Intro guide to creating a model from recordings?

[Feanor78]

Hello,

I've been messing around with CoquiTTS and trying to figure out how to get it to train a TTS model using .wav files of a speaker and a metadata.csv file that transcribes the .wav files and refers to their location in the format <file.wav>|The words in the wav file.

I've continued to run into errors when trying to get the trainer running, mostly regarding ModuleNotFound errors when I run the script that's supposed to get the trainer working. Can anyone point me to a guide to writing a script and config.json file that will work properly to train a voice model?

Thanks so much!