guizme/pulaar_corpus

该数据集包含Pulaar语（富拉语）民间故事Ndimaagu（高贵/尊严）的完整转录。故事讲述了Daado、Yero以及他们面临的关于荣誉和忠诚的挑战。数据集的结构包括每个叙事片段或对话的唯一标识符、叙事内容、原始文件名和语言代码。叙事亮点包括Daado的30天耐力测试、Yero因与狗分享食物而成功、丢失的金项链导致Yero被流放以恢复Daado的荣誉，以及关于一个 deceptive cleric被Daado（伪装成Paate）审判的子情节。数据集适用于微调大型语言模型（LLMs）、西非语言的NLP研究以及口头传统的数字化保存。

This dataset contains the full transcription of the Pulaar folktale Ndimaagu (Nobleness/Dignity). It follows the story of Daado, Yero, and the challenges they face regarding honor and loyalty. The dataset structure includes a unique identifier for each narrative segment or dialogue, the narrative content in Pulaar, the original file name, and the language code. Narrative highlights include Daados 30-day endurance test for suitors, Yeros success due to sharing food with a dog, the lost golden necklace leading to Yeros exile to recover Daados honor, and a sub-plot about a deceptive cleric judged by Daado (disguised as Paate). The dataset is suitable for fine-tuning Large Language Models (LLMs), NLP research on West African languages, and the digital preservation of oral traditions.