Skip to content
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.

Sign up for GitHub

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

Jump to bottom

q.half() issue in fast_model forward pass #108

Open
lucapericlp opened this issue Mar 24, 2024 · 0 comments
Open

q.half() issue in fast_model forward pass #108

lucapericlp opened this issue Mar 24, 2024 · 0 comments
Labels
bug Something isn't working

Comments

@lucapericlp
Copy link
Contributor

lucapericlp commented Mar 24, 2024 •
edited 

Closing this issue. 

I was able to move ahead by making the query from type float to type half, by adding the line `q = q.half()` on line 221 in `fam/llm/fast_model.py`. Not sure how this would impact the performance, as of now.

I am able to use `tts` object from `fast_inference.py` in the interactive session now.

Thanks !!

Originally posted by @saurabh-kadian in #106 (comment)

We're seeing the same issue when running from the colab on a non-GPU runtime (could only get the colab working via that)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
bug Something isn't working
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@lucapericlp