Description
Trained By with the usage of 40K rates, with up to 6 minutes of the total dataset. The AI Model is mostly based off from the Music To Be Murdered By album's hit single "Godzilla", "Unaccommodating", and "I Will" from the 6 minute dataset. It pretty much doesn't have the aggressive vocals that I wanted it to be, almost like a higher pitch singing-like tone, but if maybe the Crepe setting will help, just try your luck.
Comments

Added the Original made Index file from the training sessions from the link below https://huggingface.co/GoDZ21/Eminem_MTBMBEra_RVCv2

was literally listening to mtbmb when this model just dropped, he has some higher singing voice on "farewell", is this only side a?

This is all from the Side A era, I was thinking about mixing the two, but Side B was recorded separately, so maybe I can make a separate model for only the Side B Era, It makes more sense to me 🤓

yeah, he has a very monotone voice on side b.

Maybe or maybe not I will do a Side B Version, but There is still Kamikaze and Revival eras that has not been done yet lol <:drakelaugh:1093604722377769140>

yeah, i was gonna do revival but played too much minecraft and forgot about it.

might create one later.
Add a comment
Samples
1. Singing
Male
English
2. Singing
Female
English
3. Singing (Dry)
Female
English
4. Singing (High)
Female
English
5. Singing 2
Male
English
6. Singing (Dry)
Male
English
7. Singing (Dry, High)
Male
English
Pitch
Users also tried
More to explore
Loading more