search

rahulvigneswaran / Lottery-Ticket-Hypothesis-in-Pytorch

This repository contains a Pytorch implementation of the paper "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks" by Jonathan Frankle and Michael Carbin that can be easily adapted to any model/dataset.
325 stars 91 forks source link

Every dataset is normalized with MNIST's mean and std dev #9

Open npnbpb opened 3 years ago

npnbpb commented 3 years ago

The transform used while loading every dataset (line 37, main.py) uses MNIST's mean and standard deviation. The correct values for normalizing CIFAR 10, etc are different.

bainro commented 2 years ago

For cifar10 & cifar100: normalize = transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])