Learning walk and trot from the same objective using different types of exploration

Open in new window