Attempt to replicate some of the research work in Multi-Agent Reinforcement Learning inside the Melting Pot testing suite