This repository contains a speech dataset developed for Text-to-Speech (TTS) research and model training.
The dataset is part of an ongoing research project focused on building high-quality speech corpora for modern neural TTS systems. It is actively maintained, with continuous improvements in data quality, transcription consistency, metadata, and organization.
This repository is intended to host research data used throughout the development process and is not intended… See the full description on the dataset page:
https://huggingface.co/datasets/ShiniChien/TTSDistil-Phonology.
