Training in progress, step 972, 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 2519936480
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46f5da6a05e33204e884fab0a6fa991ee04cec04b4b9d3367891a15561083a1b
|
| 3 |
size 2519936480
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1264135916
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:54c8c692eb6c581018af077d48fe6cd6898d97810cf025603486889eeaf0aac7
|
| 3 |
size 1264135916
|
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 sha256:181c5f0270cf39930062ddfa3767a2481d0c360f120b11f8e25dbf533a1cdaba
|
| 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 10233
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19cebc5db1bd11b1e7f7900e436918182f4f7343610d76be4b3ddbb754f036cb
|
| 3 |
size 10233
|
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": 0.
|
| 6 |
"eval_steps": 485,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2564,6 +2564,868 @@
|
|
| 2564 |
"learning_rate": 0.0005,
|
| 2565 |
"loss": 1.0296608209609985,
|
| 2566 |
"step": 728
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2567 |
}
|
| 2568 |
],
|
| 2569 |
"logging_steps": 2,
|
|
@@ -2583,7 +3445,7 @@
|
|
| 2583 |
"attributes": {}
|
| 2584 |
}
|
| 2585 |
},
|
| 2586 |
-
"total_flos":
|
| 2587 |
"train_batch_size": 8,
|
| 2588 |
"trial_name": null,
|
| 2589 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.401031459515214,
|
| 6 |
"eval_steps": 485,
|
| 7 |
+
"global_step": 972,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2564 |
"learning_rate": 0.0005,
|
| 2565 |
"loss": 1.0296608209609985,
|
| 2566 |
"step": 728
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 0.3011861784424961,
|
| 2570 |
+
"grad_norm": 0.4547019302845001,
|
| 2571 |
+
"learning_rate": 0.0005,
|
| 2572 |
+
"loss": 0.7701424956321716,
|
| 2573 |
+
"step": 730
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 0.30201134605466734,
|
| 2577 |
+
"grad_norm": 0.709504246711731,
|
| 2578 |
+
"learning_rate": 0.0005,
|
| 2579 |
+
"loss": 0.8757145404815674,
|
| 2580 |
+
"step": 732
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.30283651366683856,
|
| 2584 |
+
"grad_norm": 0.4775760769844055,
|
| 2585 |
+
"learning_rate": 0.0005,
|
| 2586 |
+
"loss": 1.422513723373413,
|
| 2587 |
+
"step": 734
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.3036616812790098,
|
| 2591 |
+
"grad_norm": 0.5066021680831909,
|
| 2592 |
+
"learning_rate": 0.0005,
|
| 2593 |
+
"loss": 1.1779431104660034,
|
| 2594 |
+
"step": 736
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 0.304486848891181,
|
| 2598 |
+
"grad_norm": 0.37824875116348267,
|
| 2599 |
+
"learning_rate": 0.0005,
|
| 2600 |
+
"loss": 1.396673560142517,
|
| 2601 |
+
"step": 738
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.30531201650335227,
|
| 2605 |
+
"grad_norm": 0.47177132964134216,
|
| 2606 |
+
"learning_rate": 0.0005,
|
| 2607 |
+
"loss": 0.6887521743774414,
|
| 2608 |
+
"step": 740
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 0.3061371841155235,
|
| 2612 |
+
"grad_norm": 0.4736272692680359,
|
| 2613 |
+
"learning_rate": 0.0005,
|
| 2614 |
+
"loss": 0.951183021068573,
|
| 2615 |
+
"step": 742
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 0.3069623517276947,
|
| 2619 |
+
"grad_norm": 0.5575990080833435,
|
| 2620 |
+
"learning_rate": 0.0005,
|
| 2621 |
+
"loss": 0.5598281621932983,
|
| 2622 |
+
"step": 744
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 0.3077875193398659,
|
| 2626 |
+
"grad_norm": 0.6142449378967285,
|
| 2627 |
+
"learning_rate": 0.0005,
|
| 2628 |
+
"loss": 0.8059773445129395,
|
| 2629 |
+
"step": 746
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.30861268695203714,
|
| 2633 |
+
"grad_norm": 0.3264317512512207,
|
| 2634 |
+
"learning_rate": 0.0005,
|
| 2635 |
+
"loss": 1.2799454927444458,
|
| 2636 |
+
"step": 748
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 0.30943785456420836,
|
| 2640 |
+
"grad_norm": 0.6371709704399109,
|
| 2641 |
+
"learning_rate": 0.0005,
|
| 2642 |
+
"loss": 0.9574125409126282,
|
| 2643 |
+
"step": 750
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 0.3102630221763796,
|
| 2647 |
+
"grad_norm": 0.3501775860786438,
|
| 2648 |
+
"learning_rate": 0.0005,
|
| 2649 |
+
"loss": 1.5243263244628906,
|
| 2650 |
+
"step": 752
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 0.3110881897885508,
|
| 2654 |
+
"grad_norm": 0.7664023041725159,
|
| 2655 |
+
"learning_rate": 0.0005,
|
| 2656 |
+
"loss": 0.8435171246528625,
|
| 2657 |
+
"step": 754
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 0.311913357400722,
|
| 2661 |
+
"grad_norm": 0.47338634729385376,
|
| 2662 |
+
"learning_rate": 0.0005,
|
| 2663 |
+
"loss": 1.2635228633880615,
|
| 2664 |
+
"step": 756
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 0.31273852501289323,
|
| 2668 |
+
"grad_norm": 0.5893837213516235,
|
| 2669 |
+
"learning_rate": 0.0005,
|
| 2670 |
+
"loss": 1.2268116474151611,
|
| 2671 |
+
"step": 758
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 0.31356369262506445,
|
| 2675 |
+
"grad_norm": 0.3971027433872223,
|
| 2676 |
+
"learning_rate": 0.0005,
|
| 2677 |
+
"loss": 0.9376073479652405,
|
| 2678 |
+
"step": 760
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 0.31438886023723567,
|
| 2682 |
+
"grad_norm": 0.4974158704280853,
|
| 2683 |
+
"learning_rate": 0.0005,
|
| 2684 |
+
"loss": 1.1054447889328003,
|
| 2685 |
+
"step": 762
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 0.3152140278494069,
|
| 2689 |
+
"grad_norm": 0.4881732165813446,
|
| 2690 |
+
"learning_rate": 0.0005,
|
| 2691 |
+
"loss": 1.2376474142074585,
|
| 2692 |
+
"step": 764
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 0.31603919546157816,
|
| 2696 |
+
"grad_norm": 0.4697839915752411,
|
| 2697 |
+
"learning_rate": 0.0005,
|
| 2698 |
+
"loss": 0.9591520428657532,
|
| 2699 |
+
"step": 766
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 0.3168643630737494,
|
| 2703 |
+
"grad_norm": 0.5661628842353821,
|
| 2704 |
+
"learning_rate": 0.0005,
|
| 2705 |
+
"loss": 0.5725570321083069,
|
| 2706 |
+
"step": 768
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 0.3176895306859206,
|
| 2710 |
+
"grad_norm": 0.41950663924217224,
|
| 2711 |
+
"learning_rate": 0.0005,
|
| 2712 |
+
"loss": 0.8325156569480896,
|
| 2713 |
+
"step": 770
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.3185146982980918,
|
| 2717 |
+
"grad_norm": 0.4491046071052551,
|
| 2718 |
+
"learning_rate": 0.0005,
|
| 2719 |
+
"loss": 0.8748528957366943,
|
| 2720 |
+
"step": 772
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 0.31933986591026303,
|
| 2724 |
+
"grad_norm": 0.6069282293319702,
|
| 2725 |
+
"learning_rate": 0.0005,
|
| 2726 |
+
"loss": 0.6700723171234131,
|
| 2727 |
+
"step": 774
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 0.32016503352243425,
|
| 2731 |
+
"grad_norm": 0.387180894613266,
|
| 2732 |
+
"learning_rate": 0.0005,
|
| 2733 |
+
"loss": 1.208810806274414,
|
| 2734 |
+
"step": 776
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 0.32099020113460547,
|
| 2738 |
+
"grad_norm": 0.3678168058395386,
|
| 2739 |
+
"learning_rate": 0.0005,
|
| 2740 |
+
"loss": 0.9034237861633301,
|
| 2741 |
+
"step": 778
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 0.3218153687467767,
|
| 2745 |
+
"grad_norm": 0.39267754554748535,
|
| 2746 |
+
"learning_rate": 0.0005,
|
| 2747 |
+
"loss": 1.106095790863037,
|
| 2748 |
+
"step": 780
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.3226405363589479,
|
| 2752 |
+
"grad_norm": 0.5397934317588806,
|
| 2753 |
+
"learning_rate": 0.0005,
|
| 2754 |
+
"loss": 1.2158617973327637,
|
| 2755 |
+
"step": 782
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 0.3234657039711191,
|
| 2759 |
+
"grad_norm": 0.4539080560207367,
|
| 2760 |
+
"learning_rate": 0.0005,
|
| 2761 |
+
"loss": 1.0038114786148071,
|
| 2762 |
+
"step": 784
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 0.32429087158329034,
|
| 2766 |
+
"grad_norm": 0.49578386545181274,
|
| 2767 |
+
"learning_rate": 0.0005,
|
| 2768 |
+
"loss": 1.2935315370559692,
|
| 2769 |
+
"step": 786
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 0.32511603919546156,
|
| 2773 |
+
"grad_norm": 0.3788098394870758,
|
| 2774 |
+
"learning_rate": 0.0005,
|
| 2775 |
+
"loss": 1.3695635795593262,
|
| 2776 |
+
"step": 788
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 0.3259412068076328,
|
| 2780 |
+
"grad_norm": 0.5304418206214905,
|
| 2781 |
+
"learning_rate": 0.0005,
|
| 2782 |
+
"loss": 1.2912182807922363,
|
| 2783 |
+
"step": 790
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 0.326766374419804,
|
| 2787 |
+
"grad_norm": 0.5684431195259094,
|
| 2788 |
+
"learning_rate": 0.0005,
|
| 2789 |
+
"loss": 1.1963849067687988,
|
| 2790 |
+
"step": 792
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.32759154203197527,
|
| 2794 |
+
"grad_norm": 0.5429754853248596,
|
| 2795 |
+
"learning_rate": 0.0005,
|
| 2796 |
+
"loss": 0.9575740098953247,
|
| 2797 |
+
"step": 794
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 0.3284167096441465,
|
| 2801 |
+
"grad_norm": 0.6165698170661926,
|
| 2802 |
+
"learning_rate": 0.0005,
|
| 2803 |
+
"loss": 0.6333395838737488,
|
| 2804 |
+
"step": 796
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 0.3292418772563177,
|
| 2808 |
+
"grad_norm": 0.6064788699150085,
|
| 2809 |
+
"learning_rate": 0.0005,
|
| 2810 |
+
"loss": 0.8153479099273682,
|
| 2811 |
+
"step": 798
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 0.3300670448684889,
|
| 2815 |
+
"grad_norm": 0.4232187271118164,
|
| 2816 |
+
"learning_rate": 0.0005,
|
| 2817 |
+
"loss": 1.4265881776809692,
|
| 2818 |
+
"step": 800
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.33089221248066014,
|
| 2822 |
+
"grad_norm": 0.4814676344394684,
|
| 2823 |
+
"learning_rate": 0.0005,
|
| 2824 |
+
"loss": 1.1406737565994263,
|
| 2825 |
+
"step": 802
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.33171738009283136,
|
| 2829 |
+
"grad_norm": 0.419049471616745,
|
| 2830 |
+
"learning_rate": 0.0005,
|
| 2831 |
+
"loss": 1.4567514657974243,
|
| 2832 |
+
"step": 804
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.3325425477050026,
|
| 2836 |
+
"grad_norm": 0.7689643502235413,
|
| 2837 |
+
"learning_rate": 0.0005,
|
| 2838 |
+
"loss": 0.9619349241256714,
|
| 2839 |
+
"step": 806
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.3333677153171738,
|
| 2843 |
+
"grad_norm": 0.4174315631389618,
|
| 2844 |
+
"learning_rate": 0.0005,
|
| 2845 |
+
"loss": 0.917432963848114,
|
| 2846 |
+
"step": 808
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.334192882929345,
|
| 2850 |
+
"grad_norm": 0.5049794316291809,
|
| 2851 |
+
"learning_rate": 0.0005,
|
| 2852 |
+
"loss": 1.0135047435760498,
|
| 2853 |
+
"step": 810
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.33501805054151623,
|
| 2857 |
+
"grad_norm": 0.39736947417259216,
|
| 2858 |
+
"learning_rate": 0.0005,
|
| 2859 |
+
"loss": 1.2304755449295044,
|
| 2860 |
+
"step": 812
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.33584321815368745,
|
| 2864 |
+
"grad_norm": 0.5136732459068298,
|
| 2865 |
+
"learning_rate": 0.0005,
|
| 2866 |
+
"loss": 0.6235193014144897,
|
| 2867 |
+
"step": 814
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.33666838576585867,
|
| 2871 |
+
"grad_norm": 0.6358846426010132,
|
| 2872 |
+
"learning_rate": 0.0005,
|
| 2873 |
+
"loss": 0.9476469159126282,
|
| 2874 |
+
"step": 816
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.3374935533780299,
|
| 2878 |
+
"grad_norm": 0.36760398745536804,
|
| 2879 |
+
"learning_rate": 0.0005,
|
| 2880 |
+
"loss": 1.1756913661956787,
|
| 2881 |
+
"step": 818
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.33831872099020116,
|
| 2885 |
+
"grad_norm": 0.6921237111091614,
|
| 2886 |
+
"learning_rate": 0.0005,
|
| 2887 |
+
"loss": 0.5435713529586792,
|
| 2888 |
+
"step": 820
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.3391438886023724,
|
| 2892 |
+
"grad_norm": 0.6190113425254822,
|
| 2893 |
+
"learning_rate": 0.0005,
|
| 2894 |
+
"loss": 0.6276876926422119,
|
| 2895 |
+
"step": 822
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 0.3399690562145436,
|
| 2899 |
+
"grad_norm": 0.5015466809272766,
|
| 2900 |
+
"learning_rate": 0.0005,
|
| 2901 |
+
"loss": 1.2893644571304321,
|
| 2902 |
+
"step": 824
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 0.3407942238267148,
|
| 2906 |
+
"grad_norm": 0.6742082834243774,
|
| 2907 |
+
"learning_rate": 0.0005,
|
| 2908 |
+
"loss": 0.9958200454711914,
|
| 2909 |
+
"step": 826
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 0.34161939143888603,
|
| 2913 |
+
"grad_norm": 0.43738895654678345,
|
| 2914 |
+
"learning_rate": 0.0005,
|
| 2915 |
+
"loss": 0.8957383632659912,
|
| 2916 |
+
"step": 828
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.34244455905105725,
|
| 2920 |
+
"grad_norm": 0.40049517154693604,
|
| 2921 |
+
"learning_rate": 0.0005,
|
| 2922 |
+
"loss": 0.86797696352005,
|
| 2923 |
+
"step": 830
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 0.34326972666322847,
|
| 2927 |
+
"grad_norm": 0.678524911403656,
|
| 2928 |
+
"learning_rate": 0.0005,
|
| 2929 |
+
"loss": 1.1833425760269165,
|
| 2930 |
+
"step": 832
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 0.3440948942753997,
|
| 2934 |
+
"grad_norm": 0.3560514450073242,
|
| 2935 |
+
"learning_rate": 0.0005,
|
| 2936 |
+
"loss": 1.247701644897461,
|
| 2937 |
+
"step": 834
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 0.3449200618875709,
|
| 2941 |
+
"grad_norm": 0.7766836881637573,
|
| 2942 |
+
"learning_rate": 0.0005,
|
| 2943 |
+
"loss": 0.6366620063781738,
|
| 2944 |
+
"step": 836
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 0.3457452294997421,
|
| 2948 |
+
"grad_norm": 0.3957395553588867,
|
| 2949 |
+
"learning_rate": 0.0005,
|
| 2950 |
+
"loss": 1.452673316001892,
|
| 2951 |
+
"step": 838
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 0.34657039711191334,
|
| 2955 |
+
"grad_norm": 0.6949273347854614,
|
| 2956 |
+
"learning_rate": 0.0005,
|
| 2957 |
+
"loss": 0.6712929010391235,
|
| 2958 |
+
"step": 840
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 0.34739556472408456,
|
| 2962 |
+
"grad_norm": 0.44859611988067627,
|
| 2963 |
+
"learning_rate": 0.0005,
|
| 2964 |
+
"loss": 0.9374242424964905,
|
| 2965 |
+
"step": 842
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 0.3482207323362558,
|
| 2969 |
+
"grad_norm": 0.44476670026779175,
|
| 2970 |
+
"learning_rate": 0.0005,
|
| 2971 |
+
"loss": 0.9708834290504456,
|
| 2972 |
+
"step": 844
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 0.34904589994842705,
|
| 2976 |
+
"grad_norm": 0.5429829955101013,
|
| 2977 |
+
"learning_rate": 0.0005,
|
| 2978 |
+
"loss": 1.1306025981903076,
|
| 2979 |
+
"step": 846
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 0.34987106756059827,
|
| 2983 |
+
"grad_norm": 0.5670883655548096,
|
| 2984 |
+
"learning_rate": 0.0005,
|
| 2985 |
+
"loss": 0.9808698296546936,
|
| 2986 |
+
"step": 848
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 0.3506962351727695,
|
| 2990 |
+
"grad_norm": 0.4450462758541107,
|
| 2991 |
+
"learning_rate": 0.0005,
|
| 2992 |
+
"loss": 1.2327816486358643,
|
| 2993 |
+
"step": 850
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.3515214027849407,
|
| 2997 |
+
"grad_norm": 0.44823941588401794,
|
| 2998 |
+
"learning_rate": 0.0005,
|
| 2999 |
+
"loss": 1.214652180671692,
|
| 3000 |
+
"step": 852
|
| 3001 |
+
},
|
| 3002 |
+
{
|
| 3003 |
+
"epoch": 0.3523465703971119,
|
| 3004 |
+
"grad_norm": 0.46802523732185364,
|
| 3005 |
+
"learning_rate": 0.0005,
|
| 3006 |
+
"loss": 0.8434693813323975,
|
| 3007 |
+
"step": 854
|
| 3008 |
+
},
|
| 3009 |
+
{
|
| 3010 |
+
"epoch": 0.35317173800928314,
|
| 3011 |
+
"grad_norm": 0.4221685528755188,
|
| 3012 |
+
"learning_rate": 0.0005,
|
| 3013 |
+
"loss": 1.385030746459961,
|
| 3014 |
+
"step": 856
|
| 3015 |
+
},
|
| 3016 |
+
{
|
| 3017 |
+
"epoch": 0.35399690562145436,
|
| 3018 |
+
"grad_norm": 0.3876509368419647,
|
| 3019 |
+
"learning_rate": 0.0005,
|
| 3020 |
+
"loss": 1.3161131143569946,
|
| 3021 |
+
"step": 858
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 0.3548220732336256,
|
| 3025 |
+
"grad_norm": 0.465462327003479,
|
| 3026 |
+
"learning_rate": 0.0005,
|
| 3027 |
+
"loss": 1.246232032775879,
|
| 3028 |
+
"step": 860
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 0.3556472408457968,
|
| 3032 |
+
"grad_norm": 0.38140279054641724,
|
| 3033 |
+
"learning_rate": 0.0005,
|
| 3034 |
+
"loss": 1.1987196207046509,
|
| 3035 |
+
"step": 862
|
| 3036 |
+
},
|
| 3037 |
+
{
|
| 3038 |
+
"epoch": 0.356472408457968,
|
| 3039 |
+
"grad_norm": 0.6231861710548401,
|
| 3040 |
+
"learning_rate": 0.0005,
|
| 3041 |
+
"loss": 0.7024447917938232,
|
| 3042 |
+
"step": 864
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 0.35729757607013923,
|
| 3046 |
+
"grad_norm": 0.8739637136459351,
|
| 3047 |
+
"learning_rate": 0.0005,
|
| 3048 |
+
"loss": 0.9236916899681091,
|
| 3049 |
+
"step": 866
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 0.35812274368231045,
|
| 3053 |
+
"grad_norm": 0.8303645253181458,
|
| 3054 |
+
"learning_rate": 0.0005,
|
| 3055 |
+
"loss": 0.7618856430053711,
|
| 3056 |
+
"step": 868
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 0.35894791129448167,
|
| 3060 |
+
"grad_norm": 0.44744884967803955,
|
| 3061 |
+
"learning_rate": 0.0005,
|
| 3062 |
+
"loss": 1.1455423831939697,
|
| 3063 |
+
"step": 870
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 0.3597730789066529,
|
| 3067 |
+
"grad_norm": 0.44626978039741516,
|
| 3068 |
+
"learning_rate": 0.0005,
|
| 3069 |
+
"loss": 1.1497100591659546,
|
| 3070 |
+
"step": 872
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 0.36059824651882416,
|
| 3074 |
+
"grad_norm": 0.47076717019081116,
|
| 3075 |
+
"learning_rate": 0.0005,
|
| 3076 |
+
"loss": 1.0223735570907593,
|
| 3077 |
+
"step": 874
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 0.3614234141309954,
|
| 3081 |
+
"grad_norm": 0.5181264281272888,
|
| 3082 |
+
"learning_rate": 0.0005,
|
| 3083 |
+
"loss": 0.8119293451309204,
|
| 3084 |
+
"step": 876
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 0.3622485817431666,
|
| 3088 |
+
"grad_norm": 0.5308315753936768,
|
| 3089 |
+
"learning_rate": 0.0005,
|
| 3090 |
+
"loss": 0.8662136793136597,
|
| 3091 |
+
"step": 878
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 0.3630737493553378,
|
| 3095 |
+
"grad_norm": 0.3599519729614258,
|
| 3096 |
+
"learning_rate": 0.0005,
|
| 3097 |
+
"loss": 1.0991331338882446,
|
| 3098 |
+
"step": 880
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 0.36389891696750903,
|
| 3102 |
+
"grad_norm": 0.450533002614975,
|
| 3103 |
+
"learning_rate": 0.0005,
|
| 3104 |
+
"loss": 1.199340581893921,
|
| 3105 |
+
"step": 882
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.36472408457968025,
|
| 3109 |
+
"grad_norm": 0.41602176427841187,
|
| 3110 |
+
"learning_rate": 0.0005,
|
| 3111 |
+
"loss": 0.8881515860557556,
|
| 3112 |
+
"step": 884
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 0.36554925219185147,
|
| 3116 |
+
"grad_norm": 0.3758910894393921,
|
| 3117 |
+
"learning_rate": 0.0005,
|
| 3118 |
+
"loss": 1.0691791772842407,
|
| 3119 |
+
"step": 886
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 0.3663744198040227,
|
| 3123 |
+
"grad_norm": 0.49511483311653137,
|
| 3124 |
+
"learning_rate": 0.0005,
|
| 3125 |
+
"loss": 0.84291672706604,
|
| 3126 |
+
"step": 888
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 0.3671995874161939,
|
| 3130 |
+
"grad_norm": 0.37525108456611633,
|
| 3131 |
+
"learning_rate": 0.0005,
|
| 3132 |
+
"loss": 0.787752091884613,
|
| 3133 |
+
"step": 890
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 0.3680247550283651,
|
| 3137 |
+
"grad_norm": 0.6520910859107971,
|
| 3138 |
+
"learning_rate": 0.0005,
|
| 3139 |
+
"loss": 0.7890579700469971,
|
| 3140 |
+
"step": 892
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 0.36884992264053634,
|
| 3144 |
+
"grad_norm": 0.4799257218837738,
|
| 3145 |
+
"learning_rate": 0.0005,
|
| 3146 |
+
"loss": 0.8910934925079346,
|
| 3147 |
+
"step": 894
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 0.36967509025270756,
|
| 3151 |
+
"grad_norm": 0.3718203604221344,
|
| 3152 |
+
"learning_rate": 0.0005,
|
| 3153 |
+
"loss": 1.2266496419906616,
|
| 3154 |
+
"step": 896
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 0.3705002578648788,
|
| 3158 |
+
"grad_norm": 0.7900125980377197,
|
| 3159 |
+
"learning_rate": 0.0005,
|
| 3160 |
+
"loss": 0.808830738067627,
|
| 3161 |
+
"step": 898
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 0.37132542547705005,
|
| 3165 |
+
"grad_norm": 0.3988662660121918,
|
| 3166 |
+
"learning_rate": 0.0005,
|
| 3167 |
+
"loss": 1.3982020616531372,
|
| 3168 |
+
"step": 900
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 0.37215059308922127,
|
| 3172 |
+
"grad_norm": 0.4183046519756317,
|
| 3173 |
+
"learning_rate": 0.0005,
|
| 3174 |
+
"loss": 0.945640504360199,
|
| 3175 |
+
"step": 902
|
| 3176 |
+
},
|
| 3177 |
+
{
|
| 3178 |
+
"epoch": 0.3729757607013925,
|
| 3179 |
+
"grad_norm": 0.4638324975967407,
|
| 3180 |
+
"learning_rate": 0.0005,
|
| 3181 |
+
"loss": 0.9022805690765381,
|
| 3182 |
+
"step": 904
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 0.3738009283135637,
|
| 3186 |
+
"grad_norm": 0.4181121587753296,
|
| 3187 |
+
"learning_rate": 0.0005,
|
| 3188 |
+
"loss": 1.0504035949707031,
|
| 3189 |
+
"step": 906
|
| 3190 |
+
},
|
| 3191 |
+
{
|
| 3192 |
+
"epoch": 0.3746260959257349,
|
| 3193 |
+
"grad_norm": 0.5623753666877747,
|
| 3194 |
+
"learning_rate": 0.0005,
|
| 3195 |
+
"loss": 0.885411262512207,
|
| 3196 |
+
"step": 908
|
| 3197 |
+
},
|
| 3198 |
+
{
|
| 3199 |
+
"epoch": 0.37545126353790614,
|
| 3200 |
+
"grad_norm": 0.6528892517089844,
|
| 3201 |
+
"learning_rate": 0.0005,
|
| 3202 |
+
"loss": 0.4613204598426819,
|
| 3203 |
+
"step": 910
|
| 3204 |
+
},
|
| 3205 |
+
{
|
| 3206 |
+
"epoch": 0.37627643115007736,
|
| 3207 |
+
"grad_norm": 0.43880847096443176,
|
| 3208 |
+
"learning_rate": 0.0005,
|
| 3209 |
+
"loss": 0.6581244468688965,
|
| 3210 |
+
"step": 912
|
| 3211 |
+
},
|
| 3212 |
+
{
|
| 3213 |
+
"epoch": 0.3771015987622486,
|
| 3214 |
+
"grad_norm": 0.6024404764175415,
|
| 3215 |
+
"learning_rate": 0.0005,
|
| 3216 |
+
"loss": 0.5904361009597778,
|
| 3217 |
+
"step": 914
|
| 3218 |
+
},
|
| 3219 |
+
{
|
| 3220 |
+
"epoch": 0.3779267663744198,
|
| 3221 |
+
"grad_norm": 0.4631766676902771,
|
| 3222 |
+
"learning_rate": 0.0005,
|
| 3223 |
+
"loss": 0.8536958694458008,
|
| 3224 |
+
"step": 916
|
| 3225 |
+
},
|
| 3226 |
+
{
|
| 3227 |
+
"epoch": 0.378751933986591,
|
| 3228 |
+
"grad_norm": 0.6748937964439392,
|
| 3229 |
+
"learning_rate": 0.0005,
|
| 3230 |
+
"loss": 1.154845952987671,
|
| 3231 |
+
"step": 918
|
| 3232 |
+
},
|
| 3233 |
+
{
|
| 3234 |
+
"epoch": 0.37957710159876223,
|
| 3235 |
+
"grad_norm": 0.5579289793968201,
|
| 3236 |
+
"learning_rate": 0.0005,
|
| 3237 |
+
"loss": 1.3011925220489502,
|
| 3238 |
+
"step": 920
|
| 3239 |
+
},
|
| 3240 |
+
{
|
| 3241 |
+
"epoch": 0.38040226921093345,
|
| 3242 |
+
"grad_norm": 0.4682430326938629,
|
| 3243 |
+
"learning_rate": 0.0005,
|
| 3244 |
+
"loss": 1.104500651359558,
|
| 3245 |
+
"step": 922
|
| 3246 |
+
},
|
| 3247 |
+
{
|
| 3248 |
+
"epoch": 0.38122743682310467,
|
| 3249 |
+
"grad_norm": 0.6007438898086548,
|
| 3250 |
+
"learning_rate": 0.0005,
|
| 3251 |
+
"loss": 1.0890264511108398,
|
| 3252 |
+
"step": 924
|
| 3253 |
+
},
|
| 3254 |
+
{
|
| 3255 |
+
"epoch": 0.3820526044352759,
|
| 3256 |
+
"grad_norm": 0.3681229054927826,
|
| 3257 |
+
"learning_rate": 0.0005,
|
| 3258 |
+
"loss": 1.0576262474060059,
|
| 3259 |
+
"step": 926
|
| 3260 |
+
},
|
| 3261 |
+
{
|
| 3262 |
+
"epoch": 0.38287777204744716,
|
| 3263 |
+
"grad_norm": 0.3867166340351105,
|
| 3264 |
+
"learning_rate": 0.0005,
|
| 3265 |
+
"loss": 1.4854787588119507,
|
| 3266 |
+
"step": 928
|
| 3267 |
+
},
|
| 3268 |
+
{
|
| 3269 |
+
"epoch": 0.3837029396596184,
|
| 3270 |
+
"grad_norm": 0.6359395384788513,
|
| 3271 |
+
"learning_rate": 0.0005,
|
| 3272 |
+
"loss": 0.3282276690006256,
|
| 3273 |
+
"step": 930
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 0.3845281072717896,
|
| 3277 |
+
"grad_norm": 0.46864011883735657,
|
| 3278 |
+
"learning_rate": 0.0005,
|
| 3279 |
+
"loss": 1.2231162786483765,
|
| 3280 |
+
"step": 932
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"epoch": 0.3853532748839608,
|
| 3284 |
+
"grad_norm": 0.42163488268852234,
|
| 3285 |
+
"learning_rate": 0.0005,
|
| 3286 |
+
"loss": 0.799576997756958,
|
| 3287 |
+
"step": 934
|
| 3288 |
+
},
|
| 3289 |
+
{
|
| 3290 |
+
"epoch": 0.38617844249613204,
|
| 3291 |
+
"grad_norm": 0.5348471403121948,
|
| 3292 |
+
"learning_rate": 0.0005,
|
| 3293 |
+
"loss": 1.215818166732788,
|
| 3294 |
+
"step": 936
|
| 3295 |
+
},
|
| 3296 |
+
{
|
| 3297 |
+
"epoch": 0.38700361010830325,
|
| 3298 |
+
"grad_norm": 0.47862353920936584,
|
| 3299 |
+
"learning_rate": 0.0005,
|
| 3300 |
+
"loss": 1.119133472442627,
|
| 3301 |
+
"step": 938
|
| 3302 |
+
},
|
| 3303 |
+
{
|
| 3304 |
+
"epoch": 0.38782877772047447,
|
| 3305 |
+
"grad_norm": 0.6671462059020996,
|
| 3306 |
+
"learning_rate": 0.0005,
|
| 3307 |
+
"loss": 0.6009505987167358,
|
| 3308 |
+
"step": 940
|
| 3309 |
+
},
|
| 3310 |
+
{
|
| 3311 |
+
"epoch": 0.3886539453326457,
|
| 3312 |
+
"grad_norm": 0.654689610004425,
|
| 3313 |
+
"learning_rate": 0.0005,
|
| 3314 |
+
"loss": 0.9069526195526123,
|
| 3315 |
+
"step": 942
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 0.3894791129448169,
|
| 3319 |
+
"grad_norm": 0.6140965223312378,
|
| 3320 |
+
"learning_rate": 0.0005,
|
| 3321 |
+
"loss": 1.1941167116165161,
|
| 3322 |
+
"step": 944
|
| 3323 |
+
},
|
| 3324 |
+
{
|
| 3325 |
+
"epoch": 0.3903042805569881,
|
| 3326 |
+
"grad_norm": 0.4021962881088257,
|
| 3327 |
+
"learning_rate": 0.0005,
|
| 3328 |
+
"loss": 1.2709530591964722,
|
| 3329 |
+
"step": 946
|
| 3330 |
+
},
|
| 3331 |
+
{
|
| 3332 |
+
"epoch": 0.39112944816915934,
|
| 3333 |
+
"grad_norm": 0.4146452844142914,
|
| 3334 |
+
"learning_rate": 0.0005,
|
| 3335 |
+
"loss": 1.1812446117401123,
|
| 3336 |
+
"step": 948
|
| 3337 |
+
},
|
| 3338 |
+
{
|
| 3339 |
+
"epoch": 0.39195461578133056,
|
| 3340 |
+
"grad_norm": 0.6948921084403992,
|
| 3341 |
+
"learning_rate": 0.0005,
|
| 3342 |
+
"loss": 0.7930907011032104,
|
| 3343 |
+
"step": 950
|
| 3344 |
+
},
|
| 3345 |
+
{
|
| 3346 |
+
"epoch": 0.3927797833935018,
|
| 3347 |
+
"grad_norm": 0.4594041705131531,
|
| 3348 |
+
"learning_rate": 0.0005,
|
| 3349 |
+
"loss": 1.1281040906906128,
|
| 3350 |
+
"step": 952
|
| 3351 |
+
},
|
| 3352 |
+
{
|
| 3353 |
+
"epoch": 0.39360495100567305,
|
| 3354 |
+
"grad_norm": 0.6090090274810791,
|
| 3355 |
+
"learning_rate": 0.0005,
|
| 3356 |
+
"loss": 0.9690212607383728,
|
| 3357 |
+
"step": 954
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 0.39443011861784427,
|
| 3361 |
+
"grad_norm": 0.49895620346069336,
|
| 3362 |
+
"learning_rate": 0.0005,
|
| 3363 |
+
"loss": 0.9293103814125061,
|
| 3364 |
+
"step": 956
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 0.3952552862300155,
|
| 3368 |
+
"grad_norm": 0.5032721757888794,
|
| 3369 |
+
"learning_rate": 0.0005,
|
| 3370 |
+
"loss": 1.3478631973266602,
|
| 3371 |
+
"step": 958
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.3960804538421867,
|
| 3375 |
+
"grad_norm": 0.44382214546203613,
|
| 3376 |
+
"learning_rate": 0.0005,
|
| 3377 |
+
"loss": 1.455615520477295,
|
| 3378 |
+
"step": 960
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.3969056214543579,
|
| 3382 |
+
"grad_norm": 0.4999854266643524,
|
| 3383 |
+
"learning_rate": 0.0005,
|
| 3384 |
+
"loss": 1.059862732887268,
|
| 3385 |
+
"step": 962
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.39773078906652914,
|
| 3389 |
+
"grad_norm": 0.4809070825576782,
|
| 3390 |
+
"learning_rate": 0.0005,
|
| 3391 |
+
"loss": 0.7847961187362671,
|
| 3392 |
+
"step": 964
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.39855595667870036,
|
| 3396 |
+
"grad_norm": 0.677364706993103,
|
| 3397 |
+
"learning_rate": 0.0005,
|
| 3398 |
+
"loss": 0.3548208475112915,
|
| 3399 |
+
"step": 966
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.3993811242908716,
|
| 3403 |
+
"grad_norm": 0.35379257798194885,
|
| 3404 |
+
"learning_rate": 0.0005,
|
| 3405 |
+
"loss": 1.1698503494262695,
|
| 3406 |
+
"step": 968
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.4002062919030428,
|
| 3410 |
+
"grad_norm": 0.47841575741767883,
|
| 3411 |
+
"learning_rate": 0.0005,
|
| 3412 |
+
"loss": 0.6756126284599304,
|
| 3413 |
+
"step": 970
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.4002062919030428,
|
| 3417 |
+
"eval_loss": 0.7067621946334839,
|
| 3418 |
+
"eval_runtime": 7.3635,
|
| 3419 |
+
"eval_samples_per_second": 106.471,
|
| 3420 |
+
"eval_steps_per_second": 13.309,
|
| 3421 |
+
"step": 970
|
| 3422 |
+
},
|
| 3423 |
+
{
|
| 3424 |
+
"epoch": 0.401031459515214,
|
| 3425 |
+
"grad_norm": 0.6655641794204712,
|
| 3426 |
+
"learning_rate": 0.0005,
|
| 3427 |
+
"loss": 0.3434591293334961,
|
| 3428 |
+
"step": 972
|
| 3429 |
}
|
| 3430 |
],
|
| 3431 |
"logging_steps": 2,
|
|
|
|
| 3445 |
"attributes": {}
|
| 3446 |
}
|
| 3447 |
},
|
| 3448 |
+
"total_flos": 5.407389838685568e+16,
|
| 3449 |
"train_batch_size": 8,
|
| 3450 |
"trial_name": null,
|
| 3451 |
"trial_params": null
|