Offline Goal-conditioned Reinforcement Learning on D4RL AntMaze umaze-diverse v2
0.776Normalized ScoreContrastive RL + BC
Evaluation Results
| Method | Links | |
|---|---|---|
| Contrastive RL + BCTD Learning=false, Number of networks=52022.06 | 0.776 | |
| Contrastive RL + BCTD Learning=false, Number of networks=22022.06 | 0.754 | |
| TD3+BC*TD Learning=true2022.06 | 0.714 | |
| IQL*TD Learning=true2022.06 | 0.622 | |
| GCBCTD Learning=false2022.06 | 0.609 | |
| DTTD Learning=false2022.06 | 0.512 | |
| BCTD Learning=false2022.06 | 0.456 |