/Vision-Transformer-ViT

Pytorch ViT for Image classification on the CIFAR10 dataset

Primary LanguagePythonMIT LicenseMIT

Vit-ImageClassification

Introduction

This project uses ViT to perform image classification tasks on DATA set CIFAR10. The implement of Vit and pretrained weight are from https://github.com/asyml/vision-transformer-pytorch.

The architecture of ViT

Installation

pytorch 1.7.1 python 3.7.3

Datasets

Download the CIFAR10 from http://www.cs.toronto.edu/~kriz/cifar.html or you can get it from https://pan.baidu.com/s/1ogAFopdVzswge2Aaru_lvw (code: k5v8), creat data floder and unzip the cifar-10-python.tar.gz under './data'.

Pre_trained model

You can download the pretrained file from https://pan.baidu.com/s/1CuUj-XIXwecxWMEcLoJzPg (code: ox9n), creat Vit_weights floder and pretrained file under ./Vit_weights

Train

python main.py 

Result

Base on the pretrained weight, after one epoch, we get 98.1 Accuracy

model dataset acc
ViT-B_16 CIFAR10 98.1