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

What is the function 'patch_attention' for? #20

Open
mxjmtxrm opened this issue Oct 11, 2024 · 1 comment
Open

What is the function 'patch_attention' for? #20

mxjmtxrm opened this issue Oct 11, 2024 · 1 comment

Comments

@mxjmtxrm
Copy link

As title. Why can't we use the original forward function?
@synxlin
Copy link
Contributor

synxlin commented Nov 8, 2024

Hi,
patch_attention is used to help perform fake quantization on post-RoPE KV cache. We apply KV cache quantization by registering a forward hook on RotaryEmbedding module.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@synxlin @mxjmtxrm