Fix matcha tts

[csukuangfj]

Summary by CodeRabbit

• Refactor
  • Optimized internal code structure by reorganizing token processing sequence and eliminating duplicate operations to improve efficiency without affecting functionality.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[gemini-code-assist]

Summary of Changes

Hello @csukuangfj, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request resolves a bug in the Matcha Text-to-Speech (TTS) implementation where a critical step in token processing was inadvertently placed within a debug-only conditional block. By moving this logic to execute unconditionally, the fix ensures that the TTS system consistently receives properly prepared input, allowing it to function correctly in all operational modes.

Highlights

• Matcha TTS Token Processing: The logic responsible for concatenating input tokens into the x vector was moved outside of the if (config_.model.debug) block. This ensures that the tokens are always correctly processed and prepared for the Matcha TTS model, regardless of whether debug mode is enabled or not.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[coderabbitai]

Caution

Review failed

The pull request is closed.

Walkthrough

A refactoring that moves token vector construction earlier in the Process function of offline-tts-matcha-impl.h, eliminating a redundant insertion loop previously located in a debug section. No changes to control flow or output values.

Changes

Cohort / File(s)	Change Summary
Token Vector Construction Refactoring sherpa-onnx/csrc/offline-tts-matcha-impl.h	Moved token aggregation loop earlier in Process function and removed duplicate token concatenation logic from debug section.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

• Verify the relocated loop maintains identical behavior before and after the move
• Confirm debug output remains functionally equivalent after removing the redundant section

Possibly related PRs

• fix matcha tts zh-en model #2851: Also modifies token vector construction and debug logging in offline-tts-matcha-impl.h.

Suggested labels

size:M

Poem

🐰 A loop moved up the stack with care,
Redundant work? We'll not keep there!
Same output, cleaner flow,
Debug logs still steal the show. ✨

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

📜 Recent review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 5e7e252 and 2d4ced9.

📒 Files selected for processing (1)

• sherpa-onnx/csrc/offline-tts-matcha-impl.h (1 hunks)

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[gemini-code-assist]

Code Review

This pull request correctly fixes a bug in OfflineTtsMatchaImpl::Process where the input tensor x was only being populated in debug mode. By moving the population of x outside the if (config_.model.debug) block, the function now behaves correctly in all build configurations. The change is clear and addresses the issue effectively. I have no further comments.