You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have searched the issues and there is not a similar one.
I confirm that this is not a bug report, a feature request, or a badcase.
Description
Hi, I checked that Qwen2.5-72B-Instruct got high score in IFEval 86.6. Personally, I finetuned Qwen2 base model using high quaility datasets to get Instruct model many times. However after testing IFEval benchmark, I realized score is pretty low 56.3. Any suggestions how to improve this ability in the model? Which finetuned method did you use to improve it?
This discussion was converted from issue #946 on September 25, 2024 11:35.
Heading
Bold
Italic
Quote
Code
Link
Numbered list
Unordered list
Task list
Attach files
Mention
Reference
Menu
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Has this been raised before?
Description
Hi, I checked that Qwen2.5-72B-Instruct got high score in IFEval 86.6. Personally, I finetuned Qwen2 base model using high quaility datasets to get Instruct model many times. However after testing IFEval benchmark, I realized score is pretty low 56.3. Any suggestions how to improve this ability in the model? Which finetuned method did you use to improve it?
Beta Was this translation helpful? Give feedback.
All reactions