Orion: An Open-source Framework for Validity and ZK ML ✨

🚨⚠️🚨⚠️🚨⚠️

This project has been archived and is no longer maintained. We're now developing LuminAIR, a new and super-efficient zkML framework based on a custom AIR, proven with the STWO Prover.

We'd like to thank all the Orion contributors and hope to see you again in this new journey.

🚨⚠️🚨⚠️🚨⚠️

---

Orion is an open-source, community-driven framework dedicated to Provable Machine Learning. It provides essential components and a new ONNX runtime for building verifiable Machine Learning models using STARKs.

🤔 What is ONNX Runtime?

ONNX (Open Neural Network Exchange), is an open-source standard created to represent deep learning models. The aim of its development was to enable interoperability among diverse deep learning frameworks, like TensorFlow or PyTorch. By offering a universal file format, ONNX allows models trained in one framework to be readily applied in another for inference, eliminating the need for model conversion.

Ensuring compatibility with ONNX operators facilitates integration into the ONNX ecosystem. This enables researchers and developers to pre-train models using their preferred framework, before executing verifiable inferences with Orion.

🌱 Where to start?

You can check our official docs here.

• 🧱 Framework: The building blocks for Verifiable Machine Learning models.
• 🏛 Hub: A curated collection of ML models and spaces built by the community using Orion framework.
• 🎓 Academy: Resources and tutorials for learning how to build ValidityML models using Orion.

💖 Join the community!

Join the community and help build a safer and transparent AI in our Discord!

🚀 Orion Usage

• For an insightful overview of impressive proof of concepts, models, and tutorials created by our community, please visit Orion Usage.
• Discover a curated list of tutorials and models developed using Orion in Orion-Hub.

✍️ Authors & contributors

For a full list of all authors and contributors, see the contributors page.

License

This project is licensed under the MIT license.

See LICENSE for more information.

Contributors ✨

Thanks goes to these wonderful people:

This project follows the all-contributors specification. Contributions of any kind welcome!