So im looking for a way to load some model without acces to the class code. Is there a way? The only way to make such thing possible is to save the model with torch.jit.save/load but what if im not the person which training/designt the model?
With tensorflow im just used to use the load_model function. Is there something similary within pytorch?