Distributed/Parallel Environments Training

facebookresearch / Pearl

A Production-ready Reinforcement Learning AI Agent Library brought by the Applied Reinforcement Learning team at Meta.

MIT License

2.67k stars 165 forks source link

Distributed/Parallel Environments Training #101

Closed SestoAle closed 2 months ago

SestoAle commented 2 months ago

Hello,

I was wondering if the framework allows for parallel/distributed environments. The paper says pearl supports distributed training but I can't find any examples or documentation on how to do that. Not even for running multiple environments in the same machine.

What is the plan of pearl for supporting distributed/parallelised environments training?

Thanks!

rodrigodesalvobraz commented 2 months ago

Hi. While we have designed Pearl with distributed processing in mind, it is true that this is not currently supported yet, and probably won't be for some time. We are preparing a public roadmap indicating the next steps for Pearl and will address it there.