[https://nvbugs/6272668][infra] Unwaive DSR1 and Qwen3.5 again

[taylor-yb-lee]

@coderabbitai summary

Description

• DSR1 and Qwen3.5 tests were unwaived with fixes in PR14870 and PR14667
• At the same time, PR 14686 commented out required model registry, which caused the tests' crash again
• https://github.com/NVIDIA/TensorRT-LLM/pull/15001/changes restored those model registries
• Now re-unwaive them in this PR

Below is the history

commit 178c8f4
Author: Taylor Yeonbok Lee <249374542+taylor-yb-lee@users.noreply.github.com>
Date:   Tue Jun 2 16:30:34 2026 -0700

    [https://nvbugs/6240561][fix] Unwaive DeepSeek R1 accuracy test (#14870)

    Signed-off-by: Taylor Yeonbok Lee <249374542+taylor-yb-lee@users.noreply.github.com>

commit cd38dfb (TensorRT-LLM_gptoss_rebase_0602)
Author: Venky <23023424+venkywonka@users.noreply.github.com>
Date:   Tue Jun 2 11:00:33 2026 -0700

    [https://nvbugs/6226933][fix] canonicalize multimodal cache-key serialization to prevent hash collisions (#14800)

    Signed-off-by: venkywonka <23023424+venkywonka@users.noreply.github.com>

commit 9653450
Author: tcherckez-nvidia <127761168+tcherckez-nvidia@users.noreply.github.com>
Date:   Tue Jun 2 19:59:29 2026 +0300

    [None][chore] Update AD model list (#14686)

    Signed-off-by: Yueh-Ting (eop) Chen <yuehtingc@nvidia.com>
    Signed-off-by: tcherckez-nvidia <127761168+tcherckez-nvidia@users.noreply.github.com>
    Co-authored-by: Yueh-Ting (eop) Chen <yuehtingc@nvidia.com>

commit 66262cf
Author: Chien-Chun Hung <2679986+chienchunhung@users.noreply.github.com>
Date:   Tue Jun 2 09:05:56 2026 -0700

    [TRTLLM-12648][test] implement disagg cancel stress metrics_thread (#14807)

    Signed-off-by: Chien-Chun Hung <2679986+chienchunhung@users.noreply.github.com>

commit 6ef7b38
Author: Taylor Yeonbok Lee <249374542+taylor-yb-lee@users.noreply.github.com>
Date:   Tue Jun 2 09:04:32 2026 -0700

    [https://nvbugs/6221450][fix] AutoDeploy: Qwen3.5 400B NVFP4 accuracy regression fix (#14667)

    Signed-off-by: Taylor Yeonbok Lee <249374542+taylor-yb-lee@users.noreply.github.com> (edited)

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]

📝 Walkthrough

Walkthrough

This PR removes two test waiver entries from the integration test skip list: the DeepSeek-R1-0528-True autodeploy registry accuracy test and the Qwen3.5-MoE NVFP4 quantization test. These removals re-enable those specific model tests in CI.

Changes

Test Waiver List Update

Layer / File(s)	Summary
Remove waiver entries for DeepSeek-R1 and Qwen3.5-MoE tests tests/integration/test_lists/waives.txt	Two SKIP waiver entries are removed: TestModelRegistryAccuracy::test_autodeploy_from_registry[deepseek-ai_DeepSeek-R1-0528-True] and TestQwen3_5_397B_MoE::test_nvfp4[8]. Other skip entries in the block are retained.

Possibly related PRs

• NVIDIA/TensorRT-LLM#14870: Both PRs remove the DeepSeek R1 accuracy test waiver entry, unwaiving the same test in the runner.
• NVIDIA/TensorRT-LLM#14667: This PR unwaives the Qwen3.5-MoE NVFP4 accuracy test case, which depends on the corresponding code/config fix in the related PR.
• NVIDIA/TensorRT-LLM#15001: Both PRs remove skip waivers for the same DeepSeek-R1 and Qwen3.5-MoE models to re-enable their acceptance tests.

Suggested reviewers

• hnover-nv
• nvchenghaoz
• galagam
• chienchunhung

🎯 1 (Trivial) | ⏱️ ~2 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
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.
Description check	✅ Passed	PR description adequately explains the context and rationale for re-unwaiving tests, referencing prior PRs and the specific issue.
Title check	✅ Passed	The PR title clearly and specifically refers to unwaiving DSR1 and Qwen3.5 tests, which directly matches the primary change of removing two waiver entries from the test list.

_{✏️ 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.}

[taylor-yb-lee]

/bot run --stage-list "DGX_B200-8_GPUs-AutoDeploy-Post-Merge-1"

[tensorrt-cicd]

PR_Github #52392 [ run ] triggered by Bot. Commit: 6866cd6 Link to invocation

[taylor-yb-lee]

/bot run

[tensorrt-cicd]

PR_Github #52396 [ run ] triggered by Bot. Commit: 6866cd6 Link to invocation

[tensorrt-cicd]

PR_Github #52392 [ run ] completed with state ABORTED. Commit: 6866cd6

Link to invocation

[tensorrt-cicd]

PR_Github #52396 [ run ] completed with state FAILURE. Commit: 6866cd6
/LLM/main/L0_MergeRequest_PR pipeline #41689 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

[taylor-yb-lee]

/bot run --stage-list "DGX_B200-8_GPUs-AutoDeploy-Post-Merge-1"

[tensorrt-cicd]

PR_Github #52425 [ run ] triggered by Bot. Commit: 6866cd6 Link to invocation

[tensorrt-cicd]

PR_Github #52425 [ run ] completed with state SUCCESS. Commit: 6866cd6
/LLM/main/L0_MergeRequest_PR pipeline #41717 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[taylor-yb-lee]

/bot skip --comment "post merge job passed which was the target scope of this unwaive."

[tensorrt-cicd]

PR_Github #52458 [ skip ] triggered by Bot. Commit: 6866cd6 Link to invocation

[tensorrt-cicd]

PR_Github #52458 [ skip ] completed with state SUCCESS. Commit: 6866cd6
Skipping testing for commit 6866cd6

Link to invocation