✨│ai-help

[vapid gust]

my laptop is so shit it can’t even do it

[proven hill]

saw the settings

most probably its the model

[brazen plaza]

oh ok

[pallid thicket]

Ok snoring shorty

[sand iris]

shots fired

[half cove]

[green prairie]

is there a tool where I can modify an instrumental to be like a guitar, or piano, instead of the original?

[gilded narwhal]

hey why do i get error for converting

[violet heron]

Using?

[gilded narwhal]

rvc v2 huggingface verison

[violet heron]

Did you duplicate the space

[gilded narwhal]

no

[violet heron]

(Also I would recommend to use ilaira RVC instead since it’s faster)

[gilded narwhal]

alr

[rare goblet]

Ayo? @gilded narwhal level 1 !!!

[violet heron]

https://huggingface.co/spaces/TheStinger/Ilaria_RVC

Duplicate the space by clicking the 3 dots on the top row and click “duplicate this space”

[ivory plank]

does it matter if the vocal of the song I want to use has reverb or not

[gilded narwhal]

yeahhh about that

huggingface isnt letting me create an account

[violet heron]

Why?

[gilded narwhal]

idk but thanks for help

[brazen plaza]

@violet heron i tried a bunch of new models and its still choppy in games

[rare goblet]

Ayo? @brazen plaza level 2 !!!

[vale horizon]

I've been away for a couple of months and I have a $1,000,000 Question: What is the current situation regarding okada real-time voice changer? Has latency improved? How many milliseconds late was he?

[red wraith]

anuel aa

[spare remnant]

i don't speak any of these languages so it sounds really weird, any ide how to fix?

[rare goblet]

Ayo? @spare remnant level 3 !!!

[sage spade]

how can i use a model to make my ai cover?

[molten pecan]

You'll first need this AI tool called RVC

What's your GPU?

[molten pecan]

Send my guide next time

https://rentry.org/ilarvc_inf_guide

I included everything there

You don't have to tell them about duplicating the space

[hybrid zenith]

is there an offline RVC that doesnt use google collab? I have an rtx 2060

[violet heron]

-local

[azure marsh]

Suggestions for @hybrid zenith

All Working Local Forks 🖥️

• Applio-RVC-Fork, by IA Hispano GitHub
• Mangio-RVC-Fork, by Mangio621 Huggingface
• RVC Studio, by SayanoAI Huggingface
• AICoverGen, by SociallyIneptWeeb GitHub
• Replay, by Replay Team Website
• Original RVC, by the RVC-Project team GitHub

Need some help? 🤔

You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.

More commands: -audio, -colabs, -overtrain, -realtime, -rvc, -help

[molten pecan]

@violet heron i cant rea d yy mbad

[sage spade]

I think I don't have a GPU. An year ago I made some ai cover with SVGC with anaconda but now I canno do my model with google colab so I want to use an ai model I found here and I don't know how to use it.

[molten pecan]

Alr. Do you have the model's hugging face link?

[sage spade]

no I downloaded the file .pth

[molten pecan]

Did the model include a .index file?

[vapid gust]

whats the best uvr model

[molten pecan]

For what exactly?

[sage spade]

yes .index too. Can I use mangio?

[vapid gust]

extraction?

vocal extraction

[molten pecan]

MDX23C

Process Method: MDX-net

[half cove]

wong pinf moment

[molten pecan]

Huh

[vapid gust]

wow it takes alot of time

[half cove]

yea

[vapid gust]

i dont even have a good laptop too

fk

[molten pecan]

Yes you can use it but preferably via the cloud, not locally

[half cove]

its gunna take a while then

you still can

do it

[vapid gust]

can u do it for me

[sage spade]

is there a tutorial?

[rare goblet]

Ayo? @sage spade level 1 !!!

[molten pecan]

Use mvsep.com

[vapid gust]

is it tthe same thing as uvr?

[molten pecan]

https://rentry.org/ilarvc_inf_guide

(ilaria rvc is built using Mangio)

[molten pecan]

Yeah but via the cloud

[half cove]

there is a colab for it btw

[molten pecan]

@half cove I can't wait for my website to be done so whenever someone asks a question here you'll just link them the website

[half cove]

[sage spade]

thanks

[half cove]

https://colab.research.google.com/github/jarredou/MVSEP-MDX23-Colab_v2/blob/v2.3/MVSep-MDX23-Colab.ipynb

[vapid gust]

thank you so much

so far now it’s gained 1% and now it is at 11%

[half cove]

using mvsep or colab

[vapid gust]

using uvr

local

now im switching to colab

do i need to change any of settings here?

[half cove]

the colab should explain the settings

above it

[vapid gust]

oh ok

but is it needed?

to change settings

[half cove]

dont think so

[vapid gust]

why does installation take so long

[half cove]

its installing the model from scratch

hmm

[vapid gust]

wot

do i do

[half cove]

let me try install it rq

did you allow google drive?

@vapid gust

[inland swift]

Yo

I need help with colab

So my problem is my gpu died

It failed to train and i waited for 6 hours for it

My epochs was at 563

And i needed 750

Can i somehow preload it somehow?

My runtime gpu doesn't works

Like is gpu necessary to use to make ai voice?

[proper shale]

Yes

[proper shale]

Oh you shouldn't have waited that long

[inland swift]

I mean

[proper shale]

U can stop training after 3 hours and it'll be fine

[inland swift]

It trains 1 epoch long enough

[inland swift]

Doesn't it clears the old epochs?

[proper shale]

should've done that and exported the model

[inland swift]

Like it restarts

[proper shale]

No

[inland swift]

Or what if i restared website

[proper shale]

It does not

Unless

You changed settings

[fluid matrix]

have you checked tensorboard? do you even need 750 epochs?

[inland swift]

I already did it💀

[inland swift]

Tensor doesn't even works for me

[fluid matrix]

damn

[inland swift]

The only thing i can see in tensor is inactive

[proper shale]

I think you may have lost your model

[inland swift]

It's very important for me

[proper shale]

If you didn't stop and export the model

[inland swift]

I have the rvclog thing

[fluid matrix]

havent used colab in a while but that might be true

[inland swift]

NOOOOOOOOOOOOOOOOOOOO

I didn't export

[proper shale]

Yeah but do you have D/G

[fluid matrix]

check your drives folders

[proper shale]

Exactly

[inland swift]

I don't have d-g

[proper shale]

That's why I always say, stop model after like 1-2 hours of training, export

[inland swift]

Also

My rvclog weights 312 mb

And my zip file almost 100 mb

[proper shale]

That's just preprocessing and feature extraction

[inland swift]

Alright

So...

I need to redo all that stuff again?

[fluid matrix]

looks like it

[proper shale]

...no

[fluid matrix]

not everything tho

[inland swift]

Should i reload website?

[proper shale]

You just load back all the preprocessing and stuff files

[fluid matrix]

no

yeah that

[proper shale]

You have to wait until tomorrow

[inland swift]

Why

[fluid matrix]

yeah why 🗿

[inland swift]

And how exactly

[proper shale]

That's just how Colab GPU limits work

[proper shale]

first (for resuming) step

[inland swift]

Oh ok i got it

But i just can use another account

[fluid matrix]

what if he transfers all his files to another google account and tries again

sorry last time i used colab was in july

[inland swift]

IT CAN WORK???

[proper shale]

Yeah but you'd need to copy over all files

[fluid matrix]

yeah thats what im saying

[inland swift]

I can copy

[fluid matrix]

that should work

[inland swift]

But what do i need exactly to copy

[proper shale]

You lost the current progress of the model tho

[inland swift]

Let me screenshot

[proper shale]

so you'll start again from 0

[proper shale]

Just copy over the rvcDisconnected folder

[proper shale]

Should be fine

[inland swift]

I waited each hour for every 100 epochs...

How do i copy?
Like i need to download or what

[proper shale]

Sorry about that but I told you about this before

[proper shale]

... good question 😭

I never did that tbh so

[inland swift]

I actually don't really use google drive

So idk if ctrl + c works

Like if i gonna switch account after

[fluid matrix]

pretty sure you have to download everything, log in into another account, grant access again so it creates the folders and then upload the files in drive under the same folder structure

sharing the drive wont work

[inland swift]

Alright I'll try

I did download the folder

Now i will gonna switch account

So hold up

If i do train

I can stop it right?

And nothing gonna be lost

Like epochs do save

It means i can export the thing and after load it and start training again right

For example it gonna start from 200-250 instead of 0

[vapid gust]

mb

ye i did

[proper shale]

If you export model, yes

[proper shale]

Yes

[proper shale]

Yes

[half cove]

mine works fine

[inland swift]

Time to redo all stuff again

I forgot 💀

Gpu is also doesn't works on alt account

[sharp shore]

Hey, which software do you guys use with the voice models?

[vapid gust]

i just restarted it

how long did it take for u

?

[half cove]

took me

like 2min

[vapid gust]

oh it works for me now

ty

nvm

@half cove

[half cove]

huh

just send ur audio

ill do it

[vapid gust]

ok

sent in dms

[half cove]

i dont get how you got this error 😭 mdx23c doesnt use demucs

[vapid gust]

bro idk

are u doing it?

[half cove]

look at dms brother

[vapid gust]

ty

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[inland swift]

So ok.
When i will gonna save the d-g
How i can load them after?

I never did that before

So if i will gonna know then next time when i will gonna do it i will instantly gonna know what to do

I don't need to do it rn just asking

[proper shale]

Export model step, and to load them use the (for resuming) steps

[inland swift]

This is bit confusing but I'll try to realize and process all needed stuff

[lunar prism]

hi guys.. this is a long shot. wondering if anyone ever used that "easy 10/25" colab and if it was possible to continue/resume a previous model training ?

[molten pecan]

Idk if you can in easy 10/25

But I know you can in RVC Disconnected

[lunar prism]

thank you for your answer ! i will look into it

[molten pecan]

yw! Good luck

[hot lagoon]

-paperspace

[azure marsh]

RVC v2 Tutorial for Paperspace

💡 Click on the button below to open the guide.

Credits: LollenApe

[outer junco]

Kinda new to this stuff, got some models I wanna use but No program simple or reliable enough currently

[lunar prism]

for the record, I used the FR version of Easy 10/25 and the translator added a resume training section at the end ! it'S called "reprendre l'entrainement du model".. you input the model_name, dataset_folder, and your ngrok token and run only that cell. it works

[rare goblet]

Ayo? @lunar prism level 1 !!!

[molten pecan]

Awesome!

[molten pecan]

There are, actually. Let's start from here: do you own a PC with great specs?

Or just an average one?

[outer junco]

GPU is a 3060 currently

CPU is a ryzen 5 3600

[molten pecan]

Awesome. I recommend Mangio for you

Do you know what that is? Or how to install it?

[small umbra]

..how to install dez nuts in yo mouf?*

[molten pecan]

😦

[outer junco]

Never heard of, only really tried W-okada

[molten pecan]

Ohhh I see. Are you looking for changing your voice in real-time as you speak?
Or do you want to change the voice through audio files? (Like AI covers)

[outer junco]

Real time, and just a tts one jsut to see how some rvc sound before using them

[rare goblet]

Ayo? @outer junco level 1 !!!

[molten pecan]

I see. Well, with TTS the voice will sound plain with no emotion, that's how TTS is general.
And how RVC sounds depends on a lot of things. The output quality doesn't really depend on the software itself, but the quality of your sample, the pitch extraction algorithm, quality of the model, f0, and other stuff

-realtime

[azure marsh]

This interaction has expired, use the command /guides realtime if you wish to see it again.

[molten pecan]

There you have some guides for the real-time voice changer

[outer junco]

Thank you

[molten pecan]

Have fun!

[molten pecan]

yw!

[outer junco]

Mangio is a real time one right?

[molten pecan]

Mangio is for converting vocals through audio files.
W-Okada is for converting vocals in realtime through the mic

[outer junco]

Fair enough, I'll take a look at both

[molten pecan]

Alr have fun y'all!

[outer junco]

Will do, once I've got it figured out haha

[molten pecan]

Hehe I get it. The information is all over the place, it'll take time to fully learn. I was there before, so my tip is to learn through trial and error. Experiment with all the buttons and see what sounds best

And if you screw it up you can always reinstall it :3

[outer junco]

literally me: I wOnDeR wHaT tHiS bUtToN dOeS

[violet heron]

It explodes your computer

[molten pecan]

I once told Peter to do a cover for me and his laptop bluescreened

That happened like 3 times 😭

Speaking of Peter

He was online at one point in the past 7 days

According to Telegram

No idea what's going with him 😭

[violet heron]

But is he still alive

[molten pecan]

Yeah prob

[devout forge]

when i open start_http.bat the thing pops up but its blank

[inland swift]

Hello

[rare goblet]

Ayo? @inland swift level 9 !!!

[inland swift]

What i should do if my tensor board doesn't work

It doesn't show statistics etc

I wanna know how much epochs do i need for my voice

[civic latch]

Hey guys i just got an error like this “connect http://127.0.0.1:18888” i tried to find the file named stored_settings but seems like there’s no such file in the folder, any1 have the solution?

[inland swift]

It started to work but idk how to check how much epochs do i need

[inland swift]

Do you have spaces in your folder?

[civic latch]

Wdym spaces?

[rare goblet]

Ayo? @civic latch level 1 !!!

[inland swift]

The spaces in name

[civic latch]

Wasnt it stored_settings.jsl?

[inland swift]

What

[civic latch]

Ugh dunno how to say it but ppl on github got the same problems and they said i should delete the file named stored_settings in that folder to solve that problem

[inland swift]

I don't really understand those stuff so I'll just mind my own stuff

[civic latch]

It keeps connecting to the url and comes with a blank screen

[inland swift]

I'm new to rvc stuff

[civic latch]

Ah

[inland swift]

And now i have another problem

When i stop training and exporting the model from notebook to drive it doesn't generates the d-g files

[civic latch]

I think we have the same problem

[devout forge]

yeah same

it worked once then just stopped working

[devout forge]

i deleted the file "stored_settings" and it worked

i think that fixes it

atleast for me it did

[civic latch]

Glad it did, but i cant find that file on my part

[thin stump]

Is there a UVR setting to remove breathing? I hate how the singer mixed it in when they're already speaking

[inland swift]

I think I'll use that someday

[inland swift]

And do i need to off there smth?

[half cove]

D and g are for training

[inland swift]

I know

But

How do i use them

Like I don't wanna lose my epochs when i get the gpu error

(it looks dumb rn)

[half cove]

Try not to do too much epochs at a time also i think there is a cell for savings it to drive

[inland swift]

D and g files doesn't generate when i stop training by myself

[inland swift]

I did 750 epochs

[inland swift]

When i do export thingy it only appear the config and event stuff

But when it finishes training automatically and when i do export then i get d-g files

How does it actually work

So my problem is i'm trying to get d-g files only because 750 training goes so slow
I spent 6 hours for 600 epochs

[half cove]

I'd check but I'm on mobile rn

[inland swift]

But after 600 epochs i got gpu error

[inland swift]

And everything got erased

[inland swift]

Cuz i needed to switch account etc

[inland swift]

Do i need to change smth there

Like save_only_latest_ckpt

[wind saddle]

Anyone available to help me instal it, im running into trouble with it.

If you want more detail rvc.bat wont open correctly.

Is my pc just not good enough?

[proud elbow]

that RVC gui may be outdated and have issues

[wind saddle]

Do you have the newest link?

Or if not where do i find it.

[proud elbow]

would recommend https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/releases

[wind saddle]

Much appreciated

Now to wait another 30 minutes for an instal.

[inland swift]

How i can understand i can't stop the training of model too fast to get d-g files right

Like i need to train for 1-2 hour atleast?

[inland swift]

And i'm still waiting for the answer

[south gazelle]

how do i do this none of these new colabs make sense

[inland swift]

I got it

[inland swift]

Now i understand how it works

Offing save_only_latest_ckpt is not necessary

And i need to wait minimum 1 hour before stopping training

Because if i stop too early it doesn't makes pth file and d-g files

So it means
I can stop every hour and save it
Like when gpu error gonna came out i can wait and after start at 100-200 epochs

[onyx geyser]

alguien sabe español?

necesito ayuda

[inland swift]

No habla espanol

[onyx geyser]

😦

[inland swift]

I just know what this means

It means I don't talk spanish

[onyx geyser]

Might you help me?

[inland swift]

I'll try

Show your problem

And i can bet that you use google translator

And google translator can have inaccurate translations

Para ponértelo más fácil usaré un traductor.

[onyx geyser]

What is the best method to use AI today?

[inland swift]

Simplemente no hables español tú mismo, no lo entenderé.

[onyx geyser]

traductor

[inland swift]

Puedes ver las formas aquí. #1159513888199540817

Encuentra la pestaña create a model with rvc (disconnected)

[onyx geyser]

Yes I know, but I don't understand anything

[rare goblet]

Ayo? @onyx geyser level 1 !!!

[inland swift]

Diré de inmediato que sin saber inglés, normalmente no puedes hacer casi nada.

[onyx geyser]

yes I understand but none of it worked for me

[molten pecan]

#🌍│español

[inland swift]

Porque no lo hiciste bien

Wait a sec

Is using different languages here allowed?

To help people if they don't understand english

[onyx geyser]

Okay, could you tell me what method you recommend?

[inland swift]

Create a model with rvc disconnected (colab)

[onyx geyser]

Could you give me the link?

sorry for the inconvenience

[inland swift]

-colab

[azure marsh]

☁️ Google Colabs

• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab

🤗 Hugginface Spaces

• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HF, by r3gf Huggingface Spaces
• AICoverGen, by r3gf Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces
• Advanced RVC Inference, by neuclya Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[onyx geyser]

bro

[inland swift]

?

I'll just give straight link

[onyx geyser]

There are several links there, which one works best, because there are some that don't work.

[inland swift]

https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS#scrollTo=Yj-_npuA_HRB

[inland swift]

Click this one link

It gonna be better if you just gonna learn english language

Cuz of you not knowing english language it will make job alot harder

(saying it not in disrespectful way or any offensive way)

[onyx geyser]

thank you so much

[inland swift]

No problem

[onyx geyser]

yes I know

[inland swift]

Good luck with making a voice model,
I'll go and chill adios

[onyx geyser]

So this is the best option?

[inland swift]

In my opinion yes

[onyx geyser]

bye, thank you very much for the help

[onyx geyser]

perfecto

gracias hermano

perdón por las molestias

[rare goblet]

Ayo? @onyx geyser level 2 !!!

[onyx geyser]

besos

[onyx geyser]

sharap

[clear citrus]

can someone help me with my ai voice model, everytime i turn on my mic

and somebody speaks my own mic will speak too, its like copying what someone said

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[woven furnace]

@proven hill did you make ilaria rvc?

because i want to know what these mean

[proud elbow]

literally pitch extraction methods, rmvpe works best on almost all cases tho

[inland swift]

Guys where i can check how much epochs do i need for my model?

[narrow oriole]

There is any new easy converter like gradio?
The page that what i was using it's down ;(

[proud elbow]

for online converter, https://discord.com/channels/1159260121998827560/1167844377046040769
local ones with gradio also works, but not the colab ones

[narrow oriole]

oooh thanks!

[sick acorn]

on my f0 det i can only enable crepe_tiny and rmvpe_onnx

[river wraith]

what was the ai for music called

[neat breach]

uhh does anyone know how to fix it its always on boot now i didnt change anything
https://media.discordapp.net/attachments/1192771583899619378/1192771592242090045/image.png?ex=65aa4a1c&is=6597d51c&hm=12eca2730e02a701b97acc4b63a9ce91df8264d614b60b8bbeecebde0e192b84&=&format=webp&quality=lossless&width=850&height=304

[sand iris]

try unzipping somewhere on the C:/ drive

[neat breach]

nope still broken

its on my c drive

c:\users\owo\desktop

[sand iris]

okay, plan B, I've never seen this error in the past and I'm not sure why it can even happen

but try connecting to http://127.0.0.1:18888/ on a web browser while this is running

[pure scroll]

try deleting the “stored_settings.json” file in the “MMVCServerSIO” folder and restart the app.

[rare goblet]

Ayo? @pure scroll level 21 !!!

[solar jacinth]

@molten pecan

[copper pebble]

yo lil gojo

long time no see you

how are u?

[solar jacinth]

sukuna cut me in half

ToT

[proper shale]

[copper pebble]

[solar jacinth]

I'm feeling sick this entire month

[copper pebble]

://

[solar jacinth]

How about you Mr . helper san

where have you been

[copper pebble]

nah i'd win

[zealous shuttle]

-rvc

[azure marsh]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[copper pebble]

here

[solar jacinth]

LOL

[copper pebble]

sniffing @eager pelican ass

and

[copper pebble]

nothing thats all

[eager pelican]

im a minor

[solar jacinth]

GOD , I thought he is a minor

[copper pebble]

me too

[solar jacinth]

damm he is

[copper pebble]

me too

[eager pelican]

well i dont consent

[copper pebble]

i didnt ask you

[eager pelican]

52 year old mindset

[copper pebble]

si

[solar jacinth]

Theres no way you guys are minor

[eager pelican]

ばか

[copper pebble]

lusb he said that we are not minors

so now we can do whatever we want

[eager pelican]

oh sick
imma kick you in the nuts now

[neat breach]

mmm nohttps://media.discordapp.net/attachments/1192771583899619378/1192779673533419581/image.png?ex=65aa51a2&is=6597dca2&hm=c5b208c28a4af8aa4dc5497db2ebba5e7c21b9f487a70c5f7d80a4f048fc741c&=&format=webp&quality=lossless&width=850&height=306

[rare goblet]

Ayo? @neat breach level 1 !!!

[copper pebble]

which nuts?? DEEEZ??

[neat breach]

OH MY GOODNESS

YOOOO

YOOOOOOOOOOO

YOU MIGHT BE ONTO SOMETHING

[copper pebble]

o che cazzo succede

[sand iris]

yeah the browser instance that voice changer supplies is broken

somehow

[neat breach]

thats so fucking weird

[copper pebble]

ah sera gasato il bro

[sand iris]

it just opens up a web page, so you can connect to it via a regular web browser

but I'm really confused because it should just not break. It breaks slightly differently when it's extracted on anything but C:\ , and the workaround for it is the same

[solar jacinth]

-RVC

[azure marsh]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[molten pecan]

Yeah?

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[tulip cloak]

Hey do any of you use karafan colab? I am bit stuck i don't know what to put in the settings like if i want clean vocals how will I do that

[unique igloo]

hey does anyone know how to solve the

[VCClient] Access http://127.0.0.1:18888/
[VCClient] wait web server...0 http://127.0.0.1:18888/
[VCClient] wait web server...10 http://127.0.0.1:18888/
[VCClient] wait web server...20 http://127.0.0.1:18888/
[VCClient] wait web server...30 http://127.0.0.1:18888/
[VCClient] wait web server...40 http://127.0.0.1:18888/
[VCClient] wait web server...50 http://127.0.0.1:18888/
[VCClient] wait web server...60 http://127.0.0.1:18888/

it keeps showing those lines and comes with a white windows

ppl says deleting stored_settings.json would work but i cant find it anywhere in the extracted folder

[forest sluice]

Use the Windows search function for your best bet

[unique igloo]

the find result aint better

[tawny garnet]

-colab

[azure marsh]

☁️ Google Colabs

• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab

🤗 Hugginface Spaces

• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HF, by r3gf Huggingface Spaces
• AICoverGen, by r3gf Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces
• Advanced RVC Inference, by neuclya Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[tulip cloak]

-audio

[azure marsh]

📚 Audio Guides

• Perfecting Audio Isolation on Low-End Rigs: A Practical Guide, by Litsa The Dancer and Faze Masta Google Docs
• Gathering and Isolating Audio, by SCRFilms Google Docs
• How to make a good model All-In-One guide, by LUSBERT Rentry
• Vocal Mixing Tutorial, by Roomie YouTube

🛠️ Tools

Youtube Downloaders

• https://github.com/yt-dlp/yt-dlp
• https://github.com/dsymbol/yt-dlp-gui

Audio Separation/Isolation

• https://mvsep.com/
• https://multimedia.easeus.com/vocal-remover/

More commands: -colab, -uvr, -karafan, -overtrain, -help

[tulip cloak]

-uvr

[azure marsh]

Ultimate Vocal Remover

One of the best free and open source vocal and instrumental isolation tool.

Tip: Use /guides UVR to learn how to use it.

[pure scroll]

delete the “stored_settings.json” file in the “MMVCServerSIO” folder and restart the app.

[pure scroll]

it will be on a folder named "MMVCServerSIO"

[unique igloo]

i tried

[rare goblet]

Ayo? @unique igloo level 1 !!!

[unique igloo]

the problem is i cant find the stored_settings.json in MMVCServerSIO

[pure scroll]

send screenshot

[pure scroll]

just send the folder's screenshot

don't search it

[pure scroll]

send this folder's components screenshot

[unique igloo]

@pure scroll have u found the problem yet?

[pure scroll]

[unique igloo]

[small field]

whats diference beetwen crepe, crepe tiny, mangio crepe, mangio crepe, rvmpe, harvest

and which one is best for songs

[proper shale]

crepe, crepe tiny and mangio-crepe r all fundamentally same

crepe tiny is the smaller version, means its slightly worse quality but it's faster

crepe and mangio crepe are high quality but very sensible to noise

rvmpe is faster and has roughly the same quality as crepe and mangio crepe without being so sensible to noise

harvest isn't that susceptible to noise but honestly I would not use it as rvmpe is the better version of it

[surreal sonnet]

i forgot how to download python and pytorch

[small field]

<3

[proper shale]

you don't need to do that for original rvc

i think for most forks you don't need to do so

[plush compass]

Is there a free Colab with autosave? I'm making some experimental works, but using Colab Pro is a waste of money...

[proper shale]

-colab

[azure marsh]

☁️ Google Colabs

• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab

🤗 Hugginface Spaces

• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[proper shale]

EasyGUI

[plush compass]

Isn’t that only possible with Colab Pro?

[proper shale]

nope, there's the non WebUI part

just use that

and don't open the webUI step

[molten pecan]

What about Mangio?

[plush compass]

wow thank you so much

[proper shale]

tho i would recommend RVC disconnected but since you have to auto save...

[proper shale]

don't think you need to do that for mangio too

[molten pecan]

Thought so

Thanks

[plush compass]

I'm really sorry, but are detailed settings not possible in easy Colab? For example, settings like 44k and 48k.

[proper shale]

...yeah

it's "easy" so it doesn't give you options

I believe you can change those parameters inside the code tho

I know that in the training step it has the batch size setting set to 12

[plush compass]

Ah, thank you. Then there is a high probability that it is set to 44k as the default setting. It is literally an easy colab.

[proper shale]

Prolly set to 40k. RVC doesn't have 44khz

Only 32khz and 48

[plush compass]

Oh, I was confused between 40k and 44k, sorry haha.

This is a great colab for making experimental models. Thank you.

[grand wing]

hi guys, can someone help me fix that, in voicechanger dont work RVC, but work Beatrice

[proper shale]

You're welcome

[proper shale]

What GPU do you have

[grand wing]

NVIDIA GeForce RTX 3050 Ti Laptop GPU(4GB)

[proper shale]

hmmm check cmd

[grand wing]

sec

[rare goblet]

Ayo? @grand wing level 1 !!!

[grand wing]

check out the private messages.

[proper shale]

looks fine, try to load an rvc model and send another screenshot

[grand wing]

ok, sec

[grand wing]

check

[brittle wing]

Just want to ask how I could turn it off that I hear myself talkin twice

[proper shale]

enable the sup2, echo options

[brittle wing]

ah thanks!

[proper shale]

you're welcome

[brittle wing]

One more question, I can't plug it together with my discord, I can't talk when I set my mic to cable input

[rare goblet]

Ayo? @brittle wing level 1 !!!

[proper shale]

The input output settings should be like this in W-Okada:

Input - Your mic
Output - CABLE Input

And in Discord:

Input - CABLE Output
Output: Your headphones

this is the right outputs

[brittle wing]

Thats a fast answer

Ye that's what I did so far

Still not working

[visual widget]

my voice changer worked this morning now when I tested again, seems that it doesnt pick up my mic that clearly..

[marble drum]

yoooo

i need helpppppp

[brittle wing]

how do i stop hearing myself?

[half cove]

screenshot ur settings rq

[narrow dove]

Hey guys i got a question i wanted to upgrade my gpu from a 5700 XT to a 7800XT. Does the Voicechanger better work with a Nvidia tho? Should i rather upgrade to an equivalent Nvidia Graphics Card for that manner?

[brittle wing]

i can't send images

can you check dms

[tawdry egret]

yo anyone know why my client is just scuffed

it used to work perfectly fine with a 1660ti just with a lot of delay, i switched to a 4060 and ever since my client hasnt been working properly. The ai audio is like super delayed by like 15 seconds and cuts out

[whole fjord]

mine doesnt work straight up

everthing works fine but when i click i start i dont hear anything

even if i added the input and output

[violet heron]

#🔍│help-w-okada

[tawdry egret]

you got your monitor on?

[whole fjord]

on deafult yes

[rare goblet]

Ayo? @whole fjord level 1 !!!

[tawdry egret]

make sure its your heaphones

or what ever you use for your pc audio

[violet heron]

Send a screenshot

What’s your GPU?

[whole fjord]

1050ti

[violet heron]

You downloaded the DirectML version

Which is for AMD GPUs

[tawdry egret]

but does anyone know whats wrong with mine, should i download the latest version?

[rare goblet]

Ayo? @tawdry egret level 1 !!!

[whole fjord]

then whats for geforce?

[violet heron]

https://huggingface.co/wok000/vcclient000/resolve/main/MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.17b.zip

Download this one instead

This one is for NVidia GPUS

[whole fjord]

ty friend

[tawdry egret]

why is it so big mah boy

[whole fjord]

do i have to wait for this

oh

nvm

neat

[tawdry egret]

why does it just keep on saying this and not work

[stray wolf]

can anyone in private message help me with RVC v2 Disconnected error

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[sand iris]

scroll up and re screenshot

[sage spade]

I have some problems with ilaria rvc. After upload my model it is ok but sometimes I lost it and I have to upload it again. Another problem I had is a file speedup and pitched up...

[sand iris]

delete the pretrain directory and relaunch

make sure to let it download and not close it on next launch

[tawdry egret]

what is that?

[sand iris]

you unzipped a file right ?

[tawdry egret]

yes

[sand iris]

there's a folder called MMVCServerSIO

right under that there's another one called pretrain

[tawdry egret]

this is all I have

[sand iris]

under MMVCServerSIO

[tawdry egret]

you mean in?

[sand iris]

yeah

[tawdry egret]

so is it called pretrain directory?

or just pretrain?

[rare goblet]

Ayo? @tawdry egret level 2 !!!

[sand iris]

just pretrain

folder/directory, different word for the same thing, sorry for being confusing lol

[tawdry egret]

ok i deleted it

thanks that worked

what protect do?

[sand iris]

Oh, been a long while since I checked those

[tawdry egret]

yeah all good

but now its not letting me upload voice models

[sand iris]

huh

what's the log saying?

Might have to delete the model_dir folder ? Not sure without logs

[proper shale]

oh maybe you have low space

[sand iris]

Seems like your disk no longer has any space left

[proper shale]

clear some gigabytes up

[tawdry egret]

on what

[sand iris]

the disk voice-changer has been unzipped on

[tawdry egret]

oh

[viral flint]

would a 3060 12gb vram or a 4060 8gb vram be better for rvc?

[rare goblet]

Ayo? @viral flint level 3 !!!

[regal maple]

its also very slow, i cloned the repo but it shows 0 cuda/uda on 3060, takes several minutes just to separate audio

[rare goblet]

Ayo? @regal maple level 1 !!!

[regal maple]

nvm, im talking about aigencover from huggingface

mb its setup for huggingface thats what its on 0 cuda/uda

[molten pecan]

how long was the audio sample?

when it's longer than around a minute and a half you get that error

[azure marsh]

📚 Audio Guides

• Perfecting Audio Isolation on Low-End Rigs: A Practical Guide, by Litsa The Dancer and Faze Masta Google Docs
• Gathering and Isolating Audio, by SCRFilms Google Docs
• How to make a good model All-In-One guide, by LUSBERT Rentry
• Vocal Mixing Tutorial, by Roomie YouTube

🛠️ Tools

Youtube Downloaders

• https://github.com/yt-dlp/yt-dlp
• https://github.com/dsymbol/yt-dlp-gui

Audio Separation/Isolation

• https://mvsep.com/
• https://multimedia.easeus.com/vocal-remover/

More commands: -colab, -uvr, -karafan, -overtrain, -help

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[sage spade]

Yes...too slow... Other alternatives?

[sage spade]

But not Always

[mental salmon]

hi umm i have a probles i have a rayzen 5 and AMD radeon but the voice changer dont work

[rare goblet]

Ayo? @mental salmon level 1 !!!

[mental salmon]

idk why

[brittle wing]

Running with the runtime Python, Please wait.
Microsoft Visual C++ Redistributable is not installed, this may lead to the DLL load failure.
It can be downloaded at https://aka.ms/vs/16/release/vc_redist.x64.exe
Traceback (most recent call last):
File "C:\Users\runde\Documents\RVC-GUI-pkg\rvcgui.py", line 10, in <module>
import torch
File "C:\Users\runde\Documents\RVC-GUI-pkg\runtime\lib\site-packages\torch_init_.py", line 133, in <module>
raise err
OSError: [WinError 126] The specified module could not be found. Error loading "C:\Users\runde\Documents\RVC-GUI-pkg\runtime\lib\site-packages\torch\lib\c10.dll" or one of its dependencies.
Press any key to continue . . .

[azure marsh]

Suggestion for @uneven elk

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

Suggestion for @uneven elk

huggingface.co upload Guides

English guide

Link: https://rentry.org/fdg_guide_newer
Credits: FDG

Spanish Guide

Link: https://rentry.org/AI_Hub_guide_esp
Credits: Julia (ailen2091)

Hindi guide

Link: https://rentry.org/AI_hub_guide_hi
Credits: Enes

Turkish guide

Link: https://rentry.org/AI_hub_guide_tr
Credits: Enes

[uneven elk]

HELP! 😭

[mental salmon]

all here want help my brother all want help

what is ur problem?

[uneven elk]

So..

[mental salmon]

i have curiosity

[uneven elk]

I keep having this upload problems when I tried to upload the Voice AI Model

it keeps being error

Would you help me?

@mental salmon

[mental salmon]

but i idk nothing about this

is just curiosity

[uneven elk]

oh..

So I cant upload anymore?!

[mental salmon]

idk man

i only cant use the 4 models on the app

i cant use the app

idk why

[uneven elk]

im confused whaa

[orchid coral]

hello i have a question

how can i use my own voice

😭

i can't find a tutorial on yt

[narrow dove]

Hey guys i got a question i wanted to upgrade my gpu from a 5700 XT to a 7800XT. Does the Voicechanger better work with a Nvidia tho? Should i rather upgrade to an equivalent Nvidia Graphics Card for that manner?

[violet heron]

-rvc

[azure marsh]

Suggestion for @orchid coral

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[sand iris]

yeah works better on nvidia for the time being

[proper shale]

You should probably get an NVIDIA GPU, since compatibility with AMD exists but it's not that good

[nimble vapor]

has anyone tried the intel arc gpus with rvc?

[sand iris]

I recall that some people did

the main RVC repo should have support for them but I'm not 100% sure if it's fully compatible

[terse tartan]

is it ok if a model has 2 pth and 0 index? what to do in this case?

[violet heron]

How did you end up with 2 pths

You definitely have the wrong files

[terse tartan]

downloaded a model

2 pth, 1 config jso

[violet heron]

Who was it uploaded by

There should only be 1 pth and a index

[terse tartan]

none of this server's members so it's fine

[terse tartan]

okay

[brittle wing]

you dont need an index if realtime, if rvc technically you can use any index just turn slider to 0 to not use it, index files are universal but semi weird when mixing, as you basically are giving whatever accent/habits to a dif person, small amounts like 40 and under wont cause any negative effects though

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[untold ridge]

how to install new models?

[small umbra]

..how to install dez nuts in yo mouf?*

[half cove]

😭

[thorny shoal]

Is there like one generally speaking "superior" RVC thing or is it just kind of
few are better at specific things than others, etc
like would I just be better off just using RVC V2 or is there actually something to be gained from using some other version / fork / whatever these are technically called
strictly for audio converting btw not really interested in model making yet

[proper shale]

RVC v2 is fine

There isn't anything that "superior"

[thorny shoal]

Alright
just wanted to know cuz I've had RVC v2 working fine for a while and had issues with others just wanted to hear if there's something I'm missing out on or if I'm completely fine just keeping on using what I have

[crude flame]

Whenever i click the refresh button these get stuck in queue. The console doesnt output any errors. How do i fix it?

[granite flower]

yo for me its not showing my gpy and it says like GPU(dml) cpu gpu 0 gpu1 gpu2 more info

[viral flint]

if i want a convincing rvc, do epochs matter?

[brittle wing]

where did the localtunnel option go for the google colab?

[civic onyx]

"index -1 is out of bounds for axis 0 with size 0"
^ i get this shit when i try to use the spongebob model on AICoverGen

[proud elbow]

not enough VRAM due to having some games/apps using VRAM run (check VRAM usage in task manager)

[civic onyx]

your a supporter correct?

help me with the issue above if you can

[proud elbow]

wdym supporter?

[civic onyx]

idfk

and i dont care

do you know the solution

or nah

@proud elbow ^

[crude flame]

only 2 of my 12gb is being used and its still doing it

[proud elbow]

and GPU usage 0%? no idea on such issues in Applio, do others like original RVC or Mangio work on yours?

[crude flame]

less then 10 percent usage. The others dont work because they dont support AMD (unless they changed that)

[proud elbow]

you're using AMD GPU? it seems not well supported

[proud elbow]

-howtoask

[azure marsh]

How to troubleshoot

• Give context: Throwing your error/issue alone will not help. Tell what step are you on, what were you doing before issue, your settings, your terminal, et cetera.
• Be relevant: You come to ask for help, not to complain. You can't ask for RVC help in voice changer help and vice versa.
• Be polite: DONT DEMAND FOR HELP OR I WILL FUCKING MURDER YOU. Ask politely, the server was not made for you.
• Don't be an attention whore: If people are ignoring your help, kindly ping them. If they are ignoring you, either they don't know how to solve it, they are busy, or you are an asshole.

WITH ALL DUE RESPECT, DO NOT BE AN ASSHOLE AND HAVE COMMON SENSE.

[civic onyx]

i did all of those

[proud elbow]

explain your problem pls, I don't see yours when scrolling up

[civic onyx]

when i use AICOVERGEN

in COLLAB

i get an "error"

that says

"index -1 is out of bounds for axis 0 with size 0"

when i try to use models

not all

by a few

the error Happens

when i try to Generate

the cover

[proud elbow]

others may help you since I don't use AIcovergen yet

[civic onyx]

a h

[crude flame]

Would me telling you im on linux change anything?

[proven hill]

HIGH QUALITY FEMALE VOICE MODELS

Are you searching girl voices like e-girl, mommy, v-tuber and more?
You find them here: https://ko-fi.com/ilariaowo/shop

INCLUDING FREE ONES!!

And for paid items, if you DM me, you will get a 20% discount if you pay with

I also accept Steam gift cards as last resort

Grab yours today!!!

[leaden stag]

how do i get the software for the voices 💀

[barren plume]

i need help with the playboi carti voice not sounding how its supposed too

[thorny shoal]

idk if its okay to ask this here but since its done in RVC I guess technically this is the right place?
I feel like this question is bit complicated but what causes such inconsistency(?) when converting? Or like what are the probable causes and how to avoid such problems/fix it?
This is uh The Neighbourhood Sweater Weather with JSchlatt model (1000 epochs + RMVPE in case the link doesnt work) https://discord.com/channels/1159260121998827560/1175557559164354581
I did -5 transpose cuz it sounded more like the person I did also try -7 which works too but idk one seemed to be bit more consistent while -12 was just way too much and 0 didnt sound really right
mangio-crepe (cuz I heard that its better for smoother stuff and like idk isnt the song pretty damn smooth-) with 118 hop length
0.6 feature ratio
0.2 envelope
0.35 protect voiceless consonants

also before actually putting it into the oven I did in UVR de-echo-normal and ReverbHQ (also did separate background singers with 5_HP_Karaoke and then did same de-echo and reverbHQ to that and that has even bigger problem with it just sounding horrible after converting)

[rare goblet]

Ayo? @thorny shoal level 2 !!!

[tulip cloak]

@proud elbow i am sorry for being random but do you know how to use the karafan colab

[proud elbow]

haven't used it yet

[tulip cloak]

alright thanks

[mental salmon]

bro i need help

the app literally dont work to me

i istall it 3 times

help pls

idk what do

[narrow dove]

Is it hard to train ur own AI Voice?

[rare goblet]

Ayo? @narrow dove level 2 !!!

[solar jacinth]

@glad zealot

[violet heron]

What doesn’t work?

[forest sluice]

Which notebook?

[forest sluice]

https://github.com/Captain-FLAM/KaraFan/wiki/🚀-Install-Colab

[brittle wing]

https://colab.research.google.com/github/w-okada/voice-changer/blob/master/Realtime_Voice_Changer_on_Colab.ipynb

I just want to try localtunnel because ngrok is having issues with it constantly disconnecting

[forest sluice]

Oh, that notebook doesn't implement it

[brittle wing]

Do you have one that has localtunnel?

[tulip cloak]

yeah but what things do you put in remove music, vocals bleeding etc etc and why is there 2 rows.

[brittle wing]

Actually nvm, let me follow this tutorial I found really quick and see if that solves my issue

[rare goblet]

Ayo? @brittle wing level 1 !!!

[forest sluice]

hey, colab bans webuis

it's not ngrok as the culprit, the voicechanger app is

[tulip cloak]

there should be a tuto for karafan

[glad zealot]

AAAAAAA

[solar jacinth]

:3

[brittle wing]

Do you know how to fix the 'Pipeline is not initialized.' error I keep getting in colab?

[mental salmon]

Where i press the characters, dont talk and i found the probles bc i have a AMD

Literally is that i select a character and speak but the app dont filter my voice

[violet heron]

What version did you download?

[mental salmon]

The 16

[violet heron]

Can you send a screenshot?

[mental salmon]

All see normal the problem is when i press start and speak but i dont Hear the voice

I think is for the cpu

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[violet heron]

Can you send a screenshot

So I can make sure you 1. Got the right version 2. Have the right settings

[mental salmon]

I cant i off my pc ;_; can i send tomorrow? Pls

[rare goblet]

Ayo? @mental salmon level 2 !!!

[mental salmon]

Ooo

[violet heron]

Make sure it say as DirectML

if it doesn’t you downloaded the wrong version

You said you had a AMD GPU right?

[mental salmon]

I donde have a GPU

Dont*

Only CPU

[violet heron]

That’s why

[mental salmon]

[violet heron]

Okada cannot run not even decently with only a CPU

you can try the online google colab instead for okada

https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb

[mental salmon]

[mental salmon]

Ohhh

Tysm

Oh is the same Page i try it

Word but whit so delay

I think is only a tash pc

Ty for all my boy

I apreciate it

[hollow grail]

5.13.16b i the latest right? Cause it doesnt filter my voice it just repeats it in my normal voice.

[violet heron]

Is passthru on?

[hollow grail]

Oh yeah

[violet heron]

Turn it off

[hollow grail]

How do you check if the audio is working with a filter then.

Discord playback?

[violet heron]

Set monitor to your headphones/speakers

[hollow grail]

output should also be headphones right?

and the input is your mic

[violet heron]

Yes

Output should be the virtual audio cable

[violet heron]

(Which should be the input on discord)

[hollow grail]

in the program and discord right.

[violet heron]

The program yes

On discord the cable should the input

[ebon heath]

There is a lot of delay when using the RVC while gaming (~10 second). how would I go about setting priority to the RVC?

[violet heron]

Lower you’d game settings

[hollow grail]

Thank you!

[brittle wing]

Last Error Received:

Process: MDX-Net

If this error persists, please contact the developers with the error details.

Raw Error Details:

MemoryError: "Unable to allocate 6.18 GiB for an array with shape (2, 829691904) and data type float32"
Traceback Error: "
File "UVR.py", line 6638, in process_start
File "separate.py", line 499, in seperate
File "separate.py", line 608, in demix
File "<array_function internals>", line 180, in concatenate
"

Error Time Stamp [2024-01-06 09:58:49]

Full Application Settings:

vr_model: Choose Model
aggression_setting: 5
window_size: 512
mdx_segment_size: 256
batch_size: Default
crop_size: 256
is_tta: False
is_output_image: False
is_post_process: False
is_high_end_process: False
post_process_threshold: 0.2
vr_voc_inst_secondary_model: No Model Selected
vr_other_secondary_model: No Model Selected
vr_bass_secondary_model: No Model Selected
vr_drums_secondary_model: No Model Selected
vr_is_secondary_model_activate: False
vr_voc_inst_secondary_model_scale: 0.9
vr_other_secondary_model_scale: 0.7
vr_bass_secondary_model_scale: 0.5
vr_drums_secondary_model_scale: 0.5
demucs_model: Choose Model
segment: Default
overlap: 0.25
overlap_mdx: Default
overlap_mdx23: 8
shifts: 2
chunks_demucs: Auto
margin_demucs: 44100
is_chunk_demucs: False
is_chunk_mdxnet: False
is_primary_stem_only_Demucs: False
is_secondary_stem_only_Demucs: False
is_split_mode: True
is_demucs_combine_stems: True
is_mdx23_combine_stems: True
demucs_voc_inst_secondary_model: No Model Selected
demucs_other_secondary_model: No Model Selected
demucs_bass_secondary_model: No Model Selected
demucs_drums_secondary_model: No Model Selected
demucs_is_secondary_model_activate: False
demucs_voc_inst_secondary_model_scale: 0.9
demucs_other_secondary_model_scale: 0.7
demucs_bass_secondary_model_scale: 0.5
demucs_drums_secondary_model_scale: 0.5
demucs_pre_proc_model: No Model Selected
is_demucs_pre_proc_model_activate: False
is_demucs_pre_proc_model_inst_mix: False
mdx_net_model: UVR-MDX-NET Inst HQ 3
chunks: Auto
margin: 44100
compensate: Auto
denoise_option: Standard
is_match_frequency_pitch: True
phase_option: Automatic
phase_shifts: None
is_save_align: False
is_match_silence: True
is_spec_match: False
is_mdx_c_seg_def: False
is_invert_spec: False
is_deverb_vocals: False
deverb_vocal_opt: Main Vocals Only
voc_split_save_opt: Lead Only
is_mixer_mode: False
mdx_batch_size: Default
mdx_voc_inst_secondary_model: No Model Selected
mdx_other_secondary_model: No Model Selected
mdx_bass_secondary_model: No Model Selected
mdx_drums_secondary_model: No Model Selected
mdx_is_secondary_model_activate: False
mdx_voc_inst_secondary_model_scale: 0.9
mdx_other_secondary_model_scale: 0.7
mdx_bass_secondary_model_scale: 0.5
mdx_drums_secondary_model_scale: 0.5
is_save_all_outputs_ensemble: True
is_append_ensemble_name: False
chosen_audio_tool: Manual Ensemble
choose_algorithm: Min Spec
time_stretch_rate: 2.0
pitch_rate: 2.0
is_time_correction: True
is_gpu_conversion: True
is_primary_stem_only: False
is_secondary_stem_only: True
is_testing_audio: False
is_auto_update_model_params: True
is_add_model_name: False
is_accept_any_input: False
is_task_complete: False
is_normalization: False
is_use_opencl: False
is_wav_ensemble: False
is_create_model_folder: False
mp3_bit_set: 320k
semitone_shift: 0
save_format: WAV
wav_type_set: PCM_16
device_set: Default
help_hints_var: True
set_vocal_splitter: No Model Selected
is_set_vocal_splitter: False
is_save_inst_set_vocal_splitter: False
model_sample_mode: False
model_sample_mode_duration: 30
demucs_stems: All Stems
mdx_stems: All Stems

I keep getting this error

Last Error Received:

Process: MDX-Net

If this error persists, please contact the developers with the error details.

Raw Error Details:

MemoryError: "Unable to allocate 6.18 GiB for an array with shape (2, 829691904) and data type float32"
Traceback Error: "
File "UVR.py", line 6638, in processstart
File "separate.py", line 499, in seperate
File "separate.py", line 608, in demix
File "<_array_function internals>", line 180, in concatenate
"

Error Time Stamp [2024-01-06 09:58:49]

UVR5

[woven furnace]

hi are my models saved to the ilaria program? because i want to keep my model private.

[brave garnet]

Liste dei modelli

⠀

• Lista dei modelli italiani

Link

• Lista dei modelli internazionali

Link

• Richiedi nuovi modelli

Link

⠀

Made with love by Ilaria 💖

[woven furnace]

@proven hill

[thin stump]

@brazen bough Sorry for pinging you, I should had answered your question. The least complex one that you can do locally on your computer without an internet connection is https://www.tryreplay.io/

[brazen bough]

Thank you 🙂

[rare goblet]

Ayo? @brazen bough level 1 !!!

[thin stump]

Yw, they are associated with weights.gg. The site where you get models from

[thin stump]

https://free-mp3-download.net/

It says MP3 on the website, but it should all be Flac files

[sage spade]

best settiungs on ilaria rvc for High pitched notes?

[carmine osprey]

how many epoches would good for 15min dataset

[proper shale]

go for 500 and check the graphs

for every size of dataset

[zinc anchor]

I usually do 300 for 15 minutes

[copper pebble]

did ya duplicate the space?

[solar jacinth]

everyone

[neon valve]

Why is it that sometime when I train a model for example epoch 200, it just stop training at e95 and said it's finished. It's pretty frustrating.

Yes I'm using GG Collab because my computer does not have enough memory

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[proven hill]

if you duplicated the space its private

[woven furnace]

yes

[woven furnace]

😭

[proven hill]

then its private dont worry

[woven furnace]

thank you ❤️

[woven furnace]

i love ilaria rvc

[proven hill]

too bad its slow because huggingface use cpu

otherwise..

[woven furnace]

yea it is a bit slow but it works well

[proven hill]

at least i got something right

[copper pebble]

non so se te l'ho detto ho fatto un intero cd con Ilaria_RVC

[proven hill]

cosa cosa?? AHAHAH

[copper pebble]

common Ilaria_RVC w

[calm jackal]

How can i use it

[warped cove]

Yo guys

[woven furnace]

is it possible to delete my model in it in case i made a mistake not making it private

[proven hill]

restart the space

3 dots on top right, settings, scroll down, restart space

[warped cove]

Let’s say I alr used someone else’s model, but it doesn’t sound quite right but it’s good how do I make it sound better

[rare goblet]

Ayo? @warped cove level 1 !!!

[warped cove]

Like it sounds unnatural

[woven furnace]

thank you

i can't find that option on the three dots

[gleaming plover]

I`m getting this error in Applio-rvc

Traceback (most recent call last):
  File "D:\rvc\Applio-RVC-Fork\infer-web.py", line 60, in <module>
    from lib.modules.vc.modules import VC
  File "D:\rvc\Applio-RVC-Fork\lib\modules\vc\modules.py", line 12, in <module>
    from lib.modules.infer.infer_pack.models import (
  File "D:\rvc\Applio-RVC-Fork\lib\modules\infer\infer_pack\models.py", line 12, in <module>
    from torch.nn.utils.parametrizations import spectral_norm, weight_norm
ImportError: cannot import name 'weight_norm' from 'torch.nn.utils.parametrizations' (D:\rvc\Applio-RVC-Fork\env\lib\site-packages\torch\nn\utils\parametrizations.py)```

[forest sluice]

Have you updated Applio to latest version?

[gleaming plover]

Yes

[rare goblet]

Ayo? @gleaming plover level 1 !!!

[forest sluice]

Also, what's the model you're trying to infer?

[gleaming plover]

I'm just launching applio

And then error appears

I already reinstalled it

[gleaming plover]

two times

but issue still exist

[polar grove]

Real time voice changer is so broken

When i try installing a model it gets to a certain percent in the program then just freezes

[neon valve]

You guys are mean for leaving me hanging like that.

[woven furnace]

@proven hill

[half cove]

Screenshot ur screen and send it

[molten pecan]

check the spanish chat. I bringed another staff member to help with the docs

spanish chat is the new staff chat

[brittle wing]

hey rvc dont work on my pc can someone make me a model please, i have the audio and everything

[molten pecan]

you can make models through the cloud on any device too

check the RVC Disconnected post on #1159513888199540817

otherwise you can pay an expert to make it for you in #1191429836321849435

[woven furnace]

i tried screenshotting it but it goes away when i press a key to screenshot it

[half cove]

😭

[half cove]

Wat

[woven furnace]

yea

[half cove]

Windows shift s

[atomic gyro]

It has been downloaded successfully.
Proceeding with the extraction...

The model works for inference, and has the .index file.

im using applio and its stuck there

[half cove]

Then go on the inference page click the model refresh button then on inferring voice put the voice you downloaded and check if the index is also used

[half cove]

Screenshot your entire page

[serene ginkgo]

hi have a question how can i change so the program uses my gpu

it only uses the cpu idk why

and how to change it

[half cove]

@woven furnace u there?

[half cove]

what gpu you have

[serene ginkgo]

rx 6600

[rare goblet]

Ayo? @serene ginkgo level 1 !!!

[half cove]

and what program you using

[half cove]

you prob installed the nvidia one

[serene ginkgo]

ou

oh

fuck

how do i uninstall?

[half cove]

just delete the folder lol

https://huggingface.co/wok000/vcclient000/resolve/main/MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.17b.zip

[serene ginkgo]

oh

okk

[half cove]

this should be the right one

[serene ginkgo]

recommend any settings?

[half cove]

https://rentry.co/VoiceChangerGuide

[serene ginkgo]

wait

how do i get voices

on the program

i already have 2

[crimson basalt]

training is not working for me now for some reason with applio. ive re installed and deleated applio 3 times now, could it be somthing to do with the error? thats all i can think of... RuntimeError: Expected tensor for argument #1 'input' to have the same type as tensor for argument #2 'weight'; but type torch.cuda.FloatTensor does not equal torch.cuda.HalfTensor (while checking arguments for cudnn_batch_norm)

[half cove]

i think you click edit and then upload the model

[serene ginkgo]

ty

[serene ginkgo]

what's that program for?

or the website?

[half cove]

should be for w-okada

[serene ginkgo]

how do i use it

[half cove]

the guide

[serene ginkgo]

oh

im new to this

so that's why

[half cove]

the guide should say how to use it

[serene ginkgo]

ok ty

[rare goblet]

Ayo? @serene ginkgo level 2 !!!

[serene ginkgo]

how do i upload it

[half cove]

ask in #🔍│help-w-okada

[warped cove]

Pth as the model

Added as the index

[half cove]

oh shi he prob put both in a the same time ig

[brittle wing]

how can I just train a voice model please someone help me

[violet heron]

-upload

[azure marsh]

Suggestions for @serene ginkgo

huggingface.co upload Guides

English guide

Link: https://rentry.org/fdg_guide_newer
Credits: FDG

Spanish Guide

Link: https://rentry.org/AI_Hub_guide_esp
Credits: Julia (ailen2091)

Hindi guide

Link: https://rentry.org/AI_hub_guide_hi
Credits: Enes

Turkish guide

Link: https://rentry.org/AI_hub_guide_tr
Credits: Enes

[serene ginkgo]

thanks

i have another question, why is there so much delay

like 20 seconds

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

https://docs.aihub.wtf/v/brazil

🇪🇸 Spain

https://docs.aihub.wtf/v/spain/

🇫🇷 France

https://docs.aihub.wtf/v/france

🇩🇪 Germany

https://docs.aihub.wtf/v/germany

🇯🇵 Japan

https://docs.aihub.wtf/v/japan

🇰🇷 Korea

https://docs.aihub.wtf/v/korea

🇵🇱 Poland

https://docs.aihub.wtf/v/poland

🇷🇺 Russia

https://docs.aihub.wtf/v/russia

🇮🇹 Italy

https://docs.aihub.wtf/v/italy

[woven furnace]

it doesnt show anything that different

[rare goblet]

Ayo? @woven furnace level 4 !!!

[half cove]

your trying to reset the uhh space?

[steady fractal]

i followed all the settings and my audio sounds just lik3e bubbling and crackling what should i do

[sleek jacinth]

same

[halcyon osprey]

How do I find a specific voice model if the bot gives a link, but I can't download the model from there. Maybe there are alternative sites that have these voices?

[rare goblet]

Ayo? @halcyon osprey level 1 !!!

[proper shale]

if it isn't in #1175430844685484042, then it could be in weights.gg

[steady fractal]

i fixed it by following the gpu guid here https://rentry.org/fdg_guide_newer

[proven hill]

E-GIRL VOICE MODEL SALES!!

DM ME FOR DISCOUNT PRICES!

E-Girl, Mommy, V-Tuber, Anime girl and much more!

Wondering what models?
You find them here: https://ko-fi.com/ilariaowo/shop

INCLUDING FREE ONES!!

I also do commissions if you want aspecific voice!

I accept the following payment methods:

• Ko-Fi
• PayPal
• Steam Gift Cards
• Amazon Gift Cards
• Throne

Grab yours today!!!

[half cove]

[brittle wing]

is there a way to make it sound more like realistic?

[molten cedar]

help what does python3: can't open file '/content/src/main.py': [Errno 2] No such file or directory
2 mean

[woven furnace]

im trying to keep the models i made private and i dont know if i put it in the public or private version

[half cove]

oh

[spice wharf]

im tryna use ilaria voice model and put in a huggingface link and it says its invalid but when i paste in link in browser it starts downloading?

[half cove]

go on settings

and make sure it says you have it on private

[half cove]

send link rq

[spice wharf]

https://huggingface.co/leeloli/strong-momo-by-leelo/resolve/main/strongmomo.zip

[rare goblet]

Ayo? @spice wharf level 1 !!!

[brittle wing]

How can I make it sound more real?

[astral citrus]

The voice I have sounds good but it is like cutting in and out when I talk. Like i cant get a full sentence in without it going in and out. Is it due to my specs? I have a RX 6600

[sand iris]

https://rentry.co/VoiceChangerGuide#uploading-models-fixing-amdarc-res-issues

Updated to link to relevant section

Refer to the part that mentions "exporting to onnx"

[astral citrus]

Gotcha ty

[spice wharf]

hm?

[patent pasture]

@glad zealot, sorry for ping, I say to you that CoverGen is good but here is one bug - sometimes it don't revome vocal

[half cove]

its cuz the didnt put the folders right

[spice wharf]

oh

so i cant use that model?

[half cove]

you can

let me uhh

[patent pasture]

It might be instrumental but here is vocal too

[half cove]

reupload it

[spice wharf]

alr thanks

[half cove]

there are 2 models

what one u trying to use

[spice wharf]

any with momo on it lol

what options are there