gdrl

Why we included this project

This repo pairs with the Manning book 'Grokking Deep Reinforcement Learning' and is aimed at readers who want to stop just reading about RL and actually run the algorithms. The chapters come as runnable Jupyter notebooks in Python on PyTorch, working up from Q-learning and DQN to policy-gradient and actor-critic methods. Everything runs inside a Docker image that pre-installs the full stack, so you skip the usual dependency setup and spend your time experimenting with the agents instead of fighting installs. For students, data scientists, and engineers crossing into deep RL, it works as a self-paced lab, though the code follows the book's structure rather than standing alone as a production-ready library.