The MNIST dataset from http://yann.lecun.com/exdb/mnist/
The MNIST dataset is required when beginners write the hello world
program in ML. However, many of them in china suffer from the unstable Internet
connection caused by the strict network censorship when trying download this dataset from the offcial site. To remove the obstacle, I mirror them to github.
This respository contains 4 files:
train-images-idx3-ubyte.gz: training set images (9912422 bytes)
train-labels-idx1-ubyte.gz: training set labels (28881 bytes)
t10k-images-idx3-ubyte.gz: test set images (1648877 bytes)
t10k-labels-idx1-ubyte.gz: test set labels (4542 bytes)