Welcome to rllab
rllab is a framework for developing and evaluating reinforcement learning algorithms.
rllab is a work in progress, input is welcome. The available documentation is limited for now.
The rllab user guide explains how to install rllab, how to run experiments, and how to implement new MDPs and new algorithms.
.. toctree:: :maxdepth: 2 user/installation user/experiments user/gym_integration user/implement_env user/implement_algo_basic user/implement_algo_advanced user/cluster
If you use rllab for academic research, you are highly encouraged to cite the following paper:
- Yan Duan, Xi Chen, Rein Houthooft, John Schulman, Pieter Abbeel. "Benchmarking Deep Reinforcement Learning for Continuous Control. Proceedings of the 33rd International Conference on Machine Learning (ICML), 2016.