Folders:

• experiments: The code for our method. It also contains the code for MADDPG baseline method.
• dd0_aamas2021: DeepDrive-Zero environment.
• multiagent-particle-envs: Environments from MADDPG. We modified one of them and used it in MACRPO.
• parametersharingmadrl: the code for baselines used in our paper from Revisiting Parameter Sharing In Multi-Agent Deep Reinforcement Learning.