Update README.md

Split out onnx2torch_utils.py (converters + materialize_real) and style.css

Fix infinite scroll in HF iframe embed: drop vh, use fixed chat height

Run on ZeroGPU via PyTorch (onnx2torch conversion)

Hide Gradio default footer

Lock app to viewport height so input is always visible

Fix page growing infinitely: bound chat height with internal scroll

Run on free cpu-basic hardware (ONNX Runtime can't use ZeroGPU)

Enable CUDA on HF Spaces GPU: pin CUDA-12 ORT + preload nvidia libs

older cuda

gpu

fix

license + fix

space

initial commit