Hi all, I’m new in ray rllib. I am currently trying to train a ppo agent in a multi-agent environment. Maybe a stupid question but I was wondering if there is a way to store and have access to all the training episodes. I tried to assign a path for “output” parameter to save the experiences but I got an error: TypeError: Object of type ndarray is not JSON serializable.
Could you create a github issue with a minimal, self-sufficient reproduction script, so we can debug this?
Please tag the issue with “bug”, “rllib”, and “P2”.