Skip to content

ruforavishnu/simple_rl_env

Repository files navigation

🧠 simple_rl_env

🕹️ A Minimal Reinforcement Learning Environment Framework for Python . Built by Vishnu Ajit plus ChatGPT. Reduces a huge amount of boilerplate coding when building rl projects.

Uses the pygame library to help visualize the graphical side of things. An advantage it has over the Sam Altman's openaigym is that - openaigym forces you to see a text grid on a terminal output. but this library has better visual interpretation as animations using a pygame window. Try building a snake rl with this. The visual feel is better. Though at the end of the day you should remember that comparing this graphical engine with Sam Altman's openai gym would have the same tradeoff when comparing a Hero Honda Splendor to a Bajaj Pulsar. One engine gives better mileage, one gives better performance. Likewise this library does a brilliant job of showing what the AI is thinking 'visually' simultaneously while it calculates the epsilon values and reward values.

PyPI Version Python Version License GitHub stars


🧩 About

simple_rl_env is a lightweight Python framework for building 2D Reinforcement Learning environments — perfect for beginners learning RL concepts or developers prototyping their own games and agents.

Built from scratch with Python + Pygame, it provides:

  • 📦 Simple environment management
  • 🎨 Built-in grid rendering system
  • 🐍 Example environments (Snake RL)
  • ⚙️ Easy hooks for RL agents (DQN, Q-Learning, etc.)

🚀 Installation

You can install directly from GitHub (until it’s on PyPI):

pip install git+https://github.com/ruforavishnu/simple_rl_env.git

About

wrote a library that reduces a huge amount of boilerplate code when building rl projects. uses pygame for visualizing what the AI is thinking simultaneously while the terminal shows the epsilon values and reward values.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages