The original code from the DeepMind article + my tweaks
Reproducing the results of "Playing Atari with Deep Reinforcement Learning" by DeepMind
Open on GitHub