docker-compose.yml

services:
  qwen3-tts:
    build:
      context: .
      args:
        MODEL_NAME: ${MODEL_NAME:-0.6B}
    image: qwen3-tts-fastapi:latest
    container_name: qwen3-tts
    ports:
      - "${EXTERNAL_PORT:-8005}:8880"
    environment:
      - MODEL_NAME=${MODEL_NAME:-0.6B}
      - DEVICE=${DEVICE:-cuda:0}
      - DEFAULT_VOICE=${DEFAULT_VOICE:-default}
      - DEFAULT_FORMAT=${DEFAULT_FORMAT:-mp3}
      - MAX_INPUT_LENGTH=${MAX_INPUT_LENGTH:-4096}
    volumes:
      - ./voices:/app/voices
      - qwen3-tts-cache:/app/cache
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8880/health"]
      interval: 30s
      timeout: 30s
      retries: 3
      start_period: 120s

  gradio-ui:
    build:
      context: ./ui
    image: qwen3-tts-gradio:latest
    container_name: qwen3-tts-gradio
    ports:
      - "${GRADIO_PORT:-7860}:7860"
    environment:
      - API_URL=http://qwen3-tts:8880
    depends_on:
      qwen3-tts:
        condition: service_healthy
    restart: unless-stopped

volumes:
  qwen3-tts-cache: