https://github.com/vivek3141/super-mario-rl
Deep Reinforcement learning and Python learn how to play the original Super Mario Bros
https://github.com/vivek3141/super-mario-rl
deep-q-learning machine-learning mario reinforcement-learning
Last synced: about 1 year ago
JSON representation
Deep Reinforcement learning and Python learn how to play the original Super Mario Bros
- Host: GitHub
- URL: https://github.com/vivek3141/super-mario-rl
- Owner: vivek3141
- Created: 2018-07-31T07:02:14.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2018-09-02T05:27:21.000Z (almost 8 years ago)
- Last Synced: 2023-03-02T04:25:50.879Z (over 3 years ago)
- Topics: deep-q-learning, machine-learning, mario, reinforcement-learning
- Language: Python
- Size: 27.3 KB
- Stars: 27
- Watchers: 3
- Forks: 8
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Super Mario AI
The original Super Mario Bros has always been a classic. One of the best part about
older games is that they run on lower resolution, which makes it easier
to feed into a neural network.
## What this program does
This project uses a topic called reinforcement learning. The way it works is by
having a reward based system and the program learns on its own to get the best reward.
You can read a few articles on the internet to get a gist of it.
This program can play the original Mario Bros for the NES.
## Requirements
* Gym - `pip install gym`
* Gym Super Mario Bros - `pip install gym-super-mario-bros`
* NES-Py - `pip install nes-py`
As of right now, nes-py is only supported on linux so please run it on linux.
I have tried for hours to try to get it on Windows, to no avail. If you know how to, please let me know
so I can update this.
## Running
`python3 main.py `
Eg. If you want to run 1-1
`python3 main.py '1-1'`
## Images
Implementation of NEAT can be found [here](https://github.com/vivek3141/super-mario-neat)