Offline Reinforcement Learning on D4RL MuJoCo walker2d random v2
23.6Normalized ScoreMOBILE
Evaluation results
| Method | Links | |
|---|---|---|
| MOBILEtraining steps=1M, seeds=5, version=v22026.03 | 23.6 | |
| ROMItraining steps=1M, seeds=5, version=v22026.03 | 22.4 | |
| ADMPO-OFF2024.05 | 22.2 | |
| COUNTtraining steps=1M, seeds=5, version=v22026.03 | 19.9 | |
| MOBILE2024.05 | 17.9 | |
| CBOP2024.05 | 17.8 | |
| EDAC2024.05 | 16.6 | |
| MOPO2024.05 | 7.4 | |
| COMBO2024.05 | 7 | |
| CQL2024.05 | 5.4 | |
| IQLtraining steps=1M, seeds=5, version=v22026.03 | 5.4 | |
| CQLtraining steps=1M, seeds=5, version=v22026.03 | 5.1 | |
| TD3+BC2024.05 | 1.6 | |
| BC2024.05 | 1.3 | |
| RAMBO2024.05 | 0 | |
| MOPOtraining steps=1M, seeds=5, version=v22026.03 | 0 | |
| RAMBOtraining steps=1M, seeds=5, version=v22026.03 | 0 |