aleju/mario-ai

Playing Mario with Deep Reinforcement Learning

51
/ 100
Established

This project offers a way to train a digital agent to play the classic video game Super Mario World using only the game's visuals as input. The agent learns to press the right buttons by observing the screen, trying different actions, and remembering what leads to success or failure. This tool is for researchers or enthusiasts in artificial intelligence who want to experiment with creating smart agents for games.

695 stars. No commits in the last 6 months.

Use this if you are an AI researcher or enthusiast interested in teaching an artificial agent to master complex tasks, like playing video games, from raw visual data.

Not ideal if you are looking for a general-purpose game-playing AI, as this agent is specifically trained for the first level of Super Mario World and struggles with other parts of the game.

artificial-intelligence game-AI reinforcement-learning computer-vision game-agent-training
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 25 / 25

How are scores calculated?

Stars

695

Forks

142

Language

Lua

License

MIT

Last pushed

May 26, 2016

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/aleju/mario-ai"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.