Introduction to Hugging Face
Get introduced to Hugging Face. Learn about model hub, datasets and spaces.
What is Hugging Face?
Hugging Face is an open-source data science and machine learning platform. It is most well known for being a place where users can host their own AI models, train them and share them with others. Hugging Face makes it simple to share these models in just a few lines of code, and tailor them for your use case. However, it is not limited to just this, as there are also many datasets that can be downloaded and shared, as well as spaces to demo your work. In this lesson, we are going to look at key features of Hugging Face, as well as looking at how to use models shared by others.
Uses of Hugging Face
We will cover the three main uses for Hugging Face, namely the models, datasets and spaces. These can all be found in the top ribbon of the Hugging Face website, as shown below.
Let’s look into each one of these.
Model Hub
The Model Hub contains hundreds of thousands of models trained and uploaded by the community. These include models for things like text summarization, question and answering, text generation, translation and many more. Many of these models have billions of parameters and are truly large and powerful. Hugging Face has made it simple to filter through these models to find the one that is suitable for a particular use case, and usually has a sample code for how this can be integrated into a custom pipeline.