Upload folder using huggingface_hub
Browse files- adapter_config.json +1 -1
- adapter_model.bin +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +347 -197
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"auto_mapping": null,
|
| 3 |
-
"base_model_name_or_path": "TinyPixel/redpajama-3b-mix-
|
| 4 |
"bias": "none",
|
| 5 |
"fan_in_fan_out": false,
|
| 6 |
"inference_mode": true,
|
|
|
|
| 1 |
{
|
| 2 |
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "TinyPixel/redpajama-3b-mix-3",
|
| 4 |
"bias": "none",
|
| 5 |
"fan_in_fan_out": false,
|
| 6 |
"inference_mode": true,
|
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 335638221
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:139648f0b904e6bc4dab4277fdc88959126a75a256de27c039d49a3298e1262d
|
| 3 |
size 335638221
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 671246597
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:29f39c191399ca0fc8741dc52c3bc4003709976a180e44c71ac57df5a55ad989
|
| 3 |
size 671246597
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2511c0bbee94e55e4befeae0ccbcdffdf8672140610222080cb25ce9c8e69f6
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:daa1e3d9bed75ca5ad2539cb47d1d2a53221eb8cf13757db09a3fb8aa9c38c60
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -10,607 +10,757 @@
|
|
| 10 |
{
|
| 11 |
"epoch": 0.0,
|
| 12 |
"learning_rate": 2e-05,
|
| 13 |
-
"loss":
|
| 14 |
"step": 2
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.0,
|
| 18 |
"learning_rate": 2e-05,
|
| 19 |
-
"loss": 1.
|
| 20 |
"step": 4
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0.
|
| 24 |
"learning_rate": 2e-05,
|
| 25 |
-
"loss": 1.
|
| 26 |
"step": 6
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.01,
|
| 30 |
"learning_rate": 2e-05,
|
| 31 |
-
"loss": 1.
|
| 32 |
"step": 8
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.01,
|
| 36 |
"learning_rate": 2e-05,
|
| 37 |
-
"loss": 1.
|
| 38 |
"step": 10
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.01,
|
| 42 |
"learning_rate": 2e-05,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 12
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.01,
|
| 48 |
"learning_rate": 2e-05,
|
| 49 |
-
"loss": 1.
|
| 50 |
"step": 14
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0.
|
| 54 |
"learning_rate": 2e-05,
|
| 55 |
-
"loss":
|
| 56 |
"step": 16
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0.
|
| 60 |
"learning_rate": 2e-05,
|
| 61 |
-
"loss":
|
| 62 |
"step": 18
|
| 63 |
},
|
| 64 |
{
|
| 65 |
-
"epoch": 0.
|
| 66 |
"learning_rate": 2e-05,
|
| 67 |
-
"loss":
|
| 68 |
"step": 20
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.02,
|
| 72 |
"learning_rate": 2e-05,
|
| 73 |
-
"loss":
|
| 74 |
"step": 22
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 0.
|
| 78 |
"learning_rate": 2e-05,
|
| 79 |
-
"loss":
|
| 80 |
"step": 24
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
"learning_rate": 2e-05,
|
| 85 |
-
"loss":
|
| 86 |
"step": 26
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
"learning_rate": 2e-05,
|
| 91 |
-
"loss":
|
| 92 |
"step": 28
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
"learning_rate": 2e-05,
|
| 97 |
-
"loss":
|
| 98 |
"step": 30
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch": 0.
|
| 102 |
"learning_rate": 2e-05,
|
| 103 |
-
"loss":
|
| 104 |
"step": 32
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch": 0.
|
| 108 |
"learning_rate": 2e-05,
|
| 109 |
-
"loss":
|
| 110 |
"step": 34
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch": 0.
|
| 114 |
"learning_rate": 2e-05,
|
| 115 |
-
"loss":
|
| 116 |
"step": 36
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 0.
|
| 120 |
"learning_rate": 2e-05,
|
| 121 |
-
"loss":
|
| 122 |
"step": 38
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
"learning_rate": 2e-05,
|
| 127 |
-
"loss":
|
| 128 |
"step": 40
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
"learning_rate": 2e-05,
|
| 133 |
-
"loss": 1.
|
| 134 |
"step": 42
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch": 0.
|
| 138 |
"learning_rate": 2e-05,
|
| 139 |
-
"loss": 1.
|
| 140 |
"step": 44
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch": 0.
|
| 144 |
"learning_rate": 2e-05,
|
| 145 |
-
"loss": 1.
|
| 146 |
"step": 46
|
| 147 |
},
|
| 148 |
{
|
| 149 |
-
"epoch": 0.
|
| 150 |
"learning_rate": 2e-05,
|
| 151 |
-
"loss": 1.
|
| 152 |
"step": 48
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch": 0.
|
| 156 |
"learning_rate": 2e-05,
|
| 157 |
-
"loss": 1.
|
| 158 |
"step": 50
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch": 0.
|
| 162 |
"learning_rate": 2e-05,
|
| 163 |
-
"loss": 1.
|
| 164 |
"step": 52
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
"learning_rate": 2e-05,
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 54
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
"learning_rate": 2e-05,
|
| 175 |
-
"loss": 1.
|
| 176 |
"step": 56
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch": 0.
|
| 180 |
"learning_rate": 2e-05,
|
| 181 |
-
"loss": 1.
|
| 182 |
"step": 58
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch": 0.
|
| 186 |
"learning_rate": 2e-05,
|
| 187 |
-
"loss": 1.
|
| 188 |
"step": 60
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch": 0.
|
| 192 |
"learning_rate": 2e-05,
|
| 193 |
-
"loss": 1.
|
| 194 |
"step": 62
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch": 0.
|
| 198 |
"learning_rate": 2e-05,
|
| 199 |
-
"loss": 1.
|
| 200 |
"step": 64
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch": 0.
|
| 204 |
"learning_rate": 2e-05,
|
| 205 |
-
"loss": 1.
|
| 206 |
"step": 66
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
"learning_rate": 2e-05,
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 68
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
"learning_rate": 2e-05,
|
| 217 |
-
"loss": 1.
|
| 218 |
"step": 70
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch": 0.
|
| 222 |
"learning_rate": 2e-05,
|
| 223 |
-
"loss":
|
| 224 |
"step": 72
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"epoch": 0.
|
| 228 |
"learning_rate": 2e-05,
|
| 229 |
-
"loss": 1.
|
| 230 |
"step": 74
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch": 0.
|
| 234 |
"learning_rate": 2e-05,
|
| 235 |
-
"loss":
|
| 236 |
"step": 76
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch": 0.
|
| 240 |
"learning_rate": 2e-05,
|
| 241 |
-
"loss":
|
| 242 |
"step": 78
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch": 0.
|
| 246 |
"learning_rate": 2e-05,
|
| 247 |
-
"loss": 1.
|
| 248 |
"step": 80
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
"learning_rate": 2e-05,
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 82
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
"learning_rate": 2e-05,
|
| 259 |
-
"loss":
|
| 260 |
"step": 84
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"epoch": 0.
|
| 264 |
"learning_rate": 2e-05,
|
| 265 |
-
"loss":
|
| 266 |
"step": 86
|
| 267 |
},
|
| 268 |
{
|
| 269 |
-
"epoch": 0.
|
| 270 |
"learning_rate": 2e-05,
|
| 271 |
-
"loss": 1.
|
| 272 |
"step": 88
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch": 0.
|
| 276 |
"learning_rate": 2e-05,
|
| 277 |
-
"loss":
|
| 278 |
"step": 90
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch": 0.
|
| 282 |
"learning_rate": 2e-05,
|
| 283 |
-
"loss": 1.
|
| 284 |
"step": 92
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch": 0.
|
| 288 |
"learning_rate": 2e-05,
|
| 289 |
-
"loss": 1.
|
| 290 |
"step": 94
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 0.
|
| 294 |
"learning_rate": 2e-05,
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 96
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
"learning_rate": 2e-05,
|
| 301 |
-
"loss": 1.
|
| 302 |
"step": 98
|
| 303 |
},
|
| 304 |
{
|
| 305 |
-
"epoch": 0.
|
| 306 |
"learning_rate": 2e-05,
|
| 307 |
-
"loss": 1.
|
| 308 |
"step": 100
|
| 309 |
},
|
| 310 |
{
|
| 311 |
-
"epoch": 0.
|
| 312 |
"learning_rate": 2e-05,
|
| 313 |
-
"loss": 1.
|
| 314 |
"step": 102
|
| 315 |
},
|
| 316 |
{
|
| 317 |
-
"epoch": 0.
|
| 318 |
"learning_rate": 2e-05,
|
| 319 |
-
"loss": 1.
|
| 320 |
"step": 104
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"epoch": 0.
|
| 324 |
"learning_rate": 2e-05,
|
| 325 |
-
"loss": 1.
|
| 326 |
"step": 106
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch": 0.
|
| 330 |
"learning_rate": 2e-05,
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 108
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
"learning_rate": 2e-05,
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 110
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
"learning_rate": 2e-05,
|
| 343 |
-
"loss": 1.
|
| 344 |
"step": 112
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"epoch": 0.
|
| 348 |
"learning_rate": 2e-05,
|
| 349 |
-
"loss": 1.
|
| 350 |
"step": 114
|
| 351 |
},
|
| 352 |
{
|
| 353 |
-
"epoch": 0.
|
| 354 |
"learning_rate": 2e-05,
|
| 355 |
-
"loss": 1.
|
| 356 |
"step": 116
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch": 0.
|
| 360 |
"learning_rate": 2e-05,
|
| 361 |
-
"loss": 1.
|
| 362 |
"step": 118
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"epoch": 0.
|
| 366 |
"learning_rate": 2e-05,
|
| 367 |
-
"loss": 1.
|
| 368 |
"step": 120
|
| 369 |
},
|
| 370 |
{
|
| 371 |
-
"epoch": 0.
|
| 372 |
"learning_rate": 2e-05,
|
| 373 |
-
"loss": 1.
|
| 374 |
"step": 122
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch": 0.
|
| 378 |
"learning_rate": 2e-05,
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 124
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch": 0.
|
| 384 |
"learning_rate": 2e-05,
|
| 385 |
-
"loss":
|
| 386 |
"step": 126
|
| 387 |
},
|
| 388 |
{
|
| 389 |
-
"epoch": 0.
|
| 390 |
"learning_rate": 2e-05,
|
| 391 |
-
"loss": 1.
|
| 392 |
"step": 128
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"epoch": 0.
|
| 396 |
"learning_rate": 2e-05,
|
| 397 |
-
"loss": 1.
|
| 398 |
"step": 130
|
| 399 |
},
|
| 400 |
{
|
| 401 |
-
"epoch": 0.
|
| 402 |
"learning_rate": 2e-05,
|
| 403 |
-
"loss": 1.
|
| 404 |
"step": 132
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch": 0.
|
| 408 |
"learning_rate": 2e-05,
|
| 409 |
-
"loss": 1.
|
| 410 |
"step": 134
|
| 411 |
},
|
| 412 |
{
|
| 413 |
-
"epoch": 0.
|
| 414 |
"learning_rate": 2e-05,
|
| 415 |
-
"loss": 1.
|
| 416 |
"step": 136
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch": 0.
|
| 420 |
"learning_rate": 2e-05,
|
| 421 |
-
"loss": 1.
|
| 422 |
"step": 138
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch": 0.
|
| 426 |
"learning_rate": 2e-05,
|
| 427 |
-
"loss": 1.
|
| 428 |
"step": 140
|
| 429 |
},
|
| 430 |
{
|
| 431 |
-
"epoch": 0.
|
| 432 |
"learning_rate": 2e-05,
|
| 433 |
-
"loss": 1.
|
| 434 |
"step": 142
|
| 435 |
},
|
| 436 |
{
|
| 437 |
-
"epoch": 0.
|
| 438 |
"learning_rate": 2e-05,
|
| 439 |
-
"loss": 1.
|
| 440 |
"step": 144
|
| 441 |
},
|
| 442 |
{
|
| 443 |
-
"epoch": 0.
|
| 444 |
"learning_rate": 2e-05,
|
| 445 |
-
"loss": 1.
|
| 446 |
"step": 146
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch": 0.
|
| 450 |
"learning_rate": 2e-05,
|
| 451 |
-
"loss": 1.
|
| 452 |
"step": 148
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch": 0.
|
| 456 |
"learning_rate": 2e-05,
|
| 457 |
-
"loss": 1.
|
| 458 |
"step": 150
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 0.
|
| 462 |
"learning_rate": 2e-05,
|
| 463 |
-
"loss": 1.
|
| 464 |
"step": 152
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
"learning_rate": 2e-05,
|
| 469 |
-
"loss": 1.
|
| 470 |
"step": 154
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch": 0.
|
| 474 |
"learning_rate": 2e-05,
|
| 475 |
-
"loss": 1.
|
| 476 |
"step": 156
|
| 477 |
},
|
| 478 |
{
|
| 479 |
-
"epoch": 0.
|
| 480 |
"learning_rate": 2e-05,
|
| 481 |
-
"loss": 1.
|
| 482 |
"step": 158
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"epoch": 0.
|
| 486 |
"learning_rate": 2e-05,
|
| 487 |
-
"loss": 1.
|
| 488 |
"step": 160
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch": 0.
|
| 492 |
"learning_rate": 2e-05,
|
| 493 |
-
"loss": 1.
|
| 494 |
"step": 162
|
| 495 |
},
|
| 496 |
{
|
| 497 |
-
"epoch": 0.
|
| 498 |
"learning_rate": 2e-05,
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 164
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch": 0.
|
| 504 |
"learning_rate": 2e-05,
|
| 505 |
-
"loss": 1.
|
| 506 |
"step": 166
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch": 0.
|
| 510 |
"learning_rate": 2e-05,
|
| 511 |
-
"loss":
|
| 512 |
"step": 168
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch": 0.
|
| 516 |
"learning_rate": 2e-05,
|
| 517 |
-
"loss": 1.
|
| 518 |
"step": 170
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch": 0.
|
| 522 |
"learning_rate": 2e-05,
|
| 523 |
-
"loss": 1.
|
| 524 |
"step": 172
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch": 0.
|
| 528 |
"learning_rate": 2e-05,
|
| 529 |
-
"loss": 1.
|
| 530 |
"step": 174
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch": 0.
|
| 534 |
"learning_rate": 2e-05,
|
| 535 |
-
"loss": 1.
|
| 536 |
"step": 176
|
| 537 |
},
|
| 538 |
{
|
| 539 |
-
"epoch": 0.
|
| 540 |
"learning_rate": 2e-05,
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 178
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch": 0.
|
| 546 |
"learning_rate": 2e-05,
|
| 547 |
-
"loss": 1.
|
| 548 |
"step": 180
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch": 0.
|
| 552 |
"learning_rate": 2e-05,
|
| 553 |
-
"loss": 1.
|
| 554 |
"step": 182
|
| 555 |
},
|
| 556 |
{
|
| 557 |
-
"epoch": 0.
|
| 558 |
"learning_rate": 2e-05,
|
| 559 |
-
"loss": 1.
|
| 560 |
"step": 184
|
| 561 |
},
|
| 562 |
{
|
| 563 |
-
"epoch": 0.
|
| 564 |
"learning_rate": 2e-05,
|
| 565 |
-
"loss": 1.
|
| 566 |
"step": 186
|
| 567 |
},
|
| 568 |
{
|
| 569 |
-
"epoch": 0.
|
| 570 |
"learning_rate": 2e-05,
|
| 571 |
-
"loss": 1.
|
| 572 |
"step": 188
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch": 0.
|
| 576 |
"learning_rate": 2e-05,
|
| 577 |
-
"loss": 1.
|
| 578 |
"step": 190
|
| 579 |
},
|
| 580 |
{
|
| 581 |
-
"epoch": 0.
|
| 582 |
"learning_rate": 2e-05,
|
| 583 |
-
"loss": 1.
|
| 584 |
"step": 192
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch": 0.
|
| 588 |
"learning_rate": 2e-05,
|
| 589 |
-
"loss": 1.
|
| 590 |
"step": 194
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch": 0.
|
| 594 |
"learning_rate": 2e-05,
|
| 595 |
-
"loss": 1.
|
| 596 |
"step": 196
|
| 597 |
},
|
| 598 |
{
|
| 599 |
-
"epoch": 0.
|
| 600 |
"learning_rate": 2e-05,
|
| 601 |
-
"loss": 1.
|
| 602 |
"step": 198
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch": 0.
|
| 606 |
"learning_rate": 2e-05,
|
| 607 |
-
"loss": 1.
|
| 608 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 609 |
}
|
| 610 |
],
|
| 611 |
"max_steps": 1000,
|
| 612 |
-
"num_train_epochs":
|
| 613 |
-
"total_flos":
|
| 614 |
"trial_name": null,
|
| 615 |
"trial_params": null
|
| 616 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.26666666666666666,
|
| 5 |
+
"global_step": 250,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 10 |
{
|
| 11 |
"epoch": 0.0,
|
| 12 |
"learning_rate": 2e-05,
|
| 13 |
+
"loss": 1.5189,
|
| 14 |
"step": 2
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.0,
|
| 18 |
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 1.7638,
|
| 20 |
"step": 4
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
"learning_rate": 2e-05,
|
| 25 |
+
"loss": 1.7996,
|
| 26 |
"step": 6
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.01,
|
| 30 |
"learning_rate": 2e-05,
|
| 31 |
+
"loss": 1.8916,
|
| 32 |
"step": 8
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.01,
|
| 36 |
"learning_rate": 2e-05,
|
| 37 |
+
"loss": 1.8279,
|
| 38 |
"step": 10
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.01,
|
| 42 |
"learning_rate": 2e-05,
|
| 43 |
+
"loss": 1.7899,
|
| 44 |
"step": 12
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.01,
|
| 48 |
"learning_rate": 2e-05,
|
| 49 |
+
"loss": 1.8743,
|
| 50 |
"step": 14
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 0.02,
|
| 54 |
"learning_rate": 2e-05,
|
| 55 |
+
"loss": 1.8173,
|
| 56 |
"step": 16
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
"learning_rate": 2e-05,
|
| 61 |
+
"loss": 1.6883,
|
| 62 |
"step": 18
|
| 63 |
},
|
| 64 |
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
"learning_rate": 2e-05,
|
| 67 |
+
"loss": 1.7763,
|
| 68 |
"step": 20
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.02,
|
| 72 |
"learning_rate": 2e-05,
|
| 73 |
+
"loss": 1.6734,
|
| 74 |
"step": 22
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 0.03,
|
| 78 |
"learning_rate": 2e-05,
|
| 79 |
+
"loss": 1.7075,
|
| 80 |
"step": 24
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.03,
|
| 84 |
"learning_rate": 2e-05,
|
| 85 |
+
"loss": 1.6444,
|
| 86 |
"step": 26
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.03,
|
| 90 |
"learning_rate": 2e-05,
|
| 91 |
+
"loss": 1.4933,
|
| 92 |
"step": 28
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.03,
|
| 96 |
"learning_rate": 2e-05,
|
| 97 |
+
"loss": 1.5791,
|
| 98 |
"step": 30
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
"learning_rate": 2e-05,
|
| 103 |
+
"loss": 1.7062,
|
| 104 |
"step": 32
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 0.04,
|
| 108 |
"learning_rate": 2e-05,
|
| 109 |
+
"loss": 1.8535,
|
| 110 |
"step": 34
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 0.04,
|
| 114 |
"learning_rate": 2e-05,
|
| 115 |
+
"loss": 1.978,
|
| 116 |
"step": 36
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 0.04,
|
| 120 |
"learning_rate": 2e-05,
|
| 121 |
+
"loss": 1.8168,
|
| 122 |
"step": 38
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.04,
|
| 126 |
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 1.4516,
|
| 128 |
"step": 40
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.04,
|
| 132 |
"learning_rate": 2e-05,
|
| 133 |
+
"loss": 1.623,
|
| 134 |
"step": 42
|
| 135 |
},
|
| 136 |
{
|
| 137 |
+
"epoch": 0.05,
|
| 138 |
"learning_rate": 2e-05,
|
| 139 |
+
"loss": 1.6009,
|
| 140 |
"step": 44
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 0.05,
|
| 144 |
"learning_rate": 2e-05,
|
| 145 |
+
"loss": 1.4367,
|
| 146 |
"step": 46
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 0.05,
|
| 150 |
"learning_rate": 2e-05,
|
| 151 |
+
"loss": 1.4459,
|
| 152 |
"step": 48
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.05,
|
| 156 |
"learning_rate": 2e-05,
|
| 157 |
+
"loss": 1.2903,
|
| 158 |
"step": 50
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 0.06,
|
| 162 |
"learning_rate": 2e-05,
|
| 163 |
+
"loss": 1.3021,
|
| 164 |
"step": 52
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.06,
|
| 168 |
"learning_rate": 2e-05,
|
| 169 |
+
"loss": 1.5408,
|
| 170 |
"step": 54
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.06,
|
| 174 |
"learning_rate": 2e-05,
|
| 175 |
+
"loss": 1.595,
|
| 176 |
"step": 56
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 0.06,
|
| 180 |
"learning_rate": 2e-05,
|
| 181 |
+
"loss": 1.69,
|
| 182 |
"step": 58
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 0.06,
|
| 186 |
"learning_rate": 2e-05,
|
| 187 |
+
"loss": 1.6761,
|
| 188 |
"step": 60
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 0.07,
|
| 192 |
"learning_rate": 2e-05,
|
| 193 |
+
"loss": 1.5006,
|
| 194 |
"step": 62
|
| 195 |
},
|
| 196 |
{
|
| 197 |
+
"epoch": 0.07,
|
| 198 |
"learning_rate": 2e-05,
|
| 199 |
+
"loss": 1.7522,
|
| 200 |
"step": 64
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"epoch": 0.07,
|
| 204 |
"learning_rate": 2e-05,
|
| 205 |
+
"loss": 1.6367,
|
| 206 |
"step": 66
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.07,
|
| 210 |
"learning_rate": 2e-05,
|
| 211 |
+
"loss": 1.7298,
|
| 212 |
"step": 68
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.07,
|
| 216 |
"learning_rate": 2e-05,
|
| 217 |
+
"loss": 1.6236,
|
| 218 |
"step": 70
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.08,
|
| 222 |
"learning_rate": 2e-05,
|
| 223 |
+
"loss": 1.6526,
|
| 224 |
"step": 72
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 0.08,
|
| 228 |
"learning_rate": 2e-05,
|
| 229 |
+
"loss": 1.6137,
|
| 230 |
"step": 74
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 0.08,
|
| 234 |
"learning_rate": 2e-05,
|
| 235 |
+
"loss": 1.5818,
|
| 236 |
"step": 76
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 0.08,
|
| 240 |
"learning_rate": 2e-05,
|
| 241 |
+
"loss": 1.3845,
|
| 242 |
"step": 78
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 0.09,
|
| 246 |
"learning_rate": 2e-05,
|
| 247 |
+
"loss": 1.4508,
|
| 248 |
"step": 80
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.09,
|
| 252 |
"learning_rate": 2e-05,
|
| 253 |
+
"loss": 1.4855,
|
| 254 |
"step": 82
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.09,
|
| 258 |
"learning_rate": 2e-05,
|
| 259 |
+
"loss": 1.4927,
|
| 260 |
"step": 84
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"epoch": 0.09,
|
| 264 |
"learning_rate": 2e-05,
|
| 265 |
+
"loss": 2.1637,
|
| 266 |
"step": 86
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 0.09,
|
| 270 |
"learning_rate": 2e-05,
|
| 271 |
+
"loss": 1.5407,
|
| 272 |
"step": 88
|
| 273 |
},
|
| 274 |
{
|
| 275 |
+
"epoch": 0.1,
|
| 276 |
"learning_rate": 2e-05,
|
| 277 |
+
"loss": 1.4036,
|
| 278 |
"step": 90
|
| 279 |
},
|
| 280 |
{
|
| 281 |
+
"epoch": 0.1,
|
| 282 |
"learning_rate": 2e-05,
|
| 283 |
+
"loss": 1.3731,
|
| 284 |
"step": 92
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 0.1,
|
| 288 |
"learning_rate": 2e-05,
|
| 289 |
+
"loss": 1.3963,
|
| 290 |
"step": 94
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.1,
|
| 294 |
"learning_rate": 2e-05,
|
| 295 |
+
"loss": 1.3245,
|
| 296 |
"step": 96
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.1,
|
| 300 |
"learning_rate": 2e-05,
|
| 301 |
+
"loss": 1.2533,
|
| 302 |
"step": 98
|
| 303 |
},
|
| 304 |
{
|
| 305 |
+
"epoch": 0.11,
|
| 306 |
"learning_rate": 2e-05,
|
| 307 |
+
"loss": 1.0874,
|
| 308 |
"step": 100
|
| 309 |
},
|
| 310 |
{
|
| 311 |
+
"epoch": 0.11,
|
| 312 |
"learning_rate": 2e-05,
|
| 313 |
+
"loss": 1.2428,
|
| 314 |
"step": 102
|
| 315 |
},
|
| 316 |
{
|
| 317 |
+
"epoch": 0.11,
|
| 318 |
"learning_rate": 2e-05,
|
| 319 |
+
"loss": 1.3934,
|
| 320 |
"step": 104
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 0.11,
|
| 324 |
"learning_rate": 2e-05,
|
| 325 |
+
"loss": 1.6125,
|
| 326 |
"step": 106
|
| 327 |
},
|
| 328 |
{
|
| 329 |
+
"epoch": 0.12,
|
| 330 |
"learning_rate": 2e-05,
|
| 331 |
+
"loss": 1.5428,
|
| 332 |
"step": 108
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.12,
|
| 336 |
"learning_rate": 2e-05,
|
| 337 |
+
"loss": 1.6413,
|
| 338 |
"step": 110
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.12,
|
| 342 |
"learning_rate": 2e-05,
|
| 343 |
+
"loss": 1.7078,
|
| 344 |
"step": 112
|
| 345 |
},
|
| 346 |
{
|
| 347 |
+
"epoch": 0.12,
|
| 348 |
"learning_rate": 2e-05,
|
| 349 |
+
"loss": 1.5403,
|
| 350 |
"step": 114
|
| 351 |
},
|
| 352 |
{
|
| 353 |
+
"epoch": 0.12,
|
| 354 |
"learning_rate": 2e-05,
|
| 355 |
+
"loss": 1.673,
|
| 356 |
"step": 116
|
| 357 |
},
|
| 358 |
{
|
| 359 |
+
"epoch": 0.13,
|
| 360 |
"learning_rate": 2e-05,
|
| 361 |
+
"loss": 1.6125,
|
| 362 |
"step": 118
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 0.13,
|
| 366 |
"learning_rate": 2e-05,
|
| 367 |
+
"loss": 1.5603,
|
| 368 |
"step": 120
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 0.13,
|
| 372 |
"learning_rate": 2e-05,
|
| 373 |
+
"loss": 1.5853,
|
| 374 |
"step": 122
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 0.13,
|
| 378 |
"learning_rate": 2e-05,
|
| 379 |
+
"loss": 1.4797,
|
| 380 |
"step": 124
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.13,
|
| 384 |
"learning_rate": 2e-05,
|
| 385 |
+
"loss": 1.5056,
|
| 386 |
"step": 126
|
| 387 |
},
|
| 388 |
{
|
| 389 |
+
"epoch": 0.14,
|
| 390 |
"learning_rate": 2e-05,
|
| 391 |
+
"loss": 1.4937,
|
| 392 |
"step": 128
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 0.14,
|
| 396 |
"learning_rate": 2e-05,
|
| 397 |
+
"loss": 1.5029,
|
| 398 |
"step": 130
|
| 399 |
},
|
| 400 |
{
|
| 401 |
+
"epoch": 0.14,
|
| 402 |
"learning_rate": 2e-05,
|
| 403 |
+
"loss": 1.4266,
|
| 404 |
"step": 132
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 0.14,
|
| 408 |
"learning_rate": 2e-05,
|
| 409 |
+
"loss": 1.6567,
|
| 410 |
"step": 134
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 0.15,
|
| 414 |
"learning_rate": 2e-05,
|
| 415 |
+
"loss": 1.6873,
|
| 416 |
"step": 136
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.15,
|
| 420 |
"learning_rate": 2e-05,
|
| 421 |
+
"loss": 1.6388,
|
| 422 |
"step": 138
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.15,
|
| 426 |
"learning_rate": 2e-05,
|
| 427 |
+
"loss": 1.3007,
|
| 428 |
"step": 140
|
| 429 |
},
|
| 430 |
{
|
| 431 |
+
"epoch": 0.15,
|
| 432 |
"learning_rate": 2e-05,
|
| 433 |
+
"loss": 1.2195,
|
| 434 |
"step": 142
|
| 435 |
},
|
| 436 |
{
|
| 437 |
+
"epoch": 0.15,
|
| 438 |
"learning_rate": 2e-05,
|
| 439 |
+
"loss": 1.2056,
|
| 440 |
"step": 144
|
| 441 |
},
|
| 442 |
{
|
| 443 |
+
"epoch": 0.16,
|
| 444 |
"learning_rate": 2e-05,
|
| 445 |
+
"loss": 1.2234,
|
| 446 |
"step": 146
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 0.16,
|
| 450 |
"learning_rate": 2e-05,
|
| 451 |
+
"loss": 1.08,
|
| 452 |
"step": 148
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 0.16,
|
| 456 |
"learning_rate": 2e-05,
|
| 457 |
+
"loss": 1.0367,
|
| 458 |
"step": 150
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.16,
|
| 462 |
"learning_rate": 2e-05,
|
| 463 |
+
"loss": 1.3315,
|
| 464 |
"step": 152
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.16,
|
| 468 |
"learning_rate": 2e-05,
|
| 469 |
+
"loss": 1.2798,
|
| 470 |
"step": 154
|
| 471 |
},
|
| 472 |
{
|
| 473 |
+
"epoch": 0.17,
|
| 474 |
"learning_rate": 2e-05,
|
| 475 |
+
"loss": 1.4107,
|
| 476 |
"step": 156
|
| 477 |
},
|
| 478 |
{
|
| 479 |
+
"epoch": 0.17,
|
| 480 |
"learning_rate": 2e-05,
|
| 481 |
+
"loss": 1.5695,
|
| 482 |
"step": 158
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 0.17,
|
| 486 |
"learning_rate": 2e-05,
|
| 487 |
+
"loss": 1.6263,
|
| 488 |
"step": 160
|
| 489 |
},
|
| 490 |
{
|
| 491 |
+
"epoch": 0.17,
|
| 492 |
"learning_rate": 2e-05,
|
| 493 |
+
"loss": 1.6117,
|
| 494 |
"step": 162
|
| 495 |
},
|
| 496 |
{
|
| 497 |
+
"epoch": 0.17,
|
| 498 |
"learning_rate": 2e-05,
|
| 499 |
+
"loss": 1.6127,
|
| 500 |
"step": 164
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 0.18,
|
| 504 |
"learning_rate": 2e-05,
|
| 505 |
+
"loss": 1.5601,
|
| 506 |
"step": 166
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.18,
|
| 510 |
"learning_rate": 2e-05,
|
| 511 |
+
"loss": 1.4628,
|
| 512 |
"step": 168
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.18,
|
| 516 |
"learning_rate": 2e-05,
|
| 517 |
+
"loss": 1.6585,
|
| 518 |
"step": 170
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 0.18,
|
| 522 |
"learning_rate": 2e-05,
|
| 523 |
+
"loss": 1.4854,
|
| 524 |
"step": 172
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 0.19,
|
| 528 |
"learning_rate": 2e-05,
|
| 529 |
+
"loss": 1.5067,
|
| 530 |
"step": 174
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 0.19,
|
| 534 |
"learning_rate": 2e-05,
|
| 535 |
+
"loss": 1.4587,
|
| 536 |
"step": 176
|
| 537 |
},
|
| 538 |
{
|
| 539 |
+
"epoch": 0.19,
|
| 540 |
"learning_rate": 2e-05,
|
| 541 |
+
"loss": 1.4076,
|
| 542 |
"step": 178
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 0.19,
|
| 546 |
"learning_rate": 2e-05,
|
| 547 |
+
"loss": 1.4073,
|
| 548 |
"step": 180
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 0.19,
|
| 552 |
"learning_rate": 2e-05,
|
| 553 |
+
"loss": 1.3402,
|
| 554 |
"step": 182
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 0.2,
|
| 558 |
"learning_rate": 2e-05,
|
| 559 |
+
"loss": 1.6865,
|
| 560 |
"step": 184
|
| 561 |
},
|
| 562 |
{
|
| 563 |
+
"epoch": 0.2,
|
| 564 |
"learning_rate": 2e-05,
|
| 565 |
+
"loss": 1.5336,
|
| 566 |
"step": 186
|
| 567 |
},
|
| 568 |
{
|
| 569 |
+
"epoch": 0.2,
|
| 570 |
"learning_rate": 2e-05,
|
| 571 |
+
"loss": 1.9135,
|
| 572 |
"step": 188
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"epoch": 0.2,
|
| 576 |
"learning_rate": 2e-05,
|
| 577 |
+
"loss": 1.2408,
|
| 578 |
"step": 190
|
| 579 |
},
|
| 580 |
{
|
| 581 |
+
"epoch": 0.2,
|
| 582 |
"learning_rate": 2e-05,
|
| 583 |
+
"loss": 1.2584,
|
| 584 |
"step": 192
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 0.21,
|
| 588 |
"learning_rate": 2e-05,
|
| 589 |
+
"loss": 1.1331,
|
| 590 |
"step": 194
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 0.21,
|
| 594 |
"learning_rate": 2e-05,
|
| 595 |
+
"loss": 1.0084,
|
| 596 |
"step": 196
|
| 597 |
},
|
| 598 |
{
|
| 599 |
+
"epoch": 0.21,
|
| 600 |
"learning_rate": 2e-05,
|
| 601 |
+
"loss": 1.1522,
|
| 602 |
"step": 198
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"epoch": 0.21,
|
| 606 |
"learning_rate": 2e-05,
|
| 607 |
+
"loss": 1.0801,
|
| 608 |
"step": 200
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.22,
|
| 612 |
+
"learning_rate": 2e-05,
|
| 613 |
+
"loss": 1.3441,
|
| 614 |
+
"step": 202
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.22,
|
| 618 |
+
"learning_rate": 2e-05,
|
| 619 |
+
"loss": 1.3746,
|
| 620 |
+
"step": 204
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.22,
|
| 624 |
+
"learning_rate": 2e-05,
|
| 625 |
+
"loss": 1.4496,
|
| 626 |
+
"step": 206
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.22,
|
| 630 |
+
"learning_rate": 2e-05,
|
| 631 |
+
"loss": 1.5532,
|
| 632 |
+
"step": 208
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.22,
|
| 636 |
+
"learning_rate": 2e-05,
|
| 637 |
+
"loss": 1.5473,
|
| 638 |
+
"step": 210
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.23,
|
| 642 |
+
"learning_rate": 2e-05,
|
| 643 |
+
"loss": 1.6257,
|
| 644 |
+
"step": 212
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.23,
|
| 648 |
+
"learning_rate": 2e-05,
|
| 649 |
+
"loss": 1.5486,
|
| 650 |
+
"step": 214
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.23,
|
| 654 |
+
"learning_rate": 2e-05,
|
| 655 |
+
"loss": 1.6562,
|
| 656 |
+
"step": 216
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.23,
|
| 660 |
+
"learning_rate": 2e-05,
|
| 661 |
+
"loss": 1.6434,
|
| 662 |
+
"step": 218
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.23,
|
| 666 |
+
"learning_rate": 2e-05,
|
| 667 |
+
"loss": 1.5767,
|
| 668 |
+
"step": 220
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.24,
|
| 672 |
+
"learning_rate": 2e-05,
|
| 673 |
+
"loss": 1.5164,
|
| 674 |
+
"step": 222
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.24,
|
| 678 |
+
"learning_rate": 2e-05,
|
| 679 |
+
"loss": 1.4387,
|
| 680 |
+
"step": 224
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.24,
|
| 684 |
+
"learning_rate": 2e-05,
|
| 685 |
+
"loss": 1.4196,
|
| 686 |
+
"step": 226
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.24,
|
| 690 |
+
"learning_rate": 2e-05,
|
| 691 |
+
"loss": 1.1444,
|
| 692 |
+
"step": 228
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.25,
|
| 696 |
+
"learning_rate": 2e-05,
|
| 697 |
+
"loss": 1.4044,
|
| 698 |
+
"step": 230
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.25,
|
| 702 |
+
"learning_rate": 2e-05,
|
| 703 |
+
"loss": 1.5233,
|
| 704 |
+
"step": 232
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.25,
|
| 708 |
+
"learning_rate": 2e-05,
|
| 709 |
+
"loss": 1.3644,
|
| 710 |
+
"step": 234
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.25,
|
| 714 |
+
"learning_rate": 2e-05,
|
| 715 |
+
"loss": 1.4033,
|
| 716 |
+
"step": 236
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.25,
|
| 720 |
+
"learning_rate": 2e-05,
|
| 721 |
+
"loss": 2.0282,
|
| 722 |
+
"step": 238
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.26,
|
| 726 |
+
"learning_rate": 2e-05,
|
| 727 |
+
"loss": 1.2922,
|
| 728 |
+
"step": 240
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.26,
|
| 732 |
+
"learning_rate": 2e-05,
|
| 733 |
+
"loss": 1.2978,
|
| 734 |
+
"step": 242
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.26,
|
| 738 |
+
"learning_rate": 2e-05,
|
| 739 |
+
"loss": 1.1365,
|
| 740 |
+
"step": 244
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.26,
|
| 744 |
+
"learning_rate": 2e-05,
|
| 745 |
+
"loss": 1.0143,
|
| 746 |
+
"step": 246
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.26,
|
| 750 |
+
"learning_rate": 2e-05,
|
| 751 |
+
"loss": 1.0573,
|
| 752 |
+
"step": 248
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.27,
|
| 756 |
+
"learning_rate": 2e-05,
|
| 757 |
+
"loss": 1.057,
|
| 758 |
+
"step": 250
|
| 759 |
}
|
| 760 |
],
|
| 761 |
"max_steps": 1000,
|
| 762 |
+
"num_train_epochs": 2,
|
| 763 |
+
"total_flos": 1.216157036454912e+16,
|
| 764 |
"trial_name": null,
|
| 765 |
"trial_params": null
|
| 766 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3963
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6c41e7e7909f822e649a159e03c3918f946633f35bcdf29a5fc019fb7ac23ff5
|
| 3 |
size 3963
|