Load binding based model in Pytorch using torchscript'

I am looking for information on how models trained in different language bindings of Pytorch, can be loaded into Pytorch using torchscript.

Is using ffi of the language to call the libtorch function to save model in torchscript format enough?