TrySalient/tts-2hr

Bianca Fisher 2小时结构化文本转语音训练数据集，包含专业债务催收员Bianca Fisher的录音，总时长约1.92小时，分为868个高质量音频片段。数据集内容经过专业设计，包含问候/身份确认、身份验证、账户信息、处理/后续步骤和边缘情况/注入等五个类别，旨在训练具有同理心、冷静和积极语音合成的模型。

The Bianca Fisher 2-Hour Structured TTS Dataset is a professionally curated training dataset featuring recordings from Bianca Fisher, a top-performing debt collection agent. With a total duration of approximately 1.92 hours, divided into 868 high-quality audio clips, this dataset is structured with professional content composition designed to train empathetic, calm, and positive voice synthesis models across five categories: Greeting/Identity, Authentication, Account Information, Disposition/Next-steps, and Edge Cases/Injections.