Some implements of reinforcement leraning include DQN, TD3. PPO and Visual PPO.
The enviroments of reinforcement learning is implemented by gymnasuim, you can check all enviroments in their documents.
- All neural networks of each algorithms are implemented by Pytorch.
- In theory, you can try any discrete environments with DQN, and any both discrete and continuous environments with TD3 and PPO, while I just written continous types. :P
- Visual PPO is PPO which use screenshot(image) but not normal attributes as input, it is not complete here, there is more about visual reinforcement learning in this project.