Skip to content

Dataset · deepghs

29M anime and movie frames (Fancaps)

deepghs/fancaps-webp-4Mpixel

A massive archive of WebP-compressed screenshots from visual content for generative models and classification.

What's inside

The dataset contains exactly 29,004,217 images with resolutions up to 4 megapixels, representing a recoded version of the deepghs/fancaps_full collection. Metadata is in English, and the repository is gated, requiring confirmation via a Hugging Face token. The author recommends using the cheesechaser utility for fast image downloading.

What you can do with it

  • Training and fine-tuning (LoRA) generative models in anime and movie frame styles
  • Training image classifiers and zero-shot recognition models
Tasks
Image classification, Zero shot image classification, Text → image
Languages
English
Size
10M–100M rows
Downloads this month
53.7K
Likes
3
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("deepghs/fancaps-webp-4Mpixel")