PKU-ONELab/NLG-Eval
收藏Hugging Face2025-02-23 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/PKU-ONELab/NLG-Eval
下载链接
链接失效反馈官方服务：
资源简介：
NLG-Eval是一个大规模的自然语言生成（NLG）任务评估语料库。该数据集通过收集9个常见NLG任务的58个评估数据集，并辅以人类和GPT-4的注释构建而成。数据集包含了元信息，并分为三个文件：Meta-info.csv、NLG-Eval.jsonl和Train.jsonl，每个文件都有特定的内容和用途。
NLG-Eval is a large-scale evaluation corpus for natural language generation (NLG) tasks. It is constructed by collecting 58 evaluation datasets across 9 common NLG tasks, supplemented with annotations from both human and GPT-4. The dataset includes meta-information and is divided into three files: Meta-info.csv, NLG-Eval.jsonl, and Train.jsonl, each containing specific content and serving a particular purpose.
提供机构：
PKU-ONELab
