[TRTLLMINF-152][infra] Waive tests that have too many logs in system-out

[yiqingy0]

Summary by CodeRabbit

• Tests
  • Added new test waivers for a GPU accuracy scenario at multiple sparsity and KV-cache configurations.

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.

[yiqingy0]

/bot run --disable-fail-fast --post-merge

[yiqingy0]

/bot run --disable-fail-fast --post-merge

[coderabbitai]

📝 Walkthrough

Walkthrough

Four SKIP waiver entries are added to waives.txt for the test_skip_softmax_attention_4gpus test on DGX_B200 with the Qwen3 30B A3B Instruct 2507 model, covering target_sparsity values of 0.5 and 0.9 with fp8kv set to both False and True, referencing bug nvbugs/6390307.

Changes

Waiver List Update

Layer / File(s)	Summary
SKIP entries for Qwen3 30B softmax attention test tests/integration/test_lists/waives.txt	Adds four waiver lines for DGX_B200 test_skip_softmax_attention_4gpus parameterized by target_sparsity (0.5, 0.9) and fp8kv (False, True), all citing nvbugs/6390307.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#15077: Adds waiver entries for accuracy/test_llm_api_pytorch.py Qwen3 fp8/kv-flag variants in the same waives file.
• NVIDIA/TensorRT-LLM#15505: Adds B200-related test_llm_api_pytorch.py skip entries to the same waives file.
• NVIDIA/TensorRT-LLM#15719: Adds SKIP entries for failing LLM accuracy tests in the same waives file.

Suggested reviewers

• dongxuy04
• pcastonguay
• yuxianq

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description only contains the template and a checked checklist, with no actual title, description, or test coverage filled in.	Add the PR title in the required format and briefly describe the issue, the fix, and relevant test coverage.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and accurately summarizes the test-waiver change.
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.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/integration/test_lists/waives.txt`:
- Around line 173-176: The softmax-attention waiver is still applied globally
for one Qwen3_30B_A3B_Instruct_2507 case, even though the newer entries are
DGX_B200-scoped. Update the waiver entries in test_lists/waives.txt so the
target_sparsity_0.9-fp8kv=False skip is either removed or restricted to DGX_B200
like the other
`TestQwen3_30B_A3B_Instruct_2507::test_skip_softmax_attention_4gpus` entries,
keeping the waiver scope consistent with the intended hardware-specific skip.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: f7211f21-dedf-424f-aeff-256348a1bc0f

📥 Commits

Reviewing files that changed from the base of the PR and between 96416f1 and b1ad9d7.

📒 Files selected for processing (1)

• tests/integration/test_lists/waives.txt

[yiqingy0]

/bot run --disable-fail-fast --post-merge

[tensorrt-cicd]

PR_Github #56506 [ run ] triggered by Bot. Commit: b1ad9d7 Link to invocation

[tensorrt-cicd]

PR_Github #56506 [ run ] completed with state SUCCESS. Commit: b1ad9d7
/LLM/main/L0_MergeRequest_PR pipeline #45344 completed with status: 'SUCCESS'

CI Report

Link to invocation

[yiqingy0]

/bot run --stage-list "*" --disable-fail-fast

[tensorrt-cicd]

PR_Github #56796 [ run ] triggered by Bot. Commit: ce70c0f Link to invocation

[tensorrt-cicd]

PR_Github #56796 [ run ] completed with state FAILURE. Commit: ce70c0f
/LLM/main/L0_MergeRequest_PR pipeline #45613 (Partly Tested) 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

[yiqingy0]

/bot skip --comment "Waive tests"

[tensorrt-cicd]

PR_Github #57094 [ skip ] triggered by Bot. Commit: 231475c Link to invocation

[tensorrt-cicd]

PR_Github #57094 [ skip ] completed with state SUCCESS. Commit: 231475c
Skipping testing for commit 231475c

Link to invocation