Training in progress, step 140000
Browse files- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/pytorch_model.bin +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +203 -3
- pytorch_model.bin +1 -1
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 50044689
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dcab0d8867b5b850ae68e36d8ef2c70a5dbfa9a4b065a13a6b362e3397cf3fe2
|
3 |
size 50044689
|
last-checkpoint/pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 25761253
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5f41bd458cd3584fbcfdfac193494afb86160004f3f15f4d7c8df5b7ded36762
|
3 |
size 25761253
|
last-checkpoint/rng_state_0.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_1.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_2.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_3.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_4.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_5.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_6.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/rng_state_7.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14503
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f52dddb37e26c52f49ab4dce6c56c4f27359125886e16875691cfe50bcc29ef0
|
3 |
size 14503
|
last-checkpoint/scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 623
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ed68d92642b5c57649c135331b8243d8047b1dee7f4eb5f6f68f9dc4d2f32821
|
3 |
size 623
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch":
|
5 |
-
"global_step":
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
@@ -2606,11 +2606,211 @@
|
|
2606 |
"eval_samples_per_second": 1064.4,
|
2607 |
"eval_steps_per_second": 16.682,
|
2608 |
"step": 130000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2609 |
}
|
2610 |
],
|
2611 |
"max_steps": 250000,
|
2612 |
"num_train_epochs": 12,
|
2613 |
-
"total_flos": 2.
|
2614 |
"trial_name": null,
|
2615 |
"trial_params": null
|
2616 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 6.245538900785154,
|
5 |
+
"global_step": 140000,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
|
|
2606 |
"eval_samples_per_second": 1064.4,
|
2607 |
"eval_steps_per_second": 16.682,
|
2608 |
"step": 130000
|
2609 |
+
},
|
2610 |
+
{
|
2611 |
+
"epoch": 5.82,
|
2612 |
+
"learning_rate": 0.00030792659356955893,
|
2613 |
+
"loss": 0.3853,
|
2614 |
+
"step": 130500
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"epoch": 5.84,
|
2618 |
+
"learning_rate": 0.0003059755454143586,
|
2619 |
+
"loss": 0.3849,
|
2620 |
+
"step": 131000
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 5.84,
|
2624 |
+
"eval_loss": 0.3566107451915741,
|
2625 |
+
"eval_runtime": 2.2772,
|
2626 |
+
"eval_samples_per_second": 1008.7,
|
2627 |
+
"eval_steps_per_second": 15.809,
|
2628 |
+
"step": 131000
|
2629 |
+
},
|
2630 |
+
{
|
2631 |
+
"epoch": 5.87,
|
2632 |
+
"learning_rate": 0.00030402445458564144,
|
2633 |
+
"loss": 0.3845,
|
2634 |
+
"step": 131500
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 5.89,
|
2638 |
+
"learning_rate": 0.0003020734064304411,
|
2639 |
+
"loss": 0.3845,
|
2640 |
+
"step": 132000
|
2641 |
+
},
|
2642 |
+
{
|
2643 |
+
"epoch": 5.89,
|
2644 |
+
"eval_loss": 0.35691574215888977,
|
2645 |
+
"eval_runtime": 2.2557,
|
2646 |
+
"eval_samples_per_second": 1018.294,
|
2647 |
+
"eval_steps_per_second": 15.959,
|
2648 |
+
"step": 132000
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 5.91,
|
2652 |
+
"learning_rate": 0.00030012248629392423,
|
2653 |
+
"loss": 0.3842,
|
2654 |
+
"step": 132500
|
2655 |
+
},
|
2656 |
+
{
|
2657 |
+
"epoch": 5.93,
|
2658 |
+
"learning_rate": 0.00029817177951565793,
|
2659 |
+
"loss": 0.384,
|
2660 |
+
"step": 133000
|
2661 |
+
},
|
2662 |
+
{
|
2663 |
+
"epoch": 5.93,
|
2664 |
+
"eval_loss": 0.35665038228034973,
|
2665 |
+
"eval_runtime": 2.2062,
|
2666 |
+
"eval_samples_per_second": 1041.164,
|
2667 |
+
"eval_steps_per_second": 16.318,
|
2668 |
+
"step": 133000
|
2669 |
+
},
|
2670 |
+
{
|
2671 |
+
"epoch": 5.96,
|
2672 |
+
"learning_rate": 0.00029622137142587594,
|
2673 |
+
"loss": 0.3837,
|
2674 |
+
"step": 133500
|
2675 |
+
},
|
2676 |
+
{
|
2677 |
+
"epoch": 5.98,
|
2678 |
+
"learning_rate": 0.0002942713473417466,
|
2679 |
+
"loss": 0.3921,
|
2680 |
+
"step": 134000
|
2681 |
+
},
|
2682 |
+
{
|
2683 |
+
"epoch": 5.98,
|
2684 |
+
"eval_loss": 0.3628464341163635,
|
2685 |
+
"eval_runtime": 2.2578,
|
2686 |
+
"eval_samples_per_second": 1017.356,
|
2687 |
+
"eval_steps_per_second": 15.945,
|
2688 |
+
"step": 134000
|
2689 |
+
},
|
2690 |
+
{
|
2691 |
+
"epoch": 6.0,
|
2692 |
+
"learning_rate": 0.00029232179256364054,
|
2693 |
+
"loss": 0.3854,
|
2694 |
+
"step": 134500
|
2695 |
+
},
|
2696 |
+
{
|
2697 |
+
"epoch": 6.02,
|
2698 |
+
"learning_rate": 0.0002903727923713994,
|
2699 |
+
"loss": 0.3844,
|
2700 |
+
"step": 135000
|
2701 |
+
},
|
2702 |
+
{
|
2703 |
+
"epoch": 6.02,
|
2704 |
+
"eval_loss": 0.3565491735935211,
|
2705 |
+
"eval_runtime": 2.2392,
|
2706 |
+
"eval_samples_per_second": 1025.813,
|
2707 |
+
"eval_steps_per_second": 16.077,
|
2708 |
+
"step": 135000
|
2709 |
+
},
|
2710 |
+
{
|
2711 |
+
"epoch": 6.04,
|
2712 |
+
"learning_rate": 0.00028842443202060556,
|
2713 |
+
"loss": 0.383,
|
2714 |
+
"step": 135500
|
2715 |
+
},
|
2716 |
+
{
|
2717 |
+
"epoch": 6.07,
|
2718 |
+
"learning_rate": 0.00028647679673885255,
|
2719 |
+
"loss": 0.383,
|
2720 |
+
"step": 136000
|
2721 |
+
},
|
2722 |
+
{
|
2723 |
+
"epoch": 6.07,
|
2724 |
+
"eval_loss": 0.35466697812080383,
|
2725 |
+
"eval_runtime": 2.2168,
|
2726 |
+
"eval_samples_per_second": 1036.2,
|
2727 |
+
"eval_steps_per_second": 16.24,
|
2728 |
+
"step": 136000
|
2729 |
+
},
|
2730 |
+
{
|
2731 |
+
"epoch": 6.09,
|
2732 |
+
"learning_rate": 0.000284529971722017,
|
2733 |
+
"loss": 0.3829,
|
2734 |
+
"step": 136500
|
2735 |
+
},
|
2736 |
+
{
|
2737 |
+
"epoch": 6.11,
|
2738 |
+
"learning_rate": 0.0002825840421305321,
|
2739 |
+
"loss": 0.3828,
|
2740 |
+
"step": 137000
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"epoch": 6.11,
|
2744 |
+
"eval_loss": 0.3585571050643921,
|
2745 |
+
"eval_runtime": 2.2164,
|
2746 |
+
"eval_samples_per_second": 1036.381,
|
2747 |
+
"eval_steps_per_second": 16.243,
|
2748 |
+
"step": 137000
|
2749 |
+
},
|
2750 |
+
{
|
2751 |
+
"epoch": 6.13,
|
2752 |
+
"learning_rate": 0.00028063909308566196,
|
2753 |
+
"loss": 0.3826,
|
2754 |
+
"step": 137500
|
2755 |
+
},
|
2756 |
+
{
|
2757 |
+
"epoch": 6.16,
|
2758 |
+
"learning_rate": 0.00027869520966577874,
|
2759 |
+
"loss": 0.3824,
|
2760 |
+
"step": 138000
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 6.16,
|
2764 |
+
"eval_loss": 0.35530510544776917,
|
2765 |
+
"eval_runtime": 2.2219,
|
2766 |
+
"eval_samples_per_second": 1033.801,
|
2767 |
+
"eval_steps_per_second": 16.202,
|
2768 |
+
"step": 138000
|
2769 |
+
},
|
2770 |
+
{
|
2771 |
+
"epoch": 6.18,
|
2772 |
+
"learning_rate": 0.00027675247690264027,
|
2773 |
+
"loss": 0.3823,
|
2774 |
+
"step": 138500
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 6.2,
|
2778 |
+
"learning_rate": 0.0002748109797776715,
|
2779 |
+
"loss": 0.3825,
|
2780 |
+
"step": 139000
|
2781 |
+
},
|
2782 |
+
{
|
2783 |
+
"epoch": 6.2,
|
2784 |
+
"eval_loss": 0.3549206256866455,
|
2785 |
+
"eval_runtime": 2.1769,
|
2786 |
+
"eval_samples_per_second": 1055.166,
|
2787 |
+
"eval_steps_per_second": 16.537,
|
2788 |
+
"step": 139000
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 6.22,
|
2792 |
+
"learning_rate": 0.0002728708032182461,
|
2793 |
+
"loss": 0.382,
|
2794 |
+
"step": 139500
|
2795 |
+
},
|
2796 |
+
{
|
2797 |
+
"epoch": 6.25,
|
2798 |
+
"learning_rate": 0.0002709320320939721,
|
2799 |
+
"loss": 0.3818,
|
2800 |
+
"step": 140000
|
2801 |
+
},
|
2802 |
+
{
|
2803 |
+
"epoch": 6.25,
|
2804 |
+
"eval_loss": 0.3537048399448395,
|
2805 |
+
"eval_runtime": 2.2266,
|
2806 |
+
"eval_samples_per_second": 1031.628,
|
2807 |
+
"eval_steps_per_second": 16.168,
|
2808 |
+
"step": 140000
|
2809 |
}
|
2810 |
],
|
2811 |
"max_steps": 250000,
|
2812 |
"num_train_epochs": 12,
|
2813 |
+
"total_flos": 2.2423092482282158e+21,
|
2814 |
"trial_name": null,
|
2815 |
"trial_params": null
|
2816 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 25761253
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5f41bd458cd3584fbcfdfac193494afb86160004f3f15f4d7c8df5b7ded36762
|
3 |
size 25761253
|