Skip to content

masskro0/dqn_breakout

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Using Deep Q-Learning to win Breakout

Winning GIF

Simple and efficient way to win Breakout using Deep Q-Learning.

I didn't include the weights of the winning rounds because I generated too many weights and lost track. But within 16 hours of training, you should be able to get them by yourself.

Evaluation Results

Training Results Evaluation Results Loss Values

Setup

  • Python 3.8.10
  • NVidia GPU (GTX 1060, RTX 3060)
  • Ubuntu 20.04 & Windows

Installation

pip install -r requirements.txt

Launching

Run jupyter-notebook in your terminal. Then select main.ipynb and start running the blocks from top to bottom. You should visually see the output.