Install Ollama, run LLaMA 3.2 on your own computer and call it from Python, with practical tips from using local LLMs in a real project.