Upload folder using huggingface_hub
Browse files
model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4991459544
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81ab56e8590993681dead0f3d3cc6f45a8eb23b61469318abbe93ba8b801c546
|
| 3 |
size 4991459544
|
model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4991757456
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8cf00992d4e87e8b9f1608fabcc52dd672432390f412288836a3fa223e221dc4
|
| 3 |
size 4991757456
|
model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4947691960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:773165bd6b4f416dd506aa299b1989b6cf65e31562900a85d33e3787fd1db543
|
| 3 |
size 4947691960
|
model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1863387424
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1adc5a0afafce5cf96019957a1e54e82b326f1509d538f929f01e595d5ee2fe7
|
| 3 |
size 1863387424
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 100000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2536,6 +2536,726 @@
|
|
| 2536 |
"memory(GiB)": 75.24,
|
| 2537 |
"step": 1400,
|
| 2538 |
"train_speed(iter/s)": 0.018457
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2539 |
}
|
| 2540 |
],
|
| 2541 |
"logging_steps": 5,
|
|
@@ -2555,7 +3275,7 @@
|
|
| 2555 |
"attributes": {}
|
| 2556 |
}
|
| 2557 |
},
|
| 2558 |
-
"total_flos": 1.
|
| 2559 |
"train_batch_size": 4,
|
| 2560 |
"trial_name": null,
|
| 2561 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.42406885229004543,
|
| 5 |
"eval_steps": 100000,
|
| 6 |
+
"global_step": 1800,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2536 |
"memory(GiB)": 75.24,
|
| 2537 |
"step": 1400,
|
| 2538 |
"train_speed(iter/s)": 0.018457
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 0.3310092985930632,
|
| 2542 |
+
"grad_norm": 2.234375,
|
| 2543 |
+
"learning_rate": 7.533056236688856e-06,
|
| 2544 |
+
"loss": 0.05440799,
|
| 2545 |
+
"memory(GiB)": 75.24,
|
| 2546 |
+
"step": 1405,
|
| 2547 |
+
"train_speed(iter/s)": 0.018449
|
| 2548 |
+
},
|
| 2549 |
+
{
|
| 2550 |
+
"epoch": 0.33218726762720224,
|
| 2551 |
+
"grad_norm": 1.875,
|
| 2552 |
+
"learning_rate": 7.517079679626914e-06,
|
| 2553 |
+
"loss": 0.04769931,
|
| 2554 |
+
"memory(GiB)": 75.24,
|
| 2555 |
+
"step": 1410,
|
| 2556 |
+
"train_speed(iter/s)": 0.01845
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 0.33336523666134127,
|
| 2560 |
+
"grad_norm": 2.8125,
|
| 2561 |
+
"learning_rate": 7.501068624863622e-06,
|
| 2562 |
+
"loss": 0.06096817,
|
| 2563 |
+
"memory(GiB)": 75.24,
|
| 2564 |
+
"step": 1415,
|
| 2565 |
+
"train_speed(iter/s)": 0.01845
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.3345432056954803,
|
| 2569 |
+
"grad_norm": 2.09375,
|
| 2570 |
+
"learning_rate": 7.485023291837641e-06,
|
| 2571 |
+
"loss": 0.05072919,
|
| 2572 |
+
"memory(GiB)": 75.24,
|
| 2573 |
+
"step": 1420,
|
| 2574 |
+
"train_speed(iter/s)": 0.01845
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.3357211747296193,
|
| 2578 |
+
"grad_norm": 1.921875,
|
| 2579 |
+
"learning_rate": 7.468943900457423e-06,
|
| 2580 |
+
"loss": 0.05104024,
|
| 2581 |
+
"memory(GiB)": 75.24,
|
| 2582 |
+
"step": 1425,
|
| 2583 |
+
"train_speed(iter/s)": 0.01845
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 0.3368991437637583,
|
| 2587 |
+
"grad_norm": 1.703125,
|
| 2588 |
+
"learning_rate": 7.4528306710982135e-06,
|
| 2589 |
+
"loss": 0.05378548,
|
| 2590 |
+
"memory(GiB)": 75.24,
|
| 2591 |
+
"step": 1430,
|
| 2592 |
+
"train_speed(iter/s)": 0.018451
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.3380771127978973,
|
| 2596 |
+
"grad_norm": 1.796875,
|
| 2597 |
+
"learning_rate": 7.436683824599015e-06,
|
| 2598 |
+
"loss": 0.05540187,
|
| 2599 |
+
"memory(GiB)": 75.24,
|
| 2600 |
+
"step": 1435,
|
| 2601 |
+
"train_speed(iter/s)": 0.018451
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.33925508183203634,
|
| 2605 |
+
"grad_norm": 2.25,
|
| 2606 |
+
"learning_rate": 7.4205035822595716e-06,
|
| 2607 |
+
"loss": 0.05696809,
|
| 2608 |
+
"memory(GiB)": 75.24,
|
| 2609 |
+
"step": 1440,
|
| 2610 |
+
"train_speed(iter/s)": 0.018452
|
| 2611 |
+
},
|
| 2612 |
+
{
|
| 2613 |
+
"epoch": 0.34043305086617537,
|
| 2614 |
+
"grad_norm": 1.8671875,
|
| 2615 |
+
"learning_rate": 7.404290165837335e-06,
|
| 2616 |
+
"loss": 0.05431355,
|
| 2617 |
+
"memory(GiB)": 75.24,
|
| 2618 |
+
"step": 1445,
|
| 2619 |
+
"train_speed(iter/s)": 0.018452
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 0.3416110199003144,
|
| 2623 |
+
"grad_norm": 1.7421875,
|
| 2624 |
+
"learning_rate": 7.388043797544415e-06,
|
| 2625 |
+
"loss": 0.04478872,
|
| 2626 |
+
"memory(GiB)": 75.24,
|
| 2627 |
+
"step": 1450,
|
| 2628 |
+
"train_speed(iter/s)": 0.018453
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.34278898893445336,
|
| 2632 |
+
"grad_norm": 1.9609375,
|
| 2633 |
+
"learning_rate": 7.37176470004455e-06,
|
| 2634 |
+
"loss": 0.04893782,
|
| 2635 |
+
"memory(GiB)": 75.24,
|
| 2636 |
+
"step": 1455,
|
| 2637 |
+
"train_speed(iter/s)": 0.018453
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.3439669579685924,
|
| 2641 |
+
"grad_norm": 1.671875,
|
| 2642 |
+
"learning_rate": 7.3554530964500416e-06,
|
| 2643 |
+
"loss": 0.05598937,
|
| 2644 |
+
"memory(GiB)": 75.24,
|
| 2645 |
+
"step": 1460,
|
| 2646 |
+
"train_speed(iter/s)": 0.018454
|
| 2647 |
+
},
|
| 2648 |
+
{
|
| 2649 |
+
"epoch": 0.3451449270027314,
|
| 2650 |
+
"grad_norm": 2.453125,
|
| 2651 |
+
"learning_rate": 7.339109210318704e-06,
|
| 2652 |
+
"loss": 0.05856699,
|
| 2653 |
+
"memory(GiB)": 75.24,
|
| 2654 |
+
"step": 1465,
|
| 2655 |
+
"train_speed(iter/s)": 0.018454
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.34632289603687044,
|
| 2659 |
+
"grad_norm": 2.28125,
|
| 2660 |
+
"learning_rate": 7.322733265650793e-06,
|
| 2661 |
+
"loss": 0.04954901,
|
| 2662 |
+
"memory(GiB)": 75.24,
|
| 2663 |
+
"step": 1470,
|
| 2664 |
+
"train_speed(iter/s)": 0.018454
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 0.34750086507100947,
|
| 2668 |
+
"grad_norm": 2.046875,
|
| 2669 |
+
"learning_rate": 7.30632548688595e-06,
|
| 2670 |
+
"loss": 0.06270002,
|
| 2671 |
+
"memory(GiB)": 75.24,
|
| 2672 |
+
"step": 1475,
|
| 2673 |
+
"train_speed(iter/s)": 0.018454
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 0.34867883410514844,
|
| 2677 |
+
"grad_norm": 2.03125,
|
| 2678 |
+
"learning_rate": 7.289886098900112e-06,
|
| 2679 |
+
"loss": 0.04866944,
|
| 2680 |
+
"memory(GiB)": 75.24,
|
| 2681 |
+
"step": 1480,
|
| 2682 |
+
"train_speed(iter/s)": 0.018455
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 0.34985680313928746,
|
| 2686 |
+
"grad_norm": 1.7421875,
|
| 2687 |
+
"learning_rate": 7.273415327002431e-06,
|
| 2688 |
+
"loss": 0.04686972,
|
| 2689 |
+
"memory(GiB)": 75.24,
|
| 2690 |
+
"step": 1485,
|
| 2691 |
+
"train_speed(iter/s)": 0.018455
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.3510347721734265,
|
| 2695 |
+
"grad_norm": 1.6484375,
|
| 2696 |
+
"learning_rate": 7.256913396932196e-06,
|
| 2697 |
+
"loss": 0.05006968,
|
| 2698 |
+
"memory(GiB)": 75.24,
|
| 2699 |
+
"step": 1490,
|
| 2700 |
+
"train_speed(iter/s)": 0.018456
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.3522127412075655,
|
| 2704 |
+
"grad_norm": 1.9921875,
|
| 2705 |
+
"learning_rate": 7.240380534855729e-06,
|
| 2706 |
+
"loss": 0.0498129,
|
| 2707 |
+
"memory(GiB)": 75.24,
|
| 2708 |
+
"step": 1495,
|
| 2709 |
+
"train_speed(iter/s)": 0.018456
|
| 2710 |
+
},
|
| 2711 |
+
{
|
| 2712 |
+
"epoch": 0.35339071024170454,
|
| 2713 |
+
"grad_norm": 1.953125,
|
| 2714 |
+
"learning_rate": 7.223816967363289e-06,
|
| 2715 |
+
"loss": 0.0470607,
|
| 2716 |
+
"memory(GiB)": 75.24,
|
| 2717 |
+
"step": 1500,
|
| 2718 |
+
"train_speed(iter/s)": 0.018456
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.3545686792758435,
|
| 2722 |
+
"grad_norm": 1.9453125,
|
| 2723 |
+
"learning_rate": 7.207222921465966e-06,
|
| 2724 |
+
"loss": 0.04950109,
|
| 2725 |
+
"memory(GiB)": 75.24,
|
| 2726 |
+
"step": 1505,
|
| 2727 |
+
"train_speed(iter/s)": 0.018448
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 0.35574664830998254,
|
| 2731 |
+
"grad_norm": 1.703125,
|
| 2732 |
+
"learning_rate": 7.190598624592575e-06,
|
| 2733 |
+
"loss": 0.04311339,
|
| 2734 |
+
"memory(GiB)": 75.24,
|
| 2735 |
+
"step": 1510,
|
| 2736 |
+
"train_speed(iter/s)": 0.018448
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 0.35692461734412156,
|
| 2740 |
+
"grad_norm": 1.84375,
|
| 2741 |
+
"learning_rate": 7.173944304586529e-06,
|
| 2742 |
+
"loss": 0.04974242,
|
| 2743 |
+
"memory(GiB)": 75.24,
|
| 2744 |
+
"step": 1515,
|
| 2745 |
+
"train_speed(iter/s)": 0.018448
|
| 2746 |
+
},
|
| 2747 |
+
{
|
| 2748 |
+
"epoch": 0.3581025863782606,
|
| 2749 |
+
"grad_norm": 1.484375,
|
| 2750 |
+
"learning_rate": 7.15726018970272e-06,
|
| 2751 |
+
"loss": 0.04702036,
|
| 2752 |
+
"memory(GiB)": 75.24,
|
| 2753 |
+
"step": 1520,
|
| 2754 |
+
"train_speed(iter/s)": 0.018448
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.3592805554123996,
|
| 2758 |
+
"grad_norm": 2.15625,
|
| 2759 |
+
"learning_rate": 7.140546508604399e-06,
|
| 2760 |
+
"loss": 0.0517059,
|
| 2761 |
+
"memory(GiB)": 75.24,
|
| 2762 |
+
"step": 1525,
|
| 2763 |
+
"train_speed(iter/s)": 0.018448
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 0.3604585244465386,
|
| 2767 |
+
"grad_norm": 2.390625,
|
| 2768 |
+
"learning_rate": 7.12380349036003e-06,
|
| 2769 |
+
"loss": 0.04891534,
|
| 2770 |
+
"memory(GiB)": 75.24,
|
| 2771 |
+
"step": 1530,
|
| 2772 |
+
"train_speed(iter/s)": 0.018449
|
| 2773 |
+
},
|
| 2774 |
+
{
|
| 2775 |
+
"epoch": 0.3616364934806776,
|
| 2776 |
+
"grad_norm": 2.703125,
|
| 2777 |
+
"learning_rate": 7.107031364440157e-06,
|
| 2778 |
+
"loss": 0.05355787,
|
| 2779 |
+
"memory(GiB)": 75.24,
|
| 2780 |
+
"step": 1535,
|
| 2781 |
+
"train_speed(iter/s)": 0.018449
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.36281446251481664,
|
| 2785 |
+
"grad_norm": 2.140625,
|
| 2786 |
+
"learning_rate": 7.090230360714254e-06,
|
| 2787 |
+
"loss": 0.05681676,
|
| 2788 |
+
"memory(GiB)": 75.24,
|
| 2789 |
+
"step": 1540,
|
| 2790 |
+
"train_speed(iter/s)": 0.01845
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.36399243154895566,
|
| 2794 |
+
"grad_norm": 2.359375,
|
| 2795 |
+
"learning_rate": 7.073400709447587e-06,
|
| 2796 |
+
"loss": 0.05522673,
|
| 2797 |
+
"memory(GiB)": 75.24,
|
| 2798 |
+
"step": 1545,
|
| 2799 |
+
"train_speed(iter/s)": 0.01845
|
| 2800 |
+
},
|
| 2801 |
+
{
|
| 2802 |
+
"epoch": 0.3651704005830947,
|
| 2803 |
+
"grad_norm": 2.015625,
|
| 2804 |
+
"learning_rate": 7.056542641298042e-06,
|
| 2805 |
+
"loss": 0.04967948,
|
| 2806 |
+
"memory(GiB)": 75.24,
|
| 2807 |
+
"step": 1550,
|
| 2808 |
+
"train_speed(iter/s)": 0.018451
|
| 2809 |
+
},
|
| 2810 |
+
{
|
| 2811 |
+
"epoch": 0.36634836961723366,
|
| 2812 |
+
"grad_norm": 2.25,
|
| 2813 |
+
"learning_rate": 7.039656387312973e-06,
|
| 2814 |
+
"loss": 0.05282811,
|
| 2815 |
+
"memory(GiB)": 75.24,
|
| 2816 |
+
"step": 1555,
|
| 2817 |
+
"train_speed(iter/s)": 0.018452
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.3675263386513727,
|
| 2821 |
+
"grad_norm": 1.859375,
|
| 2822 |
+
"learning_rate": 7.022742178926034e-06,
|
| 2823 |
+
"loss": 0.04866407,
|
| 2824 |
+
"memory(GiB)": 75.24,
|
| 2825 |
+
"step": 1560,
|
| 2826 |
+
"train_speed(iter/s)": 0.018452
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 0.3687043076855117,
|
| 2830 |
+
"grad_norm": 2.359375,
|
| 2831 |
+
"learning_rate": 7.005800247954005e-06,
|
| 2832 |
+
"loss": 0.05169091,
|
| 2833 |
+
"memory(GiB)": 75.24,
|
| 2834 |
+
"step": 1565,
|
| 2835 |
+
"train_speed(iter/s)": 0.018452
|
| 2836 |
+
},
|
| 2837 |
+
{
|
| 2838 |
+
"epoch": 0.36988227671965074,
|
| 2839 |
+
"grad_norm": 1.953125,
|
| 2840 |
+
"learning_rate": 6.98883082659362e-06,
|
| 2841 |
+
"loss": 0.05216286,
|
| 2842 |
+
"memory(GiB)": 75.24,
|
| 2843 |
+
"step": 1570,
|
| 2844 |
+
"train_speed(iter/s)": 0.018453
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.37106024575378976,
|
| 2848 |
+
"grad_norm": 1.7734375,
|
| 2849 |
+
"learning_rate": 6.971834147418377e-06,
|
| 2850 |
+
"loss": 0.05442037,
|
| 2851 |
+
"memory(GiB)": 75.24,
|
| 2852 |
+
"step": 1575,
|
| 2853 |
+
"train_speed(iter/s)": 0.018454
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.3722382147879288,
|
| 2857 |
+
"grad_norm": 1.7421875,
|
| 2858 |
+
"learning_rate": 6.954810443375356e-06,
|
| 2859 |
+
"loss": 0.05119426,
|
| 2860 |
+
"memory(GiB)": 75.24,
|
| 2861 |
+
"step": 1580,
|
| 2862 |
+
"train_speed(iter/s)": 0.018454
|
| 2863 |
+
},
|
| 2864 |
+
{
|
| 2865 |
+
"epoch": 0.37341618382206776,
|
| 2866 |
+
"grad_norm": 1.59375,
|
| 2867 |
+
"learning_rate": 6.937759947782027e-06,
|
| 2868 |
+
"loss": 0.05131354,
|
| 2869 |
+
"memory(GiB)": 75.24,
|
| 2870 |
+
"step": 1585,
|
| 2871 |
+
"train_speed(iter/s)": 0.018455
|
| 2872 |
+
},
|
| 2873 |
+
{
|
| 2874 |
+
"epoch": 0.3745941528562068,
|
| 2875 |
+
"grad_norm": 1.640625,
|
| 2876 |
+
"learning_rate": 6.920682894323046e-06,
|
| 2877 |
+
"loss": 0.05248099,
|
| 2878 |
+
"memory(GiB)": 75.24,
|
| 2879 |
+
"step": 1590,
|
| 2880 |
+
"train_speed(iter/s)": 0.018456
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.3757721218903458,
|
| 2884 |
+
"grad_norm": 1.7109375,
|
| 2885 |
+
"learning_rate": 6.903579517047061e-06,
|
| 2886 |
+
"loss": 0.04946129,
|
| 2887 |
+
"memory(GiB)": 75.24,
|
| 2888 |
+
"step": 1595,
|
| 2889 |
+
"train_speed(iter/s)": 0.018456
|
| 2890 |
+
},
|
| 2891 |
+
{
|
| 2892 |
+
"epoch": 0.37695009092448484,
|
| 2893 |
+
"grad_norm": 1.796875,
|
| 2894 |
+
"learning_rate": 6.886450050363496e-06,
|
| 2895 |
+
"loss": 0.05146198,
|
| 2896 |
+
"memory(GiB)": 75.24,
|
| 2897 |
+
"step": 1600,
|
| 2898 |
+
"train_speed(iter/s)": 0.018457
|
| 2899 |
+
},
|
| 2900 |
+
{
|
| 2901 |
+
"epoch": 0.37812805995862386,
|
| 2902 |
+
"grad_norm": 2.171875,
|
| 2903 |
+
"learning_rate": 6.869294729039344e-06,
|
| 2904 |
+
"loss": 0.05650499,
|
| 2905 |
+
"memory(GiB)": 75.24,
|
| 2906 |
+
"step": 1605,
|
| 2907 |
+
"train_speed(iter/s)": 0.018447
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 0.37930602899276283,
|
| 2911 |
+
"grad_norm": 1.578125,
|
| 2912 |
+
"learning_rate": 6.8521137881959454e-06,
|
| 2913 |
+
"loss": 0.05240663,
|
| 2914 |
+
"memory(GiB)": 75.24,
|
| 2915 |
+
"step": 1610,
|
| 2916 |
+
"train_speed(iter/s)": 0.018448
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.38048399802690186,
|
| 2920 |
+
"grad_norm": 1.7421875,
|
| 2921 |
+
"learning_rate": 6.834907463305771e-06,
|
| 2922 |
+
"loss": 0.0468967,
|
| 2923 |
+
"memory(GiB)": 75.24,
|
| 2924 |
+
"step": 1615,
|
| 2925 |
+
"train_speed(iter/s)": 0.018448
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 0.3816619670610409,
|
| 2929 |
+
"grad_norm": 1.8671875,
|
| 2930 |
+
"learning_rate": 6.817675990189188e-06,
|
| 2931 |
+
"loss": 0.04780641,
|
| 2932 |
+
"memory(GiB)": 75.24,
|
| 2933 |
+
"step": 1620,
|
| 2934 |
+
"train_speed(iter/s)": 0.018449
|
| 2935 |
+
},
|
| 2936 |
+
{
|
| 2937 |
+
"epoch": 0.3828399360951799,
|
| 2938 |
+
"grad_norm": 1.625,
|
| 2939 |
+
"learning_rate": 6.800419605011232e-06,
|
| 2940 |
+
"loss": 0.05245537,
|
| 2941 |
+
"memory(GiB)": 75.24,
|
| 2942 |
+
"step": 1625,
|
| 2943 |
+
"train_speed(iter/s)": 0.01845
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 0.38401790512931894,
|
| 2947 |
+
"grad_norm": 2.53125,
|
| 2948 |
+
"learning_rate": 6.78313854427837e-06,
|
| 2949 |
+
"loss": 0.04904989,
|
| 2950 |
+
"memory(GiB)": 75.24,
|
| 2951 |
+
"step": 1630,
|
| 2952 |
+
"train_speed(iter/s)": 0.01845
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 0.3851958741634579,
|
| 2956 |
+
"grad_norm": 1.7890625,
|
| 2957 |
+
"learning_rate": 6.76583304483526e-06,
|
| 2958 |
+
"loss": 0.05040743,
|
| 2959 |
+
"memory(GiB)": 75.24,
|
| 2960 |
+
"step": 1635,
|
| 2961 |
+
"train_speed(iter/s)": 0.018451
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 0.38637384319759693,
|
| 2965 |
+
"grad_norm": 2.125,
|
| 2966 |
+
"learning_rate": 6.7485033438615e-06,
|
| 2967 |
+
"loss": 0.04769071,
|
| 2968 |
+
"memory(GiB)": 75.24,
|
| 2969 |
+
"step": 1640,
|
| 2970 |
+
"train_speed(iter/s)": 0.018451
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 0.38755181223173596,
|
| 2974 |
+
"grad_norm": 2.34375,
|
| 2975 |
+
"learning_rate": 6.731149678868385e-06,
|
| 2976 |
+
"loss": 0.0504717,
|
| 2977 |
+
"memory(GiB)": 75.24,
|
| 2978 |
+
"step": 1645,
|
| 2979 |
+
"train_speed(iter/s)": 0.018451
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 0.388729781265875,
|
| 2983 |
+
"grad_norm": 1.7578125,
|
| 2984 |
+
"learning_rate": 6.7137722876956435e-06,
|
| 2985 |
+
"loss": 0.04594471,
|
| 2986 |
+
"memory(GiB)": 75.24,
|
| 2987 |
+
"step": 1650,
|
| 2988 |
+
"train_speed(iter/s)": 0.018452
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.389907750300014,
|
| 2992 |
+
"grad_norm": 2.015625,
|
| 2993 |
+
"learning_rate": 6.696371408508185e-06,
|
| 2994 |
+
"loss": 0.04914819,
|
| 2995 |
+
"memory(GiB)": 75.24,
|
| 2996 |
+
"step": 1655,
|
| 2997 |
+
"train_speed(iter/s)": 0.018453
|
| 2998 |
+
},
|
| 2999 |
+
{
|
| 3000 |
+
"epoch": 0.391085719334153,
|
| 3001 |
+
"grad_norm": 2.21875,
|
| 3002 |
+
"learning_rate": 6.67894727979283e-06,
|
| 3003 |
+
"loss": 0.05229876,
|
| 3004 |
+
"memory(GiB)": 75.24,
|
| 3005 |
+
"step": 1660,
|
| 3006 |
+
"train_speed(iter/s)": 0.018453
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 0.392263688368292,
|
| 3010 |
+
"grad_norm": 1.6015625,
|
| 3011 |
+
"learning_rate": 6.661500140355046e-06,
|
| 3012 |
+
"loss": 0.05038055,
|
| 3013 |
+
"memory(GiB)": 75.24,
|
| 3014 |
+
"step": 1665,
|
| 3015 |
+
"train_speed(iter/s)": 0.018453
|
| 3016 |
+
},
|
| 3017 |
+
{
|
| 3018 |
+
"epoch": 0.39344165740243103,
|
| 3019 |
+
"grad_norm": 1.6484375,
|
| 3020 |
+
"learning_rate": 6.644030229315674e-06,
|
| 3021 |
+
"loss": 0.04761205,
|
| 3022 |
+
"memory(GiB)": 75.24,
|
| 3023 |
+
"step": 1670,
|
| 3024 |
+
"train_speed(iter/s)": 0.018454
|
| 3025 |
+
},
|
| 3026 |
+
{
|
| 3027 |
+
"epoch": 0.39461962643657006,
|
| 3028 |
+
"grad_norm": 1.96875,
|
| 3029 |
+
"learning_rate": 6.626537786107647e-06,
|
| 3030 |
+
"loss": 0.05017205,
|
| 3031 |
+
"memory(GiB)": 75.24,
|
| 3032 |
+
"step": 1675,
|
| 3033 |
+
"train_speed(iter/s)": 0.018455
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.3957975954707091,
|
| 3037 |
+
"grad_norm": 2.03125,
|
| 3038 |
+
"learning_rate": 6.609023050472712e-06,
|
| 3039 |
+
"loss": 0.0522933,
|
| 3040 |
+
"memory(GiB)": 75.24,
|
| 3041 |
+
"step": 1680,
|
| 3042 |
+
"train_speed(iter/s)": 0.018456
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 0.39697556450484806,
|
| 3046 |
+
"grad_norm": 2.015625,
|
| 3047 |
+
"learning_rate": 6.591486262458146e-06,
|
| 3048 |
+
"loss": 0.05544751,
|
| 3049 |
+
"memory(GiB)": 75.24,
|
| 3050 |
+
"step": 1685,
|
| 3051 |
+
"train_speed(iter/s)": 0.018456
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 0.3981535335389871,
|
| 3055 |
+
"grad_norm": 2.5625,
|
| 3056 |
+
"learning_rate": 6.573927662413462e-06,
|
| 3057 |
+
"loss": 0.05269616,
|
| 3058 |
+
"memory(GiB)": 75.24,
|
| 3059 |
+
"step": 1690,
|
| 3060 |
+
"train_speed(iter/s)": 0.018457
|
| 3061 |
+
},
|
| 3062 |
+
{
|
| 3063 |
+
"epoch": 0.3993315025731261,
|
| 3064 |
+
"grad_norm": 1.6796875,
|
| 3065 |
+
"learning_rate": 6.556347490987115e-06,
|
| 3066 |
+
"loss": 0.0480206,
|
| 3067 |
+
"memory(GiB)": 75.24,
|
| 3068 |
+
"step": 1695,
|
| 3069 |
+
"train_speed(iter/s)": 0.018457
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 0.40050947160726513,
|
| 3073 |
+
"grad_norm": 1.71875,
|
| 3074 |
+
"learning_rate": 6.53874598912321e-06,
|
| 3075 |
+
"loss": 0.04850869,
|
| 3076 |
+
"memory(GiB)": 75.24,
|
| 3077 |
+
"step": 1700,
|
| 3078 |
+
"train_speed(iter/s)": 0.018458
|
| 3079 |
+
},
|
| 3080 |
+
{
|
| 3081 |
+
"epoch": 0.40168744064140416,
|
| 3082 |
+
"grad_norm": 2.109375,
|
| 3083 |
+
"learning_rate": 6.521123398058189e-06,
|
| 3084 |
+
"loss": 0.04664425,
|
| 3085 |
+
"memory(GiB)": 75.24,
|
| 3086 |
+
"step": 1705,
|
| 3087 |
+
"train_speed(iter/s)": 0.018449
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 0.40286540967554313,
|
| 3091 |
+
"grad_norm": 1.640625,
|
| 3092 |
+
"learning_rate": 6.503479959317536e-06,
|
| 3093 |
+
"loss": 0.05617427,
|
| 3094 |
+
"memory(GiB)": 75.24,
|
| 3095 |
+
"step": 1710,
|
| 3096 |
+
"train_speed(iter/s)": 0.01845
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.40404337870968215,
|
| 3100 |
+
"grad_norm": 1.421875,
|
| 3101 |
+
"learning_rate": 6.485815914712461e-06,
|
| 3102 |
+
"loss": 0.04710643,
|
| 3103 |
+
"memory(GiB)": 75.24,
|
| 3104 |
+
"step": 1715,
|
| 3105 |
+
"train_speed(iter/s)": 0.01845
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.4052213477438212,
|
| 3109 |
+
"grad_norm": 2.0,
|
| 3110 |
+
"learning_rate": 6.468131506336584e-06,
|
| 3111 |
+
"loss": 0.05491562,
|
| 3112 |
+
"memory(GiB)": 75.24,
|
| 3113 |
+
"step": 1720,
|
| 3114 |
+
"train_speed(iter/s)": 0.01845
|
| 3115 |
+
},
|
| 3116 |
+
{
|
| 3117 |
+
"epoch": 0.4063993167779602,
|
| 3118 |
+
"grad_norm": 1.7578125,
|
| 3119 |
+
"learning_rate": 6.450426976562623e-06,
|
| 3120 |
+
"loss": 0.04522764,
|
| 3121 |
+
"memory(GiB)": 75.24,
|
| 3122 |
+
"step": 1725,
|
| 3123 |
+
"train_speed(iter/s)": 0.018451
|
| 3124 |
+
},
|
| 3125 |
+
{
|
| 3126 |
+
"epoch": 0.40757728581209923,
|
| 3127 |
+
"grad_norm": 1.8671875,
|
| 3128 |
+
"learning_rate": 6.4327025680390646e-06,
|
| 3129 |
+
"loss": 0.0509581,
|
| 3130 |
+
"memory(GiB)": 75.24,
|
| 3131 |
+
"step": 1730,
|
| 3132 |
+
"train_speed(iter/s)": 0.018451
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.4087552548462382,
|
| 3136 |
+
"grad_norm": 2.5625,
|
| 3137 |
+
"learning_rate": 6.414958523686845e-06,
|
| 3138 |
+
"loss": 0.05455146,
|
| 3139 |
+
"memory(GiB)": 75.24,
|
| 3140 |
+
"step": 1735,
|
| 3141 |
+
"train_speed(iter/s)": 0.018452
|
| 3142 |
+
},
|
| 3143 |
+
{
|
| 3144 |
+
"epoch": 0.40993322388037723,
|
| 3145 |
+
"grad_norm": 1.609375,
|
| 3146 |
+
"learning_rate": 6.397195086696021e-06,
|
| 3147 |
+
"loss": 0.04453045,
|
| 3148 |
+
"memory(GiB)": 75.24,
|
| 3149 |
+
"step": 1740,
|
| 3150 |
+
"train_speed(iter/s)": 0.018452
|
| 3151 |
+
},
|
| 3152 |
+
{
|
| 3153 |
+
"epoch": 0.41111119291451625,
|
| 3154 |
+
"grad_norm": 2.984375,
|
| 3155 |
+
"learning_rate": 6.379412500522427e-06,
|
| 3156 |
+
"loss": 0.04473322,
|
| 3157 |
+
"memory(GiB)": 75.24,
|
| 3158 |
+
"step": 1745,
|
| 3159 |
+
"train_speed(iter/s)": 0.018452
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.4122891619486553,
|
| 3163 |
+
"grad_norm": 1.609375,
|
| 3164 |
+
"learning_rate": 6.361611008884353e-06,
|
| 3165 |
+
"loss": 0.05007306,
|
| 3166 |
+
"memory(GiB)": 75.24,
|
| 3167 |
+
"step": 1750,
|
| 3168 |
+
"train_speed(iter/s)": 0.018452
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 0.4134671309827943,
|
| 3172 |
+
"grad_norm": 1.984375,
|
| 3173 |
+
"learning_rate": 6.343790855759191e-06,
|
| 3174 |
+
"loss": 0.04592868,
|
| 3175 |
+
"memory(GiB)": 75.24,
|
| 3176 |
+
"step": 1755,
|
| 3177 |
+
"train_speed(iter/s)": 0.018452
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 0.41464510001693333,
|
| 3181 |
+
"grad_norm": 1.9140625,
|
| 3182 |
+
"learning_rate": 6.3259522853801015e-06,
|
| 3183 |
+
"loss": 0.04621382,
|
| 3184 |
+
"memory(GiB)": 75.24,
|
| 3185 |
+
"step": 1760,
|
| 3186 |
+
"train_speed(iter/s)": 0.018453
|
| 3187 |
+
},
|
| 3188 |
+
{
|
| 3189 |
+
"epoch": 0.4158230690510723,
|
| 3190 |
+
"grad_norm": 1.8671875,
|
| 3191 |
+
"learning_rate": 6.3080955422326595e-06,
|
| 3192 |
+
"loss": 0.04989561,
|
| 3193 |
+
"memory(GiB)": 75.24,
|
| 3194 |
+
"step": 1765,
|
| 3195 |
+
"train_speed(iter/s)": 0.018453
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 0.41700103808521133,
|
| 3199 |
+
"grad_norm": 1.5390625,
|
| 3200 |
+
"learning_rate": 6.290220871051506e-06,
|
| 3201 |
+
"loss": 0.05395367,
|
| 3202 |
+
"memory(GiB)": 75.24,
|
| 3203 |
+
"step": 1770,
|
| 3204 |
+
"train_speed(iter/s)": 0.018454
|
| 3205 |
+
},
|
| 3206 |
+
{
|
| 3207 |
+
"epoch": 0.41817900711935035,
|
| 3208 |
+
"grad_norm": 1.8671875,
|
| 3209 |
+
"learning_rate": 6.272328516816993e-06,
|
| 3210 |
+
"loss": 0.04335098,
|
| 3211 |
+
"memory(GiB)": 75.24,
|
| 3212 |
+
"step": 1775,
|
| 3213 |
+
"train_speed(iter/s)": 0.018454
|
| 3214 |
+
},
|
| 3215 |
+
{
|
| 3216 |
+
"epoch": 0.4193569761534894,
|
| 3217 |
+
"grad_norm": 1.671875,
|
| 3218 |
+
"learning_rate": 6.2544187247518285e-06,
|
| 3219 |
+
"loss": 0.05292164,
|
| 3220 |
+
"memory(GiB)": 75.24,
|
| 3221 |
+
"step": 1780,
|
| 3222 |
+
"train_speed(iter/s)": 0.018455
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.4205349451876284,
|
| 3226 |
+
"grad_norm": 1.609375,
|
| 3227 |
+
"learning_rate": 6.2364917403177115e-06,
|
| 3228 |
+
"loss": 0.04423888,
|
| 3229 |
+
"memory(GiB)": 75.24,
|
| 3230 |
+
"step": 1785,
|
| 3231 |
+
"train_speed(iter/s)": 0.018456
|
| 3232 |
+
},
|
| 3233 |
+
{
|
| 3234 |
+
"epoch": 0.4217129142217674,
|
| 3235 |
+
"grad_norm": 2.140625,
|
| 3236 |
+
"learning_rate": 6.218547809211973e-06,
|
| 3237 |
+
"loss": 0.04801815,
|
| 3238 |
+
"memory(GiB)": 75.24,
|
| 3239 |
+
"step": 1790,
|
| 3240 |
+
"train_speed(iter/s)": 0.018456
|
| 3241 |
+
},
|
| 3242 |
+
{
|
| 3243 |
+
"epoch": 0.4228908832559064,
|
| 3244 |
+
"grad_norm": 2.15625,
|
| 3245 |
+
"learning_rate": 6.200587177364204e-06,
|
| 3246 |
+
"loss": 0.0498428,
|
| 3247 |
+
"memory(GiB)": 75.24,
|
| 3248 |
+
"step": 1795,
|
| 3249 |
+
"train_speed(iter/s)": 0.018457
|
| 3250 |
+
},
|
| 3251 |
+
{
|
| 3252 |
+
"epoch": 0.42406885229004543,
|
| 3253 |
+
"grad_norm": 2.421875,
|
| 3254 |
+
"learning_rate": 6.1826100909328855e-06,
|
| 3255 |
+
"loss": 0.05989224,
|
| 3256 |
+
"memory(GiB)": 75.24,
|
| 3257 |
+
"step": 1800,
|
| 3258 |
+
"train_speed(iter/s)": 0.018457
|
| 3259 |
}
|
| 3260 |
],
|
| 3261 |
"logging_steps": 5,
|
|
|
|
| 3275 |
"attributes": {}
|
| 3276 |
}
|
| 3277 |
},
|
| 3278 |
+
"total_flos": 1.9948649883254129e+18,
|
| 3279 |
"train_batch_size": 4,
|
| 3280 |
"trial_name": null,
|
| 3281 |
"trial_params": null
|