✨│ai-help

[cosmic spire]

you want to install it on colab

[gray condor]

this was the tut i was wsatching it on

[gray condor]

is there a link

[cosmic spire]

so you gotta run the jupyter notebook scripts first

[gray condor]

im on collab

but it only shows 2

nd i cant install it

is there any other way to install it

been tryna install it for 2 days but im either stuck or js getting impatient

[cosmic spire]

run the first, wait till it finishes and run the second after following the steps described there

[gray condor]

so i went into this server for help

[gray condor]

okay

ill tell u when im done

thanks

[cosmic spire]

[rare goblet]

Ayo? @gray condor level 2 !!!

[gray condor]

@cosmic spire

do i click continue

ion know

[brittle wing]

yea

[gray condor]

okay

ill ping when im done

[cosmic spire]

-colab

oh wait

they updated the cmd

[cosmic spire]

nah

they'll ban ur google Account

[gray condor]

too late

[cosmic spire]

has happened before

[gray condor]

wait

ill stop it

wait i couldnt

it said done

what do i do

[cosmic spire]

create a new one completely unrelated to the one u have rn and do it there

[cosmic spire]

just delete the notebook

[gray condor]

like this|?

[cosmic spire]

you can do stuff there

[gray condor]

a seperate google acc

and i redo it now

but i wont do the 2nd part

[cosmic spire]

yes

[gray condor]

oka

[cosmic spire]

if you don't mind getting it deleted

and maybe don't leak ur email

people will sign u up for naughty websites and stuff

[gray condor]

LOL

kay

mb

i did it

oh wait

nvm

[gray condor]

im done

i made another ngrok acc but i didnt do the 2nd step

what do i do now

[gleaming laurel]

i want to do a duet cover of a song, is there an 'easy' way to get only the parts sung by one artist and the other? i thought of chopping it up with audacity but it would take way too long so idk if there would be an easier way

[deft edge]

im making gigglesaurus model

[gray condor]

@cosmic spire

[sharp hazel]

guys dose 40m of datasets is good? or too much

[rare goblet]

Ayo? @sharp hazel level 1 !!!

[brittle wing]

can someone dm me to help w sum rq

[rare goblet]

Ayo? @brittle wing level 1 !!!

[cosmic spire]

it should say it in the notebook

[gray condor]

what notebook

..

[brittle wing]

can someone help

[woeful bridge]

yo

someone help me?

Im tryna troll as a e girl

but my mic be fuckin glithcing

like a reobot

someone wanna help a brotha out?

[brittle wing]

What's your graphics card, man?

[woeful bridge]

1660 super

[rare goblet]

Ayo? @woeful bridge level 1 !!!

[woeful bridge]

it worked perfectly fine

like last night

then it just fuckin started glitching like a robot

its my output virtual cable

thats the problem

it aint my mic or graphics card

I jsut need hlep with what virtual cable i should use

[random plume]

How do y'all use the ai voices

[edgy glen]

There's alot of tutorials on yt

[cosmic spire]

jupyter notebook

the colab thing

[tame finch]

does anyone have any experience with ai vtubers?

on the vtube studio side of things

[dapper grove]

Hi does anyone here uses amd gpu? I have a good gpu and it still sounds way too choppy

Does anyone knows any commands or anything at all that could improve the performance?

[tame finch]

what do u mean sounds choppy

what sounds choppy

[dapper grove]

I mean on a way it keeps cutting the voice

The first two default models for me are working fine very rarely I get some voice crack or the voice gets cut

[rare goblet]

Ayo? @dapper grove level 1 !!!

[dapper grove]

Every other models are not working properly

[brittle wing]

What settings should i use for a 1650 super?

[tame finch]

well ur doing the voice training? for rvc

im sorry im not much help here

cause idk much about it my self

[faint charm]

My voice.ai isn't opening it already took 3 hours to start , it's not working!

[brittle wing]

im just using the voice changer thingy

[tame finch]

theres a tutorial on youtube for rvc webui

[brittle wing]

yea i was just seeing if anyone had my gpu and already had good settings

[tame finch]

well i tried it once but not really had much luck

ill attempt again at some point

[faint charm]

where can i find the voice.ai video?

My voice.ai doesn't let me open it anymore

[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

[sharp rain]

how do i create a voice model

[rare goblet]

Ayo? @sharp rain level 1 !!!

[lusty ermine]

The app has access to my microphone but it is not detecting my voice and the microphone is working fine then when I looked at it I realized that Beatrice is working but rvc is not detecting my voice

[arctic sage]

how do i make my voice less delayed

[hushed tiger]

Why does it take so long for the voice changer to work when I have a good gpu

RX 6700 XT

[proper shale]

Check the https://rentry.co/VoiceChangerGuide , specifically the "installing models on AMD" part

[proper shale]

What does CMD say

[lusty ermine]

i need the send all?

[rare goblet]

Ayo? @lusty ermine level 1 !!!

[lusty ermine]

Is it writing problem?

[proper shale]

Okay so

Go on the MMVCServerSIO folder and delete the Pretrain folder, and restart W-Okada

[lusty ermine]

okey

[lusty ermine]

thanks its works

[lavish nacelle]

Hmm. Anyone know what I should do?

I made many voice models in the last perfectly fine

I just about managed to make one on collab even tough I kept getting disconnected.

But by the end of it, it now sounds nothing like the person it should.

I need two voices into models for a project I'm working on but don't really know what to do in this case?

[proper shale]

I suggest try retraining the model on batch size 4

And, I believe that you might be using an old Colab - could you send a link of the one you're using?

[lavish nacelle]

Yeah one second

Also training was smooth af so there's no way any loss happened to get it this bad sounding

[lavish nacelle]

Im using this one (the one i always used) https://colab.research.google.com/drive/1TU-kkQWVf-PLO_hSa2QCMZS1XF5xVHqs

but this is the newer one right? it confuses me lol https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw#scrollTo=Sb5fzhzEXK8X

[proper shale]

ohhh the one you always used is outdated yeah

google colab restrictions...

[lavish nacelle]

yeah it kept disconecting me but i still managed to train it and get 300 epoch's so i don't understand why it sounds nothing like who i need

[rare goblet]

Ayo? @lavish nacelle level 1 !!!

[proper shale]

you should use RVC Disconnected though, it has more options n stuff

-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]

^^

there's also a guide for it

-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)

[proper shale]

this happens sometimes, it can happen due to batch size too iirc

[lavish nacelle]

thanks. ill retry it and hope it goes better

[proper shale]

good luck on the model! :)

[lavish nacelle]

thanks, i'll send a message later and let you know

[proven cave]

Can someone help me with only hearing the okada ai voice changer like how can i make it so i kinda disable my own voice

[proper shale]

Make sure the input/output settings r correct

In W-okada:

Input - Your mic
Output - CABLE Input
Monitor (optional) - headphones

And in Discord (or any other program):

Input - CABLE Output
Output: Your headphones

You may have enabled Monitor tho

[proven cave]

yeah they are correct

yeah how to to disabl;

[rare goblet]

Ayo? @proven cave level 4 !!!

[proper shale]

So you can just... set monitor to none and don't hear the AI voice

[proven cave]

done

No i want my own voice to not go through

just the ai voice in discord

[proper shale]

it isn't going through unless you're on passthru

[proven cave]

how can i do that?:

wait then how do duckkus and others do it?

[half cove]

ima install okada so i can screenshot shit rq

[proven cave]

is it this

[proper shale]

i think you misunderstood.

You need to DISABLE passthru so the normal voice doesn't come out (if that's enabled)

[proper shale]

No

[half cove]

top right

[proper shale]

Yeah this is correct

[proper shale]

no they do not need it enabled

[proven cave]

wait

[half cove]

oh

[proven cave]

now I hear 2 off myself

[proper shale]

please for the love of god use rvmpe as ur f0 🙏

[proven cave]

im new lol

[proper shale]

disable passthru

[proven cave]

I did

[proper shale]

now send ur input output settings on discord

[proper shale]

That is correct so uh

Your AI voice is going through

[proven cave]

and also mine

[proper shale]

(or at least it should be)

[proven cave]

yeah it is

anything else?

[proper shale]

literally nothing else i can think of

everything is correct lol

[proven cave]

when i press it shows this

[proper shale]

yeah no you don't need that enabled

[proven cave]

ok disabled

[proper shale]

that just passes your real voice through

[proven cave]

how does duckkus do it

[proper shale]

bro does it the same way as everybody else

[proven cave]

idk i thought there was another program i can use

[proper shale]

might just be your microphone or your windows settings

but eh idk

[proven cave]

imma just use voice ai

[proper shale]

oooooor

there's a possibility

a slight possibility

that this is the virtual audio cable goofin out

try the VAC by Muzychenko

[proper shale]

also, paid software moment

[proven cave]

vac.muzychenko.net

??

[winged grotto]

is there any ai that does text to speech using custom voice models?

[proven cave]

Dude

Dude

dude

@proper shale I love you

[proper shale]

yeye

[proper shale]

Awww thx :) Glad that I could help

[tranquil cliff]

Hi

[proper shale]

The models in here can't do TTS, you'll need to use ElevenLabs to create a tts audio to convert in RVC

[half cove]

hai

[proper shale]

Alternatively you could try to use GPTSoVITS on Huggingface Spaces

[proven cave]

the only bad this is i have to switch between the two

[hearty ravine]

is there a way to seperate a sentence automatically or should I do it by my hands?

[proper shale]

yeah u should do by hand

im sorry

[proper shale]

at least it works 🙏

[proven cave]

eyah

[proven cave]

dude 5*

[half cove]

you know what ima install nvidia and amd okada so i can screenshot

truly amazing

[winged grotto]

Anyone know how to separte tracks cleanly? i use the spitter website but it does it not very good, so new voice is kinda trash aswell

[rare goblet]

Ayo? @winged grotto level 2 !!!

[winged grotto]

it sounds cryspi

[proper shale]

https://www.mvsep.com

remember to register too

[lavish nacelle]

@proper shale yeah sounds way better, only needed 140 epoch's but atleast it's done. thanks

[proper shale]

woo! glad it worked :)

[frosty meteor]

Hello friends, am I the only one who can't use "easy GUI" now?

[rare goblet]

Ayo? @frosty meteor level 1 !!!

[thin harbor]

how do you get the ai voice changer

[half cove]

https://rentry.co/VoiceChangerGuide

[thin harbor]

is all of it safe?

[half cove]

yep

[thin harbor]

ok

[torn seal]

Does Google collab still ban ui?

It's been a while since I been on here

[half cove]

yep for free users

[torn seal]

Anything changed?

Or any other free options?

[half cove]

we have noui colabs

and i think the applio ui colab works

[dire wagon]

Is there any way i can make covers from iPhone? I already have the model

[obtuse vapor]

bro my mic sounds like a robot how do i fix it?

[half cove]

if you use it then make sure you use an alt account

[torn seal]

Oh for real? The applio ui works for free userss?

[dire wagon]

-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

[torn seal]

I saw that they updated it, it's definitely much better. They have a good voice library

[quaint wharf]

how do i make it so RVC does it job when im Tabbed into a game for example. Currently it holds till i tab out again

[proper shale]

do you mean W-Okada?

[quaint wharf]

yes

[proper shale]

huhh it shouldnt be doing that

either way you wouldn't get a good experience while gaming with W-Okada though

[torn seal]

Holly Molly, when did hugging space come out? Is that a colab note book alternative? For ui?

[quaint wharf]

what should i be using then

[rare goblet]

Ayo? @quaint wharf level 1 !!!

[proper shale]

like, years ago at this point (ive used it since like 2022)

for ui, yeah

[proper shale]

i think if you run the voice changer colab it'd be a better experience overall, so both the game and W-Okada wouldn't slow down

it's in the pins of #🔍│help-w-okada

[sharp surge]

why dont my collab work

[crimson kettle]

all the models i downlload on realtime voice changer client are laggy as hell n keep cutting out. The pre downloaded models work tho.

[violet heron]

Which colab are you using?

[crimson kettle]

is my pc just ass?

[violet heron]

Send a screenshot

I’ll tell you what settings to change

[crimson kettle]

can i dm u i dont have perms unless i can get them quickly

[proper shale]

read the https://rentry.co/VoiceChangerGuide

[proper shale]

specifically the section about installing models on AMD

[crimson kettle]

ok thanks

[rare goblet]

Ayo? @crimson kettle level 1 !!!

[dire wagon]

Is there any way i can make covers from iPhone? I already have the model

[brave garnet]

RVC Colabs and Spaces

⠀

HuggingFace Spaces 🤗

⠀
Ilaria RVC
EasyGUI port with some improvements, by Ilaria.

RVC-HFv2
Applio port, by r3gm.

AICoverGen
AICoverGen port, by r3gm.

Advanced RVC Inference
Extended version of the GUI with advanced settings, r3gm.
⠀

Made with love by Ilaria 💖

[proper shale]

^

Ilaria RVC

also read the guide: https://www.rentry.org/ilarvc_inf_guide

[crimson kettle]

tysm man this really helped

[proper shale]

yay, glad i could help :)

[dire wagon]

Is there any way i can make covers from iPhone? I already have the model

[violet heron]

https://www.rentry.org/ilarvc_inf_guide

[dense marsh]

help

whenever I try to generate an ai cover

it just says

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

[crimson kettle]

what do you think is a good amount of epochs for a realistic sounding voice model?
ive tried 500 ones and they're all alright but im yet to find better ones (not looked very far tbf)
also, is there a way to sort by most downloaded? Most epochs? whatever?

[lusty bridge]

sound echo

[glad zealot]

depends on how clean your dataset is, if its very clean, you can even get away with just 10 epochs

[proper shale]

activate sup2 and increase s Thresh a bit

[crimson kettle]

thanks

[glad zealot]

but on most models around 200-300 is good

[cloud zealot]

My voice is lagging and says random things how can I fix that?

[proper shale]

Check https://rentry.co/W-Okada-FAQ

[pearl lantern]

Sorry for asking more questions! But uhm do I have to do anything through terminal for proper deletion with a new install?

Was hoping that a clean start would fix some of the problems I had when it first downloaded wrong

[proper shale]

nah just delete the folder and u should be good

[pearl lantern]

Okok. I scrubbed it clean twice now and then extracted it again. On start_http it will look like it's installing proper then open a blank page. Atleast before I got into the program and the Beatrice model worked, but I can't get it to boot the url

I'll do it the same way and take a picture of the screen and terminal open

[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

[lusty pivot]

Why I can't find my GPU option 🙄

[violet heron]

Does it appear as “gpu0” “gpu1” etc

[lusty pivot]

Yes

[rare goblet]

Ayo? @lusty pivot level 1 !!!

[violet heron]

What’s your GPU?

[lusty pivot]

Rtx 3060 ventus 2x 12 gb

[violet heron]

You downloaded the AMD version

That’s why

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 17b cuda

Download for AMD GPUs
Version 17b directml

Download for Intel GPUs
Version 17b directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[violet heron]

Download the AMD version instead

[lusty pivot]

Thank you so much 💕💕

[shrewd seal]

can't find a .pth file in the logs folder

[violet heron]

That’s because it’s in weights

[shrewd seal]

OH

no it's not

I didn't check the box

[fierce crown]

How do i make a model

[violet heron]

-rvc

[azure marsh]

Suggestion for @fierce crown

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)

[fierce crown]

Thanks

[pearl lantern]

I've uninstalled and reinstalled several times now just to get the same message with python and pytorch installed and without YuY

Imma just install an older version I guess, I don't see any other things like this in the help forum but don't want to keep asking so much

Shoot well older version did same thing.

[half cove]

have you tried asking in #1192011222023950368

[pearl lantern]

Yes I had a discussion in there earlier where MJ helped! I got things up and running after that, but after an uninstall and reinstall to fix some of the assets not downloading (like the four included models) it now won't boot at all no matter what I scrub and redownload

I am making a post now because it seems relevant enough ^u^

[tribal sorrel]

uh my w-okada isin't going into discord?

[jaunty raptor]

31

[half cove]

are you using a virtual audio cable

[tribal sorrel]

ye

it's working now sorry

[half cove]

okie

[unkempt ocean]

Trying to install new Applio, when running run-applio.bat it prompts me to "Please run 'run-install.bat' first to set up the environment." despite everything being installed. Can anyone please help me with this issue?

[brittle wing]

oops

im so impatient

it rendered

[half cove]

-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

[brittle wing]

which is the fastest

[half cove]

aicovergen should be a bit faster iirc

[brittle wing]

ty

[half cove]

also applio colab works

i think

try that

[half cove]

https://colab.research.google.com/github/iahispano/applio/blob/master/assets/Applio.ipynb

it should be faster as it doesnt use the pre-processing

[brittle wing]

What’s everyone training on now ?

[gritty tundra]

Guys, I have a cover here by Russian YouTubers Flo Rida - Whistle

[gray solstice]

I keep getting this error on Voice Changer navtive client Web Server Launch Exception, DLL load failed while importing beatrice_internal_api: A dynamic link library (DLL) initialization routine failed.

[dawn yacht]

M I the only one who feels like google colab manipulated the gpu. Even if we r connected to gpu we can not utilise it ?

[sand iris]

Do you know your computer's CPU model ? Just out of curiosity

I'd link the older versions of voice-changer which should work, but I'm not sure where they are now lol

[pale loom]

best settings for amd radeon rx 6600? also idk if it matters but i use the cuba version that is intented for nvidia because for some reason if i use the other ones i cant select gpu option so it uses my cpu

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for AMD GPUs

Don't forget that your models needs to be converted in ONNX!

F0 Det.: rmvpe_onnx (suggested for all series)

7xxx XT cards: 112-128 chunk | +16384 extra
6xxx XT cards: 128-192 chunk | +16384 extra
5xxx XT cards: 192-256 chunk | +8192 extra

RX 580: 192-256 chunk | +8192 extra
RX 570: 192-256 chunk | +8192 extra
RX 560: 256-384 chunk | +8192 extra

Advanced Settings

Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low

⠀

Made with love by Ilaria 💖

[half cove]

@pale loom ^

[half cove]

your using okada right?

[pale loom]

yeah

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 17b cuda

Download for AMD GPUs
Version 17b directml

Download for Intel GPUs
Version 17b directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[half cove]

use the amd download

[pale loom]

will it be better quality from the one i use now? it works somehow even tho its for nvidia

[half cove]

if you use the nvidia one it will use your cpu tho?

how tf would you use an amd gpu on the nvidia one

[pale loom]

im not sure when i downloaded the amd ones i could only choose cpu then i downloaded the nvidia one just to test and i could chose gpu

[rare goblet]

Ayo? @pale loom level 1 !!!

[half cove]

it should have options for cpu gpu0 gpu1 etc like this

[pale loom]

yeah it does but only when i use the nvidia one idk why

[half cove]

how the fuck

😭

the nvida one has a dropdown

u sure you using it?

[zealous sun]

Как найти голос Тайлера дердена?

[small umbra]

Searching for Voice Models 🔍

https://discord.com/channels/1159260121998827560/1175430844685484042

/search https://discord.com/channels/1159260121998827560/1163592055830880266
or take a look at weights.gg

do not use `/search` outside of #search-models

Commission a Custom Model 👑

https://discord.com/channels/1159260121998827560/1191429836321849435

cant find the model you want?
you can commission one from a Model Master 👑 https://discord.com/channels/1159260121998827560/1159289738314919936

[half cove]

@zealous sun ^

[thin stump]

Does kim vocals 2 introduce white noise or something? I can't find where someone talked about that here

oh that was inst hq 3

[finite prawn]

Will a microphone's idle background noise hurt the quality of a dataset?

not like metal pipes and shit banging in the background but just that really quiet static that microphones pick up? example like https://voca.ro/17qG65It5j1T

[thin stump]

I can hear it clearly with my headset. I mean the noise

[finite prawn]

i have a really bad mic lol

imma get my phone for a better example

[thin stump]

you just need to denoise the audio

unless someone else can chime in, i'm not a model maker

[finite prawn]

actually nvm ill just noise reduce yeah

[rocky bluff]

Has anyone figured out how to pass the sound to discord through voicemeeter banana?

[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

[round token]

can someone help me download

[violet heron]

Download what?

[round token]

the voice chager thing idk how to im lost

[rare goblet]

Ayo? @round token level 1 !!!

[violet heron]

Realtime or for covers?

[round token]

idk honesty im tryna open the file that says start https and it wont open

even after i press start anyway

[brittle wing]

/realtime

[edgy ledge]

How do I make the create song button appear/get rid of server not running

[rapid mist]

Hey

[violet heron]

Haven’t seen replay in a long time

[rare goblet]

Ayo? @violet heron level 93 !!!

[edgy ledge]

What do you reccomend

Replay is so quick and simple to set up im surprised so few people use it so im wondering what the problem is

[violet heron]

I personally use mainline rvc

Harder to use but works

Not really much of a problem with it

[thin stump]

lack of visibility

[edgy ledge]

https://nohello.net/en/

[thin stump]

like the only way to find out it exists is joining their discord

[median monolith]

What index settings do? What happens when i lower or "enlarge" it?

The search feature ratio more speciffically

[half cove]

It changes how much the .index file affects the infer

Index is the accents

Too high will lead to artifacts in audio

[violet heron]

0.66 is what I do

Index acts like accent

[tranquil fiber]

When I try using the voice models on colab it says "skipping voice model" anybody knows why?

[swift radish]

i need help with pytorch

it says it cant find a version that satisfies the requirement

[median monolith]

In the case that the character doesnt say anything and the samples are just sounds of him, i guess this doesnt do almost anything besides possible sound artifacts, right?

Also, idk what is "infer"

There is actually a lot of terms used in the creation of ai models, covers and such that idk what mean, and i wanted to know if there is a place where i could learn these without having to ask every single time to someone something that idk what it is or how it works. Maybe something in the guides channel?

[terse turret]

hello

i have a audio

of a voice

and is 4 minutes

its that good to make a ai voice?

[violet heron]

Good enough

[terse turret]

oh alr

should i make it on kits ai?

or RVC

[half cove]

I guess

[terse turret]

because i want it to sound good

[half cove]

When it converts the audio

[violet heron]

DO NOT

[violet heron]

RVC 100%

[terse turret]

oh

oh alr what the actual rvc?

because i see alot of rvc

the 2 or 1

[terse turret]

whats the actual rvc?

the 2 or 1?

[violet heron]

2 or 1?

[terse turret]

rvc 2 or rvc

[rare goblet]

Ayo? @terse turret level 2 !!!

[violet heron]

-local

[azure marsh]

Suggestions for @terse turret

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

[violet heron]

Original RVC

if you have a strong enough NVidia GPU

[terse turret]

oh alr

oh

how i check that?

[violet heron]

Task manager, go to performance

Click GPU 1, if there’s not 1 click GPU 0

In the corner is your GPU

[terse turret]

is a nvidia geforce GTX 1050 ti

[violet heron]

Would that be good enough for training locally

I guess that would be the minimum

How much vRAM does it have

[terse turret]

vram?

[violet heron]

Lemme just google it

4GB

not enough for local training

[terse turret]

so which one i use

if i cant rvc

[violet heron]

You can use the RVC colab for training

-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

[violet heron]

“RVC Disconnected”

That one is for training

[terse turret]

so i train the ai

then i can use it?

[violet heron]

Yes

[terse turret]

oh alr

[violet heron]

You can use RVC locally for inference

[terse turret]

i hope theres a toturial

[violet heron]

https://docs.google.com/document/d/1XuxQYiqEhYrdYeCZRRLrmV_ciMKo0bV-jTCGHu_-5Cc/edit?usp=drivesdk

[terse turret]

thanks

also

the voice i have

is from Ruben sim

if you know him

oh this is bad

oh wait hold on

there some errors

and i put the name right

oh wait

experiment name

[brittle wing]

I need some assistance

[terse turret]

it says it cant find rubensim.zip when its there

[granite loom]

What program should I download to check RVC?

[brittle wing]

So, my epochs just finished on easy GUI ver. 1/16/24 and I need help saving my model

[granite loom]

Models are certainly good, but what program is best to use to check and try them?

[brittle wing]

So did it save automatically?

Because I got some new pth files in my Google drive folder?

Nevermind, I think I found them

[rare goblet]

Ayo? @brittle wing level 3 !!!

[terse turret]

hey i have a question, i hope you can help me

[rare goblet]

Ayo? @terse turret level 3 !!!

[terse turret]

so i put the name of the file

[violet heron]

Change the name to be all lowercase

[violet heron]

RVC

[terse turret]

so i put Ruben_sim right

[violet heron]

No I’m saying on drive rename Ruben_sim to be ruben_sim

[terse turret]

oh alr

there we go

thanks

[brittle wing]

Sure, ask away

[terse turret]

i already fix it

[brittle wing]

Happy to hear that and Happy Lunar Year guys 🎈🎉

[terse turret]

how i check dataset lenght?

its how long the audio is?

so if the audio is 4 minutes

i put 4000?

[brittle wing]

4 minutes is perfect

[terse turret]

so i put 4000 at total_epochs

[brittle wing]

Use 300-400

[terse turret]

or alr

thanks

btw the video duration is 4:43

[brittle wing]

You don't want to over train your model, trust me I learned the hard way with a dataset containing 3 files xD

[terse turret]

you learn from your mistakes right?

what

oh i think i know the error

[terse turret]

nope it didnt fixed

[modern surge]

Does anyone know of any way to change the tones of an AI? In an old colab when applying the audio it let me down or raise the tones so that it sounds sharper or more serious

[modern surge]

Depends on the content of a dataset

[terse turret]

thats probably the error

how i check the dataset?

lenght

[modern surge]

The error is that you missed a step or did not put the dataset well to be processed

Read the bottom part of the photo

[terse turret]

a file is missing

[modern surge]

@violet heron ?

[violet heron]

Huh?

Never remembered that being a option

[terse turret]

hey timbo how i check the dataset lenght?

[violet heron]

You mean pitch/transpose which changed how high/low the voice is

[modern surge]

yes it was, but it had an explanation that was in case the voice was of a man or woman of the audio

[violet heron]

Open the audio file see how long it is

[terse turret]

is 4:43

[violet heron]

Yeah that’s transpose

[violet heron]

That’s the dataset length then

[modern surge]

still exists in the colab?

[terse turret]

so i put 443?

[violet heron]

Depends what colab

[modern surge]

Pitch_change pitch_change_all

[violet heron]

Yes

[modern surge]

the difference?

And how do I configure for bass tones, -1, -2?

Bass tones

Omg

The translator

[violet heron]

I would just experiment

[terse turret]

so i put 4:43

[modern surge]

Okok

[terse turret]

let me try

[modern surge]

Emm no

Just put 400 and if you know how to detect when it is overtrained only for the epoch it stayed

[terse turret]

oh alr

its still showing the error

i think a file is missing

[modern surge]

Have you started the previous cells yet?

Please see again the tutorial to make a model

[terse turret]

yes i did

[rare goblet]

Ayo? @terse turret level 4 !!!

[modern surge]

Make sure your dataset is well processed

[terse turret]

i loaded the dataset

feature extraction

and save preprocessed dataset files to google drive

what im missing

[agile ledge]

bro is making a model with a 3 second clip paid?

[terse turret]

is 4:43 minutes

do i need to load the preprocessed dataset files?

@violet heron do i load this thing?

nope it didnt fixed it

@violet heron ?

[violet heron]

No

[terse turret]

oh alr

i dont know what to do

i already loaded everything that shows the guide

[violet heron]

Did you run the first preprocessing cells?

[terse turret]

if you are talking about this yes

[violet heron]

What did you put for your training options?

[terse turret]

this

i didnt changed the batch_sizes tho

[violet heron]

Change save to be 50

What’s the training variables

[terse turret]

these?

[violet heron]

Yes

Can you show all of them

[terse turret]

ok

[violet heron]

What about the once above it

[wet pewter]

whats epoch

[violet heron]

No the other one

[terse turret]

these?

[violet heron]

No

The one where you select RVMPE

[terse turret]

oh alr

these?

oh wait

im very sorry if im annoying you @violet heron

[violet heron]

Turn off new pretrain

Rerun it t

Then try training

[terse turret]

whre is pretrain?

oh wait

[terse turret]

oh alr

nope same error

i think its something about a file

uhmm

3_feature768

do you think its because a file?

@violet heron

[rare goblet]

Ayo? @terse turret level 5 !!!

[violet heron]

Try repreprocssing the dataset

[terse turret]

ok

so i gotta do everything again?

like i press the button

and i have to also load the feature thingy?

[violet heron]

Yes

Wait

[terse turret]

oh alr

[violet heron]

Just the feature extraction

[terse turret]

oh alr

i hope this works

[violet heron]

Then try if not try restarting your runtime and try again

Idk

[terse turret]

YEEES

YESS

WOOHO!!!

IT WORKED

FINALLY

😐

also feature extraction is working

[violet heron]

Try redoing feature extraction and then doing that

If not restart your runtime enter the training variables and run it

[terse turret]

i did that but i didnt worked

so i need to restart my runtime

which is at the top right?

where is the runtime

[violet heron]

Yes

Rerun the first preprocessing cells

Wait

[violet heron]

Before you do it

K

[terse turret]

that?

[violet heron]

Wait

Did you train your model already?

[terse turret]

yeah it trained

[violet heron]

Before you delete your runtime download it

[violet heron]

Did you download it yet

[terse turret]

i need to run index

to download it i think

[violet heron]

No

I’m saying your models .pth

In the corner bottom left on colab iirc there’s a file button

[violet heron]

No

[terse turret]

oh i download the rvcdisconnected folder

[violet heron]

No

[terse turret]

nevermind

[violet heron]

Open the Mangio-rvc folder

Open weights

Download the model file with your models name

Modelname.pth

Download that

That’s your .pth file

[terse turret]

uhh

it doesnt appear

[terse turret]

it doesnt appear

[violet heron]

What’s in the weights folder?

[terse turret]

theres a folder with the name of my model

[violet heron]

Send a screenshot

[violet heron]

No on google colab

Not on your drive

[terse turret]

this?

[terse turret]

that one

[violet heron]

No

On colab

There is a little file button

[terse turret]

im on colab

OOOHH

now i see the weights folder

[violet heron]

Your model should be in there

If you didn’t delete your runtime

[terse turret]

i didnt

[violet heron]

Under whateveryounamedyourmodel.pth

[terse turret]

my model isnt appearing

[violet heron]

You sure you didn’t delete or restart your runtime

Also 1 little thing

Uhh

https://www.weights.gg/models/clm72kgyx00y2cctc13bwex2v a Ruben sim model already exists

[terse turret]

yeah i didnt

welp

what do i do

[violet heron]

Guess your going have to restart

[violet heron]

Or use this model

[terse turret]

welp

[violet heron]

Since that’s the same person your trying to train a model on right ?

[terse turret]

yeah

but this time

its from a call

here

[violet heron]

copyright

Delete it

[terse turret]

ok

wdym by copyright also

[gusty wyvern]

it just closes instantly

and i tried the command line

nothing happens

[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

[lusty pivot]

not working

how do i fixd it/

[livid marsh]

Is there an Ennard model?

[lusty pivot]

wdm?

I can hear the voice but it is not working in any apps

[livid marsh]

Nvm..

[stuck bane]

hey guyssss if anyone used applio rvc v3 yet can you please let me know if there is AUTOMATIC Over Training Detection?

[queen hearth]

hey guys, is it possible i could make my own AI voice? (like my own voice), im not the best singer right now and im wondering what AI can do.

[glad zealot]

yes

you use rvc for that

[queen hearth]

Thank you! Could you please link me a place where i could learn how to make it from my vocal and what do i have to record or whatever to do that :)

[hoary lance]

What's the best software or way to create AI song covers? I heard RVC WebUI and was wondering if that is the best free method where you can use it on your own local machine without having to pay online membership to a 3rd party site.

[brittle wing]

why isnt my virtual audio cable working

[hollow galleon]

I have a question can the new apillo run on any pc?

[brittle wing]

why isnt my virtual audio cable working

why isnt my virtual audio cable working

why isnt my virtual audio cable working

why isnt my virtual audio cable working

why isnt my virtual audio cable working

why isnt my virtual audio cable working

[safe adder]

i need help whenever i talk on my okada it sounds rlly muffled

[rare goblet]

Ayo? @safe adder level 1 !!!

[vapid dirge]

i also need help when i press http_start it just instantly closes

and yes i downloaded python and pytorch

[rapid raptor]

hey yall, uh

the /search command in search-models channel isnt working for me

[sacred estuary]

Anybody encountering a no such file or directory error with Applio?

[placid holly]

about Applio v3 (latest one)

did the training improve on all methods? like Mangio Crepe or just RMVPE?

[shadow silo]

How do i use https://discord.com/channels/1159260121998827560/1195329058255409223?

[hazy crane]

hello guys im new to this

can someone explain to me how can i use these ai models

i downloaded one recently but then what do i do

[civic meteor]

why is cable OUTPUT set to INPUT

[brittle wing]

are you using Mangio or Applio?

[shadow silo]

idk

i'm new to doing this

like

i use kits.ai

[crimson owl]

Ello why does in the colab ai voicechanger real time shows this to me?

what i should do?

[glad zealot]

you run out of gpu

[crimson owl]

what do i do now?

[glad zealot]

wait 12 hours or use a different account

[hollow galleon]

Can I run the Apollo on any pc?

[crimson owl]

i alrd used like 20 acc

[glad zealot]

wtf

[glad zealot]

as long as it has gpu i think

[crimson owl]

dint know what to do when what message appeared

[hollow galleon]

Ofc

It's navida

[glad zealot]

yup

[crimson owl]

so i decided swith and change acc

and create new ones XD

[glad zealot]

make sure you are logged in to the correct account

[hollow galleon]

I wish the apolo can run without CPU for people without a powerful CPU

[rare goblet]

Ayo? @hollow galleon level 1 !!!

[crimson owl]

so if that appears i just need wait 12 hours?

[glad zealot]

yup

or switch acc to one that can still use gpu

[crimson owl]

check. i alr have like 10 acc so i will have like 4 or 5 without the 12 hours cd

[rare goblet]

Ayo? @crimson owl level 2 !!!

[glad zealot]

you can use colab

[crimson owl]

switching to the main one XD

[hollow galleon]

How

[glad zealot]

you can only do around 4 hours of usage on each account

[glad zealot]

#📰│dev-updates message

[hollow galleon]

But is the collab fixed at least the bugs and everything

[crimson owl]

i used yesterday like 6 accounts

[crimson owl]

also btw

[glad zealot]

wut

[crimson owl]

sometimes my friends tell mes, that my voice when i talk it is slurred when speaking and sometimes it is very difficult to hear

in discord

but in games or some else i hear it good

[glad zealot]

no clue, could be different things like model, your setting, you cable buggng out

[crimson owl]

maybe

[crimson owl]

usually my mic is a little bad and it does like background sound (static and those)

[glad zealot]

could be the cable you are using or you have other electronics near your mic like your phone

[crimson owl]

my phone is dead xd

no baterry rn

and the only thing near my mic is the vr headset,

[brittle wing]

buy a better mic

[covert violet]

it doesnt work for me when i out the ai voice on the thing

[verbal palm]

I'm training a model for the first time so i'm sorry if this sounds like a stupid question but i have a question about performance:

I'm training it locally with a 25min file and a batch size of 4, it took almost 40 minutes to generate 5 epochs.. Is that really bad/normal? Should I stop it and try something different?

[glad zealot]

depends on your gpu

[verbal palm]

It's grx 1650 4gb

Gtx**

[glad zealot]

i think thats normal then

not sure about local

[sand iris]

4gb vram

https://tenor.com/view/joe-hawley-joe-hawley-its-joever-joever-gif-18282355701517291166

[brittle wing]

Any AMD GPU Trainers here? Is there even any RVC fork that supports DirectML training?

[glad zealot]

mainline does right?

[brittle wing]

Do they? 🤔

afaik they only support voice inferring

Training not currently supported, will switch to CPU training

[rare goblet]

Ayo? @brittle wing level 4 !!!

[sand iris]

On linux yeah

DirectML can't train though

[brittle wing]

Can't at all or just on linux?

[sand iris]

Can't at all with DirectML. At the moment AMD has something 1:1 to the CUDA backend of torch but only on linux for the time being

[brittle wing]

Alright, thank you

[forest sluice]

rocm is what you're looking for. DirectML is dedicated to DirectX, so obviously it isn't supported

[fringe crow]

Latest Applio has no audio seperator?

[sand iris]

Unrelated, but what the fuck has Microsoft done with DirectML

Launched it, claimed it would be the one thing people would use and now it hasn't seen much activity in like half a year

[low berry]

brrr

everytime I try to use the voice changer its insanely delayed and the voice is so laggy 😭

[glad zealot]

you have a gpu?

[low berry]

yeee

its a gtx 1650 super i think

[shadow silo]

i'm not giving up mr saleh

[glad zealot]

try this setting ig

[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

[low berry]

ooo lemme try

also crepe is jus N/A for me

[carmine osprey]

why my model is getting trained so fast today?

[glad zealot]

¯_(ツ)_/¯

dataset?

[carmine osprey]

its 20min dataset

100 epoch in 2min

[glad zealot]

ye that should be fairly fast

[glad zealot]

NVM

[carmine osprey]

is something wrong?

[glad zealot]

not sure but do test the model if its okay

[glad zealot]

damn 1 sec/epoch

[rare goblet]

Ayo? @glad zealot level 56 !!!

[carmine osprey]

yeah

[rare goblet]

Ayo? @carmine osprey level 5 !!!

[carmine osprey]

all cells runnin so fast

rn

i dunno why

[glad zealot]

ye not sure too

test the models tho

[carmine osprey]

@glad zealot

[glad zealot]

no clue i havent used that thing in a while

[half cove]

Make sure the name that you set in set training variables is the same as your zip folder

[carmine osprey]

it wasnt same while training

[half cove]

Hm?

[cold raptor]

thats why

[carmine osprey]

zip and the name were diff

now this coming

when i chnaged the name

[half cove]

Reset runtime then train index lol

And make sure the names are the same

[carmine osprey]

did already

not wokring

[half cove]

Wat

[carmine osprey]

can you join vc

ill screenshar

share

[half cove]

Did you run the other cells?

[half cove]

Not rn

[carmine osprey]

set var

than train

but error

ahh wait lemme try again

[half cove]

Run the dependencies then set training then feature extraction then index

[carmine osprey]

ok

[half cove]

Make sure the set training name is same as dataset folder

[carmine osprey]

first runtime restart

right?

[half cove]

Delete runtime but if you trained a model first make sure you save it first

[carmine osprey]

ok

the file i uploded was orginally named Gabriel (20Min).zip

which i need to enter in set var

[carmine osprey]

i did all steps as u said

still same error

[half cove]

Try ask on help forum ig

On mobile so I can't help much

[forest sluice]

UVR modules in mainline/Mangio sucked balls

You better off use Karafan, MDX-Net Colab, or mvsep.com/proper UVR5

-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
• Creating Datasets for RVC using iZotope RX, 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

[rain sigil]

Hi all! I'm using a ready-made voice model from weights.gg via Mangio-RVC, but I get consistently bad results. Does anyone have any information about the best recommendations for improving the quality of the cover?

[crimson kettle]

Why does the audio sound so much worse on discord than on the vocie changer client app?

[proper shale]

It could be the Virtual Audio Cable maybe

Or maybe just Discord's voice settings

Because it has compression and all that stuff

Unless it's like choppy anything else is normal

[crimson kettle]

ok thanks, i turned input sensitivity down to -100 and that seemeed to do the trick. One more thing, I find that when I first start to use the voice changer after booting up my pc, the lag is very short and the audio is clear, but the longer i keep it open for and the more I switch to different voice models, the laggier and worse it becomes. This continues until I reboot my pc.

[proper shale]

I think this might be normal but I'm not entirely sure

So uh, dunno how to fix it

[spare shale]

-help

[azure marsh]

Command has expired

[crimson kettle]

its fine thanks anyway

[proper shale]

If you have any more questions lmk

[lusty pivot]

where do i find best female ai voice model !

[proper shale]

...#1175430844685484042

[hoary estuary]

-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

[dusk ether]

how do i create model

[proper shale]

-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)

[proper shale]

^^^

[olive mirage]

Hey sorry to bother you peeps, I would like to make ai narrations and/or covers but when I try to use the google collab it pops up with this warning, is it going to delete my account or something like that If I use this? I super tired and cant' figure out If I'm doing something wrong.

"Warning
You may be executing code that is disallowed which may terminate your runtime without warning. Colab prioritizes interactive notebook compute and disallows some types of usage when executing code without compute units as outlined in the FAQ.

Your compute unit balance is 0. Purchase more"

[proper shale]

-colab

[azure marsh]

Suggestions for @olive mirage

☁️ 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]

Use these Colabs/Huggingface Spaces

I'll send you a guide of Ilaria RVC, which is a good inferencing (audio converting) Space

https://www.rentry.org/ilarvc_inf_guide

[olive mirage]

Thank you heaps, much appreciated

[rare goblet]

Ayo? @olive mirage level 1 !!!

[proper shale]

You're welcome :)

[tacit mesa]

What should I put as the tune for angel dust if I'm female? Because it doesn't sound right for me.

[rare goblet]

Ayo? @tacit mesa level 1 !!!

[proper shale]

If the voice is deeper than yours decrease, otherwise increase

Doesn't need to be exactly 12 or -12 too

[tacit mesa]

Thank you!

[proper shale]

You're welcome :)

[covert verge]

can i use RVC On AMD PC

[proper shale]

If you have an AMD GPU, you will only be able to do the inferencing or real time voice changing

No training

[covert verge]

yea i want to use it

but it sucks

[proper shale]

oh wait i just saw

[covert verge]

I Want to do exactly that but it sucks how can i download it

[rare goblet]

Ayo? @covert verge level 1 !!!

[proper shale]

you don't have a dedicated GPU?

[covert verge]

nah bro on vega 7

[proper shale]

and also, exactly what

[covert verge]

igpu

[proper shale]

that's integrated

[covert verge]

Realtime voice changing ?

[proper shale]

it won't be able to run realtime voice changing technically

[covert verge]

it works but too much latency like a whole second or two

[proper shale]

u can use Colab for a better experience, check the pins in #🔍│help-w-okada

[proper shale]

yeah it's gonna have bad latency

cuz cpu

[covert verge]

what can i do in that case

[proper shale]

_ _

[covert verge]

ooo

[proper shale]

theres a link for the colab notebook there

[glad zealot]

fck

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

[proper shale]

thanks hina

[covert verge]

btw? will it have lesser latency

[glad zealot]

on good internet speed lowest is 256ms

less than a sec basically

[proper shale]

^

[covert verge]

oki tysmm

[proper shale]

you're welcome :)

let us know of any errors in #🔍│help-w-okada

[hollow sleet]

Yo does anyone know why it doesnt give out anything in the output of training your model even do i did it the right way

[glad zealot]

it need to train to whatever epoch count you set it before it starts saving/making the pth

[hollow sleet]

Im on step 2a which is me putting in the audio and supposed to be decoded into audio

[proper shale]

do you mean preprocessing

[hollow sleet]

i can send a screenshot of what i mean

in dms

[pearl current]

Hello ! Do you know of any good, stable, multilingual model to test?

[proper shale]

make sure that the path to the dataset has no spaces, also the name of the model must have no spaces

[proper shale]

u can send here once u reach lvl 1

[proper shale]

no sorr

[pearl current]

ok thanks anyway

[hollow sleet]

look dms.

[proper shale]

i have seen

[proper shale]

_ _

[hollow sleet]

Yes do you get what i mean

[rare goblet]

Ayo? @hollow sleet level 1 !!!

[proper shale]

Yeah

[hollow sleet]

and i got the folder with my audio

in mp3

still does not let me process

[proper shale]

check cmd for any Tracebacks

[pearl current]

Are there any basic fine-tuning models to avoid the 2-hour wait?