Description
**Huggingface:** Trained on a dataset featuring a mixture of vocals from Toxicity and Hypnotize/Mesmerize.
Comments

Considering it's a mix of the Toxicity and the Mez/Hyp era, this sounds like Serj from shows in 2003-04

That was the intention! Mid-SOAD Serj. Kind of a mix of good elements of both.

I love Serj's operatic kind of elements in his voice but I also love his screams so a mix of those both is perfect Serj for me.

I had to share this because it's fucking beautiful 😭

I love it how with the high notes it screams.

cool af!

I know im pretty late here, lol. but do you know how much "Save Frequency" you usually put it to? I think it's like 1 - 50 right?

I put it to 25 usually

Just so it regularly saves but not too often.

I see, because I thought like the higher the number, the best quality i guess,

I was thinking of doing a Self Titled version of Serj

but thanks

That's Epochs you're thinking of. Generally yeah the more epochs you have, the more the AI has learned but you have to watch out for overtraining using tensorboard.

I was thinking about around maybe 300 - 450, is that fine?

Try that out yeh

Serj sing Unfaithful now is my fav song
Add a comment
Samples
Pitch
More to explore
Saiba Momoi (Blue Archive)

Ariana Grande AI

JENNIE of BLACKPINK [Strong Ver.]

Saiba Momoi (Blue Archive) (VA: Tokui Sora)

Hatsune Miku
SpongeBob SquarePants (Talking And Singing)
Takanashi Hoshino (from Blue Archive)

Satoru Gojo (JJK) [VA Yuichi Nakamura]

ENHYPEN Heeseung

Sunaokami Shiroko (Blue Archive)

Villager (Minecraft)

Mortis [Brawl stars]
Jungkook (BTS)

Tendou Arisu (Blue Archive)

Kanye West
Loading more