fish666/T2Retrieval
收藏Hugging Face2026-04-29 更新2026-05-03 收录
下载链接：
https://hf-mirror.com/datasets/fish666/T2Retrieval
下载链接
链接失效反馈官方服务：
资源简介：
T2Retrieval是一个大规模中文文本检索基准数据集，属于MTEB（Massive Text Embedding Benchmark）的一部分。该数据集专注于段落排名任务，覆盖医学、学术、金融、政府和非小说等多个领域。数据集包含三个配置：语料库（corpus）、默认（default）和查询（queries），每个配置都有特定的特征和开发集（dev）划分。数据集由人工标注，采用Apache-2.0许可证。
T2Retrieval is a large-scale Chinese benchmark dataset for passage ranking, part of the Massive Text Embedding Benchmark (MTEB). The dataset focuses on passage ranking tasks across multiple domains including Medical, Academic, Financial, Government, and Non-fiction. It includes three configurations: corpus, default, and queries, each with specific features and a dev split. The dataset is human-annotated and licensed under Apache-2.0.
提供机构：
fish666
