search

rll / rllab

rllab is a framework for developing and evaluating reinforcement learning algorithms, fully compatible with OpenAI Gym.
Other
2.91k stars 800 forks source link

Changin the policies to Tensorflow version #227

Open elmiraloo opened 6 years ago

elmiraloo commented 6 years ago

I was wondering if there is any specific reason behind using Lasagne and Theano in defining different policies. I can change the code to Tensorflow (seems to be straightforward) and in case of resource allocation working with one main library seems to be a better idea. Am I missing something? If there is no specific reason I can re-factor the code.