665,000+ skills indexed

BermudaLocalsCOMMUNITY

Explore all skills by this author →

Verified:Website GitHub

heartmula

Generate music from lyrics and tags using HeartMuLa models.

1|1|Updated Apr 26, 2026

Search Tags:#multilingual #open-source #lyrics #music-generation #ai-music #offline-generation #hardware-requirements

One-click install

npx skills add https://github.com/BermudaLocals/hermes-agent-lite --skill heartmula-bermudalocals

Or copy as Structured Prompt for Agent▼

Please help me install this Agent Skill.
Skill: heartmula
Source: https://github.com/BermudaLocals/hermes-agent-lite/tree/main/skills/media/heartmula
Command: npx skills add https://github.com/BermudaLocals/hermes-agent-lite --skill heartmula-bermudalocals

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

HeartMuLa provides open-source music generation models that produce music conditioned on lyrics and tags, enabling offline, local music creation with multilingual support.

Core Features & Use Cases

Generate music from lyrics and tags using the HeartMuLa family.
Includes HeartMuLa, HeartCodec, HeartTranscriptor, and HeartCLAP for end-to-end creation and alignment.
Supports offline/local generation with multi-GPU configurations and lazy-loading to reduce VRAM usage.
Suitable for demos, soundtracks, and rapid prototyping from textual descriptions.

Quick Start

Run the provided example with your lyrics and tags to produce an 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 and tags locally?▼

Music generation from lyrics and tags runs locally and offline through the HeartMuLa open-source family. It processes your textual inputs using components like HeartMuLa, HeartCodec, HeartTranscriptor, and HeartCLAP to produce audio outputs.

What hardware do I need for offline AI music generation?▼

Offline AI music generation requires CUDA-enabled GPUs with 8 to 16GB VRAM. Multi-GPU configurations are supported, and lazy-loading is utilized to reduce VRAM usage during the generation process.

Can I use multilingual lyrics for AI music generation?▼

Yes, multilingual lyrics are supported for AI music generation. The open-source models process diverse language inputs alongside tags to condition and align the generated audio tracks.

What is the best way to create music from text descriptions without internet?▼

Creating music from text descriptions offline is achieved through open-source local generation models. The HeartMuLa family provides end-to-end creation and alignment from lyrics and tags directly on your local hardware.

Does open-source music generation work with Python and multi-GPU setups?▼

Open-source music generation works with a Python environment and supports multi-GPU configurations. Installation requires following specific setup steps and checkpoints to ensure the local environment functions correctly.

Why does offline music generation require 8 to 16GB of VRAM?▼

Offline music generation requires 8 to 16GB of VRAM to load and process the AI models locally on CUDA-enabled GPUs. The system uses lazy-loading to help reduce this VRAM overhead during the generation process.

Related Skills

Explore More Software Engineering Skills→

Rawgrowth-Consulting

heartmula

Generate music from lyrics and tags using HeartMuLa and HeartCodec.

acestep

Generate original music, vocal tracks, covers, and stem extractions from text prompts.

code-to-music

Generate algorithmic music from code and export it as MP3.

heartmula

Generates full songs from user lyrics and style tags as MP3 files.