Data source

Data from GitHub (zalandoresearch), delivered clean.

4 datasets pulled from GitHub (zalandoresearch)'s releases, checked field by field and shipped the way you want them — daily, weekly, or hourly, your call.

  • 4 datasets
  • 2 industrys
  • Real rows on request

Apparel, Accessories & Luxury Goods

3
Apparel, Accessories & Luxury Goods Zalando article imagery (company product cata… · Single snapshot released August 2017

Fashion-MNIST - Zalando Research Apparel Image Dataset

Apparel, Accessories & Luxury Goods Global - web-sourced shop and consumer fashio… · Static research release

DeepFashion Dataset (CUHK MMLab)

Apparel, Accessories & Luxury Goods Web-sourced fashion photography from six phot… · Fixed research corpus published at ECCV 2020

Fashionpedia — Detection Datasets on Hugging Face

Apparel Retail

1
Apparel Retail Not geographic - Zalando e-commerce article p… · Static snapshot released in 2017

Kaggle Fashion MNIST Mirror

Pick a catch, see the rows.

Name any GitHub (zalandoresearch) dataset and we send real rows from it — not a screenshot of rows. 1,744 datasets. Pick your catch.

Get a sample

API, files, or your warehouse. Daily, weekly, or hourly.

Straight answers about GitHub (zalandoresearch) data

What does Datadory deliver from GitHub (zalandoresearch)?

One Apparel, Accessories & Luxury Goods collection - Fashion-MNIST, 70,000 labeled 28x28 grayscale product images across ten classes, split 60,000 train and 10,000 test - shipped with a verified field dictionary, stratified sample rows and a coverage statement attached.

How many images does the Fashion-MNIST corpus hold?

70,000 in total: 60,000 training examples and 10,000 test examples, roughly 30 MB compressed - 26 MB of training images, 4.3 MB of test images plus the two label files. Small enough to iterate on a laptop, large enough that class balance and model capacity actually matter.

What are the ten classes in the label set?

0 T-shirt/top, 1 Trouser, 2 Pullover, 3 Dress, 4 Coat, 5 Sandal, 6 Shirt, 7 Sneaker, 8 Bag and 9 Ankle boot - garments and accessories taken from Zalando's article inventory. Each integer pairs with one grayscale image, giving every class thousands of training examples.

Does the corpus change over time?

No - the contents are a single snapshot released in August 2017 and have stayed byte-stable since. That permanence is the feature: accuracy recorded today remains comparable with numbers reported years ago, and any re-cut of the corpus happens on your schedule, not anyone else's.