A single 200 lines of python code to demostrate DQN with Keras
Please read the following blog for details
https://yanpanlau.github.io/2016/07/10/FlappyBird-Keras.html
- Python 2.7
- Keras 1.0
- pygame
- scikit-image
CPU only
git clone https://github.com/yanpanlau/Keras-FlappyBird.git
cd Keras-FlappyBird
python qlearn.py -m "Run"
GPU version (Theano)
git clone https://github.com/yanpanlau/Keras-FlappyBird.git
cd Keras-FlappyBird
THEANO_FLAGS=device=gpu,floatX=float32,lib.cnmem=0.2 python qlearn.py -m "Run"
If you want to train the network from beginning, delete the model.h5 and run qlearn.py -m "Train"