Skip to content

Dataset · StoryAura

Danbooru tags, synonyms, and metadata CSV dataset

StoryAura/Danbooru-Dataset-csv

Hundreds of thousands of structured character, artist, and general tags without the images themselves.

What's inside

A collection of cleaned CSV tables with metadata and tag wiki descriptions from Danbooru in English, Chinese, and Japanese, distributed under the MIT license. It contains no images, only text tables: over 106,000 general tags with hierarchy, nearly 330,000 character tags, about 577,000 artist tags, and synonym relations. The dataset was gathered for developing automatic tagging systems and database management.

What you can do with it

  • Training and tuning automated tagging and prompt normalization systems for image generation
  • Creating synonym dictionaries and text tag classifiers in three languages
  • Building relation graphs between characters, titles, and stylistic tags
Tasks
Text classification
Languages
English, Chinese, Japanese
Size
100K–1M rows
Downloads this month
985
Likes
4

How to load it

from datasets import load_dataset

ds = load_dataset("StoryAura/Danbooru-Dataset-csv")