Torchvision.__version__ causing "attribute error - module has no attribute __version__"

vision
#1

Hi, I’m wondering what is the alternative for __version__ in torchvision? It used to work normally but now it doesn’t and I’m not able to find the alternative in the documentation.

Error received when calling print(torchvision.__version__) is

AttributeError: module ‘torchvision’ has no attribute ‘version’

Thanks!

#2

It might be a tough question, but do you remember which version you’ve installed (or when from which source)?
The command still works for 0.5.0a0+28003e9.
Are you able to use torchvision besides this missing piece?

#3

I switched my anaconda environment and updated my packages and now it is working with torchvision 0.4.2 and torch 1.3.1.

I was running torchvision version is 0.4.1, it seems to be the latest I can get through pip for some reason.

pip install --upgrade torchvision

returns

Requirement already up-to-date: torchvision in c:\users\...\lib\site-packages (0.4.1)
Requirement already satisfied, skipping upgrade: pillow>=4.1.1 in c:\users\...\lib\site-packages (from torchvision) (6.2.1)
Requirement already satisfied, skipping upgrade: torch in c:\users\...\lib\site-packages (from torchvision) (1.0.1)
Requirement already satisfied, skipping upgrade: numpy in c:\users\...\lib\site-packages (from torchvision) (1.17.4)
Requirement already satisfied, skipping upgrade: six in c:\users\...\lib\site-packages (from torchvision) (1.13.0)

Everything is working now, however I’m still unable to find the command in the documentation, which is why I thought the syntax might have changed.