Skip to content

[BugFix] [Spec Decode] Remove LlamaForCausalLMEagle3 to fix CI #22611

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
merged 8 commits into from
Aug 11, 2025

Conversation

22quinn
Copy link
Collaborator

@22quinn 22quinn commented Aug 11, 2025

Essential Elements of an Effective PR Description Checklist

  • The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
  • The test plan, such as providing test command.
  • The test results, such as pasting the results comparison before and after, or e2e results
  • (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.

Purpose

"Basic Models Test" has been failing due to some eagle models not named with "eagle-" or "eagle3-" prefix.

This is a short-term fix to make CI green. In the long-term, maybe eagle models should use model_type instead of the name to identify itself, similar to medusa models. But that requires updating config.json in HF repo cc @Liyuhui-12 @LiuXiaoxuanPKU

Test Plan

pytest tests/models/test_initialization.py

Test Result

passed

(Optional) Documentation Update

Copy link

👋 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.

🚀

@mergify mergify bot added the llama Related to Llama models label Aug 11, 2025
Copy link
Contributor

Warning

Gemini encountered an error creating the review. You can try again by commenting /gemini review.

@22quinn 22quinn changed the title [BugFix] Fix initialization for LlamaForCausalLMEagle3 [BugFix] [Spec Decode] Fix initialization for LlamaForCausalLMEagle3 Aug 11, 2025
@22quinn 22quinn added speculative-decoding ready ONLY add when PR is ready to merge/full CI is needed ci-failure Issue about an unexpected test failure in CI labels Aug 11, 2025
@DarkLight1337
Copy link
Member

The test is still failing, PTAL

@22quinn 22quinn changed the title [BugFix] [Spec Decode] Fix initialization for LlamaForCausalLMEagle3 [BugFix] [Spec Decode] Remove LlamaForCausalLMEagle3 to fix CI Aug 11, 2025
@DarkLight1337 DarkLight1337 enabled auto-merge (squash) August 11, 2025 07:52
@22quinn
Copy link
Collaborator Author

22quinn commented Aug 11, 2025

@DarkLight1337 sorry this becomes more involved than I expected, so I removed it to unblock CI.
I got more errors like ValueError: Duplicate layer name: model.layers.1.self_attn.attn. I'll leave it to the original author @lec77 to sort it out.

@hmellor
Copy link
Member

hmellor commented Aug 11, 2025

Should we comment out the model instead of deleting it? (Makes it easier for future maintainers to know what to add back)

Signed-off-by: 22quinn <[email protected]>
Signed-off-by: 22quinn <[email protected]>
auto-merge was automatically disabled August 11, 2025 08:58

Head branch was pushed to by a user without write access

@mergify mergify bot added new-model Requests to new models v1 labels Aug 11, 2025
@DarkLight1337
Copy link
Member

@22quinn can you address @hmellor 's comment?

Signed-off-by: 22quinn <[email protected]>
Signed-off-by: 22quinn <[email protected]>
@vllm-bot vllm-bot merged commit 807d21b into vllm-project:main Aug 11, 2025
5 of 10 checks passed
jingyu-ml pushed a commit to jingyu-ml/vllm that referenced this pull request Aug 11, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ci-failure Issue about an unexpected test failure in CI llama Related to Llama models new-model Requests to new models ready ONLY add when PR is ready to merge/full CI is needed speculative-decoding v1
Projects
Status: Done
Development

Successfully merging this pull request may close these issues.

4 participants