ResearchBenchmarksOffline Reinforcement Learning on D4RL Medium-Expert v2Follow90.7HalfCheetah Normalized ScoreTD3+BC86.64487.69788.7589.803Jun 12, 2021Evaluation ResultsMethodMethodLinksHalfCheetah Normalized ScoreHopper Normalized ScoreWalker2d Normalized ScoreTD3+BCseeds=5, hyperparamete...seeds=5, hyperparameter tuning=none, evaluation protocol=average over final 10 evaluations2021.0690.798110.1DTseeds=3seeds=32021.0686.8107.6108.1