[https://nvbugs/6631019][test] waive Nemotron Nano FP8 CUDA graph test on DGX B200

[allisonlim-nv]

Description

accuracy/test_llm_api_pytorch_multimodal.py::TestNanoV3Omni::test_auto_dtype[fp8_mmmu_encoder_cuda_graph] consistently fails on DGX_B200-PyTorch-5 during Nemotron Nano / RADIO encoder CUDA-graph initialization with:

AssertionError: Model extra attrs is not set

The same failure occurred on unrelated pre-merge PRs #16839 and #17562, indicating a main-side issue rather than a PR-specific regression. The initial x86_64 single-GPU failure blocks or aborts downstream L0 stages.

This change adds one hardware-scoped waiver for DGX_B200 only; the test remains enabled on other platforms. The underlying issue is tracked in nvbugs/6631019.

Test Coverage

Test-list-only change; no source, build, or API changes.

The waiver targets only:

full:DGX_B200/accuracy/test_llm_api_pytorch_multimodal.py::TestNanoV3Omni::test_auto_dtype[fp8_mmmu_encoder_cuda_graph]

Dev Engineer Review

• Added one DGX_B200 waiver in tests/integration/test_lists/waives.txt.
• The waiver targets accuracy/test_llm_api_pytorch_multimodal.py::TestNanoV3Omni::test_auto_dtype[fp8_mmmu_encoder_cuda_graph].
• The waiver references nvbugs/6631019.
• The change is limited to DGX B200 and does not affect other platforms.
• No source, build, API, or performance changes are included.
• No duplicate or unintended scope changes were identified.

QA Engineer Review

• Modified tests/integration/test_lists/waives.txt.
• Added one waiver entry.
• No test code was added, modified, or removed.
• CBTS coverage data is unavailable.
• Verdict: needs follow-up.

[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: c07f2d5b-afaf-4404-8daa-53b3f27252a9

📥 Commits

Reviewing files that changed from the base of the PR and between 6a53222 and 8ada6a0.

📒 Files selected for processing (1)

• tests/integration/test_lists/waives.txt

Included review availability: Your plan provides up to 12 included reviews per hour; 11 remain after this review.

---

Walkthrough

The integration waiver list adds a DGX B200 skip entry for the NanoV3 Omni test_auto_dtype case using the FP8 MMMU encoder CUDA-graph configuration.

Changes

Integration waiver

Layer / File(s)	Summary
Add DGX B200 test waiver tests/integration/test_lists/waives.txt	Adds a skip entry for TestNanoV3Omni::test_auto_dtype[fp8_mmmu_encoder_cuda_graph] with tracking reference nvbugs/6631019.

Estimated code review effort: 1 (Trivial) | ~2 minutes

Merge Risk: ⚪ Minimal · up to 8ada6

This narrowly disables one consistently failing test on DGX_B200 while leaving other platforms unchanged; no actionable merge-blocking risk remains after normal checks and review.

Suggested reviewers: bowenfu

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the DGX B200-specific waiver for the Nemotron Nano FP8 CUDA graph test.
Description check	✅ Passed	The description explains the failure, scope, tracking bug, solution, and targeted test coverage; the omitted checklist is non-critical.
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.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[allisonlim-nv]

/bot skip --comment "waives-only PR"

[tensorrt-cicd]

PR_Github #67136 [ skip ] triggered by Bot. Commit: 88ce429 Link to invocation

[tensorrt-cicd]

PR_Github #67136 [ skip ] completed with state SUCCESS. Commit: 88ce429
Skipping testing for commit 88ce429

Link to invocation