When I use pytorch datasets like:
dataset_train = datasets.CIFAR10(root='../data/CIFAR10', train=True, download=True, transform=train_transform)
it downloads the dataset the first time and will not do it again.
I sometimes need to copy the entire code. In the meantime, I want to avoid downloading or copying the dataset again so I try to create shortcut folder named “data” in the copied code (hopefully it can replace the real “data” folder that stores the dataset). However, it always downloads the dataset and create another folder name “data” despite that a shortcut with such name has been present.
Any idea to fix this? Thank you.
I am quite new to pytorch, so this might be a trivial question. I apologize for any inconvenience caused.