Katy's Ollama Quickstart

This is just meant to be a small repository to keep my Ollama docker-compose.yml file in sync so I can keep the same environment on both of my machines.

Feel free to take it, use it, fork it, whatever. 🙂

Setup

Prerequisites:

• Docker
• Docker Compose
• NVIDIA Container Toolkit (nvidia-container-toolkit — needed for runtime: nvidia).

Getting it running:

• cp .env.example .env (open up .env and run openssl rand[..] to generate a secret key)
• Make your bind targets with mkdir -p /path/to/docker_volumes/{ollama_data,openwebui_data}.
• Run docker compose up -d.
• Open OpenWebUI at http://localhost:3000.

Recommended models

These are the models I typically pull down:

• glm4:9b
• qwen3.5:9b
• deepseek-coder:6.7b
• translategemma:4b

Links

• ollama/ollama on Docker Hub
• open-webui on GHCR
• ollama/ollama releases
• open-webui releases