Skip to content
/ D4PG-ray Public

Distributed PyTorch implementation of D4PG with ray. Using a SOTA IQN Critic instead of C51. Implementation includes also the extensions Munchausen RL and D2RL which can be added to D4PG to improve its performance.

2 stars 1 fork Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

BY571/D4PG-ray

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
scripts
scripts
 
 
README.md
README.md
 
 
run.py
run.py
 
 

Repository files navigation

D4PG-ray

TODO:

  • Clear / clean code
  • Do testruns for LunarLander // Pendulum
  • Make comparisons of all features
  • Make Readme
  • Add PER

About

Distributed PyTorch implementation of D4PG with ray. Using a SOTA IQN Critic instead of C51. Implementation includes also the extensions Munchausen RL and D2RL which can be added to D4PG to improve its performance.

Topics

reinforcement-learning distributed-computing reinforcement-learning-algorithms ray ddpg iqn state-of-the-art continuous-action-space d4pg

Resources

Readme
Activity

Stars

2 stars

Watchers

2 watching

Forks

1 fork
Report repository

Releases

No releases published

Packages

No packages published

Languages