Skip to content

Dataset · ma-xu

Corpus for Text-to-Image Fine-Tuning

ma-xu/fine-t2i

A large open collection of English image-text pairs for high-quality text-to-image fine-tuning.

What's inside

An open, large-scale, and diverse dataset for high-quality text-to-image fine-tuning. It contains images and English texts with a volume of 100,000 to 1 million examples. Authored by Xu Ma, Yitian Zhang, Qihua Dong, and Yun Fu from Northeastern University, with an accompanying arXiv paper. The card also notes its top ranking in Hugging Face Datasets Trending.

What you can do with it

  • Fine-tune a text-to-image model or LoRA on image-text pairs
  • Check generation quality based on text prompts
  • Build a text-to-image fine-tuning pipeline
Tasks
Image → text, Text → image
Languages
English
Size
100K–1M rows
Downloads this month
37.9K
Likes
136

How to load it

from datasets import load_dataset

ds = load_dataset("ma-xu/fine-t2i")