From c4ae7cd924cc76b3c3f456008ce4df80d3d8c746 Mon Sep 17 00:00:00 2001 From: Yash Suresh Chandra Date: Sun, 23 Feb 2025 04:47:59 +0530 Subject: [PATCH] pettingzoo example observation_space fix (#1243) Had to make few changes to run tictactoe example (was getting errors otherwise) --------- Co-authored-by: YASH CHANDRA --- test/pettingzoo/tic_tac_toe.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/test/pettingzoo/tic_tac_toe.py b/test/pettingzoo/tic_tac_toe.py index cac615896..9e74c003e 100644 --- a/test/pettingzoo/tic_tac_toe.py +++ b/test/pettingzoo/tic_tac_toe.py @@ -104,7 +104,7 @@ def get_agents( ) -> tuple[BasePolicy, torch.optim.Optimizer | None, list]: env = get_env() observation_space = ( - env.observation_space["observation"] + env.observation_space.spaces["observation"] if isinstance(env.observation_space, gymnasium.spaces.Dict) else env.observation_space ) @@ -235,5 +235,5 @@ def watch( policy, optim, agents = get_agents(args, agent_learn=agent_learn, agent_opponent=agent_opponent) policy.policies[agents[args.agent_id - 1]].set_eps(args.eps_test) collector = Collector[CollectStats](policy, env, exploration_noise=True) - result = collector.collect(n_episode=1, render=args.render) + result = collector.collect(n_episode=1, render=args.render, reset_before_collect=True) result.pprint_asdict()