maknee/

bioasq_bier_1024_1m

Generated embeddings dataset for vector database training and evaluation.

This dataset contains 1,000,000 text samples with vector embeddings (1024 dimensions) generated from the BioASQ-BEIR dataset using Qwen/Qwen3-Embedding-8B.

Base dataset: 1,000,000 samples with embeddings Embedding dimension: 1024

	Repository Structure




	
	parquet/

base.parquet - Main… See the full description on the dataset page: https://huggingface.co/datasets/maknee/bioasq_bier_1024_1m.