This repository contains a PyTorch implementation of Monte Carlo policy gradient reinforcement (REINFORCE) for discrete action spaces.
🚧 🛠️👷♀️ 🛑 Under construction...
Install the required dependencies using the following command:
pip install -r requirements.txtYou can run the algorithm on any supported Gymnasium environment. For example:
python main.py --env 'LunarLander-v2'|
CartPole-v1 
|
MountainCar-v0 
|
Acrobot-v1 
|
|
|
|
|
LunarLander-v2 
|
Asteroids-v5 
|
Breakout-v5 
|
|
|
|
|
BeamRider-v5 
|
Centipede-v5 
|
DonkeyKong-v5 
|
|
|
|
|
Frogger-v5 
|
KungFuMaster-v5 
|
MarioBros-v5 
|
|
|
|
|
SpaceInvaders-v5 
|
Tetris-v5 
|
Gopher-v5 
|
|
|
|
|
MsPacman-v5 
|
Pong-v5 
|
Seaquest-v5 
|
|
|
|
Special thanks to Phil Tabor, an excellent teacher! I highly recommend his Youtube channel.