Data Science Special Topics: Working with Local LLMS

Join us for a hands-on introduction to running AI models on your own computer instead of someone else's. We'll use Ollama, a free tool that makes downloading and running open-source models straightforward, and work through the whole path: installing it, pulling your first model, chatting with it, and swapping in a second to see how models differ. Whether you code or not, you'll come away understanding why researchers increasingly want a local option — sensitive data never leaves your machine, there's no per-use cost, and results stay reproducible because the model doesn't change under you — along with the honest trade-offs in speed and capability. This session takes place online; you'll leave with a model running locally and enough footing to try others on your own.

Before the session: please look up your computer's specifications, in particular how much memory (RAM) it has and whether it has a dedicated graphics card. On a Mac, check the Apple menu → About This Mac; on Windows, open Settings → System → About. What you can run locally depends on your hardware, so knowing this lets us help you choose a model that will actually run well on your machine. You don't need a powerful computer to attend — there are models that run comfortably on modest hardware, and you're welcome to follow along or simply watch.