arcee-ai/The-Tome
收藏Hugging Face2024-08-15 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/arcee-ai/The-Tome
下载链接
链接失效反馈官方服务：
资源简介：
The Tome是一个精心策划的数据集，旨在训练大型语言模型，特别是指令跟随模型。该数据集由9个公开数据集组成，总样本量为175万。数据集经过重新排序、教育价值评分和综合评分等筛选过程，以确保高质量内容。该数据集在Nova模型的开发中起到了关键作用，并最终与Qwen2-72B-Instruct模型合并。
The Tome is a curated dataset designed for training large language models with a focus on instruction following. It consists of 1.75 million samples compiled from 9 publicly available datasets. The dataset includes various specific task datasets such as instruction following and educational value scoring. The dataset underwent reranking and educational value scoring to ensure high-quality content. It played a crucial role in the development of the Nova model, which was subsequently merged with the Qwen2-72B-Instruct model.
提供机构：
arcee-ai
