Training in progress, step 400, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 42486824
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c88af2419b3f12c58d97223868b714417b3562213eda68e2297759a277e92c7
|
| 3 |
size 42486824
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 21735482
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0679fc935e48b88de3c3b3200db950165ee89a59b24b30aac4c75b4df5f49562
|
| 3 |
size 21735482
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d479b1d847f42fe27487720674f2794600f0b0766d2243b40acdd39821b0150f
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:705cabf5cbc3a6ab0feb67c77b9b453d59efcc939ce90d310af96e621810f990
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 4.
|
| 3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
| 4 |
-
"epoch": 2.
|
| 5 |
"eval_steps": 50,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2521,6 +2521,364 @@
|
|
| 2521 |
"eval_samples_per_second": 89.433,
|
| 2522 |
"eval_steps_per_second": 22.513,
|
| 2523 |
"step": 350
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2524 |
}
|
| 2525 |
],
|
| 2526 |
"logging_steps": 1,
|
|
@@ -2544,12 +2902,12 @@
|
|
| 2544 |
"should_evaluate": false,
|
| 2545 |
"should_log": false,
|
| 2546 |
"should_save": true,
|
| 2547 |
-
"should_training_stop":
|
| 2548 |
},
|
| 2549 |
"attributes": {}
|
| 2550 |
}
|
| 2551 |
},
|
| 2552 |
-
"total_flos":
|
| 2553 |
"train_batch_size": 8,
|
| 2554 |
"trial_name": null,
|
| 2555 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 4.339465618133545,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-400",
|
| 4 |
+
"epoch": 2.3222060957910013,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2521 |
"eval_samples_per_second": 89.433,
|
| 2522 |
"eval_steps_per_second": 22.513,
|
| 2523 |
"step": 350
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.0377358490566038,
|
| 2527 |
+
"grad_norm": 3.483829975128174,
|
| 2528 |
+
"learning_rate": 3.844650207332562e-06,
|
| 2529 |
+
"loss": 26.4342,
|
| 2530 |
+
"step": 351
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 2.043541364296081,
|
| 2534 |
+
"grad_norm": 3.365922451019287,
|
| 2535 |
+
"learning_rate": 3.691267552111183e-06,
|
| 2536 |
+
"loss": 24.6714,
|
| 2537 |
+
"step": 352
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 2.049346879535559,
|
| 2541 |
+
"grad_norm": 3.5824503898620605,
|
| 2542 |
+
"learning_rate": 3.54088980417534e-06,
|
| 2543 |
+
"loss": 25.5254,
|
| 2544 |
+
"step": 353
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 2.0551523947750363,
|
| 2548 |
+
"grad_norm": 3.657639265060425,
|
| 2549 |
+
"learning_rate": 3.393526721321616e-06,
|
| 2550 |
+
"loss": 22.559,
|
| 2551 |
+
"step": 354
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 2.0609579100145137,
|
| 2555 |
+
"grad_norm": 3.1590425968170166,
|
| 2556 |
+
"learning_rate": 3.249187865729264e-06,
|
| 2557 |
+
"loss": 21.4447,
|
| 2558 |
+
"step": 355
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 2.0667634252539915,
|
| 2562 |
+
"grad_norm": 3.292860746383667,
|
| 2563 |
+
"learning_rate": 3.1078826033397843e-06,
|
| 2564 |
+
"loss": 24.0399,
|
| 2565 |
+
"step": 356
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 2.072568940493469,
|
| 2569 |
+
"grad_norm": 3.7838804721832275,
|
| 2570 |
+
"learning_rate": 2.9696201032491434e-06,
|
| 2571 |
+
"loss": 22.5607,
|
| 2572 |
+
"step": 357
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 2.078374455732946,
|
| 2576 |
+
"grad_norm": 3.6551146507263184,
|
| 2577 |
+
"learning_rate": 2.8344093371128424e-06,
|
| 2578 |
+
"loss": 24.2265,
|
| 2579 |
+
"step": 358
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 2.084179970972424,
|
| 2583 |
+
"grad_norm": 4.6166605949401855,
|
| 2584 |
+
"learning_rate": 2.70225907856374e-06,
|
| 2585 |
+
"loss": 19.9128,
|
| 2586 |
+
"step": 359
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 2.0899854862119014,
|
| 2590 |
+
"grad_norm": 4.374910831451416,
|
| 2591 |
+
"learning_rate": 2.573177902642726e-06,
|
| 2592 |
+
"loss": 22.2403,
|
| 2593 |
+
"step": 360
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 2.0957910014513788,
|
| 2597 |
+
"grad_norm": 4.862795352935791,
|
| 2598 |
+
"learning_rate": 2.4471741852423237e-06,
|
| 2599 |
+
"loss": 22.3289,
|
| 2600 |
+
"step": 361
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 2.101596516690856,
|
| 2604 |
+
"grad_norm": 4.927355766296387,
|
| 2605 |
+
"learning_rate": 2.324256102563188e-06,
|
| 2606 |
+
"loss": 24.9095,
|
| 2607 |
+
"step": 362
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.107402031930334,
|
| 2611 |
+
"grad_norm": 4.585765838623047,
|
| 2612 |
+
"learning_rate": 2.204431630583548e-06,
|
| 2613 |
+
"loss": 25.2274,
|
| 2614 |
+
"step": 363
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 2.1132075471698113,
|
| 2618 |
+
"grad_norm": 5.47407865524292,
|
| 2619 |
+
"learning_rate": 2.087708544541689e-06,
|
| 2620 |
+
"loss": 19.1767,
|
| 2621 |
+
"step": 364
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.1190130624092887,
|
| 2625 |
+
"grad_norm": 3.833303928375244,
|
| 2626 |
+
"learning_rate": 1.974094418431388e-06,
|
| 2627 |
+
"loss": 16.9155,
|
| 2628 |
+
"step": 365
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 2.1248185776487665,
|
| 2632 |
+
"grad_norm": 4.477059364318848,
|
| 2633 |
+
"learning_rate": 1.8635966245104664e-06,
|
| 2634 |
+
"loss": 18.9644,
|
| 2635 |
+
"step": 366
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 2.130624092888244,
|
| 2639 |
+
"grad_norm": 4.299328804016113,
|
| 2640 |
+
"learning_rate": 1.7562223328224325e-06,
|
| 2641 |
+
"loss": 14.2111,
|
| 2642 |
+
"step": 367
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 2.136429608127721,
|
| 2646 |
+
"grad_norm": 4.06097936630249,
|
| 2647 |
+
"learning_rate": 1.6519785107311891e-06,
|
| 2648 |
+
"loss": 16.2717,
|
| 2649 |
+
"step": 368
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 2.142235123367199,
|
| 2653 |
+
"grad_norm": 4.192160606384277,
|
| 2654 |
+
"learning_rate": 1.5508719224689717e-06,
|
| 2655 |
+
"loss": 16.403,
|
| 2656 |
+
"step": 369
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 2.1480406386066764,
|
| 2660 |
+
"grad_norm": 3.730713367462158,
|
| 2661 |
+
"learning_rate": 1.4529091286973995e-06,
|
| 2662 |
+
"loss": 14.7868,
|
| 2663 |
+
"step": 370
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 2.1538461538461537,
|
| 2667 |
+
"grad_norm": 3.103394031524658,
|
| 2668 |
+
"learning_rate": 1.358096486081778e-06,
|
| 2669 |
+
"loss": 15.1416,
|
| 2670 |
+
"step": 371
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 2.1596516690856316,
|
| 2674 |
+
"grad_norm": 4.092171669006348,
|
| 2675 |
+
"learning_rate": 1.2664401468786114e-06,
|
| 2676 |
+
"loss": 15.9898,
|
| 2677 |
+
"step": 372
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 2.165457184325109,
|
| 2681 |
+
"grad_norm": 3.479367971420288,
|
| 2682 |
+
"learning_rate": 1.1779460585363944e-06,
|
| 2683 |
+
"loss": 13.835,
|
| 2684 |
+
"step": 373
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 2.1712626995645863,
|
| 2688 |
+
"grad_norm": 3.9154791831970215,
|
| 2689 |
+
"learning_rate": 1.0926199633097157e-06,
|
| 2690 |
+
"loss": 14.9181,
|
| 2691 |
+
"step": 374
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 2.1770682148040637,
|
| 2695 |
+
"grad_norm": 4.3506364822387695,
|
| 2696 |
+
"learning_rate": 1.0104673978866164e-06,
|
| 2697 |
+
"loss": 12.7378,
|
| 2698 |
+
"step": 375
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 2.1828737300435415,
|
| 2702 |
+
"grad_norm": 4.499413013458252,
|
| 2703 |
+
"learning_rate": 9.314936930293283e-07,
|
| 2704 |
+
"loss": 13.1438,
|
| 2705 |
+
"step": 376
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 2.188679245283019,
|
| 2709 |
+
"grad_norm": 5.159627914428711,
|
| 2710 |
+
"learning_rate": 8.557039732283944e-07,
|
| 2711 |
+
"loss": 13.8225,
|
| 2712 |
+
"step": 377
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 2.194484760522496,
|
| 2716 |
+
"grad_norm": 4.514315605163574,
|
| 2717 |
+
"learning_rate": 7.83103156370113e-07,
|
| 2718 |
+
"loss": 11.8075,
|
| 2719 |
+
"step": 378
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 2.200290275761974,
|
| 2723 |
+
"grad_norm": 4.398647785186768,
|
| 2724 |
+
"learning_rate": 7.136959534174592e-07,
|
| 2725 |
+
"loss": 12.3044,
|
| 2726 |
+
"step": 379
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 2.2060957910014514,
|
| 2730 |
+
"grad_norm": 3.818833827972412,
|
| 2731 |
+
"learning_rate": 6.474868681043578e-07,
|
| 2732 |
+
"loss": 11.623,
|
| 2733 |
+
"step": 380
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 2.2119013062409287,
|
| 2737 |
+
"grad_norm": 5.551044940948486,
|
| 2738 |
+
"learning_rate": 5.844801966434832e-07,
|
| 2739 |
+
"loss": 9.8352,
|
| 2740 |
+
"step": 381
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 2.2177068214804065,
|
| 2744 |
+
"grad_norm": 5.551724433898926,
|
| 2745 |
+
"learning_rate": 5.246800274474439e-07,
|
| 2746 |
+
"loss": 8.6958,
|
| 2747 |
+
"step": 382
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 2.223512336719884,
|
| 2751 |
+
"grad_norm": 5.6569437980651855,
|
| 2752 |
+
"learning_rate": 4.680902408635335e-07,
|
| 2753 |
+
"loss": 7.2665,
|
| 2754 |
+
"step": 383
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 2.2293178519593613,
|
| 2758 |
+
"grad_norm": 6.576578617095947,
|
| 2759 |
+
"learning_rate": 4.1471450892189846e-07,
|
| 2760 |
+
"loss": 8.245,
|
| 2761 |
+
"step": 384
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 2.235123367198839,
|
| 2765 |
+
"grad_norm": 5.94155216217041,
|
| 2766 |
+
"learning_rate": 3.6455629509730136e-07,
|
| 2767 |
+
"loss": 6.5846,
|
| 2768 |
+
"step": 385
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 2.2409288824383164,
|
| 2772 |
+
"grad_norm": 5.242124557495117,
|
| 2773 |
+
"learning_rate": 3.1761885408435054e-07,
|
| 2774 |
+
"loss": 6.9586,
|
| 2775 |
+
"step": 386
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 2.246734397677794,
|
| 2779 |
+
"grad_norm": 5.365730285644531,
|
| 2780 |
+
"learning_rate": 2.7390523158633554e-07,
|
| 2781 |
+
"loss": 8.1522,
|
| 2782 |
+
"step": 387
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 2.252539912917271,
|
| 2786 |
+
"grad_norm": 3.7668797969818115,
|
| 2787 |
+
"learning_rate": 2.334182641175686e-07,
|
| 2788 |
+
"loss": 13.7066,
|
| 2789 |
+
"step": 388
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 2.258345428156749,
|
| 2793 |
+
"grad_norm": 3.5341620445251465,
|
| 2794 |
+
"learning_rate": 1.9616057881935436e-07,
|
| 2795 |
+
"loss": 20.6591,
|
| 2796 |
+
"step": 389
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 2.2641509433962264,
|
| 2800 |
+
"grad_norm": 2.6963229179382324,
|
| 2801 |
+
"learning_rate": 1.6213459328950352e-07,
|
| 2802 |
+
"loss": 21.3613,
|
| 2803 |
+
"step": 390
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 2.269956458635704,
|
| 2807 |
+
"grad_norm": 2.5437979698181152,
|
| 2808 |
+
"learning_rate": 1.3134251542544774e-07,
|
| 2809 |
+
"loss": 25.4584,
|
| 2810 |
+
"step": 391
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 2.2757619738751815,
|
| 2814 |
+
"grad_norm": 2.9237968921661377,
|
| 2815 |
+
"learning_rate": 1.0378634328099269e-07,
|
| 2816 |
+
"loss": 22.8008,
|
| 2817 |
+
"step": 392
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 2.281567489114659,
|
| 2821 |
+
"grad_norm": 3.5814568996429443,
|
| 2822 |
+
"learning_rate": 7.946786493666647e-08,
|
| 2823 |
+
"loss": 25.1581,
|
| 2824 |
+
"step": 393
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 2.2873730043541363,
|
| 2828 |
+
"grad_norm": 3.174384593963623,
|
| 2829 |
+
"learning_rate": 5.838865838366792e-08,
|
| 2830 |
+
"loss": 26.3703,
|
| 2831 |
+
"step": 394
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 2.293178519593614,
|
| 2835 |
+
"grad_norm": 3.5474047660827637,
|
| 2836 |
+
"learning_rate": 4.055009142152067e-08,
|
| 2837 |
+
"loss": 25.6042,
|
| 2838 |
+
"step": 395
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 2.2989840348330914,
|
| 2842 |
+
"grad_norm": 3.0559208393096924,
|
| 2843 |
+
"learning_rate": 2.595332156925534e-08,
|
| 2844 |
+
"loss": 25.875,
|
| 2845 |
+
"step": 396
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 2.304789550072569,
|
| 2849 |
+
"grad_norm": 3.6101160049438477,
|
| 2850 |
+
"learning_rate": 1.4599295990352924e-08,
|
| 2851 |
+
"loss": 23.3597,
|
| 2852 |
+
"step": 397
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 2.3105950653120466,
|
| 2856 |
+
"grad_norm": 3.7612712383270264,
|
| 2857 |
+
"learning_rate": 6.488751431266149e-09,
|
| 2858 |
+
"loss": 19.0682,
|
| 2859 |
+
"step": 398
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 2.316400580551524,
|
| 2863 |
+
"grad_norm": 3.835679054260254,
|
| 2864 |
+
"learning_rate": 1.622214173602199e-09,
|
| 2865 |
+
"loss": 23.2814,
|
| 2866 |
+
"step": 399
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 2.3222060957910013,
|
| 2870 |
+
"grad_norm": 3.5171422958374023,
|
| 2871 |
+
"learning_rate": 0.0,
|
| 2872 |
+
"loss": 23.0591,
|
| 2873 |
+
"step": 400
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 2.3222060957910013,
|
| 2877 |
+
"eval_loss": 4.339465618133545,
|
| 2878 |
+
"eval_runtime": 3.2269,
|
| 2879 |
+
"eval_samples_per_second": 89.868,
|
| 2880 |
+
"eval_steps_per_second": 22.622,
|
| 2881 |
+
"step": 400
|
| 2882 |
}
|
| 2883 |
],
|
| 2884 |
"logging_steps": 1,
|
|
|
|
| 2902 |
"should_evaluate": false,
|
| 2903 |
"should_log": false,
|
| 2904 |
"should_save": true,
|
| 2905 |
+
"should_training_stop": true
|
| 2906 |
},
|
| 2907 |
"attributes": {}
|
| 2908 |
}
|
| 2909 |
},
|
| 2910 |
+
"total_flos": 7519492758306816.0,
|
| 2911 |
"train_batch_size": 8,
|
| 2912 |
"trial_name": null,
|
| 2913 |
"trial_params": null
|