This voicebank was configured (labeled & trained) by PixPrucer! I highly recommend Pix if you're looking to get a voicebank made!
🔗 bsky.app/profile/pixprucer.b…
Makoto's 2.0 DiffSinger model is trained solely off of vocal data from myself, Makoto's voice provider.
honestly the one time i tuned diffsinger on a song i failed miserably because it was some last minute tuning for vocagathering >.< i do intend on learning to tune utaus and by extent diffs when i make my official debut as a vocap