irds/beir_msmarco_train

---
pretty_name: '`beir/msmarco/train`'
viewer: false
source_datasets: ['irds/beir_msmarco']
task_categories:
- text-retrieval
---

# Dataset Card for `beir/msmarco/train`

The `beir/msmarco/train` dataset, provided by the [ir-datasets](https://ir-datasets.com/) package.
For more information about the dataset, see the [documentation](https://ir-datasets.com/beir#beir/msmarco/train).

# Data

This dataset provides:
- `queries` (i.e., topics); count=502,939
- `qrels`: (relevance assessments); count=532,751

- For `docs`, use [`irds/beir_msmarco`](https://huggingface.co/datasets/irds/beir_msmarco)

## Usage

```python
from datasets import load_dataset

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

qrels = load_dataset('irds/beir_msmarco_train', 'qrels')
for record in qrels:
record # {'query_id': ..., 'doc_id': ..., 'relevance': ..., 'iteration': ...}

```

Note that calling `load_dataset` will download the dataset (or provide access instructions when it's not public) and make a copy of the
data in 🤗 Dataset format.

## Citation Information

```
@inproceedings{Bajaj2016Msmarco,
title={MS MARCO: A Human Generated MAchine Reading COmprehension Dataset},
author={Payal Bajaj, Daniel Campos, Nick Craswell, Li Deng, Jianfeng Gao, Xiaodong Liu, Rangan Majumder, Andrew McNamara, Bhaskar Mitra, Tri Nguyen, Mir Rosenberg, Xia Song, Alina Stoica, Saurabh Tiwary, Tong Wang},
booktitle={InCoCo@NIPS},
year={2016}
}
@article{Thakur2021Beir,
title = "BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models",
author = "Thakur, Nandan and Reimers, Nils and Rücklé, Andreas and Srivastava, Abhishek and Gurevych, Iryna",
journal= "arXiv preprint arXiv:2104.08663",
month = "4",
year = "2021",
url = "https://arxiv.org/abs/2104.08663",
}
```

原始信息汇总

数据集概述

数据集名称

beir/msmarco/train

数据集来源

由ir-datasets包提供。

数据集内容

• queries：查询语句，数量为502,939。
• qrels：相关性评估，数量为532,751。
• docs：文档数据，需从irds/beir_msmarco获取。

数据集使用示例

python from datasets import load_dataset

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

qrels = load_dataset(irds/beir_msmarco_train, qrels) for record in qrels: record # {query_id: ..., doc_id: ..., relevance: ..., iteration: ...}

引用信息

@inproceedings{Bajaj2016Msmarco, title={MS MARCO: A Human Generated MAchine Reading COmprehension Dataset}, author={Payal Bajaj, Daniel Campos, Nick Craswell, Li Deng, Jianfeng Gao, Xiaodong Liu, Rangan Majumder, Andrew McNamara, Bhaskar Mitra, Tri Nguyen, Mir Rosenberg, Xia Song, Alina Stoica, Saurabh Tiwary, Tong Wang}, booktitle={InCoCo@NIPS}, year={2016} } @article{Thakur2021Beir, title = "BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models", author = "Thakur, Nandan and Reimers, Nils and Rücklé, Andreas and Srivastava, Abhishek and Gurevych, Iryna", journal= "arXiv preprint arXiv:2104.08663", month = "4", year = "2021", url = "https://arxiv.org/abs/2104.08663", }

搜集汇总

背景与挑战

背景概述

该数据集是BEIR基准中的'beir/msmarco/train'训练集，用于信息检索模型的零样本评估，包含约50.3万条查询和53.3万条相关性评估数据，文档部分需通过'irds/beir_msmarco'数据集获取。它基于MS MARCO数据集构建，适用于文本检索任务的研究和开发。