Buck26/

tts-french-dataset

This dataset contains French speech audio paired with clean transcriptions, intended for training text-to-speech models such as Spark-TTS or Coqui TTS.

dataset.parquet — metadata file with audio paths, transcriptions, speaker info Audio/ — directory of all .wav files used for training

The dataset.parquet file includes the following columns:

Column Description

audio Path to .wav file

text… See the full description on the dataset page: https://huggingface.co/datasets/Buck26/tts-french-dataset.