search

rll / rllab

rllab is a framework for developing and evaluating reinforcement learning algorithms, fully compatible with OpenAI Gym.
Other
2.91k stars 801 forks source link

Gaussian Policy - no inputs #245

Open aespielberg opened 6 years ago

aespielberg commented 6 years ago

Hi all,

I was looking to make a Gaussian Policy that took in zero inputs. rllab doesn't seem to have that capability, and I'm having trouble making a a policy similar to the GaussianMLPPolicy that can use 0 inputs. Is this possible? Any suggestions on how to get started? I tried making my own class, but unless I use Lasagne, it seems to act a little strange.