2 ms·

I was looking at video on training a custom voice with Piper, following a tutorial at https://www.youtube.com/watch?v=b_we_jma220 https://www.youtube.com/watch?

I was looking at video on training a custom voice with Piper, following a tutorial at https://www.youtube.com/watch?v=b_we_jma220 https://www.youtube.com/watch?v=b_we_jma220, and noticed how the datasets required metadata of the text for the source audio files. This training method by Collabora seems to automate that process and only requires an audio file for training.

jpcl 3y ago
Yup, we are using Whisper to transcribe automatically so we can train the model on just speech recordings, without human transcripts. This works for any language that is well supported by the OpenAI Whisper model.
- deskamess 3y ago
  Where can we find the latest OpenAI language model rankings?
  - jpcl 3y ago
    There is a plot of language performance on their repo: https://github.com/openai/whisper https://github.com/openai/whisper I am not aware of a multi-lingual leaderboard for speech recognition models.
gmerc 3y ago
Whisper solves it, that’s its purpose.