services:
  llm:
    build: .
    container_name: server
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    restart: unless-stopped
    environment:
      - NVIDIA_VISIBLE_DEVICES=all
    volumes:
      - ./models:/models
    ports:
      - 3000:8080
    command: --models-max 1 --models-preset /models/model-presets.ini --fit on --fit-target 1024