[](https://travis-ci.com/hill-a/stable-baselines) [](https://stable-baselines.readthedocs.io/en/master/?badge=master) # Torchy Baselines PyTorch version of [Stable Baselines](https://github.com/hill-a/stable-baselines), a set of improved implementations of reinforcement learning algorithms. TODO: - save/load - automatic choice for action distribution - predict - better rescale (min + action * range) - documentation - flexible mlp - logger - better monitor wrapper? Later: - get_parameters / set_parameters - CNN policies + normalization - tensorboard support - DQN - TRPO - A2C - ACER - HER -> use stable-baselines because does not depends on tf?