CUB200 dataset does not download automatically
francesco-p opened this issue · 4 comments
To reproduce
Just run your colab and replace MNIST with CUB200 and you get the error during the extraction of the tar file.
edit:
even by cloning the repo and pip installing it throws the error.
Unfortunately, it seems that the quote of download for this dataset as exceed.
Here is the content of the downloaded file (which should have been a .tgz archive):
Too many users have viewed or downloaded this file recently. Please try accessing the file again later. If the file you are trying to access is particularly large or is shared with many people, it may take up to 24 hours to be able to view or download the file. If you still can't access a file after 24 hours, contact your domain administrator.
I cannot host by myself the file archive, and the official download page (http://www.vision.caltech.edu/visipedia/CUB-200.html) only gave this google drive link. So you may want to contact the author to tell them about that, because continuum cannot do anything about that I'm afraid.
OK, if you donwload it manually it works. Maybe you can specify this thing in the documentation. Thanks anyway.
I'll add a warning message in the code, because it SHOULD work automatically, but I guess a few times it won't.
Well I was able right now to download cub200 with continuum...
Anyway, I've added an error message in case this bug re-appears: 9c29d8a