Description

if you use this model for any projects i would prefer you credit the original VA (Cary Huang) instead of me. this one tends to be iffy but produces mostly passable results despite the fact he only has 1:10 of data from across all speaking roles ever. Grassy speaks in a very specific way so if you just talk like a human it tends not to sound like him. for speaking roles aim to mimick his way of speech model: