def __init__(self, name, max_step=None):
if name == "simple_spread":
self.env = simple_spread_v2.env()
elif name == "waterworld":
self.env = waterworld_v3.env()
elif name == "multiwalker":
self.env = multiwalker_v6.env()
else:
assert AssertionError, "wrong env name."
self.max_step = max_step
self.curr_step = 0
self.name = name
self.agents = self.env.possible_agents
self.env.reset()
from ray import tune
from ray.tune.registry import register_env
from ray.rllib.env.wrappers.pettingzoo_env import PettingZooEnv
from pettingzoo.sisl import waterworld_v3
# Based on code from github.com/parametersharingmadrl/parametersharingmadrl
if __name__ == "__main__":
# RDQN - Rainbow DQN
# ADQN - Apex DQN
register_env("waterworld", lambda _: PettingZooEnv(waterworld_v3.env()))
tune.run(
"APEX_DDPG",
stop={"episodes_total": 60000},
checkpoint_freq=10,
config={
# Enviroment specific.
"env": "waterworld",
# General
"num_gpus": 1,
"num_workers": 2,
"num_envs_per_worker": 8,
"learning_starts": 1000,
"buffer_size": int(1e5),
"compress_observations": True,
"rollout_fragment_length": 20,
"train_batch_size": 512,
"gamma": 0.99,
"n_step": 3,
def env_creator(args):
return PettingZooEnv(waterworld_v3.env())