Configuration files and model definitions for distributed OpenWebUI deployment.
services/
├── openwebui/ # Main application configuration
│ ├── .env # Environment variables
│ └── compose.yaml # Docker Compose configuration
├── ollama/ # Model management
│ ├── models/ # Model pull scripts by category
│ └── planning/ # Model inventory and planning
└── searxng/ # Search engine configuration
├── settings.yml # SearXNG configuration
└── uwsgi.ini # UWSGI server settings
- Configure environment variables in
.envfiles - Deploy services via Docker Compose
- Pull models using category-specific scripts
- OpenWebUI: Primary interface with GPU acceleration
- Ollama: Model serving across multiple hosts
- pgvector: Vector database for embeddings
- SearXNG: Privacy-focused search integration
- Auxiliary services: Tika, Playwright, Gotenberg for document processing
- Docker with GPU support
- NVIDIA runtime
- ZFS datasets for persistent storage
- Network configuration for multi-host deployment