Polygeemma

Software / App

An open-source, three-billion-parameter vision-language model used as a foundation for a robot control system. It takes images and language commands as input to predict future actions.

Mentioned in 1 video