Commit
·
de5d682
1
Parent(s):
1fa4da1
Upload folder using huggingface_hub
Browse files- adapter_model.bin +1 -1
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +750 -2
- training_args.bin +1 -1
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 75641741
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9af4c3a65b7a773b24b082fa3d2bdf3b889f0c61459c6c6dc53f6e339785bcd4
|
| 3 |
size 75641741
|
optimizer.pt
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 sha256:461887c9ec08fe4b1692b9fa6def1741d6a60b64bb08c37d3fdc064aa795bd26
|
| 3 |
+
size 151222021
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdda7482499b855a06935901db3ef5c0346fd12eb58a510bec30c9e4dab13b1e
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:749e3338e97b1ab6783e7f614fd50b2475fd9e06f92f74c2d73b555d52907729
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -739,6 +739,754 @@
|
|
| 739 |
"rewards/margins": -0.20669928193092346,
|
| 740 |
"rewards/rejected": -5.946280479431152,
|
| 741 |
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 742 |
}
|
| 743 |
],
|
| 744 |
"logging_steps": 10,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.19331142470520007,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 739 |
"rewards/margins": -0.20669928193092346,
|
| 740 |
"rewards/rejected": -5.946280479431152,
|
| 741 |
"step": 500
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.1,
|
| 745 |
+
"learning_rate": 0.0004777992277992278,
|
| 746 |
+
"logits/chosen": -2.692495584487915,
|
| 747 |
+
"logits/rejected": -2.6421055793762207,
|
| 748 |
+
"logps/chosen": -172.01806640625,
|
| 749 |
+
"logps/rejected": -165.9178466796875,
|
| 750 |
+
"loss": 1.2994,
|
| 751 |
+
"rewards/accuracies": 0.574999988079071,
|
| 752 |
+
"rewards/chosen": -4.898122787475586,
|
| 753 |
+
"rewards/margins": 0.04086846113204956,
|
| 754 |
+
"rewards/rejected": -4.938991546630859,
|
| 755 |
+
"step": 510
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.1,
|
| 759 |
+
"learning_rate": 0.0004874517374517375,
|
| 760 |
+
"logits/chosen": -2.7154297828674316,
|
| 761 |
+
"logits/rejected": -2.5865087509155273,
|
| 762 |
+
"logps/chosen": -134.430419921875,
|
| 763 |
+
"logps/rejected": -144.77151489257812,
|
| 764 |
+
"loss": 1.0245,
|
| 765 |
+
"rewards/accuracies": 0.574999988079071,
|
| 766 |
+
"rewards/chosen": -3.9665591716766357,
|
| 767 |
+
"rewards/margins": 0.17584654688835144,
|
| 768 |
+
"rewards/rejected": -4.1424055099487305,
|
| 769 |
+
"step": 520
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.1,
|
| 773 |
+
"learning_rate": 0.0004961389961389962,
|
| 774 |
+
"logits/chosen": -2.863107442855835,
|
| 775 |
+
"logits/rejected": -2.8306002616882324,
|
| 776 |
+
"logps/chosen": -269.8907165527344,
|
| 777 |
+
"logps/rejected": -273.8556823730469,
|
| 778 |
+
"loss": 3.3909,
|
| 779 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 780 |
+
"rewards/chosen": -13.715133666992188,
|
| 781 |
+
"rewards/margins": 0.3205181956291199,
|
| 782 |
+
"rewards/rejected": -14.035652160644531,
|
| 783 |
+
"step": 530
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.1,
|
| 787 |
+
"learning_rate": 0.0004999979503849796,
|
| 788 |
+
"logits/chosen": -3.0216221809387207,
|
| 789 |
+
"logits/rejected": -3.014930248260498,
|
| 790 |
+
"logps/chosen": -194.19422912597656,
|
| 791 |
+
"logps/rejected": -201.08251953125,
|
| 792 |
+
"loss": 2.3037,
|
| 793 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 794 |
+
"rewards/chosen": -7.459778785705566,
|
| 795 |
+
"rewards/margins": -0.22785942256450653,
|
| 796 |
+
"rewards/rejected": -7.231919288635254,
|
| 797 |
+
"step": 540
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.11,
|
| 801 |
+
"learning_rate": 0.0004999854250815602,
|
| 802 |
+
"logits/chosen": -2.967331647872925,
|
| 803 |
+
"logits/rejected": -2.933845043182373,
|
| 804 |
+
"logps/chosen": -215.81240844726562,
|
| 805 |
+
"logps/rejected": -204.9075927734375,
|
| 806 |
+
"loss": 2.1232,
|
| 807 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 808 |
+
"rewards/chosen": -7.889649868011475,
|
| 809 |
+
"rewards/margins": 0.17769476771354675,
|
| 810 |
+
"rewards/rejected": -8.067344665527344,
|
| 811 |
+
"step": 550
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 0.11,
|
| 815 |
+
"learning_rate": 0.00049996151371953,
|
| 816 |
+
"logits/chosen": -3.052489757537842,
|
| 817 |
+
"logits/rejected": -3.0332350730895996,
|
| 818 |
+
"logps/chosen": -181.4810791015625,
|
| 819 |
+
"logps/rejected": -172.9596405029297,
|
| 820 |
+
"loss": 1.3367,
|
| 821 |
+
"rewards/accuracies": 0.5,
|
| 822 |
+
"rewards/chosen": -5.431277275085449,
|
| 823 |
+
"rewards/margins": 0.03274815157055855,
|
| 824 |
+
"rewards/rejected": -5.464025020599365,
|
| 825 |
+
"step": 560
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.11,
|
| 829 |
+
"learning_rate": 0.0004999262173879769,
|
| 830 |
+
"logits/chosen": -3.040531635284424,
|
| 831 |
+
"logits/rejected": -3.037515163421631,
|
| 832 |
+
"logps/chosen": -179.90150451660156,
|
| 833 |
+
"logps/rejected": -185.0644989013672,
|
| 834 |
+
"loss": 1.4866,
|
| 835 |
+
"rewards/accuracies": 0.574999988079071,
|
| 836 |
+
"rewards/chosen": -6.061680316925049,
|
| 837 |
+
"rewards/margins": 0.6980069279670715,
|
| 838 |
+
"rewards/rejected": -6.759686470031738,
|
| 839 |
+
"step": 570
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 0.11,
|
| 843 |
+
"learning_rate": 0.0004998795376945392,
|
| 844 |
+
"logits/chosen": -3.021232843399048,
|
| 845 |
+
"logits/rejected": -2.9936585426330566,
|
| 846 |
+
"logps/chosen": -175.13389587402344,
|
| 847 |
+
"logps/rejected": -156.2028350830078,
|
| 848 |
+
"loss": 1.7235,
|
| 849 |
+
"rewards/accuracies": 0.375,
|
| 850 |
+
"rewards/chosen": -5.4631123542785645,
|
| 851 |
+
"rewards/margins": -0.5610149502754211,
|
| 852 |
+
"rewards/rejected": -4.902098178863525,
|
| 853 |
+
"step": 580
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 0.11,
|
| 857 |
+
"learning_rate": 0.0004998214767653319,
|
| 858 |
+
"logits/chosen": -2.944594621658325,
|
| 859 |
+
"logits/rejected": -3.0027194023132324,
|
| 860 |
+
"logps/chosen": -189.1737823486328,
|
| 861 |
+
"logps/rejected": -175.7019500732422,
|
| 862 |
+
"loss": 1.9699,
|
| 863 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 864 |
+
"rewards/chosen": -6.499063014984131,
|
| 865 |
+
"rewards/margins": -0.3209795355796814,
|
| 866 |
+
"rewards/rejected": -6.178082466125488,
|
| 867 |
+
"step": 590
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.12,
|
| 871 |
+
"learning_rate": 0.0004997520372448494,
|
| 872 |
+
"logits/chosen": -2.8185458183288574,
|
| 873 |
+
"logits/rejected": -2.798320770263672,
|
| 874 |
+
"logps/chosen": -262.44500732421875,
|
| 875 |
+
"logps/rejected": -245.83889770507812,
|
| 876 |
+
"loss": 3.8268,
|
| 877 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 878 |
+
"rewards/chosen": -12.636558532714844,
|
| 879 |
+
"rewards/margins": -0.5035432577133179,
|
| 880 |
+
"rewards/rejected": -12.133015632629395,
|
| 881 |
+
"step": 600
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.12,
|
| 885 |
+
"eval_logits/chosen": -2.9935925006866455,
|
| 886 |
+
"eval_logits/rejected": -2.961137533187866,
|
| 887 |
+
"eval_logps/chosen": -184.5051727294922,
|
| 888 |
+
"eval_logps/rejected": -181.61184692382812,
|
| 889 |
+
"eval_loss": 1.3598365783691406,
|
| 890 |
+
"eval_rewards/accuracies": 0.5193312168121338,
|
| 891 |
+
"eval_rewards/chosen": -6.594781875610352,
|
| 892 |
+
"eval_rewards/margins": 0.12169010192155838,
|
| 893 |
+
"eval_rewards/rejected": -6.7164716720581055,
|
| 894 |
+
"eval_runtime": 1314.8357,
|
| 895 |
+
"eval_samples_per_second": 0.728,
|
| 896 |
+
"eval_steps_per_second": 0.728,
|
| 897 |
+
"step": 600
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.12,
|
| 901 |
+
"learning_rate": 0.0004996712222958462,
|
| 902 |
+
"logits/chosen": -2.9863028526306152,
|
| 903 |
+
"logits/rejected": -2.841834306716919,
|
| 904 |
+
"logps/chosen": -234.49893188476562,
|
| 905 |
+
"logps/rejected": -215.51123046875,
|
| 906 |
+
"loss": 3.4529,
|
| 907 |
+
"rewards/accuracies": 0.574999988079071,
|
| 908 |
+
"rewards/chosen": -11.542932510375977,
|
| 909 |
+
"rewards/margins": -1.085506796836853,
|
| 910 |
+
"rewards/rejected": -10.457425117492676,
|
| 911 |
+
"step": 610
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.12,
|
| 915 |
+
"learning_rate": 0.0004995790355991916,
|
| 916 |
+
"logits/chosen": -2.9887356758117676,
|
| 917 |
+
"logits/rejected": -2.887108564376831,
|
| 918 |
+
"logps/chosen": -202.21853637695312,
|
| 919 |
+
"logps/rejected": -198.64749145507812,
|
| 920 |
+
"loss": 1.6552,
|
| 921 |
+
"rewards/accuracies": 0.574999988079071,
|
| 922 |
+
"rewards/chosen": -7.725579261779785,
|
| 923 |
+
"rewards/margins": 0.2432982176542282,
|
| 924 |
+
"rewards/rejected": -7.96887731552124,
|
| 925 |
+
"step": 620
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.12,
|
| 929 |
+
"learning_rate": 0.0004994754813537031,
|
| 930 |
+
"logits/chosen": -3.116293430328369,
|
| 931 |
+
"logits/rejected": -3.126661539077759,
|
| 932 |
+
"logps/chosen": -201.38604736328125,
|
| 933 |
+
"logps/rejected": -192.84515380859375,
|
| 934 |
+
"loss": 1.9923,
|
| 935 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 936 |
+
"rewards/chosen": -8.627758979797363,
|
| 937 |
+
"rewards/margins": -0.4443356990814209,
|
| 938 |
+
"rewards/rejected": -8.183423042297363,
|
| 939 |
+
"step": 630
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.12,
|
| 943 |
+
"learning_rate": 0.000499372567166064,
|
| 944 |
+
"logits/chosen": -3.480961561203003,
|
| 945 |
+
"logits/rejected": -3.476128339767456,
|
| 946 |
+
"logps/chosen": -226.2942657470703,
|
| 947 |
+
"logps/rejected": -198.20462036132812,
|
| 948 |
+
"loss": 3.5697,
|
| 949 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 950 |
+
"rewards/chosen": -10.370187759399414,
|
| 951 |
+
"rewards/margins": -1.4912974834442139,
|
| 952 |
+
"rewards/rejected": -8.878890037536621,
|
| 953 |
+
"step": 640
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.13,
|
| 957 |
+
"learning_rate": 0.0004992474279997049,
|
| 958 |
+
"logits/chosen": -3.380039930343628,
|
| 959 |
+
"logits/rejected": -3.3540236949920654,
|
| 960 |
+
"logps/chosen": -183.65228271484375,
|
| 961 |
+
"logps/rejected": -192.43350219726562,
|
| 962 |
+
"loss": 2.0561,
|
| 963 |
+
"rewards/accuracies": 0.512499988079071,
|
| 964 |
+
"rewards/chosen": -6.808587551116943,
|
| 965 |
+
"rewards/margins": 1.2175410985946655,
|
| 966 |
+
"rewards/rejected": -8.026129722595215,
|
| 967 |
+
"step": 650
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.13,
|
| 971 |
+
"learning_rate": 0.0004991109363882065,
|
| 972 |
+
"logits/chosen": -2.5853641033172607,
|
| 973 |
+
"logits/rejected": -2.6103484630584717,
|
| 974 |
+
"logps/chosen": -464.68310546875,
|
| 975 |
+
"logps/rejected": -447.2250061035156,
|
| 976 |
+
"loss": 11.4738,
|
| 977 |
+
"rewards/accuracies": 0.5625,
|
| 978 |
+
"rewards/chosen": -34.34156036376953,
|
| 979 |
+
"rewards/margins": -1.95975661277771,
|
| 980 |
+
"rewards/rejected": -32.381797790527344,
|
| 981 |
+
"step": 660
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.13,
|
| 985 |
+
"learning_rate": 0.0004989630985483375,
|
| 986 |
+
"logits/chosen": -2.7174885272979736,
|
| 987 |
+
"logits/rejected": -2.707397937774658,
|
| 988 |
+
"logps/chosen": -447.7940979003906,
|
| 989 |
+
"logps/rejected": -407.99493408203125,
|
| 990 |
+
"loss": 12.9625,
|
| 991 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 992 |
+
"rewards/chosen": -32.60967254638672,
|
| 993 |
+
"rewards/margins": -3.206895112991333,
|
| 994 |
+
"rewards/rejected": -29.40277671813965,
|
| 995 |
+
"step": 670
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.13,
|
| 999 |
+
"learning_rate": 0.0004988203490218075,
|
| 1000 |
+
"logits/chosen": -2.946742296218872,
|
| 1001 |
+
"logits/rejected": -2.8993031978607178,
|
| 1002 |
+
"logps/chosen": -445.21142578125,
|
| 1003 |
+
"logps/rejected": -432.021240234375,
|
| 1004 |
+
"loss": 10.7741,
|
| 1005 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 1006 |
+
"rewards/chosen": -32.348819732666016,
|
| 1007 |
+
"rewards/margins": -1.2561819553375244,
|
| 1008 |
+
"rewards/rejected": -31.092632293701172,
|
| 1009 |
+
"step": 680
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 0.13,
|
| 1013 |
+
"learning_rate": 0.0004986509723258511,
|
| 1014 |
+
"logits/chosen": -3.1020426750183105,
|
| 1015 |
+
"logits/rejected": -3.133068561553955,
|
| 1016 |
+
"logps/chosen": -427.81756591796875,
|
| 1017 |
+
"logps/rejected": -416.2047424316406,
|
| 1018 |
+
"loss": 11.4242,
|
| 1019 |
+
"rewards/accuracies": 0.4375,
|
| 1020 |
+
"rewards/chosen": -31.230037689208984,
|
| 1021 |
+
"rewards/margins": -0.655289351940155,
|
| 1022 |
+
"rewards/rejected": -30.57474708557129,
|
| 1023 |
+
"step": 690
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.14,
|
| 1027 |
+
"learning_rate": 0.0004984702703514565,
|
| 1028 |
+
"logits/chosen": -3.0160446166992188,
|
| 1029 |
+
"logits/rejected": -3.0138049125671387,
|
| 1030 |
+
"logps/chosen": -433.4644470214844,
|
| 1031 |
+
"logps/rejected": -405.3623046875,
|
| 1032 |
+
"loss": 10.8165,
|
| 1033 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1034 |
+
"rewards/chosen": -30.836090087890625,
|
| 1035 |
+
"rewards/margins": -2.514590263366699,
|
| 1036 |
+
"rewards/rejected": -28.321496963500977,
|
| 1037 |
+
"step": 700
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 0.14,
|
| 1041 |
+
"learning_rate": 0.0004982782513290365,
|
| 1042 |
+
"logits/chosen": -3.1978919506073,
|
| 1043 |
+
"logits/rejected": -3.197380542755127,
|
| 1044 |
+
"logps/chosen": -422.511962890625,
|
| 1045 |
+
"logps/rejected": -402.30938720703125,
|
| 1046 |
+
"loss": 12.8189,
|
| 1047 |
+
"rewards/accuracies": 0.4375,
|
| 1048 |
+
"rewards/chosen": -30.785781860351562,
|
| 1049 |
+
"rewards/margins": -2.3963069915771484,
|
| 1050 |
+
"rewards/rejected": -28.389474868774414,
|
| 1051 |
+
"step": 710
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.14,
|
| 1055 |
+
"learning_rate": 0.0004980749240044603,
|
| 1056 |
+
"logits/chosen": -3.1342532634735107,
|
| 1057 |
+
"logits/rejected": -3.1338047981262207,
|
| 1058 |
+
"logps/chosen": -403.13494873046875,
|
| 1059 |
+
"logps/rejected": -357.790771484375,
|
| 1060 |
+
"loss": 11.5675,
|
| 1061 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1062 |
+
"rewards/chosen": -28.892202377319336,
|
| 1063 |
+
"rewards/margins": -3.723827362060547,
|
| 1064 |
+
"rewards/rejected": -25.168371200561523,
|
| 1065 |
+
"step": 720
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.14,
|
| 1069 |
+
"learning_rate": 0.0004978602976386554,
|
| 1070 |
+
"logits/chosen": -3.0739312171936035,
|
| 1071 |
+
"logits/rejected": -3.0738184452056885,
|
| 1072 |
+
"logps/chosen": -381.2265625,
|
| 1073 |
+
"logps/rejected": -378.4680480957031,
|
| 1074 |
+
"loss": 11.5225,
|
| 1075 |
+
"rewards/accuracies": 0.5,
|
| 1076 |
+
"rewards/chosen": -27.621994018554688,
|
| 1077 |
+
"rewards/margins": -0.9019744992256165,
|
| 1078 |
+
"rewards/rejected": -26.720022201538086,
|
| 1079 |
+
"step": 730
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.14,
|
| 1083 |
+
"learning_rate": 0.0004976343820071849,
|
| 1084 |
+
"logits/chosen": -3.166983127593994,
|
| 1085 |
+
"logits/rejected": -3.1671650409698486,
|
| 1086 |
+
"logps/chosen": -408.42071533203125,
|
| 1087 |
+
"logps/rejected": -387.2364196777344,
|
| 1088 |
+
"loss": 13.9818,
|
| 1089 |
+
"rewards/accuracies": 0.5,
|
| 1090 |
+
"rewards/chosen": -29.39678382873535,
|
| 1091 |
+
"rewards/margins": -2.448943614959717,
|
| 1092 |
+
"rewards/rejected": -26.94784164428711,
|
| 1093 |
+
"step": 740
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 0.14,
|
| 1097 |
+
"learning_rate": 0.0004973971873998035,
|
| 1098 |
+
"logits/chosen": -3.0561656951904297,
|
| 1099 |
+
"logits/rejected": -3.0557007789611816,
|
| 1100 |
+
"logps/chosen": -417.0025329589844,
|
| 1101 |
+
"logps/rejected": -349.56463623046875,
|
| 1102 |
+
"loss": 12.3073,
|
| 1103 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 1104 |
+
"rewards/chosen": -29.997058868408203,
|
| 1105 |
+
"rewards/margins": -5.650521278381348,
|
| 1106 |
+
"rewards/rejected": -24.346534729003906,
|
| 1107 |
+
"step": 750
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.15,
|
| 1111 |
+
"learning_rate": 0.0004971487246199875,
|
| 1112 |
+
"logits/chosen": -3.0265376567840576,
|
| 1113 |
+
"logits/rejected": -3.0265283584594727,
|
| 1114 |
+
"logps/chosen": -434.55419921875,
|
| 1115 |
+
"logps/rejected": -381.22808837890625,
|
| 1116 |
+
"loss": 12.0398,
|
| 1117 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1118 |
+
"rewards/chosen": -30.918407440185547,
|
| 1119 |
+
"rewards/margins": -4.258307456970215,
|
| 1120 |
+
"rewards/rejected": -26.66009521484375,
|
| 1121 |
+
"step": 760
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 0.15,
|
| 1125 |
+
"learning_rate": 0.000496889004984444,
|
| 1126 |
+
"logits/chosen": -2.8932366371154785,
|
| 1127 |
+
"logits/rejected": -2.895204544067383,
|
| 1128 |
+
"logps/chosen": -396.3167419433594,
|
| 1129 |
+
"logps/rejected": -428.53839111328125,
|
| 1130 |
+
"loss": 9.4104,
|
| 1131 |
+
"rewards/accuracies": 0.5625,
|
| 1132 |
+
"rewards/chosen": -28.4284610748291,
|
| 1133 |
+
"rewards/margins": 1.9928890466690063,
|
| 1134 |
+
"rewards/rejected": -30.42134666442871,
|
| 1135 |
+
"step": 770
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 0.15,
|
| 1139 |
+
"learning_rate": 0.0004966180403225946,
|
| 1140 |
+
"logits/chosen": -2.895068407058716,
|
| 1141 |
+
"logits/rejected": -2.894937753677368,
|
| 1142 |
+
"logps/chosen": -395.245849609375,
|
| 1143 |
+
"logps/rejected": -378.4429626464844,
|
| 1144 |
+
"loss": 10.2846,
|
| 1145 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 1146 |
+
"rewards/chosen": -28.237285614013672,
|
| 1147 |
+
"rewards/margins": -1.503316879272461,
|
| 1148 |
+
"rewards/rejected": -26.733972549438477,
|
| 1149 |
+
"step": 780
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.15,
|
| 1153 |
+
"learning_rate": 0.0004963358429760368,
|
| 1154 |
+
"logits/chosen": -2.551323652267456,
|
| 1155 |
+
"logits/rejected": -2.5523290634155273,
|
| 1156 |
+
"logps/chosen": -477.17327880859375,
|
| 1157 |
+
"logps/rejected": -427.474853515625,
|
| 1158 |
+
"loss": 12.1626,
|
| 1159 |
+
"rewards/accuracies": 0.42500001192092896,
|
| 1160 |
+
"rewards/chosen": -35.30048370361328,
|
| 1161 |
+
"rewards/margins": -4.13530158996582,
|
| 1162 |
+
"rewards/rejected": -31.165185928344727,
|
| 1163 |
+
"step": 790
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 0.15,
|
| 1167 |
+
"learning_rate": 0.0004960424257979822,
|
| 1168 |
+
"logits/chosen": -2.7914838790893555,
|
| 1169 |
+
"logits/rejected": -2.790367841720581,
|
| 1170 |
+
"logps/chosen": -478.9364318847656,
|
| 1171 |
+
"logps/rejected": -461.02655029296875,
|
| 1172 |
+
"loss": 10.3404,
|
| 1173 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 1174 |
+
"rewards/chosen": -35.25132369995117,
|
| 1175 |
+
"rewards/margins": -1.6512939929962158,
|
| 1176 |
+
"rewards/rejected": -33.60003662109375,
|
| 1177 |
+
"step": 800
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 0.15,
|
| 1181 |
+
"eval_logits/chosen": -3.0016679763793945,
|
| 1182 |
+
"eval_logits/rejected": -3.0014870166778564,
|
| 1183 |
+
"eval_logps/chosen": -441.87054443359375,
|
| 1184 |
+
"eval_logps/rejected": -399.2597961425781,
|
| 1185 |
+
"eval_loss": 11.33322811126709,
|
| 1186 |
+
"eval_rewards/accuracies": 0.4555903971195221,
|
| 1187 |
+
"eval_rewards/chosen": -32.331321716308594,
|
| 1188 |
+
"eval_rewards/margins": -3.850048303604126,
|
| 1189 |
+
"eval_rewards/rejected": -28.481277465820312,
|
| 1190 |
+
"eval_runtime": 1312.1317,
|
| 1191 |
+
"eval_samples_per_second": 0.729,
|
| 1192 |
+
"eval_steps_per_second": 0.729,
|
| 1193 |
+
"step": 800
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.16,
|
| 1197 |
+
"learning_rate": 0.0004957378021526705,
|
| 1198 |
+
"logits/chosen": -2.9228155612945557,
|
| 1199 |
+
"logits/rejected": -2.925412654876709,
|
| 1200 |
+
"logps/chosen": -481.1300354003906,
|
| 1201 |
+
"logps/rejected": -465.1952209472656,
|
| 1202 |
+
"loss": 11.3707,
|
| 1203 |
+
"rewards/accuracies": 0.4375,
|
| 1204 |
+
"rewards/chosen": -35.0019416809082,
|
| 1205 |
+
"rewards/margins": -1.9142730236053467,
|
| 1206 |
+
"rewards/rejected": -33.08766555786133,
|
| 1207 |
+
"step": 810
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.16,
|
| 1211 |
+
"learning_rate": 0.0004954219859147614,
|
| 1212 |
+
"logits/chosen": -3.0219039916992188,
|
| 1213 |
+
"logits/rejected": -3.0174221992492676,
|
| 1214 |
+
"logps/chosen": -364.8722839355469,
|
| 1215 |
+
"logps/rejected": -295.5256652832031,
|
| 1216 |
+
"loss": 12.3607,
|
| 1217 |
+
"rewards/accuracies": 0.38749998807907104,
|
| 1218 |
+
"rewards/chosen": -25.810903549194336,
|
| 1219 |
+
"rewards/margins": -5.978564262390137,
|
| 1220 |
+
"rewards/rejected": -19.832340240478516,
|
| 1221 |
+
"step": 820
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.16,
|
| 1225 |
+
"learning_rate": 0.0004950949914687023,
|
| 1226 |
+
"logits/chosen": -3.118417263031006,
|
| 1227 |
+
"logits/rejected": -3.1218113899230957,
|
| 1228 |
+
"logps/chosen": -469.4906311035156,
|
| 1229 |
+
"logps/rejected": -412.5103454589844,
|
| 1230 |
+
"loss": 11.9496,
|
| 1231 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 1232 |
+
"rewards/chosen": -33.402565002441406,
|
| 1233 |
+
"rewards/margins": -4.156603813171387,
|
| 1234 |
+
"rewards/rejected": -29.245960235595703,
|
| 1235 |
+
"step": 830
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.16,
|
| 1239 |
+
"learning_rate": 0.0004947568337080732,
|
| 1240 |
+
"logits/chosen": -3.0231635570526123,
|
| 1241 |
+
"logits/rejected": -3.0243794918060303,
|
| 1242 |
+
"logps/chosen": -382.8542785644531,
|
| 1243 |
+
"logps/rejected": -346.9595031738281,
|
| 1244 |
+
"loss": 9.7701,
|
| 1245 |
+
"rewards/accuracies": 0.4375,
|
| 1246 |
+
"rewards/chosen": -27.092443466186523,
|
| 1247 |
+
"rewards/margins": -3.154633045196533,
|
| 1248 |
+
"rewards/rejected": -23.93781089782715,
|
| 1249 |
+
"step": 840
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.16,
|
| 1253 |
+
"learning_rate": 0.0004944075280349084,
|
| 1254 |
+
"logits/chosen": -3.034963846206665,
|
| 1255 |
+
"logits/rejected": -3.0339653491973877,
|
| 1256 |
+
"logps/chosen": -385.3253173828125,
|
| 1257 |
+
"logps/rejected": -367.23638916015625,
|
| 1258 |
+
"loss": 9.2328,
|
| 1259 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 1260 |
+
"rewards/chosen": -27.111583709716797,
|
| 1261 |
+
"rewards/margins": -1.6716159582138062,
|
| 1262 |
+
"rewards/rejected": -25.43996810913086,
|
| 1263 |
+
"step": 850
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.17,
|
| 1267 |
+
"learning_rate": 0.0004940470903589948,
|
| 1268 |
+
"logits/chosen": -3.1586310863494873,
|
| 1269 |
+
"logits/rejected": -3.128281831741333,
|
| 1270 |
+
"logps/chosen": -487.345703125,
|
| 1271 |
+
"logps/rejected": -389.9637756347656,
|
| 1272 |
+
"loss": 12.7587,
|
| 1273 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 1274 |
+
"rewards/chosen": -34.73926544189453,
|
| 1275 |
+
"rewards/margins": -6.996462821960449,
|
| 1276 |
+
"rewards/rejected": -27.7428035736084,
|
| 1277 |
+
"step": 860
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.17,
|
| 1281 |
+
"learning_rate": 0.0004936755370971475,
|
| 1282 |
+
"logits/chosen": -2.9109790325164795,
|
| 1283 |
+
"logits/rejected": -2.888288974761963,
|
| 1284 |
+
"logps/chosen": -463.9207458496094,
|
| 1285 |
+
"logps/rejected": -376.2655334472656,
|
| 1286 |
+
"loss": 13.8618,
|
| 1287 |
+
"rewards/accuracies": 0.36250001192092896,
|
| 1288 |
+
"rewards/chosen": -34.083106994628906,
|
| 1289 |
+
"rewards/margins": -7.244679927825928,
|
| 1290 |
+
"rewards/rejected": -26.838430404663086,
|
| 1291 |
+
"step": 870
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.17,
|
| 1295 |
+
"learning_rate": 0.0004932928851724621,
|
| 1296 |
+
"logits/chosen": -2.8432798385620117,
|
| 1297 |
+
"logits/rejected": -2.8495278358459473,
|
| 1298 |
+
"logps/chosen": -365.90679931640625,
|
| 1299 |
+
"logps/rejected": -361.8638000488281,
|
| 1300 |
+
"loss": 8.7379,
|
| 1301 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1302 |
+
"rewards/chosen": -26.055978775024414,
|
| 1303 |
+
"rewards/margins": -0.3134794235229492,
|
| 1304 |
+
"rewards/rejected": -25.742502212524414,
|
| 1305 |
+
"step": 880
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.17,
|
| 1309 |
+
"learning_rate": 0.0004928991520135436,
|
| 1310 |
+
"logits/chosen": -2.6536898612976074,
|
| 1311 |
+
"logits/rejected": -2.6361289024353027,
|
| 1312 |
+
"logps/chosen": -519.7141723632812,
|
| 1313 |
+
"logps/rejected": -389.46575927734375,
|
| 1314 |
+
"loss": 15.3084,
|
| 1315 |
+
"rewards/accuracies": 0.38749998807907104,
|
| 1316 |
+
"rewards/chosen": -37.92496871948242,
|
| 1317 |
+
"rewards/margins": -10.107343673706055,
|
| 1318 |
+
"rewards/rejected": -27.817623138427734,
|
| 1319 |
+
"step": 890
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.17,
|
| 1323 |
+
"learning_rate": 0.0004924943555537128,
|
| 1324 |
+
"logits/chosen": -3.1115312576293945,
|
| 1325 |
+
"logits/rejected": -3.0791449546813965,
|
| 1326 |
+
"logps/chosen": -469.66351318359375,
|
| 1327 |
+
"logps/rejected": -401.14508056640625,
|
| 1328 |
+
"loss": 13.3952,
|
| 1329 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 1330 |
+
"rewards/chosen": -33.486183166503906,
|
| 1331 |
+
"rewards/margins": -5.449090480804443,
|
| 1332 |
+
"rewards/rejected": -28.037090301513672,
|
| 1333 |
+
"step": 900
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.18,
|
| 1337 |
+
"learning_rate": 0.0004920785142301893,
|
| 1338 |
+
"logits/chosen": -2.454453945159912,
|
| 1339 |
+
"logits/rejected": -2.4585988521575928,
|
| 1340 |
+
"logps/chosen": -463.24517822265625,
|
| 1341 |
+
"logps/rejected": -402.23773193359375,
|
| 1342 |
+
"loss": 13.7904,
|
| 1343 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1344 |
+
"rewards/chosen": -34.2692985534668,
|
| 1345 |
+
"rewards/margins": -5.860762596130371,
|
| 1346 |
+
"rewards/rejected": -28.40853500366211,
|
| 1347 |
+
"step": 910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.18,
|
| 1351 |
+
"learning_rate": 0.0004916516469832524,
|
| 1352 |
+
"logits/chosen": -2.8716561794281006,
|
| 1353 |
+
"logits/rejected": -2.8634109497070312,
|
| 1354 |
+
"logps/chosen": -347.35198974609375,
|
| 1355 |
+
"logps/rejected": -340.46942138671875,
|
| 1356 |
+
"loss": 7.8634,
|
| 1357 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1358 |
+
"rewards/chosen": -24.902542114257812,
|
| 1359 |
+
"rewards/margins": -1.0167404413223267,
|
| 1360 |
+
"rewards/rejected": -23.885799407958984,
|
| 1361 |
+
"step": 920
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.18,
|
| 1365 |
+
"learning_rate": 0.0004912137732553772,
|
| 1366 |
+
"logits/chosen": -3.2272415161132812,
|
| 1367 |
+
"logits/rejected": -3.2235121726989746,
|
| 1368 |
+
"logps/chosen": -450.973876953125,
|
| 1369 |
+
"logps/rejected": -462.0083923339844,
|
| 1370 |
+
"loss": 9.2953,
|
| 1371 |
+
"rewards/accuracies": 0.512499988079071,
|
| 1372 |
+
"rewards/chosen": -32.52794647216797,
|
| 1373 |
+
"rewards/margins": 0.38686689734458923,
|
| 1374 |
+
"rewards/rejected": -32.91481399536133,
|
| 1375 |
+
"step": 930
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.18,
|
| 1379 |
+
"learning_rate": 0.0004907649129903504,
|
| 1380 |
+
"logits/chosen": -2.647204875946045,
|
| 1381 |
+
"logits/rejected": -2.6482150554656982,
|
| 1382 |
+
"logps/chosen": -373.737060546875,
|
| 1383 |
+
"logps/rejected": -383.5912780761719,
|
| 1384 |
+
"loss": 7.0717,
|
| 1385 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 1386 |
+
"rewards/chosen": -26.5350284576416,
|
| 1387 |
+
"rewards/margins": 0.7499195337295532,
|
| 1388 |
+
"rewards/rejected": -27.284948348999023,
|
| 1389 |
+
"step": 940
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.18,
|
| 1393 |
+
"learning_rate": 0.0004903050866323608,
|
| 1394 |
+
"logits/chosen": -3.079465627670288,
|
| 1395 |
+
"logits/rejected": -3.0794615745544434,
|
| 1396 |
+
"logps/chosen": -395.6184997558594,
|
| 1397 |
+
"logps/rejected": -406.3051452636719,
|
| 1398 |
+
"loss": 9.1411,
|
| 1399 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 1400 |
+
"rewards/chosen": -29.57635498046875,
|
| 1401 |
+
"rewards/margins": 0.8788874745368958,
|
| 1402 |
+
"rewards/rejected": -30.45524024963379,
|
| 1403 |
+
"step": 950
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.19,
|
| 1407 |
+
"learning_rate": 0.000489834315125069,
|
| 1408 |
+
"logits/chosen": -3.1981568336486816,
|
| 1409 |
+
"logits/rejected": -3.1923341751098633,
|
| 1410 |
+
"logps/chosen": -453.5596618652344,
|
| 1411 |
+
"logps/rejected": -425.4774475097656,
|
| 1412 |
+
"loss": 11.2943,
|
| 1413 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1414 |
+
"rewards/chosen": -32.61115646362305,
|
| 1415 |
+
"rewards/margins": -2.69136118888855,
|
| 1416 |
+
"rewards/rejected": -29.9197998046875,
|
| 1417 |
+
"step": 960
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.19,
|
| 1421 |
+
"learning_rate": 0.0004893526199106531,
|
| 1422 |
+
"logits/chosen": -2.876206874847412,
|
| 1423 |
+
"logits/rejected": -2.881593942642212,
|
| 1424 |
+
"logps/chosen": -433.71636962890625,
|
| 1425 |
+
"logps/rejected": -391.15692138671875,
|
| 1426 |
+
"loss": 10.9992,
|
| 1427 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 1428 |
+
"rewards/chosen": -31.739410400390625,
|
| 1429 |
+
"rewards/margins": -3.6809983253479004,
|
| 1430 |
+
"rewards/rejected": -28.058406829833984,
|
| 1431 |
+
"step": 970
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.19,
|
| 1435 |
+
"learning_rate": 0.0004888600229288316,
|
| 1436 |
+
"logits/chosen": -2.865589141845703,
|
| 1437 |
+
"logits/rejected": -2.8664183616638184,
|
| 1438 |
+
"logps/chosen": -359.43023681640625,
|
| 1439 |
+
"logps/rejected": -321.70599365234375,
|
| 1440 |
+
"loss": 7.9915,
|
| 1441 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1442 |
+
"rewards/chosen": -25.906265258789062,
|
| 1443 |
+
"rewards/margins": -3.289003849029541,
|
| 1444 |
+
"rewards/rejected": -22.617259979248047,
|
| 1445 |
+
"step": 980
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.19,
|
| 1449 |
+
"learning_rate": 0.0004883565466158652,
|
| 1450 |
+
"logits/chosen": -2.8116517066955566,
|
| 1451 |
+
"logits/rejected": -2.782489776611328,
|
| 1452 |
+
"logps/chosen": -494.1553649902344,
|
| 1453 |
+
"logps/rejected": -430.222412109375,
|
| 1454 |
+
"loss": 13.6529,
|
| 1455 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1456 |
+
"rewards/chosen": -36.09266662597656,
|
| 1457 |
+
"rewards/margins": -5.284867286682129,
|
| 1458 |
+
"rewards/rejected": -30.807796478271484,
|
| 1459 |
+
"step": 990
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.19,
|
| 1463 |
+
"learning_rate": 0.0004878422139035341,
|
| 1464 |
+
"logits/chosen": -2.4114489555358887,
|
| 1465 |
+
"logits/rejected": -2.377622604370117,
|
| 1466 |
+
"logps/chosen": -482.96856689453125,
|
| 1467 |
+
"logps/rejected": -437.3125,
|
| 1468 |
+
"loss": 11.137,
|
| 1469 |
+
"rewards/accuracies": 0.5,
|
| 1470 |
+
"rewards/chosen": -34.935813903808594,
|
| 1471 |
+
"rewards/margins": -3.0001707077026367,
|
| 1472 |
+
"rewards/rejected": -31.93564224243164,
|
| 1473 |
+
"step": 1000
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.19,
|
| 1477 |
+
"eval_logits/chosen": -2.2959094047546387,
|
| 1478 |
+
"eval_logits/rejected": -2.2838947772979736,
|
| 1479 |
+
"eval_logps/chosen": -438.8805236816406,
|
| 1480 |
+
"eval_logps/rejected": -399.5718688964844,
|
| 1481 |
+
"eval_loss": 10.402800559997559,
|
| 1482 |
+
"eval_rewards/accuracies": 0.45036572217941284,
|
| 1483 |
+
"eval_rewards/chosen": -32.032310485839844,
|
| 1484 |
+
"eval_rewards/margins": -3.519833564758301,
|
| 1485 |
+
"eval_rewards/rejected": -28.51247787475586,
|
| 1486 |
+
"eval_runtime": 1313.6418,
|
| 1487 |
+
"eval_samples_per_second": 0.729,
|
| 1488 |
+
"eval_steps_per_second": 0.729,
|
| 1489 |
+
"step": 1000
|
| 1490 |
}
|
| 1491 |
],
|
| 1492 |
"logging_steps": 10,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5958bf8104a088a450e0b4310623d9428611ebdb5fcbb92ab0f567ab6883bfbd
|
| 3 |
size 4091
|