Is your feature request related to a problem? Please describe.
So I used your script to build my own onnx backend and integrated it successfully into the triton server I use. However this introduces a dependency, namely cuda 11-8. Since ort 1.23.0 should support cuda 12.x - is there a way to mitigate this? like a flag during building?
Describe the solution you'd like
I dont know if the solution is already there, I would just like to have the libraries i need after building.