This repository contains Deep Q-Networks and Double Deep Q-Networks implementation in tensorflow for Open AI Gym environments such as Cartpole problem and the mountain car problems. Its a plug and play code with no preprocessing required and just the main code has to be run. Different hyperparameters can be changed as per the user.
The Deep Q-Networks algorithm implemented in the code is a direct implementation of the paper " Playing Atari with Deep Reinforcement Learning" by Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Alex Graves, Ioannis Antonoglou, Daan Wierstra, Martin Riedmiller. Archive Link: https://arxiv.org/pdf/1312.5602.pdf
The code is currently by default set for running in tensorflow gpu and thus requires tensorflow gpu installation but can be easily modified for making it run in CPU
Libraries Required:
- Tensorflow GPU
- Open AI Gym (full package installation)
Note: If someone can write some code to save the trained model it would be great and I would love to add it to the main branch.