Skip to content

joaopedrofn/learnToWalk

Repository files navigation

Learn to Walk

A project with the only goal to apply new knowledge about Q-Learning through studies to develop the undergraduate thesis.

The project consists in a little game where the character must walk to the grid world to get the flag avoiding the walls and the little demon.

There are three buttons, one to train the Q-Learning algorithm, one to make the Q-Learning take the next action and the last one to reset the world.

Installation

Clone the project git clone https://github.com/joaopedrofn/learnToWalk.git.

Install the dependencies python3 -m pip install gym numpy

Install the new gym enviroment

    cd ./learnToWalk/gym-sample
    python3 -m pip install -e .

Usage

Just execute the start.sh script ./start.sh

And access the http://localhost:3000 at your browser.

About

Part of studies to implement Q-Learning into my undergradute thesis. Objective is to character learn to walk until the objective avoiding obstacles and lose point.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors