2021 Call for Code Awards: Live from New York, with SNL’s Colin Jost! Learn more


The Fashion-MNIST dataset contains 60,000 training images (and 10,000 test images) of fashion and clothing items, taken from 10 classes. Each image is a standardized 28×28 size in grayscale (784 total pixels). Fashion-MNIST was created by Zalando as a compatible replacement for the original MNIST dataset of handwritten digits. This version of the dataset has been converted to CSV to enable easier loading in common data science tools (see https://www.kaggle.com/zalando-research/fashionmnist).

Dataset Metadata

Field Value
Format CSV
License MIT
Domain Image Classification
Number of Records 60,000 (images)
Size 37 MB
Author Kashif Rasul, Han Xiao
Dataset Origin Zalando Research
Dataset Version Version 1 – 9/12/2019
Dataset Coverage A dataset of standardized images of fashion items from 10 classes: T-shirt/top, Trouser, Pullover, Dress, Coat, Sandal, Shirt, Sneaker, Bag, Ankle boot
Business Use Case Retail – Classify images of clothing available for sale in a store.

Dataset Archive Contents

File or Folder Description
fashion-mnist_train.csv Training split
fashion-mnist_test.csv Test split
LICENSE.txt Terms of Use

Data Glossary and Preview

Click here to explore the data glossary, sample records, and additional dataset metadata.

Use the Dataset

This dataset is complemented by data exploration, data analysis, and modeling Python notebooks to help you get started:

Quick access in Python (requires the pardata pypi package):

$ pip install pardata

import pardata
data = pardata.load_dataset('fashion_mnist')


author="Xiao, Han
and Rasul, Kashif
and Vollgraf, Roland
title="Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms",