Ollama + OpenWebUI + SearXNG Stack

This repository contains a Docker Compose setup for running Ollama (LLM server), OpenWebUI (UI for interacting with LLMs), and SearXNG (privacy-focused search engine) together.

Features

• 🚀 One-click deployment of the entire AI stack
• 🧠 Run open-source language models locally
• 🌐 Beautiful web interface for chatting with models
• 🔍 Integrated search capabilities
• 🎮 NVIDIA GPU acceleration support
• 💾 Persistent storage for all components
• 🔄 Easy update and backup scripts

Prerequisites

• Docker and Docker Compose
• NVIDIA Docker runtime (for GPU acceleration)
• Git

Quick Start

1. Clone this repository:

   git clone https://github.com/yourusername/ollama-webui-stack.git
   cd ollama-webui-stack

2. Start the stack:

   ./scripts/start-stack.sh

3. Access the interfaces:

   • OpenWebUI: http://localhost:3000
   • SearXNG: http://localhost:8080
   • Ollama API: http://localhost:11434

Configuration

Copy the example environment file and modify as needed:

cp .env.example .env

Edit the .env file to configure:

• Ports
• Base URLs
• GPU settings

Directory Structure

ollama-webui-stack/
├── docker-compose.yml     # Main configuration file
├── .env.example           # Example environment variables
├── .gitignore             # Git ignore file
├── README.md              # This file
├── data/                  # OpenWebUI data directory
├── ollama-data/           # Ollama data directory
├── searxng-data/          # SearXNG configuration directory
└── scripts/               # Utility scripts
    ├── backup.sh          # Backup script
    ├── run-ollama-docker.sh # Run Ollama standalone
    ├── start-stack.sh     # Start the entire stack
    └── update.sh          # Update Docker images

Usage

Starting the Stack

./scripts/start-stack.sh

Updating the Stack

./scripts/update.sh

Backing Up Data

./scripts/backup.sh [backup_directory]

Models

To download and use models with Ollama:

1. Access OpenWebUI at http://localhost:3000
2. Navigate to the Models section
3. Choose from available models to download

Popular models include:

• llama3
• mistral
• phi
• gemma

Troubleshooting

GPU Issues

If you're having GPU issues:

1. Ensure NVIDIA drivers are properly installed
2. Verify NVIDIA Docker runtime is configured
3. Check Docker logs: docker-compose logs ollama

Connection Issues

If components can't connect to each other:

1. Check that all containers are running: docker-compose ps
2. Inspect logs: docker-compose logs
3. Verify network configuration in docker-compose.yml

Maintenance

Stopping the Stack

docker-compose down

Checking Logs

docker-compose logs -f

License

This project is distributed under the MIT License. See the LICENSE file for more information.

Acknowledgements

• Ollama
• OpenWebUI
• SearXNG