RicardoDominguez / PyCREPS

Contextual Relative Entropy Policy Search for Reinforcement Learning in Python
15 stars 1 forks source link

Theano do not use reshape #8

Closed RicardoDominguez closed 6 years ago

RicardoDominguez commented 6 years ago

Make use of dcol or drow to avoid reshaping.

RicardoDominguez commented 6 years ago

For consistency with the other implementations its best not to change it.