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.

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

Add blocked flash decode kernel #380

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

harsh-nod
Copy link
Contributor

This PR adds a version of the decode attention kernel that only reads chunks of a fixed size when reading the key and value tensors.

@harsh-nod harsh-nod force-pushed the blocked_page_attn branch 5 times, most recently from 7a4ef8f to 1078110 Compare January 12, 2025 20:52
This PR adds a version of the decode attention kernel
that only reads chunks of a fixed size when reading the
key and value tensors.

Signed-off-by: Harsh Menon <[email protected]>
@harsh-nod harsh-nod requested review from martin-luecke, raikonenfnu and Hardcode84 and removed request for martin-luecke January 13, 2025 17:12
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant