unimind-kronika
收藏资源简介:
该数据集是一个小规模的对话或问答记录集合,包含5个样本。数据以结构化形式组织,每个样本包含六个字段:id(唯一标识符)、ts(时间戳)、user(用户标识)、q(用户提出的问题或查询)、a(系统或助手给出的回答或响应)、flag(用于标记样本状态的标志位)。数据集仅提供训练集拆分,数据格式为文本字符串,适用于对话系统训练、问答模型构建或对话数据分析和研究等场景。
This dataset is a small-scale collection of conversation and question-answering records, comprising 5 samples. The data is structured in a standardized format, with each sample containing six fields: id (unique identifier), ts (timestamp), user (user identifier), q (user's question or query), a (system or assistant's answer or response), and flag (a status marker for the sample). Only the training set split is provided for this dataset. The data is formatted as plain text strings, and is suitable for applications including dialogue system training, question-answering model development, conversational data analysis and related research.
- 数据集名称:unimind-kronika
- 数据集地址:https://huggingface.co/datasets/StrawberryJelly/unimind-kronika
- 数据集大小:下载大小为3340字节,数据集大小为1828字节
- 数据集结构:包含1个配置(default),1个数据划分(train)
- 数据划分:训练集包含15个样本,1828字节
- 数据字段:
- id(int64):唯一标识符
- ts(large_string):时间戳
- user(large_string):用户信息
- q(large_string):问题
- a(large_string):回答
- flag(large_string):标志位
- 数据文件:训练数据文件位于 data/train-*




