[https://nvbugs/6507109][infra] Split slow DGX B300 attention unit tests

[yuxianq]

Dev Engineer Review

• Updated tests/integration/test_lists/test-db/l0_dgx_b300.yml to run the attention suite while excluding four slow sparse attention test files for independent CI execution.
• The change is limited to test selection configuration; no code, APIs, dependencies, or production behavior are affected.
• Test paths and ignore entries are consistent with the stated objective, with no apparent scope expansion or duplicate declarations.

QA Engineer Review

• Modified tests/integration/test_lists/test-db/l0_dgx_b300.yml.
• Removed the broad exclusion of unittest/_torch/attention.
• Added the attention suite with four explicit sparse attention file exclusions: two FP8/FP4 paged MQA logits tests and two GVR top-k decode tests.
• Verdict: needs follow-up pending CBTS coverage data.

Description

Apply the attention test split from #16709 to the DGX B300 post-merge test list. The main attention entry excludes the three slow sparse CuTe DSL files, and each excluded file runs as a separate CI entry.

Test Coverage

• pre-commit run --files tests/integration/test_lists/test-db/l0_dgx_b300.yml
• python scripts/check_test_list.py --validate
• git diff --check

PR Checklist

• PR description clearly explains what and why.
• PR follows the TRT-LLM coding guidelines.
• No new code paths, APIs, or dependencies are introduced.
• No CODEOWNERS, documentation, or architecture updates are required.
• 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, comment /bot help.

[yuxianq]

/bot run --disable-fail-fast

[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: 53329788-e55d-4a83-adcb-3a939484f192

📥 Commits

Reviewing files that changed from the base of the PR and between 75b39d4 and dd3d98a.

📒 Files selected for processing (1)

• tests/integration/test_lists/test-db/l0_dgx_b300.yml

---

Walkthrough

The DGX B300 post-merge test configuration now runs the attention test suite while excluding four specified sparse attention test files.

Changes

DGX B300 attention test selection

Layer / File(s)	Summary
Attention test selection tests/integration/test_lists/test-db/l0_dgx_b300.yml	The post-merge selection targets unittest/_torch/attention and ignores four sparse attention test files.

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

Suggested reviewers: joyang-nv

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title matches the main change and follows the required [ticket][type] format.
Description check	✅ Passed	The description includes the required summary, description, test coverage, and checklist sections.
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.}

[tensorrt-cicd]

PR_Github #61564 [ run ] triggered by Bot. Commit: dd3d98a Link to invocation

[tensorrt-cicd]

PR_Github #61564 [ run ] completed with state SUCCESS. Commit: dd3d98a
/LLM/main/L0_MergeRequest_PR pipeline #49775 completed with status: 'SUCCESS'

CI Report

Link to invocation

[longlee0622]

Approved for multi-gpu run