Generate MP3 songs from lyrics and style tags using HeartMuLa checkpoints.
npx skills add https://github.com/ZardLi1115/zedclaw --skill heartmula-zardli1115
Please help me install this Agent Skill. Skill: heartmula Source: https://github.com/ZardLi1115/zedclaw/tree/main/skills/media/heartmula Command: npx skills add https://github.com/ZardLi1115/zedclaw --skill heartmula-zardli1115
HeartMuLa turns written lyrics and stylistic tags into full-length, music-like audio so you can create song drafts without composing melodies, harmonies, or instrumentation manually.
Ask the skill to generate an MP3 song using your provided lyrics file and comma-separated tags, loading the 3B model from your local checkpoint directory and saving the output to an MP3 path.
High-intent search queries and answers about installing and using this skill.
To generate MP3 songs from lyrics and style tags, provide a lyrics file formatted with structural sections like Intro and Chorus alongside comma-separated tags. The skill processes these inputs locally to output a full-length MP3 track without manual composition.
Yes, you can run audio synthesis locally and offline. The skill requires local HeartMuLa checkpoints and supports configurable device or dtype settings, allowing music generation entirely on your own machine without needing external API calls.
Music generation does work with CPU if you lack GPU VRAM. The skill supports both GPU and CPU execution and includes optional lazy loading for VRAM efficiency, allowing you to configure model precision to fit your local hardware constraints.
The best way to structure lyrics for multilingual music generation is to format them with explicit structural sections such as Intro, Verse, and Chorus. This conditions the audio synthesis to align accurately with your specified lyrical progression.
You need local checkpoints for offline song drafting because the skill relies on open-source foundation models like HeartMuLa for generation and HeartCodec for audio reconstruction. Loading these models locally ensures deterministic generation without internet dependency.
cabeznoGenerate MP3 songs from lyrics files and comma-separated tags using HeartMuLa with lazy loading enabled.
dakshrawat298-gifTranscribe local audio files into text or SRT/VTT captions using the Whisper CLI.
Cam10001110101Generate algorithmic music from code and export it as MP3.
JamesFincherGenerate full songs from lyrics and tags using HeartMuLa.
