I have a torch script model that I load and run in C++. I recently found that the torch script file itself (which appears to be a zip file) contains several Python source files which prompted the following question
When running a torch script model using libtorch, is Python runtime involved in any way or do all operations run natively in C++? For instance, are the Python source files in the serialized model actually executed in a Python (or Python-like) interpreter?
I would appreciate any clarification here.