Deep Q-Learning

From

• Playing Atari with Deep Reinforcement Learning, V. Mnih et al., 2013
• Deep Reinforcement Learning with Double Q-learning H. van Hasselt et al., 2015

The game is a simple "GridWorld" style game, where the agent has to move the green square to the red square, in a minimum number of steps.

For using the code, first install the requirements with

$ pip install -r requirements.txt

Then run the training code by:

$ python main.py OPTIONS

Options:

• -d for using Double Q-Learning
• --gpu for using GPU acceleration

For testing a model:

$ python test.py -f FILENAME

Options:

• --gpu for using GPU acceleration