Benchmarking Recurrent Architectures for Robust Continuous Control
The code in this repository is based on an (outdated) fork of rllab. Please refer to that repository for full documentation.
The main contribution of this repository is additional (Theano) RNN policy implementations on top of the ones provided by rllab: