services:
  autorag-api:
    image: autorag-base
    container_name: autorag-api
    environment:
      - CUDA_VISIBLE_DEVICES=0
      - OPENAI_API_KEY=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
      - OLLAMA_API_BASE_URL=http://autorag-ollama:11434 # Ollama 컨테이너로 연결결
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
      - ./:/usr/src/app/
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    stdin_open: true
    tty: true
    working_dir: /usr/src/app
    depends_on:
      - autorag-ollama # Ollama가 먼저 실행되도록 설정
    networks:
      - autorag_network
    command: ["bash"] # 기본 실행 명령

  autorag-ollama:
    image: ollama/ollama
    container_name: autorag-ollama
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    ports:
      - "11434:11434" # Ollama API 포트 설정
    networks:
      - autorag_network
    restart: always

networks:
  autorag_network:
    driver: bridge