Multi-Agent Control on SMAX SMACv2 (smacv2_10_units)
75Average Win RateGPAE
Evaluation results
| Method | Links | |
|---|---|---|
| GPAESampling Protocol=off-policy, Timesteps=1 x 10^7, Random Seeds=52026.03 | 75 | |
| GPAESampling Protocol=on-policy, Timesteps=1 x 10^7, Random Seeds=52026.03 | 69 | |
| DAETimesteps=1 x 10^7, Random Seeds=52026.03 | 65.7 | |
| MAPPOTimesteps=1 x 10^7, Random Seeds=52026.03 | 63.5 | |
| QMIXTimesteps=1 x 10^7, Random Seeds=52026.03 | 62.1 | |
| COMATimesteps=1 x 10^7, Random Seeds=52026.03 | 47.4 | |
| VDNTimesteps=1 x 10^7, Random Seeds=52026.03 | 29.5 |