Offline Reinforcement Learning on Hopper Medium-Replay JointNoise Shift
93.704Average Return1T10S
Evaluation Results
| Method | Links | |
|---|---|---|
| 1T10SFramework=QT2024.10 | 93.704 | |
| QTAugmentation Strategy=1T10S2024.10 | 93.704 | |
| REAG*MVFramework=QT2024.10 | 93.409 | |
| QTAugmentation Strategy=REAG_MV2024.10 | 93.409 | |
| REAG*DaraFramework=DT2024.10 | 83.525 | |
| DTAugmentation Strategy=REAG_Dara2024.10 | 83.525 | |
| REAG*MVFramework=DT2024.10 | 77.825 | |
| DTAugmentation Strategy=REAG_MV2024.10 | 77.825 | |
| 1T10SFramework=DT2024.10 | 61.87 | |
| DTAugmentation Strategy=1T10S2024.10 | 61.87 | |
| REAG*DaraFramework=Reinformer2024.10 | 52.052 | |
| ReinformerAugmentation Strategy=REAG_Dara2024.10 | 52.052 | |
| REAG*DaraFramework=QT2024.10 | 51.456 | |
| QTAugmentation Strategy=REAG_Dara2024.10 | 51.456 | |
| REAG*MVFramework=Reinformer2024.10 | 43.985 | |
| ReinformerAugmentation Strategy=REAG_MV2024.10 | 43.985 | |
| 1T10SFramework=Reinformer2024.10 | 41.82 | |
| ReinformerAugmentation Strategy=1T10S2024.10 | 41.82 |