Skip to content

fix(omni): Preserve caller-provided runtime device mapping (e.g. "0,1")#9034

Merged
rmccorm4 merged 9 commits into
ai-dynamo:mainfrom
tthakkal:preserve_runtime-device-mapping
May 11, 2026
Merged

fix(omni): Preserve caller-provided runtime device mapping (e.g. "0,1")#9034
rmccorm4 merged 9 commits into
ai-dynamo:mainfrom
tthakkal:preserve_runtime-device-mapping

Conversation

@tthakkal

@tthakkal tthakkal commented May 1, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Overview:

Preserve caller-provided runtime device mapping (e.g. "0,1,2,3")
so tensor-parallel workers can initialize on all visible devices.
Fallback to "0" only when devices are not specified.

Details:

Modified the _stage_config_to_dict function to conditionally set the "devices" field to "0" only when absent, instead of forcibly overwriting any preconfigured runtime device selection.

Where should the reviewer start?

components/src/dynamo/vllm/omni/stage_worker.py

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

  • closes GitHub issue: #xxx
Open in Devin Review

Summary by CodeRabbit

Bug Fixes

  • Fixed runtime device configuration to preserve user-defined settings instead of being overwritten with default values, allowing custom device selections to persist correctly.

@tthakkal
Update stage_worker.py
0009096 
Signed-off-by: Thanaji Rao Thakkalapelli <thanaji.rao.thakkalapelli@intel.com>
@tthakkal tthakkal requested review from a team as code owners May 1, 2026 22:03
@copy-pr-bot

copy-pr-bot Bot commented May 1, 2026

Copy link
Copy Markdown

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@github-actions

github-actions Bot commented May 1, 2026

Copy link
Copy Markdown
Contributor

👋 Hi tthakkal! Thank you for contributing to ai-dynamo/dynamo.

Just a reminder: The NVIDIA Test Github Validation CI runs an essential subset of the testing framework to quickly catch errors.Your PR reviewers may elect to test the changes comprehensively before approving your changes.

🚀

@github-actions github-actions Bot added backend::vllm Relates to the vllm backend multimodal fix external-contribution Pull request is from an external contributor labels May 1, 2026
@coderabbitai

coderabbitai Bot commented May 1, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Walkthrough

Modified the _stage_config_to_dict function to conditionally set the "devices" field to "0" only when absent, instead of forcibly overwriting any preconfigured runtime device selection.

Changes

Cohort / File(s) Summary
Runtime Device Configuration
components/src/dynamo/vllm/omni/stage_worker.py		 Updated _stage_config_to_dict to preserve existing "devices" field values, setting it to "0" only if absent from the stage's runtime config.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 5
✅ Passed checks (5 passed)
Check name Status Explanation
Title check ✅ Passed The title accurately describes the main change: preserving caller-provided runtime device mapping instead of overwriting it.
Docstring Coverage ✅ Passed Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.
Description check ✅ Passed The pull request description includes all required template sections with substantive content: overview, details, reviewer starting point, and related issues placeholder.

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

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

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

@rmccorm4

rmccorm4 commented May 5, 2026

Copy link
Copy Markdown
Contributor

/ok to test c903a67

furionw
furionw approved these changes May 7, 2026
View reviewed changes

@furionw furionw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you

@rmccorm4 rmccorm4 enabled auto-merge (squash) May 11, 2026 05:14
@rmccorm4

rmccorm4 commented May 11, 2026

Copy link
Copy Markdown
Contributor

/ok to test db072f7

@rmccorm4 rmccorm4 merged commit 67dcd6e into ai-dynamo:main May 11, 2026
80 of 81 checks passed
@ptarasiewiczNV ptarasiewiczNV mentioned this pull request May 11, 2026
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@furionw furionw furionw approved these changes

@dynamo-ops dynamo-ops dynamo-ops approved these changes

Assignees

No one assigned

Labels

backend::vllm Relates to the vllm backend external-contribution Pull request is from an external contributor fix multimodal size/XS

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@tthakkal @rmccorm4 @furionw @dynamo-ops