Abdullah804/yoruba-subset

Yoruba Subset Dataset是尼日利亚约鲁巴语语音数据集的100小时子集，包含101,391个样本，已转换为16kHz WAV格式。数据集包括路径、文本、说话者ID、性别、年龄范围和持续时间等详细元数据，适用于自动语音识别和文本到语音的微调。Yoruba Subset Dataset is a 100-hour subset of the Naija Voices Yoruba speech dataset, containing 101,391 samples converted to 16kHz WAV format. The dataset includes detailed metadata such as path, text, speaker_id, gender, age_range, and duration, and is prepared for ASR/TTS finetuning.