wassname/DeepRL

0 Fork 0

mirror of https://github.com/wassname/DeepRL.git synced 2026-10-07 11:13:51 +08:00

T

Shangtong Zhang 7c4919d6ac A3C

2017-05-13 13:48:18 -06:00

.gitignore

Refactor interfaces

2017-05-05 22:24:51 -06:00

async_agent.py

N-Step Q-Learning and One-Step Sarsa

2017-05-11 23:50:41 -06:00

bootstrap.py

A3C

2017-05-13 13:48:18 -06:00

dqn_agent.py

Update Readme

2017-05-11 21:30:04 -06:00

LICENSE

Initial commit

2017-04-20 13:59:53 -06:00

main.py

A3C

2017-05-13 13:48:18 -06:00

network.py

A3C

2017-05-13 13:48:18 -06:00

policy.py

A3C

2017-05-13 13:48:18 -06:00

README.md

A3C

2017-05-13 13:48:18 -06:00

replay.py

Add copyright declaration

2017-04-20 15:20:23 -06:00

task.py

Update Readme

2017-05-11 21:30:04 -06:00

README.md

DeepRL

Highly modularized implementation of popular deep RL algorithms powered by PyTorch

Deep Q-Learning
Asynchronous One-Step Q-Learning
Asynchronous One-Step Sarsa
Asynchronous N-Step Q-Learning
Asynchronous Advantage Actor Critic (A3C)

Benchmarked by classical control tasks (CartPole, LunarLander). Atari games will make it difficult to replicate in a regular laptop without a good GPU. However it's fairly easy to adapt the components to fit Atari games.

Try it out from main.py!