golaxy/KnowCoder-Schema-Following-Data
收藏Hugging Face2024-04-09 更新2024-06-11 收录
下载链接:
https://hf-mirror.com/datasets/golaxy/KnowCoder-Schema-Following-Data
下载链接
链接失效反馈官方服务:
资源简介:
Schema Following Data是基于UniversalNER、InstructIE和LSEE构建的数据集,用于将结构化知识编码到大型语言模型(LLMs)中,以实现通用信息提取。该数据集包含多个任务的数据样本,但由于数据保护的原因,每个任务仅提供了100个样本供参考。
Schema Following Data is a dataset constructed based on UniversalNER, InstructIE, and LSEE. It is developed for encoding structured knowledge into Large Language Models (LLMs) to enable general information extraction. This dataset includes data samples across multiple tasks, yet due to data protection considerations, only 100 samples per task are provided for reference.
提供机构:
golaxy
原始信息汇总
数据集概述
数据集配置
- 配置名称: default
- 数据文件:
- split: schema_following_ner
- 路径: instruction-tuning/NER/train-prompt.json
- split: schema_following_re
- 路径: instruction-tuning/RE/train-prompt.json
- split: schema_following_ed
- 路径: instruction-tuning/ED/train-prompt.json
- split: schema_following_eae
- 路径: instruction-tuning/EAE/train-prompt.json
- split: schema_following_ner
许可证
- 类型: apache-2.0
语言
- 支持语言: en
标签
- 标签: instruction tuning
大小分类
- 范围: 100K<n<1M
数据来源
- 基于以下资源构建:
数据样例
- 由于数据保护考虑,每个任务仅提供100个数据样本供参考。



