Hugging Face, Hands On

Hugging Face is where most open models, datasets and demo apps live, and it is the fastest way to go from "I read about this model" to "it is running on my machine". This module starts with the fun part: in the first lesson you try a live app, chat with a small model in your terminal, and run three-line Python programs that label text, recognise images and transcribe speech, before any theory.

Then it teaches the skills that make you independent. You learn to read a model page the way an engineer does (who made it, what it does, what license it carries, which file to download and which file not to trust), to work out whether a model fits your laptop before you download gigabytes, and to choose between the four ways to run a model: a hosted API, a local app, Python, or the browser. The last two lessons cover accounts, tokens, the download cache and datasets, and end with you building and sharing your own Space. Every command and code sample in this module was run for real, and the outputs shown are the ones it produced.