Files
Realtime-Voice-Clone-Chinese/SKILL.md
2026-03-02 22:16:56 +08:00

26 lines
865 B
Markdown

---
name: MockingBird
description: AI voice cloning and TTS toolkit. The only active skill is `speak`.
---
# MockingBird
Voice cloning / text-to-speech toolkit built on encoder-synthesizer-vocoder pipeline.
## Available Skills
| Skill | Path | Purpose |
|-------|------|---------|
| speak | `skills/speak/SKILL.md` | Text-to-speech (Kokoro local / Noiz cloud), timeline rendering, voice cloning |
## Project Layout
- `models/` — encoder, synthesizer, vocoder, ppg2mel, ppg_extractor
- `control/` — CLI and GUI entry points (toolbox, mkgui)
- `skills/speak/` — TTS skill scripts (`tts.sh`, `noiz_tts.py`, `render_timeline.py`, `text_to_srt.py`)
- `data/ckpt/` — pretrained checkpoints
## When to Use
If the user asks for anything related to speech synthesis, TTS, voice cloning, dubbing, or audio generation, read and follow `skills/speak/SKILL.md`.