Skip to main content
oaaic
Projects
openhermes-dpo
Reports
DPOpenHermes
Log in
Sign up
Share
Comment
Star
Share
Comment
Star
DPOpenHermes
Wing Lian
Created on December 2
|
Last edited on December 2
Comment
Section 1
train/logps/chosen, train/logps/rejected
train/logps/chosen, train/logps/rejected
0
500
1k
1.5k
2k
2.5k
Step
-600
-500
-400
-300
-200
-100
train/logps/chosen
train/logps/chosen
0
500
1k
1.5k
2k
2.5k
Step
-600
-500
-400
-300
-200
-100
train/logps/rejected
train/logps/rejected
0
500
1k
1.5k
2k
2.5k
Step
-600
-500
-400
-300
-200
Run: DPO intel x argilla
1
Run: DPO intel x argilla
1
Run: DPO intel x argilla
1
Add a comment