Hi there! At AutoAI, we're building gears for the future of machine learning - A more automatic, more accessible and more flexible machine learning. Initially we were funded by the innovator fellowship from ETH Library Lab for the project AID, and later on continued as a free community for anyone to work together on demoncratizing machine learning.

• Project AID: With Project AID, we are trying to build a platform that everyone can upload, share and reuse machine learning models. It is like the App Store, the NPM, and the Python Package Index, but for machine learning models. Documents and Model Hub.
• ShiFTML: ShiftML is an efficient and flexible search engine for transfer learning. Documents.
• TinyNet: TinyML is a simple, lightweight machine learning framework implemented in pure Python, for educational purpose. Documents.

• Data Cleaning Challenge @ DataPerf: With the data cleaning challenge, we hope to direct human attention (cleaning efforts) toward data examples that matter more to the improvement of ML models.

• GMService: A modular framework for serving generative models.
• FFMZip: FFMZip is a system for efficiently serving finetuned foundation models.
• FMEngine: Experimental Training/Serving framework for foundation models.

Please stay tuned for more news and updates :)

Join our Discord and talk to us!