D4RL
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
D4RL hopper-medium v2 (test)
99.4Normalized Reward
8
D4RL hopper-medium-replay v2 (test)
9,600Normalized Reward
8
D4RL MuJoCo Walker2d standard (mixed)
89.9Normalized Score
8
D4RL MuJoCo Walker2d standard (random)
17.9Normalized Score
8
D4RL MuJoCo Hopper standard (mixed)
103.9Normalized Score
8
D4RL MuJoCo Halfcheetah standard (mixed)
74.9Normalized Score
8
D4RL MuJoCo Halfcheetah standard (medium)
78.7Normalized Score
8
D4RL MuJoCo Halfcheetah standard (random)
39.5Normalized Score
8
D4RL Walker2d Simultaneous Adversarial Corruption
17.21Avg Score
8
D4RL Halfcheetah Simultaneous Random Corruption
33.04Average Score
8
D4RL MuJoCo v0
35.4HalfCheetah Return (Random)
8
D4RL MuJoCo halfcheetah-medium-expert v0
105Normalized Score
8
D4RL MuJoCo walker2d-medium v0
82Normalized Score
8
D4RL MuJoCo halfcheetah-medium v0
45.1Normalized Score
8
D4RL MuJoCo walker2d-medium-replay v0
42.6Normalized Score
8
D4RL MuJoCo halfcheetah-medium-replay v0
53.1Normalized Score
8
D4RL MuJoCo walker2d-random v0
13.6Normalized Score
8
D4RL MuJoCo halfcheetah-random v0
35.4Normalized Average Score
8
D4RL Average
73.5Average Reward
8
D4RL Walker Medium-Replay
62.9Reward
8
D4RL Walker Medium-Expert
108Reward
8
Stochastic D4RL Walker2d Medium MuJoCo
1,537.3Mean Return
8
Stochastic D4RL Hopper MuJoCo (Mixed)
1,551.2Mean Return
8
D4RL Cheetah Stochastic MuJoCo (Mixed)
396.4Mean Return
8
Stochastic D4RL Cheetah MuJoCo (Medium)
361.4Mean Return
8