将 ZipEnhancer 模型从 ModelScope pipeline 中剥离,用纯 PyTorch 实现推理流程,封装为 pip 可安装的降噪库和 FastAPI 批量语音降噪,为 CosyVoice 提供干净的人声输入。
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).