• I have marked all applicable categories:
  • exception-raising bug
  • RL algorithm bug
  • documentation request (i.e. "X is missing from the documentation.")
  • new feature request
• I have visited the source website
• I have searched through the issue tracker for duplicates
• I have mentioned version numbers, operating system and environment, where applicable:

  import tianshou, torch, numpy, sys
  print(tianshou.__version__, torch.__version__, numpy.__version__, sys.version, sys.platform)

@jkterry1 @benblack769
Following the discussions in issues #399 and #121 (comment), I have been working on a wrapper to use Pettingzoo multiagent environments with the tianshou API. However, I will not be able to continue working from now on and I will appreciate if someone else could take it from here. The additions made are:

• A pettingzoo environment wrapper for Agent Environment Cycle (AEC) Games. https://github.com/rodrigodelazcano/tianshou/blob/d40e6c6234dae1e0846dc34b5c11cc554a819bd4/tianshou/env/pettingzoo_env.py
• Changed the policies list in the multiagent policy manager for a dictionary relating agent name (key) and the corresponding policy (value)
• An example of the training of a vectorized pettingzoo environment can be run here. https://github.com/rodrigodelazcano/tianshou/blob/d40e6c6234dae1e0846dc34b5c11cc554a819bd4/test/pettingzoo/pistonball_train.py

I will really appreciate if someone else could continue working on this issue and create a stable multiagent feature for this amazing library that tianshou is. Thanks!

Some of the tasks moving forward can be the ones stated in this comment #399 (comment)