Kukedlc/arg-spanish-tts

arg-spanish-tts是一个用于阿根廷西班牙语（es-AR）的统一、去重语音语料库，通过合并三个公共数据集（giandiego_ar、ylacombe和fabricio3g）并去除跨源重复项构建。所有音频已重采样为24 kHz单声道，包含10,747行数据、12.18小时总音频时长、65个独特说话人以及性别分布（女性7,419条，男性3,328条）。该数据集专为多说话人文本到语音（TTS）预训练设计，适用于在微调到目标声音之前进行模型训练，以提升单说话人训练效果。

arg-spanish-tts is a unified, deduplicated speech corpus for Argentine Spanish (es-AR), built by merging three public datasets and stripping cross-source duplicates. All audio is resampled to 24 kHz mono. It contains 10,747 rows, 12.18 hours of total audio, 65 unique speakers, and a gender split (female: 7,419, male: 3,328). Intended for multi-speaker TTS pre-training before fine-tuning to a target voice, a workflow that generally beats single-speaker training from scratch.