Document how to configure to use a self hosted ollama llm

[GregHilston]

For example, say I'm running ollama on the same machine as this container. Or even another machine, how can J configure this project to use that llm? And skip using Claude and chat gpt