Which images are used for the representative models trained in pytorch?

Where in pytorch are the images that VGG, MobileNet, etc. are learning?

The pretrained classification models should have been trained on ILSVRC2012 if I’m not mistaken.