qrl-qai is a python framework built on top of Gymnasium, PennyLane, and PyTorch to serve as a central platform for everything quantum reinforcement learning.
The current release (0.2.0) offers 5 native quantum RL style environments - BlochSphereV0, CompilerV0, ErrorChannelV0, ExpressibilityV0, ProbabilityV0.
Each significant release has an asssociated Google Colab and Lightning AI Studio template to help users experiment faster (see the table below). Lightning AI Studio has additional Streamlit based webapp for no-code experimentation and is recommended!
You can have a look at the documentation for each version too. However, it is recommended to start with Google Colab or Lightning AI Studio!
| Version | Google Colab | Lightning AI Studio | Documentation |
|---|---|---|---|
| 0.1.0 | - | - | Read The Docs |
| 0.2.0 | Read The Docs |
Install the package directly from PyPI:
pip install qrl-qaiTo save the episodes as mp4 videos, it is essential to have ffmpeg installed on your system. In the absence of ffmpeg, you can run the environments but save the episodes as gif only and not mp4.
To check if you have ffmpeg installed (Windows/Linux/Mac), you can try:
ffmpeg -versionTo install ffmpeg, you can follow these steps:
You can also install FFmpeg within a conda environment:
conda install -c conda-forge ffmpegOption 1: Using Chocolatey (Recommended)
If you have Chocolatey installed:
choco install ffmpeg✅ Option 2: Manual installation
-
Go to the official FFmpeg site → Click Windows → gyan.dev builds or BtbN builds.
-
Download the latest release full build ZIP.
-
Extract it (e.g., to C:\ffmpeg).
-
Add the bin folder to your PATH:
4.1 Press Win + R → sysdm.cpl → Advanced → Environment Variables
4.2 Edit Path → Add new entry: C:\ffmpeg\bin
Debian / Ubuntu:
sudo apt update
sudo apt install ffmpeg -yFedora:
sudo dnf install ffmpeg -yArch Linux:
sudo pacman -S ffmpegUsing Homebrew:
brew install ffmpegContributions are welcome! 🎉
See our CONTRIBUTING.md for guidelines.
This project is licensed under the Apache 2.0 License
Initiated by Jay Shah
Email: [email protected]

