Open linshuxi opened 12 months ago
I get some error: Action space MultiDiscrete([11 5 1 2]) is not supported for DQN.
ray: 2.6.3
from ray.rllib.algorithms.dqn import DQNConfig algo = DQNConfig().framework("torch").environment(env = myenv).multi_agent( policies={ "A":(None, a.observation_space, a.action_space, {"gamma": 0.80}), "B":(None, a.observation_space, a.action_space, {"gamma": 0.80}) }, policy_mapping_fn = lambda agent_id: f"{agent_id}",
).build()
None
I am experiencing the same problem. May I ask how you solved it?
What happened + What you expected to happen
I get some error: Action space MultiDiscrete([11 5 1 2]) is not supported for DQN.
Versions / Dependencies
ray: 2.6.3
Reproduction script
from ray.rllib.algorithms.dqn import DQNConfig algo = DQNConfig().framework("torch").environment(env = myenv).multi_agent( policies={ "A":(None, a.observation_space, a.action_space, {"gamma": 0.80}), "B":(None, a.observation_space, a.action_space, {"gamma": 0.80}) }, policy_mapping_fn = lambda agent_id: f"{agent_id}",
).build()
Issue Severity
None