heartmula

Generate music from lyrics and tags using open-source models.

Updated Apr 15, 2026
One-click install
npx skills add https://github.com/juliuss1907/knowledge-base --skill heartmula-juliuss1907
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: heartmula
Source: https://github.com/juliuss1907/knowledge-base/tree/main/.hermes/skills/media/heartmula
Command: npx skills add https://github.com/juliuss1907/knowledge-base --skill heartmula-juliuss1907

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires transformers, datasets, torch, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill enables the generation of music from provided lyrics and tags, offering an open-source alternative to commercial music generation tools.

Core Features & Use Cases

  • Lyric-Based Music Generation: Create music directly from user-provided lyrics.
  • Tag-Driven Customization: Generate music that matches specified tags and moods.
  • Open-Source Models: Utilizes open-source models like HeartMuLa, HeartCodec, and HeartTranscriptor.
  • Use Case: Generate a romantic piano ballad from the lyrics "I love you" with tags like "piano", "romantic", and "ballad".

Quick Start

Generate a song from the lyrics "I love you" with tags "piano", "romantic", "ballad" and save the output as 'song.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 using open-source models?

To generate music from lyrics, you provide text descriptions and tags to condition open-source models like HeartMuLa, which outputs an audio file such as an MP3. The generation process supports multilingual lyrics and utilizes specific models for audio reconstruction and text alignment.

Can I customize AI music generation with specific mood tags?

You can customize AI music generation by providing specific tags like "piano", "romantic", or "ballad" alongside your lyrics. The models condition the audio output to match these specified moods and instrumental styles.

Do I need CUDA support and Python 3.10 to run lyric-based music generation?

Yes, you need Python 3.10 and CUDA support to run this lyric-based music generation process. The Skill also requires specific libraries including transformers, datasets, and PyTorch to execute the audio-text alignment and generation models.

What is the best open-source alternative for generating music from text?

An open-source alternative for generating music from text is using models like HeartMuLa and HeartCodec. They condition audio reconstruction and lyrics transcription directly on text inputs, providing a non-commercial pathway for generating audio from multilingual lyrics.

Does the music generation process support multilingual lyrics transcription?

Yes, the music generation process supports multilingual lyrics transcription. It uses open-source models like HeartTranscriptor to align the audio output with the provided text, ensuring accurate transcription across different languages.