Online Reinforcement Learning on Antmaze umaze-diverse
99.9ScoreBC-PEX
Evaluation Results
| Method | Links | |
|---|---|---|
| BC-PEXEnvironment Steps=500k2023.10 | 99.9 | |
| RLPDReward Setting=with true reward, Environment Steps=500k2023.10 | 99.9 | |
| UBEREnvironment Steps=500k2023.10 | 98.8 | |
| BC-PEXEnvironment Steps=100k2023.10 | 95.5 | |
| UBEREnvironment Steps=100k2023.10 | 94.5 | |
| RLPDReward Setting=with true reward, Environment Steps=100k2023.10 | 83.3 | |
| RLPDReward Setting=without true reward, Environment Steps=500k2023.10 | 0 | |
| RLPDReward Setting=without true reward, Environment Steps=100k2023.10 | 0 |