遇见数据集

Suryanshg/SUPER-NATURALINSTRUCTIONS-english-filtered-100x-augmented-paraphased

收藏
Hugging Face2026-05-15 更新2026-05-31 收录
官方服务:

资源简介:

该数据集是一个多语言任务处理数据集,包含训练集和测试集,分别有2,790,891和458,488个示例。数据集特征包括任务名称、指令、输入和输出语言、类别、输入内容、输出内容、总标记数以及释义指令列表,支持多语言交互任务。数据以文件形式存储,总大小约为17.1 GB,下载大小为440 MB。

This dataset is a multilingual task processing dataset comprising training and test splits with 2,790,891 and 458,488 examples, respectively. It includes features such as task name, instruction, input and output languages, categories, input content, output content, total tokens, and paraphrased instructions lists, designed for multilingual interaction tasks. The data is stored in file formats, with a total size of approximately 17.1 GB and a download size of 440 MB.

提供机构:
Suryanshg
二维码
社区交流群
二维码
科研交流群
商业服务