When
Curious about running AI models directly on your laptop without paying subscriptions or sending personal data to third-party clouds? This beginner-friendly, hands-on workshop walks you through the entire process step-by-step. You will see how open-source models work locally and learn how to manage them on standard personal hardware.
During the session, we will run and compare two different model sizes—an agile lightweight model (~1B) and an intermediate model (~12B)—so you can directly observe how model scale influences response quality, execution speed, and computer memory. In-person support will be available throughout the room to help get everything up and running.
What We Will Cover:
- Why Run Locally: Understand the practical benefits of open-source AI: full data privacy, offline availability, zero API fees, and total system control.
- Guided Installation: Step-by-step setup using Ollama, the easiest runtime for running models on macOS, Windows, and Linux.
- Hands-On Comparison: Running 1B vs. 12B parameter models side by side to experience firsthand how different model sizes perform on your hardware.
- Demystifying Hardware: Clear rules of thumb for how much RAM, CPU, or GPU power different models actually require.
What to Bring:
- Just bring your laptop (Mac, Windows, or Linux) and charger! (If you do not have one, we have 6-year-old Laptops ready to play)
- Optional: If you want to jump right in without waiting on room Wi-Fi, you are welcome to install Ollama from ollama.com beforehand, but we will happily walk through setup together during the session.
Where: AI+ Studio, Weaver Science-Engineering Library, Room 212