toloka/JEEM
Dataset Card for JEEM ๐ Curated by: Toloka, MBZUAI ๐ Language(s): Modern Standard Arabic, ๐ฏ๐ด Jordanian dialect, ๐ช๐ฌ Egyptian dialect, ๐ฆ๐ช Emirati dialect, ๐ฒ๐ฆ Moroccan dialect ๐ Dataset Description JEEM is a benchmark dataset designed to evaluate Vision-Language Models (VLMs) in the context of Arabic dialectal diversity. It includes one representative dialect from each major Arabic dialectal region: ๐ฏ๐ด Jordanian (Levantine)โฆ See the full description on the dataset page: https://huggingface.co/datasets/toloka/JEEM.
| Author / organization | toloka |
|---|---|
| License (as declared) | mit |
| Downloads | 141 (as reported by Hugging Face) |
| Likes | 14 |
| Created | 28 February 2025 |
| Last updated at source | 11 May 2025 |
Popularity, last 2 days
141 now (-3 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 17 Oct 2026 Description: Dataset Card for JEEM ๐ Curated by: Toloka, MBZUAI ๐ Language(s): Modern Standard Arabic, ๐ฏ๐ด Jordanian dialect, ๐ช๐ฌ Egyptian dialect, ๐ฆ๐ช Emirati dialect, ๐ฒ๐ฆ Moroccan dialect ๐ Dataset Description JEEM is a benchmark dataset design โ Dataset Card for JEEM ๐ Curated by: Toloka, MBZUAI ๐ Language(s): Modern Standard Arabic, ๐ฏ๐ด Jordanian dialect, ๐ช๐ฌ Egyptian dialect, ๐ฆ๐ช Emirati dialect, ๐ฒ๐ฆ Moroccan dialect ๐ Dataset Description JEEM is a benc
Changes detected by Silytics between refreshes of the source.
More from toloka
- toloka/crowdkit-datasetsDataset
- toloka/beemoDataset
- toloka/VOX-DUBDataset
- toloka/WSDMCup2023Dataset
- toloka/u-mathDataset
- toloka/homer-v2Dataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.