Model save
Browse files
.ipynb_checkpoints/aqlm_2bit_training-checkpoint.ipynb
CHANGED
|
@@ -387,8 +387,8 @@
|
|
| 387 |
"\n",
|
| 388 |
" <div>\n",
|
| 389 |
" \n",
|
| 390 |
-
" <progress value='
|
| 391 |
-
" [
|
| 392 |
" </div>\n",
|
| 393 |
" <table border=\"1\" class=\"dataframe\">\n",
|
| 394 |
" <thead>\n",
|
|
@@ -402,6 +402,154 @@
|
|
| 402 |
" <td>1</td>\n",
|
| 403 |
" <td>5.558500</td>\n",
|
| 404 |
" </tr>\n",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 405 |
" </tbody>\n",
|
| 406 |
"</table><p>"
|
| 407 |
],
|
|
@@ -411,6 +559,22 @@
|
|
| 411 |
},
|
| 412 |
"metadata": {},
|
| 413 |
"output_type": "display_data"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 414 |
}
|
| 415 |
],
|
| 416 |
"source": [
|
|
|
|
| 387 |
"\n",
|
| 388 |
" <div>\n",
|
| 389 |
" \n",
|
| 390 |
+
" <progress value='929' max='10000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
|
| 391 |
+
" [ 929/10000 7:12:36 < 70:33:12, 0.04 it/s, Epoch 0.30/4]\n",
|
| 392 |
" </div>\n",
|
| 393 |
" <table border=\"1\" class=\"dataframe\">\n",
|
| 394 |
" <thead>\n",
|
|
|
|
| 402 |
" <td>1</td>\n",
|
| 403 |
" <td>5.558500</td>\n",
|
| 404 |
" </tr>\n",
|
| 405 |
+
" <tr>\n",
|
| 406 |
+
" <td>25</td>\n",
|
| 407 |
+
" <td>4.310400</td>\n",
|
| 408 |
+
" </tr>\n",
|
| 409 |
+
" <tr>\n",
|
| 410 |
+
" <td>50</td>\n",
|
| 411 |
+
" <td>1.984600</td>\n",
|
| 412 |
+
" </tr>\n",
|
| 413 |
+
" <tr>\n",
|
| 414 |
+
" <td>75</td>\n",
|
| 415 |
+
" <td>1.548100</td>\n",
|
| 416 |
+
" </tr>\n",
|
| 417 |
+
" <tr>\n",
|
| 418 |
+
" <td>100</td>\n",
|
| 419 |
+
" <td>1.286000</td>\n",
|
| 420 |
+
" </tr>\n",
|
| 421 |
+
" <tr>\n",
|
| 422 |
+
" <td>125</td>\n",
|
| 423 |
+
" <td>1.133400</td>\n",
|
| 424 |
+
" </tr>\n",
|
| 425 |
+
" <tr>\n",
|
| 426 |
+
" <td>150</td>\n",
|
| 427 |
+
" <td>1.040200</td>\n",
|
| 428 |
+
" </tr>\n",
|
| 429 |
+
" <tr>\n",
|
| 430 |
+
" <td>175</td>\n",
|
| 431 |
+
" <td>0.977800</td>\n",
|
| 432 |
+
" </tr>\n",
|
| 433 |
+
" <tr>\n",
|
| 434 |
+
" <td>200</td>\n",
|
| 435 |
+
" <td>0.913900</td>\n",
|
| 436 |
+
" </tr>\n",
|
| 437 |
+
" <tr>\n",
|
| 438 |
+
" <td>225</td>\n",
|
| 439 |
+
" <td>0.909900</td>\n",
|
| 440 |
+
" </tr>\n",
|
| 441 |
+
" <tr>\n",
|
| 442 |
+
" <td>250</td>\n",
|
| 443 |
+
" <td>0.854600</td>\n",
|
| 444 |
+
" </tr>\n",
|
| 445 |
+
" <tr>\n",
|
| 446 |
+
" <td>275</td>\n",
|
| 447 |
+
" <td>0.851700</td>\n",
|
| 448 |
+
" </tr>\n",
|
| 449 |
+
" <tr>\n",
|
| 450 |
+
" <td>300</td>\n",
|
| 451 |
+
" <td>0.832200</td>\n",
|
| 452 |
+
" </tr>\n",
|
| 453 |
+
" <tr>\n",
|
| 454 |
+
" <td>325</td>\n",
|
| 455 |
+
" <td>0.810900</td>\n",
|
| 456 |
+
" </tr>\n",
|
| 457 |
+
" <tr>\n",
|
| 458 |
+
" <td>350</td>\n",
|
| 459 |
+
" <td>0.816500</td>\n",
|
| 460 |
+
" </tr>\n",
|
| 461 |
+
" <tr>\n",
|
| 462 |
+
" <td>375</td>\n",
|
| 463 |
+
" <td>0.796300</td>\n",
|
| 464 |
+
" </tr>\n",
|
| 465 |
+
" <tr>\n",
|
| 466 |
+
" <td>400</td>\n",
|
| 467 |
+
" <td>0.810300</td>\n",
|
| 468 |
+
" </tr>\n",
|
| 469 |
+
" <tr>\n",
|
| 470 |
+
" <td>425</td>\n",
|
| 471 |
+
" <td>0.767200</td>\n",
|
| 472 |
+
" </tr>\n",
|
| 473 |
+
" <tr>\n",
|
| 474 |
+
" <td>450</td>\n",
|
| 475 |
+
" <td>0.767100</td>\n",
|
| 476 |
+
" </tr>\n",
|
| 477 |
+
" <tr>\n",
|
| 478 |
+
" <td>475</td>\n",
|
| 479 |
+
" <td>0.772500</td>\n",
|
| 480 |
+
" </tr>\n",
|
| 481 |
+
" <tr>\n",
|
| 482 |
+
" <td>500</td>\n",
|
| 483 |
+
" <td>0.788000</td>\n",
|
| 484 |
+
" </tr>\n",
|
| 485 |
+
" <tr>\n",
|
| 486 |
+
" <td>525</td>\n",
|
| 487 |
+
" <td>0.741900</td>\n",
|
| 488 |
+
" </tr>\n",
|
| 489 |
+
" <tr>\n",
|
| 490 |
+
" <td>550</td>\n",
|
| 491 |
+
" <td>0.757600</td>\n",
|
| 492 |
+
" </tr>\n",
|
| 493 |
+
" <tr>\n",
|
| 494 |
+
" <td>575</td>\n",
|
| 495 |
+
" <td>0.732800</td>\n",
|
| 496 |
+
" </tr>\n",
|
| 497 |
+
" <tr>\n",
|
| 498 |
+
" <td>600</td>\n",
|
| 499 |
+
" <td>0.741600</td>\n",
|
| 500 |
+
" </tr>\n",
|
| 501 |
+
" <tr>\n",
|
| 502 |
+
" <td>625</td>\n",
|
| 503 |
+
" <td>0.749000</td>\n",
|
| 504 |
+
" </tr>\n",
|
| 505 |
+
" <tr>\n",
|
| 506 |
+
" <td>650</td>\n",
|
| 507 |
+
" <td>0.723700</td>\n",
|
| 508 |
+
" </tr>\n",
|
| 509 |
+
" <tr>\n",
|
| 510 |
+
" <td>675</td>\n",
|
| 511 |
+
" <td>0.735200</td>\n",
|
| 512 |
+
" </tr>\n",
|
| 513 |
+
" <tr>\n",
|
| 514 |
+
" <td>700</td>\n",
|
| 515 |
+
" <td>0.731500</td>\n",
|
| 516 |
+
" </tr>\n",
|
| 517 |
+
" <tr>\n",
|
| 518 |
+
" <td>725</td>\n",
|
| 519 |
+
" <td>0.711800</td>\n",
|
| 520 |
+
" </tr>\n",
|
| 521 |
+
" <tr>\n",
|
| 522 |
+
" <td>750</td>\n",
|
| 523 |
+
" <td>0.702200</td>\n",
|
| 524 |
+
" </tr>\n",
|
| 525 |
+
" <tr>\n",
|
| 526 |
+
" <td>775</td>\n",
|
| 527 |
+
" <td>0.714100</td>\n",
|
| 528 |
+
" </tr>\n",
|
| 529 |
+
" <tr>\n",
|
| 530 |
+
" <td>800</td>\n",
|
| 531 |
+
" <td>0.705400</td>\n",
|
| 532 |
+
" </tr>\n",
|
| 533 |
+
" <tr>\n",
|
| 534 |
+
" <td>825</td>\n",
|
| 535 |
+
" <td>0.711800</td>\n",
|
| 536 |
+
" </tr>\n",
|
| 537 |
+
" <tr>\n",
|
| 538 |
+
" <td>850</td>\n",
|
| 539 |
+
" <td>0.687200</td>\n",
|
| 540 |
+
" </tr>\n",
|
| 541 |
+
" <tr>\n",
|
| 542 |
+
" <td>875</td>\n",
|
| 543 |
+
" <td>0.708400</td>\n",
|
| 544 |
+
" </tr>\n",
|
| 545 |
+
" <tr>\n",
|
| 546 |
+
" <td>900</td>\n",
|
| 547 |
+
" <td>0.690700</td>\n",
|
| 548 |
+
" </tr>\n",
|
| 549 |
+
" <tr>\n",
|
| 550 |
+
" <td>925</td>\n",
|
| 551 |
+
" <td>0.697200</td>\n",
|
| 552 |
+
" </tr>\n",
|
| 553 |
" </tbody>\n",
|
| 554 |
"</table><p>"
|
| 555 |
],
|
|
|
|
| 559 |
},
|
| 560 |
"metadata": {},
|
| 561 |
"output_type": "display_data"
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"name": "stderr",
|
| 565 |
+
"output_type": "stream",
|
| 566 |
+
"text": [
|
| 567 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 568 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 569 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 570 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 571 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 572 |
+
"/usr/local/lib/python3.10/dist-packages/torch/_dynamo/eval_frame.py:600: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.4 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.\n",
|
| 573 |
+
" return fn(*args, **kwargs)\n",
|
| 574 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 575 |
+
"/usr/local/lib/python3.10/dist-packages/torch/utils/checkpoint.py:295: FutureWarning: `torch.cpu.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cpu', args...)` instead.\n",
|
| 576 |
+
" with torch.enable_grad(), device_autocast_ctx, torch.cpu.amp.autocast(**ctx.cpu_autocast_kwargs): # type: ignore[attr-defined]\n"
|
| 577 |
+
]
|
| 578 |
}
|
| 579 |
],
|
| 580 |
"source": [
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167832240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afdc8b2b69fa2c74f372ba9d4b5c6478695d88a02448bf9f25b262d40ca93538
|
| 3 |
size 167832240
|
aqlm_2bit_training.ipynb
CHANGED
|
@@ -368,7 +368,9 @@
|
|
| 368 |
{
|
| 369 |
"cell_type": "code",
|
| 370 |
"execution_count": null,
|
| 371 |
-
"metadata": {
|
|
|
|
|
|
|
| 372 |
"outputs": [
|
| 373 |
{
|
| 374 |
"name": "stderr",
|
|
@@ -387,8 +389,8 @@
|
|
| 387 |
"\n",
|
| 388 |
" <div>\n",
|
| 389 |
" \n",
|
| 390 |
-
" <progress value='
|
| 391 |
-
" [
|
| 392 |
" </div>\n",
|
| 393 |
" <table border=\"1\" class=\"dataframe\">\n",
|
| 394 |
" <thead>\n",
|
|
@@ -482,6 +484,82 @@
|
|
| 482 |
" <td>500</td>\n",
|
| 483 |
" <td>0.788000</td>\n",
|
| 484 |
" </tr>\n",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 485 |
" </tbody>\n",
|
| 486 |
"</table><p>"
|
| 487 |
],
|
|
|
|
| 368 |
{
|
| 369 |
"cell_type": "code",
|
| 370 |
"execution_count": null,
|
| 371 |
+
"metadata": {
|
| 372 |
+
"scrolled": true
|
| 373 |
+
},
|
| 374 |
"outputs": [
|
| 375 |
{
|
| 376 |
"name": "stderr",
|
|
|
|
| 389 |
"\n",
|
| 390 |
" <div>\n",
|
| 391 |
" \n",
|
| 392 |
+
" <progress value='994' max='10000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
|
| 393 |
+
" [ 994/10000 7:43:00 < 70:03:24, 0.04 it/s, Epoch 0.32/4]\n",
|
| 394 |
" </div>\n",
|
| 395 |
" <table border=\"1\" class=\"dataframe\">\n",
|
| 396 |
" <thead>\n",
|
|
|
|
| 484 |
" <td>500</td>\n",
|
| 485 |
" <td>0.788000</td>\n",
|
| 486 |
" </tr>\n",
|
| 487 |
+
" <tr>\n",
|
| 488 |
+
" <td>525</td>\n",
|
| 489 |
+
" <td>0.741900</td>\n",
|
| 490 |
+
" </tr>\n",
|
| 491 |
+
" <tr>\n",
|
| 492 |
+
" <td>550</td>\n",
|
| 493 |
+
" <td>0.757600</td>\n",
|
| 494 |
+
" </tr>\n",
|
| 495 |
+
" <tr>\n",
|
| 496 |
+
" <td>575</td>\n",
|
| 497 |
+
" <td>0.732800</td>\n",
|
| 498 |
+
" </tr>\n",
|
| 499 |
+
" <tr>\n",
|
| 500 |
+
" <td>600</td>\n",
|
| 501 |
+
" <td>0.741600</td>\n",
|
| 502 |
+
" </tr>\n",
|
| 503 |
+
" <tr>\n",
|
| 504 |
+
" <td>625</td>\n",
|
| 505 |
+
" <td>0.749000</td>\n",
|
| 506 |
+
" </tr>\n",
|
| 507 |
+
" <tr>\n",
|
| 508 |
+
" <td>650</td>\n",
|
| 509 |
+
" <td>0.723700</td>\n",
|
| 510 |
+
" </tr>\n",
|
| 511 |
+
" <tr>\n",
|
| 512 |
+
" <td>675</td>\n",
|
| 513 |
+
" <td>0.735200</td>\n",
|
| 514 |
+
" </tr>\n",
|
| 515 |
+
" <tr>\n",
|
| 516 |
+
" <td>700</td>\n",
|
| 517 |
+
" <td>0.731500</td>\n",
|
| 518 |
+
" </tr>\n",
|
| 519 |
+
" <tr>\n",
|
| 520 |
+
" <td>725</td>\n",
|
| 521 |
+
" <td>0.711800</td>\n",
|
| 522 |
+
" </tr>\n",
|
| 523 |
+
" <tr>\n",
|
| 524 |
+
" <td>750</td>\n",
|
| 525 |
+
" <td>0.702200</td>\n",
|
| 526 |
+
" </tr>\n",
|
| 527 |
+
" <tr>\n",
|
| 528 |
+
" <td>775</td>\n",
|
| 529 |
+
" <td>0.714100</td>\n",
|
| 530 |
+
" </tr>\n",
|
| 531 |
+
" <tr>\n",
|
| 532 |
+
" <td>800</td>\n",
|
| 533 |
+
" <td>0.705400</td>\n",
|
| 534 |
+
" </tr>\n",
|
| 535 |
+
" <tr>\n",
|
| 536 |
+
" <td>825</td>\n",
|
| 537 |
+
" <td>0.711800</td>\n",
|
| 538 |
+
" </tr>\n",
|
| 539 |
+
" <tr>\n",
|
| 540 |
+
" <td>850</td>\n",
|
| 541 |
+
" <td>0.687200</td>\n",
|
| 542 |
+
" </tr>\n",
|
| 543 |
+
" <tr>\n",
|
| 544 |
+
" <td>875</td>\n",
|
| 545 |
+
" <td>0.708400</td>\n",
|
| 546 |
+
" </tr>\n",
|
| 547 |
+
" <tr>\n",
|
| 548 |
+
" <td>900</td>\n",
|
| 549 |
+
" <td>0.690700</td>\n",
|
| 550 |
+
" </tr>\n",
|
| 551 |
+
" <tr>\n",
|
| 552 |
+
" <td>925</td>\n",
|
| 553 |
+
" <td>0.697200</td>\n",
|
| 554 |
+
" </tr>\n",
|
| 555 |
+
" <tr>\n",
|
| 556 |
+
" <td>950</td>\n",
|
| 557 |
+
" <td>0.698000</td>\n",
|
| 558 |
+
" </tr>\n",
|
| 559 |
+
" <tr>\n",
|
| 560 |
+
" <td>975</td>\n",
|
| 561 |
+
" <td>0.681700</td>\n",
|
| 562 |
+
" </tr>\n",
|
| 563 |
" </tbody>\n",
|
| 564 |
"</table><p>"
|
| 565 |
],
|