MtCelesteMa/multiglue

--- license: cc-by-4.0 task_categories: - text-classification size_categories: - 100K<n<1M language: - en multilinguality: - monolingual pretty_name: MultiGLUE source_datasets: - extended|glue language_creators: - found annotations_creators: - found --- # Dataset Card for MultiGLUE ## Dataset Description - **Homepage:** - **Repository:** - **Paper:** - **Leaderboard:** - **Point of Contact:** ### Dataset Summary This dataset is a combination of the cola, mrpc, qnli, qqp, rte, sst2, and wnli subsets of the GLUE dataset. Its intended use is to benchmark language models on multitask binary classification. ### Supported Tasks and Leaderboards [More Information Needed] ### Languages Like the GLUE dataset, this dataset is in English. ## Dataset Structure ### Data Instances An example instance looks like this: ``` { "label": 1, "task": "cola", "sentence1": "The sailors rode the breeze clear of the rocks.", "sentence2": null } ``` ### Data Fields - `task`: A `string` feature, indicating the GLUE task the instance is from. - `sentence1`: A `string` feature. - `sentence2`: A `string` feature. - `label`: A classification label, either 0 or 1. ### Data Splits - `train`: 551,282 instances - `validation`: 48,564 instances - `test`: 404,183 instances, no classification label (same as GLUE) ## Dataset Creation ### Curation Rationale [More Information Needed] ### Source Data #### Initial Data Collection and Normalization This dataset is created by combining the cola, mrpc, qnli, qqp, rte, sst2, and wnli subsets of the GLUE dataset. #### Who are the source language producers? [More Information Needed] ### Annotations #### Annotation process [More Information Needed] #### Who are the annotators? [More Information Needed] ### Personal and Sensitive Information [More Information Needed] ## Considerations for Using the Data ### Social Impact of Dataset [More Information Needed] ### Discussion of Biases [More Information Needed] ### Other Known Limitations [More Information Needed] ## Additional Information ### Dataset Curators [More Information Needed] ### Licensing Information [More Information Needed] ### Citation Information [More Information Needed] ### Contributions [More Information Needed]

原始信息汇总

数据集卡片 MultiGLUE

数据集描述

数据集概述

该数据集是GLUE数据集的cola、mrpc、qnli、qqp、rte、sst2和wnli子集的组合。其目的是用于在多任务二分类上基准测试语言模型。

支持的任务和排行榜

[更多信息需要]

语言

与GLUE数据集一样，该数据集是英文的。

数据集结构

数据实例

一个示例实例如下： json { "label": 1, "task": "cola", "sentence1": "The sailors rode the breeze clear of the rocks.", "sentence2": null }

数据字段

• task: 字符串特征，指示实例来自的GLUE任务。
• sentence1: 字符串特征。
• sentence2: 字符串特征。
• label: 分类标签，0或1。

数据分割

• train: 551,282个实例
• validation: 48,564个实例
• test: 404,183个实例，无分类标签（与GLUE相同）

数据集创建

数据集来源

初始数据收集和规范化

该数据集是通过组合GLUE数据集的cola、mrpc、qnli、qqp、rte、sst2和wnli子集创建的。

源语言生产者

[更多信息需要]

注释

注释过程

[更多信息需要]

注释者

[更多信息需要]

个人和敏感信息

[更多信息需要]

使用数据集的注意事项

数据集的社会影响

[更多信息需要]

偏见的讨论

[更多信息需要]

其他已知限制

[更多信息需要]

附加信息

数据集策展人

[更多信息需要]

许可信息

[更多信息需要]

引用信息

[更多信息需要]

贡献

[更多信息需要]