-
Notifications
You must be signed in to change notification settings - Fork 81
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Revise SFT File #793
base: main
Are you sure you want to change the base?
Revise SFT File #793
Conversation
src/levanter/main/sft.py
Outdated
|
||
|
||
@dataclass | ||
class SFTConfig(TrainLmConfig): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
imho this should really not inherit from TrainLmConfig but just grab the parts you need
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
don't want to do this one?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I end up needing a lot of shared content with TrainLMConfig since I want to explicitly declare all variables in the marin experiment so i think it's better to inherit
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
generally speaking it's a code smell when you inherit and don't use the stuff in the super
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Got it, I just changed it to not inherit and copied the attributes I need
PR to revise SFT file to avoid breaking changes to marin and for a request from @dlwh