Book Image

Deep Reinforcement Learning Hands-On

By : Maxim Lapan
Book Image

Deep Reinforcement Learning Hands-On

By: Maxim Lapan

Overview of this book

Deep Reinforcement Learning Hands-On is a comprehensive guide to the very latest DL tools and their limitations. You will evaluate methods including Cross-entropy and policy gradients, before applying them to real-world environments. Take on both the Atari set of virtual games and family favorites such as Connect4. The book provides an introduction to the basics of RL, giving you the know-how to code intelligent learning agents to take on a formidable array of practical tasks. Discover how to implement Q-learning on 'grid world' environments, teach your agent to buy and trade stocks, and find out how natural language models are driving the boom in chatbots.
Table of Contents (23 chapters)
Deep Reinforcement Learning Hands-On
Other Books You May Enjoy

Basic DQN

By combining all the above, we can reimplement the same DQN agent in a much shorter, but still flexible, way, which will become handy later, when we'll start to modify and change various DQN parts to make the DQN better.

In the basic DQN implementation we have three modules:

  • Chapter07/lib/ The DQN neural network, which is the same as we've seen in the previous chapter

  • Chapter07/lib/ Common functions used in this chapter's examples, but too specialized to be moved to PTAN

  • Chapter07/ The creation of all used pieces and the training loop

Let's start with the contents of lib/ First of all, we have here hyperparameters for our Pong environment, that was introduced in the previous chapter. The hyperparameters are stored in the dict, with keys as the configuration name and values as a dict of parameters. This makes it easy to add another configuration set for more complicated Atari games.

    'pong': {
        'env_name': ...