Giter Site home page Giter Site logo

atari-pong-rl's Introduction

DQN agent for Wimblepong

This project is a DQN agent able to play in the pong-v0 OpenAI Gym environment. The environment is used in the Reinforcement Learning course at Aalto University, Finland.

๐Ÿ”ง How to use

  • Clone the repository: https://github.com/MattiaMolon/Atari-Pong-RL.
  • Install the dependencies listed in requirements.txt.
  • to train: check the train_basic.py to train the agent against simpleAI or train_hibrid.py to train the agent one game against simpleAI and one game against himself (This helps the agent to not overfit against simpleAI alone). Run both files with the flag --train True.
  • to test: check the train_basic.py to test the agent against simoleAI or train_hibrid.py to test the agent against another agent. Run both the files without the flag --train True.

๐Ÿค– Agents

  • The SimpleAI agent is an agent that uses the absolute ball and player positions to follow the ball and reflect it in random directions.
  • The agent implemented in this repository consists in a DQN agent with experience replay and can be found in agent.py. The weights in ./weights/hibrid_tuned_best.ai can reach a winrate of 80% against simpleAI.

๐Ÿ“œ Report

The report pdf_files/RL_final_project.pdf includes all the details about the implementation of the agent and the training procedures followed to reach the aformentioned scores.

๐ŸŒ Environment

Additional details on the environment used and how to interact with it are available in the official repository of the course.

๐ŸŒ  Future works:

  • implementation DQN
  • implementation experience replay, buffers, and preprocessing
  • implementation of PPO
  • implementation of actor-critic

atari-pong-rl's People

Contributors

mattiamolon avatar

Stargazers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.