05,
|
| 560 |
+
"loss": 2.602,
|
| 561 |
"step": 184
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 2.77,
|
| 565 |
+
"learning_rate": 1.172764398989118e-05,
|
| 566 |
+
"loss": 1.7423,
|
| 567 |
"step": 186
|
| 568 |
},
|
| 569 |
{
|
| 570 |
"epoch": 2.8,
|
| 571 |
+
"learning_rate": 1.1568332844650623e-05,
|
| 572 |
+
"loss": 1.7171,
|
| 573 |
"step": 188
|
| 574 |
},
|
| 575 |
{
|
| 576 |
"epoch": 2.83,
|
| 577 |
+
"learning_rate": 1.1408612543027963e-05,
|
| 578 |
+
"loss": 1.7039,
|
| 579 |
"step": 190
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 2.86,
|
| 583 |
+
"learning_rate": 1.1248524753844325e-05,
|
| 584 |
+
"loss": 1.3292,
|
| 585 |
"step": 192
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 2.89,
|
| 589 |
+
"learning_rate": 1.1088111241793258e-05,
|
| 590 |
+
"loss": 1.7061,
|
| 591 |
"step": 194
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 2.92,
|
| 595 |
+
"learning_rate": 1.0927413856544906e-05,
|
| 596 |
+
"loss": 1.9157,
|
| 597 |
"step": 196
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 2.95,
|
| 601 |
+
"learning_rate": 1.0766474521828022e-05,
|
| 602 |
+
"loss": 1.8149,
|
| 603 |
"step": 198
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 2.98,
|
| 607 |
+
"learning_rate": 1.0605335224492617e-05,
|
| 608 |
+
"loss": 1.4948,
|
| 609 |
"step": 200
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 3.01,
|
| 613 |
+
"learning_rate": 1.0444038003556201e-05,
|
| 614 |
+
"loss": 1.8947,
|
| 615 |
+
"step": 202
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 3.04,
|
| 619 |
+
"learning_rate": 1.0282624939236367e-05,
|
| 620 |
+
"loss": 2.2848,
|
| 621 |
+
"step": 204
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 3.07,
|
| 625 |
+
"learning_rate": 1.0121138141972649e-05,
|
| 626 |
+
"loss": 1.6915,
|
| 627 |
+
"step": 206
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 3.1,
|
| 631 |
+
"learning_rate": 9.959619741440486e-06,
|
| 632 |
+
"loss": 1.6911,
|
| 633 |
+
"step": 208
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 3.13,
|
| 637 |
+
"learning_rate": 9.798111875560167e-06,
|
| 638 |
+
"loss": 1.6485,
|
| 639 |
+
"step": 210
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.16,
|
| 643 |
+
"learning_rate": 9.636656679503647e-06,
|
| 644 |
+
"loss": 1.4867,
|
| 645 |
+
"step": 212
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 3.19,
|
| 649 |
+
"learning_rate": 9.475296274702044e-06,
|
| 650 |
+
"loss": 1.7333,
|
| 651 |
+
"step": 214
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 3.22,
|
| 655 |
+
"learning_rate": 9.314072757856752e-06,
|
| 656 |
+
"loss": 1.9008,
|
| 657 |
+
"step": 216
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 3.25,
|
| 661 |
+
"learning_rate": 9.153028189956986e-06,
|
| 662 |
+
"loss": 1.8957,
|
| 663 |
+
"step": 218
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 3.28,
|
| 667 |
+
"learning_rate": 8.99220458530664e-06,
|
| 668 |
+
"loss": 2.1675,
|
| 669 |
+
"step": 220
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 3.31,
|
| 673 |
+
"learning_rate": 8.831643900563372e-06,
|
| 674 |
+
"loss": 1.6982,
|
| 675 |
+
"step": 222
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 3.34,
|
| 679 |
+
"learning_rate": 8.671388023792642e-06,
|
| 680 |
+
"loss": 1.7398,
|
| 681 |
+
"step": 224
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.37,
|
| 685 |
+
"learning_rate": 8.511478763539737e-06,
|
| 686 |
+
"loss": 1.9094,
|
| 687 |
+
"step": 226
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.4,
|
| 691 |
+
"learning_rate": 8.351957837922467e-06,
|
| 692 |
+
"loss": 1.4749,
|
| 693 |
+
"step": 228
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 3.43,
|
| 697 |
+
"learning_rate": 8.192866863747516e-06,
|
| 698 |
+
"loss": 1.6732,
|
| 699 |
+
"step": 230
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 3.46,
|
| 703 |
+
"learning_rate": 8.034247345653148e-06,
|
| 704 |
+
"loss": 1.8966,
|
| 705 |
+
"step": 232
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 3.49,
|
| 709 |
+
"learning_rate": 7.876140665281273e-06,
|
| 710 |
+
"loss": 1.9921,
|
| 711 |
+
"step": 234
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 3.52,
|
| 715 |
+
"learning_rate": 7.718588070481501e-06,
|
| 716 |
+
"loss": 1.9,
|
| 717 |
+
"step": 236
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 3.55,
|
| 721 |
+
"learning_rate": 7.561630664550179e-06,
|
| 722 |
+
"loss": 1.6396,
|
| 723 |
+
"step": 238
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.58,
|
| 727 |
+
"learning_rate": 7.405309395507098e-06,
|
| 728 |
+
"loss": 1.672,
|
| 729 |
+
"step": 240
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.61,
|
| 733 |
+
"learning_rate": 7.249665045412704e-06,
|
| 734 |
+
"loss": 1.6482,
|
| 735 |
+
"step": 242
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 3.64,
|
| 739 |
+
"learning_rate": 7.0947382197286566e-06,
|
| 740 |
+
"loss": 1.5786,
|
| 741 |
+
"step": 244
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 3.67,
|
| 745 |
+
"learning_rate": 6.94056933672439e-06,
|
| 746 |
+
"loss": 1.7509,
|
| 747 |
+
"step": 246
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 3.7,
|
| 751 |
+
"learning_rate": 6.787198616932571e-06,
|
| 752 |
+
"loss": 1.803,
|
| 753 |
+
"step": 248
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 3.73,
|
| 757 |
+
"learning_rate": 6.634666072656097e-06,
|
| 758 |
+
"loss": 1.5227,
|
| 759 |
+
"step": 250
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 3.76,
|
| 763 |
+
"learning_rate": 6.483011497529457e-06,
|
| 764 |
+
"loss": 2.5859,
|
| 765 |
+
"step": 252
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.79,
|
| 769 |
+
"learning_rate": 6.332274456137097e-06,
|
| 770 |
+
"loss": 1.7433,
|
| 771 |
+
"step": 254
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.82,
|
| 775 |
+
"learning_rate": 6.182494273691602e-06,
|
| 776 |
+
"loss": 1.6223,
|
| 777 |
+
"step": 256
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 3.85,
|
| 781 |
+
"learning_rate": 6.033710025774253e-06,
|
| 782 |
+
"loss": 1.5475,
|
| 783 |
+
"step": 258
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 3.88,
|
| 787 |
+
"learning_rate": 5.885960528140784e-06,
|
| 788 |
+
"loss": 1.3609,
|
| 789 |
+
"step": 260
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 3.91,
|
| 793 |
+
"learning_rate": 5.739284326594845e-06,
|
| 794 |
+
"loss": 1.6793,
|
| 795 |
+
"step": 262
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 3.94,
|
| 799 |
+
"learning_rate": 5.59371968693198e-06,
|
| 800 |
+
"loss": 1.8558,
|
| 801 |
+
"step": 264
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 3.97,
|
| 805 |
+
"learning_rate": 5.449304584956582e-06,
|
| 806 |
+
"loss": 1.8424,
|
| 807 |
+
"step": 266
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.0,
|
| 811 |
+
"learning_rate": 5.306076696574522e-06,
|
| 812 |
+
"loss": 1.672,
|
| 813 |
+
"step": 268
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 4.03,
|
| 817 |
+
"learning_rate": 5.164073387964057e-06,
|
| 818 |
+
"loss": 2.6413,
|
| 819 |
+
"step": 270
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 4.06,
|
| 823 |
+
"learning_rate": 5.023331705827477e-06,
|
| 824 |
+
"loss": 1.8991,
|
| 825 |
+
"step": 272
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 4.09,
|
| 829 |
+
"learning_rate": 4.883888367726153e-06,
|
| 830 |
+
"loss": 1.7208,
|
| 831 |
+
"step": 274
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 4.12,
|
| 835 |
+
"learning_rate": 4.74577975250143e-06,
|
| 836 |
+
"loss": 1.6942,
|
| 837 |
+
"step": 276
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 4.15,
|
| 841 |
+
"learning_rate": 4.609041890783882e-06,
|
| 842 |
+
"loss": 1.5046,
|
| 843 |
+
"step": 278
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.18,
|
| 847 |
+
"learning_rate": 4.473710455593416e-06,
|
| 848 |
+
"loss": 1.484,
|
| 849 |
+
"step": 280
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.21,
|
| 853 |
+
"learning_rate": 4.339820753032692e-06,
|
| 854 |
+
"loss": 1.7197,
|
| 855 |
+
"step": 282
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 4.23,
|
| 859 |
+
"learning_rate": 4.207407713076221e-06,
|
| 860 |
+
"loss": 1.9587,
|
| 861 |
+
"step": 284
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 4.26,
|
| 865 |
+
"learning_rate": 4.076505880457642e-06,
|
| 866 |
+
"loss": 2.2569,
|
| 867 |
+
"step": 286
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 4.29,
|
| 871 |
+
"learning_rate": 3.947149405657469e-06,
|
| 872 |
+
"loss": 1.8487,
|
| 873 |
+
"step": 288
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 4.32,
|
| 877 |
+
"learning_rate": 3.8193720359936905e-06,
|
| 878 |
+
"loss": 1.7394,
|
| 879 |
+
"step": 290
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 4.35,
|
| 883 |
+
"learning_rate": 3.69320710681758e-06,
|
| 884 |
+
"loss": 1.7156,
|
| 885 |
+
"step": 292
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.38,
|
| 889 |
+
"learning_rate": 3.5686875328169513e-06,
|
| 890 |
+
"loss": 1.3529,
|
| 891 |
+
"step": 294
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.41,
|
| 895 |
+
"learning_rate": 3.4458457994291763e-06,
|
| 896 |
+
"loss": 1.6026,
|
| 897 |
+
"step": 296
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 4.44,
|
| 901 |
+
"learning_rate": 3.324713954366171e-06,
|
| 902 |
+
"loss": 1.8169,
|
| 903 |
+
"step": 298
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 4.47,
|
| 907 |
+
"learning_rate": 3.2053235992536137e-06,
|
| 908 |
+
"loss": 1.5503,
|
| 909 |
+
"step": 300
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 4.5,
|
| 913 |
+
"learning_rate": 3.0877058813864856e-06,
|
| 914 |
+
"loss": 2.1322,
|
| 915 |
+
"step": 302
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.53,
|
| 919 |
+
"learning_rate": 2.9718914856032033e-06,
|
| 920 |
+
"loss": 1.6235,
|
| 921 |
+
"step": 304
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 4.56,
|
| 925 |
+
"learning_rate": 2.8579106262803467e-06,
|
| 926 |
+
"loss": 1.7218,
|
| 927 |
+
"step": 306
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.59,
|
| 931 |
+
"learning_rate": 2.7457930394501564e-06,
|
| 932 |
+
"loss": 1.6783,
|
| 933 |
+
"step": 308
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.62,
|
| 937 |
+
"learning_rate": 2.635567975042809e-06,
|
| 938 |
+
"loss": 1.6301,
|
| 939 |
+
"step": 310
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 4.65,
|
| 943 |
+
"learning_rate": 2.527264189255507e-06,
|
| 944 |
+
"loss": 1.6768,
|
| 945 |
+
"step": 312
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 4.68,
|
| 949 |
+
"learning_rate": 2.420909937050405e-06,
|
| 950 |
+
"loss": 1.7345,
|
| 951 |
+
"step": 314
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 4.71,
|
| 955 |
+
"learning_rate": 2.3165329647832525e-06,
|
| 956 |
+
"loss": 1.8255,
|
| 957 |
+
"step": 316
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 4.74,
|
| 961 |
+
"learning_rate": 2.214160502964783e-06,
|
| 962 |
+
"loss": 2.1617,
|
| 963 |
+
"step": 318
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 4.77,
|
| 967 |
+
"learning_rate": 2.1138192591566177e-06,
|
| 968 |
+
"loss": 2.0365,
|
| 969 |
+
"step": 320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.8,
|
| 973 |
+
"learning_rate": 2.0155354110036607e-06,
|
| 974 |
+
"loss": 1.7143,
|
| 975 |
+
"step": 322
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.83,
|
| 979 |
+
"learning_rate": 1.9193345994046965e-06,
|
| 980 |
+
"loss": 1.7533,
|
| 981 |
+
"step": 324
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 4.86,
|
| 985 |
+
"learning_rate": 1.8252419218230389e-06,
|
| 986 |
+
"loss": 1.6113,
|
| 987 |
+
"step": 326
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 4.89,
|
| 991 |
+
"learning_rate": 1.7332819257389388e-06,
|
| 992 |
+
"loss": 1.3852,
|
| 993 |
+
"step": 328
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 4.92,
|
| 997 |
+
"learning_rate": 1.6434786022455073e-06,
|
| 998 |
+
"loss": 1.6639,
|
| 999 |
+
"step": 330
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.95,
|
| 1003 |
+
"learning_rate": 1.5558553797897469e-06,
|
| 1004 |
+
"loss": 1.6325,
|
| 1005 |
+
"step": 332
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 4.98,
|
| 1009 |
+
"learning_rate": 1.4704351180604126e-06,
|
| 1010 |
+
"loss": 1.7177,
|
| 1011 |
+
"step": 334
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 5.01,
|
| 1015 |
+
"learning_rate": 1.3872401020242222e-06,
|
| 1016 |
+
"loss": 1.9718,
|
| 1017 |
+
"step": 336
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 5.04,
|
| 1021 |
+
"learning_rate": 1.3062920361120224e-06,
|
| 1022 |
+
"loss": 2.3978,
|
| 1023 |
+
"step": 338
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 5.07,
|
| 1027 |
+
"learning_rate": 1.2276120385564006e-06,
|
| 1028 |
+
"loss": 1.6484,
|
| 1029 |
+
"step": 340
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 5.1,
|
| 1033 |
+
"learning_rate": 1.1512206358822264e-06,
|
| 1034 |
+
"loss": 1.6426,
|
| 1035 |
+
"step": 342
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 5.13,
|
| 1039 |
+
"learning_rate": 1.077137757551573e-06,
|
| 1040 |
+
"loss": 1.5856,
|
| 1041 |
+
"step": 344
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 5.16,
|
| 1045 |
+
"learning_rate": 1.005382730764386e-06,
|
| 1046 |
+
"loss": 1.5543,
|
| 1047 |
+
"step": 346
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 5.19,
|
| 1051 |
+
"learning_rate": 9.359742754162926e-07,
|
| 1052 |
+
"loss": 1.5946,
|
| 1053 |
+
"step": 348
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 5.22,
|
| 1057 |
+
"learning_rate": 8.689304992148285e-07,
|
| 1058 |
+
"loss": 1.6595,
|
| 1059 |
+
"step": 350
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.25,
|
| 1063 |
+
"learning_rate": 8.042688929554076e-07,
|
| 1064 |
+
"loss": 1.8412,
|
| 1065 |
+
"step": 352
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 5.28,
|
| 1069 |
+
"learning_rate": 7.420063259581856e-07,
|
| 1070 |
+
"loss": 2.3088,
|
| 1071 |
+
"step": 354
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 5.31,
|
| 1075 |
+
"learning_rate": 6.821590416671108e-07,
|
| 1076 |
+
"loss": 1.8214,
|
| 1077 |
+
"step": 356
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 5.34,
|
| 1081 |
+
"learning_rate": 6.247426534122292e-07,
|
| 1082 |
+
"loss": 1.8167,
|
| 1083 |
+
"step": 358
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 5.37,
|
| 1087 |
+
"learning_rate": 5.697721403363699e-07,
|
| 1088 |
+
"loss": 1.5789,
|
| 1089 |
+
"step": 360
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 5.4,
|
| 1093 |
+
"learning_rate": 5.172618434873112e-07,
|
| 1094 |
+
"loss": 1.7484,
|
| 1095 |
+
"step": 362
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 5.43,
|
| 1099 |
+
"learning_rate": 4.672254620763839e-07,
|
| 1100 |
+
"loss": 1.728,
|
| 1101 |
+
"step": 364
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.46,
|
| 1105 |
+
"learning_rate": 4.196760499045505e-07,
|
| 1106 |
+
"loss": 1.9324,
|
| 1107 |
+
"step": 366
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 5.49,
|
| 1111 |
+
"learning_rate": 3.746260119568368e-07,
|
| 1112 |
+
"loss": 1.4902,
|
| 1113 |
+
"step": 368
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 5.52,
|
| 1117 |
+
"learning_rate": 3.320871011660498e-07,
|
| 1118 |
+
"loss": 2.3606,
|
| 1119 |
+
"step": 370
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 5.55,
|
| 1123 |
+
"learning_rate": 2.920704153465936e-07,
|
| 1124 |
+
"loss": 1.7926,
|
| 1125 |
+
"step": 372
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 5.58,
|
| 1129 |
+
"learning_rate": 2.5458639429921105e-07,
|
| 1130 |
+
"loss": 1.8292,
|
| 1131 |
+
"step": 374
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 5.61,
|
| 1135 |
+
"learning_rate": 2.196448170873755e-07,
|
| 1136 |
+
"loss": 1.6128,
|
| 1137 |
+
"step": 376
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 5.64,
|
| 1141 |
+
"learning_rate": 1.8725479948607515e-07,
|
| 1142 |
+
"loss": 1.3591,
|
| 1143 |
+
"step": 378
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.67,
|
| 1147 |
+
"learning_rate": 1.5742479160362978e-07,
|
| 1148 |
+
"loss": 1.4791,
|
| 1149 |
+
"step": 380
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 5.7,
|
| 1153 |
+
"learning_rate": 1.3016257567717295e-07,
|
| 1154 |
+
"loss": 1.8504,
|
| 1155 |
+
"step": 382
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 5.73,
|
| 1159 |
+
"learning_rate": 1.054752640423784e-07,
|
| 1160 |
+
"loss": 1.7373,
|
| 1161 |
+
"step": 384
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 5.76,
|
| 1165 |
+
"learning_rate": 8.336929727794318e-08,
|
| 1166 |
+
"loss": 2.1627,
|
| 1167 |
+
"step": 386
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 5.79,
|
| 1171 |
+
"learning_rate": 6.385044252533723e-08,
|
| 1172 |
+
"loss": 1.6952,
|
| 1173 |
+
"step": 388
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 5.82,
|
| 1177 |
+
"learning_rate": 4.692379198422803e-08,
|
| 1178 |
+
"loss": 1.7199,
|
| 1179 |
+
"step": 390
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 5.85,
|
| 1183 |
+
"learning_rate": 3.259376158400329e-08,
|
| 1184 |
+
"loss": 1.7157,
|
| 1185 |
+
"step": 392
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.88,
|
| 1189 |
+
"learning_rate": 2.0864089831711398e-08,
|
| 1190 |
+
"loss": 1.4137,
|
| 1191 |
+
"step": 394
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 5.9,
|
| 1195 |
+
"learning_rate": 1.1737836836737126e-08,
|
| 1196 |
+
"loss": 1.6499,
|
| 1197 |
+
"step": 396
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 5.93,
|
| 1201 |
+
"learning_rate": 5.217383512463592e-09,
|
| 1202 |
+
"loss": 1.7672,
|
| 1203 |
+
"step": 398
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 5.96,
|
| 1207 |
+
"learning_rate": 1.3044309551213385e-09,
|
| 1208 |
+
"loss": 1.5814,
|
| 1209 |
+
"step": 400
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 5.99,
|
| 1213 |
+
"learning_rate": 0.0,
|
| 1214 |
+
"loss": 1.6152,
|
| 1215 |
+
"step": 402
|
| 1216 |
}
|
| 1217 |
],
|
| 1218 |
"logging_steps": 2,
|
| 1219 |
+
"max_steps": 402,
|
| 1220 |
+
"num_train_epochs": 6,
|
| 1221 |
"save_steps": 500,
|
| 1222 |
+
"total_flos": 1.7781494793805824e+16,
|
| 1223 |
"trial_name": null,
|
| 1224 |
"trial_params": null
|
| 1225 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4600
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:239e6ed22c6da37e9830926f0efad13bc0fcad9d2c09e0f0964ed8fccabea5b1
|
| 3 |
size 4600
|