create a docker-compose.yml for piper neural text with tts and rvc and api endpoint to generate audio and clone audio

@ahmedtahash

I am certain that I can design a production-ready docker-compose setup integrating Piper TTS, RVC voice cloning, and an API layer for audio generation and voice conversion. The stack will include isolated services, GPU support (if available), shared volume management for models, and a clean REST endpoint for TTS and cloning requests. I will ensure scalable container orchestration, proper dependency handling, and clear documentation for deployment and model management. Will this deployment require GPU acceleration (CUDA) for real-time RVC inference, or should it be optimized for CPU-only environments? Regards, Ahmad Al-Ashery.