Here I implement Deep Q-Network (for discrete action space) and Deep Deterministic Policy Gradient (for continuous action space) using pytorch package, both on a sample gym environment(MountainCar).
-
Notifications
You must be signed in to change notification settings - Fork 0
mhtb32/Deep-Reinforcement-Learning
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Solving Mountain Car using Deep Reinforcement Learning Algorithms
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published