Training in progress, step 200, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +802 -2272
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 373077376
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c63902d8b0744436654604cbbb61ea6622249c7b321940fc5dce3ca6b189ab84
|
| 3 |
size 373077376
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 373225035
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f74586ca15fa24e8f1f52d970ae432e5c7d4cd98179b069c5c5b3185e749a36e
|
| 3 |
size 373225035
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1401
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:081c3b4e71cca86d6e9989fa363a77944af063f9fac1ce35c3d68967db489b61
|
| 3 |
size 1401
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -4,2884 +4,1414 @@
|
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
"epoch": 10.0,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
-
"loss": 10.
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm": 2.
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss": 10.
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss": 10.
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss": 10.
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm": 2.
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss": 10.
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm": 2.
|
| 50 |
-
"learning_rate": 0.
|
| 51 |
-
"loss": 10.
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm": 2.
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss": 10.
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm": 2.
|
| 64 |
-
"learning_rate": 0.
|
| 65 |
-
"loss": 10.
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 0.
|
| 72 |
-
"loss": 10.
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"grad_norm": 2.
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 10.
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"grad_norm": 2.
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss": 10.
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm": 2.
|
| 92 |
-
"learning_rate": 0.
|
| 93 |
-
"loss": 10.
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"grad_norm": 2.
|
| 99 |
-
"learning_rate": 0.
|
| 100 |
-
"loss": 10.
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch": 0.
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 0.
|
| 107 |
-
"loss": 10.
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch": 0.
|
| 112 |
-
"grad_norm": 2.
|
| 113 |
-
"learning_rate": 0.
|
| 114 |
-
"loss": 10.
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch": 0.
|
| 119 |
-
"grad_norm": 3.
|
| 120 |
-
"learning_rate": 0.
|
| 121 |
-
"loss": 10.
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
-
"grad_norm":
|
| 127 |
-
"learning_rate": 0.
|
| 128 |
-
"loss": 10.
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate": 0.
|
| 135 |
-
"loss": 10.
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch": 0.
|
| 140 |
-
"grad_norm": 2.
|
| 141 |
-
"learning_rate": 0.
|
| 142 |
-
"loss": 10.
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch":
|
| 147 |
-
"grad_norm": 2.
|
| 148 |
-
"learning_rate": 0.
|
| 149 |
-
"loss": 10.
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch":
|
| 154 |
-
"grad_norm": 3.
|
| 155 |
-
"learning_rate": 0.
|
| 156 |
-
"loss": 10.
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch":
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate": 0.
|
| 163 |
-
"loss": 10.
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch":
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate": 0.
|
| 170 |
-
"loss": 10.
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"grad_norm": 2.
|
| 176 |
-
"learning_rate": 0.
|
| 177 |
-
"loss": 10.
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch":
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate": 0.
|
| 184 |
-
"loss": 10.
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch":
|
| 189 |
-
"grad_norm": 3.
|
| 190 |
-
"learning_rate": 0.
|
| 191 |
-
"loss":
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch":
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate": 0.
|
| 198 |
-
"loss":
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch":
|
| 203 |
-
"grad_norm":
|
| 204 |
-
"learning_rate": 0.
|
| 205 |
-
"loss": 10.
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch":
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate": 0.
|
| 212 |
-
"loss":
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate": 0.
|
| 219 |
-
"loss":
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch":
|
| 224 |
-
"grad_norm":
|
| 225 |
-
"learning_rate": 0.
|
| 226 |
-
"loss":
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch":
|
| 231 |
-
"grad_norm":
|
| 232 |
-
"learning_rate": 0.
|
| 233 |
-
"loss":
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch":
|
| 238 |
-
"grad_norm": 5.
|
| 239 |
-
"learning_rate": 0.
|
| 240 |
-
"loss":
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch":
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate": 0.
|
| 247 |
-
"loss": 10.
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch":
|
| 252 |
-
"grad_norm":
|
| 253 |
-
"learning_rate": 0.
|
| 254 |
-
"loss":
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate": 0.
|
| 261 |
-
"loss": 10.
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
-
"epoch":
|
| 266 |
-
"grad_norm": 4.
|
| 267 |
-
"learning_rate": 0.
|
| 268 |
-
"loss":
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch":
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate": 0.
|
| 275 |
-
"loss":
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch":
|
| 280 |
-
"grad_norm":
|
| 281 |
-
"learning_rate": 0.
|
| 282 |
-
"loss":
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
-
"epoch":
|
| 287 |
-
"grad_norm":
|
| 288 |
-
"learning_rate": 0.
|
| 289 |
-
"loss":
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch":
|
| 294 |
-
"grad_norm":
|
| 295 |
-
"learning_rate": 0.
|
| 296 |
-
"loss":
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch":
|
| 301 |
-
"grad_norm":
|
| 302 |
-
"learning_rate": 0.
|
| 303 |
-
"loss": 10.
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
-
"epoch":
|
| 308 |
-
"grad_norm":
|
| 309 |
-
"learning_rate": 0.
|
| 310 |
-
"loss":
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch":
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate": 0.
|
| 317 |
-
"loss": 10.
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
-
"epoch":
|
| 322 |
-
"grad_norm":
|
| 323 |
-
"learning_rate": 0.
|
| 324 |
-
"loss": 9.
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
-
"epoch":
|
| 329 |
-
"grad_norm":
|
| 330 |
-
"learning_rate": 0.
|
| 331 |
-
"loss": 9.
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch":
|
| 336 |
-
"grad_norm":
|
| 337 |
-
"learning_rate": 0.
|
| 338 |
-
"loss": 9.
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch":
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate": 0.
|
| 345 |
-
"loss": 9.
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch":
|
| 350 |
-
"grad_norm": 3.
|
| 351 |
-
"learning_rate": 0.
|
| 352 |
-
"loss":
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch":
|
| 357 |
-
"grad_norm":
|
| 358 |
-
"learning_rate": 0.
|
| 359 |
-
"loss":
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch":
|
| 364 |
-
"grad_norm": 3.
|
| 365 |
-
"learning_rate": 0.
|
| 366 |
-
"loss":
|
| 367 |
"step": 51
|
| 368 |
},
|
| 369 |
{
|
| 370 |
-
"epoch":
|
| 371 |
-
"grad_norm": 3.
|
| 372 |
-
"learning_rate": 0.
|
| 373 |
-
"loss":
|
| 374 |
"step": 52
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch":
|
| 378 |
-
"grad_norm":
|
| 379 |
-
"learning_rate": 0.
|
| 380 |
-
"loss":
|
| 381 |
"step": 53
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch":
|
| 385 |
-
"grad_norm": 4.
|
| 386 |
-
"learning_rate": 0.
|
| 387 |
-
"loss": 9.
|
| 388 |
"step": 54
|
| 389 |
},
|
| 390 |
{
|
| 391 |
-
"epoch":
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate": 0.
|
| 394 |
-
"loss": 9.
|
| 395 |
"step": 55
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch":
|
| 399 |
-
"grad_norm":
|
| 400 |
-
"learning_rate": 0.
|
| 401 |
-
"loss": 8.
|
| 402 |
"step": 56
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch":
|
| 406 |
-
"grad_norm":
|
| 407 |
-
"learning_rate": 0.
|
| 408 |
-
"loss":
|
| 409 |
"step": 57
|
| 410 |
},
|
| 411 |
{
|
| 412 |
-
"epoch":
|
| 413 |
"grad_norm": 4.21875,
|
| 414 |
-
"learning_rate": 0.
|
| 415 |
-
"loss":
|
| 416 |
"step": 58
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch":
|
| 420 |
-
"grad_norm": 4.
|
| 421 |
-
"learning_rate": 0.
|
| 422 |
-
"loss": 9.
|
| 423 |
"step": 59
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"grad_norm": 4.
|
| 428 |
-
"learning_rate": 0.
|
| 429 |
-
"loss":
|
| 430 |
"step": 60
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch":
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate": 0.
|
| 436 |
-
"loss":
|
| 437 |
"step": 61
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate": 0.
|
| 443 |
-
"loss":
|
| 444 |
"step": 62
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch":
|
| 448 |
-
"grad_norm": 4.
|
| 449 |
-
"learning_rate": 0.
|
| 450 |
-
"loss":
|
| 451 |
"step": 63
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch":
|
| 455 |
-
"grad_norm":
|
| 456 |
-
"learning_rate": 0.
|
| 457 |
-
"loss": 9.
|
| 458 |
"step": 64
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"grad_norm":
|
| 463 |
-
"learning_rate": 0.
|
| 464 |
-
"loss":
|
| 465 |
"step": 65
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"grad_norm": 3.
|
| 470 |
-
"learning_rate": 0.
|
| 471 |
-
"loss":
|
| 472 |
"step": 66
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch":
|
| 476 |
-
"grad_norm":
|
| 477 |
-
"learning_rate": 0.
|
| 478 |
-
"loss": 9.
|
| 479 |
"step": 67
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch":
|
| 483 |
-
"grad_norm":
|
| 484 |
-
"learning_rate": 0.
|
| 485 |
-
"loss": 9.
|
| 486 |
"step": 68
|
| 487 |
},
|
| 488 |
{
|
| 489 |
-
"epoch":
|
| 490 |
-
"grad_norm": 5.
|
| 491 |
-
"learning_rate": 0.
|
| 492 |
-
"loss":
|
| 493 |
"step": 69
|
| 494 |
},
|
| 495 |
{
|
| 496 |
-
"epoch":
|
| 497 |
-
"grad_norm": 4.
|
| 498 |
-
"learning_rate": 0.
|
| 499 |
-
"loss":
|
| 500 |
"step": 70
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch":
|
| 504 |
-
"grad_norm":
|
| 505 |
-
"learning_rate": 0.
|
| 506 |
-
"loss":
|
| 507 |
"step": 71
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
-
"grad_norm":
|
| 512 |
-
"learning_rate": 0.
|
| 513 |
-
"loss": 9.
|
| 514 |
"step": 72
|
| 515 |
},
|
| 516 |
{
|
| 517 |
-
"epoch":
|
| 518 |
-
"grad_norm":
|
| 519 |
-
"learning_rate": 0.
|
| 520 |
-
"loss": 8.
|
| 521 |
"step": 73
|
| 522 |
},
|
| 523 |
{
|
| 524 |
-
"epoch":
|
| 525 |
-
"grad_norm":
|
| 526 |
-
"learning_rate": 0.
|
| 527 |
-
"loss":
|
| 528 |
"step": 74
|
| 529 |
},
|
| 530 |
{
|
| 531 |
-
"epoch":
|
| 532 |
-
"grad_norm":
|
| 533 |
-
"learning_rate": 0.
|
| 534 |
-
"loss": 9.
|
| 535 |
"step": 75
|
| 536 |
},
|
| 537 |
{
|
| 538 |
-
"epoch":
|
| 539 |
-
"grad_norm":
|
| 540 |
-
"learning_rate": 0.
|
| 541 |
-
"loss": 9.
|
| 542 |
"step": 76
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch":
|
| 546 |
-
"grad_norm":
|
| 547 |
-
"learning_rate": 0.
|
| 548 |
-
"loss":
|
| 549 |
"step": 77
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch":
|
| 553 |
-
"grad_norm": 3.
|
| 554 |
-
"learning_rate": 0.
|
| 555 |
-
"loss":
|
| 556 |
"step": 78
|
| 557 |
},
|
| 558 |
{
|
| 559 |
-
"epoch":
|
| 560 |
-
"grad_norm":
|
| 561 |
-
"learning_rate": 0.
|
| 562 |
-
"loss": 8.
|
| 563 |
"step": 79
|
| 564 |
},
|
| 565 |
{
|
| 566 |
-
"epoch":
|
| 567 |
-
"grad_norm":
|
| 568 |
-
"learning_rate": 0.
|
| 569 |
-
"loss": 8.
|
| 570 |
"step": 80
|
| 571 |
},
|
| 572 |
{
|
| 573 |
-
"epoch":
|
| 574 |
-
"grad_norm": 3.
|
| 575 |
-
"learning_rate": 0.
|
| 576 |
-
"loss":
|
| 577 |
"step": 81
|
| 578 |
},
|
| 579 |
{
|
| 580 |
-
"epoch":
|
| 581 |
-
"grad_norm": 3.
|
| 582 |
-
"learning_rate": 0.
|
| 583 |
-
"loss":
|
| 584 |
"step": 82
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch":
|
| 588 |
-
"grad_norm":
|
| 589 |
-
"learning_rate": 0.
|
| 590 |
-
"loss": 8.
|
| 591 |
"step": 83
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
-
"grad_norm":
|
| 596 |
-
"learning_rate": 0.
|
| 597 |
-
"loss":
|
| 598 |
"step": 84
|
| 599 |
},
|
| 600 |
{
|
| 601 |
-
"epoch":
|
| 602 |
-
"grad_norm": 4.
|
| 603 |
-
"learning_rate": 0.
|
| 604 |
-
"loss":
|
| 605 |
"step": 85
|
| 606 |
},
|
| 607 |
{
|
| 608 |
-
"epoch":
|
| 609 |
-
"grad_norm":
|
| 610 |
-
"learning_rate": 0.
|
| 611 |
-
"loss":
|
| 612 |
"step": 86
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch":
|
| 616 |
-
"grad_norm":
|
| 617 |
-
"learning_rate": 0.
|
| 618 |
-
"loss":
|
| 619 |
"step": 87
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch":
|
| 623 |
-
"grad_norm": 4.
|
| 624 |
-
"learning_rate": 0.
|
| 625 |
-
"loss":
|
| 626 |
"step": 88
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch":
|
| 630 |
-
"grad_norm": 4.
|
| 631 |
-
"learning_rate": 0.
|
| 632 |
-
"loss":
|
| 633 |
"step": 89
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"grad_norm":
|
| 638 |
-
"learning_rate": 0.
|
| 639 |
-
"loss":
|
| 640 |
"step": 90
|
| 641 |
},
|
| 642 |
{
|
| 643 |
-
"epoch":
|
| 644 |
-
"grad_norm":
|
| 645 |
-
"learning_rate": 0.
|
| 646 |
-
"loss":
|
| 647 |
"step": 91
|
| 648 |
},
|
| 649 |
{
|
| 650 |
-
"epoch":
|
| 651 |
-
"grad_norm":
|
| 652 |
-
"learning_rate": 0.
|
| 653 |
-
"loss": 9.
|
| 654 |
"step": 92
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch":
|
| 658 |
-
"grad_norm":
|
| 659 |
-
"learning_rate": 0.
|
| 660 |
-
"loss": 8.
|
| 661 |
"step": 93
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch":
|
| 665 |
-
"grad_norm":
|
| 666 |
-
"learning_rate": 0.
|
| 667 |
-
"loss":
|
| 668 |
"step": 94
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch":
|
| 672 |
-
"grad_norm":
|
| 673 |
-
"learning_rate": 0.
|
| 674 |
-
"loss":
|
| 675 |
"step": 95
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch":
|
| 679 |
-
"grad_norm":
|
| 680 |
-
"learning_rate": 0.
|
| 681 |
-
"loss": 8.
|
| 682 |
"step": 96
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"epoch":
|
| 686 |
-
"grad_norm":
|
| 687 |
-
"learning_rate": 0.
|
| 688 |
-
"loss":
|
| 689 |
"step": 97
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch":
|
| 693 |
-
"grad_norm":
|
| 694 |
-
"learning_rate": 0.
|
| 695 |
-
"loss":
|
| 696 |
"step": 98
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch":
|
| 700 |
-
"grad_norm":
|
| 701 |
-
"learning_rate": 0.
|
| 702 |
-
"loss": 9.
|
| 703 |
"step": 99
|
| 704 |
},
|
| 705 |
{
|
| 706 |
-
"epoch":
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate": 0.
|
| 709 |
-
"loss":
|
| 710 |
"step": 100
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch":
|
| 714 |
-
"grad_norm": 3.
|
| 715 |
-
"learning_rate": 0.
|
| 716 |
-
"loss":
|
| 717 |
"step": 101
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch":
|
| 721 |
-
"grad_norm":
|
| 722 |
-
"learning_rate": 0.
|
| 723 |
-
"loss":
|
| 724 |
"step": 102
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"epoch":
|
| 728 |
-
"grad_norm":
|
| 729 |
-
"learning_rate": 0.
|
| 730 |
-
"loss":
|
| 731 |
"step": 103
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch":
|
| 735 |
-
"grad_norm":
|
| 736 |
-
"learning_rate": 0.
|
| 737 |
-
"loss": 8.
|
| 738 |
"step": 104
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch":
|
| 742 |
-
"grad_norm":
|
| 743 |
-
"learning_rate": 0.
|
| 744 |
-
"loss":
|
| 745 |
"step": 105
|
| 746 |
},
|
| 747 |
{
|
| 748 |
-
"epoch":
|
| 749 |
-
"grad_norm": 3.
|
| 750 |
-
"learning_rate": 0.
|
| 751 |
-
"loss":
|
| 752 |
"step": 106
|
| 753 |
},
|
| 754 |
{
|
| 755 |
-
"epoch":
|
| 756 |
-
"grad_norm":
|
| 757 |
-
"learning_rate": 0.
|
| 758 |
-
"loss":
|
| 759 |
"step": 107
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch":
|
| 763 |
-
"grad_norm": 3.
|
| 764 |
-
"learning_rate": 0.
|
| 765 |
-
"loss":
|
| 766 |
"step": 108
|
| 767 |
},
|
| 768 |
{
|
| 769 |
-
"epoch":
|
| 770 |
-
"grad_norm":
|
| 771 |
-
"learning_rate": 0.
|
| 772 |
-
"loss": 8.
|
| 773 |
"step": 109
|
| 774 |
},
|
| 775 |
{
|
| 776 |
-
"epoch":
|
| 777 |
-
"grad_norm": 2.
|
| 778 |
-
"learning_rate": 0.
|
| 779 |
-
"loss":
|
| 780 |
"step": 110
|
| 781 |
},
|
| 782 |
{
|
| 783 |
-
"epoch":
|
| 784 |
-
"grad_norm":
|
| 785 |
-
"learning_rate": 0.
|
| 786 |
-
"loss": 8.
|
| 787 |
"step": 111
|
| 788 |
},
|
| 789 |
{
|
| 790 |
-
"epoch":
|
| 791 |
-
"grad_norm":
|
| 792 |
-
"learning_rate": 0.
|
| 793 |
-
"loss":
|
| 794 |
"step": 112
|
| 795 |
},
|
| 796 |
{
|
| 797 |
-
"epoch":
|
| 798 |
-
"grad_norm":
|
| 799 |
-
"learning_rate": 0.
|
| 800 |
-
"loss":
|
| 801 |
"step": 113
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch":
|
| 805 |
-
"grad_norm":
|
| 806 |
-
"learning_rate": 0.
|
| 807 |
-
"loss":
|
| 808 |
"step": 114
|
| 809 |
},
|
| 810 |
{
|
| 811 |
-
"epoch":
|
| 812 |
-
"grad_norm":
|
| 813 |
-
"learning_rate": 0.
|
| 814 |
-
"loss":
|
| 815 |
"step": 115
|
| 816 |
},
|
| 817 |
{
|
| 818 |
-
"epoch":
|
| 819 |
-
"grad_norm":
|
| 820 |
-
"learning_rate": 0.
|
| 821 |
-
"loss":
|
| 822 |
"step": 116
|
| 823 |
},
|
| 824 |
{
|
| 825 |
-
"epoch":
|
| 826 |
-
"grad_norm":
|
| 827 |
-
"learning_rate": 0.
|
| 828 |
-
"loss": 7.
|
| 829 |
"step": 117
|
| 830 |
},
|
| 831 |
{
|
| 832 |
-
"epoch":
|
| 833 |
-
"grad_norm":
|
| 834 |
-
"learning_rate": 0.
|
| 835 |
-
"loss":
|
| 836 |
"step": 118
|
| 837 |
},
|
| 838 |
{
|
| 839 |
-
"epoch":
|
| 840 |
-
"grad_norm":
|
| 841 |
-
"learning_rate": 0.
|
| 842 |
-
"loss":
|
| 843 |
"step": 119
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch":
|
| 847 |
-
"grad_norm":
|
| 848 |
-
"learning_rate": 0.
|
| 849 |
-
"loss": 8.
|
| 850 |
"step": 120
|
| 851 |
},
|
| 852 |
{
|
| 853 |
-
"epoch":
|
| 854 |
-
"grad_norm":
|
| 855 |
-
"learning_rate": 0.
|
| 856 |
-
"loss":
|
| 857 |
"step": 121
|
| 858 |
},
|
| 859 |
{
|
| 860 |
-
"epoch":
|
| 861 |
-
"grad_norm":
|
| 862 |
-
"learning_rate": 0.
|
| 863 |
-
"loss":
|
| 864 |
"step": 122
|
| 865 |
},
|
| 866 |
{
|
| 867 |
-
"epoch":
|
| 868 |
-
"grad_norm":
|
| 869 |
-
"learning_rate": 0.
|
| 870 |
-
"loss":
|
| 871 |
"step": 123
|
| 872 |
},
|
| 873 |
{
|
| 874 |
-
"epoch":
|
| 875 |
-
"grad_norm":
|
| 876 |
-
"learning_rate": 0.
|
| 877 |
-
"loss": 7.
|
| 878 |
"step": 124
|
| 879 |
},
|
| 880 |
{
|
| 881 |
-
"epoch":
|
| 882 |
-
"grad_norm":
|
| 883 |
-
"learning_rate": 0.
|
| 884 |
-
"loss":
|
| 885 |
"step": 125
|
| 886 |
},
|
| 887 |
{
|
| 888 |
-
"epoch":
|
| 889 |
-
"grad_norm":
|
| 890 |
-
"learning_rate": 0.
|
| 891 |
-
"loss":
|
| 892 |
"step": 126
|
| 893 |
},
|
| 894 |
{
|
| 895 |
-
"epoch":
|
| 896 |
-
"grad_norm":
|
| 897 |
-
"learning_rate": 0.
|
| 898 |
-
"loss":
|
| 899 |
"step": 127
|
| 900 |
},
|
| 901 |
{
|
| 902 |
-
"epoch":
|
| 903 |
-
"grad_norm":
|
| 904 |
-
"learning_rate": 0.
|
| 905 |
-
"loss": 8.
|
| 906 |
"step": 128
|
| 907 |
},
|
| 908 |
{
|
| 909 |
-
"epoch":
|
| 910 |
-
"grad_norm":
|
| 911 |
-
"learning_rate": 0.
|
| 912 |
-
"loss":
|
| 913 |
"step": 129
|
| 914 |
},
|
| 915 |
{
|
| 916 |
-
"epoch":
|
| 917 |
-
"grad_norm":
|
| 918 |
-
"learning_rate": 0.
|
| 919 |
-
"loss": 7.
|
| 920 |
"step": 130
|
| 921 |
},
|
| 922 |
{
|
| 923 |
-
"epoch":
|
| 924 |
-
"grad_norm":
|
| 925 |
-
"learning_rate": 0.
|
| 926 |
-
"loss":
|
| 927 |
"step": 131
|
| 928 |
},
|
| 929 |
{
|
| 930 |
-
"epoch":
|
| 931 |
-
"grad_norm":
|
| 932 |
-
"learning_rate": 0.
|
| 933 |
-
"loss": 7.
|
| 934 |
"step": 132
|
| 935 |
},
|
| 936 |
{
|
| 937 |
-
"epoch":
|
| 938 |
-
"grad_norm":
|
| 939 |
-
"learning_rate": 0.
|
| 940 |
-
"loss": 8.
|
| 941 |
"step": 133
|
| 942 |
},
|
| 943 |
{
|
| 944 |
-
"epoch":
|
| 945 |
-
"grad_norm":
|
| 946 |
-
"learning_rate": 0.
|
| 947 |
-
"loss":
|
| 948 |
"step": 134
|
| 949 |
},
|
| 950 |
{
|
| 951 |
-
"epoch":
|
| 952 |
-
"grad_norm":
|
| 953 |
-
"learning_rate": 0.
|
| 954 |
-
"loss":
|
| 955 |
"step": 135
|
| 956 |
},
|
| 957 |
{
|
| 958 |
-
"epoch":
|
| 959 |
-
"grad_norm":
|
| 960 |
-
"learning_rate": 0.
|
| 961 |
-
"loss": 7.
|
| 962 |
"step": 136
|
| 963 |
},
|
| 964 |
{
|
| 965 |
-
"epoch":
|
| 966 |
-
"grad_norm":
|
| 967 |
-
"learning_rate": 0.
|
| 968 |
-
"loss":
|
| 969 |
"step": 137
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch":
|
| 973 |
-
"grad_norm":
|
| 974 |
-
"learning_rate": 0.
|
| 975 |
-
"loss":
|
| 976 |
"step": 138
|
| 977 |
},
|
| 978 |
{
|
| 979 |
-
"epoch":
|
| 980 |
-
"grad_norm":
|
| 981 |
-
"learning_rate": 0.
|
| 982 |
-
"loss": 8.
|
| 983 |
"step": 139
|
| 984 |
},
|
| 985 |
{
|
| 986 |
-
"epoch":
|
| 987 |
-
"grad_norm":
|
| 988 |
-
"learning_rate": 0.
|
| 989 |
-
"loss": 7.
|
| 990 |
"step": 140
|
| 991 |
},
|
| 992 |
{
|
| 993 |
-
"epoch":
|
| 994 |
-
"grad_norm":
|
| 995 |
-
"learning_rate": 0.
|
| 996 |
-
"loss": 9.
|
| 997 |
"step": 141
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
-
"epoch":
|
| 1001 |
-
"grad_norm":
|
| 1002 |
-
"learning_rate": 0.
|
| 1003 |
-
"loss":
|
| 1004 |
"step": 142
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"epoch":
|
| 1008 |
-
"grad_norm":
|
| 1009 |
-
"learning_rate": 0.
|
| 1010 |
-
"loss":
|
| 1011 |
"step": 143
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
-
"epoch":
|
| 1015 |
-
"grad_norm":
|
| 1016 |
-
"learning_rate": 0.
|
| 1017 |
-
"loss": 8.
|
| 1018 |
"step": 144
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
-
"epoch":
|
| 1022 |
-
"grad_norm":
|
| 1023 |
-
"learning_rate": 0.
|
| 1024 |
-
"loss":
|
| 1025 |
"step": 145
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
-
"epoch":
|
| 1029 |
-
"grad_norm":
|
| 1030 |
-
"learning_rate": 0.
|
| 1031 |
-
"loss": 8.
|
| 1032 |
"step": 146
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
-
"epoch":
|
| 1036 |
-
"grad_norm":
|
| 1037 |
-
"learning_rate": 0.
|
| 1038 |
-
"loss":
|
| 1039 |
"step": 147
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
-
"epoch":
|
| 1043 |
-
"grad_norm":
|
| 1044 |
-
"learning_rate": 0.
|
| 1045 |
-
"loss": 8.
|
| 1046 |
"step": 148
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
-
"epoch":
|
| 1050 |
-
"grad_norm":
|
| 1051 |
-
"learning_rate": 0.
|
| 1052 |
-
"loss":
|
| 1053 |
"step": 149
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
-
"epoch":
|
| 1057 |
-
"grad_norm":
|
| 1058 |
-
"learning_rate": 0.
|
| 1059 |
-
"loss":
|
| 1060 |
"step": 150
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
-
"epoch":
|
| 1064 |
-
"grad_norm":
|
| 1065 |
-
"learning_rate": 0.
|
| 1066 |
-
"loss": 8.
|
| 1067 |
"step": 151
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
-
"epoch":
|
| 1071 |
-
"grad_norm":
|
| 1072 |
-
"learning_rate": 0.
|
| 1073 |
-
"loss":
|
| 1074 |
"step": 152
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
-
"epoch":
|
| 1078 |
-
"grad_norm": 3.
|
| 1079 |
-
"learning_rate": 0.
|
| 1080 |
-
"loss":
|
| 1081 |
"step": 153
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
-
"epoch":
|
| 1085 |
-
"grad_norm":
|
| 1086 |
-
"learning_rate": 0.
|
| 1087 |
-
"loss":
|
| 1088 |
"step": 154
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
-
"epoch":
|
| 1092 |
-
"grad_norm":
|
| 1093 |
-
"learning_rate": 0.
|
| 1094 |
-
"loss": 8.
|
| 1095 |
"step": 155
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
-
"epoch":
|
| 1099 |
-
"grad_norm":
|
| 1100 |
-
"learning_rate": 0.
|
| 1101 |
-
"loss":
|
| 1102 |
"step": 156
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
-
"epoch":
|
| 1106 |
-
"grad_norm":
|
| 1107 |
-
"learning_rate": 0.
|
| 1108 |
-
"loss": 8.
|
| 1109 |
"step": 157
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
-
"epoch":
|
| 1113 |
-
"grad_norm":
|
| 1114 |
-
"learning_rate": 0.
|
| 1115 |
-
"loss": 7.
|
| 1116 |
"step": 158
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
-
"epoch":
|
| 1120 |
-
"grad_norm":
|
| 1121 |
-
"learning_rate": 0.
|
| 1122 |
-
"loss": 7.
|
| 1123 |
"step": 159
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
-
"epoch":
|
| 1127 |
-
"grad_norm":
|
| 1128 |
-
"learning_rate": 0.
|
| 1129 |
-
"loss":
|
| 1130 |
"step": 160
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
-
"epoch":
|
| 1134 |
-
"grad_norm":
|
| 1135 |
-
"learning_rate": 0.
|
| 1136 |
-
"loss":
|
| 1137 |
"step": 161
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
-
"epoch":
|
| 1141 |
-
"grad_norm":
|
| 1142 |
-
"learning_rate": 0.
|
| 1143 |
-
"loss":
|
| 1144 |
"step": 162
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
-
"epoch":
|
| 1148 |
-
"grad_norm":
|
| 1149 |
-
"learning_rate":
|
| 1150 |
-
"loss": 8.
|
| 1151 |
"step": 163
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
-
"epoch":
|
| 1155 |
-
"grad_norm":
|
| 1156 |
-
"learning_rate":
|
| 1157 |
-
"loss": 8.
|
| 1158 |
"step": 164
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
-
"epoch":
|
| 1162 |
-
"grad_norm":
|
| 1163 |
-
"learning_rate":
|
| 1164 |
-
"loss": 7.
|
| 1165 |
"step": 165
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
-
"epoch":
|
| 1169 |
-
"grad_norm":
|
| 1170 |
-
"learning_rate":
|
| 1171 |
-
"loss":
|
| 1172 |
"step": 166
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
-
"epoch":
|
| 1176 |
-
"grad_norm":
|
| 1177 |
-
"learning_rate":
|
| 1178 |
-
"loss":
|
| 1179 |
"step": 167
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
-
"epoch":
|
| 1183 |
-
"grad_norm": 3.
|
| 1184 |
-
"learning_rate":
|
| 1185 |
-
"loss":
|
| 1186 |
"step": 168
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
-
"epoch":
|
| 1190 |
-
"grad_norm":
|
| 1191 |
-
"learning_rate":
|
| 1192 |
-
"loss": 8.
|
| 1193 |
"step": 169
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
-
"epoch":
|
| 1197 |
-
"grad_norm":
|
| 1198 |
-
"learning_rate":
|
| 1199 |
-
"loss":
|
| 1200 |
"step": 170
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
-
"epoch":
|
| 1204 |
-
"grad_norm":
|
| 1205 |
-
"learning_rate":
|
| 1206 |
-
"loss":
|
| 1207 |
"step": 171
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
-
"epoch":
|
| 1211 |
-
"grad_norm":
|
| 1212 |
-
"learning_rate":
|
| 1213 |
-
"loss": 7.
|
| 1214 |
"step": 172
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
-
"epoch":
|
| 1218 |
-
"grad_norm":
|
| 1219 |
-
"learning_rate":
|
| 1220 |
-
"loss":
|
| 1221 |
"step": 173
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
-
"epoch":
|
| 1225 |
-
"grad_norm":
|
| 1226 |
-
"learning_rate":
|
| 1227 |
-
"loss":
|
| 1228 |
"step": 174
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
-
"epoch":
|
| 1232 |
-
"grad_norm":
|
| 1233 |
-
"learning_rate":
|
| 1234 |
-
"loss": 7.
|
| 1235 |
"step": 175
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
-
"epoch":
|
| 1239 |
-
"grad_norm":
|
| 1240 |
-
"learning_rate":
|
| 1241 |
-
"loss": 8.
|
| 1242 |
"step": 176
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
-
"epoch":
|
| 1246 |
-
"grad_norm":
|
| 1247 |
-
"learning_rate":
|
| 1248 |
-
"loss":
|
| 1249 |
"step": 177
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
-
"epoch":
|
| 1253 |
-
"grad_norm":
|
| 1254 |
-
"learning_rate":
|
| 1255 |
-
"loss":
|
| 1256 |
"step": 178
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
-
"epoch":
|
| 1260 |
-
"grad_norm":
|
| 1261 |
-
"learning_rate":
|
| 1262 |
-
"loss": 8.
|
| 1263 |
"step": 179
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
-
"epoch":
|
| 1267 |
-
"grad_norm":
|
| 1268 |
-
"learning_rate":
|
| 1269 |
-
"loss": 8.
|
| 1270 |
"step": 180
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
-
"epoch":
|
| 1274 |
-
"grad_norm":
|
| 1275 |
-
"learning_rate":
|
| 1276 |
-
"loss": 7.
|
| 1277 |
"step": 181
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
-
"epoch":
|
| 1281 |
-
"grad_norm":
|
| 1282 |
-
"learning_rate":
|
| 1283 |
-
"loss":
|
| 1284 |
"step": 182
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
-
"epoch":
|
| 1288 |
-
"grad_norm":
|
| 1289 |
-
"learning_rate":
|
| 1290 |
-
"loss":
|
| 1291 |
"step": 183
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
-
"epoch":
|
| 1295 |
-
"grad_norm":
|
| 1296 |
-
"learning_rate":
|
| 1297 |
-
"loss":
|
| 1298 |
"step": 184
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
-
"epoch":
|
| 1302 |
-
"grad_norm":
|
| 1303 |
-
"learning_rate":
|
| 1304 |
-
"loss":
|
| 1305 |
"step": 185
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
-
"epoch":
|
| 1309 |
-
"grad_norm":
|
| 1310 |
-
"learning_rate":
|
| 1311 |
-
"loss":
|
| 1312 |
"step": 186
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
-
"epoch":
|
| 1316 |
-
"grad_norm":
|
| 1317 |
-
"learning_rate":
|
| 1318 |
-
"loss":
|
| 1319 |
"step": 187
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
-
"epoch":
|
| 1323 |
-
"grad_norm":
|
| 1324 |
-
"learning_rate":
|
| 1325 |
-
"loss":
|
| 1326 |
"step": 188
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
-
"epoch":
|
| 1330 |
-
"grad_norm":
|
| 1331 |
-
"learning_rate":
|
| 1332 |
-
"loss":
|
| 1333 |
"step": 189
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
-
"epoch":
|
| 1337 |
-
"grad_norm":
|
| 1338 |
-
"learning_rate":
|
| 1339 |
-
"loss": 7.
|
| 1340 |
"step": 190
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
-
"epoch":
|
| 1344 |
-
"grad_norm":
|
| 1345 |
-
"learning_rate":
|
| 1346 |
-
"loss": 8.
|
| 1347 |
"step": 191
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
-
"epoch":
|
| 1351 |
-
"grad_norm":
|
| 1352 |
-
"learning_rate":
|
| 1353 |
-
"loss": 7.
|
| 1354 |
"step": 192
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
-
"epoch":
|
| 1358 |
-
"grad_norm":
|
| 1359 |
-
"learning_rate":
|
| 1360 |
-
"loss":
|
| 1361 |
"step": 193
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
-
"epoch":
|
| 1365 |
-
"grad_norm":
|
| 1366 |
-
"learning_rate":
|
| 1367 |
-
"loss":
|
| 1368 |
"step": 194
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
-
"epoch":
|
| 1372 |
-
"grad_norm":
|
| 1373 |
-
"learning_rate":
|
| 1374 |
-
"loss":
|
| 1375 |
"step": 195
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
-
"epoch":
|
| 1379 |
-
"grad_norm":
|
| 1380 |
-
"learning_rate":
|
| 1381 |
-
"loss": 7.
|
| 1382 |
"step": 196
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
-
"epoch":
|
| 1386 |
-
"grad_norm":
|
| 1387 |
-
"learning_rate":
|
| 1388 |
-
"loss":
|
| 1389 |
"step": 197
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
-
"epoch":
|
| 1393 |
-
"grad_norm":
|
| 1394 |
-
"learning_rate":
|
| 1395 |
-
"loss":
|
| 1396 |
"step": 198
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
-
"epoch":
|
| 1400 |
-
"grad_norm":
|
| 1401 |
-
"learning_rate":
|
| 1402 |
-
"loss":
|
| 1403 |
"step": 199
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
-
"epoch":
|
| 1407 |
-
"grad_norm":
|
| 1408 |
-
"learning_rate":
|
| 1409 |
-
"loss":
|
| 1410 |
"step": 200
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 4.91358024691358,
|
| 1414 |
-
"grad_norm": 5.40625,
|
| 1415 |
-
"learning_rate": 0.0005624262376922163,
|
| 1416 |
-
"loss": 6.8278,
|
| 1417 |
-
"step": 201
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 4.938271604938271,
|
| 1421 |
-
"grad_norm": 5.59375,
|
| 1422 |
-
"learning_rate": 0.0005584178049718315,
|
| 1423 |
-
"loss": 6.1303,
|
| 1424 |
-
"step": 202
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 4.962962962962963,
|
| 1428 |
-
"grad_norm": 3.75,
|
| 1429 |
-
"learning_rate": 0.0005544055620896629,
|
| 1430 |
-
"loss": 9.2713,
|
| 1431 |
-
"step": 203
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 4.987654320987654,
|
| 1435 |
-
"grad_norm": 4.34375,
|
| 1436 |
-
"learning_rate": 0.000550389770734668,
|
| 1437 |
-
"loss": 7.6376,
|
| 1438 |
-
"step": 204
|
| 1439 |
-
},
|
| 1440 |
-
{
|
| 1441 |
-
"epoch": 5.0,
|
| 1442 |
-
"grad_norm": 4.9375,
|
| 1443 |
-
"learning_rate": 0.0005463706928272452,
|
| 1444 |
-
"loss": 8.5016,
|
| 1445 |
-
"step": 205
|
| 1446 |
-
},
|
| 1447 |
-
{
|
| 1448 |
-
"epoch": 5.0246913580246915,
|
| 1449 |
-
"grad_norm": 3.765625,
|
| 1450 |
-
"learning_rate": 0.0005423485905021508,
|
| 1451 |
-
"loss": 7.1441,
|
| 1452 |
-
"step": 206
|
| 1453 |
-
},
|
| 1454 |
-
{
|
| 1455 |
-
"epoch": 5.049382716049383,
|
| 1456 |
-
"grad_norm": 3.9375,
|
| 1457 |
-
"learning_rate": 0.000538323726091401,
|
| 1458 |
-
"loss": 6.4224,
|
| 1459 |
-
"step": 207
|
| 1460 |
-
},
|
| 1461 |
-
{
|
| 1462 |
-
"epoch": 5.074074074074074,
|
| 1463 |
-
"grad_norm": 3.984375,
|
| 1464 |
-
"learning_rate": 0.0005342963621071624,
|
| 1465 |
-
"loss": 5.6542,
|
| 1466 |
-
"step": 208
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 5.098765432098766,
|
| 1470 |
-
"grad_norm": 4.0625,
|
| 1471 |
-
"learning_rate": 0.0005302667612246308,
|
| 1472 |
-
"loss": 8.4893,
|
| 1473 |
-
"step": 209
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 5.1234567901234565,
|
| 1477 |
-
"grad_norm": 3.53125,
|
| 1478 |
-
"learning_rate": 0.0005262351862648978,
|
| 1479 |
-
"loss": 7.3327,
|
| 1480 |
-
"step": 210
|
| 1481 |
-
},
|
| 1482 |
-
{
|
| 1483 |
-
"epoch": 5.148148148148148,
|
| 1484 |
-
"grad_norm": 4.40625,
|
| 1485 |
-
"learning_rate": 0.00052220190017781,
|
| 1486 |
-
"loss": 6.0517,
|
| 1487 |
-
"step": 211
|
| 1488 |
-
},
|
| 1489 |
-
{
|
| 1490 |
-
"epoch": 5.172839506172839,
|
| 1491 |
-
"grad_norm": 5.84375,
|
| 1492 |
-
"learning_rate": 0.0005181671660248179,
|
| 1493 |
-
"loss": 7.6619,
|
| 1494 |
-
"step": 212
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 5.197530864197531,
|
| 1498 |
-
"grad_norm": 4.0625,
|
| 1499 |
-
"learning_rate": 0.0005141312469618183,
|
| 1500 |
-
"loss": 8.5013,
|
| 1501 |
-
"step": 213
|
| 1502 |
-
},
|
| 1503 |
-
{
|
| 1504 |
-
"epoch": 5.222222222222222,
|
| 1505 |
-
"grad_norm": 3.546875,
|
| 1506 |
-
"learning_rate": 0.0005100944062219918,
|
| 1507 |
-
"loss": 7.2512,
|
| 1508 |
-
"step": 214
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 5.246913580246914,
|
| 1512 |
-
"grad_norm": 3.75,
|
| 1513 |
-
"learning_rate": 0.0005060569070986324,
|
| 1514 |
-
"loss": 8.8302,
|
| 1515 |
-
"step": 215
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 5.271604938271605,
|
| 1519 |
-
"grad_norm": 4.40625,
|
| 1520 |
-
"learning_rate": 0.0005020190129279759,
|
| 1521 |
-
"loss": 8.1647,
|
| 1522 |
-
"step": 216
|
| 1523 |
-
},
|
| 1524 |
-
{
|
| 1525 |
-
"epoch": 5.296296296296296,
|
| 1526 |
-
"grad_norm": 5.125,
|
| 1527 |
-
"learning_rate": 0.0004979809870720242,
|
| 1528 |
-
"loss": 4.7292,
|
| 1529 |
-
"step": 217
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 5.320987654320987,
|
| 1533 |
-
"grad_norm": 3.46875,
|
| 1534 |
-
"learning_rate": 0.0004939430929013677,
|
| 1535 |
-
"loss": 8.1476,
|
| 1536 |
-
"step": 218
|
| 1537 |
-
},
|
| 1538 |
-
{
|
| 1539 |
-
"epoch": 5.345679012345679,
|
| 1540 |
-
"grad_norm": 3.953125,
|
| 1541 |
-
"learning_rate": 0.0004899055937780083,
|
| 1542 |
-
"loss": 8.0866,
|
| 1543 |
-
"step": 219
|
| 1544 |
-
},
|
| 1545 |
-
{
|
| 1546 |
-
"epoch": 5.37037037037037,
|
| 1547 |
-
"grad_norm": 3.828125,
|
| 1548 |
-
"learning_rate": 0.00048586875303818176,
|
| 1549 |
-
"loss": 8.235,
|
| 1550 |
-
"step": 220
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 5.395061728395062,
|
| 1554 |
-
"grad_norm": 4.25,
|
| 1555 |
-
"learning_rate": 0.00048183283397518233,
|
| 1556 |
-
"loss": 8.2042,
|
| 1557 |
-
"step": 221
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 5.419753086419753,
|
| 1561 |
-
"grad_norm": 3.921875,
|
| 1562 |
-
"learning_rate": 0.0004777980998221901,
|
| 1563 |
-
"loss": 7.2502,
|
| 1564 |
-
"step": 222
|
| 1565 |
-
},
|
| 1566 |
-
{
|
| 1567 |
-
"epoch": 5.444444444444445,
|
| 1568 |
-
"grad_norm": 3.671875,
|
| 1569 |
-
"learning_rate": 0.0004737648137351022,
|
| 1570 |
-
"loss": 8.9586,
|
| 1571 |
-
"step": 223
|
| 1572 |
-
},
|
| 1573 |
-
{
|
| 1574 |
-
"epoch": 5.469135802469136,
|
| 1575 |
-
"grad_norm": 3.9375,
|
| 1576 |
-
"learning_rate": 0.0004697332387753693,
|
| 1577 |
-
"loss": 7.2468,
|
| 1578 |
-
"step": 224
|
| 1579 |
-
},
|
| 1580 |
-
{
|
| 1581 |
-
"epoch": 5.493827160493828,
|
| 1582 |
-
"grad_norm": 3.53125,
|
| 1583 |
-
"learning_rate": 0.0004657036378928376,
|
| 1584 |
-
"loss": 7.3585,
|
| 1585 |
-
"step": 225
|
| 1586 |
-
},
|
| 1587 |
-
{
|
| 1588 |
-
"epoch": 5.518518518518518,
|
| 1589 |
-
"grad_norm": 4.46875,
|
| 1590 |
-
"learning_rate": 0.00046167627390859926,
|
| 1591 |
-
"loss": 6.9372,
|
| 1592 |
-
"step": 226
|
| 1593 |
-
},
|
| 1594 |
-
{
|
| 1595 |
-
"epoch": 5.54320987654321,
|
| 1596 |
-
"grad_norm": 3.546875,
|
| 1597 |
-
"learning_rate": 0.00045765140949784926,
|
| 1598 |
-
"loss": 8.0293,
|
| 1599 |
-
"step": 227
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 5.567901234567901,
|
| 1603 |
-
"grad_norm": 3.484375,
|
| 1604 |
-
"learning_rate": 0.0004536293071727548,
|
| 1605 |
-
"loss": 9.1196,
|
| 1606 |
-
"step": 228
|
| 1607 |
-
},
|
| 1608 |
-
{
|
| 1609 |
-
"epoch": 5.592592592592593,
|
| 1610 |
-
"grad_norm": 4.09375,
|
| 1611 |
-
"learning_rate": 0.000449610229265332,
|
| 1612 |
-
"loss": 7.2503,
|
| 1613 |
-
"step": 229
|
| 1614 |
-
},
|
| 1615 |
-
{
|
| 1616 |
-
"epoch": 5.617283950617284,
|
| 1617 |
-
"grad_norm": 4.75,
|
| 1618 |
-
"learning_rate": 0.00044559443791033734,
|
| 1619 |
-
"loss": 5.6725,
|
| 1620 |
-
"step": 230
|
| 1621 |
-
},
|
| 1622 |
-
{
|
| 1623 |
-
"epoch": 5.6419753086419755,
|
| 1624 |
-
"grad_norm": 4.1875,
|
| 1625 |
-
"learning_rate": 0.0004415821950281686,
|
| 1626 |
-
"loss": 7.2648,
|
| 1627 |
-
"step": 231
|
| 1628 |
-
},
|
| 1629 |
-
{
|
| 1630 |
-
"epoch": 5.666666666666667,
|
| 1631 |
-
"grad_norm": 4.21875,
|
| 1632 |
-
"learning_rate": 0.00043757376230778387,
|
| 1633 |
-
"loss": 7.2724,
|
| 1634 |
-
"step": 232
|
| 1635 |
-
},
|
| 1636 |
-
{
|
| 1637 |
-
"epoch": 5.6913580246913575,
|
| 1638 |
-
"grad_norm": 3.828125,
|
| 1639 |
-
"learning_rate": 0.00043356940118963205,
|
| 1640 |
-
"loss": 8.7947,
|
| 1641 |
-
"step": 233
|
| 1642 |
-
},
|
| 1643 |
-
{
|
| 1644 |
-
"epoch": 5.716049382716049,
|
| 1645 |
-
"grad_norm": 3.75,
|
| 1646 |
-
"learning_rate": 0.00042956937284860196,
|
| 1647 |
-
"loss": 8.0334,
|
| 1648 |
-
"step": 234
|
| 1649 |
-
},
|
| 1650 |
-
{
|
| 1651 |
-
"epoch": 5.7407407407407405,
|
| 1652 |
-
"grad_norm": 4.8125,
|
| 1653 |
-
"learning_rate": 0.0004255739381769868,
|
| 1654 |
-
"loss": 7.9606,
|
| 1655 |
-
"step": 235
|
| 1656 |
-
},
|
| 1657 |
-
{
|
| 1658 |
-
"epoch": 5.765432098765432,
|
| 1659 |
-
"grad_norm": 4.09375,
|
| 1660 |
-
"learning_rate": 0.00042158335776746903,
|
| 1661 |
-
"loss": 6.513,
|
| 1662 |
-
"step": 236
|
| 1663 |
-
},
|
| 1664 |
-
{
|
| 1665 |
-
"epoch": 5.790123456790123,
|
| 1666 |
-
"grad_norm": 3.59375,
|
| 1667 |
-
"learning_rate": 0.00041759789189612333,
|
| 1668 |
-
"loss": 8.1531,
|
| 1669 |
-
"step": 237
|
| 1670 |
-
},
|
| 1671 |
-
{
|
| 1672 |
-
"epoch": 5.814814814814815,
|
| 1673 |
-
"grad_norm": 4.78125,
|
| 1674 |
-
"learning_rate": 0.00041361780050544104,
|
| 1675 |
-
"loss": 7.1614,
|
| 1676 |
-
"step": 238
|
| 1677 |
-
},
|
| 1678 |
-
{
|
| 1679 |
-
"epoch": 5.839506172839506,
|
| 1680 |
-
"grad_norm": 3.71875,
|
| 1681 |
-
"learning_rate": 0.0004096433431873757,
|
| 1682 |
-
"loss": 8.6922,
|
| 1683 |
-
"step": 239
|
| 1684 |
-
},
|
| 1685 |
-
{
|
| 1686 |
-
"epoch": 5.864197530864198,
|
| 1687 |
-
"grad_norm": 3.40625,
|
| 1688 |
-
"learning_rate": 0.00040567477916641164,
|
| 1689 |
-
"loss": 8.9935,
|
| 1690 |
-
"step": 240
|
| 1691 |
-
},
|
| 1692 |
-
{
|
| 1693 |
-
"epoch": 5.888888888888889,
|
| 1694 |
-
"grad_norm": 4.5,
|
| 1695 |
-
"learning_rate": 0.0004017123672826574,
|
| 1696 |
-
"loss": 8.3437,
|
| 1697 |
-
"step": 241
|
| 1698 |
-
},
|
| 1699 |
-
{
|
| 1700 |
-
"epoch": 5.91358024691358,
|
| 1701 |
-
"grad_norm": 3.453125,
|
| 1702 |
-
"learning_rate": 0.0003977563659749629,
|
| 1703 |
-
"loss": 8.1862,
|
| 1704 |
-
"step": 242
|
| 1705 |
-
},
|
| 1706 |
-
{
|
| 1707 |
-
"epoch": 5.938271604938271,
|
| 1708 |
-
"grad_norm": 3.84375,
|
| 1709 |
-
"learning_rate": 0.0003938070332640636,
|
| 1710 |
-
"loss": 7.3413,
|
| 1711 |
-
"step": 243
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 5.962962962962963,
|
| 1715 |
-
"grad_norm": 4.0,
|
| 1716 |
-
"learning_rate": 0.00038986462673575153,
|
| 1717 |
-
"loss": 6.4921,
|
| 1718 |
-
"step": 244
|
| 1719 |
-
},
|
| 1720 |
-
{
|
| 1721 |
-
"epoch": 5.987654320987654,
|
| 1722 |
-
"grad_norm": 3.890625,
|
| 1723 |
-
"learning_rate": 0.000385929403524075,
|
| 1724 |
-
"loss": 8.07,
|
| 1725 |
-
"step": 245
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 6.0,
|
| 1729 |
-
"grad_norm": 5.84375,
|
| 1730 |
-
"learning_rate": 0.00038200162029456797,
|
| 1731 |
-
"loss": 6.4634,
|
| 1732 |
-
"step": 246
|
| 1733 |
-
},
|
| 1734 |
-
{
|
| 1735 |
-
"epoch": 6.0246913580246915,
|
| 1736 |
-
"grad_norm": 3.78125,
|
| 1737 |
-
"learning_rate": 0.0003780815332275089,
|
| 1738 |
-
"loss": 6.1432,
|
| 1739 |
-
"step": 247
|
| 1740 |
-
},
|
| 1741 |
-
{
|
| 1742 |
-
"epoch": 6.049382716049383,
|
| 1743 |
-
"grad_norm": 3.734375,
|
| 1744 |
-
"learning_rate": 0.00037416939800121303,
|
| 1745 |
-
"loss": 6.2527,
|
| 1746 |
-
"step": 248
|
| 1747 |
-
},
|
| 1748 |
-
{
|
| 1749 |
-
"epoch": 6.074074074074074,
|
| 1750 |
-
"grad_norm": 3.453125,
|
| 1751 |
-
"learning_rate": 0.00037026546977535484,
|
| 1752 |
-
"loss": 8.8918,
|
| 1753 |
-
"step": 249
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 6.098765432098766,
|
| 1757 |
-
"grad_norm": 3.65625,
|
| 1758 |
-
"learning_rate": 0.0003663700031743277,
|
| 1759 |
-
"loss": 7.2045,
|
| 1760 |
-
"step": 250
|
| 1761 |
-
},
|
| 1762 |
-
{
|
| 1763 |
-
"epoch": 6.1234567901234565,
|
| 1764 |
-
"grad_norm": 3.796875,
|
| 1765 |
-
"learning_rate": 0.0003624832522706352,
|
| 1766 |
-
"loss": 5.8116,
|
| 1767 |
-
"step": 251
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 6.148148148148148,
|
| 1771 |
-
"grad_norm": 3.34375,
|
| 1772 |
-
"learning_rate": 0.0003586054705683208,
|
| 1773 |
-
"loss": 8.6758,
|
| 1774 |
-
"step": 252
|
| 1775 |
-
},
|
| 1776 |
-
{
|
| 1777 |
-
"epoch": 6.172839506172839,
|
| 1778 |
-
"grad_norm": 3.671875,
|
| 1779 |
-
"learning_rate": 0.0003547369109864328,
|
| 1780 |
-
"loss": 7.0771,
|
| 1781 |
-
"step": 253
|
| 1782 |
-
},
|
| 1783 |
-
{
|
| 1784 |
-
"epoch": 6.197530864197531,
|
| 1785 |
-
"grad_norm": 3.71875,
|
| 1786 |
-
"learning_rate": 0.00035087782584252905,
|
| 1787 |
-
"loss": 6.3009,
|
| 1788 |
-
"step": 254
|
| 1789 |
-
},
|
| 1790 |
-
{
|
| 1791 |
-
"epoch": 6.222222222222222,
|
| 1792 |
-
"grad_norm": 3.65625,
|
| 1793 |
-
"learning_rate": 0.00034702846683621946,
|
| 1794 |
-
"loss": 7.8653,
|
| 1795 |
-
"step": 255
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 6.246913580246914,
|
| 1799 |
-
"grad_norm": 3.4375,
|
| 1800 |
-
"learning_rate": 0.00034318908503275057,
|
| 1801 |
-
"loss": 5.3887,
|
| 1802 |
-
"step": 256
|
| 1803 |
-
},
|
| 1804 |
-
{
|
| 1805 |
-
"epoch": 6.271604938271605,
|
| 1806 |
-
"grad_norm": 3.265625,
|
| 1807 |
-
"learning_rate": 0.0003393599308466285,
|
| 1808 |
-
"loss": 7.9237,
|
| 1809 |
-
"step": 257
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 6.296296296296296,
|
| 1813 |
-
"grad_norm": 3.40625,
|
| 1814 |
-
"learning_rate": 0.0003355412540252887,
|
| 1815 |
-
"loss": 8.8507,
|
| 1816 |
-
"step": 258
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 6.320987654320987,
|
| 1820 |
-
"grad_norm": 4.21875,
|
| 1821 |
-
"learning_rate": 0.00033173330363280484,
|
| 1822 |
-
"loss": 6.5917,
|
| 1823 |
-
"step": 259
|
| 1824 |
-
},
|
| 1825 |
-
{
|
| 1826 |
-
"epoch": 6.345679012345679,
|
| 1827 |
-
"grad_norm": 3.21875,
|
| 1828 |
-
"learning_rate": 0.0003279363280336449,
|
| 1829 |
-
"loss": 8.7578,
|
| 1830 |
-
"step": 260
|
| 1831 |
-
},
|
| 1832 |
-
{
|
| 1833 |
-
"epoch": 6.37037037037037,
|
| 1834 |
-
"grad_norm": 3.546875,
|
| 1835 |
-
"learning_rate": 0.0003241505748764728,
|
| 1836 |
-
"loss": 6.149,
|
| 1837 |
-
"step": 261
|
| 1838 |
-
},
|
| 1839 |
-
{
|
| 1840 |
-
"epoch": 6.395061728395062,
|
| 1841 |
-
"grad_norm": 3.90625,
|
| 1842 |
-
"learning_rate": 0.0003203762910779944,
|
| 1843 |
-
"loss": 7.1028,
|
| 1844 |
-
"step": 262
|
| 1845 |
-
},
|
| 1846 |
-
{
|
| 1847 |
-
"epoch": 6.419753086419753,
|
| 1848 |
-
"grad_norm": 3.53125,
|
| 1849 |
-
"learning_rate": 0.00031661372280685483,
|
| 1850 |
-
"loss": 6.1741,
|
| 1851 |
-
"step": 263
|
| 1852 |
-
},
|
| 1853 |
-
{
|
| 1854 |
-
"epoch": 6.444444444444445,
|
| 1855 |
-
"grad_norm": 3.265625,
|
| 1856 |
-
"learning_rate": 0.0003128631154675818,
|
| 1857 |
-
"loss": 7.9666,
|
| 1858 |
-
"step": 264
|
| 1859 |
-
},
|
| 1860 |
-
{
|
| 1861 |
-
"epoch": 6.469135802469136,
|
| 1862 |
-
"grad_norm": 3.328125,
|
| 1863 |
-
"learning_rate": 0.00030912471368458005,
|
| 1864 |
-
"loss": 8.0334,
|
| 1865 |
-
"step": 265
|
| 1866 |
-
},
|
| 1867 |
-
{
|
| 1868 |
-
"epoch": 6.493827160493828,
|
| 1869 |
-
"grad_norm": 3.390625,
|
| 1870 |
-
"learning_rate": 0.0003053987612861753,
|
| 1871 |
-
"loss": 9.5293,
|
| 1872 |
-
"step": 266
|
| 1873 |
-
},
|
| 1874 |
-
{
|
| 1875 |
-
"epoch": 6.518518518518518,
|
| 1876 |
-
"grad_norm": 4.09375,
|
| 1877 |
-
"learning_rate": 0.00030168550128871263,
|
| 1878 |
-
"loss": 6.234,
|
| 1879 |
-
"step": 267
|
| 1880 |
-
},
|
| 1881 |
-
{
|
| 1882 |
-
"epoch": 6.54320987654321,
|
| 1883 |
-
"grad_norm": 3.390625,
|
| 1884 |
-
"learning_rate": 0.00029798517588070536,
|
| 1885 |
-
"loss": 7.8696,
|
| 1886 |
-
"step": 268
|
| 1887 |
-
},
|
| 1888 |
-
{
|
| 1889 |
-
"epoch": 6.567901234567901,
|
| 1890 |
-
"grad_norm": 3.578125,
|
| 1891 |
-
"learning_rate": 0.0002942980264070392,
|
| 1892 |
-
"loss": 7.3372,
|
| 1893 |
-
"step": 269
|
| 1894 |
-
},
|
| 1895 |
-
{
|
| 1896 |
-
"epoch": 6.592592592592593,
|
| 1897 |
-
"grad_norm": 3.515625,
|
| 1898 |
-
"learning_rate": 0.0002906242933532304,
|
| 1899 |
-
"loss": 7.0152,
|
| 1900 |
-
"step": 270
|
| 1901 |
-
},
|
| 1902 |
-
{
|
| 1903 |
-
"epoch": 6.617283950617284,
|
| 1904 |
-
"grad_norm": 3.953125,
|
| 1905 |
-
"learning_rate": 0.0002869642163297422,
|
| 1906 |
-
"loss": 7.0756,
|
| 1907 |
-
"step": 271
|
| 1908 |
-
},
|
| 1909 |
-
{
|
| 1910 |
-
"epoch": 6.6419753086419755,
|
| 1911 |
-
"grad_norm": 3.328125,
|
| 1912 |
-
"learning_rate": 0.0002833180340563554,
|
| 1913 |
-
"loss": 7.9405,
|
| 1914 |
-
"step": 272
|
| 1915 |
-
},
|
| 1916 |
-
{
|
| 1917 |
-
"epoch": 6.666666666666667,
|
| 1918 |
-
"grad_norm": 3.5,
|
| 1919 |
-
"learning_rate": 0.00027968598434659897,
|
| 1920 |
-
"loss": 8.7255,
|
| 1921 |
-
"step": 273
|
| 1922 |
-
},
|
| 1923 |
-
{
|
| 1924 |
-
"epoch": 6.6913580246913575,
|
| 1925 |
-
"grad_norm": 3.359375,
|
| 1926 |
-
"learning_rate": 0.00027606830409223937,
|
| 1927 |
-
"loss": 8.9699,
|
| 1928 |
-
"step": 274
|
| 1929 |
-
},
|
| 1930 |
-
{
|
| 1931 |
-
"epoch": 6.716049382716049,
|
| 1932 |
-
"grad_norm": 3.9375,
|
| 1933 |
-
"learning_rate": 0.00027246522924782905,
|
| 1934 |
-
"loss": 8.8373,
|
| 1935 |
-
"step": 275
|
| 1936 |
-
},
|
| 1937 |
-
{
|
| 1938 |
-
"epoch": 6.7407407407407405,
|
| 1939 |
-
"grad_norm": 3.421875,
|
| 1940 |
-
"learning_rate": 0.00026887699481531815,
|
| 1941 |
-
"loss": 7.9459,
|
| 1942 |
-
"step": 276
|
| 1943 |
-
},
|
| 1944 |
-
{
|
| 1945 |
-
"epoch": 6.765432098765432,
|
| 1946 |
-
"grad_norm": 4.03125,
|
| 1947 |
-
"learning_rate": 0.0002653038348287261,
|
| 1948 |
-
"loss": 6.8735,
|
| 1949 |
-
"step": 277
|
| 1950 |
-
},
|
| 1951 |
-
{
|
| 1952 |
-
"epoch": 6.790123456790123,
|
| 1953 |
-
"grad_norm": 3.65625,
|
| 1954 |
-
"learning_rate": 0.0002617459823388778,
|
| 1955 |
-
"loss": 6.9867,
|
| 1956 |
-
"step": 278
|
| 1957 |
-
},
|
| 1958 |
-
{
|
| 1959 |
-
"epoch": 6.814814814814815,
|
| 1960 |
-
"grad_norm": 3.796875,
|
| 1961 |
-
"learning_rate": 0.0002582036693982028,
|
| 1962 |
-
"loss": 7.8286,
|
| 1963 |
-
"step": 279
|
| 1964 |
-
},
|
| 1965 |
-
{
|
| 1966 |
-
"epoch": 6.839506172839506,
|
| 1967 |
-
"grad_norm": 3.9375,
|
| 1968 |
-
"learning_rate": 0.00025467712704560117,
|
| 1969 |
-
"loss": 6.0116,
|
| 1970 |
-
"step": 280
|
| 1971 |
-
},
|
| 1972 |
-
{
|
| 1973 |
-
"epoch": 6.864197530864198,
|
| 1974 |
-
"grad_norm": 4.0,
|
| 1975 |
-
"learning_rate": 0.00025116658529137384,
|
| 1976 |
-
"loss": 8.3119,
|
| 1977 |
-
"step": 281
|
| 1978 |
-
},
|
| 1979 |
-
{
|
| 1980 |
-
"epoch": 6.888888888888889,
|
| 1981 |
-
"grad_norm": 3.609375,
|
| 1982 |
-
"learning_rate": 0.00024767227310222074,
|
| 1983 |
-
"loss": 7.1145,
|
| 1984 |
-
"step": 282
|
| 1985 |
-
},
|
| 1986 |
-
{
|
| 1987 |
-
"epoch": 6.91358024691358,
|
| 1988 |
-
"grad_norm": 3.921875,
|
| 1989 |
-
"learning_rate": 0.00024419441838630763,
|
| 1990 |
-
"loss": 6.5803,
|
| 1991 |
-
"step": 283
|
| 1992 |
-
},
|
| 1993 |
-
{
|
| 1994 |
-
"epoch": 6.938271604938271,
|
| 1995 |
-
"grad_norm": 4.0,
|
| 1996 |
-
"learning_rate": 0.0002407332479784005,
|
| 1997 |
-
"loss": 6.6936,
|
| 1998 |
-
"step": 284
|
| 1999 |
-
},
|
| 2000 |
-
{
|
| 2001 |
-
"epoch": 6.962962962962963,
|
| 2002 |
-
"grad_norm": 3.890625,
|
| 2003 |
-
"learning_rate": 0.00023728898762507146,
|
| 2004 |
-
"loss": 5.3851,
|
| 2005 |
-
"step": 285
|
| 2006 |
-
},
|
| 2007 |
-
{
|
| 2008 |
-
"epoch": 6.987654320987654,
|
| 2009 |
-
"grad_norm": 3.4375,
|
| 2010 |
-
"learning_rate": 0.00023386186196997443,
|
| 2011 |
-
"loss": 7.0678,
|
| 2012 |
-
"step": 286
|
| 2013 |
-
},
|
| 2014 |
-
{
|
| 2015 |
-
"epoch": 7.0,
|
| 2016 |
-
"grad_norm": 4.25,
|
| 2017 |
-
"learning_rate": 0.00023045209453919407,
|
| 2018 |
-
"loss": 9.7393,
|
| 2019 |
-
"step": 287
|
| 2020 |
-
},
|
| 2021 |
-
{
|
| 2022 |
-
"epoch": 7.0246913580246915,
|
| 2023 |
-
"grad_norm": 3.40625,
|
| 2024 |
-
"learning_rate": 0.00022705990772666558,
|
| 2025 |
-
"loss": 8.4961,
|
| 2026 |
-
"step": 288
|
| 2027 |
-
},
|
| 2028 |
-
{
|
| 2029 |
-
"epoch": 7.049382716049383,
|
| 2030 |
-
"grad_norm": 3.34375,
|
| 2031 |
-
"learning_rate": 0.00022368552277967081,
|
| 2032 |
-
"loss": 6.9223,
|
| 2033 |
-
"step": 289
|
| 2034 |
-
},
|
| 2035 |
-
{
|
| 2036 |
-
"epoch": 7.074074074074074,
|
| 2037 |
-
"grad_norm": 3.234375,
|
| 2038 |
-
"learning_rate": 0.00022032915978440748,
|
| 2039 |
-
"loss": 7.0141,
|
| 2040 |
-
"step": 290
|
| 2041 |
-
},
|
| 2042 |
-
{
|
| 2043 |
-
"epoch": 7.098765432098766,
|
| 2044 |
-
"grad_norm": 3.125,
|
| 2045 |
-
"learning_rate": 0.0002169910376516346,
|
| 2046 |
-
"loss": 7.7613,
|
| 2047 |
-
"step": 291
|
| 2048 |
-
},
|
| 2049 |
-
{
|
| 2050 |
-
"epoch": 7.1234567901234565,
|
| 2051 |
-
"grad_norm": 3.109375,
|
| 2052 |
-
"learning_rate": 0.00021367137410239402,
|
| 2053 |
-
"loss": 8.7319,
|
| 2054 |
-
"step": 292
|
| 2055 |
-
},
|
| 2056 |
-
{
|
| 2057 |
-
"epoch": 7.148148148148148,
|
| 2058 |
-
"grad_norm": 3.578125,
|
| 2059 |
-
"learning_rate": 0.00021037038565381106,
|
| 2060 |
-
"loss": 8.2093,
|
| 2061 |
-
"step": 293
|
| 2062 |
-
},
|
| 2063 |
-
{
|
| 2064 |
-
"epoch": 7.172839506172839,
|
| 2065 |
-
"grad_norm": 3.125,
|
| 2066 |
-
"learning_rate": 0.000207088287604972,
|
| 2067 |
-
"loss": 7.2345,
|
| 2068 |
-
"step": 294
|
| 2069 |
-
},
|
| 2070 |
-
{
|
| 2071 |
-
"epoch": 7.197530864197531,
|
| 2072 |
-
"grad_norm": 3.640625,
|
| 2073 |
-
"learning_rate": 0.00020382529402288209,
|
| 2074 |
-
"loss": 6.2056,
|
| 2075 |
-
"step": 295
|
| 2076 |
-
},
|
| 2077 |
-
{
|
| 2078 |
-
"epoch": 7.222222222222222,
|
| 2079 |
-
"grad_norm": 3.421875,
|
| 2080 |
-
"learning_rate": 0.00020058161772850336,
|
| 2081 |
-
"loss": 7.035,
|
| 2082 |
-
"step": 296
|
| 2083 |
-
},
|
| 2084 |
-
{
|
| 2085 |
-
"epoch": 7.246913580246914,
|
| 2086 |
-
"grad_norm": 3.078125,
|
| 2087 |
-
"learning_rate": 0.00019735747028287344,
|
| 2088 |
-
"loss": 8.6249,
|
| 2089 |
-
"step": 297
|
| 2090 |
-
},
|
| 2091 |
-
{
|
| 2092 |
-
"epoch": 7.271604938271605,
|
| 2093 |
-
"grad_norm": 3.859375,
|
| 2094 |
-
"learning_rate": 0.00019415306197330802,
|
| 2095 |
-
"loss": 7.2435,
|
| 2096 |
-
"step": 298
|
| 2097 |
-
},
|
| 2098 |
-
{
|
| 2099 |
-
"epoch": 7.296296296296296,
|
| 2100 |
-
"grad_norm": 4.6875,
|
| 2101 |
-
"learning_rate": 0.0001909686017996845,
|
| 2102 |
-
"loss": 3.2569,
|
| 2103 |
-
"step": 299
|
| 2104 |
-
},
|
| 2105 |
-
{
|
| 2106 |
-
"epoch": 7.320987654320987,
|
| 2107 |
-
"grad_norm": 3.734375,
|
| 2108 |
-
"learning_rate": 0.0001878042974608109,
|
| 2109 |
-
"loss": 7.8251,
|
| 2110 |
-
"step": 300
|
| 2111 |
-
},
|
| 2112 |
-
{
|
| 2113 |
-
"epoch": 7.345679012345679,
|
| 2114 |
-
"grad_norm": 3.921875,
|
| 2115 |
-
"learning_rate": 0.000184660355340879,
|
| 2116 |
-
"loss": 5.5533,
|
| 2117 |
-
"step": 301
|
| 2118 |
-
},
|
| 2119 |
-
{
|
| 2120 |
-
"epoch": 7.37037037037037,
|
| 2121 |
-
"grad_norm": 3.421875,
|
| 2122 |
-
"learning_rate": 0.00018153698049600342,
|
| 2123 |
-
"loss": 6.109,
|
| 2124 |
-
"step": 302
|
| 2125 |
-
},
|
| 2126 |
-
{
|
| 2127 |
-
"epoch": 7.395061728395062,
|
| 2128 |
-
"grad_norm": 3.25,
|
| 2129 |
-
"learning_rate": 0.00017843437664084754,
|
| 2130 |
-
"loss": 7.0764,
|
| 2131 |
-
"step": 303
|
| 2132 |
-
},
|
| 2133 |
-
{
|
| 2134 |
-
"epoch": 7.419753086419753,
|
| 2135 |
-
"grad_norm": 3.34375,
|
| 2136 |
-
"learning_rate": 0.00017535274613533663,
|
| 2137 |
-
"loss": 7.7965,
|
| 2138 |
-
"step": 304
|
| 2139 |
-
},
|
| 2140 |
-
{
|
| 2141 |
-
"epoch": 7.444444444444445,
|
| 2142 |
-
"grad_norm": 3.03125,
|
| 2143 |
-
"learning_rate": 0.0001722922899714588,
|
| 2144 |
-
"loss": 8.0387,
|
| 2145 |
-
"step": 305
|
| 2146 |
-
},
|
| 2147 |
-
{
|
| 2148 |
-
"epoch": 7.469135802469136,
|
| 2149 |
-
"grad_norm": 3.609375,
|
| 2150 |
-
"learning_rate": 0.0001692532077601569,
|
| 2151 |
-
"loss": 5.9805,
|
| 2152 |
-
"step": 306
|
| 2153 |
-
},
|
| 2154 |
-
{
|
| 2155 |
-
"epoch": 7.493827160493828,
|
| 2156 |
-
"grad_norm": 3.328125,
|
| 2157 |
-
"learning_rate": 0.00016623569771830854,
|
| 2158 |
-
"loss": 6.0428,
|
| 2159 |
-
"step": 307
|
| 2160 |
-
},
|
| 2161 |
-
{
|
| 2162 |
-
"epoch": 7.518518518518518,
|
| 2163 |
-
"grad_norm": 3.171875,
|
| 2164 |
-
"learning_rate": 0.00016323995665579817,
|
| 2165 |
-
"loss": 6.9212,
|
| 2166 |
-
"step": 308
|
| 2167 |
-
},
|
| 2168 |
-
{
|
| 2169 |
-
"epoch": 7.54320987654321,
|
| 2170 |
-
"grad_norm": 3.484375,
|
| 2171 |
-
"learning_rate": 0.00016026617996268066,
|
| 2172 |
-
"loss": 8.4905,
|
| 2173 |
-
"step": 309
|
| 2174 |
-
},
|
| 2175 |
-
{
|
| 2176 |
-
"epoch": 7.567901234567901,
|
| 2177 |
-
"grad_norm": 3.109375,
|
| 2178 |
-
"learning_rate": 0.00015731456159643687,
|
| 2179 |
-
"loss": 6.9324,
|
| 2180 |
-
"step": 310
|
| 2181 |
-
},
|
| 2182 |
-
{
|
| 2183 |
-
"epoch": 7.592592592592593,
|
| 2184 |
-
"grad_norm": 3.375,
|
| 2185 |
-
"learning_rate": 0.00015438529406932428,
|
| 2186 |
-
"loss": 7.8715,
|
| 2187 |
-
"step": 311
|
| 2188 |
-
},
|
| 2189 |
-
{
|
| 2190 |
-
"epoch": 7.617283950617284,
|
| 2191 |
-
"grad_norm": 3.265625,
|
| 2192 |
-
"learning_rate": 0.00015147856843582004,
|
| 2193 |
-
"loss": 6.2495,
|
| 2194 |
-
"step": 312
|
| 2195 |
-
},
|
| 2196 |
-
{
|
| 2197 |
-
"epoch": 7.6419753086419755,
|
| 2198 |
-
"grad_norm": 3.515625,
|
| 2199 |
-
"learning_rate": 0.00014859457428016015,
|
| 2200 |
-
"loss": 6.6685,
|
| 2201 |
-
"step": 313
|
| 2202 |
-
},
|
| 2203 |
-
{
|
| 2204 |
-
"epoch": 7.666666666666667,
|
| 2205 |
-
"grad_norm": 3.34375,
|
| 2206 |
-
"learning_rate": 0.00014573349970397432,
|
| 2207 |
-
"loss": 6.9555,
|
| 2208 |
-
"step": 314
|
| 2209 |
-
},
|
| 2210 |
-
{
|
| 2211 |
-
"epoch": 7.6913580246913575,
|
| 2212 |
-
"grad_norm": 3.328125,
|
| 2213 |
-
"learning_rate": 0.00014289553131401734,
|
| 2214 |
-
"loss": 7.19,
|
| 2215 |
-
"step": 315
|
| 2216 |
-
},
|
| 2217 |
-
{
|
| 2218 |
-
"epoch": 7.716049382716049,
|
| 2219 |
-
"grad_norm": 3.328125,
|
| 2220 |
-
"learning_rate": 0.00014008085420999834,
|
| 2221 |
-
"loss": 6.6798,
|
| 2222 |
-
"step": 316
|
| 2223 |
-
},
|
| 2224 |
-
{
|
| 2225 |
-
"epoch": 7.7407407407407405,
|
| 2226 |
-
"grad_norm": 3.09375,
|
| 2227 |
-
"learning_rate": 0.00013728965197250782,
|
| 2228 |
-
"loss": 9.4656,
|
| 2229 |
-
"step": 317
|
| 2230 |
-
},
|
| 2231 |
-
{
|
| 2232 |
-
"epoch": 7.765432098765432,
|
| 2233 |
-
"grad_norm": 3.328125,
|
| 2234 |
-
"learning_rate": 0.00013452210665104442,
|
| 2235 |
-
"loss": 7.7878,
|
| 2236 |
-
"step": 318
|
| 2237 |
-
},
|
| 2238 |
-
{
|
| 2239 |
-
"epoch": 7.790123456790123,
|
| 2240 |
-
"grad_norm": 3.484375,
|
| 2241 |
-
"learning_rate": 0.00013177839875214042,
|
| 2242 |
-
"loss": 6.3782,
|
| 2243 |
-
"step": 319
|
| 2244 |
-
},
|
| 2245 |
-
{
|
| 2246 |
-
"epoch": 7.814814814814815,
|
| 2247 |
-
"grad_norm": 3.109375,
|
| 2248 |
-
"learning_rate": 0.00012905870722758973,
|
| 2249 |
-
"loss": 7.0306,
|
| 2250 |
-
"step": 320
|
| 2251 |
-
},
|
| 2252 |
-
{
|
| 2253 |
-
"epoch": 7.839506172839506,
|
| 2254 |
-
"grad_norm": 3.046875,
|
| 2255 |
-
"learning_rate": 0.00012636320946277535,
|
| 2256 |
-
"loss": 8.773,
|
| 2257 |
-
"step": 321
|
| 2258 |
-
},
|
| 2259 |
-
{
|
| 2260 |
-
"epoch": 7.864197530864198,
|
| 2261 |
-
"grad_norm": 3.15625,
|
| 2262 |
-
"learning_rate": 0.0001236920812651003,
|
| 2263 |
-
"loss": 7.7132,
|
| 2264 |
-
"step": 322
|
| 2265 |
-
},
|
| 2266 |
-
{
|
| 2267 |
-
"epoch": 7.888888888888889,
|
| 2268 |
-
"grad_norm": 3.078125,
|
| 2269 |
-
"learning_rate": 0.00012104549685252026,
|
| 2270 |
-
"loss": 7.8435,
|
| 2271 |
-
"step": 323
|
| 2272 |
-
},
|
| 2273 |
-
{
|
| 2274 |
-
"epoch": 7.91358024691358,
|
| 2275 |
-
"grad_norm": 3.125,
|
| 2276 |
-
"learning_rate": 0.00011842362884218167,
|
| 2277 |
-
"loss": 8.5793,
|
| 2278 |
-
"step": 324
|
| 2279 |
-
},
|
| 2280 |
-
{
|
| 2281 |
-
"epoch": 7.938271604938271,
|
| 2282 |
-
"grad_norm": 3.453125,
|
| 2283 |
-
"learning_rate": 0.00011582664823916261,
|
| 2284 |
-
"loss": 6.9349,
|
| 2285 |
-
"step": 325
|
| 2286 |
-
},
|
| 2287 |
-
{
|
| 2288 |
-
"epoch": 7.962962962962963,
|
| 2289 |
-
"grad_norm": 3.359375,
|
| 2290 |
-
"learning_rate": 0.00011325472442531948,
|
| 2291 |
-
"loss": 7.377,
|
| 2292 |
-
"step": 326
|
| 2293 |
-
},
|
| 2294 |
-
{
|
| 2295 |
-
"epoch": 7.987654320987654,
|
| 2296 |
-
"grad_norm": 3.046875,
|
| 2297 |
-
"learning_rate": 0.00011070802514823913,
|
| 2298 |
-
"loss": 6.9828,
|
| 2299 |
-
"step": 327
|
| 2300 |
-
},
|
| 2301 |
-
{
|
| 2302 |
-
"epoch": 8.0,
|
| 2303 |
-
"grad_norm": 4.8125,
|
| 2304 |
-
"learning_rate": 0.00010818671651029855,
|
| 2305 |
-
"loss": 6.7613,
|
| 2306 |
-
"step": 328
|
| 2307 |
-
},
|
| 2308 |
-
{
|
| 2309 |
-
"epoch": 8.024691358024691,
|
| 2310 |
-
"grad_norm": 3.046875,
|
| 2311 |
-
"learning_rate": 0.00010569096295783088,
|
| 2312 |
-
"loss": 6.0772,
|
| 2313 |
-
"step": 329
|
| 2314 |
-
},
|
| 2315 |
-
{
|
| 2316 |
-
"epoch": 8.049382716049383,
|
| 2317 |
-
"grad_norm": 3.65625,
|
| 2318 |
-
"learning_rate": 0.00010322092727039989,
|
| 2319 |
-
"loss": 3.4719,
|
| 2320 |
-
"step": 330
|
| 2321 |
-
},
|
| 2322 |
-
{
|
| 2323 |
-
"epoch": 8.074074074074074,
|
| 2324 |
-
"grad_norm": 2.859375,
|
| 2325 |
-
"learning_rate": 0.00010077677055018303,
|
| 2326 |
-
"loss": 7.9138,
|
| 2327 |
-
"step": 331
|
| 2328 |
-
},
|
| 2329 |
-
{
|
| 2330 |
-
"epoch": 8.098765432098766,
|
| 2331 |
-
"grad_norm": 3.21875,
|
| 2332 |
-
"learning_rate": 9.83586522114639e-05,
|
| 2333 |
-
"loss": 7.2392,
|
| 2334 |
-
"step": 332
|
| 2335 |
-
},
|
| 2336 |
-
{
|
| 2337 |
-
"epoch": 8.123456790123457,
|
| 2338 |
-
"grad_norm": 3.078125,
|
| 2339 |
-
"learning_rate": 9.596672997023481e-05,
|
| 2340 |
-
"loss": 7.6866,
|
| 2341 |
-
"step": 333
|
| 2342 |
-
},
|
| 2343 |
-
{
|
| 2344 |
-
"epoch": 8.148148148148149,
|
| 2345 |
-
"grad_norm": 3.046875,
|
| 2346 |
-
"learning_rate": 9.360115983391032e-05,
|
| 2347 |
-
"loss": 7.9637,
|
| 2348 |
-
"step": 334
|
| 2349 |
-
},
|
| 2350 |
-
{
|
| 2351 |
-
"epoch": 8.17283950617284,
|
| 2352 |
-
"grad_norm": 3.15625,
|
| 2353 |
-
"learning_rate": 9.126209609115193e-05,
|
| 2354 |
-
"loss": 5.9937,
|
| 2355 |
-
"step": 335
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 8.197530864197532,
|
| 2359 |
-
"grad_norm": 3.34375,
|
| 2360 |
-
"learning_rate": 8.894969130180453e-05,
|
| 2361 |
-
"loss": 6.9129,
|
| 2362 |
-
"step": 336
|
| 2363 |
-
},
|
| 2364 |
-
{
|
| 2365 |
-
"epoch": 8.222222222222221,
|
| 2366 |
-
"grad_norm": 3.25,
|
| 2367 |
-
"learning_rate": 8.666409628694693e-05,
|
| 2368 |
-
"loss": 9.3835,
|
| 2369 |
-
"step": 337
|
| 2370 |
-
},
|
| 2371 |
-
{
|
| 2372 |
-
"epoch": 8.246913580246913,
|
| 2373 |
-
"grad_norm": 3.328125,
|
| 2374 |
-
"learning_rate": 8.440546011905426e-05,
|
| 2375 |
-
"loss": 7.1956,
|
| 2376 |
-
"step": 338
|
| 2377 |
-
},
|
| 2378 |
-
{
|
| 2379 |
-
"epoch": 8.271604938271604,
|
| 2380 |
-
"grad_norm": 3.203125,
|
| 2381 |
-
"learning_rate": 8.217393011227537e-05,
|
| 2382 |
-
"loss": 9.5874,
|
| 2383 |
-
"step": 339
|
| 2384 |
-
},
|
| 2385 |
-
{
|
| 2386 |
-
"epoch": 8.296296296296296,
|
| 2387 |
-
"grad_norm": 2.890625,
|
| 2388 |
-
"learning_rate": 7.996965181282417e-05,
|
| 2389 |
-
"loss": 8.7425,
|
| 2390 |
-
"step": 340
|
| 2391 |
-
},
|
| 2392 |
-
{
|
| 2393 |
-
"epoch": 8.320987654320987,
|
| 2394 |
-
"grad_norm": 3.515625,
|
| 2395 |
-
"learning_rate": 7.779276898948734e-05,
|
| 2396 |
-
"loss": 5.3657,
|
| 2397 |
-
"step": 341
|
| 2398 |
-
},
|
| 2399 |
-
{
|
| 2400 |
-
"epoch": 8.345679012345679,
|
| 2401 |
-
"grad_norm": 3.109375,
|
| 2402 |
-
"learning_rate": 7.564342362424714e-05,
|
| 2403 |
-
"loss": 6.8744,
|
| 2404 |
-
"step": 342
|
| 2405 |
-
},
|
| 2406 |
-
{
|
| 2407 |
-
"epoch": 8.37037037037037,
|
| 2408 |
-
"grad_norm": 3.421875,
|
| 2409 |
-
"learning_rate": 7.352175590302063e-05,
|
| 2410 |
-
"loss": 6.373,
|
| 2411 |
-
"step": 343
|
| 2412 |
-
},
|
| 2413 |
-
{
|
| 2414 |
-
"epoch": 8.395061728395062,
|
| 2415 |
-
"grad_norm": 3.21875,
|
| 2416 |
-
"learning_rate": 7.14279042065169e-05,
|
| 2417 |
-
"loss": 6.7817,
|
| 2418 |
-
"step": 344
|
| 2419 |
-
},
|
| 2420 |
-
{
|
| 2421 |
-
"epoch": 8.419753086419753,
|
| 2422 |
-
"grad_norm": 3.203125,
|
| 2423 |
-
"learning_rate": 6.93620051012111e-05,
|
| 2424 |
-
"loss": 6.8337,
|
| 2425 |
-
"step": 345
|
| 2426 |
-
},
|
| 2427 |
-
{
|
| 2428 |
-
"epoch": 8.444444444444445,
|
| 2429 |
-
"grad_norm": 2.953125,
|
| 2430 |
-
"learning_rate": 6.732419333043755e-05,
|
| 2431 |
-
"loss": 8.7099,
|
| 2432 |
-
"step": 346
|
| 2433 |
-
},
|
| 2434 |
-
{
|
| 2435 |
-
"epoch": 8.469135802469136,
|
| 2436 |
-
"grad_norm": 2.984375,
|
| 2437 |
-
"learning_rate": 6.531460180560111e-05,
|
| 2438 |
-
"loss": 6.9555,
|
| 2439 |
-
"step": 347
|
| 2440 |
-
},
|
| 2441 |
-
{
|
| 2442 |
-
"epoch": 8.493827160493828,
|
| 2443 |
-
"grad_norm": 3.8125,
|
| 2444 |
-
"learning_rate": 6.333336159750869e-05,
|
| 2445 |
-
"loss": 7.0301,
|
| 2446 |
-
"step": 348
|
| 2447 |
-
},
|
| 2448 |
-
{
|
| 2449 |
-
"epoch": 8.518518518518519,
|
| 2450 |
-
"grad_norm": 3.46875,
|
| 2451 |
-
"learning_rate": 6.138060192782002e-05,
|
| 2452 |
-
"loss": 6.0988,
|
| 2453 |
-
"step": 349
|
| 2454 |
-
},
|
| 2455 |
-
{
|
| 2456 |
-
"epoch": 8.54320987654321,
|
| 2457 |
-
"grad_norm": 3.0625,
|
| 2458 |
-
"learning_rate": 5.945645016062001e-05,
|
| 2459 |
-
"loss": 7.8126,
|
| 2460 |
-
"step": 350
|
| 2461 |
-
},
|
| 2462 |
-
{
|
| 2463 |
-
"epoch": 8.567901234567902,
|
| 2464 |
-
"grad_norm": 3.046875,
|
| 2465 |
-
"learning_rate": 5.7561031794111316e-05,
|
| 2466 |
-
"loss": 6.7753,
|
| 2467 |
-
"step": 351
|
| 2468 |
-
},
|
| 2469 |
-
{
|
| 2470 |
-
"epoch": 8.592592592592592,
|
| 2471 |
-
"grad_norm": 3.359375,
|
| 2472 |
-
"learning_rate": 5.569447045242931e-05,
|
| 2473 |
-
"loss": 6.4113,
|
| 2474 |
-
"step": 352
|
| 2475 |
-
},
|
| 2476 |
-
{
|
| 2477 |
-
"epoch": 8.617283950617283,
|
| 2478 |
-
"grad_norm": 3.125,
|
| 2479 |
-
"learning_rate": 5.385688787757864e-05,
|
| 2480 |
-
"loss": 6.0227,
|
| 2481 |
-
"step": 353
|
| 2482 |
-
},
|
| 2483 |
-
{
|
| 2484 |
-
"epoch": 8.641975308641975,
|
| 2485 |
-
"grad_norm": 3.171875,
|
| 2486 |
-
"learning_rate": 5.204840392149329e-05,
|
| 2487 |
-
"loss": 6.8429,
|
| 2488 |
-
"step": 354
|
| 2489 |
-
},
|
| 2490 |
-
{
|
| 2491 |
-
"epoch": 8.666666666666666,
|
| 2492 |
-
"grad_norm": 3.109375,
|
| 2493 |
-
"learning_rate": 5.02691365382193e-05,
|
| 2494 |
-
"loss": 7.7754,
|
| 2495 |
-
"step": 355
|
| 2496 |
-
},
|
| 2497 |
-
{
|
| 2498 |
-
"epoch": 8.691358024691358,
|
| 2499 |
-
"grad_norm": 3.140625,
|
| 2500 |
-
"learning_rate": 4.851920177622143e-05,
|
| 2501 |
-
"loss": 7.8351,
|
| 2502 |
-
"step": 356
|
| 2503 |
-
},
|
| 2504 |
-
{
|
| 2505 |
-
"epoch": 8.716049382716049,
|
| 2506 |
-
"grad_norm": 2.96875,
|
| 2507 |
-
"learning_rate": 4.6798713770814624e-05,
|
| 2508 |
-
"loss": 8.5431,
|
| 2509 |
-
"step": 357
|
| 2510 |
-
},
|
| 2511 |
-
{
|
| 2512 |
-
"epoch": 8.74074074074074,
|
| 2513 |
-
"grad_norm": 3.203125,
|
| 2514 |
-
"learning_rate": 4.5107784736718946e-05,
|
| 2515 |
-
"loss": 6.7501,
|
| 2516 |
-
"step": 358
|
| 2517 |
-
},
|
| 2518 |
-
{
|
| 2519 |
-
"epoch": 8.765432098765432,
|
| 2520 |
-
"grad_norm": 3.265625,
|
| 2521 |
-
"learning_rate": 4.344652496074142e-05,
|
| 2522 |
-
"loss": 8.5514,
|
| 2523 |
-
"step": 359
|
| 2524 |
-
},
|
| 2525 |
-
{
|
| 2526 |
-
"epoch": 8.790123456790123,
|
| 2527 |
-
"grad_norm": 3.1875,
|
| 2528 |
-
"learning_rate": 4.181504279458287e-05,
|
| 2529 |
-
"loss": 8.5261,
|
| 2530 |
-
"step": 360
|
| 2531 |
-
},
|
| 2532 |
-
{
|
| 2533 |
-
"epoch": 8.814814814814815,
|
| 2534 |
-
"grad_norm": 3.484375,
|
| 2535 |
-
"learning_rate": 4.021344464777038e-05,
|
| 2536 |
-
"loss": 5.1709,
|
| 2537 |
-
"step": 361
|
| 2538 |
-
},
|
| 2539 |
-
{
|
| 2540 |
-
"epoch": 8.839506172839506,
|
| 2541 |
-
"grad_norm": 3.234375,
|
| 2542 |
-
"learning_rate": 3.864183498071699e-05,
|
| 2543 |
-
"loss": 6.7079,
|
| 2544 |
-
"step": 362
|
| 2545 |
-
},
|
| 2546 |
-
{
|
| 2547 |
-
"epoch": 8.864197530864198,
|
| 2548 |
-
"grad_norm": 3.140625,
|
| 2549 |
-
"learning_rate": 3.710031629790928e-05,
|
| 2550 |
-
"loss": 6.9687,
|
| 2551 |
-
"step": 363
|
| 2552 |
-
},
|
| 2553 |
-
{
|
| 2554 |
-
"epoch": 8.88888888888889,
|
| 2555 |
-
"grad_norm": 3.28125,
|
| 2556 |
-
"learning_rate": 3.558898914122088e-05,
|
| 2557 |
-
"loss": 6.912,
|
| 2558 |
-
"step": 364
|
| 2559 |
-
},
|
| 2560 |
-
{
|
| 2561 |
-
"epoch": 8.91358024691358,
|
| 2562 |
-
"grad_norm": 3.0,
|
| 2563 |
-
"learning_rate": 3.4107952083355545e-05,
|
| 2564 |
-
"loss": 9.5947,
|
| 2565 |
-
"step": 365
|
| 2566 |
-
},
|
| 2567 |
-
{
|
| 2568 |
-
"epoch": 8.938271604938272,
|
| 2569 |
-
"grad_norm": 2.984375,
|
| 2570 |
-
"learning_rate": 3.265730172141757e-05,
|
| 2571 |
-
"loss": 9.588,
|
| 2572 |
-
"step": 366
|
| 2573 |
-
},
|
| 2574 |
-
{
|
| 2575 |
-
"epoch": 8.962962962962964,
|
| 2576 |
-
"grad_norm": 4.28125,
|
| 2577 |
-
"learning_rate": 3.1237132670611455e-05,
|
| 2578 |
-
"loss": 5.5477,
|
| 2579 |
-
"step": 367
|
| 2580 |
-
},
|
| 2581 |
-
{
|
| 2582 |
-
"epoch": 8.987654320987655,
|
| 2583 |
-
"grad_norm": 3.015625,
|
| 2584 |
-
"learning_rate": 2.9847537558071126e-05,
|
| 2585 |
-
"loss": 6.8754,
|
| 2586 |
-
"step": 368
|
| 2587 |
-
},
|
| 2588 |
-
{
|
| 2589 |
-
"epoch": 9.0,
|
| 2590 |
-
"grad_norm": 4.46875,
|
| 2591 |
-
"learning_rate": 2.848860701681849e-05,
|
| 2592 |
-
"loss": 5.978,
|
| 2593 |
-
"step": 369
|
| 2594 |
-
},
|
| 2595 |
-
{
|
| 2596 |
-
"epoch": 9.024691358024691,
|
| 2597 |
-
"grad_norm": 3.0625,
|
| 2598 |
-
"learning_rate": 2.7160429679851938e-05,
|
| 2599 |
-
"loss": 7.7472,
|
| 2600 |
-
"step": 370
|
| 2601 |
-
},
|
| 2602 |
-
{
|
| 2603 |
-
"epoch": 9.049382716049383,
|
| 2604 |
-
"grad_norm": 3.1875,
|
| 2605 |
-
"learning_rate": 2.5863092174365555e-05,
|
| 2606 |
-
"loss": 7.6777,
|
| 2607 |
-
"step": 371
|
| 2608 |
-
},
|
| 2609 |
-
{
|
| 2610 |
-
"epoch": 9.074074074074074,
|
| 2611 |
-
"grad_norm": 3.84375,
|
| 2612 |
-
"learning_rate": 2.4596679116099086e-05,
|
| 2613 |
-
"loss": 6.3852,
|
| 2614 |
-
"step": 372
|
| 2615 |
-
},
|
| 2616 |
-
{
|
| 2617 |
-
"epoch": 9.098765432098766,
|
| 2618 |
-
"grad_norm": 3.0,
|
| 2619 |
-
"learning_rate": 2.336127310381919e-05,
|
| 2620 |
-
"loss": 7.4578,
|
| 2621 |
-
"step": 373
|
| 2622 |
-
},
|
| 2623 |
-
{
|
| 2624 |
-
"epoch": 9.123456790123457,
|
| 2625 |
-
"grad_norm": 3.328125,
|
| 2626 |
-
"learning_rate": 2.2156954713931974e-05,
|
| 2627 |
-
"loss": 4.2617,
|
| 2628 |
-
"step": 374
|
| 2629 |
-
},
|
| 2630 |
-
{
|
| 2631 |
-
"epoch": 9.148148148148149,
|
| 2632 |
-
"grad_norm": 3.0,
|
| 2633 |
-
"learning_rate": 2.0983802495227522e-05,
|
| 2634 |
-
"loss": 8.7434,
|
| 2635 |
-
"step": 375
|
| 2636 |
-
},
|
| 2637 |
-
{
|
| 2638 |
-
"epoch": 9.17283950617284,
|
| 2639 |
-
"grad_norm": 3.28125,
|
| 2640 |
-
"learning_rate": 1.984189296375699e-05,
|
| 2641 |
-
"loss": 7.671,
|
| 2642 |
-
"step": 376
|
| 2643 |
-
},
|
| 2644 |
-
{
|
| 2645 |
-
"epoch": 9.197530864197532,
|
| 2646 |
-
"grad_norm": 2.9375,
|
| 2647 |
-
"learning_rate": 1.873130059784184e-05,
|
| 2648 |
-
"loss": 7.8471,
|
| 2649 |
-
"step": 377
|
| 2650 |
-
},
|
| 2651 |
-
{
|
| 2652 |
-
"epoch": 9.222222222222221,
|
| 2653 |
-
"grad_norm": 3.03125,
|
| 2654 |
-
"learning_rate": 1.7652097833216195e-05,
|
| 2655 |
-
"loss": 6.0614,
|
| 2656 |
-
"step": 378
|
| 2657 |
-
},
|
| 2658 |
-
{
|
| 2659 |
-
"epoch": 9.246913580246913,
|
| 2660 |
-
"grad_norm": 3.21875,
|
| 2661 |
-
"learning_rate": 1.660435505830249e-05,
|
| 2662 |
-
"loss": 6.3376,
|
| 2663 |
-
"step": 379
|
| 2664 |
-
},
|
| 2665 |
-
{
|
| 2666 |
-
"epoch": 9.271604938271604,
|
| 2667 |
-
"grad_norm": 3.15625,
|
| 2668 |
-
"learning_rate": 1.5588140609620326e-05,
|
| 2669 |
-
"loss": 7.9326,
|
| 2670 |
-
"step": 380
|
| 2671 |
-
},
|
| 2672 |
-
{
|
| 2673 |
-
"epoch": 9.296296296296296,
|
| 2674 |
-
"grad_norm": 3.265625,
|
| 2675 |
-
"learning_rate": 1.4603520767329681e-05,
|
| 2676 |
-
"loss": 5.339,
|
| 2677 |
-
"step": 381
|
| 2678 |
-
},
|
| 2679 |
-
{
|
| 2680 |
-
"epoch": 9.320987654320987,
|
| 2681 |
-
"grad_norm": 3.078125,
|
| 2682 |
-
"learning_rate": 1.3650559750907732e-05,
|
| 2683 |
-
"loss": 5.9893,
|
| 2684 |
-
"step": 382
|
| 2685 |
-
},
|
| 2686 |
-
{
|
| 2687 |
-
"epoch": 9.345679012345679,
|
| 2688 |
-
"grad_norm": 3.265625,
|
| 2689 |
-
"learning_rate": 1.2729319714960552e-05,
|
| 2690 |
-
"loss": 7.7263,
|
| 2691 |
-
"step": 383
|
| 2692 |
-
},
|
| 2693 |
-
{
|
| 2694 |
-
"epoch": 9.37037037037037,
|
| 2695 |
-
"grad_norm": 3.171875,
|
| 2696 |
-
"learning_rate": 1.1839860745168628e-05,
|
| 2697 |
-
"loss": 7.8087,
|
| 2698 |
-
"step": 384
|
| 2699 |
-
},
|
| 2700 |
-
{
|
| 2701 |
-
"epoch": 9.395061728395062,
|
| 2702 |
-
"grad_norm": 3.4375,
|
| 2703 |
-
"learning_rate": 1.0982240854368775e-05,
|
| 2704 |
-
"loss": 5.376,
|
| 2705 |
-
"step": 385
|
| 2706 |
-
},
|
| 2707 |
-
{
|
| 2708 |
-
"epoch": 9.419753086419753,
|
| 2709 |
-
"grad_norm": 3.09375,
|
| 2710 |
-
"learning_rate": 1.0156515978769597e-05,
|
| 2711 |
-
"loss": 6.8875,
|
| 2712 |
-
"step": 386
|
| 2713 |
-
},
|
| 2714 |
-
{
|
| 2715 |
-
"epoch": 9.444444444444445,
|
| 2716 |
-
"grad_norm": 3.359375,
|
| 2717 |
-
"learning_rate": 9.362739974303758e-06,
|
| 2718 |
-
"loss": 6.0189,
|
| 2719 |
-
"step": 387
|
| 2720 |
-
},
|
| 2721 |
-
{
|
| 2722 |
-
"epoch": 9.469135802469136,
|
| 2723 |
-
"grad_norm": 3.09375,
|
| 2724 |
-
"learning_rate": 8.600964613114937e-06,
|
| 2725 |
-
"loss": 6.8963,
|
| 2726 |
-
"step": 388
|
| 2727 |
-
},
|
| 2728 |
-
{
|
| 2729 |
-
"epoch": 9.493827160493828,
|
| 2730 |
-
"grad_norm": 3.203125,
|
| 2731 |
-
"learning_rate": 7.871239580181489e-06,
|
| 2732 |
-
"loss": 8.6681,
|
| 2733 |
-
"step": 389
|
| 2734 |
-
},
|
| 2735 |
-
{
|
| 2736 |
-
"epoch": 9.518518518518519,
|
| 2737 |
-
"grad_norm": 2.875,
|
| 2738 |
-
"learning_rate": 7.173612470075419e-06,
|
| 2739 |
-
"loss": 7.8249,
|
| 2740 |
-
"step": 390
|
| 2741 |
-
},
|
| 2742 |
-
{
|
| 2743 |
-
"epoch": 9.54320987654321,
|
| 2744 |
-
"grad_norm": 3.265625,
|
| 2745 |
-
"learning_rate": 6.508128783858647e-06,
|
| 2746 |
-
"loss": 6.0478,
|
| 2747 |
-
"step": 391
|
| 2748 |
-
},
|
| 2749 |
-
{
|
| 2750 |
-
"epoch": 9.567901234567902,
|
| 2751 |
-
"grad_norm": 3.109375,
|
| 2752 |
-
"learning_rate": 5.874831926114932e-06,
|
| 2753 |
-
"loss": 8.4802,
|
| 2754 |
-
"step": 392
|
| 2755 |
-
},
|
| 2756 |
-
{
|
| 2757 |
-
"epoch": 9.592592592592592,
|
| 2758 |
-
"grad_norm": 3.515625,
|
| 2759 |
-
"learning_rate": 5.2737632021189196e-06,
|
| 2760 |
-
"loss": 6.0807,
|
| 2761 |
-
"step": 393
|
| 2762 |
-
},
|
| 2763 |
-
{
|
| 2764 |
-
"epoch": 9.617283950617283,
|
| 2765 |
-
"grad_norm": 2.96875,
|
| 2766 |
-
"learning_rate": 4.704961815142406e-06,
|
| 2767 |
-
"loss": 9.5407,
|
| 2768 |
-
"step": 394
|
| 2769 |
-
},
|
| 2770 |
-
{
|
| 2771 |
-
"epoch": 9.641975308641975,
|
| 2772 |
-
"grad_norm": 3.546875,
|
| 2773 |
-
"learning_rate": 4.168464863897159e-06,
|
| 2774 |
-
"loss": 6.9195,
|
| 2775 |
-
"step": 395
|
| 2776 |
-
},
|
| 2777 |
-
{
|
| 2778 |
-
"epoch": 9.666666666666666,
|
| 2779 |
-
"grad_norm": 3.328125,
|
| 2780 |
-
"learning_rate": 3.664307340115247e-06,
|
| 2781 |
-
"loss": 6.8413,
|
| 2782 |
-
"step": 396
|
| 2783 |
-
},
|
| 2784 |
-
{
|
| 2785 |
-
"epoch": 9.691358024691358,
|
| 2786 |
-
"grad_norm": 3.234375,
|
| 2787 |
-
"learning_rate": 3.1925221262668615e-06,
|
| 2788 |
-
"loss": 6.9524,
|
| 2789 |
-
"step": 397
|
| 2790 |
-
},
|
| 2791 |
-
{
|
| 2792 |
-
"epoch": 9.716049382716049,
|
| 2793 |
-
"grad_norm": 3.078125,
|
| 2794 |
-
"learning_rate": 2.753139993415532e-06,
|
| 2795 |
-
"loss": 9.5263,
|
| 2796 |
-
"step": 398
|
| 2797 |
-
},
|
| 2798 |
-
{
|
| 2799 |
-
"epoch": 9.74074074074074,
|
| 2800 |
-
"grad_norm": 3.15625,
|
| 2801 |
-
"learning_rate": 2.3461895992114014e-06,
|
| 2802 |
-
"loss": 6.9148,
|
| 2803 |
-
"step": 399
|
| 2804 |
-
},
|
| 2805 |
-
{
|
| 2806 |
-
"epoch": 9.765432098765432,
|
| 2807 |
-
"grad_norm": 3.1875,
|
| 2808 |
-
"learning_rate": 1.971697486021773e-06,
|
| 2809 |
-
"loss": 7.6293,
|
| 2810 |
-
"step": 400
|
| 2811 |
-
},
|
| 2812 |
-
{
|
| 2813 |
-
"epoch": 9.790123456790123,
|
| 2814 |
-
"grad_norm": 3.171875,
|
| 2815 |
-
"learning_rate": 1.6296880792001646e-06,
|
| 2816 |
-
"loss": 7.6517,
|
| 2817 |
-
"step": 401
|
| 2818 |
-
},
|
| 2819 |
-
{
|
| 2820 |
-
"epoch": 9.814814814814815,
|
| 2821 |
-
"grad_norm": 3.828125,
|
| 2822 |
-
"learning_rate": 1.3201836854931926e-06,
|
| 2823 |
-
"loss": 7.0111,
|
| 2824 |
-
"step": 402
|
| 2825 |
-
},
|
| 2826 |
-
{
|
| 2827 |
-
"epoch": 9.839506172839506,
|
| 2828 |
-
"grad_norm": 2.984375,
|
| 2829 |
-
"learning_rate": 1.0432044915855699e-06,
|
| 2830 |
-
"loss": 8.7232,
|
| 2831 |
-
"step": 403
|
| 2832 |
-
},
|
| 2833 |
-
{
|
| 2834 |
-
"epoch": 9.864197530864198,
|
| 2835 |
-
"grad_norm": 2.9375,
|
| 2836 |
-
"learning_rate": 7.987685627836027e-07,
|
| 2837 |
-
"loss": 8.6614,
|
| 2838 |
-
"step": 404
|
| 2839 |
-
},
|
| 2840 |
-
{
|
| 2841 |
-
"epoch": 9.88888888888889,
|
| 2842 |
-
"grad_norm": 3.0625,
|
| 2843 |
-
"learning_rate": 5.868918418368563e-07,
|
| 2844 |
-
"loss": 7.741,
|
| 2845 |
-
"step": 405
|
| 2846 |
-
},
|
| 2847 |
-
{
|
| 2848 |
-
"epoch": 9.91358024691358,
|
| 2849 |
-
"grad_norm": 3.265625,
|
| 2850 |
-
"learning_rate": 4.07588147898319e-07,
|
| 2851 |
-
"loss": 4.983,
|
| 2852 |
-
"step": 406
|
| 2853 |
-
},
|
| 2854 |
-
{
|
| 2855 |
-
"epoch": 9.938271604938272,
|
| 2856 |
-
"grad_norm": 2.875,
|
| 2857 |
-
"learning_rate": 2.6086917562317956e-07,
|
| 2858 |
-
"loss": 7.8299,
|
| 2859 |
-
"step": 407
|
| 2860 |
-
},
|
| 2861 |
-
{
|
| 2862 |
-
"epoch": 9.962962962962964,
|
| 2863 |
-
"grad_norm": 3.21875,
|
| 2864 |
-
"learning_rate": 1.4674449440582606e-07,
|
| 2865 |
-
"loss": 7.1008,
|
| 2866 |
-
"step": 408
|
| 2867 |
-
},
|
| 2868 |
-
{
|
| 2869 |
-
"epoch": 9.987654320987655,
|
| 2870 |
-
"grad_norm": 3.28125,
|
| 2871 |
-
"learning_rate": 6.522154775606693e-08,
|
| 2872 |
-
"loss": 6.8156,
|
| 2873 |
-
"step": 409
|
| 2874 |
-
},
|
| 2875 |
-
{
|
| 2876 |
-
"epoch": 10.0,
|
| 2877 |
-
"grad_norm": 3.96875,
|
| 2878 |
-
"learning_rate": 1.6305652813353166e-08,
|
| 2879 |
-
"loss": 5.9423,
|
| 2880 |
-
"step": 410
|
| 2881 |
}
|
| 2882 |
],
|
| 2883 |
"logging_steps": 1,
|
| 2884 |
-
"max_steps":
|
| 2885 |
"num_input_tokens_seen": 0,
|
| 2886 |
"num_train_epochs": 10,
|
| 2887 |
"save_steps": 1000,
|
|
@@ -2897,8 +1427,8 @@
|
|
| 2897 |
"attributes": {}
|
| 2898 |
}
|
| 2899 |
},
|
| 2900 |
-
"total_flos": 2.
|
| 2901 |
-
"train_batch_size":
|
| 2902 |
"trial_name": null,
|
| 2903 |
"trial_params": null
|
| 2904 |
}
|
|
|
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
"epoch": 10.0,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 200,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.05,
|
| 14 |
+
"grad_norm": 2.328125,
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
+
"loss": 10.8757,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.1,
|
| 21 |
+
"grad_norm": 2.953125,
|
| 22 |
+
"learning_rate": 0.0001,
|
| 23 |
+
"loss": 10.8666,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.15,
|
| 28 |
+
"grad_norm": 3.125,
|
| 29 |
+
"learning_rate": 0.0002,
|
| 30 |
+
"loss": 10.8665,
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.2,
|
| 35 |
+
"grad_norm": 2.078125,
|
| 36 |
+
"learning_rate": 0.0003,
|
| 37 |
+
"loss": 10.8806,
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.25,
|
| 42 |
+
"grad_norm": 2.015625,
|
| 43 |
+
"learning_rate": 0.0004,
|
| 44 |
+
"loss": 10.8593,
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.3,
|
| 49 |
+
"grad_norm": 2.03125,
|
| 50 |
+
"learning_rate": 0.0005,
|
| 51 |
+
"loss": 10.8283,
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.35,
|
| 56 |
+
"grad_norm": 2.4375,
|
| 57 |
+
"learning_rate": 0.0006,
|
| 58 |
+
"loss": 10.7791,
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.4,
|
| 63 |
+
"grad_norm": 2.3125,
|
| 64 |
+
"learning_rate": 0.0007,
|
| 65 |
+
"loss": 10.7935,
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.45,
|
| 70 |
+
"grad_norm": 2.546875,
|
| 71 |
+
"learning_rate": 0.0008,
|
| 72 |
+
"loss": 10.7611,
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.5,
|
| 77 |
+
"grad_norm": 2.125,
|
| 78 |
+
"learning_rate": 0.0009000000000000001,
|
| 79 |
+
"loss": 10.7505,
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.55,
|
| 84 |
+
"grad_norm": 2.34375,
|
| 85 |
+
"learning_rate": 0.001,
|
| 86 |
+
"loss": 10.6993,
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.6,
|
| 91 |
+
"grad_norm": 2.234375,
|
| 92 |
+
"learning_rate": 0.0009999316524962345,
|
| 93 |
+
"loss": 10.6673,
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 0.65,
|
| 98 |
+
"grad_norm": 2.578125,
|
| 99 |
+
"learning_rate": 0.000999726628670463,
|
| 100 |
+
"loss": 10.539,
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 0.7,
|
| 105 |
+
"grad_norm": 2.359375,
|
| 106 |
+
"learning_rate": 0.0009993849845741523,
|
| 107 |
+
"loss": 10.6568,
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 0.75,
|
| 112 |
+
"grad_norm": 2.71875,
|
| 113 |
+
"learning_rate": 0.0009989068136093873,
|
| 114 |
+
"loss": 10.4355,
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 0.8,
|
| 119 |
+
"grad_norm": 3.515625,
|
| 120 |
+
"learning_rate": 0.000998292246503335,
|
| 121 |
+
"loss": 10.3068,
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.85,
|
| 126 |
+
"grad_norm": 2.859375,
|
| 127 |
+
"learning_rate": 0.0009975414512725057,
|
| 128 |
+
"loss": 10.4485,
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.9,
|
| 133 |
+
"grad_norm": 2.15625,
|
| 134 |
+
"learning_rate": 0.0009966546331768192,
|
| 135 |
+
"loss": 10.7021,
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 0.95,
|
| 140 |
+
"grad_norm": 2.390625,
|
| 141 |
+
"learning_rate": 0.0009956320346634876,
|
| 142 |
+
"loss": 10.591,
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 1.0,
|
| 147 |
+
"grad_norm": 2.765625,
|
| 148 |
+
"learning_rate": 0.0009944739353007343,
|
| 149 |
+
"loss": 10.4046,
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 1.05,
|
| 154 |
+
"grad_norm": 3.03125,
|
| 155 |
+
"learning_rate": 0.0009931806517013613,
|
| 156 |
+
"loss": 10.2317,
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 1.1,
|
| 161 |
+
"grad_norm": 2.4375,
|
| 162 |
+
"learning_rate": 0.0009917525374361913,
|
| 163 |
+
"loss": 10.3033,
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 1.15,
|
| 168 |
+
"grad_norm": 2.75,
|
| 169 |
+
"learning_rate": 0.0009901899829374047,
|
| 170 |
+
"loss": 10.2943,
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 1.2,
|
| 175 |
+
"grad_norm": 2.65625,
|
| 176 |
+
"learning_rate": 0.0009884934153917997,
|
| 177 |
+
"loss": 10.229,
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 1.25,
|
| 182 |
+
"grad_norm": 2.828125,
|
| 183 |
+
"learning_rate": 0.000986663298624003,
|
| 184 |
+
"loss": 10.1952,
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 1.3,
|
| 189 |
+
"grad_norm": 3.46875,
|
| 190 |
+
"learning_rate": 0.0009847001329696653,
|
| 191 |
+
"loss": 9.9897,
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 1.35,
|
| 196 |
+
"grad_norm": 3.5,
|
| 197 |
+
"learning_rate": 0.0009826044551386743,
|
| 198 |
+
"loss": 9.9053,
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 1.4,
|
| 203 |
+
"grad_norm": 3.375,
|
| 204 |
+
"learning_rate": 0.0009803768380684242,
|
| 205 |
+
"loss": 10.0425,
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 1.45,
|
| 210 |
+
"grad_norm": 3.6875,
|
| 211 |
+
"learning_rate": 0.0009780178907671788,
|
| 212 |
+
"loss": 9.8623,
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 1.5,
|
| 217 |
+
"grad_norm": 3.515625,
|
| 218 |
+
"learning_rate": 0.0009755282581475768,
|
| 219 |
+
"loss": 9.8943,
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 1.55,
|
| 224 |
+
"grad_norm": 3.296875,
|
| 225 |
+
"learning_rate": 0.0009729086208503173,
|
| 226 |
+
"loss": 9.8964,
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 1.6,
|
| 231 |
+
"grad_norm": 3.578125,
|
| 232 |
+
"learning_rate": 0.0009701596950580807,
|
| 233 |
+
"loss": 9.753,
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 1.65,
|
| 238 |
+
"grad_norm": 5.125,
|
| 239 |
+
"learning_rate": 0.0009672822322997304,
|
| 240 |
+
"loss": 9.5822,
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 1.7,
|
| 245 |
+
"grad_norm": 2.96875,
|
| 246 |
+
"learning_rate": 0.0009642770192448536,
|
| 247 |
+
"loss": 10.0307,
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 1.75,
|
| 252 |
+
"grad_norm": 4.1875,
|
| 253 |
+
"learning_rate": 0.0009611448774886924,
|
| 254 |
+
"loss": 9.7922,
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 1.8,
|
| 259 |
+
"grad_norm": 2.5,
|
| 260 |
+
"learning_rate": 0.0009578866633275287,
|
| 261 |
+
"loss": 10.2784,
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 1.85,
|
| 266 |
+
"grad_norm": 4.40625,
|
| 267 |
+
"learning_rate": 0.0009545032675245813,
|
| 268 |
+
"loss": 9.7457,
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 1.9,
|
| 273 |
+
"grad_norm": 3.796875,
|
| 274 |
+
"learning_rate": 0.0009509956150664796,
|
| 275 |
+
"loss": 9.8809,
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 1.95,
|
| 280 |
+
"grad_norm": 3.625,
|
| 281 |
+
"learning_rate": 0.0009473646649103818,
|
| 282 |
+
"loss": 9.7999,
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 2.0,
|
| 287 |
+
"grad_norm": 3.15625,
|
| 288 |
+
"learning_rate": 0.000943611409721806,
|
| 289 |
+
"loss": 10.0666,
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 2.05,
|
| 294 |
+
"grad_norm": 2.765625,
|
| 295 |
+
"learning_rate": 0.0009397368756032445,
|
| 296 |
+
"loss": 9.9986,
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 2.1,
|
| 301 |
+
"grad_norm": 2.890625,
|
| 302 |
+
"learning_rate": 0.0009357421218136386,
|
| 303 |
+
"loss": 10.1341,
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
+
"epoch": 2.15,
|
| 308 |
+
"grad_norm": 3.5,
|
| 309 |
+
"learning_rate": 0.000931628240478787,
|
| 310 |
+
"loss": 9.6153,
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
+
"epoch": 2.2,
|
| 315 |
+
"grad_norm": 2.53125,
|
| 316 |
+
"learning_rate": 0.0009273963562927695,
|
| 317 |
+
"loss": 10.0933,
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
+
"epoch": 2.25,
|
| 322 |
+
"grad_norm": 3.703125,
|
| 323 |
+
"learning_rate": 0.0009230476262104677,
|
| 324 |
+
"loss": 9.8219,
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 2.3,
|
| 329 |
+
"grad_norm": 3.75,
|
| 330 |
+
"learning_rate": 0.0009185832391312643,
|
| 331 |
+
"loss": 9.56,
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 2.35,
|
| 336 |
+
"grad_norm": 3.5625,
|
| 337 |
+
"learning_rate": 0.00091400441557401,
|
| 338 |
+
"loss": 9.5864,
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 2.4,
|
| 343 |
+
"grad_norm": 3.90625,
|
| 344 |
+
"learning_rate": 0.0009093124073433463,
|
| 345 |
+
"loss": 9.2842,
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
+
"epoch": 2.45,
|
| 350 |
+
"grad_norm": 3.46875,
|
| 351 |
+
"learning_rate": 0.0009045084971874737,
|
| 352 |
+
"loss": 9.6458,
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 2.5,
|
| 357 |
+
"grad_norm": 4.75,
|
| 358 |
+
"learning_rate": 0.0008995939984474624,
|
| 359 |
+
"loss": 8.7347,
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 2.55,
|
| 364 |
+
"grad_norm": 3.859375,
|
| 365 |
+
"learning_rate": 0.0008945702546981969,
|
| 366 |
+
"loss": 9.2305,
|
| 367 |
"step": 51
|
| 368 |
},
|
| 369 |
{
|
| 370 |
+
"epoch": 2.6,
|
| 371 |
+
"grad_norm": 3.515625,
|
| 372 |
+
"learning_rate": 0.0008894386393810563,
|
| 373 |
+
"loss": 9.3789,
|
| 374 |
"step": 52
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 2.65,
|
| 378 |
+
"grad_norm": 3.125,
|
| 379 |
+
"learning_rate": 0.0008842005554284296,
|
| 380 |
+
"loss": 9.7892,
|
| 381 |
"step": 53
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 2.7,
|
| 385 |
+
"grad_norm": 4.25,
|
| 386 |
+
"learning_rate": 0.0008788574348801675,
|
| 387 |
+
"loss": 9.0562,
|
| 388 |
"step": 54
|
| 389 |
},
|
| 390 |
{
|
| 391 |
+
"epoch": 2.75,
|
| 392 |
+
"grad_norm": 3.734375,
|
| 393 |
+
"learning_rate": 0.000873410738492077,
|
| 394 |
+
"loss": 9.3907,
|
| 395 |
"step": 55
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 2.8,
|
| 399 |
+
"grad_norm": 4.3125,
|
| 400 |
+
"learning_rate": 0.0008678619553365659,
|
| 401 |
+
"loss": 8.7691,
|
| 402 |
"step": 56
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"epoch": 2.85,
|
| 406 |
+
"grad_norm": 4.0625,
|
| 407 |
+
"learning_rate": 0.0008622126023955446,
|
| 408 |
+
"loss": 9.2092,
|
| 409 |
"step": 57
|
| 410 |
},
|
| 411 |
{
|
| 412 |
+
"epoch": 2.9,
|
| 413 |
"grad_norm": 4.21875,
|
| 414 |
+
"learning_rate": 0.0008564642241456986,
|
| 415 |
+
"loss": 9.3601,
|
| 416 |
"step": 58
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 2.95,
|
| 420 |
+
"grad_norm": 4.0625,
|
| 421 |
+
"learning_rate": 0.0008506183921362443,
|
| 422 |
+
"loss": 9.2664,
|
| 423 |
"step": 59
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 3.0,
|
| 427 |
+
"grad_norm": 4.625,
|
| 428 |
+
"learning_rate": 0.0008446767045592829,
|
| 429 |
+
"loss": 8.8939,
|
| 430 |
"step": 60
|
| 431 |
},
|
| 432 |
{
|
| 433 |
+
"epoch": 3.05,
|
| 434 |
+
"grad_norm": 5.3125,
|
| 435 |
+
"learning_rate": 0.0008386407858128706,
|
| 436 |
+
"loss": 8.3351,
|
| 437 |
"step": 61
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 3.1,
|
| 441 |
+
"grad_norm": 3.40625,
|
| 442 |
+
"learning_rate": 0.0008325122860569241,
|
| 443 |
+
"loss": 8.9346,
|
| 444 |
"step": 62
|
| 445 |
},
|
| 446 |
{
|
| 447 |
+
"epoch": 3.15,
|
| 448 |
+
"grad_norm": 4.09375,
|
| 449 |
+
"learning_rate": 0.0008262928807620843,
|
| 450 |
+
"loss": 9.1784,
|
| 451 |
"step": 63
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 3.2,
|
| 455 |
+
"grad_norm": 3.421875,
|
| 456 |
+
"learning_rate": 0.0008199842702516583,
|
| 457 |
+
"loss": 9.1151,
|
| 458 |
"step": 64
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 3.25,
|
| 462 |
+
"grad_norm": 3.765625,
|
| 463 |
+
"learning_rate": 0.0008135881792367685,
|
| 464 |
+
"loss": 8.8613,
|
| 465 |
"step": 65
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 3.3,
|
| 469 |
+
"grad_norm": 3.5625,
|
| 470 |
+
"learning_rate": 0.000807106356344834,
|
| 471 |
+
"loss": 9.067,
|
| 472 |
"step": 66
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"epoch": 3.35,
|
| 476 |
+
"grad_norm": 3.515625,
|
| 477 |
+
"learning_rate": 0.0008005405736415125,
|
| 478 |
+
"loss": 9.3725,
|
| 479 |
"step": 67
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 3.4,
|
| 483 |
+
"grad_norm": 3.421875,
|
| 484 |
+
"learning_rate": 0.0007938926261462366,
|
| 485 |
+
"loss": 9.2532,
|
| 486 |
"step": 68
|
| 487 |
},
|
| 488 |
{
|
| 489 |
+
"epoch": 3.45,
|
| 490 |
+
"grad_norm": 5.09375,
|
| 491 |
+
"learning_rate": 0.0007871643313414718,
|
| 492 |
+
"loss": 8.1986,
|
| 493 |
"step": 69
|
| 494 |
},
|
| 495 |
{
|
| 496 |
+
"epoch": 3.5,
|
| 497 |
+
"grad_norm": 4.28125,
|
| 498 |
+
"learning_rate": 0.0007803575286758365,
|
| 499 |
+
"loss": 8.7623,
|
| 500 |
"step": 70
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 3.55,
|
| 504 |
+
"grad_norm": 2.671875,
|
| 505 |
+
"learning_rate": 0.0007734740790612135,
|
| 506 |
+
"loss": 9.9666,
|
| 507 |
"step": 71
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 3.6,
|
| 511 |
+
"grad_norm": 2.890625,
|
| 512 |
+
"learning_rate": 0.0007665158643639969,
|
| 513 |
+
"loss": 9.8191,
|
| 514 |
"step": 72
|
| 515 |
},
|
| 516 |
{
|
| 517 |
+
"epoch": 3.65,
|
| 518 |
+
"grad_norm": 4.0625,
|
| 519 |
+
"learning_rate": 0.0007594847868906076,
|
| 520 |
+
"loss": 8.8212,
|
| 521 |
"step": 73
|
| 522 |
},
|
| 523 |
{
|
| 524 |
+
"epoch": 3.7,
|
| 525 |
+
"grad_norm": 4.0625,
|
| 526 |
+
"learning_rate": 0.000752382768867422,
|
| 527 |
+
"loss": 8.804,
|
| 528 |
"step": 74
|
| 529 |
},
|
| 530 |
{
|
| 531 |
+
"epoch": 3.75,
|
| 532 |
+
"grad_norm": 3.484375,
|
| 533 |
+
"learning_rate": 0.0007452117519152541,
|
| 534 |
+
"loss": 9.2155,
|
| 535 |
"step": 75
|
| 536 |
},
|
| 537 |
{
|
| 538 |
+
"epoch": 3.8,
|
| 539 |
+
"grad_norm": 3.609375,
|
| 540 |
+
"learning_rate": 0.0007379736965185368,
|
| 541 |
+
"loss": 9.0182,
|
| 542 |
"step": 76
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 3.85,
|
| 546 |
+
"grad_norm": 4.5,
|
| 547 |
+
"learning_rate": 0.000730670581489344,
|
| 548 |
+
"loss": 8.6178,
|
| 549 |
"step": 77
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 3.9,
|
| 553 |
+
"grad_norm": 3.734375,
|
| 554 |
+
"learning_rate": 0.0007233044034264033,
|
| 555 |
+
"loss": 8.6258,
|
| 556 |
"step": 78
|
| 557 |
},
|
| 558 |
{
|
| 559 |
+
"epoch": 3.95,
|
| 560 |
+
"grad_norm": 4.0,
|
| 561 |
+
"learning_rate": 0.0007158771761692464,
|
| 562 |
+
"loss": 8.3774,
|
| 563 |
"step": 79
|
| 564 |
},
|
| 565 |
{
|
| 566 |
+
"epoch": 4.0,
|
| 567 |
+
"grad_norm": 3.9375,
|
| 568 |
+
"learning_rate": 0.0007083909302476452,
|
| 569 |
+
"loss": 8.5808,
|
| 570 |
"step": 80
|
| 571 |
},
|
| 572 |
{
|
| 573 |
+
"epoch": 4.05,
|
| 574 |
+
"grad_norm": 3.84375,
|
| 575 |
+
"learning_rate": 0.0007008477123264848,
|
| 576 |
+
"loss": 8.2265,
|
| 577 |
"step": 81
|
| 578 |
},
|
| 579 |
{
|
| 580 |
+
"epoch": 4.1,
|
| 581 |
+
"grad_norm": 3.109375,
|
| 582 |
+
"learning_rate": 0.0006932495846462261,
|
| 583 |
+
"loss": 8.9861,
|
| 584 |
"step": 82
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 4.15,
|
| 588 |
+
"grad_norm": 3.3125,
|
| 589 |
+
"learning_rate": 0.0006855986244591103,
|
| 590 |
+
"loss": 8.6625,
|
| 591 |
"step": 83
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 4.2,
|
| 595 |
+
"grad_norm": 3.03125,
|
| 596 |
+
"learning_rate": 0.0006778969234612583,
|
| 597 |
+
"loss": 9.1608,
|
| 598 |
"step": 84
|
| 599 |
},
|
| 600 |
{
|
| 601 |
+
"epoch": 4.25,
|
| 602 |
+
"grad_norm": 4.25,
|
| 603 |
+
"learning_rate": 0.0006701465872208216,
|
| 604 |
+
"loss": 8.5075,
|
| 605 |
"step": 85
|
| 606 |
},
|
| 607 |
{
|
| 608 |
+
"epoch": 4.3,
|
| 609 |
+
"grad_norm": 2.796875,
|
| 610 |
+
"learning_rate": 0.0006623497346023419,
|
| 611 |
+
"loss": 9.1438,
|
| 612 |
"step": 86
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"epoch": 4.35,
|
| 616 |
+
"grad_norm": 3.53125,
|
| 617 |
+
"learning_rate": 0.0006545084971874737,
|
| 618 |
+
"loss": 8.6505,
|
| 619 |
"step": 87
|
| 620 |
},
|
| 621 |
{
|
| 622 |
+
"epoch": 4.4,
|
| 623 |
+
"grad_norm": 4.46875,
|
| 624 |
+
"learning_rate": 0.0006466250186922324,
|
| 625 |
+
"loss": 7.3011,
|
| 626 |
"step": 88
|
| 627 |
},
|
| 628 |
{
|
| 629 |
+
"epoch": 4.45,
|
| 630 |
+
"grad_norm": 4.0625,
|
| 631 |
+
"learning_rate": 0.0006387014543809223,
|
| 632 |
+
"loss": 7.7608,
|
| 633 |
"step": 89
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 4.5,
|
| 637 |
+
"grad_norm": 3.671875,
|
| 638 |
+
"learning_rate": 0.0006307399704769099,
|
| 639 |
+
"loss": 8.3767,
|
| 640 |
"step": 90
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 4.55,
|
| 644 |
+
"grad_norm": 4.125,
|
| 645 |
+
"learning_rate": 0.0006227427435703996,
|
| 646 |
+
"loss": 7.8244,
|
| 647 |
"step": 91
|
| 648 |
},
|
| 649 |
{
|
| 650 |
+
"epoch": 4.6,
|
| 651 |
+
"grad_norm": 2.609375,
|
| 652 |
+
"learning_rate": 0.0006147119600233758,
|
| 653 |
+
"loss": 9.6235,
|
| 654 |
"step": 92
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 4.65,
|
| 658 |
+
"grad_norm": 3.609375,
|
| 659 |
+
"learning_rate": 0.0006066498153718734,
|
| 660 |
+
"loss": 8.2429,
|
| 661 |
"step": 93
|
| 662 |
},
|
| 663 |
{
|
| 664 |
+
"epoch": 4.7,
|
| 665 |
+
"grad_norm": 3.9375,
|
| 666 |
+
"learning_rate": 0.0005985585137257401,
|
| 667 |
+
"loss": 8.3386,
|
| 668 |
"step": 94
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 4.75,
|
| 672 |
+
"grad_norm": 3.703125,
|
| 673 |
+
"learning_rate": 0.000590440267166055,
|
| 674 |
+
"loss": 8.0486,
|
| 675 |
"step": 95
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 4.8,
|
| 679 |
+
"grad_norm": 3.078125,
|
| 680 |
+
"learning_rate": 0.000582297295140367,
|
| 681 |
+
"loss": 8.9146,
|
| 682 |
"step": 96
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 4.85,
|
| 686 |
+
"grad_norm": 3.34375,
|
| 687 |
+
"learning_rate": 0.000574131823855921,
|
| 688 |
+
"loss": 8.872,
|
| 689 |
"step": 97
|
| 690 |
},
|
| 691 |
{
|
| 692 |
+
"epoch": 4.9,
|
| 693 |
+
"grad_norm": 4.15625,
|
| 694 |
+
"learning_rate": 0.0005659460856710345,
|
| 695 |
+
"loss": 7.5625,
|
| 696 |
"step": 98
|
| 697 |
},
|
| 698 |
{
|
| 699 |
+
"epoch": 4.95,
|
| 700 |
+
"grad_norm": 2.90625,
|
| 701 |
+
"learning_rate": 0.0005577423184847932,
|
| 702 |
+
"loss": 9.1913,
|
| 703 |
"step": 99
|
| 704 |
},
|
| 705 |
{
|
| 706 |
+
"epoch": 5.0,
|
| 707 |
+
"grad_norm": 3.109375,
|
| 708 |
+
"learning_rate": 0.0005495227651252315,
|
| 709 |
+
"loss": 8.8559,
|
| 710 |
"step": 100
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 5.05,
|
| 714 |
+
"grad_norm": 3.0,
|
| 715 |
+
"learning_rate": 0.0005412896727361663,
|
| 716 |
+
"loss": 8.5212,
|
| 717 |
"step": 101
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 5.1,
|
| 721 |
+
"grad_norm": 3.40625,
|
| 722 |
+
"learning_rate": 0.0005330452921628497,
|
| 723 |
+
"loss": 8.0584,
|
| 724 |
"step": 102
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"epoch": 5.15,
|
| 728 |
+
"grad_norm": 2.4375,
|
| 729 |
+
"learning_rate": 0.0005247918773366112,
|
| 730 |
+
"loss": 9.0202,
|
| 731 |
"step": 103
|
| 732 |
},
|
| 733 |
{
|
| 734 |
+
"epoch": 5.2,
|
| 735 |
+
"grad_norm": 3.0,
|
| 736 |
+
"learning_rate": 0.0005165316846586541,
|
| 737 |
+
"loss": 8.7377,
|
| 738 |
"step": 104
|
| 739 |
},
|
| 740 |
{
|
| 741 |
+
"epoch": 5.25,
|
| 742 |
+
"grad_norm": 3.703125,
|
| 743 |
+
"learning_rate": 0.0005082669723831793,
|
| 744 |
+
"loss": 7.4581,
|
| 745 |
"step": 105
|
| 746 |
},
|
| 747 |
{
|
| 748 |
+
"epoch": 5.3,
|
| 749 |
+
"grad_norm": 3.9375,
|
| 750 |
+
"learning_rate": 0.0005,
|
| 751 |
+
"loss": 7.3085,
|
| 752 |
"step": 106
|
| 753 |
},
|
| 754 |
{
|
| 755 |
+
"epoch": 5.35,
|
| 756 |
+
"grad_norm": 3.3125,
|
| 757 |
+
"learning_rate": 0.0004917330276168208,
|
| 758 |
+
"loss": 7.968,
|
| 759 |
"step": 107
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 5.4,
|
| 763 |
+
"grad_norm": 3.171875,
|
| 764 |
+
"learning_rate": 0.0004834683153413459,
|
| 765 |
+
"loss": 8.05,
|
| 766 |
"step": 108
|
| 767 |
},
|
| 768 |
{
|
| 769 |
+
"epoch": 5.45,
|
| 770 |
+
"grad_norm": 3.359375,
|
| 771 |
+
"learning_rate": 0.0004752081226633888,
|
| 772 |
+
"loss": 8.1395,
|
| 773 |
"step": 109
|
| 774 |
},
|
| 775 |
{
|
| 776 |
+
"epoch": 5.5,
|
| 777 |
+
"grad_norm": 2.953125,
|
| 778 |
+
"learning_rate": 0.00046695470783715033,
|
| 779 |
+
"loss": 7.9415,
|
| 780 |
"step": 110
|
| 781 |
},
|
| 782 |
{
|
| 783 |
+
"epoch": 5.55,
|
| 784 |
+
"grad_norm": 3.5,
|
| 785 |
+
"learning_rate": 0.0004587103272638339,
|
| 786 |
+
"loss": 8.5547,
|
| 787 |
"step": 111
|
| 788 |
},
|
| 789 |
{
|
| 790 |
+
"epoch": 5.6,
|
| 791 |
+
"grad_norm": 2.765625,
|
| 792 |
+
"learning_rate": 0.0004504772348747687,
|
| 793 |
+
"loss": 9.3524,
|
| 794 |
"step": 112
|
| 795 |
},
|
| 796 |
{
|
| 797 |
+
"epoch": 5.65,
|
| 798 |
+
"grad_norm": 2.875,
|
| 799 |
+
"learning_rate": 0.00044225768151520694,
|
| 800 |
+
"loss": 9.0078,
|
| 801 |
"step": 113
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 5.7,
|
| 805 |
+
"grad_norm": 3.296875,
|
| 806 |
+
"learning_rate": 0.0004340539143289655,
|
| 807 |
+
"loss": 7.8546,
|
| 808 |
"step": 114
|
| 809 |
},
|
| 810 |
{
|
| 811 |
+
"epoch": 5.75,
|
| 812 |
+
"grad_norm": 3.1875,
|
| 813 |
+
"learning_rate": 0.00042586817614407896,
|
| 814 |
+
"loss": 8.1999,
|
| 815 |
"step": 115
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 5.8,
|
| 819 |
+
"grad_norm": 3.796875,
|
| 820 |
+
"learning_rate": 0.00041770270485963295,
|
| 821 |
+
"loss": 7.4018,
|
| 822 |
"step": 116
|
| 823 |
},
|
| 824 |
{
|
| 825 |
+
"epoch": 5.85,
|
| 826 |
+
"grad_norm": 3.46875,
|
| 827 |
+
"learning_rate": 0.0004095597328339452,
|
| 828 |
+
"loss": 7.6116,
|
| 829 |
"step": 117
|
| 830 |
},
|
| 831 |
{
|
| 832 |
+
"epoch": 5.9,
|
| 833 |
+
"grad_norm": 2.859375,
|
| 834 |
+
"learning_rate": 0.0004014414862742599,
|
| 835 |
+
"loss": 8.2606,
|
| 836 |
"step": 118
|
| 837 |
},
|
| 838 |
{
|
| 839 |
+
"epoch": 5.95,
|
| 840 |
+
"grad_norm": 3.453125,
|
| 841 |
+
"learning_rate": 0.00039335018462812664,
|
| 842 |
+
"loss": 7.5184,
|
| 843 |
"step": 119
|
| 844 |
},
|
| 845 |
{
|
| 846 |
+
"epoch": 6.0,
|
| 847 |
+
"grad_norm": 3.0625,
|
| 848 |
+
"learning_rate": 0.00038528803997662424,
|
| 849 |
+
"loss": 8.5412,
|
| 850 |
"step": 120
|
| 851 |
},
|
| 852 |
{
|
| 853 |
+
"epoch": 6.05,
|
| 854 |
+
"grad_norm": 2.78125,
|
| 855 |
+
"learning_rate": 0.00037725725642960046,
|
| 856 |
+
"loss": 7.8766,
|
| 857 |
"step": 121
|
| 858 |
},
|
| 859 |
{
|
| 860 |
+
"epoch": 6.1,
|
| 861 |
+
"grad_norm": 2.625,
|
| 862 |
+
"learning_rate": 0.00036926002952309013,
|
| 863 |
+
"loss": 9.2624,
|
| 864 |
"step": 122
|
| 865 |
},
|
| 866 |
{
|
| 867 |
+
"epoch": 6.15,
|
| 868 |
+
"grad_norm": 3.28125,
|
| 869 |
+
"learning_rate": 0.00036129854561907783,
|
| 870 |
+
"loss": 7.9263,
|
| 871 |
"step": 123
|
| 872 |
},
|
| 873 |
{
|
| 874 |
+
"epoch": 6.2,
|
| 875 |
+
"grad_norm": 2.90625,
|
| 876 |
+
"learning_rate": 0.0003533749813077677,
|
| 877 |
+
"loss": 7.3967,
|
| 878 |
"step": 124
|
| 879 |
},
|
| 880 |
{
|
| 881 |
+
"epoch": 6.25,
|
| 882 |
+
"grad_norm": 3.03125,
|
| 883 |
+
"learning_rate": 0.00034549150281252633,
|
| 884 |
+
"loss": 7.4602,
|
| 885 |
"step": 125
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 6.3,
|
| 889 |
+
"grad_norm": 2.5,
|
| 890 |
+
"learning_rate": 0.0003376502653976583,
|
| 891 |
+
"loss": 8.51,
|
| 892 |
"step": 126
|
| 893 |
},
|
| 894 |
{
|
| 895 |
+
"epoch": 6.35,
|
| 896 |
+
"grad_norm": 2.75,
|
| 897 |
+
"learning_rate": 0.0003298534127791785,
|
| 898 |
+
"loss": 8.091,
|
| 899 |
"step": 127
|
| 900 |
},
|
| 901 |
{
|
| 902 |
+
"epoch": 6.4,
|
| 903 |
+
"grad_norm": 2.96875,
|
| 904 |
+
"learning_rate": 0.0003221030765387417,
|
| 905 |
+
"loss": 8.3851,
|
| 906 |
"step": 128
|
| 907 |
},
|
| 908 |
{
|
| 909 |
+
"epoch": 6.45,
|
| 910 |
+
"grad_norm": 2.5625,
|
| 911 |
+
"learning_rate": 0.0003144013755408895,
|
| 912 |
+
"loss": 9.1584,
|
| 913 |
"step": 129
|
| 914 |
},
|
| 915 |
{
|
| 916 |
+
"epoch": 6.5,
|
| 917 |
+
"grad_norm": 3.1875,
|
| 918 |
+
"learning_rate": 0.000306750415353774,
|
| 919 |
+
"loss": 7.3288,
|
| 920 |
"step": 130
|
| 921 |
},
|
| 922 |
{
|
| 923 |
+
"epoch": 6.55,
|
| 924 |
+
"grad_norm": 2.9375,
|
| 925 |
+
"learning_rate": 0.0002991522876735154,
|
| 926 |
+
"loss": 7.5572,
|
| 927 |
"step": 131
|
| 928 |
},
|
| 929 |
{
|
| 930 |
+
"epoch": 6.6,
|
| 931 |
+
"grad_norm": 2.8125,
|
| 932 |
+
"learning_rate": 0.0002916090697523549,
|
| 933 |
+
"loss": 7.794,
|
| 934 |
"step": 132
|
| 935 |
},
|
| 936 |
{
|
| 937 |
+
"epoch": 6.65,
|
| 938 |
+
"grad_norm": 2.46875,
|
| 939 |
+
"learning_rate": 0.0002841228238307536,
|
| 940 |
+
"loss": 8.5126,
|
| 941 |
"step": 133
|
| 942 |
},
|
| 943 |
{
|
| 944 |
+
"epoch": 6.7,
|
| 945 |
+
"grad_norm": 2.953125,
|
| 946 |
+
"learning_rate": 0.00027669559657359676,
|
| 947 |
+
"loss": 7.7882,
|
| 948 |
"step": 134
|
| 949 |
},
|
| 950 |
{
|
| 951 |
+
"epoch": 6.75,
|
| 952 |
+
"grad_norm": 2.546875,
|
| 953 |
+
"learning_rate": 0.00026932941851065615,
|
| 954 |
+
"loss": 8.8431,
|
| 955 |
"step": 135
|
| 956 |
},
|
| 957 |
{
|
| 958 |
+
"epoch": 6.8,
|
| 959 |
+
"grad_norm": 2.984375,
|
| 960 |
+
"learning_rate": 0.0002620263034814632,
|
| 961 |
+
"loss": 7.5375,
|
| 962 |
"step": 136
|
| 963 |
},
|
| 964 |
{
|
| 965 |
+
"epoch": 6.85,
|
| 966 |
+
"grad_norm": 3.515625,
|
| 967 |
+
"learning_rate": 0.0002547882480847461,
|
| 968 |
+
"loss": 6.797,
|
| 969 |
"step": 137
|
| 970 |
},
|
| 971 |
{
|
| 972 |
+
"epoch": 6.9,
|
| 973 |
+
"grad_norm": 2.984375,
|
| 974 |
+
"learning_rate": 0.0002476172311325783,
|
| 975 |
+
"loss": 7.3026,
|
| 976 |
"step": 138
|
| 977 |
},
|
| 978 |
{
|
| 979 |
+
"epoch": 6.95,
|
| 980 |
+
"grad_norm": 2.5,
|
| 981 |
+
"learning_rate": 0.00024051521310939256,
|
| 982 |
+
"loss": 8.4932,
|
| 983 |
"step": 139
|
| 984 |
},
|
| 985 |
{
|
| 986 |
+
"epoch": 7.0,
|
| 987 |
+
"grad_norm": 2.828125,
|
| 988 |
+
"learning_rate": 0.00023348413563600325,
|
| 989 |
+
"loss": 7.7623,
|
| 990 |
"step": 140
|
| 991 |
},
|
| 992 |
{
|
| 993 |
+
"epoch": 7.05,
|
| 994 |
+
"grad_norm": 2.25,
|
| 995 |
+
"learning_rate": 0.00022652592093878665,
|
| 996 |
+
"loss": 9.0713,
|
| 997 |
"step": 141
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
+
"epoch": 7.1,
|
| 1001 |
+
"grad_norm": 2.59375,
|
| 1002 |
+
"learning_rate": 0.0002196424713241637,
|
| 1003 |
+
"loss": 7.4665,
|
| 1004 |
"step": 142
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
+
"epoch": 7.15,
|
| 1008 |
+
"grad_norm": 2.609375,
|
| 1009 |
+
"learning_rate": 0.0002128356686585282,
|
| 1010 |
+
"loss": 7.6801,
|
| 1011 |
"step": 143
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
+
"epoch": 7.2,
|
| 1015 |
+
"grad_norm": 2.46875,
|
| 1016 |
+
"learning_rate": 0.00020610737385376348,
|
| 1017 |
+
"loss": 8.419,
|
| 1018 |
"step": 144
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
+
"epoch": 7.25,
|
| 1022 |
+
"grad_norm": 2.828125,
|
| 1023 |
+
"learning_rate": 0.00019945942635848747,
|
| 1024 |
+
"loss": 7.8005,
|
| 1025 |
"step": 145
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
+
"epoch": 7.3,
|
| 1029 |
+
"grad_norm": 2.515625,
|
| 1030 |
+
"learning_rate": 0.00019289364365516608,
|
| 1031 |
+
"loss": 8.2083,
|
| 1032 |
"step": 146
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
+
"epoch": 7.35,
|
| 1036 |
+
"grad_norm": 2.6875,
|
| 1037 |
+
"learning_rate": 0.00018641182076323148,
|
| 1038 |
+
"loss": 7.6043,
|
| 1039 |
"step": 147
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
+
"epoch": 7.4,
|
| 1043 |
+
"grad_norm": 2.640625,
|
| 1044 |
+
"learning_rate": 0.00018001572974834167,
|
| 1045 |
+
"loss": 8.0956,
|
| 1046 |
"step": 148
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
+
"epoch": 7.45,
|
| 1050 |
+
"grad_norm": 2.5625,
|
| 1051 |
+
"learning_rate": 0.00017370711923791565,
|
| 1052 |
+
"loss": 7.9467,
|
| 1053 |
"step": 149
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
+
"epoch": 7.5,
|
| 1057 |
+
"grad_norm": 2.921875,
|
| 1058 |
+
"learning_rate": 0.00016748771394307582,
|
| 1059 |
+
"loss": 7.1884,
|
| 1060 |
"step": 150
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
+
"epoch": 7.55,
|
| 1064 |
+
"grad_norm": 2.59375,
|
| 1065 |
+
"learning_rate": 0.00016135921418712956,
|
| 1066 |
+
"loss": 8.8472,
|
| 1067 |
"step": 151
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
+
"epoch": 7.6,
|
| 1071 |
+
"grad_norm": 2.34375,
|
| 1072 |
+
"learning_rate": 0.0001553232954407171,
|
| 1073 |
+
"loss": 9.2069,
|
| 1074 |
"step": 152
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
+
"epoch": 7.65,
|
| 1078 |
+
"grad_norm": 3.1875,
|
| 1079 |
+
"learning_rate": 0.00014938160786375572,
|
| 1080 |
+
"loss": 6.9255,
|
| 1081 |
"step": 153
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
+
"epoch": 7.7,
|
| 1085 |
+
"grad_norm": 2.578125,
|
| 1086 |
+
"learning_rate": 0.0001435357758543015,
|
| 1087 |
+
"loss": 7.9529,
|
| 1088 |
"step": 154
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
+
"epoch": 7.75,
|
| 1092 |
+
"grad_norm": 2.515625,
|
| 1093 |
+
"learning_rate": 0.00013778739760445552,
|
| 1094 |
+
"loss": 8.0858,
|
| 1095 |
"step": 155
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
+
"epoch": 7.8,
|
| 1099 |
+
"grad_norm": 3.03125,
|
| 1100 |
+
"learning_rate": 0.0001321380446634342,
|
| 1101 |
+
"loss": 6.9282,
|
| 1102 |
"step": 156
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
+
"epoch": 7.85,
|
| 1106 |
+
"grad_norm": 2.453125,
|
| 1107 |
+
"learning_rate": 0.0001265892615079232,
|
| 1108 |
+
"loss": 8.348,
|
| 1109 |
"step": 157
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
+
"epoch": 7.9,
|
| 1113 |
+
"grad_norm": 2.71875,
|
| 1114 |
+
"learning_rate": 0.00012114256511983274,
|
| 1115 |
+
"loss": 7.1049,
|
| 1116 |
"step": 158
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
+
"epoch": 7.95,
|
| 1120 |
+
"grad_norm": 2.8125,
|
| 1121 |
+
"learning_rate": 0.0001157994445715706,
|
| 1122 |
+
"loss": 7.2993,
|
| 1123 |
"step": 159
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
+
"epoch": 8.0,
|
| 1127 |
+
"grad_norm": 2.984375,
|
| 1128 |
+
"learning_rate": 0.00011056136061894385,
|
| 1129 |
+
"loss": 7.6333,
|
| 1130 |
"step": 160
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
+
"epoch": 8.05,
|
| 1134 |
+
"grad_norm": 2.625,
|
| 1135 |
+
"learning_rate": 0.00010542974530180327,
|
| 1136 |
+
"loss": 7.6652,
|
| 1137 |
"step": 161
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
+
"epoch": 8.1,
|
| 1141 |
+
"grad_norm": 2.671875,
|
| 1142 |
+
"learning_rate": 0.00010040600155253765,
|
| 1143 |
+
"loss": 7.818,
|
| 1144 |
"step": 162
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
+
"epoch": 8.15,
|
| 1148 |
+
"grad_norm": 2.421875,
|
| 1149 |
+
"learning_rate": 9.549150281252633e-05,
|
| 1150 |
+
"loss": 8.3813,
|
| 1151 |
"step": 163
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
+
"epoch": 8.2,
|
| 1155 |
+
"grad_norm": 2.390625,
|
| 1156 |
+
"learning_rate": 9.068759265665382e-05,
|
| 1157 |
+
"loss": 8.4012,
|
| 1158 |
"step": 164
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
+
"epoch": 8.25,
|
| 1162 |
+
"grad_norm": 2.78125,
|
| 1163 |
+
"learning_rate": 8.599558442598998e-05,
|
| 1164 |
+
"loss": 7.5276,
|
| 1165 |
"step": 165
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
+
"epoch": 8.3,
|
| 1169 |
+
"grad_norm": 2.609375,
|
| 1170 |
+
"learning_rate": 8.141676086873573e-05,
|
| 1171 |
+
"loss": 7.6272,
|
| 1172 |
"step": 166
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
+
"epoch": 8.35,
|
| 1176 |
+
"grad_norm": 2.671875,
|
| 1177 |
+
"learning_rate": 7.695237378953224e-05,
|
| 1178 |
+
"loss": 7.8184,
|
| 1179 |
"step": 167
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
+
"epoch": 8.4,
|
| 1183 |
+
"grad_norm": 3.0625,
|
| 1184 |
+
"learning_rate": 7.260364370723043e-05,
|
| 1185 |
+
"loss": 6.4234,
|
| 1186 |
"step": 168
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
+
"epoch": 8.45,
|
| 1190 |
+
"grad_norm": 2.3125,
|
| 1191 |
+
"learning_rate": 6.837175952121305e-05,
|
| 1192 |
+
"loss": 8.6891,
|
| 1193 |
"step": 169
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
+
"epoch": 8.5,
|
| 1197 |
+
"grad_norm": 3.09375,
|
| 1198 |
+
"learning_rate": 6.425787818636131e-05,
|
| 1199 |
+
"loss": 6.0584,
|
| 1200 |
"step": 170
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
+
"epoch": 8.55,
|
| 1204 |
+
"grad_norm": 2.84375,
|
| 1205 |
+
"learning_rate": 6.026312439675552e-05,
|
| 1206 |
+
"loss": 6.8717,
|
| 1207 |
"step": 171
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
+
"epoch": 8.6,
|
| 1211 |
+
"grad_norm": 2.796875,
|
| 1212 |
+
"learning_rate": 5.63885902781941e-05,
|
| 1213 |
+
"loss": 7.7985,
|
| 1214 |
"step": 172
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
+
"epoch": 8.65,
|
| 1218 |
+
"grad_norm": 2.359375,
|
| 1219 |
+
"learning_rate": 5.2635335089618264e-05,
|
| 1220 |
+
"loss": 8.1566,
|
| 1221 |
"step": 173
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
+
"epoch": 8.7,
|
| 1225 |
+
"grad_norm": 2.921875,
|
| 1226 |
+
"learning_rate": 4.900438493352055e-05,
|
| 1227 |
+
"loss": 7.2313,
|
| 1228 |
"step": 174
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
+
"epoch": 8.75,
|
| 1232 |
+
"grad_norm": 2.765625,
|
| 1233 |
+
"learning_rate": 4.5496732475418745e-05,
|
| 1234 |
+
"loss": 7.7624,
|
| 1235 |
"step": 175
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
+
"epoch": 8.8,
|
| 1239 |
+
"grad_norm": 2.265625,
|
| 1240 |
+
"learning_rate": 4.211333667247125e-05,
|
| 1241 |
+
"loss": 8.999,
|
| 1242 |
"step": 176
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
+
"epoch": 8.85,
|
| 1246 |
+
"grad_norm": 2.328125,
|
| 1247 |
+
"learning_rate": 3.885512251130763e-05,
|
| 1248 |
+
"loss": 8.7584,
|
| 1249 |
"step": 177
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
+
"epoch": 8.9,
|
| 1253 |
+
"grad_norm": 2.640625,
|
| 1254 |
+
"learning_rate": 3.572298075514652e-05,
|
| 1255 |
+
"loss": 7.7007,
|
| 1256 |
"step": 178
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
+
"epoch": 8.95,
|
| 1260 |
+
"grad_norm": 2.421875,
|
| 1261 |
+
"learning_rate": 3.271776770026963e-05,
|
| 1262 |
+
"loss": 8.0224,
|
| 1263 |
"step": 179
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
+
"epoch": 9.0,
|
| 1267 |
+
"grad_norm": 2.34375,
|
| 1268 |
+
"learning_rate": 2.9840304941919415e-05,
|
| 1269 |
+
"loss": 8.7696,
|
| 1270 |
"step": 180
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
+
"epoch": 9.05,
|
| 1274 |
+
"grad_norm": 2.671875,
|
| 1275 |
+
"learning_rate": 2.709137914968268e-05,
|
| 1276 |
+
"loss": 7.6793,
|
| 1277 |
"step": 181
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
+
"epoch": 9.1,
|
| 1281 |
+
"grad_norm": 2.859375,
|
| 1282 |
+
"learning_rate": 2.4471741852423235e-05,
|
| 1283 |
+
"loss": 7.1734,
|
| 1284 |
"step": 182
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
+
"epoch": 9.15,
|
| 1288 |
+
"grad_norm": 2.390625,
|
| 1289 |
+
"learning_rate": 2.198210923282118e-05,
|
| 1290 |
+
"loss": 8.3581,
|
| 1291 |
"step": 183
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
+
"epoch": 9.2,
|
| 1295 |
+
"grad_norm": 2.28125,
|
| 1296 |
+
"learning_rate": 1.9623161931575928e-05,
|
| 1297 |
+
"loss": 9.1404,
|
| 1298 |
"step": 184
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
+
"epoch": 9.25,
|
| 1302 |
+
"grad_norm": 2.359375,
|
| 1303 |
+
"learning_rate": 1.7395544861325717e-05,
|
| 1304 |
+
"loss": 8.7077,
|
| 1305 |
"step": 185
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
+
"epoch": 9.3,
|
| 1309 |
+
"grad_norm": 2.65625,
|
| 1310 |
+
"learning_rate": 1.5299867030334813e-05,
|
| 1311 |
+
"loss": 7.635,
|
| 1312 |
"step": 186
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
+
"epoch": 9.35,
|
| 1316 |
+
"grad_norm": 2.453125,
|
| 1317 |
+
"learning_rate": 1.3336701375997129e-05,
|
| 1318 |
+
"loss": 8.3723,
|
| 1319 |
"step": 187
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
+
"epoch": 9.4,
|
| 1323 |
+
"grad_norm": 2.5625,
|
| 1324 |
+
"learning_rate": 1.1506584608200365e-05,
|
| 1325 |
+
"loss": 7.618,
|
| 1326 |
"step": 188
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
+
"epoch": 9.45,
|
| 1330 |
+
"grad_norm": 2.453125,
|
| 1331 |
+
"learning_rate": 9.810017062595322e-06,
|
| 1332 |
+
"loss": 8.0419,
|
| 1333 |
"step": 189
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
+
"epoch": 9.5,
|
| 1337 |
+
"grad_norm": 2.5625,
|
| 1338 |
+
"learning_rate": 8.247462563808816e-06,
|
| 1339 |
+
"loss": 7.9436,
|
| 1340 |
"step": 190
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
+
"epoch": 9.55,
|
| 1344 |
+
"grad_norm": 2.546875,
|
| 1345 |
+
"learning_rate": 6.819348298638839e-06,
|
| 1346 |
+
"loss": 8.1992,
|
| 1347 |
"step": 191
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
+
"epoch": 9.6,
|
| 1351 |
+
"grad_norm": 2.515625,
|
| 1352 |
+
"learning_rate": 5.5260646992657535e-06,
|
| 1353 |
+
"loss": 7.6467,
|
| 1354 |
"step": 192
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
+
"epoch": 9.65,
|
| 1358 |
+
"grad_norm": 2.984375,
|
| 1359 |
+
"learning_rate": 4.367965336512402e-06,
|
| 1360 |
+
"loss": 6.9724,
|
| 1361 |
"step": 193
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
+
"epoch": 9.7,
|
| 1365 |
+
"grad_norm": 2.4375,
|
| 1366 |
+
"learning_rate": 3.3453668231809286e-06,
|
| 1367 |
+
"loss": 8.3873,
|
| 1368 |
"step": 194
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
+
"epoch": 9.75,
|
| 1372 |
+
"grad_norm": 2.484375,
|
| 1373 |
+
"learning_rate": 2.458548727494292e-06,
|
| 1374 |
+
"loss": 8.2895,
|
| 1375 |
"step": 195
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
+
"epoch": 9.8,
|
| 1379 |
+
"grad_norm": 2.484375,
|
| 1380 |
+
"learning_rate": 1.7077534966650766e-06,
|
| 1381 |
+
"loss": 7.6052,
|
| 1382 |
"step": 196
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
+
"epoch": 9.85,
|
| 1386 |
+
"grad_norm": 2.5625,
|
| 1387 |
+
"learning_rate": 1.0931863906127326e-06,
|
| 1388 |
+
"loss": 7.6291,
|
| 1389 |
"step": 197
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
+
"epoch": 9.9,
|
| 1393 |
+
"grad_norm": 2.78125,
|
| 1394 |
+
"learning_rate": 6.150154258476315e-07,
|
| 1395 |
+
"loss": 7.2794,
|
| 1396 |
"step": 198
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
+
"epoch": 9.95,
|
| 1400 |
+
"grad_norm": 2.703125,
|
| 1401 |
+
"learning_rate": 2.733713295369755e-07,
|
| 1402 |
+
"loss": 7.7223,
|
| 1403 |
"step": 199
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
+
"epoch": 10.0,
|
| 1407 |
+
"grad_norm": 3.046875,
|
| 1408 |
+
"learning_rate": 6.834750376549792e-08,
|
| 1409 |
+
"loss": 6.5375,
|
| 1410 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1411 |
}
|
| 1412 |
],
|
| 1413 |
"logging_steps": 1,
|
| 1414 |
+
"max_steps": 200,
|
| 1415 |
"num_input_tokens_seen": 0,
|
| 1416 |
"num_train_epochs": 10,
|
| 1417 |
"save_steps": 1000,
|
|
|
|
| 1427 |
"attributes": {}
|
| 1428 |
}
|
| 1429 |
},
|
| 1430 |
+
"total_flos": 2.54723638689792e+16,
|
| 1431 |
+
"train_batch_size": 8,
|
| 1432 |
"trial_name": null,
|
| 1433 |
"trial_params": null
|
| 1434 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df267d6b2fcad0a5830380bd77a4fe79e5db8d6bdfcba0116a1e8a253f5ddaf9
|
| 3 |
size 5777
|