Skip to content

[NeurIPS'19] Policy Continuation with Hindsight Inverse Dynamics

Notifications You must be signed in to change notification settings

decisionforce/PCHID

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PCHID_code

Code for Policy Continuation with Hindsight Inverse Dynamics (PCHID)

There are three environments

  • Bitflip

The bitflip environment was introduced by Hindsight experience replay (HER)

Our implementation of bitflip and HER refers RL-Adventure-2 and Baselines

  • GridWorld

Our experiment on the GridWorld domain is based on previous work of Value Iteration Networks (VIN) and the PyTorch implementation of VIN

  • FetchReach

The FetchReach environment is one of four OpenAI Fetch envs. As for the other three environments (Push, Slide, PickAndPlace), synchoronous improvement is needed to further improve PCHID's learning efficiency.

About

[NeurIPS'19] Policy Continuation with Hindsight Inverse Dynamics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published