[Core][TPU] Support TPU Data Parallalism

[wenxindongwork]

This pull request introduces a small but important update to the data parallel initialization logic in vllm/entrypoints/llm.py. The change ensures that the warning and error for unsupported single-process data parallel usage are not triggered when running on TPU platforms.

TPU DP support added in vllm-project/tpu-inference#865

• 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.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft in the Google Doc.

[gemini-code-assist]

Code Review

This pull request introduces a get_dynamic_token_budget method to the Scheduler class, providing a hook for subclasses to implement dynamic scheduling constraints. This is a good extension point. The new method is then used in the schedule method to constrain the number of tokens for both running and waiting requests.

My main feedback is to simplify the newly added logic. The three lines added in two places to apply the dynamic budget can be condensed into a single line. This will make the code more concise and avoid duplication, improving maintainability.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[yaochengji]

LGTM!

[yaochengji]

@njhill could you please take a look at the small change?