[https://nvbugs/6478715][fix] Renamed the worker's forward() to _forward_impl() so the base…

[chenfeiz0326]

Summary

• Root cause: MTPEagleDynamicTreeWorker overrode SpecWorkerBase.forward, which the init_subclass guard (nvbugs/6442074) forbids, raising TypeError at import time and collecting 0 tests.
• Fix: Renamed the worker's forward() to _forward_impl() so the base SpecWorkerBase.forward wrapper drives it and retains its attn-metadata cleanup guarantee; no behavioral change. Staged ONLY the source file to avoid the LFS-binary pollution from prior rejected commits.
• Automated fix generated by repair-bot

Test plan

• Verify fix on the same GPU type as the original failure
• Check for regressions in related tests

Links

• Bug: https://nvbugs/6478715

Dev Engineer Review

• Renamed MTPEagleDynamicTreeWorker.forward() to _forward_impl() so SpecWorkerBase can retain control of the guarded forward() wrapper and perform attention-metadata cleanup.
• No behavioral, configuration, or test-list changes were made.

QA Engineer Review

No test changes.

[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: b75e806b-c90d-482c-8076-230e6fa76978

📥 Commits

Reviewing files that changed from the base of the PR and between 9cd6a49 and ac73231.

📒 Files selected for processing (1)

• tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py

---

Walkthrough

The MTP dynamic tree worker’s top-level method is renamed from forward to _forward_impl; its parameters and implementation remain unchanged.

Changes

MTP dynamic tree worker

Layer / File(s)	Summary
Rename worker entry point tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py	Renames MTPEagleDynamicTreeWorker.forward(...) to _forward_impl(...) without changing the method signature or body.

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

Possibly related PRs

• NVIDIA/TensorRT-LLM#16724: Performs the same method-signature rename.
• NVIDIA/TensorRT-LLM#16727: Performs the same method rename in the same file.
• NVIDIA/TensorRT-LLM#16728: Renames the same method with unchanged logic.

Suggested reviewers: lori-ren

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description has Summary/Test plan/Links, but it omits the required Description and PR Checklist sections from the template.	Add a Description section explaining the issue and fix, plus a completed PR Checklist; keep the Test Coverage section aligned with the template.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title matches the change and includes the required ticket/type prefix.
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.}

[BowenFu]

This is already on main. The identical one-line rename landed in #16724 (1fbd240363, nvbugs/6442074) — tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py now defines _forward_impl at line 643 and no longer has a def forward at all.

#16730 (nvbugs/6478715) and #16731 (nvbugs/6445316) are byte-identical to each other and to what merged, so both are no-ops. Suggest closing both and resolving the two bugs as duplicates of 6442074.