docker-compose.yml

services:
  toy-llm:
    build:
      context: .
      dockerfile: dockerfiles/Dockerfile
    image: toy-llm:latest  # 构建后镜像的名称和标签   
    environment:
      APP: toy-llm
      TAVILY_API_KEY: ${TAVILY_API_KEY}
      OLLAMA_BASE_URL: http://host.docker.internal:11434
    extra_hosts:
      - "host.docker.internal:host-gateway"
    networks:
      - default  
    volumes:
      - "~/.ssh:/root/.ssh:ro" # 挂载宿主机的ssh配置文件，方便推拉代码（只读）
      - "/etc/localtime:/etc/localtime:ro" # 挂载宿主机时间到容器
      - ".:/root/toy-llm"
    working_dir: /root/toy-llm
    command: ["sleep", 'infinity']

  
  # ollama-server:
  #   image: ollama/ollama:latest
  #   pull_policy: always
  #   restart: unless-stopped
  #   entrypoint: ["/usr/bin/bash", "/entrypoint.sh"]
  #   environment:
  #     - OLLAMA_MODELS=${OLLAMA_MODELS:-qwen2.5:3b}
  #     # '/Users/henryzhu/data/qwen2.5:0.5b'
  #   networks:
  #     - default
  #   volumes:
  #     - ./scripts/entrypoint.sh:/entrypoint.sh # 用于修改 entrypoint.sh 文件
  #     # - ollama_volume:/root/.ollama # 貌似不需要挂载   
  #   ports:
  #     - "11434:11434"
  #   deploy:
  #     resources:
  #       reservations:
  #         devices:
  #           - driver: nvidia
  #             count: 1
  #             capabilities: [gpu]
  #   healthcheck:
  #     test: ["CMD-SHELL", "curl --silent --fail http://localhost:11434/ || exit 1"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 5
  #     start_period: 30s
  #   logging:
  #     driver: "json-file"
  #     options:
  #       max-size: "10m"
  #       max-file: "3"

  open-webui:
    image: ghcr.io/open-webui/open-webui:main
    ports:
      - "3000:8080" # http://localhost:3000
    environment:
      - OLLAMA_BASE_URL=http://ollama-server:11434
      - WEBUI_NAME="Your Local AI Assistant" # Sets the main WebUI name. Appends (Open WebUI) if overridden.
      - WEBUI_AUTH=False
      - WEBUI_SECRET_KEY=t0p-s3cr3t
    restart: unless-stopped


# -- 网络名称，用于容器间通信
networks:
  default:
    name: toyllm-network