The situation of "multiple learners" is in which code file?

[zonewenke]

Hello, community!
I came across Key Graphs on the introduction page of the impala algorithm (https://di-engine-docs.readthedocs.io/zh-cn/latest/12_policies/impala.html) and am quite interested in the scenario of multiple learners. However, I couldn't find any code files related to multiple learners in the DI-engine project's codebase.

I am currently running an experiment with SpaceInvaders (in the new pipeline), and the configuration file is (https://github.com/opendilab/DI-engine/blob/main/dizoo/atari/config/serial/spaceinvaders/spaceinvaders_impala_config.py), but I am not sure whether the configuration in this file is for a single learner or for multiple learners.
If anyone knows, please let me know, I would be very grateful!

[PaParaZz1]

This demo spaceinvaders_impala_config.py is a single-worker example. For the multi-learner case, here is a basic example implemented with PyTorch DDP, you can refer to it for more details. If you are interested in other distributed RL cases, you can continue to ask in this discussion.

[zonewenke]

Thank you very much for your answer; it's very timely and helpful to me.

[zonewenke]

Thank you very much for your timely response, it has been a great help to me!

…

---Original--- From: ***@***.***> Date: Mon, Jul 22, 2024 10:54 AM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [opendilab/DI-engine] The situation of "multiple learners" is inwhich code file? (Discussion #819) This demo spaceinvaders_impala_config.py is a single-worker example. For the multi-learner case, here is a basic example implemented with PyTorch DDP, you can refer to it for more details. If you are interested in other distributed RL cases, you can continue to ask in this discussion. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>