Lala8383/ms-marco-qa-10k

--- language: - en license: other license_name: microsoft-research-license license_link: https://microsoft.github.io/msmarco/ tags: - question-answering - ms-marco - reading-comprehension source_datasets: - microsoft/ms_marco task_categories: - question-answering dataset_info: config_name: default pretty_name: MS MARCO QA Subset (10K) --- # MS MARCO QA Subset (10K) This is a **subset** of the [MS MARCO v1.1 dataset](https://huggingface.co/datasets/microsoft/ms_marco) by Microsoft, sampled for lightweight experimentation. ## Source - **Original dataset**: [microsoft/ms_marco](https://huggingface.co/datasets/microsoft/ms_marco) (v1.1) - **Original paper**: [MS MARCO: A Human Generated MAchine Reading COmprehension Dataset](https://arxiv.org/abs/1611.09268) - **Original authors**: Tri Nguyen, Mir Rosenberg, Xia Song, Jianfeng Gao, Saurabh Tiwary, Rangan Majumder, Li Deng (Microsoft) ## What was changed - Randomly sampled **10,000** examples from the train split (seed=42) - Randomly sampled **1,000** examples from the validation split (seed=42) - No other modifications were made to the data ## License This dataset is derived from MS MARCO, which is released under the [Microsoft Research License](https://microsoft.github.io/msmarco/). Please refer to the original license terms before use. ## Citation If you use this dataset, please cite the original MS MARCO paper: ```bibtex @article{nguyen2016ms, title={MS MARCO: A Human Generated MAchine Reading COmprehension Dataset}, author={Nguyen, Tri and Rosenberg, Mir and Song, Xia and Gao, Jianfeng and Tiwary, Saurabh and Majumder, Rangan and Deng, Li}, journal={arXiv preprint arXiv:1611.09268}, year={2016} } ```language: - 英语 license: other license_name: 微软研究许可证（microsoft-research-license） license_link: https://microsoft.github.io/msmarco/ tags: - 问答（question-answering） - ms-marco ms-marco - 阅读理解（reading-comprehension） source_datasets: - microsoft/ms_marco task_categories: - 问答（question-answering） dataset_info: config_name: 默认配置 pretty_name: MS MARCO 问答子集（10K） # MS MARCO 问答子集（10K） 本数据集为微软发布的[MS MARCO v1.1数据集（MS MARCO v1.1 dataset）](https://huggingface.co/datasets/microsoft/ms_marco)的子集，经采样以支持轻量化实验。 ## 数据集来源 - **原始数据集**：[microsoft/ms_marco](https://huggingface.co/datasets/microsoft/ms_marco)（v1.1版本） - **原始论文**：[MS MARCO：一个人工生成的机器阅读理解数据集（MS MARCO: A Human Generated MAchine Reading COmprehension Dataset）](https://arxiv.org/abs/1611.09268) - **原始作者**：Tri Nguyen、Mir Rosenberg、Xia Song、Gao Jianfeng、Saurabh Tiwary、Rangan Majumder、Li Deng（均来自微软） ## 数据集修改说明 - 从训练划分中随机采样**10000**条样本（随机种子设为42） - 从验证划分中随机采样**1000**条样本（随机种子设为42） - 未对数据进行其他修改 ## 许可证说明 本数据集衍生自MS MARCO，后者基于[微软研究许可证（Microsoft Research License）](https://microsoft.github.io/msmarco/)发布。使用前请查阅原始许可证条款。 ## 引用说明 若使用本数据集，请引用原始MS MARCO论文： bibtex @article{nguyen2016ms, title={MS MARCO: A Human Generated MAchine Reading COmprehension Dataset}, author={Nguyen, Tri and Rosenberg, Mir and Song, Xia and Gao, Jianfeng and Tiwary, Saurabh and Majumder, Rangan and Deng, Li}, journal={arXiv preprint arXiv:1611.09268}, year={2016} }