heartmula

Generate full songs from lyrics and tags using HeartMuLa and HeartCodec.

1|Updated Apr 10, 2026
One-click install
npx skills add https://github.com/zerwiz/wayofpi --skill heartmula-zerwiz
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: heartmula
Source: https://github.com/zerwiz/wayofpi/tree/main/.hermes/hermes-agent/skills/media/heartmula
Command: npx skills add https://github.com/zerwiz/wayofpi --skill heartmula-zerwiz

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It removes the friction of setting up and running an open-source music generation stack so you can turn lyrics and style tags into complete songs.

Core Features & Use Cases

  • Full-song generation from lyrics plus comma-separated tags.
  • Model stack guidance for HeartMuLa, HeartCodec, and transcription or alignment components.
  • Practical deployment help for local GPU use, lazy loading, precision choices, and checkpoint setup.
  • Use case: Produce a multilingual demo track, then iterate on tags, structure, and generation settings until the result fits your project.

Quick Start

Ask the skill to help you install HeartMuLa, prepare the checkpoints, and generate a song from your lyrics and tags.

Frequently Asked Questions about heartmula

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a full song from lyrics and tags locally?

You can generate a full song from lyrics and tags by configuring a CUDA-aware GPU environment, installing the HeartMuLa and HeartCodec runtime stack, loading checkpoints, and passing your lyrics with comma-separated style tags to the generation workflow.

What is needed to set up open-source music generation on a local GPU?

Setting up open-source music generation on a local GPU requires a CUDA-aware configuration, the HeartMuLa and HeartCodec runtime stack, proper checkpoint downloads, and dependency fixes to ensure controlled precision and loading behavior for audio synthesis.

Does this music generation workflow support multilingual songwriting?

Yes, the music generation workflow supports multilingual songwriting, allowing you to produce demo tracks with lyrics in various languages and iterate on tags, structure, and generation settings.

How do I fix dependency and loading issues when running audio synthesis models?

To fix dependency and loading issues during audio synthesis, apply required dependency fixes and utilize lazy loading behavior within the HeartMuLa and HeartCodec stack to manage precision choices and checkpoint loading effectively.

Can I iterate on audio generation settings after creating an initial song?

Yes, you can iterate on audio generation settings after creating an initial song by adjusting the comma-separated style tags, song structure, and generation parameters until the resulting audio fits your project requirements.