/ShuffleNet-1g8-Pytorch

Model shared. Top1:67.408/Top5:87.258

Primary LanguagePythonBSD 2-Clause "Simplified" LicenseBSD-2-Clause

ShuffleNet-1g8-Pytorch 
Introduction

This is a Pytorch implementation of faceplusplus's ShuffleNet-1g8. For details, please read the following papers: 
	ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices

Pretrained Models on ImageNet

We provide pretrained ShuffleNet-1g8 models on ImageNet, which achieve nearly accuracy with the original ones reported in the paper.

The top-1/5 accuracy rates by using single center crop (crop size: 224x224, image size: 256xN): 
Network 	Top-1 	Top-5	Top-1(reported in the paper) 
ShuffleNet-1g8 	67.408 	87.258	67.60

Evaluate Models python eval.py -a shufflenet --evaluate ./ShuffleNet_1g8_Top1_67.408_Top5_87.258.pth.tar ./ILSVRC2012/

Dataset prepare Refer to https://github.com/facebook/fb.resnet.torch/blob/master/INSTALL.md#download-the-imagenet-dataset