-
-
Notifications
You must be signed in to change notification settings - Fork 5.9k
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
[WIP][Kernel] Flashinfer MLA support #13630
base: main
Are you sure you want to change the base?
[WIP][Kernel] Flashinfer MLA support #13630
Conversation
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add 🚀 |
Hi @LucasWilkinson , thank you for the great work! Would you mind trying the https://github.com/flashinfer-ai/flashinfer/releases/tag/v0.2.2 where we fixed some potential bugs for cudagraph and also add a new template for hopper (flashinfer-ai/flashinfer#887). |
This pull request has merge conflicts that must be resolved before it can be |
Signed-off-by: Lucas Wilkinson <[email protected]> wip flash-infer Signed-off-by: Lucas Wilkinson <[email protected]> wip debugging Signed-off-by: Lucas Wilkinson <[email protected]>
1df44c3
to
92a8bbb
Compare
Signed-off-by: Lucas Wilkinson <[email protected]>
Thank you for your amazing work, big fan of flashinfer! Very quickly tested it and im still having coda-graph issues but ill definitely play with it more tmrw and keep you up to date here 👍 |
This pull request has merge conflicts that must be resolved before it can be |
CUDA graphs still broken
Based off: #12639