Training in progress, step 7000, checkpoint
Browse files
last-checkpoint/config.json
CHANGED
|
@@ -26,7 +26,7 @@
|
|
| 26 |
"rope_type": "default"
|
| 27 |
},
|
| 28 |
"tie_word_embeddings": false,
|
| 29 |
-
"transformers_version": "5.8.
|
| 30 |
"use_cache": false,
|
| 31 |
"vocab_size": 50257
|
| 32 |
}
|
|
|
|
| 26 |
"rope_type": "default"
|
| 27 |
},
|
| 28 |
"tie_word_embeddings": false,
|
| 29 |
+
"transformers_version": "5.8.1",
|
| 30 |
"use_cache": false,
|
| 31 |
"vocab_size": 50257
|
| 32 |
}
|
last-checkpoint/generation_config.json
CHANGED
|
@@ -5,6 +5,6 @@
|
|
| 5 |
"output_attentions": false,
|
| 6 |
"output_hidden_states": false,
|
| 7 |
"pad_token_id": 0,
|
| 8 |
-
"transformers_version": "5.8.
|
| 9 |
"use_cache": true
|
| 10 |
}
|
|
|
|
| 5 |
"output_attentions": false,
|
| 6 |
"output_hidden_states": false,
|
| 7 |
"pad_token_id": 0,
|
| 8 |
+
"transformers_version": "5.8.1",
|
| 9 |
"use_cache": true
|
| 10 |
}
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4523108832
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5357a9d40da12fb2def448cb62f5c3aed9fa0739e8f2c2c0ec7fec7354241c0
|
| 3 |
size 4523108832
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2912179275
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:107bb493c2caa0cd6fa40a129559390214dfd1f7e807e09d56c5d2aebb1c47e9
|
| 3 |
size 2912179275
|
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 sha256:6bbe17a5fe328cdb084ec10afd49d2cb7bb9035b6736206385c2c19bf9f66c41
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1831663cfc396872c81aa655b9b01b2d24ce27a5ca47e7cf38d581dbb6adb922
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86a40e24c6251fa4890a76644fa4c014362ca533cd573d7139b9e226b18d8d04
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -4558,6 +4558,356 @@
|
|
| 4558 |
"learning_rate": 7.004e-05,
|
| 4559 |
"loss": 0.030304345488548278,
|
| 4560 |
"step": 6500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4561 |
}
|
| 4562 |
],
|
| 4563 |
"logging_steps": 10,
|
|
@@ -4577,7 +4927,7 @@
|
|
| 4577 |
"attributes": {}
|
| 4578 |
}
|
| 4579 |
},
|
| 4580 |
-
"total_flos": 5.
|
| 4581 |
"train_batch_size": 1,
|
| 4582 |
"trial_name": null,
|
| 4583 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0188,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 7000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 4558 |
"learning_rate": 7.004e-05,
|
| 4559 |
"loss": 0.030304345488548278,
|
| 4560 |
"step": 6500
|
| 4561 |
+
},
|
| 4562 |
+
{
|
| 4563 |
+
"epoch": 1.001,
|
| 4564 |
+
"grad_norm": 0.17337968945503235,
|
| 4565 |
+
"learning_rate": 6.984000000000001e-05,
|
| 4566 |
+
"loss": 0.04341032803058624,
|
| 4567 |
+
"step": 6510
|
| 4568 |
+
},
|
| 4569 |
+
{
|
| 4570 |
+
"epoch": 1.002,
|
| 4571 |
+
"grad_norm": 0.14447154104709625,
|
| 4572 |
+
"learning_rate": 6.964e-05,
|
| 4573 |
+
"loss": 0.02765703499317169,
|
| 4574 |
+
"step": 6520
|
| 4575 |
+
},
|
| 4576 |
+
{
|
| 4577 |
+
"epoch": 1.003,
|
| 4578 |
+
"grad_norm": 0.1359182447195053,
|
| 4579 |
+
"learning_rate": 6.944e-05,
|
| 4580 |
+
"loss": 0.027671322226524353,
|
| 4581 |
+
"step": 6530
|
| 4582 |
+
},
|
| 4583 |
+
{
|
| 4584 |
+
"epoch": 1.004,
|
| 4585 |
+
"grad_norm": 0.1486900895833969,
|
| 4586 |
+
"learning_rate": 6.924000000000001e-05,
|
| 4587 |
+
"loss": 0.024222801625728606,
|
| 4588 |
+
"step": 6540
|
| 4589 |
+
},
|
| 4590 |
+
{
|
| 4591 |
+
"epoch": 1.005,
|
| 4592 |
+
"grad_norm": 0.17833292484283447,
|
| 4593 |
+
"learning_rate": 6.904e-05,
|
| 4594 |
+
"loss": 0.024685877561569213,
|
| 4595 |
+
"step": 6550
|
| 4596 |
+
},
|
| 4597 |
+
{
|
| 4598 |
+
"epoch": 1.006,
|
| 4599 |
+
"grad_norm": 0.1444295197725296,
|
| 4600 |
+
"learning_rate": 6.884e-05,
|
| 4601 |
+
"loss": 0.027162906527519227,
|
| 4602 |
+
"step": 6560
|
| 4603 |
+
},
|
| 4604 |
+
{
|
| 4605 |
+
"epoch": 1.007,
|
| 4606 |
+
"grad_norm": 0.18214137852191925,
|
| 4607 |
+
"learning_rate": 6.864000000000001e-05,
|
| 4608 |
+
"loss": 0.02484011501073837,
|
| 4609 |
+
"step": 6570
|
| 4610 |
+
},
|
| 4611 |
+
{
|
| 4612 |
+
"epoch": 1.008,
|
| 4613 |
+
"grad_norm": 0.16924667358398438,
|
| 4614 |
+
"learning_rate": 6.844e-05,
|
| 4615 |
+
"loss": 0.025604432821273802,
|
| 4616 |
+
"step": 6580
|
| 4617 |
+
},
|
| 4618 |
+
{
|
| 4619 |
+
"epoch": 1.009,
|
| 4620 |
+
"grad_norm": 0.16066834330558777,
|
| 4621 |
+
"learning_rate": 6.824e-05,
|
| 4622 |
+
"loss": 0.025216665863990784,
|
| 4623 |
+
"step": 6590
|
| 4624 |
+
},
|
| 4625 |
+
{
|
| 4626 |
+
"epoch": 1.01,
|
| 4627 |
+
"grad_norm": 0.14215150475502014,
|
| 4628 |
+
"learning_rate": 6.804e-05,
|
| 4629 |
+
"loss": 0.024471497535705565,
|
| 4630 |
+
"step": 6600
|
| 4631 |
+
},
|
| 4632 |
+
{
|
| 4633 |
+
"epoch": 1.011,
|
| 4634 |
+
"grad_norm": 0.128558948636055,
|
| 4635 |
+
"learning_rate": 6.784e-05,
|
| 4636 |
+
"loss": 0.02575332820415497,
|
| 4637 |
+
"step": 6610
|
| 4638 |
+
},
|
| 4639 |
+
{
|
| 4640 |
+
"epoch": 1.012,
|
| 4641 |
+
"grad_norm": 0.33806535601615906,
|
| 4642 |
+
"learning_rate": 6.764000000000001e-05,
|
| 4643 |
+
"loss": 0.025676625967025756,
|
| 4644 |
+
"step": 6620
|
| 4645 |
+
},
|
| 4646 |
+
{
|
| 4647 |
+
"epoch": 1.013,
|
| 4648 |
+
"grad_norm": 0.16000686585903168,
|
| 4649 |
+
"learning_rate": 6.744e-05,
|
| 4650 |
+
"loss": 0.025471991300582884,
|
| 4651 |
+
"step": 6630
|
| 4652 |
+
},
|
| 4653 |
+
{
|
| 4654 |
+
"epoch": 1.014,
|
| 4655 |
+
"grad_norm": 0.19417209923267365,
|
| 4656 |
+
"learning_rate": 6.724e-05,
|
| 4657 |
+
"loss": 0.0291011780500412,
|
| 4658 |
+
"step": 6640
|
| 4659 |
+
},
|
| 4660 |
+
{
|
| 4661 |
+
"epoch": 1.015,
|
| 4662 |
+
"grad_norm": 0.2124422937631607,
|
| 4663 |
+
"learning_rate": 6.704000000000001e-05,
|
| 4664 |
+
"loss": 0.02770865261554718,
|
| 4665 |
+
"step": 6650
|
| 4666 |
+
},
|
| 4667 |
+
{
|
| 4668 |
+
"epoch": 1.016,
|
| 4669 |
+
"grad_norm": 0.16012416779994965,
|
| 4670 |
+
"learning_rate": 6.684e-05,
|
| 4671 |
+
"loss": 0.03058260679244995,
|
| 4672 |
+
"step": 6660
|
| 4673 |
+
},
|
| 4674 |
+
{
|
| 4675 |
+
"epoch": 1.017,
|
| 4676 |
+
"grad_norm": 0.1718786507844925,
|
| 4677 |
+
"learning_rate": 6.664e-05,
|
| 4678 |
+
"loss": 0.02695387601852417,
|
| 4679 |
+
"step": 6670
|
| 4680 |
+
},
|
| 4681 |
+
{
|
| 4682 |
+
"epoch": 1.018,
|
| 4683 |
+
"grad_norm": 0.18105700612068176,
|
| 4684 |
+
"learning_rate": 6.644000000000001e-05,
|
| 4685 |
+
"loss": 0.026461568474769593,
|
| 4686 |
+
"step": 6680
|
| 4687 |
+
},
|
| 4688 |
+
{
|
| 4689 |
+
"epoch": 1.019,
|
| 4690 |
+
"grad_norm": 0.15785875916481018,
|
| 4691 |
+
"learning_rate": 6.624e-05,
|
| 4692 |
+
"loss": 0.02598581314086914,
|
| 4693 |
+
"step": 6690
|
| 4694 |
+
},
|
| 4695 |
+
{
|
| 4696 |
+
"epoch": 1.02,
|
| 4697 |
+
"grad_norm": 0.15605275332927704,
|
| 4698 |
+
"learning_rate": 6.604e-05,
|
| 4699 |
+
"loss": 0.026878923177719116,
|
| 4700 |
+
"step": 6700
|
| 4701 |
+
},
|
| 4702 |
+
{
|
| 4703 |
+
"epoch": 1.021,
|
| 4704 |
+
"grad_norm": 0.1681290715932846,
|
| 4705 |
+
"learning_rate": 6.584e-05,
|
| 4706 |
+
"loss": 0.026533681154251098,
|
| 4707 |
+
"step": 6710
|
| 4708 |
+
},
|
| 4709 |
+
{
|
| 4710 |
+
"epoch": 1.022,
|
| 4711 |
+
"grad_norm": 0.17714430391788483,
|
| 4712 |
+
"learning_rate": 6.564e-05,
|
| 4713 |
+
"loss": 0.026042383909225465,
|
| 4714 |
+
"step": 6720
|
| 4715 |
+
},
|
| 4716 |
+
{
|
| 4717 |
+
"epoch": 1.023,
|
| 4718 |
+
"grad_norm": 0.17928007245063782,
|
| 4719 |
+
"learning_rate": 6.544e-05,
|
| 4720 |
+
"loss": 0.026147887110710144,
|
| 4721 |
+
"step": 6730
|
| 4722 |
+
},
|
| 4723 |
+
{
|
| 4724 |
+
"epoch": 1.024,
|
| 4725 |
+
"grad_norm": 0.21016575396060944,
|
| 4726 |
+
"learning_rate": 6.524e-05,
|
| 4727 |
+
"loss": 0.026558607816696167,
|
| 4728 |
+
"step": 6740
|
| 4729 |
+
},
|
| 4730 |
+
{
|
| 4731 |
+
"epoch": 1.025,
|
| 4732 |
+
"grad_norm": 0.19502075016498566,
|
| 4733 |
+
"learning_rate": 6.504e-05,
|
| 4734 |
+
"loss": 0.02726261019706726,
|
| 4735 |
+
"step": 6750
|
| 4736 |
+
},
|
| 4737 |
+
{
|
| 4738 |
+
"epoch": 1.026,
|
| 4739 |
+
"grad_norm": 0.1563853919506073,
|
| 4740 |
+
"learning_rate": 6.484e-05,
|
| 4741 |
+
"loss": 0.025972676277160645,
|
| 4742 |
+
"step": 6760
|
| 4743 |
+
},
|
| 4744 |
+
{
|
| 4745 |
+
"epoch": 1.027,
|
| 4746 |
+
"grad_norm": 0.15548229217529297,
|
| 4747 |
+
"learning_rate": 6.464e-05,
|
| 4748 |
+
"loss": 0.027564069628715514,
|
| 4749 |
+
"step": 6770
|
| 4750 |
+
},
|
| 4751 |
+
{
|
| 4752 |
+
"epoch": 1.028,
|
| 4753 |
+
"grad_norm": 0.18457072973251343,
|
| 4754 |
+
"learning_rate": 6.444e-05,
|
| 4755 |
+
"loss": 0.027495378255844118,
|
| 4756 |
+
"step": 6780
|
| 4757 |
+
},
|
| 4758 |
+
{
|
| 4759 |
+
"epoch": 1.029,
|
| 4760 |
+
"grad_norm": 0.15842361748218536,
|
| 4761 |
+
"learning_rate": 6.424e-05,
|
| 4762 |
+
"loss": 0.026171448826789855,
|
| 4763 |
+
"step": 6790
|
| 4764 |
+
},
|
| 4765 |
+
{
|
| 4766 |
+
"epoch": 1.03,
|
| 4767 |
+
"grad_norm": 0.19476866722106934,
|
| 4768 |
+
"learning_rate": 6.404e-05,
|
| 4769 |
+
"loss": 0.031318637728691104,
|
| 4770 |
+
"step": 6800
|
| 4771 |
+
},
|
| 4772 |
+
{
|
| 4773 |
+
"epoch": 1.031,
|
| 4774 |
+
"grad_norm": 0.1272721141576767,
|
| 4775 |
+
"learning_rate": 6.384e-05,
|
| 4776 |
+
"loss": 0.026132452487945556,
|
| 4777 |
+
"step": 6810
|
| 4778 |
+
},
|
| 4779 |
+
{
|
| 4780 |
+
"epoch": 2.0008,
|
| 4781 |
+
"grad_norm": 0.14176732301712036,
|
| 4782 |
+
"learning_rate": 6.364e-05,
|
| 4783 |
+
"loss": 0.03041217029094696,
|
| 4784 |
+
"step": 6820
|
| 4785 |
+
},
|
| 4786 |
+
{
|
| 4787 |
+
"epoch": 2.0018,
|
| 4788 |
+
"grad_norm": 0.14128795266151428,
|
| 4789 |
+
"learning_rate": 6.344e-05,
|
| 4790 |
+
"loss": 0.02396068423986435,
|
| 4791 |
+
"step": 6830
|
| 4792 |
+
},
|
| 4793 |
+
{
|
| 4794 |
+
"epoch": 2.0028,
|
| 4795 |
+
"grad_norm": 0.14129574596881866,
|
| 4796 |
+
"learning_rate": 6.324e-05,
|
| 4797 |
+
"loss": 0.02366064041852951,
|
| 4798 |
+
"step": 6840
|
| 4799 |
+
},
|
| 4800 |
+
{
|
| 4801 |
+
"epoch": 2.0038,
|
| 4802 |
+
"grad_norm": 0.1310533732175827,
|
| 4803 |
+
"learning_rate": 6.303999999999999e-05,
|
| 4804 |
+
"loss": 0.02287290096282959,
|
| 4805 |
+
"step": 6850
|
| 4806 |
+
},
|
| 4807 |
+
{
|
| 4808 |
+
"epoch": 2.0048,
|
| 4809 |
+
"grad_norm": 0.11340674012899399,
|
| 4810 |
+
"learning_rate": 6.284e-05,
|
| 4811 |
+
"loss": 0.02178637236356735,
|
| 4812 |
+
"step": 6860
|
| 4813 |
+
},
|
| 4814 |
+
{
|
| 4815 |
+
"epoch": 2.0058,
|
| 4816 |
+
"grad_norm": 0.10936598479747772,
|
| 4817 |
+
"learning_rate": 6.264e-05,
|
| 4818 |
+
"loss": 0.02486586421728134,
|
| 4819 |
+
"step": 6870
|
| 4820 |
+
},
|
| 4821 |
+
{
|
| 4822 |
+
"epoch": 2.0068,
|
| 4823 |
+
"grad_norm": 0.14033988118171692,
|
| 4824 |
+
"learning_rate": 6.244e-05,
|
| 4825 |
+
"loss": 0.022671811282634735,
|
| 4826 |
+
"step": 6880
|
| 4827 |
+
},
|
| 4828 |
+
{
|
| 4829 |
+
"epoch": 2.0078,
|
| 4830 |
+
"grad_norm": 0.1289263367652893,
|
| 4831 |
+
"learning_rate": 6.224e-05,
|
| 4832 |
+
"loss": 0.021725392341613768,
|
| 4833 |
+
"step": 6890
|
| 4834 |
+
},
|
| 4835 |
+
{
|
| 4836 |
+
"epoch": 2.0088,
|
| 4837 |
+
"grad_norm": 0.13452701270580292,
|
| 4838 |
+
"learning_rate": 6.204e-05,
|
| 4839 |
+
"loss": 0.02472696304321289,
|
| 4840 |
+
"step": 6900
|
| 4841 |
+
},
|
| 4842 |
+
{
|
| 4843 |
+
"epoch": 2.0098,
|
| 4844 |
+
"grad_norm": 0.14499792456626892,
|
| 4845 |
+
"learning_rate": 6.184e-05,
|
| 4846 |
+
"loss": 0.022322843968868255,
|
| 4847 |
+
"step": 6910
|
| 4848 |
+
},
|
| 4849 |
+
{
|
| 4850 |
+
"epoch": 2.0108,
|
| 4851 |
+
"grad_norm": 0.1398196518421173,
|
| 4852 |
+
"learning_rate": 6.164e-05,
|
| 4853 |
+
"loss": 0.022544071078300476,
|
| 4854 |
+
"step": 6920
|
| 4855 |
+
},
|
| 4856 |
+
{
|
| 4857 |
+
"epoch": 2.0118,
|
| 4858 |
+
"grad_norm": 0.1454416811466217,
|
| 4859 |
+
"learning_rate": 6.144e-05,
|
| 4860 |
+
"loss": 0.022087126970291138,
|
| 4861 |
+
"step": 6930
|
| 4862 |
+
},
|
| 4863 |
+
{
|
| 4864 |
+
"epoch": 2.0128,
|
| 4865 |
+
"grad_norm": 0.12601584196090698,
|
| 4866 |
+
"learning_rate": 6.124e-05,
|
| 4867 |
+
"loss": 0.023739957809448244,
|
| 4868 |
+
"step": 6940
|
| 4869 |
+
},
|
| 4870 |
+
{
|
| 4871 |
+
"epoch": 2.0138,
|
| 4872 |
+
"grad_norm": 0.13553965091705322,
|
| 4873 |
+
"learning_rate": 6.104000000000001e-05,
|
| 4874 |
+
"loss": 0.024189202487468718,
|
| 4875 |
+
"step": 6950
|
| 4876 |
+
},
|
| 4877 |
+
{
|
| 4878 |
+
"epoch": 2.0148,
|
| 4879 |
+
"grad_norm": 0.17853738367557526,
|
| 4880 |
+
"learning_rate": 6.084000000000001e-05,
|
| 4881 |
+
"loss": 0.024137826263904573,
|
| 4882 |
+
"step": 6960
|
| 4883 |
+
},
|
| 4884 |
+
{
|
| 4885 |
+
"epoch": 2.0158,
|
| 4886 |
+
"grad_norm": 0.1292908936738968,
|
| 4887 |
+
"learning_rate": 6.064000000000001e-05,
|
| 4888 |
+
"loss": 0.029603716731071473,
|
| 4889 |
+
"step": 6970
|
| 4890 |
+
},
|
| 4891 |
+
{
|
| 4892 |
+
"epoch": 2.0168,
|
| 4893 |
+
"grad_norm": 0.13389606773853302,
|
| 4894 |
+
"learning_rate": 6.044000000000001e-05,
|
| 4895 |
+
"loss": 0.02270825654268265,
|
| 4896 |
+
"step": 6980
|
| 4897 |
+
},
|
| 4898 |
+
{
|
| 4899 |
+
"epoch": 2.0178,
|
| 4900 |
+
"grad_norm": 0.10673966258764267,
|
| 4901 |
+
"learning_rate": 6.0240000000000006e-05,
|
| 4902 |
+
"loss": 0.0229933500289917,
|
| 4903 |
+
"step": 6990
|
| 4904 |
+
},
|
| 4905 |
+
{
|
| 4906 |
+
"epoch": 2.0188,
|
| 4907 |
+
"grad_norm": 0.13245923817157745,
|
| 4908 |
+
"learning_rate": 6.004000000000001e-05,
|
| 4909 |
+
"loss": 0.02357790470123291,
|
| 4910 |
+
"step": 7000
|
| 4911 |
}
|
| 4912 |
],
|
| 4913 |
"logging_steps": 10,
|
|
|
|
| 4927 |
"attributes": {}
|
| 4928 |
}
|
| 4929 |
},
|
| 4930 |
+
"total_flos": 5.92542131384107e+17,
|
| 4931 |
"train_batch_size": 1,
|
| 4932 |
"trial_name": null,
|
| 4933 |
"trial_params": null
|