search

jgamper / intrinsic-dimensionality

Use this package to compute intrinsic dimensionality of your task given a fixed neural network in PYTORCH! :fire:
MIT License
33 stars 3 forks source link

codes does not run #6

Open dorost1234 opened 3 years ago

dorost1234 commented 3 years ago

Hi

when I run the codes, this goes to infinite loop and freezes, additionally the codes have a bug: FastFoodWrap does not exist and this is with small f for food,

could you please suggest me any resource to understand the implementations? I think there is also a fast CUDA version of this code, do you have any idea about it?

Thanks a lot for sharing these codes also, this is wonderful and very clean

thanks

dorost1234 commented 3 years ago

could you please also explain these lines of init:

           while "." in localname:
                    prefix, localname = localname.split(".", 1)
                    base = base.__getattr__(prefix)

                self.name_base_localname.append((name, base, localname))

        for name, base, localname in self.name_base_localname:
            delattr(base, localname)

what is the goal here? thanks