ResearchBenchmarksOffline Reinforcement Learning on D4RL Medium-Replay v2Follow44.6HalfCheetah ScoreTD3+BC36.2838.4440.642.76Jun 12, 2021Evaluation ResultsMethodMethodLinksHalfCheetah ScoreHopper ScoreWalker2d ScoreTD3+BCseeds=5, hyperparamete...seeds=5, hyperparameter tuning=none, evaluation protocol=average over final 10 evaluations2021.0644.660.981.8DTseeds=3seeds=32021.0636.682.766.6