Hugging Face Spaces
Hugging Face Spaces is a hosting service for publishing machine learning demo apps directly on your profile.
Overview
Hugging Face Spaces is a service that hosts machine learning demo apps directly on a personal or organization profile. It also works as an AI app directory where you can browse apps others have uploaded by category, such as image generation, video generation, speech synthesis, OCR and chatbots.
It supports the Gradio SDK out of the box so you can build apps quickly in Python. You can also use Docker to upload any Dockerfile as is, or build a static Space with JavaScript and HTML. If needed, a Space can be upgraded to run on accelerated hardware such as GPUs.
It suits researchers and developers who want others to try their models firsthand. It can also be used to build a machine learning portfolio or to collaborate with others.
What it is used for
- Turn a model or project into a demo app and publish it as part of a machine learning portfolio.
- Demo a project at a conference or to stakeholders.
- Find and try public apps in areas such as image and video generation, speech synthesis and OCR.
- Build and maintain Spaces together with others in the ML ecosystem.
- Link a Space to an arXiv paper or embed a Space in another web page.
Pricing and license
See the official website
Getting started
- Sign up for Hugging Face.
- Click New Space to create a new Space.
- Choose how to build the app: Gradio, Docker or static HTML.
- Upload your code to deploy the app, and if needed, upgrade to GPU hardware in the settings.
Links
Related terms and companies
Sources
Compiled from public sources (official pages, READMEs). Pricing and features change, so check the official site.