Dataset · metmuseum
Metropolitan Museum of Art open access collection
metmuseum/openaccess
Images and text descriptions of exhibits from the New York Metropolitan Museum of Art open access collection in English.
What's inside
The dataset collects images and texts representing over 5,000 years of art from the Metropolitan Museum of Art (The Met) collection in New York. It contains between 100k and 1M English records. The data is sourced from the museum open access program.
What you can do with it
- Training multimodal models (VLM) for art description and attribution
- Training generative models (LoRA) for classical painting and sculpture styling
- Building visual search systems for museum archives
- Languages
- English
- Size
- 100K–1M rows
- Downloads this month
- 9.4K
- Likes
- 9
How to load it
from datasets import load_dataset
ds = load_dataset("metmuseum/openaccess")