Sign inGet started
Dataset · Hugging Face

mteb/IWSLT2017BitextMining

IWSLT2017BitextMining An MTEB dataset Massive Text Embedding Benchmark The IWSLT 2017 Multilingual Task addresses text translation, including zero-shot translation, with a single MT system across all directions including English, German, Dutch, Italian and Romanian. Task category t2t Domains Non-fiction, Fiction, Written Reference https://aclanthology.org/2017.iwslt-1.1/ How to evaluate on this task You can evaluate an embedding model on this… See the full description on the dataset page: https://huggingface.co/datasets/mteb/IWSLT2017BitextMining.

Author / organizationmteb
License (as declared)cc-by-nc-nd-4.0
Downloads199 (as reported by Hugging Face)
Likes0
Created30 November 2024
Last updated at source4 May 2025
mtebtext
Embed this entry
<iframe src="https://silytics.si/embed/item/8d8b6415-bb2c-46ee-a10e-015135c959c8" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 2 days

199 now (-12 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Recent changes

  • 17 Oct 2026 Description: IWSLT2017BitextMining An MTEB dataset Massive Text Embedding Benchmark The IWSLT 2017 Multilingual Task addresses text translation, including zero-shot translation, with a single MT system across all directions including English, German, Du → IWSLT2017BitextMining An MTEB dataset Massive Text Embedding Benchmark The IWSLT 2017 Multilingual Task addresses text translation, including zero-shot translation, with a single MT system across all directions including English, German, Du

Changes detected by Silytics between refreshes of the source.

More from mteb

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 Information may be incomplete or out of date.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.