Simple implementation of Deep Reinforcement Learning agents DQN (tested on the CartPole environment from OpenAI Gym) and DDPG (tested on environments Pendulum and LunarLander).

Includes some useful functions to track experiments: save parameters used, log files and plots in a results folder.

Uses TensorFlow.
To install a working conda environment: conda env create -f environment.yml (more info here: conda.io)