PlatformMarketplaceSolutionsResourcesOpen DatasetsCommunityCompany

Welcome to Graviti AI Community 2.0

Published at2022-05-05

Hello World,

We are excited to announce that Graviti AI Community 2.0 is now launched and we’re here to share more details with you on the community's future plan!

Why community 2.0?

Since our inception in 2019, Graviti AI community has been devoted to making datasets more accessible and interoperable to AI developers, fostering a supportive community for building machine learning applications. As of now, we have 1,300+ datasets and 27K+ registered users globally. Thank you for being with us!

Through standardizing dataset formats, our community enables machine learning researchers and practitioners to easily upload your datasets or folk open datasets, and easily train models with experimenting tools on datasets. We are in a vision to push the boundaries of AI, as revealed by the community slogan, so we broaden our scope from datasets to machine learning models and applications, and more importantly, advocate community-generated datasets, models and applications in a decentralized way ( 🌟Community members' success is mission-critical to us! 🌟). Are you all ready for the voyage with us?

What’s exactly in for you?

Vitalizing open datasets is our first step to pushing the boundaries of AI. Here are some of the existing dataset features in the community.

  1. Unify data formats to improve the interoperability of open datasets. Community developers can easily experiment on nearly 100 classic datasets and train models without the need of understanding various data structures.
  2. Out-of-box tools for dataset contributors to curate, upload and exhibit your own datasets, and tools for dataset users to visualize, assess and train on datasets.
  3. (Coming soon) Our open-sourced schema language Portex will enable a new way of organizing data - a columnar data table for unstructured data. (Image all binary data, metadata, and semantic data of one dataset organized in one single place with customized query functions.)

Dataset always comes with models. Advocating community members' machine learning models and applications is another mission. We believe great work comes from the crowd when they have sufficient support and a stage to shine, which in a way will solve many long-tailed problems in AI production.

  1. Experiment models through our evaluation tool Sextant and work orchestration tool Action.
  2. Share your machine learning experience (Wanna polish your skills? Let the community help you) and your completed projects or applications on our Notebook. We will advocate the top-upvoted works from the community on our platform and beyond our community.
  3. (On our roadmap) We will partner with challenge hosts to provide opportunities for the community members to showcase their models on the benchmark dataset.
  4. (On our roadmap) Similar to Github Apps, we would like to turn community members' great applications into ready-to-run code or APIs and promote them on and beyond our platform (Marketplace). The contributor will be compensated in a way for each call!

How can you participate in the community?

  • Connect with the community on Discord.  Join our Discord server! We'd love to hear your feedback to help you, and help us!
  • Sign up today to try our cool features.
  • Want to contribute? Here are the guides to get started👇
  1. How to contribute datasets
  2. How to contribute Machine Learning code and projects

If you prefer emails to communicate, please contact us at or subscribe to our newsletters!

-Graviti AI Community Team