QuantaScriptor/QGAN_Project ? reverse-engineered prompt

Build me a Python project for a Quantum Generative Adversarial Network that trains a quantum generator and discriminator together, using PennyLane and PyTorch. I want a simple but complete app that I can run locally to train the model, tweak the number of qubits and layers, and see basic training output as it learns.

Please include a clean training flow, a way to start it from the command line, and a small utility for working with molecules too, since I want to try it on simple chemistry examples like water. If there’s a deployment script or an easy setup path, make that work too.

Keep it practical and easy to understand, with sensible defaults, comments where needed, and enough structure that I can experiment with quantum data generation without having to build everything from scratch. If you need current docs for PennyLane or PyTorch, look them up online.