Reinforcement Learning with Imagined Goals
April 6, 2019 ยท View on GitHub
Implementation of reinforcement learning with imagined goals (RIG) To find out more, see any of the following links:
- arXiv: https://arxiv.org/abs/1807.04742
- Website: https://sites.google.com/site/visualrlwithimaginedgoals/
- Blog Post: https://bair.berkeley.edu/blog/2018/09/06/rig/
To see the original implementation, checkout version v0.1.2 of this repo.
In versions 0.2+, RIG is a special case of Skew-Fit with the
power set to 0.