Multi-Agent Reinforcement Learning on Google Football Research ca_hard
0.8581Winning RateATD-MAPPO
Evaluation results
| Method | Links | |
|---|---|---|
| ATD-MAPPOTime steps=50M, Evaluation processes=32, Seeds=102026.05 | 0.8581 | |
| MAPPOTime steps=50M, Evaluation processes=32, Seeds=102026.05 | 0.8006 | |
| OW-QMIXTime steps=50M, Evaluation processes=32, Seeds=102026.05 | 0.0855 | |
| QMIXTime steps=50M, Evaluation processes=32, Seeds=102026.05 | 0.0234 | |
| QPLEXTime steps=50M, Evaluation processes=32, Seeds=102026.05 | 0.0021 | |
| MACPFTime steps=50M, Evaluation processes=32, Seeds=102026.05 | -0.0123 |