Flappy Bird as a Farama Gymnasium environment.

pip install flappy-bird-env

1. Play it by running

python -m flappy_bird_env

Press space to flap the wings.

2. Import it to train your RL model

import flappy_bird_env  # noqa
env = gymnasium.make("FlappyBird-v0")

The package relies on import side-effects to register the environment name so, even though the package is never explicitly used, its import is necessary to access the environment.

Flappy Bird has the action space Discrete(2).

The observation will be the RGB image that is displayed to a human player with observation space Box(low=0, high=255, shape=(800, 576, 3), dtype=np.uint8).

You get +1 every time you pass a pipe, otherwise +0.001 for each frame where you don't collide against the top and bottom bounds, or against a pipe.

• v0: initial version release