Dataset · Hugging Face

facebook/wiki_dpr

Name: wiki_dpr
Creator: facebook
License: cc-by-nc-4.0
Keywords: task_categories:fill-mask, task_categories:text-generation, task_ids:language-modeling, task_ids:masked-language-modeling, annotations_creators:no-annotation, language_creators:crowdsourced, multilinguality:multilingual, source_datasets:original, language:en, license:cc-by-nc-4.0, size_categories:10M<n<100M, arxiv:2004.04906, region:us, text-search

This is the wikipedia split used to evaluate the Dense Passage Retrieval (DPR) model. It contains 21M passages from wikipedia along with their DPR embeddings. The wikipedia articles were split into multiple, disjoint text blocks of 100 words as passages.

View on Hugging FaceReport an issue

Author / organization	facebook
License (as declared)	cc-by-nc-4.0
Downloads	26.6K (as reported by Hugging Face)
Likes	46
Created	2 March 2022
Last updated at source	12 March 2024

text-search

Embed this entry

<iframe src="https://silytics.si/embed/item/9efb6212-8e3d-48c4-9e28-594c41d0d30e" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

More from facebook

facebook/uco3dDataset
facebook/seamless-interactionDataset
facebook/anliDataset
facebook/multilingual_librispeechDataset
facebook/belebeleDataset
facebook/xnliDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 15 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.