yguooo/summarize_from_feedback_tldr_3_filtered_oai_preprocessing_llama3
收藏Hugging Face2024-11-04 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/yguooo/summarize_from_feedback_tldr_3_filtered_oai_preprocessing_llama3
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是从OpenAI的Summarize from Feedback任务中提取的，主要用于文本摘要任务。数据集包含Reddit帖子的标题、正文、摘要等信息，并经过预处理生成了查询、参考响应等字段。数据集分为训练集、验证集和测试集，每个集都包含大量的字节和样本。预处理脚本对原始数据进行了长度限制和填充处理，以确保输入数据的统一性。
The dataset is directly taken from OpenAIs Summarize from Feedback task, primarily used for training and evaluating models for summarization based on feedback. The dataset includes multiple fields extracted from Reddits TL;DR dataset, such as id, subreddit, title, post, and summary. Additionally, some fields generated by the preprocessing script, such as query and query_token, are added to support model training and evaluation. The dataset is divided into training, validation, and test sets, containing 116722, 6447, and 6553 samples respectively.
提供机构：
yguooo
