home_stack/ai_stack/ollama/docker-compose.yml

services:
  ollama:
    image: docker.io/ollama/ollama:latest
    # network_mode: host
    ports:
      - "11434:11434" # Map host port 11434 to container port 11434
    volumes:
      - ./models:/models # Mount a local directory for models
      - ollama_data:/root/.ollama # Persistent storage for Ollama data
    container_name: ollama-server
    environment:
      - OLLAMA_HOST=0.0.0.0
    restart: unless-stopped
    # Uncomment the following lines if you have NVIDIA GPU support
    deploy:
      resources:
        reservations:
          devices:
            - driver: cdi
              capabilities: [gpu]
              device_ids:
                - nvidia.com/gpu=all

  # postgres:
  #   image: docker.io/pgvector/pgvector:pg17
  #   container_name: ollama-postgres
  #   # network_mode: host
  #   ports:
  #     - "54320:5432" # Map host port 54320 to container port 5432
  #   environment:
  #     - POSTGRES_PASSWORD=postgres
  #     - POSTGRES_PORT=54320
  #   volumes:
  #     - ./postgres-data:/var/lib/postgresql/data
  #   restart: unless-stopped

  # speaches:
  #   image: ghcr.io/speaches-ai/speaches:latest-cpu
  #   container_name: speaches-server
  #   # network_mode: host
  #   ports:
  #     - "8900:8000" # Map host port 8900 to container port 8000
  #   volumes:
  #     - hf-hub-cache:/home/ubuntu/.cache/huggingface/hub
  #   environment:
  #     - PORT=8900
    # deploy:
    #   resources:
    #     reservations:
    #       # WARN: requires Docker Compose 2.24.2
    #       # https://docs.docker.com/reference/compose-file/merge/#replace-value
    #       devices: !override
    #         - capabilities: ["gpu"]
    #           driver: cdi
    #           device_ids:
    #             - nvidia.com/gpu=all

volumes:
  ollama_data:
  hf-hub-cache: