Offline Reinforcement Learning on Adroit v1 (cloned-total)
66.3Average Normalized ScoreTSRL
Evaluation Results
| Method | Links | |
|---|---|---|
| TSRLRatio=1, Size=500k2023.06 | 66.3 | |
| BCRatio=1, Size=500k2023.06 | 57.5 | |
| DOGERatio=1, Size=500k2023.06 | 56.7 | |
| TSRLRatio=1/50, Size=10k2023.06 | 44.9 | |
| CQLRatio=1, Size=500k2023.06 | 41.6 | |
| TD3+BCRatio=1, Size=500k2023.06 | 41.1 | |
| IQLRatio=1, Size=500k2023.06 | 40.8 | |
| IQLRatio=1/50, Size=10k2023.06 | 32.7 | |
| DOGERatio=1/50, Size=10k2023.06 | 29.7 | |
| BCRatio=1/50, Size=10k2023.06 | 29.5 | |
| MOPORatio=1, Size=500k2023.06 | 1.2 | |
| CQLRatio=1/50, Size=10k2023.06 | 0.6 | |
| TD3+BCRatio=1/50, Size=10k2023.06 | 0.2 | |
| MOPORatio=1/50, Size=10k2023.06 | -1.7 |