Buildathons.ai

HuggingFace: Model Hub & Inference API Course

Discover the world's largest open-source model repository. Deploy models via the free Inference API, host demos with Spaces, and fine-tune foundation models.…

Course overview

Discover the world's largest open-source model repository. Deploy models via the free Inference API, host demos with Spaces, and fine-tune foundation models.

What is HuggingFace?

HuggingFace is the GitHub of AI — a platform hosting 500,000+ pre-trained models, 50,000+ datasets, and interactive ML demos called Spaces. It powers everything from ChatGPT alternatives to image generators. Key concepts: Model Hub: Browse, download and deploy pre-trained models (text, image, audio, video) Spaces: Free hosting for Gradio and Streamlit apps — perfect for demos Inference API: REST API to run any Hub model in the cloud instantly, no GPU required

Using the Inference API

The Inference API lets you call models without any setup. Just grab a free token from huggingface.co/settings/tokens and hit the API.

Deploying a HuggingFace Space

Spaces let you ship a live web app backed by any model in minutes. You can use Gradio (Python) or Streamlit — both deploy free on HuggingFace's GPU infrastructure.