Offline Reinforcement Learning on Antmaze umaze
96.7Average ReturnRORL
Evaluation results
| Method | Links | |
|---|---|---|
| RORL2022.06 | 96.7 | |
| SPQRBase algorithm=CQL-Min2024.01 | 93.3 | |
| IQL2023.07 | 92.8 | |
| IQL+smoothing2022.06 | 92.3 | |
| GCIQL2023.07 | 91.6 | |
| WGCSL2023.07 | 90.8 | |
| CQL2023.07 | 88.2 | |
| IQL2022.06 | 87.5 | |
| IQL2024.01 | 87.5 | |
| TD3+BC2022.06 | 78.6 | |
| CQL2024.02 | 74 | |
| CQL2022.06 | 74 | |
| CQL-Min2024.01 | 74 | |
| BEAR2024.01 | 73 | |
| DWSL2023.07 | 71.2 | |
| GCPC2023.07 | 71.2 | |
| RvS-G2023.07 | 70.4 | |
| ROMI+BCQ2022.06 | 68.7 | |
| BC2024.01 | 65 | |
| BC2023.07 | 63.4 | |
| AWAC2022.06 | 56.7 | |
| DT2023.07 | 55.6 | |
| BC2022.06 | 54.6 | |
| SAC-Min2024.01 | 0 |