ollama-docker

This repository contains configuration files to start two Docker containers: one for Ollama and one for Open WebUI. Both services are based on Docker images with NVIDIA support.

• Docker version 20.10.7 or higher.
• Docker Compose version 1.29.2 or higher.
• NVIDIA drivers installed for Docker (if necessary).
• NVIDIA Container Toolkit installed and configured.

The Ollama service uses the ollama/ollama image and is exposed on port 11434.

The Open WebUI service uses the ghcr.io/open-webui/open-webui:cuda image and is exposed on port 3000 (mapped internally to port 8080 in the container).

To start the services, run the following command in the directory containing the docker-compose.yml file:

docker compose up -d

Now you have a Docker container ready to run the Open WebUI. Access the Open WebUI application at http://localhost:3000 in your web browser.

To stop the services, run the following command:

docker compose down

You can view the logs of the containers with the following commands:

docker logs ollama
docker logs open-webui

• ollama/: Local directory for Ollama configuration and data.
• open-webui/: Local directory for Open WebUI configuration and data.

• NVIDIA Container Toolkit Installation Guide
• Docker Official Website

This project is the core of manzolo's Local AI Lab — a family of self-hosted AI projects (LLM, voice, vision & documents) that share the same conventions.

This stack owns the shared local-ai-net Docker network: start it first, then attach any sibling project with its docker-compose.local-ai.yml override. From inside the network, the Ollama API is always at http://ollama:11434.

Explore the whole family: topic:local-ai