An explanation of Hugging Face

In this video, TechTarget editor Jen English explains the Hugging Face machine learning and data science platform.

Looking for free AI models that don't belong to Google or OpenAI? Hugging Face is a machine learning and data science platform that lets users build, deploy and train machine learning models. It hosts thousands of open source ML models, data sets and demos. This means you can see and use the code behind the models, unlike Google's Bard and OpenAI's ChatGPT, which are closed sourced.

Developers can use Hugging Face to create and post the code for their own AI models, so its repository is continuously growing. Hugging Face also hosts a public LLM leaderboard that tracks, ranks and evaluates the LLMs and chatbots on the platform.

But it's not limited to language models. It also has computer vision models, audio models and image models -- so you can build text to image, or image to image generators, for example.

The biggest upside of Hugging Face is that it's an open source community with thousands of developers iterating and improving upon each other's work. But it can be difficult to find the right model because there are so many to choose from. Models on Hugging Face also share the same limitations as any other machine learning model -- they're subject to bias and can be costly to run and maintain.

Hugging Face is free to sign up for but has a paid enterprise offering, with additional features like dedicated customer support.

Ben Lutkevich is the site editor for Software Quality. Previously, he wrote definitions and features for Whatis.com.