OroloRL is a collection of various deep reinforcement learning algorithms. Each alogorthim is implented in single file using pytorch and tested on OpenAI's gym library ( pybullet, atari,... )
- Currently the algorithms are design to run on a cpu ( as I don't have a GPU yet) but a quick crtl+shift+f 'cpu' to device should fix fix that. Where:
device = T.device('cuda:0' if T.cuda.is_available() else 'cpu')
- PyTorch
- OpenAI GYM
- [1] Youtube-Code-Repository https://github.com/philtabor/Youtube-Code-Repository
- [2] minimalRL-pytorch https://github.com/seungeunrho/minimalRL
- [3] CleanRL (Clean Implementation of RL Algorithms) https://github.com/vwxyzjn/cleanrl