Upgrade xgrammar to 0.1.23

[russellb]

• Bump xgrammar to 0.1.23
• Revert "[V1] Resolve failed concurrent structured output requests ([V1] Resolve failed concurrent structred output requests #19565)"

Upgrade Xgrammar to 0.1.23, which includes some important fixes which allow us
to revert the workaround introduced by #19565.

[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]

Code Review

This pull request upgrades xgrammar to version 0.1.23 and reverts a workaround related to concurrent structured output requests. The changes are straightforward: the dependency version is updated in requirements/common.txt, and in vllm/v1/worker/gpu_model_runner.py, the code is simplified by removing the forced use of a specific xgrammar kernel implementation. This is done under the assumption that the new xgrammar version fixes the underlying issue, as stated in the pull request description. The code changes are consistent with this goal and appear correct. I have no further comments.

[Jialin]

Sweet!