Training in progress, step 4000
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- runs/Jun07_12-33-16_DESKTOP-69FPKCK/events.out.tfevents.1717788805.DESKTOP-69FPKCK +2 -2
- scheduler.pt +1 -1
- trainer_state.json +712 -3
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 44644496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:873ecad1ac407c43e7aa61ab18acd884cc44d434b7b2b60712e4ef84776ac1ea
|
| 3 |
size 44644496
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 11230198
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a46c2ece7ae2c9d715653461803e265a77857b8a386b5c1c56337c331e66ce7
|
| 3 |
size 11230198
|
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 sha256:9196a1e708bf24d6abba41cce3f8558820acc3e50f9394c5955e29eb41ffea3d
|
| 3 |
size 14244
|
runs/Jun07_12-33-16_DESKTOP-69FPKCK/events.out.tfevents.1717788805.DESKTOP-69FPKCK
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dcbb91f0746df5740ecb933d29845f7ab764aa5e5df8ed19411f1412a367cc36
|
| 3 |
+
size 47069
|
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 sha256:4ecff5e57c3ae91a65eb709110640874ace55a10b2c31d6b25cbf636312baf7e
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 2000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -716,6 +716,715 @@
|
|
| 716 |
"eval_samples_per_second": 2823.711,
|
| 717 |
"eval_steps_per_second": 11.033,
|
| 718 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 719 |
}
|
| 720 |
],
|
| 721 |
"logging_steps": 20,
|
|
@@ -723,7 +1432,7 @@
|
|
| 723 |
"num_input_tokens_seen": 0,
|
| 724 |
"num_train_epochs": 3,
|
| 725 |
"save_steps": 100,
|
| 726 |
-
"total_flos":
|
| 727 |
"train_batch_size": 256,
|
| 728 |
"trial_name": null,
|
| 729 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.01809799157538492,
|
| 5 |
"eval_steps": 2000,
|
| 6 |
+
"global_step": 4000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 716 |
"eval_samples_per_second": 2823.711,
|
| 717 |
"eval_steps_per_second": 11.033,
|
| 718 |
"step": 2000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.009139485745569385,
|
| 722 |
+
"grad_norm": 2.2706515789031982,
|
| 723 |
+
"learning_rate": 2.7418333182517418e-05,
|
| 724 |
+
"loss": 9.2604,
|
| 725 |
+
"step": 2020
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.00922997570344631,
|
| 729 |
+
"grad_norm": 2.297621011734009,
|
| 730 |
+
"learning_rate": 2.7689801827888878e-05,
|
| 731 |
+
"loss": 9.2367,
|
| 732 |
+
"step": 2040
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.009320465661323234,
|
| 736 |
+
"grad_norm": 2.049971342086792,
|
| 737 |
+
"learning_rate": 2.7961270473260337e-05,
|
| 738 |
+
"loss": 9.2545,
|
| 739 |
+
"step": 2060
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.009410955619200159,
|
| 743 |
+
"grad_norm": 2.3538951873779297,
|
| 744 |
+
"learning_rate": 2.82327391186318e-05,
|
| 745 |
+
"loss": 9.2511,
|
| 746 |
+
"step": 2080
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.009501445577077083,
|
| 750 |
+
"grad_norm": 3.1383931636810303,
|
| 751 |
+
"learning_rate": 2.8504207764003254e-05,
|
| 752 |
+
"loss": 9.2319,
|
| 753 |
+
"step": 2100
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.00959193553495401,
|
| 757 |
+
"grad_norm": 2.6480958461761475,
|
| 758 |
+
"learning_rate": 2.8775676409374717e-05,
|
| 759 |
+
"loss": 9.2353,
|
| 760 |
+
"step": 2120
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.009682425492830934,
|
| 764 |
+
"grad_norm": 2.3209128379821777,
|
| 765 |
+
"learning_rate": 2.9047145054746177e-05,
|
| 766 |
+
"loss": 9.241,
|
| 767 |
+
"step": 2140
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.009772915450707858,
|
| 771 |
+
"grad_norm": 2.3225491046905518,
|
| 772 |
+
"learning_rate": 2.9318613700117634e-05,
|
| 773 |
+
"loss": 9.2133,
|
| 774 |
+
"step": 2160
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.009863405408584783,
|
| 778 |
+
"grad_norm": 2.0134568214416504,
|
| 779 |
+
"learning_rate": 2.9590082345489093e-05,
|
| 780 |
+
"loss": 9.2188,
|
| 781 |
+
"step": 2180
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.009953895366461707,
|
| 785 |
+
"grad_norm": 3.033569574356079,
|
| 786 |
+
"learning_rate": 2.9861550990860557e-05,
|
| 787 |
+
"loss": 9.2131,
|
| 788 |
+
"step": 2200
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.010044385324338632,
|
| 792 |
+
"grad_norm": 2.8993263244628906,
|
| 793 |
+
"learning_rate": 3.0133019636232017e-05,
|
| 794 |
+
"loss": 9.2119,
|
| 795 |
+
"step": 2220
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.010134875282215556,
|
| 799 |
+
"grad_norm": 2.718588352203369,
|
| 800 |
+
"learning_rate": 3.0404488281603473e-05,
|
| 801 |
+
"loss": 9.2187,
|
| 802 |
+
"step": 2240
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.01022536524009248,
|
| 806 |
+
"grad_norm": 2.635470390319824,
|
| 807 |
+
"learning_rate": 3.0675956926974936e-05,
|
| 808 |
+
"loss": 9.1953,
|
| 809 |
+
"step": 2260
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.010315855197969405,
|
| 813 |
+
"grad_norm": 2.6032440662384033,
|
| 814 |
+
"learning_rate": 3.094742557234639e-05,
|
| 815 |
+
"loss": 9.1967,
|
| 816 |
+
"step": 2280
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.01040634515584633,
|
| 820 |
+
"grad_norm": 2.4713950157165527,
|
| 821 |
+
"learning_rate": 3.121889421771785e-05,
|
| 822 |
+
"loss": 9.1881,
|
| 823 |
+
"step": 2300
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.010496835113723254,
|
| 827 |
+
"grad_norm": 2.4573025703430176,
|
| 828 |
+
"learning_rate": 3.149036286308931e-05,
|
| 829 |
+
"loss": 9.1827,
|
| 830 |
+
"step": 2320
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.010587325071600179,
|
| 834 |
+
"grad_norm": 2.6169447898864746,
|
| 835 |
+
"learning_rate": 3.1761831508460776e-05,
|
| 836 |
+
"loss": 9.1865,
|
| 837 |
+
"step": 2340
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.010677815029477103,
|
| 841 |
+
"grad_norm": 2.6744954586029053,
|
| 842 |
+
"learning_rate": 3.203330015383223e-05,
|
| 843 |
+
"loss": 9.1829,
|
| 844 |
+
"step": 2360
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.010768304987354028,
|
| 848 |
+
"grad_norm": 2.766223907470703,
|
| 849 |
+
"learning_rate": 3.230476879920369e-05,
|
| 850 |
+
"loss": 9.177,
|
| 851 |
+
"step": 2380
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.010858794945230952,
|
| 855 |
+
"grad_norm": 2.8083655834198,
|
| 856 |
+
"learning_rate": 3.257623744457515e-05,
|
| 857 |
+
"loss": 9.1853,
|
| 858 |
+
"step": 2400
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.010949284903107877,
|
| 862 |
+
"grad_norm": 4.484155178070068,
|
| 863 |
+
"learning_rate": 3.284770608994661e-05,
|
| 864 |
+
"loss": 9.1655,
|
| 865 |
+
"step": 2420
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.011039774860984803,
|
| 869 |
+
"grad_norm": 3.5152087211608887,
|
| 870 |
+
"learning_rate": 3.311917473531807e-05,
|
| 871 |
+
"loss": 9.1516,
|
| 872 |
+
"step": 2440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.011130264818861728,
|
| 876 |
+
"grad_norm": 2.3122165203094482,
|
| 877 |
+
"learning_rate": 3.339064338068953e-05,
|
| 878 |
+
"loss": 9.1552,
|
| 879 |
+
"step": 2460
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.011220754776738652,
|
| 883 |
+
"grad_norm": 3.0563108921051025,
|
| 884 |
+
"learning_rate": 3.3662112026060985e-05,
|
| 885 |
+
"loss": 9.1494,
|
| 886 |
+
"step": 2480
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.011311244734615577,
|
| 890 |
+
"grad_norm": 3.926668882369995,
|
| 891 |
+
"learning_rate": 3.393358067143245e-05,
|
| 892 |
+
"loss": 9.1425,
|
| 893 |
+
"step": 2500
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.011401734692492501,
|
| 897 |
+
"grad_norm": 2.7006709575653076,
|
| 898 |
+
"learning_rate": 3.420504931680391e-05,
|
| 899 |
+
"loss": 9.1328,
|
| 900 |
+
"step": 2520
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.011492224650369426,
|
| 904 |
+
"grad_norm": 3.1082751750946045,
|
| 905 |
+
"learning_rate": 3.447651796217537e-05,
|
| 906 |
+
"loss": 9.1316,
|
| 907 |
+
"step": 2540
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.01158271460824635,
|
| 911 |
+
"grad_norm": 2.744490385055542,
|
| 912 |
+
"learning_rate": 3.4747986607546824e-05,
|
| 913 |
+
"loss": 9.1193,
|
| 914 |
+
"step": 2560
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.011673204566123275,
|
| 918 |
+
"grad_norm": 2.8441922664642334,
|
| 919 |
+
"learning_rate": 3.501945525291829e-05,
|
| 920 |
+
"loss": 9.1174,
|
| 921 |
+
"step": 2580
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.011763694524000199,
|
| 925 |
+
"grad_norm": 3.7371647357940674,
|
| 926 |
+
"learning_rate": 3.529092389828975e-05,
|
| 927 |
+
"loss": 9.1217,
|
| 928 |
+
"step": 2600
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.011854184481877124,
|
| 932 |
+
"grad_norm": 3.0141730308532715,
|
| 933 |
+
"learning_rate": 3.556239254366121e-05,
|
| 934 |
+
"loss": 9.0999,
|
| 935 |
+
"step": 2620
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.011944674439754048,
|
| 939 |
+
"grad_norm": 2.9731669425964355,
|
| 940 |
+
"learning_rate": 3.5833861189032664e-05,
|
| 941 |
+
"loss": 9.1044,
|
| 942 |
+
"step": 2640
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.012035164397630973,
|
| 946 |
+
"grad_norm": 3.166254997253418,
|
| 947 |
+
"learning_rate": 3.610532983440413e-05,
|
| 948 |
+
"loss": 9.103,
|
| 949 |
+
"step": 2660
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.012125654355507897,
|
| 953 |
+
"grad_norm": 2.949646472930908,
|
| 954 |
+
"learning_rate": 3.6376798479775584e-05,
|
| 955 |
+
"loss": 9.1026,
|
| 956 |
+
"step": 2680
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.012216144313384822,
|
| 960 |
+
"grad_norm": 2.762843132019043,
|
| 961 |
+
"learning_rate": 3.664826712514705e-05,
|
| 962 |
+
"loss": 9.1047,
|
| 963 |
+
"step": 2700
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.012306634271261746,
|
| 967 |
+
"grad_norm": 3.188957929611206,
|
| 968 |
+
"learning_rate": 3.6919735770518503e-05,
|
| 969 |
+
"loss": 9.0968,
|
| 970 |
+
"step": 2720
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.01239712422913867,
|
| 974 |
+
"grad_norm": 4.116425037384033,
|
| 975 |
+
"learning_rate": 3.719120441588996e-05,
|
| 976 |
+
"loss": 9.0993,
|
| 977 |
+
"step": 2740
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.012487614187015597,
|
| 981 |
+
"grad_norm": 2.7521297931671143,
|
| 982 |
+
"learning_rate": 3.746267306126142e-05,
|
| 983 |
+
"loss": 9.063,
|
| 984 |
+
"step": 2760
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.012578104144892521,
|
| 988 |
+
"grad_norm": 3.1481823921203613,
|
| 989 |
+
"learning_rate": 3.7734141706632886e-05,
|
| 990 |
+
"loss": 9.062,
|
| 991 |
+
"step": 2780
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.012668594102769446,
|
| 995 |
+
"grad_norm": 2.48091721534729,
|
| 996 |
+
"learning_rate": 3.800561035200434e-05,
|
| 997 |
+
"loss": 9.0727,
|
| 998 |
+
"step": 2800
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.01275908406064637,
|
| 1002 |
+
"grad_norm": 3.0816426277160645,
|
| 1003 |
+
"learning_rate": 3.82770789973758e-05,
|
| 1004 |
+
"loss": 9.0525,
|
| 1005 |
+
"step": 2820
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.012849574018523295,
|
| 1009 |
+
"grad_norm": 2.86342191696167,
|
| 1010 |
+
"learning_rate": 3.854854764274726e-05,
|
| 1011 |
+
"loss": 9.0447,
|
| 1012 |
+
"step": 2840
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.01294006397640022,
|
| 1016 |
+
"grad_norm": 2.769746780395508,
|
| 1017 |
+
"learning_rate": 3.8820016288118726e-05,
|
| 1018 |
+
"loss": 9.0524,
|
| 1019 |
+
"step": 2860
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.013030553934277144,
|
| 1023 |
+
"grad_norm": 3.4716339111328125,
|
| 1024 |
+
"learning_rate": 3.909148493349018e-05,
|
| 1025 |
+
"loss": 9.0453,
|
| 1026 |
+
"step": 2880
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.013121043892154068,
|
| 1030 |
+
"grad_norm": 4.585721969604492,
|
| 1031 |
+
"learning_rate": 3.936295357886164e-05,
|
| 1032 |
+
"loss": 9.0466,
|
| 1033 |
+
"step": 2900
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.013211533850030993,
|
| 1037 |
+
"grad_norm": 3.7394728660583496,
|
| 1038 |
+
"learning_rate": 3.96344222242331e-05,
|
| 1039 |
+
"loss": 9.0405,
|
| 1040 |
+
"step": 2920
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.013302023807907917,
|
| 1044 |
+
"grad_norm": 3.9100561141967773,
|
| 1045 |
+
"learning_rate": 3.990589086960456e-05,
|
| 1046 |
+
"loss": 9.0415,
|
| 1047 |
+
"step": 2940
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.013392513765784842,
|
| 1051 |
+
"grad_norm": 2.94941782951355,
|
| 1052 |
+
"learning_rate": 4.017735951497602e-05,
|
| 1053 |
+
"loss": 9.0265,
|
| 1054 |
+
"step": 2960
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.013483003723661766,
|
| 1058 |
+
"grad_norm": 2.6733226776123047,
|
| 1059 |
+
"learning_rate": 4.044882816034748e-05,
|
| 1060 |
+
"loss": 9.0195,
|
| 1061 |
+
"step": 2980
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.01357349368153869,
|
| 1065 |
+
"grad_norm": 3.4839463233947754,
|
| 1066 |
+
"learning_rate": 4.0720296805718935e-05,
|
| 1067 |
+
"loss": 9.0204,
|
| 1068 |
+
"step": 3000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.013663983639415615,
|
| 1072 |
+
"grad_norm": 3.460050344467163,
|
| 1073 |
+
"learning_rate": 4.09917654510904e-05,
|
| 1074 |
+
"loss": 9.0086,
|
| 1075 |
+
"step": 3020
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.01375447359729254,
|
| 1079 |
+
"grad_norm": 4.007343769073486,
|
| 1080 |
+
"learning_rate": 4.126323409646186e-05,
|
| 1081 |
+
"loss": 9.0185,
|
| 1082 |
+
"step": 3040
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.013844963555169464,
|
| 1086 |
+
"grad_norm": 3.917860746383667,
|
| 1087 |
+
"learning_rate": 4.153470274183331e-05,
|
| 1088 |
+
"loss": 9.0032,
|
| 1089 |
+
"step": 3060
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.01393545351304639,
|
| 1093 |
+
"grad_norm": 3.5258123874664307,
|
| 1094 |
+
"learning_rate": 4.1806171387204775e-05,
|
| 1095 |
+
"loss": 8.9983,
|
| 1096 |
+
"step": 3080
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.014025943470923315,
|
| 1100 |
+
"grad_norm": 3.002183198928833,
|
| 1101 |
+
"learning_rate": 4.207764003257624e-05,
|
| 1102 |
+
"loss": 8.9898,
|
| 1103 |
+
"step": 3100
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.01411643342880024,
|
| 1107 |
+
"grad_norm": 3.2682976722717285,
|
| 1108 |
+
"learning_rate": 4.23491086779477e-05,
|
| 1109 |
+
"loss": 8.9932,
|
| 1110 |
+
"step": 3120
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.014206923386677164,
|
| 1114 |
+
"grad_norm": 3.7955832481384277,
|
| 1115 |
+
"learning_rate": 4.262057732331915e-05,
|
| 1116 |
+
"loss": 8.9879,
|
| 1117 |
+
"step": 3140
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.014297413344554089,
|
| 1121 |
+
"grad_norm": 3.3697524070739746,
|
| 1122 |
+
"learning_rate": 4.2892045968690614e-05,
|
| 1123 |
+
"loss": 8.9757,
|
| 1124 |
+
"step": 3160
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.014387903302431013,
|
| 1128 |
+
"grad_norm": 3.756788730621338,
|
| 1129 |
+
"learning_rate": 4.316351461406208e-05,
|
| 1130 |
+
"loss": 8.9811,
|
| 1131 |
+
"step": 3180
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.014478393260307938,
|
| 1135 |
+
"grad_norm": 3.024722099304199,
|
| 1136 |
+
"learning_rate": 4.3434983259433534e-05,
|
| 1137 |
+
"loss": 8.9613,
|
| 1138 |
+
"step": 3200
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.014568883218184862,
|
| 1142 |
+
"grad_norm": 3.258375406265259,
|
| 1143 |
+
"learning_rate": 4.3706451904805e-05,
|
| 1144 |
+
"loss": 8.9614,
|
| 1145 |
+
"step": 3220
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.014659373176061787,
|
| 1149 |
+
"grad_norm": 2.970426559448242,
|
| 1150 |
+
"learning_rate": 4.3977920550176454e-05,
|
| 1151 |
+
"loss": 8.9624,
|
| 1152 |
+
"step": 3240
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.014749863133938711,
|
| 1156 |
+
"grad_norm": 4.601590156555176,
|
| 1157 |
+
"learning_rate": 4.424938919554791e-05,
|
| 1158 |
+
"loss": 8.9615,
|
| 1159 |
+
"step": 3260
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.014840353091815636,
|
| 1163 |
+
"grad_norm": 4.773068428039551,
|
| 1164 |
+
"learning_rate": 4.4520857840919373e-05,
|
| 1165 |
+
"loss": 8.9668,
|
| 1166 |
+
"step": 3280
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.01493084304969256,
|
| 1170 |
+
"grad_norm": 3.182677984237671,
|
| 1171 |
+
"learning_rate": 4.479232648629084e-05,
|
| 1172 |
+
"loss": 8.933,
|
| 1173 |
+
"step": 3300
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.015021333007569485,
|
| 1177 |
+
"grad_norm": 3.160553455352783,
|
| 1178 |
+
"learning_rate": 4.5063795131662286e-05,
|
| 1179 |
+
"loss": 8.9409,
|
| 1180 |
+
"step": 3320
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.015111822965446409,
|
| 1184 |
+
"grad_norm": 3.0617620944976807,
|
| 1185 |
+
"learning_rate": 4.533526377703375e-05,
|
| 1186 |
+
"loss": 8.95,
|
| 1187 |
+
"step": 3340
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.015202312923323334,
|
| 1191 |
+
"grad_norm": 3.1966211795806885,
|
| 1192 |
+
"learning_rate": 4.560673242240521e-05,
|
| 1193 |
+
"loss": 8.9379,
|
| 1194 |
+
"step": 3360
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.015292802881200258,
|
| 1198 |
+
"grad_norm": 2.3314368724823,
|
| 1199 |
+
"learning_rate": 4.587820106777667e-05,
|
| 1200 |
+
"loss": 8.9246,
|
| 1201 |
+
"step": 3380
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.015383292839077184,
|
| 1205 |
+
"grad_norm": 3.1242740154266357,
|
| 1206 |
+
"learning_rate": 4.6149669713148126e-05,
|
| 1207 |
+
"loss": 8.9409,
|
| 1208 |
+
"step": 3400
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.015473782796954109,
|
| 1212 |
+
"grad_norm": 3.042051315307617,
|
| 1213 |
+
"learning_rate": 4.642113835851959e-05,
|
| 1214 |
+
"loss": 8.9204,
|
| 1215 |
+
"step": 3420
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.015564272754831033,
|
| 1219 |
+
"grad_norm": 4.102015495300293,
|
| 1220 |
+
"learning_rate": 4.669260700389105e-05,
|
| 1221 |
+
"loss": 8.8915,
|
| 1222 |
+
"step": 3440
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.015654762712707958,
|
| 1226 |
+
"grad_norm": 3.2991299629211426,
|
| 1227 |
+
"learning_rate": 4.696407564926251e-05,
|
| 1228 |
+
"loss": 8.8897,
|
| 1229 |
+
"step": 3460
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.01574525267058488,
|
| 1233 |
+
"grad_norm": 3.501094102859497,
|
| 1234 |
+
"learning_rate": 4.7235544294633965e-05,
|
| 1235 |
+
"loss": 8.9223,
|
| 1236 |
+
"step": 3480
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.015835742628461807,
|
| 1240 |
+
"grad_norm": 6.248113632202148,
|
| 1241 |
+
"learning_rate": 4.750701294000543e-05,
|
| 1242 |
+
"loss": 8.8925,
|
| 1243 |
+
"step": 3500
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.01592623258633873,
|
| 1247 |
+
"grad_norm": 4.329127788543701,
|
| 1248 |
+
"learning_rate": 4.7778481585376885e-05,
|
| 1249 |
+
"loss": 8.8891,
|
| 1250 |
+
"step": 3520
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.016016722544215656,
|
| 1254 |
+
"grad_norm": 3.575141191482544,
|
| 1255 |
+
"learning_rate": 4.804995023074835e-05,
|
| 1256 |
+
"loss": 8.8741,
|
| 1257 |
+
"step": 3540
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.01610721250209258,
|
| 1261 |
+
"grad_norm": 3.301194429397583,
|
| 1262 |
+
"learning_rate": 4.832141887611981e-05,
|
| 1263 |
+
"loss": 8.8965,
|
| 1264 |
+
"step": 3560
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.016197702459969505,
|
| 1268 |
+
"grad_norm": 3.7364182472229004,
|
| 1269 |
+
"learning_rate": 4.859288752149126e-05,
|
| 1270 |
+
"loss": 8.8899,
|
| 1271 |
+
"step": 3580
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.01628819241784643,
|
| 1275 |
+
"grad_norm": 5.336267471313477,
|
| 1276 |
+
"learning_rate": 4.8864356166862725e-05,
|
| 1277 |
+
"loss": 8.8959,
|
| 1278 |
+
"step": 3600
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.016378682375723354,
|
| 1282 |
+
"grad_norm": 4.769089221954346,
|
| 1283 |
+
"learning_rate": 4.913582481223419e-05,
|
| 1284 |
+
"loss": 8.8981,
|
| 1285 |
+
"step": 3620
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.01646917233360028,
|
| 1289 |
+
"grad_norm": 3.369799852371216,
|
| 1290 |
+
"learning_rate": 4.9407293457605645e-05,
|
| 1291 |
+
"loss": 8.8954,
|
| 1292 |
+
"step": 3640
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.016559662291477203,
|
| 1296 |
+
"grad_norm": 3.063030481338501,
|
| 1297 |
+
"learning_rate": 4.96787621029771e-05,
|
| 1298 |
+
"loss": 8.8694,
|
| 1299 |
+
"step": 3660
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.01665015224935413,
|
| 1303 |
+
"grad_norm": 4.988938331604004,
|
| 1304 |
+
"learning_rate": 4.9950230748348564e-05,
|
| 1305 |
+
"loss": 8.8611,
|
| 1306 |
+
"step": 3680
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.016740642207231052,
|
| 1310 |
+
"grad_norm": 3.5118601322174072,
|
| 1311 |
+
"learning_rate": 5.022169939372003e-05,
|
| 1312 |
+
"loss": 8.8525,
|
| 1313 |
+
"step": 3700
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.016831132165107978,
|
| 1317 |
+
"grad_norm": 4.257157325744629,
|
| 1318 |
+
"learning_rate": 5.0493168039091484e-05,
|
| 1319 |
+
"loss": 8.8547,
|
| 1320 |
+
"step": 3720
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.0169216221229849,
|
| 1324 |
+
"grad_norm": 3.7021615505218506,
|
| 1325 |
+
"learning_rate": 5.076463668446294e-05,
|
| 1326 |
+
"loss": 8.8572,
|
| 1327 |
+
"step": 3740
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.017012112080861827,
|
| 1331 |
+
"grad_norm": 4.868439197540283,
|
| 1332 |
+
"learning_rate": 5.1036105329834404e-05,
|
| 1333 |
+
"loss": 8.8684,
|
| 1334 |
+
"step": 3760
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.01710260203873875,
|
| 1338 |
+
"grad_norm": 6.547580718994141,
|
| 1339 |
+
"learning_rate": 5.130757397520586e-05,
|
| 1340 |
+
"loss": 8.828,
|
| 1341 |
+
"step": 3780
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.017193091996615676,
|
| 1345 |
+
"grad_norm": 5.9254374504089355,
|
| 1346 |
+
"learning_rate": 5.1579042620577324e-05,
|
| 1347 |
+
"loss": 8.838,
|
| 1348 |
+
"step": 3800
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 0.0172835819544926,
|
| 1352 |
+
"grad_norm": 6.061065196990967,
|
| 1353 |
+
"learning_rate": 5.185051126594879e-05,
|
| 1354 |
+
"loss": 8.8405,
|
| 1355 |
+
"step": 3820
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 0.017374071912369525,
|
| 1359 |
+
"grad_norm": 6.026751518249512,
|
| 1360 |
+
"learning_rate": 5.2121979911320237e-05,
|
| 1361 |
+
"loss": 8.8305,
|
| 1362 |
+
"step": 3840
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 0.017464561870246448,
|
| 1366 |
+
"grad_norm": 4.982965469360352,
|
| 1367 |
+
"learning_rate": 5.23934485566917e-05,
|
| 1368 |
+
"loss": 8.8316,
|
| 1369 |
+
"step": 3860
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 0.017555051828123374,
|
| 1373 |
+
"grad_norm": 9.080221176147461,
|
| 1374 |
+
"learning_rate": 5.266491720206316e-05,
|
| 1375 |
+
"loss": 8.8267,
|
| 1376 |
+
"step": 3880
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.0176455417860003,
|
| 1380 |
+
"grad_norm": 6.644583225250244,
|
| 1381 |
+
"learning_rate": 5.293638584743462e-05,
|
| 1382 |
+
"loss": 8.8331,
|
| 1383 |
+
"step": 3900
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.017736031743877223,
|
| 1387 |
+
"grad_norm": 6.022925853729248,
|
| 1388 |
+
"learning_rate": 5.3207854492806076e-05,
|
| 1389 |
+
"loss": 8.8198,
|
| 1390 |
+
"step": 3920
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 0.01782652170175415,
|
| 1394 |
+
"grad_norm": 4.794320583343506,
|
| 1395 |
+
"learning_rate": 5.347932313817754e-05,
|
| 1396 |
+
"loss": 8.8075,
|
| 1397 |
+
"step": 3940
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 0.017917011659631072,
|
| 1401 |
+
"grad_norm": 5.949656963348389,
|
| 1402 |
+
"learning_rate": 5.3750791783548996e-05,
|
| 1403 |
+
"loss": 8.8175,
|
| 1404 |
+
"step": 3960
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.018007501617508,
|
| 1408 |
+
"grad_norm": 7.972283840179443,
|
| 1409 |
+
"learning_rate": 5.402226042892046e-05,
|
| 1410 |
+
"loss": 8.8263,
|
| 1411 |
+
"step": 3980
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 0.01809799157538492,
|
| 1415 |
+
"grad_norm": 6.132015228271484,
|
| 1416 |
+
"learning_rate": 5.4293729074291916e-05,
|
| 1417 |
+
"loss": 8.8035,
|
| 1418 |
+
"step": 4000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.01809799157538492,
|
| 1422 |
+
"eval_accuracy": 0.10955227810888264,
|
| 1423 |
+
"eval_loss": 8.793069839477539,
|
| 1424 |
+
"eval_runtime": 217.825,
|
| 1425 |
+
"eval_samples_per_second": 2790.497,
|
| 1426 |
+
"eval_steps_per_second": 10.903,
|
| 1427 |
+
"step": 4000
|
| 1428 |
}
|
| 1429 |
],
|
| 1430 |
"logging_steps": 20,
|
|
|
|
| 1432 |
"num_input_tokens_seen": 0,
|
| 1433 |
"num_train_epochs": 3,
|
| 1434 |
"save_steps": 100,
|
| 1435 |
+
"total_flos": 1438362107904000.0,
|
| 1436 |
"train_batch_size": 256,
|
| 1437 |
"trial_name": null,
|
| 1438 |
"trial_params": null
|