Support for creating dataset text files from English and Japanese audio too.

[nekogecko2]

Sorry if it's already on your todo list. The ASR tool can read english audio ok but it's not as good as whisper. It would be convenient if whisper or something was added to it so everything could be done in the same program. I use another program for creating dataset txt files right now.

Also for any English users reading this,

This tool is actually LEGIT. It only takes a few minutes to train a model on my RTX 3060. Like this is finally a REALLY real for real open source alternative to eleven labs.

Here's a example with Finn's voice from adventure time

https://vocaroo.com/133yE3LKENkT https://vocaroo.com/1aYbqO6yxatg