Multi-task Reinforcement Learning on Meta-World MT50 (MT50-rand) V2 (Near-optimal)
61.32Avg Success RateMTDIFF-P-ONEHOT
Evaluation Results
| Method | Links | |
|---|---|---|
| MTDIFF-P-ONEHOTTraining Protocol=Offline2023.05 | 61.32 | |
| MTBCTraining Protocol=Offline2023.05 | 60.39 | |
| MTDIFF-PTraining Protocol=Offline2023.05 | 59.53 | |
| PaCoTraining Protocol=Online2023.05 | 57.3 | |
| MTIQLTraining Protocol=Offline2023.05 | 56.21 | |
| CARETraining Protocol=Online2023.05 | 50.8 | |
| PromptDTTraining Protocol=Offline2023.05 | 45.68 | |
| MTDTTraining Protocol=Offline2023.05 | 20.99 |