WebKit

 Introducing our latest RVC Model, "Beavis (RVC v2)" - a state-of-the-art AI Voice Model specifically designed using Retrieval-Based Voice Conversion technology. With over 1,00

This model was trained on close to 20 minutes worth of audio from reaction clips from the Beavis & Butt-head reboot, as well as some classic Beavis voice clips from the older show. This model should be decently versatile allow when he sustains notes for a long time, it can sound a bit robotic, since he doesn't really sing on the show. This was trained using the mangio-crepe method, 64 HOP. Alright, well that's all I have to say, have fun with the model!  CHECK OUT THE DEMOS BELOW TO SEE HOW THE MODEL PERFORMS!

Beavis

Beavis (RVC v2) 1K Epochs 45K Steps

ell0er

Here's the training total loss graph, it overtrained slightly, but I tested the model every 250 epochs and they sound around the same in terms of fidelity.

🎙️ Original Vocal Impression from me with the AI Model 🎙️

💀 Beavis Raps "everybody dies" By J Cole 💀

🎙️ Original Vocal Sample Using Both of my Ai Voices (Butt-Head will be retrained soon!) 🎙️

Very epic video "MrSneakyDick666" on YT made using my models: https://youtu.be/CZbr4Xt9VnI

insect14

Joe cool

Shuaiby Aslam

This is an updated model based on the French Narrator from SpongeBob that I originally trained on v1 crepe on a smaller dataset. This one is trained on RVC v2 using the mangio-crepe method. This was trained locally and is my first model trained to 2K Epochs. This was trained on around 20 minutes of audio from the show and video games. This is ideally good for a regular talking voice, and maybe even Realtime, Sounds pretty accurate. Works great making impressions sound more authentic! Google Drive Face Link (Mirror):  CHECK OUT SAMPLES BELOW FOR DEMOS OF THE MODEL!

French Narrator (From SpongeBob)

French Narrator (From SpongeBob) (RVC v2) 500 Epochs

Hey, I'm finally back to making some public models after a pretty long hiatus. This model came out okay in comparision to my other models, but not the best because of it mostly being audio rips from the show, rather than Video Game voice clips. Should be decently versatile but you may have to adjust the feature index ratio to a lower value reduce background noise or pitch to improve similarity. Alright, Have Fun! Oh, also I have two differently trained versions of this model, one on Mang-Crepe and one on Harvest. Mangio-Crepe Version: Drive Face Version: Drive Face  CHECK OUT THE DEMOS BELOW FOR SAMPLES OF THE MODEL

Flying Dutchman (From SpongeBob)

Flying Dutchman (From SpongeBob) RVCv2 500 Epochs

This model was trained on mangio-crepe on a long varied dataset, which included some of dunkeys joke rap song songs he made a while back, as well as regular vocals from his actual videos. This model is a bit iffy on 0 transpose, so you may have to fiddle with the transposition to get ideal results with this model. Alright Have Fun! CHECK OUT DEMOS BELOW FOR SAMPLES OF THE MODEL!

videogamedunkey (From YouTube)

videogamedunkey (From YouTube) (RVCv2) (500 Epochs)

Trained on nearly an hour of Rick Hunter's voice overs for postal dude from POSTAL 2, Paradise Lost, & Postal 4. I forgot to add the voice lines from apocalypse weekend, but it shouldn't matter too much. This model is mostly for making memes and regular vocal inferences. It can rap, but can't sing very well. It also may require you to transpose audio to a lower or higher pitch depending on the audio you are using for inferencing. Alright, have fun with the model boyz! CHECK OUT DEMOS BELOW FOR SAMPLES OF THE MODEL!!

POSTAL Dude (From The Postal Series)

POSTAL Dude (From The Postal Series) (RVC v2) 600 Epochs

This was trained on over an hour of voice lines from the OG Deus Ex Video Game. This model retains some of the tongue clicking you would expect from JC, but it is very random when it occurs. This model should pretty much be perfectly trained, stopped right when it started to increase total loss. The model favors a very low semitones so you may have to mess with the semitones. This model isn't best for singing but is good for regular vocals and maybe rapping. Alright Have fun!  CHECK OUT THE DEMOS BELOW TO SEE EXAMPLES OF THE MODEL!

Beavis

Description

Samples

Saiba Momoi (Blue Archive)

Adel imam (Egyptian Actor) عادل امام (ممثل مصري)

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

Ariana Grande AI

Hu Tao JP (Genshin Impact)

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)

Goku (Mario Castañeda) (Latin American Spanish) (UPDATE)

Villager (Minecraft)

Mortis [Brawl stars]

Jungkook (BTS)

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

Tendou Arisu (Blue Archive)

Kanye West