This repository was archived by the owner on Oct 31, 2023. It is now read-only.

Description
Hi,
Thanks so much for sharing the excellent work.
I am wondering if the codes of the experiments on MNIST in the supplementary material can also be shared.
Also, the paper states "We trained our models on 8 Nvidia GPUs approximately for 8 hours for 1000 epochs". May I ask what type of GPUs specifically?