Abstract: Offline reinforcement learning (RL) aims to learn effective agents from previously collected datasets, facilitating the safety and efficiency of RL by ...
It is not only Wrapper Offline, it is the rest of the programs that were taken from here.