Offline Imitation Learning with Suboptimal Demonstrations via Relaxed Distribution Matching

Open in new window