Datasets:
How to prepare a custom dataset from .mp3 or .wav audio files and their corresponding transcription in parquet format for training?
#2
by arshneo - opened
I am new in the field of voice cloning. I have several Hindi audio files in .mp3 and .wav formats which runs into hundreds of hours along with their corresponding transcription. The goal is to clone the voice of the speaker. Where can i find python script to do the job please?
It's depend on your tts also but in general as much speaker is there you have to give them unique identity so as ref_audio during training u can take other audio from same speaker.I hope you will understand when u will go in deep Dear.