754,000+ skills indexed

JamesFincherCOMMUNITY

Explore all skills by this author →

Verified:GitHub

heartmula

Generate full songs from lyrics and tags using HeartMuLa.

Updated May 4, 2026

Search Tags:#open source #music generation #audio synthesis #multilingual #lyrics to song #gpu inference #mp3 output

One-click install

npx skills add https://github.com/JamesFincher/gengar --skill heartmula-jamesfincher

Or copy as Structured Prompt for Agent▼

Please help me install this Agent Skill.
Skill: heartmula
Source: https://github.com/JamesFincher/gengar/tree/main/skills/media/heartmula
Command: npx skills add https://github.com/JamesFincher/gengar --skill heartmula-jamesfincher

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) components.

What problem does it solve?

HeartMuLa solves the challenge of turning lyrics and tags into complete, music-ready songs without relying on closed commercial tools.

Core Features & Use Cases

Lyrics + Tags Conditioned Generation: Produces full songs from provided lyrics with structural tags (e.g., Intro/Verse/Chorus) and comma-separated style tags.
Multicomponent Open-Source Pipeline: Uses HeartMuLa for generation, HeartCodec for high-fidelity audio reconstruction, HeartTranscriptor for lyrics transcription (Whisper-based), and HeartCLAP for alignment.
Local/Offline Workflow: Enables self-hosted music generation on your own compute, including multi-GPU splitting to reduce bottlenecks.

Quick Start

Use the heartmula skill to generate a song from your lyrics file assets/lyrics.txt and tags file assets/tags.txt and save the result to assets/output.mp3.

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 without using commercial tools?▼

Generate songs from lyrics offline by conditioning an open-source music model on your text structure and style tags. This local workflow outputs a complete MP3 file without relying on closed commercial platforms.

Can I use multilingual lyrics for AI audio synthesis?▼

Yes, multilingual lyrics are supported for AI audio synthesis. The generation pipeline conditions on diverse text inputs, enabling experimentation with multiple languages to produce full-length audio outputs for creative projects.

What GPU settings do I need for offline music generation?▼

Offline music generation requires appropriate GPU settings and a checkpoint download. You can use optional lazy loading to manage VRAM usage efficiently, and multi-GPU splitting is supported to reduce processing bottlenecks.

How do I structure lyrics and style tags for AI song generation?▼

Structure lyrics for AI song generation by adding structural tags like Intro, Verse, and Chorus within the text. Pair this with a comma-separated tags file to define the musical style, which conditions the model to produce the desired audio output.

What's the best way to convert text into full-length audio demos?▼

The best way to convert text into full-length audio demos is using a multicomponent open-source pipeline. It aligns lyrics transcription with generation and high-fidelity audio reconstruction to produce music-ready MP3 outputs for creative projects.

Related Skills

Explore More Design & Creative Skills→

heartmula

Generate complete songs from lyrics and tags using HeartMuLa components.

audio-jingle

Generate music, voiceovers, and sound effects as MP3 or WAV files.

code-to-music

Generate algorithmic music from code and export it as MP3.

heartmula

Generate complete AI music tracks from lyrics and style descriptions.