chore: Update vLLM compilation config to vLLM v0.14.1

[ptarasiewiczNV]

Overview:

Update compilation config to follow changes made in vllm-project/vllm#29646

Details:

Where should the reviewer start?

Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)

• closes GitHub issue: #xxx

Summary by CodeRabbit

• Chores
  • Updated VLLM compilation configuration to enhance performance and efficiency in deployment settings.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Walkthrough

VLLM compilation configuration flags in the deploy_hopper_16gpu.yaml file were updated. The previous flags (enable_fusion, enable_attn_fusion, enable_noop, and custom_ops with rms_norm) were replaced with fuse_norm_quant and eliminate_noops flags while preserving cudagraph_mode.

Changes

Cohort / File(s)	Summary
VLLM Configuration recipes/deepseek-r1/vllm/disagg/deploy_hopper_16gpu.yaml	Updated compilation configuration flags by replacing enable_fusion, enable_attn_fusion, enable_noop, and custom_ops entries with fuse_norm_quant: true and eliminate_noops: true; cudagraph_mode retained.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Poem

🐰 Hops of joy for configs new,
Flags optimized, tried and true!
Norm and quant in harmony,
Hoppy times for DeepSeek spree! ✨

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description is incomplete. While the Overview section is filled with relevant context, the Details and Where should the reviewer start sections are empty, lacking explanation of what changed and which files need review.	Complete the Details section by describing the specific compilation config changes (fuse_norm_quant and eliminate_noops flags), and specify deploy_hopper_16gpu.yaml in the reviewer focus section.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and specifically describes the main change: updating vLLM compilation config to a specific version (v0.14.1), which aligns with the actual changes made to the deploy_hopper_16gpu.yaml file.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

Tip

🧪 Unit Test Generation v2 is now available!

We have significantly improved our unit test generation capabilities.

To enable: Add this to your .coderabbit.yaml configuration:

reviews:
  finishing_touches:
    unit_tests:
      enabled: true

Try it out by using the @coderabbitai generate unit tests command on your code files or under ✨ Finishing Touches on the walkthrough!

Have feedback? Share your thoughts on our Discord thread!

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}