Is it possible to train agents by competitive self-play in 5v5 or 11v11 scenario currently? If yes, can anyone give a short guide-line to get started?
--
You received this message because you are subscribed to the Google Groups "Google Research Football" group.
To unsubscribe from this group and stop receiving emails from it, send an email to google-research-fo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/google-research-football/e0b3a7a8-54cc-4611-b509-f026320a5bab%40googlegroups.com.