tinycompany/Instruct-Is-All-You-Need
收藏Hugging Face2025-03-17 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/tinycompany/Instruct-Is-All-You-Need
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多种来源的指令数据，包括纯英文、纯印地语以及混合语言的数据。数据集中的特征包括消息内容、角色和来源。训练集包含了大约355万个样本，整个数据集大小约为12.9TB。数据来源于不同的项目，例如BAAI、BhabhaAI、smol-magpie-ultra-short等，每个来源都有不同的样本数量和特点。
The dataset comprises instruction data from various sources, including pure English, pure Hindi, and mixed language data. The dataset features include message content, role, and source. The training set contains approximately 3.55 million samples, and the entire dataset is about 12.9TB in size. The data sources range from projects like BAAI, BhabhaAI, smol-magpie-ultra-short, and others, each with different sample counts and characteristics.
提供机构：
tinycompany
