metadata
license: apache-2.0
base_model: k2-fsa/OmniVoice
library_name: mlx-audio
pipeline_tag: text-to-speech
tags:
- mlx
OmniVoice in MLX (bf16). For mlx-audio on Apple Silicon.
Usage
pip install mlx-audio
python -m mlx_audio.tts.generate \
--model commandeaw/OmniVoice-MLX-bf16 \
--text "สวัสดีค่ะ ยินดีต้อนรับ" \
--lang_code th \
--ref_audio reference.wav
--ref_audio is optional (zero-shot voice cloning); drop it for the default voice.
Size
| variant | repo | vs base (fp32) |
|---|---|---|
| bf16 | 1.6 GB | -50% |
| 8bit | 1.1 GB | -68% |
| 4bit | 0.75 GB | -77% |