Hands-On Neuroevolution with Python

Experiment configuration file

Here is the file that provides configuration parameters for the experiment runner. For our experiment, it has the following content:

{    "game": "frostbite",    "model": "LargeModel",    "num_validation_episodes": 30,    "num_test_episodes": 200,    "population_size": 1000,    "episode_cutoff_mode": 5000,    "timesteps": 1.5e9,    "validation_threshold": 10,    "mutation_power": 0.002,    "selection_threshold": 20}

The configuration parameters are as follows:

• The game parameter is the name of the game, as registered in the ALE. The full list of supported games is available at https://github.com/PacktPublishing/Hands-on-Neuroevolution-with-Python/blob/master/Chapter10/gym_tensorflow/atari/tf_atari.py.
• The model parameter designates the name ...