complex task training

[Saluze]

Hello, have you ever tried applying multi-objective reinforcement learning algorithms to complex tasks? I am currently attempting to use them in robot locomotion, but due to the complexity of the rewards, I don't know how to utilize this method.