sash-a / CleanRL.jl Public

Notifications You must be signed in to change notification settings
Fork 0
Star 20

Simple single file implementations of Reinforcement Learning algorithms in Julia

20 stars 0 forks Branches Tags Activity

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
logs		logs
src		src
.gitignore		.gitignore
LICENSE		LICENSE
Manifest.toml		Manifest.toml
Project.toml		Project.toml
README.md		README.md

Repository files navigation

CleanRL.jl

Simple single file implementations of Reinforcement Learning algorithms in Julia.
Inspired by CleanRL

TODO (Algorithms):

~~Simple DQN~~
~~A2C~~
Rainbow
~~PPO~~
~~DDPG~~
SAC

TODO (Utils):

~~General replay buffer~~
~~CLI for hyperparameters~~
~~Support loggers~~
GPU training
~~Multi-thread PPO~~
Vectorized envs
~~Plotting~~
~~Multi-loggers (file/console/Tensorboard)~~
~~Make nn inputs F32 - F32 env wrapper?~~ - done for PPO
Make individual file runners e.g experiments/run_ppo.(jl/sh)
Better logging - log interval and always log at correct step

TODO (Investigate):

Profile PPO

About

Simple single file implementations of Reinforcement Learning algorithms in Julia

Report repository

Releases

No releases published

Packages

No packages published

Languages

Julia 100.0%