Training in progress, step 20000, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1520630616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:375e4b9cfa9de09d8057f42e98dbc192a0866e06789a8ec7b0e9091572c996e2
|
| 3 |
size 1520630616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3041448587
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbebb029cca703c9435c8129fccc4b3f8d45e60881ec4b04f2d6acf25bec8c42
|
| 3 |
size 3041448587
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d849fd0ed2b4c55b2499c77003ae5987968969429cd3a8cafdd43ae46b463c3e
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e17b5806922786f9c39beaa8475b0a348452b2fd43fba768c2f31b4cb13e074a
|
| 3 |
size 1465
|
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": 5.
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2528,6 +2528,286 @@
|
|
| 2528 |
"learning_rate": 0.00027787805899681976,
|
| 2529 |
"loss": 1.6079,
|
| 2530 |
"step": 18000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2531 |
}
|
| 2532 |
],
|
| 2533 |
"logging_steps": 50,
|
|
@@ -2547,7 +2827,7 @@
|
|
| 2547 |
"attributes": {}
|
| 2548 |
}
|
| 2549 |
},
|
| 2550 |
-
"total_flos":
|
| 2551 |
"train_batch_size": 16,
|
| 2552 |
"trial_name": null,
|
| 2553 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.030437158469946,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 20000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2528 |
"learning_rate": 0.00027787805899681976,
|
| 2529 |
"loss": 1.6079,
|
| 2530 |
"step": 18000
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 5.009125683060109,
|
| 2534 |
+
"grad_norm": 0.640625,
|
| 2535 |
+
"learning_rate": 0.00027773963715299957,
|
| 2536 |
+
"loss": 1.6429,
|
| 2537 |
+
"step": 18050
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 5.009672131147541,
|
| 2541 |
+
"grad_norm": 0.640625,
|
| 2542 |
+
"learning_rate": 0.00027760081829144044,
|
| 2543 |
+
"loss": 1.6671,
|
| 2544 |
+
"step": 18100
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 5.010218579234973,
|
| 2548 |
+
"grad_norm": 0.6484375,
|
| 2549 |
+
"learning_rate": 0.0002774616028435946,
|
| 2550 |
+
"loss": 1.6619,
|
| 2551 |
+
"step": 18150
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 5.010765027322404,
|
| 2555 |
+
"grad_norm": 0.546875,
|
| 2556 |
+
"learning_rate": 0.00027732199124214676,
|
| 2557 |
+
"loss": 1.6221,
|
| 2558 |
+
"step": 18200
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 5.011311475409836,
|
| 2562 |
+
"grad_norm": 0.63671875,
|
| 2563 |
+
"learning_rate": 0.0002771819839210131,
|
| 2564 |
+
"loss": 1.6311,
|
| 2565 |
+
"step": 18250
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 5.011857923497268,
|
| 2569 |
+
"grad_norm": 0.54296875,
|
| 2570 |
+
"learning_rate": 0.0002770415813153396,
|
| 2571 |
+
"loss": 1.669,
|
| 2572 |
+
"step": 18300
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 5.012404371584699,
|
| 2576 |
+
"grad_norm": 0.609375,
|
| 2577 |
+
"learning_rate": 0.00027690078386150084,
|
| 2578 |
+
"loss": 1.6177,
|
| 2579 |
+
"step": 18350
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 5.012950819672131,
|
| 2583 |
+
"grad_norm": 0.59375,
|
| 2584 |
+
"learning_rate": 0.0002767595919970984,
|
| 2585 |
+
"loss": 1.622,
|
| 2586 |
+
"step": 18400
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 5.013497267759563,
|
| 2590 |
+
"grad_norm": 0.609375,
|
| 2591 |
+
"learning_rate": 0.00027661800616096,
|
| 2592 |
+
"loss": 1.6329,
|
| 2593 |
+
"step": 18450
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 5.014043715846994,
|
| 2597 |
+
"grad_norm": 0.58203125,
|
| 2598 |
+
"learning_rate": 0.00027647602679313764,
|
| 2599 |
+
"loss": 1.6369,
|
| 2600 |
+
"step": 18500
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 5.014590163934426,
|
| 2604 |
+
"grad_norm": 0.55078125,
|
| 2605 |
+
"learning_rate": 0.0002763336543349065,
|
| 2606 |
+
"loss": 1.6624,
|
| 2607 |
+
"step": 18550
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 5.015136612021858,
|
| 2611 |
+
"grad_norm": 0.5859375,
|
| 2612 |
+
"learning_rate": 0.0002761908892287633,
|
| 2613 |
+
"loss": 1.6422,
|
| 2614 |
+
"step": 18600
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 5.015683060109289,
|
| 2618 |
+
"grad_norm": 0.5703125,
|
| 2619 |
+
"learning_rate": 0.0002760477319184255,
|
| 2620 |
+
"loss": 1.6101,
|
| 2621 |
+
"step": 18650
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 5.016229508196721,
|
| 2625 |
+
"grad_norm": 0.57421875,
|
| 2626 |
+
"learning_rate": 0.0002759041828488292,
|
| 2627 |
+
"loss": 1.6355,
|
| 2628 |
+
"step": 18700
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 5.016775956284153,
|
| 2632 |
+
"grad_norm": 0.58984375,
|
| 2633 |
+
"learning_rate": 0.0002757602424661283,
|
| 2634 |
+
"loss": 1.6114,
|
| 2635 |
+
"step": 18750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 5.017322404371584,
|
| 2639 |
+
"grad_norm": 0.59375,
|
| 2640 |
+
"learning_rate": 0.00027561591121769277,
|
| 2641 |
+
"loss": 1.6274,
|
| 2642 |
+
"step": 18800
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 5.017868852459016,
|
| 2646 |
+
"grad_norm": 0.57421875,
|
| 2647 |
+
"learning_rate": 0.00027547118955210747,
|
| 2648 |
+
"loss": 1.6139,
|
| 2649 |
+
"step": 18850
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 5.018415300546448,
|
| 2653 |
+
"grad_norm": 0.55859375,
|
| 2654 |
+
"learning_rate": 0.0002753260779191706,
|
| 2655 |
+
"loss": 1.6027,
|
| 2656 |
+
"step": 18900
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 5.018961748633879,
|
| 2660 |
+
"grad_norm": 0.62109375,
|
| 2661 |
+
"learning_rate": 0.0002751805767698927,
|
| 2662 |
+
"loss": 1.6036,
|
| 2663 |
+
"step": 18950
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 5.019508196721311,
|
| 2667 |
+
"grad_norm": 0.59375,
|
| 2668 |
+
"learning_rate": 0.0002750346865564944,
|
| 2669 |
+
"loss": 1.6073,
|
| 2670 |
+
"step": 19000
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 5.020054644808743,
|
| 2674 |
+
"grad_norm": 0.54296875,
|
| 2675 |
+
"learning_rate": 0.0002748884077324061,
|
| 2676 |
+
"loss": 1.6285,
|
| 2677 |
+
"step": 19050
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 5.020601092896175,
|
| 2681 |
+
"grad_norm": 0.59765625,
|
| 2682 |
+
"learning_rate": 0.0002747417407522656,
|
| 2683 |
+
"loss": 1.6254,
|
| 2684 |
+
"step": 19100
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 5.021147540983606,
|
| 2688 |
+
"grad_norm": 0.5546875,
|
| 2689 |
+
"learning_rate": 0.00027459468607191723,
|
| 2690 |
+
"loss": 1.6058,
|
| 2691 |
+
"step": 19150
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 5.021693989071038,
|
| 2695 |
+
"grad_norm": 0.60546875,
|
| 2696 |
+
"learning_rate": 0.00027444724414841046,
|
| 2697 |
+
"loss": 1.623,
|
| 2698 |
+
"step": 19200
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 5.02224043715847,
|
| 2702 |
+
"grad_norm": 0.59765625,
|
| 2703 |
+
"learning_rate": 0.00027429941543999814,
|
| 2704 |
+
"loss": 1.623,
|
| 2705 |
+
"step": 19250
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 5.0227868852459014,
|
| 2709 |
+
"grad_norm": 0.5390625,
|
| 2710 |
+
"learning_rate": 0.0002741512004061353,
|
| 2711 |
+
"loss": 1.601,
|
| 2712 |
+
"step": 19300
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 5.023333333333333,
|
| 2716 |
+
"grad_norm": 0.56640625,
|
| 2717 |
+
"learning_rate": 0.0002740025995074777,
|
| 2718 |
+
"loss": 1.5935,
|
| 2719 |
+
"step": 19350
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 5.023879781420765,
|
| 2723 |
+
"grad_norm": 0.578125,
|
| 2724 |
+
"learning_rate": 0.00027385361320588034,
|
| 2725 |
+
"loss": 1.5831,
|
| 2726 |
+
"step": 19400
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 5.0244262295081965,
|
| 2730 |
+
"grad_norm": 0.59765625,
|
| 2731 |
+
"learning_rate": 0.0002737042419643961,
|
| 2732 |
+
"loss": 1.5974,
|
| 2733 |
+
"step": 19450
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 5.0249726775956285,
|
| 2737 |
+
"grad_norm": 0.61328125,
|
| 2738 |
+
"learning_rate": 0.0002735544862472742,
|
| 2739 |
+
"loss": 1.5629,
|
| 2740 |
+
"step": 19500
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 5.0255191256830605,
|
| 2744 |
+
"grad_norm": 0.640625,
|
| 2745 |
+
"learning_rate": 0.00027340434651995887,
|
| 2746 |
+
"loss": 1.6589,
|
| 2747 |
+
"step": 19550
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 5.026065573770492,
|
| 2751 |
+
"grad_norm": 0.59375,
|
| 2752 |
+
"learning_rate": 0.0002732538232490879,
|
| 2753 |
+
"loss": 1.6576,
|
| 2754 |
+
"step": 19600
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 5.0266120218579236,
|
| 2758 |
+
"grad_norm": 0.5859375,
|
| 2759 |
+
"learning_rate": 0.0002731029169024911,
|
| 2760 |
+
"loss": 1.6112,
|
| 2761 |
+
"step": 19650
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 5.0271584699453555,
|
| 2765 |
+
"grad_norm": 0.609375,
|
| 2766 |
+
"learning_rate": 0.00027295162794918875,
|
| 2767 |
+
"loss": 1.6232,
|
| 2768 |
+
"step": 19700
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 5.027704918032787,
|
| 2772 |
+
"grad_norm": 0.6328125,
|
| 2773 |
+
"learning_rate": 0.00027279995685939055,
|
| 2774 |
+
"loss": 1.6242,
|
| 2775 |
+
"step": 19750
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 5.028251366120219,
|
| 2779 |
+
"grad_norm": 0.546875,
|
| 2780 |
+
"learning_rate": 0.00027264790410449363,
|
| 2781 |
+
"loss": 1.6029,
|
| 2782 |
+
"step": 19800
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 5.028797814207651,
|
| 2786 |
+
"grad_norm": 0.6875,
|
| 2787 |
+
"learning_rate": 0.0002724954701570816,
|
| 2788 |
+
"loss": 1.5935,
|
| 2789 |
+
"step": 19850
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 5.029344262295082,
|
| 2793 |
+
"grad_norm": 0.6640625,
|
| 2794 |
+
"learning_rate": 0.00027234265549092257,
|
| 2795 |
+
"loss": 1.6099,
|
| 2796 |
+
"step": 19900
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 5.029890710382514,
|
| 2800 |
+
"grad_norm": 0.6015625,
|
| 2801 |
+
"learning_rate": 0.00027218946058096805,
|
| 2802 |
+
"loss": 1.6366,
|
| 2803 |
+
"step": 19950
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 5.030437158469946,
|
| 2807 |
+
"grad_norm": 0.57421875,
|
| 2808 |
+
"learning_rate": 0.0002720358859033514,
|
| 2809 |
+
"loss": 1.6249,
|
| 2810 |
+
"step": 20000
|
| 2811 |
}
|
| 2812 |
],
|
| 2813 |
"logging_steps": 50,
|
|
|
|
| 2827 |
"attributes": {}
|
| 2828 |
}
|
| 2829 |
},
|
| 2830 |
+
"total_flos": 1.0695590988837028e+19,
|
| 2831 |
"train_batch_size": 16,
|
| 2832 |
"trial_name": null,
|
| 2833 |
"trial_params": null
|