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

No local network synchronization #22

Open rbrigden opened 7 years ago

rbrigden commented 7 years ago

I'm interested as to why you decided not to create a local copy of the variables in the worker threads and sync them with the global network at the end of the rollout. Does that create issues with the global network (being used for inference in the rollout) being updated in the middle of rollout? Is there a reason why you changed your algorithm from the one described in the Async methods for RL paper?