search

takuseno / d3rlpy

An offline deep reinforcement learning library
https://takuseno.github.io/d3rlpy
MIT License
1.32k stars 243 forks source link

[REQUEST] Offline Reproduction of AWR #154

Closed im-Kitsch closed 2 years ago

im-Kitsch commented 2 years ago

Hello,

thanks a lot for your work. I wanna to reproduce the result of AWR, however, it seems there is no implementation in ./reproductions/offlines/ folder. Would you like to add reproduce of AWR algorithm?

Thanks a lot

takuseno commented 2 years ago

Hello, thanks for the issue. Unfortunately, AWR reproduction is not planned and AWR will be dropped in the future because it has a slightly different training scheme and is less powerful than the other methods. Most importantly, AWR is not really used in recent papers. Sorry about the inconvenience.

takuseno commented 2 years ago

Let me close this issue for now.