###本工程文件适用于系统辨识小组作业
Flappy Bird Game trained on 3 models DQN algorithm: Deep Q-Network, Double DQN (DDQN) and Dueling DDQN with Prioritized Experience Replay implemented using Pytorch.
You will need Python 3.x.x with some packages which you can install direclty using requirements.txt.
pip install -r requirements.txt
first ,change the network form in paly_game.py(DQN or DDQN) THEN ,Use the following command to run the game where '--model' indicates the location of saved DQN model.
python play_game.py --model checkpoints/new/best_DQN.dat