nassimjp/

NanoBEIR-ps

	NanoBEIR-ps




	


	Dataset Summary

NanoBEIR-ps is a Pashto retrieval benchmark inspired by the NanoBEIR benchmark. It is designed for evaluating semantic search, dense retrieval, sparse retrieval, hybrid search, and Retrieval-Augmented Generation (RAG) systems in the Pashto language. The dataset provides a lightweight benchmark that can be used to compare embedding models and information retrieval systems without requiring a large-scale corpus.