[https://nvbugs/6478723][fix] Unwaive TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16 after the MTPEagleDynamicTreeWorker fix by zhaoyangwang-nvidia · Pull Request #17855 · NVIDIA/TensorRT-LLM

zhaoyangwang-nvidia · 2026-08-18T01:54:46Z

Summary

Root cause: MTPEagleDynamicTreeWorker overrode SpecWorkerBase.forward, which __init_subclass__ forbids — the resulting class-creation TypeError surfaced in L0_PostMerge as RuntimeError: Executor worker returned error on TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16 (nvbugs/6478723), and the test was waived.
Fix: The root cause was already fixed in [https://nvbugs/6478723][fix] Rename forward -> _forward_impl AND stash the extra dynamic-tree state on… #16733 (rename forward -> _forward_impl and restore the extra dynamic-tree state via _ensure_spec_dec_state_restored). This PR only removes the now-stale waive so the test runs again on main.
No product code change: one line removed from tests/integration/test_lists/waives.txt.

Test plan

DGX_H100-4_GPUs-PyTorch-Others-1: accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16 passes with the [https://nvbugs/6478723][fix] Rename forward -> _forward_impl AND stash the extra dynamic-tree state on… #16733 fix on main.
Note: the CI report flagged this test as flaky (2 failures / 30 days), so the CI run is also used to confirm it is stable before closing the bug.

Links

Dev Engineer Review

Removed the stale waiver for TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16.
The change affects only tests/integration/test_lists/waives.txt.
The entry removal uses the expected format and has no unintended scope.
The change has no API, performance, or error-handling impact.
The test is re-enabled because PR #16733 fixed the underlying MTPEagleDynamicTreeWorker issue.

QA Engineer Review

Modified tests/integration/test_lists/waives.txt.
Removed the waiver entry for TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16.
No test-db/ or qa/ files were modified.
Verdict: needs follow-up because CBTS coverage data is unavailable.

coderabbitai · 2026-08-18T01:56:36Z

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: 7fe78d1a-9d84-41fb-a4ed-cc44043a1ce8

📥 Commits

Reviewing files that changed from the base of the PR and between bef843e and 5cd2a94.

📒 Files selected for processing (1)

tests/integration/test_lists/waives.txt

💤 Files with no reviewable changes (1)

tests/integration/test_lists/waives.txt

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

Walkthrough

The integration test list no longer contains the waiver for TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16.

Changes

Nemotron V3 Super test waiver

Layer / File(s)	Summary
Remove test waiver `tests/integration/test_lists/waives.txt`	Removed the skip waiver for the NVFP4 four-GPU Hopper test.

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

Merge Risk: ⚪ Minimal · up to 5cd2a

This change only re-enables an existing integration test and does not alter product behavior; no actionable merge-blocking risk remains beyond normal checks and review.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title uses the required NVBugs and fix format and clearly states that the specified test waiver was removed after the fix.
Description check	✅ Passed	The description explains the root cause, references the existing fix, identifies the one-line change, and provides relevant test coverage.
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.}

zhaoyangwang-nvidia · 2026-08-18T01:58:34Z

/bot run

tensorrt-cicd · 2026-08-18T02:04:38Z

PR_Github #66918 [ run ] triggered by Bot. Commit: c64ff46 Link to invocation

tensorrt-cicd · 2026-08-18T02:34:12Z

PR_Github #66918 [ run ] completed with state FAILURE. Commit: c64ff46
/LLM/main/L0_MergeRequest_PR pipeline #54471 completed with status: 'UNSTABLE'

CI Report

⚠️ Multi-GPU Label Required:
Multi-GPU tests require the ci: full pre-merge approved label on this PR. Ask a member of NVIDIA/trt-llm-ci-approvers to add the label, then re-trigger CI with the same bot command (no rebase needed).

⚠️ 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

…_4gpus_hopper_w4a16 The waive was added while MTPEagleDynamicTreeWorker illegally overrode SpecWorkerBase.forward, which raised a TypeError at class creation and surfaced as 'Executor worker returned error'. That root cause was fixed in NVIDIA#16733, so re-enable the test to confirm it passes on main. Signed-off-by: ZhaoyangWang <zhaoyangw@nvidia.com>

zhaoyangwang-nvidia · 2026-08-18T03:29:02Z

Rebased onto latest main, which now contains the unwaive from #17830 (test_beam_search_e2e[multi_process-TorchSampler-no_cuda_graph_and_overlap-stop_token_ids0-1-1-True-True-True], nvbugs/6581048). Explicitly requesting the stages that own both unwaived tests so neither is silently skipped:

DGX_H100-4_GPUs-PyTorch-Others-1 -> TestNemotronV3Super::test_nvfp4_4gpus_hopper_w4a16 (this PR, nvbugs/6478723)
A30-PyTorch-1 / A30-PyTorch-2 -> unittest/_torch/sampler/test_beam_search.py ([https://nvbugs/6581048][test] Unwaive TorchSampler beam-search e2e case #17830, nvbugs/6581048)

zhaoyangwang-nvidia · 2026-08-18T03:29:03Z

/bot run --extra-stage "DGX_H100-4_GPUs-PyTorch-Others-1, A30-PyTorch-1, A30-PyTorch-2"

coderabbitai · 2026-08-18T03:29:12Z

Note

GitHub couldn't provide a complete incremental comparison for this pull request, so CodeRabbit is performing a full review instead. This review may take a little longer.

tensorrt-cicd · 2026-08-18T03:34:32Z

PR_Github #66960 [ run ] triggered by Bot. Commit: 5cd2a94 Link to invocation

tensorrt-cicd · 2026-08-18T05:17:11Z

PR_Github #66960 [ run ] completed with state FAILURE. Commit: 5cd2a94
/LLM/main/L0_MergeRequest_PR pipeline #54508 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

zhaoyangwang-nvidia · 2026-08-18T08:12:05Z

/bot run --extra-stage "DGX_H100-4_GPUs-PyTorch-Others-1, A30-PyTorch-1, A30-PyTorch-2"

tensorrt-cicd · 2026-08-18T08:17:52Z

PR_Github #67038 [ run ] triggered by Bot. Commit: 5cd2a94 Link to invocation

tensorrt-cicd · 2026-08-18T12:49:57Z

PR_Github #67038 [ run ] completed with state SUCCESS. Commit: 5cd2a94
/LLM/main/L0_MergeRequest_PR pipeline #54581 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

zhaoyangwang-nvidia · 2026-08-18T14:11:22Z

/bot run --disable-fail-fast --extra-stage "DGX_H100-4_GPUs-PyTorch-Others-1, A30-PyTorch-1, A30-PyTorch-2"

tensorrt-cicd · 2026-08-18T14:17:58Z

PR_Github #67087 [ run ] triggered by Bot. Commit: 5cd2a94 Link to invocation

tensorrt-cicd · 2026-08-18T19:02:52Z

PR_Github #67087 [ run ] completed with state SUCCESS. Commit: 5cd2a94
/LLM/main/L0_MergeRequest_PR pipeline #54625 completed with status: 'SUCCESS'

CI Report

Link to invocation

github-actions Bot assigned zhaoyangwang-nvidia Aug 18, 2026

zhaoyangwang-nvidia force-pushed the unwaive-6478723 branch from c64ff46 to 5cd2a94 Compare August 18, 2026 03:28

xxi-nv added the ci: full pre-merge approved label Aug 18, 2026

lori-ren approved these changes Aug 19, 2026

View reviewed changes

zhaoyangwang-nvidia merged commit bd90276 into NVIDIA:main Aug 19, 2026
13 checks passed

Uh oh!

Conversation

zhaoyangwang-nvidia commented Aug 18, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Links

Dev Engineer Review

QA Engineer Review

Uh oh!

coderabbitai Bot commented Aug 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Uh oh!

zhaoyangwang-nvidia commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

zhaoyangwang-nvidia commented Aug 18, 2026

Uh oh!

zhaoyangwang-nvidia commented Aug 18, 2026

Uh oh!

coderabbitai Bot commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

zhaoyangwang-nvidia commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

zhaoyangwang-nvidia commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

tensorrt-cicd commented Aug 18, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

zhaoyangwang-nvidia commented Aug 18, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented Aug 18, 2026 •

edited

Loading