> #llm

An interactive tutorial about various compute aspects of LLMs, like hosting or quantizing them. Made by (I think) Charles Frye

This is like an interactive 3Blue1Brown visualization and explanation of complete GPT-style model inference running in your browser. Incredible!