[https://nvbugs/6273846][test] gate GPT-OSS TRTLLM Gen MoE tests to SM100/SM103

[dongfengy]

Summary by CodeRabbit

Tests

• Refined test skip conditions for TRTLLM MoE backend to be more specific to GPU hardware capabilities, improving test accuracy based on compute capability versions.

Description

Test Coverage

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)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• 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]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 61775331-95ef-4c8a-8f0b-9f103a1adac2

📥 Commits

Reviewing files that changed from the base of the PR and between b222246 and c5a66a5.

📒 Files selected for processing (1)

• tests/integration/defs/accuracy/test_llm_api_pytorch.py

---

📝 Walkthrough

Walkthrough

This PR refines the skip condition for TRTLLM MoE backend tests in TestGPTOSS. A new class-level skip_no_trtllm_gen_moe_support marker checks for specific SM versions (100, 103) and is applied consistently across multiple moe_backend parametrizations.

Changes

TRTLLM MoE Skip Condition Refinement

Layer / File(s)	Summary
Skip condition for TRTLLM MoE support tests/integration/defs/accuracy/test_llm_api_pytorch.py	Introduces skip_no_trtllm_gen_moe_support pytest skipif marker that checks SM version membership, replacing device-level preconditions for TRTLLM MoE backend support.
Apply skip marker to remaining MoE backend parametrizations tests/integration/defs/accuracy/test_llm_api_pytorch.py	Updates five test methods to use the new SM version-specific skip marker in their moe_backend parametrizations.

Possibly Related PRs

• NVIDIA/TensorRT-LLM#14884: Both PRs affect tests/integration/defs/accuracy/test_llm_api_pytorch.py::TestGPTOSS test execution by changing skip/waiver behavior for specific GPT-OSS test cases.

Suggested Reviewers

• jieli-matrix

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description only contains the template with minimal customization; the Description, Test Coverage, and checklist sections are empty or incomplete, lacking substantive explanation of the issue and solution.	Complete the Description section explaining the motivation and solution, and fill in the Test Coverage section listing relevant tests that safeguard the changes.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the main change: gating GPT-OSS TRTLLM Gen MoE tests to SM100/SM103, which directly matches the raw summary describing the skip condition refinement for those specific SM versions.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
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.

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

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[longlee0622]

/bot run

[tensorrt-cicd]

PR_Github #53023 [ run ] triggered by Bot. Commit: a1e8b5b Link to invocation

[tensorrt-cicd]

PR_Github #53023 [ run ] completed with state SUCCESS. Commit: a1e8b5b
/LLM/main/L0_MergeRequest_PR pipeline #42246 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

CI Agent Failure Analysis

Link to invocation

[longlee0622]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #53076 [ run ] triggered by Bot. Commit: a1e8b5b Link to invocation

[tensorrt-cicd]

PR_Github #53076 [ run ] completed with state FAILURE. Commit: a1e8b5b
/LLM/main/L0_MergeRequest_PR pipeline #42288 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

CI Agent Failure Analysis

Link to invocation

[longlee0622]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #53257 [ run ] triggered by Bot. Commit: a1e8b5b Link to invocation

[tensorrt-cicd]

PR_Github #53257 [ run ] completed with state SUCCESS. Commit: a1e8b5b
/LLM/main/L0_MergeRequest_PR pipeline #42451 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

CI Agent Failure Analysis

Link to invocation

[longlee0622]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #53400 [ run ] triggered by Bot. Commit: a1e8b5b Link to invocation

[tensorrt-cicd]

PR_Github #53400 [ run ] completed with state FAILURE. Commit: a1e8b5b
/LLM/main/L0_MergeRequest_PR pipeline #42573 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

CI Agent Failure Analysis

Link to invocation

[longlee0622]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #53543 [ run ] triggered by Bot. Commit: a1e8b5b Link to invocation

[tensorrt-cicd]

PR_Github #53543 [ run ] completed with state FAILURE. Commit: a1e8b5b
/LLM/main/L0_MergeRequest_PR pipeline #42694 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

CI Agent Failure Analysis

Link to invocation

[dongfengy]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #55013 [ run ] triggered by Bot. Commit: 0f1c610 Link to invocation

[tensorrt-cicd]

PR_Github #55013 [ run ] completed with state SUCCESS. Commit: 0f1c610
/LLM/main/L0_MergeRequest_PR pipeline #44005 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

[dongfengy]

/bot skip --comment "low risk test only change, ci has been failing due to unrelated changes so many times"

[tensorrt-cicd]

PR_Github #55057 [ skip ] triggered by Bot. Commit: 0f1c610 Link to invocation

[tensorrt-cicd]

PR_Github #55057 [ skip ] completed with state SUCCESS. Commit: 0f1c610
Skipping testing for commit 0f1c610

Link to invocation