Skip to content

[BugFix][AMD] Compatible patch for AITER lib after 04/20 #17912

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

Merged

Conversation

qli88
Copy link
Contributor

@qli88 qli88 commented May 9, 2025

  1. Changes to adapt new AITER MLA API (MTP is not enabled)
  2. Some other bug fixes.

This PR is for AITER API changes introduced by commit 939f741fc37f46694e48c32c7164f49eae2584c4 (merged on 04/20/2025). AITER versions after this commit require this patch to work.

Copy link

github-actions bot commented May 9, 2025

👋 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 fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

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 ready label to the PR or enable auto-merge.

🚀

Copy link
Collaborator

@houseroad houseroad left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Overall, looks reasonable. Could you paste your test plan/results to the PR description?

@houseroad houseroad added the ready ONLY add when PR is ready to merge/full CI is needed label May 12, 2025
@tjtanaa
Copy link
Contributor

tjtanaa commented May 13, 2025

@hongxiayang after merging this PR, is there any plan to expedite the checking of existing AITER kernels and fixes for them to work on this new AITER commit? and especially the MLA V1.

For documentation purposes: I have quoted the AITER commit mentioned in previous PR (#17864)
AITER commit used here is : c1debd87ce0391aa27438d9e07e76e4fea7c4b70

@DarkLight1337
Copy link
Member

I'm able to run deepseekV2-lite-chat on NVIDIA GPUs in V0 mode on this branch. So we just need to verify AMD side.

@DarkLight1337 DarkLight1337 requested a review from hongxiayang May 13, 2025 07:44
@tjtanaa
Copy link
Contributor

tjtanaa commented May 13, 2025

Validated the PR, it works.

Moreover, at this point in time, the integrated AITER kernels are working with this new AITER commit. The ones that have issues in new AITER commits are:

  • HipLoad fmoe_fp8_g1u1_subGU_512_silu_tkw1.co
  • AITER MLA V1

Copy link
Collaborator

@houseroad houseroad left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good.

@houseroad houseroad merged commit 4f8b373 into vllm-project:main May 14, 2025
60 checks passed
aarnphm pushed a commit to aarnphm/vllm that referenced this pull request May 15, 2025
gshtras pushed a commit to ROCm/vllm that referenced this pull request May 15, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ready ONLY add when PR is ready to merge/full CI is needed
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants