TAGARELA dataset pt-PT

#2
by rjsimoes - opened

Hi!
First of all, thank you for putting together the TAGARELA dataset.
I'm working on fine-tuning a TTS model for European Portuguese (PT-PT) voice cloning. While the dataset is clearly a great resource for Portuguese in general, I'm wondering about the PT-PT subset.
Given that the dataset contains ~842 hours of PT-PT audio (9% of the total), would it be possible to make that subset available separately, or provide a metadata filter/file list to identify PT-PT speakers?
Thank you!

Hello, thank you very much for your interest in our dataset!

We are currently working on adding metadata such as language, gender, and audio quality to this repository. We plan to make this information available as soon as possible, so please stay tuned for upcoming updates.

@rjsimoes a column containing the accent was included.

freds0 changed discussion status to closed

Sign up or log in to comment