awjuliani / DeepRL-Agents

A set of Deep Reinforcement Learning Agents implemented in Tensorflow.
MIT License
2.23k stars 825 forks source link

A3C-Doom, is threading can make real parallelism? #68

Open GoingMyWay opened 5 years ago

GoingMyWay commented 5 years ago

Hi, your repo is very helpful for me to learn RL in action. In A3C-Doom.ipynb, I found the code spawns multi-threads to train the model, in Python threading cannot make real parallelism due to GIL, and here is threading can make the A3C model running in parallel?