Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Error installing with pip - ERROR: No matching distribution found for torchvision==0.8.2 #1

Open
roboserg opened this issue Aug 25, 2021 · 3 comments

Comments

@roboserg
Copy link

roboserg commented Aug 25, 2021

After doing pip install --upgrade git+https://github.com/rbrtwlz/fastai_object_detection
ERROR: No matching distribution found for torchvision==0.8.2

I already have torchvision 0.10 installed, does this absolutely require 0.8.2? It wants older fastai too (current ver. is 2.5.2).

Is there a way to install this without downgrading?

WindowsTerminal_PfbSSU2x6h

@roboserg
Copy link
Author

Ok, I created new empty environment and manually installed everything from setup.py. Everything install except
mean_average_precision @ git+https://github.com/bes-dev/mean_average_precision

It doesnt work (package not found error). I then tried pip install your package (the only thing I was missing was mean_average_precision ) and of course your repo fails installing it as well
WindowsTerminal_OGGXMHWo52

@roboserg
Copy link
Author

roboserg commented Aug 26, 2021

Installing Microsoft Visual Studio Build Tools 2019 didnt help - https://i.imgur.com/nYAPf41.png

@rbrtwlz
Copy link
Owner

rbrtwlz commented Aug 28, 2021

Hi roboserg,

currently there is a error with fastai's type dispatch when using torch > 1.7.1, so torchvision==8.2 is required. I recommend using a virtual environment.

And mean_average_precision is just required for testing now, by default it uses my port of it (from numpy to pytorch).

But your error is regarding to the installation of pycocotools. I never tested it on windows, but according to this issue you could try to install pycocotools-windows.

Btw, if you don't want to create a dataset with CocoData class, then you don't need pycocotools.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants