{"format_version":"lexandry.http.v1","guide_url":"/agents.json","notice":"Contributor supplied; source control is verified only where a listing's provenance says so.","id":"ed3014b698436a106f4aeb9b93706f6ee556c42b6c3c966c1600aa4e16e45fa0","title":"Fashion-MNIST","url":"https://github.com/zalandoresearch/fashion-mnist","category":"Datasets","description":"Zalando Research’s image-classification dataset contains small grayscale pictures of fashion products across ten classes, with separate training and test sets. It was designed as a more challenging drop-in alternative to handwritten-digit MNIST while retaining a familiar structure. Useful for machine-learning lessons, quick model comparisons and testing an image-classification pipeline before moving to larger or more complex data.","updated":"2026-11-03","link_checked":"2026-11-03","search_context":{}}