datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
zerochan_full
Zerochan Full Dataset
This is the full dataset of zerochan.net. And all the original images are maintained here.
Thanks to @AngelBottomless for supporting the computational resources for the creation of this dataset.
How to Painlessly Use This
Use cheesechaser to quickly get images from this repository.
Before using this code, you have to grant the access from this gated repository. And then set your personal HuggingFace token into HF_TOKEN environment variable to give… See the full description on the dataset page: https://huggingface.co/datasets/deepghs/zerochan_full.zerochan-wip-testzerochan
CommonArch - Zerochan
Description
This dataset contains a dump of image resources from the website Zerochan, encompassing a wide variety of anime-style illustrations.
It aims to provide a comprehensive collection for training anime-style image generation models and other related research.
This dataset contains 3,843,124 illustrations scraped up to 2024-10-19 11:57:31 UTC.and is regularly updated every 3 months via web crawling.
The next update is scheduled for the end… See the full description on the dataset page: https://huggingface.co/datasets/zerochan-archive/zerochan.zerochan-character-w640-ws-m50-full
Zerochan Single-Character Webdataset Full Dataset
This is the webdataset subset dataset for animetimm/zerochan-character-w640.
All the images here are guaranteed to be non-monochrome, single-person, single-headed, single-faced and have one primary character.
Can be used for single-label anime character classification training.
Images here are resized to min(width, height) <= 640.
Bounding boxes of faces, heads and persons are provided in the json metadata, in the form of (x0, y0… See the full description on the dataset page: https://huggingface.co/datasets/animetimm/zerochan-character-w640-ws-m50-full.zerochan-webp-4Mpixel
Zerochan Newest 4M Re-encoded Dataset
This is the re-encoded dataset of deepghs/zerochan_full. And all the resized images are maintained here.
There are 4106075 images in total. The maximum ID of these images is 4588400. Last updated at 2025-09-22 02:45:45 JST.
How to Painlessly Use This
Use cheesechaser to quickly get images from this repository.
Before using this code, you have to grant the access from this gated repository. And then set your personal HuggingFace token… See the full description on the dataset page: https://huggingface.co/datasets/deepghs/zerochan-webp-4Mpixel.zerochan-wipzerochan-512pxzerochan-character-w640-ws-m200-100k
Zerochan Single-Character Webdataset 100k Sub-Dataset
This is the webdataset subset dataset for animetimm/zerochan-character-w640.
All the images here are guaranteed to be non-monochrome, single-person, single-headed, single-faced and have one primary character.
Can be used for single-label anime character classification training.
Images here are resized to min(width, height) <= 640.
Bounding boxes of faces, heads and persons are provided in the json metadata, in the form of (x0… See the full description on the dataset page: https://huggingface.co/datasets/animetimm/zerochan-character-w640-ws-m200-100k.Zerochan-webp-examplezerochan-metazerochan-webdataset