Reinforcement Learning on DMControl finger-spin (expert)
98.42Averaged Normalized ScoreIQL
Evaluation Results
| Method | Links | |
|---|---|---|
| IQLBase Algorithm=IQL, S2P Augmentation=false2022.09 | 98.42 | |
| CQL+S2PBase Algorithm=CQL, S2P Augmentation=true2022.09 | 97.25 | |
| IQL+S2PBase Algorithm=IQL, S2P Augmentation=true2022.09 | 94.42 | |
| SLAC-off+S2PBase Algorithm=SLAC-off, S2P Augmentation=true2022.09 | 81.05 | |
| CQLBase Algorithm=CQL, S2P Augmentation=false2022.09 | 73.07 | |
| SLAC-offBase Algorithm=SLAC-off, S2P Augmentation=false2022.09 | 24.75 |