irds/neumarco_ru_train_judged

`neumarco/ru/train/judged`数据集由ir-datasets包提供，主要用于文本检索任务。该数据集包含502,939个查询（即主题），但文档、相关性和文档对需要分别从`irds/neumarco_ru`和`irds/neumarco_ru_train`数据集中获取。

The `neumarco/ru/train/judged` dataset is distributed via the ir-datasets package and is primarily utilized for text retrieval tasks. It contains 502,939 queries (referred to as topics in information retrieval contexts), while the associated documents, relevance judgments, and query-document pairs must be separately acquired from the `irds/neumarco_ru` and `irds/neumarco_ru_train` datasets, respectively.

原始信息汇总

数据集概述

数据集名称

neumarco/ru/train/judged

数据来源

• irds/neumarco_ru
• irds/neumarco_ru_train

数据内容

• queries: 包含502,939个查询（topics）。
• docs: 数据源为irds/neumarco_ru。
• qrels: 数据源为irds/neumarco_ru_train。
• docpairs: 数据源为irds/neumarco_ru_train。

使用示例

python from datasets import load_dataset

queries = load_dataset(irds/neumarco_ru_train_judged, queries) for record in queries: record # {query_id: ..., text: ...}