styletts2-api Text-To-Speech Inference Server for StyleTTS2 Usage docker compose up curl -X POST "http://localhost:4321/generate" \ -H "Content-Type: application/json" \ -d '{"text": "Hello world"}' \ --output audio.mp3