Offline Reinforcement Learning on DMC medium-expert
848WW to WDCoWorld
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| CoWorld2023.05 | 848 | 774 | 919 | 871 | 475 | 844 | 789 | |
| DrQ+BC2023.05 | 808 | 762 | 808 | 862 | 454 | 730 | 737 | |
| DV2 Finetunemode=fine-tuned2023.05 | 784 | 671 | 851 | 858 | 428 | 833 | 738 | |
| LOMPO2023.05 | 548 | 449 | 688 | 174 | 19 | 113 | 332 | |
| Offline DV2mode=offline2023.05 | 450 | 141 | 214 | 248 | 3 | 48 | 184 |