Fine-tune GR00T-N1 from Nvidia on a LeRobot dataset
Let's you train an ACT policy from a LeRobot dataset