def env_creator(config):
env = pistonball_v5.env()
env = dtype_v0(env, dtype=np.float32)
env = color_reduction_v0(env, mode="R")
env = normalize_obs_v0(env)
return env
import os
from os.path import exists
import numpy as np
import supersuit as ss
from array2gif import write_gif
from pettingzoo.butterfly import pistonball_v5
from stable_baselines3 import PPO
n_agents = 20
env = pistonball_v5.env()
env = ss.color_reduction_v0(env, mode="B")
env = ss.resize_v0(env, x_size=84, y_size=84)
env = ss.frame_stack_v1(env, 3)
policies = os.listdir("./optimization_policies/")
for policy in policies:
filepath = "./optimization_policies/" + policy + "/best_model"
if not exists(filepath + '.zip'):
continue
print("Loading new policy ", filepath)
model = PPO.load(filepath)
obs_list = []
i = 0
env.reset()
total_reward = 0
try:
while True: