coreylynch / async-rl

Tensorflow + Keras + OpenAI Gym implementation of 1-step Q Learning from "Asynchronous Methods for Deep Reinforcement Learning"
MIT License
1.01k stars 174 forks source link