services:
  bert-all-services:
    build:
      context: .
      dockerfile: Dockerfile
    container_name: bert-all-services
    ports:
      - "5002-5010:5002-5010"
    env_file:
      - .env
    environment:
      - DEVICE=${DEVICE:-cuda}
      - MAX_WORKERS=${MAX_WORKERS:-4}
      - BATCH_SIZE=${BATCH_SIZE:-32}
      - NVIDIA_VISIBLE_DEVICES=${NVIDIA_VISIBLE_DEVICES:-all}
      - NVIDIA_DRIVER_CAPABILITIES=${NVIDIA_DRIVER_CAPABILITIES:-all}
      - MODEL_CACHE_DIR=/app/model_cache
      # 传递代理环境变量（如果构建时使用了代理）
      - HTTP_PROXY=${HTTP_PROXY:-}
      - HTTPS_PROXY=${HTTPS_PROXY:-}
      - http_proxy=${http_proxy:-}
      - https_proxy=${https_proxy:-}
      - NO_PROXY=${NO_PROXY:-}
      - no_proxy=${no_proxy:-}
    volumes:
      - ./logs:/app/logs
      - ./models:/app/models
      - ./model_cache:/app/model_cache
    restart: unless-stopped
    deploy:
      resources:
        limits:
          memory: 8G
        reservations:
          memory: 4G
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]