TigreGotico/

tts-train-synthetic-miro_nl-NL

This is a single-speaker synthetic speech dataset for Dutch. It carries the Miro voice, a male voice. The audio was synthesized with text-to-speech and adapted to the Miro speaker identity, to train a Miro voice model for Dutch. The dataset contains 1150 recordings, with a metadata.csv transcript file and one WAV file per line.

Models trained on this dataset:

OpenVoiceOS/pipertts_nl-NL_miro

Dataset collection:… See the full description on the dataset page: https://huggingface.co/datasets/TigreGotico/tts-train-synthetic-miro_nl-NL.