All 10 datasets of the multilingual-vlm-conflict suite in one place, in English plus 22
languages translated with CohereLabs/aya-expanse-32b
(Arabic, Czech, German, Greek, Spanish, Persian, French, Hebrew, Hindi, Indonesian, Italian,
Japanese, Korean, Dutch, Polish, Portuguese, Romanian, Russian, Turkish, Ukrainian, Vietnamese,
Simplified Chinese). 10 datasets x 23 languages x 100 examples = 23,000 rows.
Each example pairs an image with a… See the full description on the dataset page:
https://huggingface.co/datasets/multilingual-vlm-conflict/combined.
