/conceptual-12m

Conceptual 12M is a dataset containing (image-URL, caption) pairs collected for vision-and-language pre-training.

OtherNOASSERTION

No issues in this repository yet.