How Resize transform influences training?

Hi. I want to use Resize transform on 256256 images to get 6464 images, since my GPU has only 4GB memory. Will the resize affect significantly the training process ?