Solve Pytorch data sets, such as data download slow MNIST
Data is the soul of the depth of learning, and in the download data sets, downloading the half stalled ......
Then look at the download directory you set in MNISTthere will be two folders: processedandraw
Open rawcan see already downloaded the data set.
In fact MNIST data set is small, direct access to the site, manually download the data set: http://yann.lecun.com/exdb/mnist/
The four downloaded .gzcopy files to a rawfolder (can not unpack).
Such data can be regarded as a good download. Then DOWNLOAD_MNIST= False, you can run the program directly.
