Are you planning to support Multiagent Reinforcement Learning frameworks?

[kailashg26]

Hi, thanks for open-sourcing the Prim benchmarks. I'm doing some analysis on MARL (Multi-agent Reinforcement learning) frameworks. I see that most of the MARL algorithms have relatively smaller network size. But there some more compute intensive operations which are performed in MARL algorithms. Is there any guide or starting point you could provide if PIM really is a good fit for MARL algorithms to resolve the scalability issues?