-
-
Notifications
You must be signed in to change notification settings - Fork 9.6k
update flashinfer to v0.2.9rc1 #21485
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
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 🚀 |
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.
Code Review
This pull request updates the FlashInfer dependency to version v0.2.9rc1. The code changes correctly adapt the calls to trtllm_batch_decode_with_kv_cache
to match the updated API in the new version. The changes are consistent across the affected files and appear to be correct.
Signed-off-by: Weiliang Liu <[email protected]>
aeec6ab
to
6289473
Compare
This pull request has merge conflicts that must be resolved before it can be |
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.
Aren't there other updates needed for the latest release? I remember the flashinfer fp4 moe utils also have some changes cc @wenscarl
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.
#21408 is already merged. No more changes needed. |
Signed-off-by: Weiliang Liu <[email protected]> Signed-off-by: shuw <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]> Signed-off-by: x22x22 <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]> Signed-off-by: Jinzhen Lin <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]> Signed-off-by: Paul Pak <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]> Signed-off-by: Boyuan Feng <[email protected]>
Signed-off-by: Weiliang Liu <[email protected]> Signed-off-by: Diego-Castan <[email protected]>
Essential Elements of an Effective PR Description Checklist
supported_models.md
andexamples
for a new model.Purpose
updata flashinfer and modify it's trtllm-gen call to use latest API.
Test Plan
test llama4 with lm_eval
Test Result
(Optional) Documentation Update