Hey all! We just wanted to let you know we have shared two things:
- the inverse dynamics model for predicting players' actions from unlabeled video, and
- an example behavioural cloning code to further train the models
Hey all! We just wanted to let you know we have shared two things: