RECAPTIONING BILLIONS OF IMAGES WITH LLAMA 3

Summary from the TLDR AI newsletter

Recaptioning web scraped images with a vision model dramatically improves downstream model performance. This is especially true in models such as CLIP.