Performance of pretrained models in torchvision

Hi all!

Is it possible to know what’s the actual performance of the pretrained models on the ImageNet dataset without having to test them by myself?
In particular i’m interested in the performance of the DenseNet models.


You can find the performance of the models here.