nectec/nrct_instruction_dataset

NRCT指令数据集(V1)是一个用于自然语言处理模型训练的数据集，每个样本包含一个唯一标识符、任务类型、指令、三个不同回答以及每个回答的质量评分。任务类型包括但不限于拼写校正、翻译、问题回答等。

NRCT Instruction Dataset (V1) is a dataset for training natural language processing models, with each sample including a unique identifier, task type, instruction, three different answers, and quality scores for each answer. Task types include but are not limited to spell correction, translation, question-answering, etc.