For context, I built this on my local rig (RTX 3080 10GB, Ryzen 5, 16GB RAM).
I also tested the backend on Lightning AI (T4 GPU) to make sure it wasn't just working on my machine. I'll update the readme now.