D4RL Walker2d

Benchmarks

Task Name	Dataset Name	SOTA Result
Offline Reinforcement Learning	D4RL Walker2d Medium v2	Normalized Return94.2	67
Locomotion	D4RL Walker2d medium-offline	Normalized Score37.45	36
Offline Reinforcement Learning	D4RL walker2d medium-replay v2	Normalized Score100.6	36
Locomotion Control	D4RL Walker2d medium-expert	Normalized Return111.2	23
Offline Reinforcement Learning	D4RL Walker2D Expert	Mean Normalized Score117.4	22
Offline Reinforcement Learning	D4RL Walker2d Medium	Normalized Avg Return87.7	18
Continuous Control	D4RL Walker2d medium	Normalized Return81.9	14
Offline Reinforcement Learning	D4RL Walker2d Med-Replay	Normalized Average Return82.6	11
Offline Behavior Distillation	D4RL Walker2D (medium-expert)	Normalized Return109	8
Offline Behavior Distillation	D4RL Walker2D medium	Normalized Return84	8
Offline Reinforcement Learning	D4RL Walker2d Simultaneous Random Corruption	Average Score23.62	8
Offline Reinforcement Learning	D4RL Walker2d Stochastic MuJoCo (Mixed)	Mean Return450	8
Offline Policy Evaluation	D4RL Walker2d medium	RMSE149	7
Offline Reinforcement Learning	D4RL Walker2d random v0	Return412	6
Continuous Control	D4RL Walker2d expert	Normalized Return125.7	5
Offline Inverse Reinforcement Learning	D4RL Walker2d Medium-Expert v2	Cumulative Reward4,049.43	4
Offline Inverse Reinforcement Learning	D4RL Walker2d Medium v2	Cumulative Reward4,121.68	4
Locomotion	D4RL Walker2d-expert	Normalized Score (100k Steps)120.35	3
Off-policy Evaluation	D4RL Walker2D-medium-expert	RMAE0.252	2

Showing 19 of 19 rows