Deep Reinforcement Learning algorithms for Starcraft II (Pytorch).
Mini games: Move to Beacon and Collect Mineral Shards (You need to download these beforehand)
| __ __TABLE | MoveToBeacon(DQN) | MoveToBeacon(A2C) | Vanilla Deepmind Scores |
|---|---|---|---|
| MEAN | 21.5 | 23.75 | 26 |
| MAX | 28 | 29 | 45 |
We mainly followed StarCraft II: A New Challenge for Reinforcement Learning paper