We ran a workshop at Dinalab about how to run your own “open” AI model locally on your own laptop! This means no paying more billionaires lots more money and no giving away your private info to them! There’s an added bonus that deepseek is supposed to be a LOT more efficient, and thus uses less energy to process prompts!
Follow these instructions below to run your own deepseek locally!
ollama run deepseek-r1or you can download other specific models with this syntax
ollama run deepseek-r1:1.5bmake sure python is installed on your computer so you can use pip:
https://www.python.org/downloads/
pip install open-webui
open-webui usually takes many minutes to install, but afterwards, just run:
open-webui serve
now go to this in your browser:
http://localhost:8080/
create username and password
you are in!
You can download Visual Studio Code to help code and run python scripts too if you want!