heartmula

Generate music from lyrics and tags using HeartMuLa and HeartCodec.

Updated Apr 23, 2026
One-click install
npx skills add https://github.com/Rawgrowth-Consulting/rawclaw-agent --skill heartmula-rawgrowth-consulting
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: heartmula
Source: https://github.com/Rawgrowth-Consulting/rawclaw-agent/tree/main/skills/media/heartmula
Command: npx skills add https://github.com/Rawgrowth-Consulting/rawclaw-agent --skill heartmula-rawgrowth-consulting

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

HeartMuLa provides an open-source solution to generate songs from lyrics and descriptive tags, enabling artists and researchers to create music locally without relying on proprietary services.

Core Features & Use Cases

  • Music generation conditioned on lyrics and tags via HeartMuLa and HeartCodec.
  • Offline/local deployment with CPU/GPU options and lazy loading to minimize VRAM usage.
  • Includes HeartTranscriptor for lyrics transcription and HeartCLAP for alignment to sync audio with text.
  • Use Case: indie musicians producing songs from text prompts and researchers exploring open-source music pipelines.

Quick Start

Run HeartMuLa's music generation workflow by supplying lyrics and tags to produce an audio file.

Frequently Asked Questions about heartmula

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

FAQPage Schema
How do I generate music from lyrics locally?

You can generate songs from lyrics locally by supplying text lyrics and descriptive tags to HeartMuLa, which uses HeartCodec to produce high-fidelity audio files on your local GPU.

What GPU specs are needed for local AI music generation?

Local AI music generation requires a CUDA-enabled PyTorch environment with 8–16GB VRAM, supporting 3B and 7B models along with multi-GPU setups for offline production.

How does open-source music generation sync audio with text?

Open-source music generation syncs audio with text using HeartTranscriptor for lyrics transcription and HeartCLAP for alignment, ensuring generated audio matches the provided text prompts.

Can I run offline music production without high VRAM usage?

Offline music production can minimize VRAM usage through lazy loading and CPU fallback options, though a CUDA-enabled GPU with 8–16GB VRAM is recommended for high-fidelity audio.

Is local AI song creation suitable for indie musicians and research?

Local AI song creation suits indie musicians producing songs from text prompts and researchers exploring open-source music pipelines, providing a proprietary-free alternative for audio generation.