donglinkang2021/reinforce_tuto

Minimal, clean and simple reinforcement learning tutorial with `gymnasium`, `numpy` and `torch`.

Reinforcement Learning Tutorial

Minimal, clean and simple reinforcement learning tutorial with gymnasium, numpy and torch.

Refer

• Gymnasium (Deep) Reinforcement Learning Tutorials
• Gymnasium
• the original code document of gymnasium

FrozenLake

intro see gymnasium.envs.toy_text.frozen_lake

You can already run the test code on pretrained Q table.

python .\code\FrozenLake\test_Qlearning.py
python .\code\FrozenLake\test_Sarsa.py 

Qlearning	Sarsa

Or you can train the model by yourself.

python .\code\FrozenLake\train_Qlearning.py
python .\code\FrozenLake\train_Sarsa.py

MountainCar

intro see gymnasium.envs.classic_control.mountain_car

python .\code\MountainCar\test_Qlearning.py
python .\code\MountainCar\test_Sarsa.py 

Qlearning	Sarsa