I've re-adapted the implementation from ACT to work on low_cost_robot.
record_episodes.py: stores teleoperation episodes in .hdf5 format
train.py: trains the policies and saves checkpoints in checkpoints/eval.py: evaluates the policy on the robot
Happy to do modifications if there's any feedback/request.
I've re-adapted the implementation from ACT to work on
low_cost_robot.record_episodes.py: stores teleoperation episodes in .hdf5 formattrain.py: trains the policies and saves checkpoints incheckpoints/eval.py: evaluates the policy on the robotHappy to do modifications if there's any feedback/request.