[spec decoding] fix inkling multi layer mtp draft extend cuda graph

[Qiaolin-Yu]

Motivation

Modifications

Accuracy Tests

Speed Tests and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

Review and Merge Process

1. Ping Merge Oncalls to start the process. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments or contact authorized users to do so.
   • Common commands include /tag-and-rerun-ci, /tag-run-ci-label, /rerun-failed-ci
4. After green CI and required approvals, ask Merge Oncalls or people with Write permission to merge the PR.

---

CI States

Latest PR Test (Base): ❌ Run #30046404681
Latest PR Test (Extra): ❌ Run #30046404529

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[Qiaolin-Yu]

/rerun-test test/registered/models_e2e/test_step3p5_flash_chain_mtp.py

[github-actions]

Results for /rerun-test test/registered/models_e2e/test_step3p5_flash_chain_mtp.py:

🚀 8-gpu-h200 (1 test): ✅ View workflow run

cd test/ && python3 registered/models_e2e/test_step3p5_flash_chain_mtp.py