somosnlp-hackathon-2025/ibero-tales-es

这是一个基于伊比利亚美洲神话和传说生成的合成故事数据集，旨在为语言模型在文化敏感叙事方面的训练和对齐提供支持。数据集包含了根据伊比利亚美洲神话和传说中的人物和主题生成的故事，每个故事都包括一个与故事相关的问题和答案。此数据集目前处于积极开发中，计划大幅增加记录数量并改进生成过程以提高叙事的多样性和质量。

This is a dataset of synthetic stories generated from myths and legends of Iberoamerica, curated and structured to support training and alignment of language models in culturally sensitive narrative. The dataset includes stories generated based on characters and themes from Iberoamerican myths and legends, each story accompanied by a question related to the story and an answer. This dataset is currently under active development, with plans to significantly increase the number of records and refine the generation process to enhance the diversity and quality of the narratives.