Offline Reinforcement Learning on AntMaze Large-Play v0
39.6Normalized ScoreIQL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| IQLrandom seeds=52024.10 | 39.6 | — | — | |
| QDQrandom seeds=52024.10 | 35.6 | — | — | |
| CQLrandom seeds=52024.10 | 15.8 | — | — | |
| TD3+BCrandom seeds=52024.10 | 0.2 | — | — | |
| BCrandom seeds=52024.10 | 0 | — | — | |
| DTrandom seeds=52024.10 | 0 | — | — | |
| Onestep RLrandom seeds=52024.10 | 0 | — | — | |
| AWACrandom seeds=52024.10 | 0 | — | — | |
| diff-QLseeds=102023.10 | — | 1.6 | 49 | |
| IQLseeds=10, checkpoint_selection=last 10 average2023.10 | — | 39.6 | — | |
| MSGseeds=10, checkpoint_selection=last 10 average2023.10 | — | 46.8 | — | |
| SEEMseeds=102023.10 | — | 65.4 | 74 | |
| sfBCseeds=10, checkpoint_selection=last 10 average2023.10 | — | 63.3 | — | |
| TD3+BCseeds=10, checkpoint_selection=last 10 average2023.10 | — | 0 | — |