A grassroots agent trainer community! Think pokemon, but for small AIs you can run and train on your PC and phone. [ join us ]
RL stands for Reinforcement Learning. It is the most general setting to formulate and study learning in machines and animals alike. In its modern, most advanced form it amounts to training algorithms based on deep neural networks to behave as to maximize reward over time in a given environment.
Our project aims to provide software and infrastructure to compile an open library of trained agents, with the help of volunteers like you and me.