Skip to content

A library of probabilistic model based RL algorithms in pytorch

License

Notifications You must be signed in to change notification settings

mcgillmrl/prob_mbrl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

prob_mbrl

Implementation of Deep-PILCO and variants for probabilistic Model Based RL. This is an (in progress) re-implementation of the algorithms in https://github.com/mcgillmrl/kusanagi. We also aim to add other probabilistic model-based RL methods to this library.

Recommended way to install:

Install the Miniconda 3 distribution: https://conda.io/miniconda.html

conda install pytorch cuda90 cudnn -c pytorch
conda install tqdm

To run the mc-pilco cartpole examples, you'll need to also install the kusanagi library (https://github.com/mcgillmrl/kusanagi). We plan to remove this dependency in the future.

For example on how to use this library, take a look at the notbooks folder. currently, we have an example for using the BNN models for regression, and an example of MC PILCO

About

A library of probabilistic model based RL algorithms in pytorch

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published