withmartian/cs12_dataset

该数据集是一个包含表格数据和SQL语句的复杂结构化数据集，旨在用于训练和评估自然语言处理模型在理解自然语言指令并生成相应的SQL查询语句方面的能力。数据集包含训练集、验证集和测试集，提供了表格名称、指令短语、列名等字段，以及用于评估SQL语句有效性的验证信息和LLM模型判断结果。

This dataset is a complex structured dataset containing table data and SQL statements, designed for training and evaluating NLP models in their ability to understand natural language instructions and generate corresponding SQL queries. The dataset includes train, validation, and test sets, providing fields such as table names, instruction phrases, column names, and validation information for assessing the validity of SQL statements along with LLM model judgment results.