Python Deep Learning Projects

Testing the agent

Once the agent has been trained, we evaluate its performance over 100 new episodes. This can be done by calling the .test function and feeding the arguments environment and number of episodes on which to test:

# Evaluate the agent on 100 new episodesscores = sarsa.test(env, nb_episodes=100, visualize=False)print('Average score over 100 test games: {}'.format(np.mean(scores.history['episode_reward'])))

To view the CartPole game on your screen when testing, set the visualize argument to True inside the .test function.

The following is the output after testing 100 episodes:

Following the the output at the end of the code execution: ...