-
Notifications
You must be signed in to change notification settings - Fork 26.1k
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
[Gemma2
] Support FA2 softcapping
#31887
Conversation
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
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.
OK, looks good to me! 2.6.0 was released 3 hours ago, let's go
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.
LGTM - thanks for adding!
Good to see this. Can we use it for model fine-tuning, or is it just for inference? Google recommends fine-tuning in 'eager' mode. |
Now you can use it for finetuning as well if you have the correct version of FA2. Not sure if finetuning "requires" it |
Great! Any plans for sdpa support as well? |
Sdpa is a bit more complicated, we need to use flex attention, did not have time to implement. Do you want to open a PR? |
Hi @ArthurZucker, should we also add the sliding window and soft-capping to transformers/src/transformers/models/gemma2/modeling_gemma2.py Lines 466 to 469 in fc35907
just like transformers/src/transformers/models/mistral/modeling_mistral.py Lines 519 to 528 in fc35907
|
It should be here on main: https://github.com/huggingface/transformers/blob/main/src/transformers/models/gemma2/modeling_gemma2.py#L361 we updated the whole FA2 integration . On the release branch it was there AFAIK |
Get it, thanks for replying! |
* Support softcapping * strictly greater than * update
* Support softcapping * strictly greater than * update
* Support softcapping * strictly greater than * update
What does this PR do?
Adds support for the new FA2 softcapping following Dao-AILab/flash-attention#1025