[PYTHON] Deep Q-learning for playing Flappy Bird
Introduction
Here is my python source code for training an agent to play flappy bird. It could be seen as a very basic example of Reinforcement Learning's application.
How to use my code
With my code, you can:
- Train your model from scratch by running python train.py
- Test your trained model by running python test.py
Trained models
You could find my trained model at trained_models/flappy_bird
Requirements
- python 3.6
- pygame
- cv2
- pytorch
- numpy