README.md

metadata

license: other
task_categories:
  - image-classification
  - zero-shot-image-classification
  - text-to-image
language:
  - en
tags:
  - art
  - anime
  - not-for-all-audiences
size_categories:
  - 100K<n<1M
annotations_creators:
  - no-annotation
source_datasets:
  - danbooru

Danbooru Newest 4M Re-encoded Dataset

This is the re-encoded dataset of deepghs/danbooru_newest. And all the resized images are maintained here.

There are 826816 images in total. The maximum ID of these images is 8203522. Last updated at 2024-09-27 20:07:56 UTC.

This dataset only contains newest webp images, if you are looking for old webp images, just see KBlueLeaf/danbooru2023-webp-4Mpixel.

How to Painlessly Use This

Use cheesechaser to quickly get images from this repository and KBlueLeaf/danbooru2023-webp-4Mpixel.

from cheesechaser.datapool import DanbooruNewestWebpDataPool

pool = DanbooruNewestWebpDataPool()
pool.batch_download_to_directory(
    resource_ids=[
        # older images, from KBlueLeaf/danbooru2023-webp-4Mpixel
        *range(7382200, 7382288),

        # newest images, from deepghs/danbooru_newest-webp-4Mpixel
        *range(7782200, 7782288),
    ],

    # save to directory /data/webp_danbooru
    dst_dir='/data/webp_danbooru',
)