Reinforcement Learning on DMControl finger-spin (mixed)
98.54Avg Normalized ScoreCQL
Evaluation Results
| Method | Links | |
|---|---|---|
| CQLBase Algorithm=CQL, S2P Augmentation=false2022.09 | 98.54 | |
| IQLBase Algorithm=IQL, S2P Augmentation=false2022.09 | 98.18 | |
| IQL+S2PBase Algorithm=IQL, S2P Augmentation=true2022.09 | 94.78 | |
| CQL+S2PBase Algorithm=CQL, S2P Augmentation=true2022.09 | 87.17 | |
| SLAC-off+S2PBase Algorithm=SLAC-off, S2P Augmentation=true2022.09 | 83.31 | |
| SLAC-offBase Algorithm=SLAC-off, S2P Augmentation=false2022.09 | 64.41 |