Nope...

The model sounds like the dataset wasn't cleaned up and currently sounds like a cry for help.

imo it isn't worth it just using brawl voice lines, just a general jason griffith model would suffice

Even if you wanted to, they need to be cleaned up to remove any noise and trained to a good epoch.

Both this model lack.

In other news, this model is sponsored by my own Jason Sonic (https://discord.com/channels/1089076875999072296/1121201406053982269) and Werehog models (https://discord.com/channels/1089076875999072296/1118046396176662578) <:joedevious8:1093355644075131041>

surely this needs improvement. It can be hard to extract vocals from mixed audio files