I used a dataset of 1 minute and 51 seconds...I think it didn't come out very good...in future i'm going to do it with more minutes and at least 250-300 Epochs. (I tested it with two different Speaking, one is in Italian and another one is in English, maybe I had to test it with more or less pitch and with a song) Credit: <@294135272097185792> If you want to use it, you have to give me credit, thanks! 🙂 Link: <https://huggingface.co/FanFierik/GokuOcean/resolve/main/Goku.zip>