[Chore] dedup logs

[aarnphm]

This PR dedups triton logs given that we are already warning about this during eager import check.

Also moves the custom all-reduce warning to debug, given that this is not useful for end-users (only for debugging purposes here.)

Signed-off-by: Aaron Pham contact@aarnphm.xyz

[github-actions]

👋 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.

🚀

[gemini-code-assist]

Summary of Changes

Hello @aarnphm, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request focuses on refining the logging output to improve clarity and reduce noise. My aim is to make the logs more concise and relevant by removing redundant warnings and adjusting the verbosity of certain messages, ensuring that end-users only see information critical to their operation while detailed debugging information is available at a lower log level.

Highlights

• Log Verbosity Adjustment: I've adjusted the log level for the custom all-reduce kernel message in vllm/config.py from INFO to DEBUG. This change makes the message less prominent for end-users, as it's primarily useful for debugging purposes.
• Log Deduplication: I've removed a redundant warning_once log message about Triton not being installed from vllm/triton_utils/importing.py. This warning is already handled by an eager import check, so removing it helps deduplicate log output.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This PR effectively dedups logs and moves a custom all-reduce warning to debug, improving log clarity and reducing noise for end-users. The changes are well-justified and contribute to a better debugging experience.

[gemini-code-assist]

Moving this to debug is good since it's not for end-users. However, consider adding more context to the debug message to aid in debugging, such as the specific platform or reason for disabling.

Suggested change

	logger.debug(
	"Disabled the custom all-reduce kernel because it is not "
	"supported on current platform.")
	logger.debug(
	"Disabled the custom all-reduce kernel because it is not "
	"supported on current platform: {}".format(current_platform.__class__.__name__))