voco-fishspeech

Fish Speech plugin for VOCO - multilingual text-to-speech with voice cloning.

Install

pip install -e .

Usage

from voco.core import AudioRouter

router = AudioRouter()
router.load("fishspeech", alias="tts")

for result in router.infer(
    "tts",
    text="Hello world",
    ref_audio="reference.wav",
    ref_text="Reference transcript"
):
    print(result.audio.shape)

Features

• Multilingual support
• Voice cloning capabilities
• High-quality speech synthesis
• Flexible reference encoding

Requirements

• Python >=3.10
• PyTorch with CUDA support recommended