Skip to content

Latest commit

 

History

History
executable file
·
11 lines (9 loc) · 805 Bytes

README.md

File metadata and controls

executable file
·
11 lines (9 loc) · 805 Bytes

Multi-Agent Reinforcement Learning

DQN, DDPG and MADDPG implementations to play tag in OpenAI's multi-agent particle environment.

DQN 1 vs 1 DQN 1 vs 2 DQN 2 vs 1
DDPG 1 vs 1 DDPG 1 vs 2 DDPG 2 vs 1
MADDPG 1 vs 1 MADDPG 1 vs 2 MADDPG 2 vs 1