Description
I know there's already like 2 Kleiner models on the server but I found out there's like 10 minutes of early voicelines in the HL2 beta lol. So the dataset is 27 minutes in total. Yeah I know, only 260 epochs. The training fucked itself and there was no way I was starting over again. I'll post some more demos later. But enjoy.
Comments
mares_
You say the training fucked itself, but it honestly sounds really good
mares_
Scuffed singing, but good at talking

resizable
i mean it tried

resizable
If you good n long sounding dataset, you can have a model that sounds good
Big Mitch
By "fucked itself" i met colab just stopped saving the weight backups to google drive. So I couldn't continue training
Add a comment
Samples
New
Classic
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
More to explore
Saiba Momoi (Blue Archive)

LordDavis778
452k
Ariana Grande AI

moonlight_18
128k
JENNIE of BLACKPINK [Strong Ver.]

leelo li
84k
Saiba Momoi (Blue Archive) (VA: Tokui Sora)

ryzusaku
76k
Hatsune Miku
293k
SpongeBob SquarePants (Talking And Singing)
Adrian Ramsey
188k
Takanashi Hoshino (from Blue Archive)

LordDavis778
150k
Satoru Gojo (JJK) [VA Yuichi Nakamura]

_phant0m
116k
ENHYPEN Heeseung

moonlight_18
100k
Sunaokami Shiroko (Blue Archive)

LordDavis778
82k
Villager (Minecraft)

AI Bakery
252k
Mortis [Brawl stars]
fergg209
155k
Jungkook (BTS)

mentosandrice
143k
Tendou Arisu (Blue Archive)

LordDavis778
94k
Kanye West
80k
Loading more