Training in progress, step 352, checkpoint
Browse files
last-checkpoint/model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4969539560
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:096439a15e0b06aee20e3753c9fc04b16d480adc998310ef78cc635894ce0439
|
| 3 |
size 4969539560
|
last-checkpoint/model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1912795688
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:baf6e5b35338a330f50e7e5f5405f807ca31d11aeffe1893d6a171ddc95e8f71
|
| 3 |
size 1912795688
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 16,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2408,6 +2408,246 @@
|
|
| 2408 |
"eval_samples_per_second": 17.552,
|
| 2409 |
"eval_steps_per_second": 17.552,
|
| 2410 |
"step": 320
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2411 |
}
|
| 2412 |
],
|
| 2413 |
"logging_steps": 1,
|
|
@@ -2427,7 +2667,7 @@
|
|
| 2427 |
"attributes": {}
|
| 2428 |
}
|
| 2429 |
},
|
| 2430 |
-
"total_flos": 1.
|
| 2431 |
"train_batch_size": 1,
|
| 2432 |
"trial_name": null,
|
| 2433 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 88.0,
|
| 6 |
"eval_steps": 16,
|
| 7 |
+
"global_step": 352,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2408 |
"eval_samples_per_second": 17.552,
|
| 2409 |
"eval_steps_per_second": 17.552,
|
| 2410 |
"step": 320
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 80.2909090909091,
|
| 2414 |
+
"grad_norm": 3.312922954559326,
|
| 2415 |
+
"learning_rate": 2.0528000059645996e-08,
|
| 2416 |
+
"loss": 0.6742,
|
| 2417 |
+
"step": 321
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 80.58181818181818,
|
| 2421 |
+
"grad_norm": 3.8569176197052,
|
| 2422 |
+
"learning_rate": 2.027689374755261e-08,
|
| 2423 |
+
"loss": 0.6589,
|
| 2424 |
+
"step": 322
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 80.87272727272727,
|
| 2428 |
+
"grad_norm": 3.100782871246338,
|
| 2429 |
+
"learning_rate": 2.0028431734436306e-08,
|
| 2430 |
+
"loss": 0.6384,
|
| 2431 |
+
"step": 323
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 81.0,
|
| 2435 |
+
"grad_norm": 3.488448143005371,
|
| 2436 |
+
"learning_rate": 1.9782632941641373e-08,
|
| 2437 |
+
"loss": 0.6523,
|
| 2438 |
+
"step": 324
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 81.2909090909091,
|
| 2442 |
+
"grad_norm": 3.713844060897827,
|
| 2443 |
+
"learning_rate": 1.9539516087697516e-08,
|
| 2444 |
+
"loss": 0.7186,
|
| 2445 |
+
"step": 325
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 81.58181818181818,
|
| 2449 |
+
"grad_norm": 3.2771830558776855,
|
| 2450 |
+
"learning_rate": 1.9299099686894422e-08,
|
| 2451 |
+
"loss": 0.7155,
|
| 2452 |
+
"step": 326
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 81.87272727272727,
|
| 2456 |
+
"grad_norm": 3.013705015182495,
|
| 2457 |
+
"learning_rate": 1.9061402047871834e-08,
|
| 2458 |
+
"loss": 0.5796,
|
| 2459 |
+
"step": 327
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 82.0,
|
| 2463 |
+
"grad_norm": 3.9129855632781982,
|
| 2464 |
+
"learning_rate": 1.8826441272225223e-08,
|
| 2465 |
+
"loss": 0.5744,
|
| 2466 |
+
"step": 328
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 82.2909090909091,
|
| 2470 |
+
"grad_norm": 3.432311773300171,
|
| 2471 |
+
"learning_rate": 1.8594235253127372e-08,
|
| 2472 |
+
"loss": 0.6681,
|
| 2473 |
+
"step": 329
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 82.58181818181818,
|
| 2477 |
+
"grad_norm": 3.2865946292877197,
|
| 2478 |
+
"learning_rate": 1.8364801673965642e-08,
|
| 2479 |
+
"loss": 0.7103,
|
| 2480 |
+
"step": 330
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 82.87272727272727,
|
| 2484 |
+
"grad_norm": 3.1664698123931885,
|
| 2485 |
+
"learning_rate": 1.8138158006995365e-08,
|
| 2486 |
+
"loss": 0.5704,
|
| 2487 |
+
"step": 331
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 83.0,
|
| 2491 |
+
"grad_norm": 3.5827629566192627,
|
| 2492 |
+
"learning_rate": 1.7914321512009295e-08,
|
| 2493 |
+
"loss": 0.7128,
|
| 2494 |
+
"step": 332
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 83.2909090909091,
|
| 2498 |
+
"grad_norm": 3.228314161300659,
|
| 2499 |
+
"learning_rate": 1.7693309235023126e-08,
|
| 2500 |
+
"loss": 0.6072,
|
| 2501 |
+
"step": 333
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 83.58181818181818,
|
| 2505 |
+
"grad_norm": 3.2263855934143066,
|
| 2506 |
+
"learning_rate": 1.7475138006977434e-08,
|
| 2507 |
+
"loss": 0.6525,
|
| 2508 |
+
"step": 334
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 83.87272727272727,
|
| 2512 |
+
"grad_norm": 3.2911698818206787,
|
| 2513 |
+
"learning_rate": 1.7259824442455922e-08,
|
| 2514 |
+
"loss": 0.7323,
|
| 2515 |
+
"step": 335
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 84.0,
|
| 2519 |
+
"grad_norm": 3.629072666168213,
|
| 2520 |
+
"learning_rate": 1.704738493842015e-08,
|
| 2521 |
+
"loss": 0.6316,
|
| 2522 |
+
"step": 336
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 84.0,
|
| 2526 |
+
"eval_loss": 0.6387331485748291,
|
| 2527 |
+
"eval_runtime": 0.761,
|
| 2528 |
+
"eval_samples_per_second": 17.082,
|
| 2529 |
+
"eval_steps_per_second": 17.082,
|
| 2530 |
+
"step": 336
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 84.2909090909091,
|
| 2534 |
+
"grad_norm": 3.0507638454437256,
|
| 2535 |
+
"learning_rate": 1.6837835672960833e-08,
|
| 2536 |
+
"loss": 0.6576,
|
| 2537 |
+
"step": 337
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 84.58181818181818,
|
| 2541 |
+
"grad_norm": 3.1356823444366455,
|
| 2542 |
+
"learning_rate": 1.663119260406585e-08,
|
| 2543 |
+
"loss": 0.6216,
|
| 2544 |
+
"step": 338
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 84.87272727272727,
|
| 2548 |
+
"grad_norm": 3.6667861938476562,
|
| 2549 |
+
"learning_rate": 1.642747146840495e-08,
|
| 2550 |
+
"loss": 0.6975,
|
| 2551 |
+
"step": 339
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 85.0,
|
| 2555 |
+
"grad_norm": 3.397160768508911,
|
| 2556 |
+
"learning_rate": 1.6226687780131337e-08,
|
| 2557 |
+
"loss": 0.6022,
|
| 2558 |
+
"step": 340
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 85.2909090909091,
|
| 2562 |
+
"grad_norm": 3.163560152053833,
|
| 2563 |
+
"learning_rate": 1.602885682970026e-08,
|
| 2564 |
+
"loss": 0.6316,
|
| 2565 |
+
"step": 341
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 85.58181818181818,
|
| 2569 |
+
"grad_norm": 3.1787586212158203,
|
| 2570 |
+
"learning_rate": 1.5833993682704515e-08,
|
| 2571 |
+
"loss": 0.6725,
|
| 2572 |
+
"step": 342
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 85.87272727272727,
|
| 2576 |
+
"grad_norm": 3.379927158355713,
|
| 2577 |
+
"learning_rate": 1.5642113178727193e-08,
|
| 2578 |
+
"loss": 0.6319,
|
| 2579 |
+
"step": 343
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 86.0,
|
| 2583 |
+
"grad_norm": 3.783219337463379,
|
| 2584 |
+
"learning_rate": 1.5453229930211566e-08,
|
| 2585 |
+
"loss": 0.72,
|
| 2586 |
+
"step": 344
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 86.2909090909091,
|
| 2590 |
+
"grad_norm": 3.678173065185547,
|
| 2591 |
+
"learning_rate": 1.5267358321348288e-08,
|
| 2592 |
+
"loss": 0.7343,
|
| 2593 |
+
"step": 345
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 86.58181818181818,
|
| 2597 |
+
"grad_norm": 3.0277297496795654,
|
| 2598 |
+
"learning_rate": 1.5084512506980025e-08,
|
| 2599 |
+
"loss": 0.6112,
|
| 2600 |
+
"step": 346
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 86.87272727272727,
|
| 2604 |
+
"grad_norm": 2.9994826316833496,
|
| 2605 |
+
"learning_rate": 1.490470641152345e-08,
|
| 2606 |
+
"loss": 0.6377,
|
| 2607 |
+
"step": 347
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 87.0,
|
| 2611 |
+
"grad_norm": 3.4868507385253906,
|
| 2612 |
+
"learning_rate": 1.4727953727908877e-08,
|
| 2613 |
+
"loss": 0.603,
|
| 2614 |
+
"step": 348
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 87.2909090909091,
|
| 2618 |
+
"grad_norm": 3.3681607246398926,
|
| 2619 |
+
"learning_rate": 1.4554267916537493e-08,
|
| 2620 |
+
"loss": 0.6832,
|
| 2621 |
+
"step": 349
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 87.58181818181818,
|
| 2625 |
+
"grad_norm": 3.123229742050171,
|
| 2626 |
+
"learning_rate": 1.438366220425628e-08,
|
| 2627 |
+
"loss": 0.6086,
|
| 2628 |
+
"step": 350
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 87.87272727272727,
|
| 2632 |
+
"grad_norm": 3.2728774547576904,
|
| 2633 |
+
"learning_rate": 1.4216149583350753e-08,
|
| 2634 |
+
"loss": 0.6735,
|
| 2635 |
+
"step": 351
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 88.0,
|
| 2639 |
+
"grad_norm": 3.185558319091797,
|
| 2640 |
+
"learning_rate": 1.405174281055556e-08,
|
| 2641 |
+
"loss": 0.6232,
|
| 2642 |
+
"step": 352
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 88.0,
|
| 2646 |
+
"eval_loss": 0.6380465030670166,
|
| 2647 |
+
"eval_runtime": 0.6934,
|
| 2648 |
+
"eval_samples_per_second": 18.748,
|
| 2649 |
+
"eval_steps_per_second": 18.748,
|
| 2650 |
+
"step": 352
|
| 2651 |
}
|
| 2652 |
],
|
| 2653 |
"logging_steps": 1,
|
|
|
|
| 2667 |
"attributes": {}
|
| 2668 |
}
|
| 2669 |
},
|
| 2670 |
+
"total_flos": 1.1805489168777216e+17,
|
| 2671 |
"train_batch_size": 1,
|
| 2672 |
"trial_name": null,
|
| 2673 |
"trial_params": null
|