Introducing the Master Roshi DB/DBZ/DBS RVC v2 (Crepe) model, now with even more powerful AI capabilities! This version has been trained for 300 epochs, resulting in a refined and enhanced ability to generate authentic and exciting 'BOOM BOOM HEHEHEHE BOOOM BOOOOOM' style AI music. With this model, you can easily create incredible AI covers that will leave your audience in awe. Plus, it's perfect for text-to-speech applications, giving life to any project with its unique sound. Try out our free AI tools and experience the power of the RVC v2 model for yourself.

Master Roshi (DB/DBZ/DBS)

Master Roshi (DB/DBZ/DBS) (RVC v2) (Crepe) - 300 Epochs

SyberG

Sample: Jason Derulo - Wiggle (this is actually good wtf)

100 Epochs > Trained with 30 minutes and 5 seconds of dataset > Pretrained with TITAN > Normalized + Echo and Noise Removed > Voiced by Jeannie Tirado > 8.3k Steps > RVMPE/64 Hop-Length > 40k Target Sample > Batch Size 8

Android 21 (Evil) [DB-FighterZ | 🇺🇸] [TITAN Retrained]

Android 21 (Evil) [DB-FighterZ | 🇺🇸] [TITAN Retrained] (RVC v2 | 100 Epochs)

100 Epochs > Trained with 43 minutes and 25 seconds of dataset > Pretrained with TITAN > Normalized + Echo Removed > Voiced by James Marsters > 11.5 Steps > RVMPE/64 Hop-Length > 40k Target Sample

Fused Zamasu (Echo Removed) [DB-FighterZ | 🇺🇸]

Fused Zamasu v2 (Echo Removed) [DB-FighterZ | 🇺🇸]

150 Epochs > Trained with 8 minutes and 1 seconds of dataset > Pretrained with TITAN > Normalized + Echo and Noise Removed > Voiced by (unknown) > 4 350 Steps > RVMPE/64 Hop-Length > 40k Target Sample > Batch Size 8 > He's English so I'll put the English tag

Narrator [DB-FighterZ | 🇺🇸]

Narrator [DB-FighterZ | 🇺🇸] (RVC v2 | 150 Epochs)

450 Epochs > Trained with 1 minute and 36 seconds of dataset > Pretrained with TITAN > Normalized > Voiced by (unknown) > 2.7k Steps > RVMPE/64 Hop-Length > 40k Target Sample > Batch Size 8

Male 15 [DB-Xenoverse 2 | 🇺🇸]

450 Epochs > Trained with 1 minute and 43 seconds of dataset > Pretrained with TITAN > Normalized > Voiced by (unknown) > 2.7k Steps > RVMPE/64 Hop-Length > 40k Target Sample > Batch Size 8

Male 14 [DB-Xenoverse 2 | 🇺🇸]

Reinstated!!!!!!! Sorry for originally deleting it!!! (Originally Created on 05/14/23) English VA; Justin Briner 33 Min Dataset Used voicelines from S1-6 and My Hero's Ones Justice (Provided by REU Music) Huggingface link (Alt)

Izuku Midoriya (Deku) (From My Hero Academia)

Izuku Midoriya (Deku) (From My Hero Academia) (RVC) 300 Epochs

with 22 minutes and 48 seconds of dataset Voiced by Christopher Sabat 18.9k Steps RVMPE/64 Hop-Length 40k Target Sample Batch Size 7 Yes I know Vegeta was already made a few times nobody made a Ov2Super version of him so I decided to make one myself based of his voice lines in FighterZ

Vegeta [DB-FighterZ | 🇺🇸]

Vegeta [DB-FighterZ | 🇺🇸] (RVC v2 | Ov2Super | 300 Epochs)

300 Epochs. Voice clips from Team Sonic Racing and Sonic Twitter Takeover 5 & 7 used to make up this model. If you can it's recommended that it's used with the AI Cover Gen WebUI for male/female pitch adjustment depending on how high pitched or low pitched a song is.

Sonic The Hedgehog [Roger Craig Smith]

A Mii in a Fedora

with 25 minutes and 1 second of dataset Voiced by Sonny Strait 22.2k Steps (Totally not steps during his training) RVMPE/64 Hop-Length 40k Target Sample Batch Size 7 Would consider this as v2 because he sounds way better (because of Ov2Super) and less bass boosted than the previous model I've made of him.

Krillin [DB-FighterZ | 🇺🇸]

Krillin v2 [DB-FighterZ | 🇺🇸] (RVC v2 | Ov2Super | 300 Epochs)

Katuski Bakugo (Kacchan) (From My Hero Academia)

Katuski Bakugo (Kacchan) (From My Hero Academia) (RVC) 350 Epochs

This is my first model trained using the RVC v2 model arch. Should be fairly versatile. Took around 9 hours to train. Trained on a dataset of around a little less than 30 minutes of video game voice lines from Plankton's character. Hugging Face Link (Mirror): Google Drive OUT THE DEMOS BELOW!

Plankton (SpongeBob)

Plankton (SpongeBob) (RVC v2) 1K Epochs 69K Steps

ell0er

with 35 minutes and 9 seconds of dataset of all of his voice lines (all transformations combined) Pretrained with TITAN Echo Removed Voiced by Sean Schemmel 8.8k Steps RVMPE/64 Hop-Length 40k Target Sample Batch Size 8 He sounds more serious compared to the first & old Goku model way back in the OG AI Hub server

Son Goku [DB-FighterZ | 🇺🇸] [All Transformations Combined]

Goku (Re-Upload)

Goku (RVC) (250 Epochs) (Crepe) (Re-Upload)

xxx_spxrks

Trained by me (Tost) Dataset based entirely off of Dragonball FighterZ Link to model:

Master Roshi (DB/DBZ/DBS)

Description

Add a comment

Izuku Midoriya (Deku) (From My Hero Academia)

Vegeta [DB-FighterZ | 🇺🇸]

Sonic The Hedgehog [Roger Craig Smith]

Krillin [DB-FighterZ | 🇺🇸]

Katuski Bakugo (Kacchan) (From My Hero Academia)

Plankton (SpongeBob)

Son Goku [DB-FighterZ | 🇺🇸] [All Transformations Combined]

Goku (Re-Upload)

Frieza -ENG Voice- (From Dragonball FighterZ)

Saiba Momoi (Blue Archive)

Villager (Minecraft)

Sas (Counter-strike global offensive)

FIXED HATSUNE MIKU V3 - VOCALOID [SCROLL DOWN FOR LINK]

Fang [Brawl Stars]

Ariana Grande AI

Satoru Gojo (JJK) [VA Yuichi Nakamura]

Mortis [Brawl stars]

Jungkook (BTS)

ENHYPEN Heeseung

Sonic The Hedgehog [Roger Craig Smith]

CaseOh

Takanashi Hoshino (from Blue Archive)

Surge (brawl Stars)

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

SpongeBob SquarePants (Talking And Singing)

Melodie - (Brawl Stars) - (English & Korean Voices)

Miss Circle (Fundamental Paper Education)

Jay (ENHYPEN)

JP Furina / Focalors CV: Inori Minase (Genshin Impact)