Update README

Add a link to other implementation.

Author: ikostrikov2

README.md

@@ -3,7 +3,7 @@
 This is a PyTorch implementation of Asynchronous Advantage Actor Critic (A3C) from ["Asynchronous Methods for Deep Reinforcement Learning"](https://arxiv.org/pdf/1602.01783v1.pdf).
 
 This implementation is inspired by [Universe Starter Agent](https://github.com/openai/universe-starter-agent).
-
+As in the starter agent, I don't share parameters of the optimizers between threads. If you want to have the same optimizer as in the original paper by DeepMind, you might want to check [this implementation.](https://github.com/rarilurelo/pytorch_a3c)
 
 ## Contibutions