You can deploy this model and start earning money today!

JinChao1022/
fanoutqa-retrieval

information-retrieval

Views

No views yet

FanOutQA Retrieval Dataset

A multi-document retrieval dataset built from FanOutQA, a multi-hop QA benchmark that uses English Wikipedia as its knowledge base.

Dataset Summary

Corpus: 1,594 full Wikipedia pages (averaging ~12,000 words each) Queries: 310 multi-hop questions requiring evidence from multiple documents Query-to-Docs: Each query maps to 5-46 Wikipedia pages (avg 6.9)

	Configs

Config Rows Description

corpus 1,594 Full… See the full description on the dataset page: https://huggingface.co/datasets/JinChao1022/fanoutqa-retrieval.