I am looking to experiment with deep reinforcement learning. In particular training with evolutionary strategies.

Which python framework is most suited to this?

From what I can gather so far:

  • inference performance (on relatively small networks) is key
  • auto-gradient methods are irrelevant

Any advice / experience?

  • $\begingroup$ You should note that if you are using evolutionary strategies outlined in the article, then you are not using deep reinforcement learning. So you are looking for something more like NEAT I guess, or perhaps some approach recommended in the blog you linked? If so, you should probably remove the references to RL in your question, other than to note you want an alternative to RL. $\endgroup$ – Neil Slater Sep 26 '19 at 9:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.