[#12257][fix] Use the first non-None result returned by hf download workers

[kev-bi]

Summary by CodeRabbit

• Bug Fixes
  • Enhanced robustness of model directory selection in distributed model download operations.

Description

This PR addresses the issue reported in #12257.

There seems to be an edge case when running the TensorRT-LLM backend in dynamo where model_dirs[0] does not hold the rank-0 result. So rather than always use the model_dirs[0] result change the code to iterate through non-None values in model_dirs and use the first is encountered. Since only the rank-0 worker is still downloading the model and all other workers will return None there should be only one non-None value in model_dirs to use. Now we should handle both the assumed case where model_dirs[0] does hold the result and the edge case where some other entry in the list holds the result.

Test Coverage

The existing tests should cover, the model_dir is still fetched and assigned, it's just that the manner it is fetched is different. Let me know if it would be preferable to add a test where the edge case is triggered.

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[coderabbitai]

📝 Walkthrough

Walkthrough

Enhanced robustness in model directory selection by filtering out None values when distributing models across nodes, and simplified the accompanying docstring for clarity.

Changes

Cohort / File(s)	Summary
Model directory selection robustness tensorrt_llm/llmapi/llm_utils.py	Updated _download_hf_model_if_needed to select the first non-None model directory instead of always using the first element, preventing potential None value handling issues. Simplified docstring by removing rank-specific note.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~5 minutes

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the specific fix: using the first non-None result from hf download workers instead of assuming the first element. It's concise, references the issue #12257, and accurately reflects the main change.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Description check	✅ Passed	The PR description adequately explains the issue, solution, test coverage, and includes all required checklist items from the template.

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

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

📝 Coding Plan

• Generate coding plan for human review comments

---

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

Tip

CodeRabbit can enforce grammar and style rules using `languagetool`.

Configure the reviews.tools.languagetool setting to enable/disable rules and categories. Refer to the LanguageTool Community to learn more.

[Superjomn]

LGTM, thanks for the contribution.

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39724 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39724 [ run ] completed with state SUCCESS. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30920 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39739 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39739 [ run ] completed with state SUCCESS. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30934 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39742 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39742 [ run ] completed with state FAILURE. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30937 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39745 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39745 [ run ] completed with state FAILURE. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30940 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39748 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39748 [ run ] completed with state FAILURE. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30943 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39752 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39752 [ run ] completed with state FAILURE. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30947 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39755 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #39755 [ run ] completed with state FAILURE. Commit: 233d8f9
/LLM/main/L0_MergeRequest_PR pipeline #30950 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #39758 [ run ] triggered by Bot. Commit: 233d8f9 Link to invocation

[tensorrt-cicd]

PR_Github #41656 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32561 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #41660 [ run ] triggered by Bot. Commit: bc28949 Link to invocation

[tensorrt-cicd]

PR_Github #41660 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32565 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #41662 [ run ] triggered by Bot. Commit: bc28949 Link to invocation

[tensorrt-cicd]

PR_Github #41662 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32567 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #41669 [ run ] triggered by Bot. Commit: bc28949 Link to invocation

[tensorrt-cicd]

PR_Github #41669 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32574 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #41676 [ run ] triggered by Bot. Commit: bc28949 Link to invocation

[tensorrt-cicd]

PR_Github #41676 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32580 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[Superjomn]

/bot run

[tensorrt-cicd]

PR_Github #41682 [ run ] triggered by Bot. Commit: bc28949 Link to invocation

[tensorrt-cicd]

PR_Github #41682 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32584 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[karljang]

/bot run

[tensorrt-cicd]

PR_Github #41716 [ run ] triggered by Bot. Commit: bc28949 Link to invocation

[tensorrt-cicd]

PR_Github #41716 [ run ] completed with state SUCCESS. Commit: bc28949
/LLM/main/L0_MergeRequest_PR pipeline #32618 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[karljang]

/bot run

[tensorrt-cicd]

PR_Github #41720 [ run ] triggered by Bot. Commit: c1a440e Link to invocation

[tensorrt-cicd]

PR_Github #41720 [ run ] completed with state SUCCESS. Commit: c1a440e
/LLM/main/L0_MergeRequest_PR pipeline #32622 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[karljang]

/bot run

[tensorrt-cicd]

PR_Github #41756 [ run ] triggered by Bot. Commit: c1a440e Link to invocation

[tensorrt-cicd]

PR_Github #41756 [ run ] completed with state SUCCESS. Commit: c1a440e
/LLM/main/L0_MergeRequest_PR pipeline #32656 completed with status: 'SUCCESS'

CI Report

Link to invocation