Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.
assets initial commit Apr 2, 2017
demo initial commit Apr 2, 2017
.gitignore fix: protocol Apr 5, 2017 initial commit Apr 2, 2017 web version Apr 4, 2017
load.npy fix: use a random seed to increase reproducibility Apr 3, 2017
long.npy fix: increase space between bird and wall Apr 3, 2017 initial commit Apr 2, 2017 fix: no need for gtk Apr 4, 2017
requirements fix requirements Apr 4, 2017

Playing Flappy Bird using Evolution Strategies

After reading Evolution Strategies as a Scalable Alternative to Reinforcement Learning, I wanted to experiment something using Evolution Strategies, and Flappy Bird has always been one of my favorites when it comes to Game experiments. A simple yet challenging game.

The model learns to play very well after 3000 epochs, but not completely flawless and it rarely loses in difficult cases (high difference between two wall entrances). Training process is pretty fast as there is no backpropagation, and is not very costy in terms of memory as there is no need to record actions as in policy gradients.

Here is a demonstration of the model after 3000 epochs (~5 minutes on an Intel(R) Core(TM) i7-4770HQ CPU @ 2.20GHz):

after training

also see: Before training

There is also a a web version available for ease of access.

For each frame the bird stays alive, +0.1 score is given to him. For each wall he passes, +10 score is given.

Try it yourself

You need python3 and pip for installing and running the code.

First, install dependencies (you might want to create a virtualenv):

pip install -r requirements

The pretrained parameters are in a file named load.npy and will be loaded when you run or will train the model, saving the parameters to saves/<TIMESTAMP>/save-<ITERATION>. shows the game in a GTK window so you can see how the AI actually plays (like the GIF above). if you feel like playing the game yourself, space: jump, once lost, press enter to play again. 😁

pro tip: reach 100 score and you will become THUG FOR LIFE 🚬


It seems training past a maximum point reduces performance, learning rate decay might help with that. My interpretation is that after finding a local maximum for accumulated reward and being able to receive high rewards, the updates become pretty large and will pull the model too much to sides, thus the model will enter a state of oscillation.

To try it yourself, there is a long.npy file, rename it to load.npy (backup load.npy before doing so) and run, you will see the bird failing more often than not. was trained for only 100 more epochs than load.npy.