This repository is based on discor-pytorch and [1]. I am going to add on-policy reweight[2] and probably some other components.
We would like to thank Samarth Sinha for kindly providing the source code of [2].
[1] Kumar, Aviral, Abhishek Gupta, and Sergey Levine. "Discor: Corrective feedback in reinforcement learning via distribution correction." arXiv preprint arXiv:2003.07305 (2020).
[2] Sinha, Samarth, et al. "Experience Replay with Likelihood-free Importance Weights." arXiv preprint arXiv:2006.13169 (2020).