import gym
env = gym.make("LunarLander-v2", render_mode="human")
observation, info = env.reset(seed=42)
for _ in range(1000):
action = policy(observation) # User-defined policy function
observation, reward, terminated, truncated, info = env.step(action)
if terminated or truncated:
observation, info = env.reset()
env.close()
