✨│ai-help

[tidal trail]

refresh and select the model and index

sometimes does not appear automatically

[fiery raptor]

as in the button wont press

[tidal trail]

mmm weird

[fiery raptor]

ikr

restarted applio

it works now

[odd shale]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[shell gulch]

need a rvc with gui please

[rare goblet]

Ayo? @shell gulch level 2 !!!

[shell gulch]

forgot how do to it lol

[violet heron]

-local

[azure marsh]

Suggestions for @shell gulch

🖥️ Local stuff

• 🍏 Applio, 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
• GPT-SoVITS, by RVC-Boss GitHub

Credits to Faze Masta and Antasma for compiling these links.

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

[steady ferry]

#✨│ai-help I've been trying to install AICoverGen but it keeps giving me this error
`` building 'fairseq.libbleu' extension
creating build\temp.win-amd64-cpython-310
creating build\temp.win-amd64-cpython-310\Release
creating build\temp.win-amd64-cpython-310\Release\fairseq
creating build\temp.win-amd64-cpython-310\Release\fairseq\clib
creating build\temp.win-amd64-cpython-310\Release\fairseq\clib\libbleu
"C:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.40.33807\bin\HostX86\x64\cl.exe" /c /nologo /O2 /W3 /GL /DNDEBUG /MD -IC:\Users\JoshPC\AppData\Local\Programs\Python\Python310\include -IC:\Users\JoshPC\AppData\Local\Programs\Python\Python310\Include "-IC:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.40.33807\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.40.33807\ATLMFC\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Auxiliary\VS\include" /EHsc /Tpfairseq/clib/libbleu/libbleu.cpp /Fobuild\temp.win-amd64-cpython-310\Release\fairseq/clib/libbleu/libbleu.obj -std=c++11 -O3
cl : Command line warning D9002 : ignoring unknown option '-std=c++11'
cl : Command line warning D9002 : ignoring unknown option '-O3'
libbleu.cpp
C:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.40.33807\include\yvals.h(20): fatal error C1083: Cannot open include file: 'crtdbg.h': No such file or directory
error: command 'C:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.40.33807\bin\HostX86\x64\cl.exe' failed with exit code 2
[end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for fairseq
Failed to build fairseq
ERROR: Could not build wheels for fairseq, which is required to install pyproject.toml-based projects``

i've tried multiple versions of visual studio (2019 and 2022) and python (3.9.0 and 3.10.10) but the error always remained the same.
im currently using visual studio build tools 2022 with desktop development with c++ already selected and with python 3.10.10 installed and included to the user system paths
Sox-14-4-2 and FFmpeg are also added to the user and system paths
any ideas on what is happening? i've been trying to fix this for months but I don't know much about this stuff

[rare goblet]

Ayo? @steady ferry level 1 !!!

[noble flare]

when i download the okada voice changer it goes on windows 32 is that normal

[cosmic epoch]

what does this error mean? "/content/RVC

Traceback (most recent call last):
File "/content/RVC/demo.py", line 1, in <module>
from original import *
File "/content/RVC/original.py", line 3, in <module>
from dotenv import load_dotenv
ModuleNotFoundError: No module named 'dotenv'
it happens to me on this colab https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw#scrollTo=DZDKirCM0F9g

[tidal trail]

easygui is broken

[tidal trail]

-colab

[azure marsh]

Suggestions for @cosmic epoch

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[tidal trail]

use one of those, guides: https://docs.aihub.wtf/

@cosmic epoch

[fossil vigil]

hey why isnt my voice working in voice calls

i have my input on the voice changer as my mic

and output as vb audio cable

and on apps i have my mic as vb audio cable and output as speakers

but it doesnt work

it worked a few seconds ago

[vivid pewter]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[fossil vigil]

Should this help or is it not for me

[vivid pewter]

???

what

[fossil vigil]

Sorry

I meant as in for my issue I sent above

[visual elbow]

I have a problem :c!! please idk what else do to I try to select the audio input device and it sends me to this error..

Looks like there's a bit of a problem.
unknwon message

If you clear the information being managed by this app, it may be recoverable.

Initialize

Reload without initialize

Error
unhandledrejection
no error stack
NotReadableError: Could not start audio source
no stack

[crystal lagoon]

Traceback (most recent call last):
File "G:@ai\Applio-3.2.3\app.py", line 9, in <module>
from tabs.inference.inference import inference_tab
File "G:@ai\Applio-3.2.3\tabs\inference\inference.py", line 8, in <module>
from core import (
File "G:@ai\Applio-3.2.3\core.py", line 22, in <module>
from rvc.lib.tools.model_download import model_download_pipeline
File "G:@ai\Applio-3.2.3\rvc\lib\tools\model_download.py", line 9, in <module>
from bs4 import BeautifulSoup
ModuleNotFoundError: No module named 'bs4'

Press any key to continue . . .

Anyone got any ideas how to fix?

[quaint nacelle]

Is there any collab with rvc v2 I where where I can edit the hz low and highs? I can’t find it anymore 😦

[cerulean lagoon]

is their any way i can just download the rvc and not do colab

colab doesnt alow you to use the model for long

[spring orchid]

download local applio

use gpu to train/infer

[cerulean lagoon]

how do you do that

ive been on the github but the only download link is for old versions

[spring orchid]

https://github.com/IAHispano/Applio/releases

[dusky quartz]

hey i had an little problem with RVC v2 Disconnected, you see i was making a model, until right the program is about to closed it says that my runtime got disconnected even though i active the Disconnected prevention method, so i'm not sure if it is a problem to someone or something, but sombody please fix the error, does this happen to anyone?

i click the play button that you see on an screen

[rare goblet]

Ayo? @dusky quartz level 2 !!!

[dusky quartz]

also is there anyone, that can make the model train go faster in RVC v2 Disconnected?

[patent acorn]

im really new to this could someone assist me with the prerequisites i need for this

[thin stump]

Depends on what you need https://docs.aihub.wtf/

[thin stump]

I recommend this instead https://rentry.co/RVC-Mainline-Kaggle

[patent acorn]

im trying to catch pedos on discord, im watching a video by duckus on yt and like more than half things hes talking about makes no sense

[open stag]

• colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[dusky quartz]

umm, not that one, something it doesn't have with google colab?

also not with kangle

[patent acorn]

can i go on vc with someone, i dont understand this at all

[rare goblet]

Ayo? @patent acorn level 1 !!!

[pastel oak]

Correct because the video is quite outdated and misinformed

https://rentry.co/ForkVoiceChangerGuide read this

[patent acorn]

im good now, i have a shitty laptop so i cant even do it yet

[pastel oak]

This is a rare issue, this can happen when:

• your mic is being accessed by another windows device
• could be your mic's sample rate

Can you provide a screenshot of the console logs when this error happens? Send it in #🔍│help-w-okada instead.

[pastel oak]

Could be a few reasons, can you check these solutions:

• Select Start (on windows) ,then select Settings > Privacy > Microphone . Select Change, then turn on Allow apps to access your microphone (YOUR NORMAL MICROPHONE)
• If this didnt work, can you install the other virtual cable? Please read "virtual audio cable" step on the guide https://rentry.co/ForkVoiceChangerGuide#virtual-audio-cable and use this instead of the VB Cable. Its called Line 1 on the app

[knotty moth]

I wish those outdated videos could be labeled as "misinformation"
(YT has been testing twitter-style community notes for a few select channels/videos)

[white owl]

i havnt used this in a long time and forgot how to start it can anyone help?

[hazy holly]

hello!, quick question!
how do i update the w-okada changer? i have version 1.5.3.18a but i feel it's outdated, as it has a LOT of delay and it wont even let me pick my GPU

(my delay is of almost 6 seconds, even tho i have a ryzen 7 with a RTX 3060TI)

[pastel oak]

6 seconds delay on a 3060ti is not normal

That is the most recent wokada version but a new modified has been made, i use 3060ti with this modified and get less than half a second iirc

https://rentry.co/ForkVoiceChangerGuide

[pastel oak]

Voice model Training or inference, ai cover?

[white owl]

no clue what any of that means

ill install the link just sent and lyk if i need help from there

[pastel oak]

You didnt specify what you want to use so im guessing by naming a few things lol

That thing i sent is realtime voice changer if thats what ur after yeah

[hazy holly]

guess i download where it says "download nvidia"?

[pastel oak]

Yes

[hazy holly]

once i install it, will it remove the old version?

or do i need to uninstall it manually?

[pastel oak]

No, you need to delete the old folder, that takes care of it. But wait until this ones installed and running if you like

You can copy paste the models from old to new

By copy pasting over the "model_dir" folder

This new one runs on webui (on a browser), but if you use the voice-changer-native-client.exe from the old wokada, you can make it run as an app.

For this, you first run the modified version like it is instructed on the guide. After the webUI opens, you run the native-client.exe from the old wokada, and it will open the modified version as an app

I recommend webui more though

[dusky quartz]

well it turns out to be a google colab page instead, is there any other google colab ai projects that can train faster?

train ai voice models faster

[white owl]

i click the exe right?

[pastel oak]

Yep, that starts the voice changer

[white owl]

ok thank you

[rare goblet]

Ayo? @white owl level 2 !!!

[white owl]

it runs in my browser now?

[pastel oak]

Yes

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

Don't forget to thank Ilaria if you find it useful! 💖

[slim anchor]

what do i do next

[flint geyser]

did you put the model where its supposed to be?

[slim anchor]

i havent putten anything for now

just

getting all settings ready

[flint geyser]

o

[slim anchor]

or anything u guys recommend me to do

[flint geyser]

ill brb

shad'll help you

[pastel oak]

F0 det: rmvpe
Chunk: 96 (if laggy, raise it to 112)
Extra: 16384

If you want to use it on apps, you need to select "output" on the app as the virtual cable. Do you have virtual cable installed?

Tune is the pitch, so if you have male voice and using female voice, raise it to around 12 and depending on if you hit the right pitch, go up or lower

[slim anchor]

i can go upto 131072 and till 64 in low one

[pastel oak]

Sup 2 for noise suppression if its a problem
S. Threshold is noise gating, push to the right if noise sup gets picked up

[slim anchor]

no i havent but watching tuto he will teach

[pastel oak]

Dont trust youtube tutorials too much

Filled with misinformation, probably recommending you the bad virtual cable too

[slim anchor]

https://youtu.be/81KYc8AAmus?si=JgQup65iQW0e3SRb

[pastel oak]

Doesnt matter, original wokada has an Extra bug so you have to use a low one, like 16384

[slim anchor]

okkkk

[pastel oak]

There is a modified version thats being updated atm, if you want to try that out instead. Can reach lower latency and higher quality

[slim anchor]

sureee im happy to try

[pastel oak]

https://rentry.co/ForkVoiceChangerGuide

[slim anchor]

i download this both?

[rare goblet]

Ayo? @slim anchor level 2 !!!

[pastel oak]

Yes

[slim anchor]

should i delete this exisiting one

[pastel oak]

Dont delete it yet

[slim anchor]

okk

[knotty moth]

extract it first

[pastel oak]

Also i saw you asked about voice models, theres a few to try out at the bottom of the guide

[slim anchor]

its downloading for now

[plush compass]

Is it impossible to separate the guitar from UVR? If possible, what algorithm should I use?

[knotty moth]

extracting a guitar stem?

[plush compass]

YES

[knotty moth]

mvsep has some guitar model, but dango.ai (paid) seems the best so far
actually depend on if it's acoustic or electric one you want

[plush compass]

Is UVR's guitar extraction not very good?

It's an acoustic guitar.

[knotty moth]

no models for that in UVR, but yeah dango.ai is good for the acooustic

[plush compass]

Thank you. I will look for it.

[viscid palm]

can you name your audio files anything or do they have to be numbers?

[rare goblet]

Ayo? @viscid palm level 2 !!!

[pastel oak]

Wont know until you try:)

as long as its not special characters

[viscid palm]

Does that include underscores

[pastel oak]

Underscores work.

[slim anchor]

i did now what]

[pastel oak]

The guide explains the next steps

Run MMVCServerSIO.exe

[slim anchor]

yess i did it

[pastel oak]

Ok, did a webui open

[slim anchor]

audio cable what should i use

[pastel oak]

Wait that out until its done

[slim anchor]

okk

[pastel oak]

Download VAC Lite from the guide if you havent done that yet

[slim anchor]

whats VAC?

[pastel oak]

Virtual audio cable

[slim anchor]

okk

[pastel oak]

Then read the first part of Audio Setup to know what to use as input and output

[slim anchor]

its paid>

[pastel oak]

There is a download on the guide....

to the free version

[knotty moth]

download the lite version, not trial

[slim anchor]

ah

wait lemme

try to find

[pastel oak]

https://rentry.co/forkvoicechangerguide#virtual-audio-cable

[slim anchor]

this?

[pastel oak]

yes

[slim anchor]

i downloaded it

[umbral lantern]

Heyo @pastel oak can you point me in the best direction for someone whose tryna making translated audio with ai?

I've already isolated the voice of the person speaking from the background music and noise now i'm trying to convert the monologue to a different language

brazilian portuguese to be specific, cannot be portuguese from portugal because the accent is simply alien in contrast with brazilian portuguese

[slim anchor]

should i install @pastel oak

[pastel oak]

Yes

[slim anchor]

there are 3 setup

which one

[pastel oak]

I'd wait for someone else, I have not looked into that before

[knotty moth]

setup64

the vanilla is 32-bit and 64a is for some ARM laptop

[umbral lantern]

aight thanks for answering me

[slim anchor]

done

installed

and that app thing opened a website

[pastel oak]

Yes

Thats where the UI runs

[proper shale]

best strat is to clone the voice with rvc and do the dubbing part with your own voice

[knotty moth]

haven't seen any decent free AI dubbing

[slim anchor]

okayy great

now what do i do

[proper shale]

if you only wanna do it with open source tools ye

[pastel oak]

Upload voice model, then read the first part of Audio Setup, then read Best Settings

[knotty moth]

also most likely dubbing skill issue, and AI can't completely replace localization with just typical machine translation

[umbral lantern]

i came to that conclusion after using 11labs and the result was, a cloned voice but the accent was simply impossible

[rare goblet]

Ayo? @umbral lantern level 1 !!!

[umbral lantern]

aight, onward to stack epochs then

[proper shale]

🙏

lmk how that goes

[slim anchor]

is this good enough? @pastel oak

[rare goblet]

Ayo? @slim anchor level 3 !!!

[pastel oak]

So you will always want to look at that "perf" number on the left with the colors. If its red, raise the chunk above that number

Atm its 72, so raise it up to around 110

until its green

If you play a game, the number could raise up again because of resources used, the color will always tell you what to do

green - stable
yellow - sort of stable, but running something intensive in the game might become unstable
red - bad

i forgot you have 4070 laptop so the settings there might not be accurate for you, so 110 - 130 ms should do the trick for you

[pastel oak]

F0 det: rmvpe is better for nvidia

[slim anchor]

okk

[pastel oak]

If you have male voice and using girl model, then increase the pitch aswell to match the voice

[slim anchor]

how much should i increase

[pastel oak]

If you have a normal male voice, increase it to 12 ; then if the pitch is too high, lower it by a few numbers. This depends on your voice

Or other way around, if the voice is too low pitch, increase it

[slim anchor]

aw hell nahh

this shi too accurate

[pastel oak]

💀

[slim anchor]

😭

[pastel oak]

bro got excited

[slim anchor]

i think i might be in love with

my own voice

a

ai

[pastel oak]

LOL

Which voice model are you using?

[slim anchor]

the one which was recommended

accurate girl or something in guide

[pastel oak]

the natural girl one by mares

[slim anchor]

though bro voice is pretty low any way to increase the gain

[pastel oak]

that ones 💯

[pastel oak]

increase the "out" underneath the start stop button

the out is for what others hear, the mon would be for your own headphones

[slim anchor]

and in?

[pastel oak]

in is your own microphone when it goes inside the voice changer, before it puts out the ai voice

i would leave it at default 100%

Increasing this too much can cause background noise

[slim anchor]

okkk

[pastel oak]

sup2 is noise suppression if you need it

[slim anchor]

nah im using nvidia broadcast

[pastel oak]

in. sens. is a threshold of when the ai starts converting voice, if you have a lot of background noise you can put this to the right too

[slim anchor]

pretty ogod

good

[pastel oak]

ok

if you want to reduce more delay, you can try:

Audio: SERVER

[Wasapi] microphone
[Wasapi] Line 1
[Wasapi] headphones

There was a new update which makes using wasapi work without needing to change audio device settings, so you should be able to do what i told you right away

[pastel oak]

oh ye S.R. 48000

[slim anchor]

done bro

shad just wanna say im really thankfull to you

for teaching me this

[pastel oak]

❤️

[slim anchor]

its hard to learn everything through guide and yt isnt always best place so u taking some time off for me really means alot <3

if u find any cool models pls let me know :)

im gonna go troll some pepes on val

[pastel oak]

im gonna make a yt tutorial when im back in country

should be helpful i hope

[slim anchor]

yess

people would be glad :) send me i will support u with 5 alt accounts i hvae

[pastel oak]

hahahaha

[agile gale]

@pastel oak HWO DO I AJUST ECHOS

do i need a certain gui

[pastel oak]

Thats caused by your mic picking up your headphones

enable sup2, move s.threshold/in. sens. to the right, lower your headphones volume, move mic further away from headphones if possible

Else need a 3rd party program like steelseries sonar for noise gate and noise reduction

[agile gale]

no i mean like

[agile gale]

\

how do i change it to 1000 epochs for example

[pastel oak]

You can not adjust epochs of a voice model you downloaded from the internet if that is what you mean

That is how many cycles of training a voice model went through

1000 does not mean its better than 300

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[agile gale]

how do i

og im slow

[gray igloo]

hey i'm looking for an rvc that works so i can train a voice i've been searching for hours and everything i've tried isn't working well.. do you have any recommendations? something that's not local i hope i'm asking in the right place 🙏🏻

[pastel oak]

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)

[main lichen]

hello there

i have another prob with my vc changer

it kinda has headphone leak so whenever other people talk, they can hear themselves in my vc changer

[pastel oak]

Enable sup2, move s.threshold to the right, lower your headphones volume, move mic further away from headphones if possible

[main lichen]

i actually did that but it still happening, any other suggestions?

[pastel oak]

Are you using a headset

[main lichen]

yup

[pastel oak]

download Steelseries Sonar and enable noise gating and ai noise cancellation

[main lichen]

alr got it ill try it rn

[wary spoke]

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[main lichen]

alright heres what it looks like

[rare goblet]

Ayo? @main lichen level 7 !!!

[main lichen]

where do i go from now?

[pastel oak]

Select your own mic on the "Mic" thing

Then go to Mic section, enabling the AI feature is probably enough already but feel free to mess around with the noise gating

Then on the voice changer you select SteelSeries Sonar Microphone as the input instead of your normal mic

[main lichen]

oh okay

[main lichen]

Sorry but when my friend used the steel series, i can hear him through his stream and through his mic, any fixes?

[pastel oak]

Ohh yh steelseries kinda wack in that regard, this only happens if you discord screenshare all application

open sound devices and disable all the playback steelseries stuff

unfortunately steelseries always sets itself as default device when u reboot so its needed again and again xd i didnt look for a fix to automate process yet

[fiery raptor]

when youre trying to use ilaria rvc with no gpu

[pastel oak]

Why does it not have a GPU? Can you reload the page again then try

[fiery raptor]

it doesnt because according to hf i dont

[pastel oak]

Must be on your end, because it is using a GPU for me

Can you check what it says for you

[fiery raptor]

i dont have huggingface premium or wtvr

[fiery raptor]

well I didnt pay for any gpu so it makes sense

[pastel oak]

I dont have premium either, im not logged in either, this is free

This is for everyone to use

[fiery raptor]

really?

when i try it it says i need pro

[pastel oak]

@proven hill ilariaaaaa can you help 😭

Says he needs hugginface pro to use a gpu on huggingface

any fixes?

[fiery raptor]

nope

OHH

only if i duplicate the space

its like that

i get it

[pastel oak]

Just to make sure, you are using the right space right?
https://huggingface.co/spaces/TheStinger/Ilaria_RVC

[fiery raptor]

yes

i just duplicated it because thats how the old ilaria rvc used to be like

[pastel oak]

Ahhhh

So you got it sorted?

[fiery raptor]

yes

[pastel oak]

Aight cool, enjoy:)

[fiery raptor]

thanks

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

Don't forget to thank Ilaria if you find it useful! 💖

[sacred temple]

Can you help me?I have an error named:"Noontype"object has no attritube "Setdefault"

[low shard]

specify what you are using, like google colab link

you can also send screenshots here

[sacred temple]

Now it's not loading in.Please wait a moment.

[low shard]

what i meant is giving more info on what you are doing, be sure it isnt from a yt video, most yt tuts are outdated

[sacred temple]

It isn't.But the page is not loading.I use the ai generator named hugging face.

[low shard]

hugging face is an ai platform

Are you using Ilaria RVC Zero

Send the exact link you are using

[sacred temple]

hold on for a moment

BUT THE AI PAGE IS NOT LOADING

[pastel oak]

Copy paste the URL into this chat

[low shard]

just paste here the url

don't need to show a screenshot

[sacred temple]

https://huggingface.co/spaces/r3gm/AICoverGen do you mean this?

[rare goblet]

Ayo? @sacred temple level 2 !!!

[sacred temple]

The page is now logged in

[low shard]

yes

Did you duplicate the space? you shouldn't duplicate ZeroGPU spaces

[sacred temple]

Now it's starting downloading.

[low shard]

ZeroGPU is a paid gpu by the space owner

[sacred temple]

There was an error and it stoppeddownloading

[low shard]

I think its better you use Ilaria RVC Zero

that Hugging face space you sent seems broken

[sacred temple]

Okay but I guess I have to learn this.

I mean the llaria rvc zero

[low shard]

i sent you the guide, just read it yea

[sacred temple]

I'm reading it

[proven hill]

user error, again

[pastel oak]

yep

didnt expect people to just duplicate the space when its right there working but they were used to it

[sacred temple]

Thank you for helping.

[proven hill]

old space worked like this

[low shard]

yw

[lime smelt]

I'm trying to get a Cheap card for Dual GPU for Dedicated RVC use & Dedicated Gaming / Streaming use for each for RVC's minimal latency.
I'm thinking 2060, What do you guys think?
Do you have a better suggestion for the Budget Dedicated RVC use-GPU?

[jaunty shale]

So I usually have problem with converting piece of singing audio into a specific voice. The real audio contains growly voice sometimes while singing, but its still sounds monotone'ish either way.

Everytime I use any voice model for that, it makes an artifact, because of that part in the original audio vocal. Is there any way to teak some setting in Applio?

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

RTX 40-series: 80-96 chunk | +16384 extra
RTX 30-series: 96-112 chunk | +16384 extra
RTX 20-series: 112-128 chunk | +16384 extra
GTX 16-series: 128-192 chunk | +8192 extra
GTX 10-series: 128-192 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 💖

[mighty atlas]

How to make your own voice model ?

[proper shale]

https://docs.aihub.wtf/

Has everything you need to know

[empty kestrel]

when i click gpu the only option i can choose is cpu and my graphics card doesnt show on it

anyone know how to fix this

[proper shale]

You may have gotten the wrong version of the voice changer for your GPU

https://rentry.co/forkvoicechangerguide

Use this guide ^

[empty kestrel]

ty

[steel cloud]

my applio doesnt save a .pth file

[odd shale]

Do you want to make a model of someone/a character or yourself?

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

RTX 40-series: 80-96 chunk | +16384 extra
RTX 30-series: 96-112 chunk | +16384 extra
RTX 20-series: 112-128 chunk | +16384 extra
GTX 16-series: 128-192 chunk | +8192 extra
GTX 10-series: 128-192 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 💖

[tawny knoll]

could someone help me? Now that I have a graph I would like to try creating the AI ​​voices locally, (only voices, without music, it is not a cover)

[proper shale]

http://docs.aihub.wtf/rvc/local/mainline/#training-

Should help

[odd shale]

Read the docs buddy.

MJ provided you a link.

[tawny knoll]

i´m sorry, thaks

[feral pilot]

when i press start my original voice comes up

i have a voice changer that makes me sound like a boy on and it doesnt work

[pastel oak]

Is passthru green or red

[feral pilot]

red

[pastel oak]

make it green

red puts your normal voice through, green voice changer

[feral pilot]

now i cant hear nothing

[pastel oak]

Whats your input, output on wokada and input output on discord

u cant send pics yet so have to type it

[feral pilot]

my input

are

the headpohgnes that im using

same with the output

[pastel oak]

Do u have Virtual cable installed

[feral pilot]

I do

[pastel oak]

On the voice changer app:

• Input: Your microphone
• Output: Line 1 Virtual Cable
• Monitor (if you wish to hear the voice changer yourself aswell): Your headphones

On discord and games, you do:

• Input: Line 1 Virtual Cable
• Output: Your headphones

[feral pilot]

when i open the setup it shows

[pastel oak]

ur virtual cable might be called CABLE instead

[feral pilot]

"remove driver"

[pastel oak]

what setup are you opening

[feral pilot]

vb audio virtual client

[rare goblet]

Ayo? @feral pilot level 1 !!!

[feral pilot]

cable client mb

[pastel oak]

You dont need to open the virtual cable related app

[feral pilot]

okay

[pastel oak]

Can you send screenshot on wokada now

[feral pilot]

wokada?

wjats tjat

oh

what screenshot

[pastel oak]

The voice changers bame

name*

[feral pilot]

voice changer client demo

[pastel oak]

of the whole thing

[feral pilot]

my whole sceren?

screen?

[pastel oak]

The voice changer screen

[main lichen]

omg sorry for such a late reply but wym by playback steelseries in sound settings?

[feral pilot]

actually lemme use one of teh samples as well

[pastel oak]

On the playback tab of sound settings theres a bunch of steelseries related devices, turn them off

[feral pilot]

when it actually works il use the one i have

[main lichen]

then it doesnt work

[feral pilot]

dms

i mean

wokanda

[main lichen]

it says theres an error

[feral pilot]

mb

#🔍│help-w-okada

[pastel oak]

Why dms just put here

ah

yes

[main lichen]

i turned them off but then it says theres an error in the app and my mic doesnt work

[rough lake]

Applio no UI not working

[valid heath]

"Do you have an nvidia rtx gpu?" i do yes, @pastel oak

[pastel oak]

Then you can train your model locally

https://docs.aihub.wtf/rvc/local/mainline/

Here is everything you need, from instructions to download links

[valid heath]

Ok! sorry im not used to alot of terms, hopefully I wont have to ask more questions and look dumber, thank you!

[pastel oak]

And this section on how to prepare the dataset if you dont know how :
https://docs.aihub.wtf/essentials/how-to-make-voice-models/

[pastel oak]

This is what the help channel is for, dont stress it

[valid heath]

i truly know nothing, will this be a big struggle?

[pastel oak]

Nope not really.

Important is that you record yourself for around 15 minutes, no background noises or other unwanted sounds (2nd link shows how to clean the audio in case you do have noise)

Then you pretty much follow first link instructions and let it train, dont have to do anything while it trains

[valid heath]

oh :O

[pastel oak]

If your audio is clean from the getgo you just put it in the RVC Training and wait

I recommend reading the epochs & tensorboard section aswell on thise docs guides.

Those explain when your voice model is roughly ready

Also called finding out when the overtraining point is

[slate tinsel]

I can't use my external voices; every voice I install separately is terrible, lagging, and has many issues functioning properly. My graphics card is an RX580 (no, I didn't buy it from China in 2021). I believe it's because CUDA cores are better processed by Nvidia. Is there any solution for this, or will I really have to deal with this reality?

[odd shale]

Maybe it's your PC specs then.

[slate tinsel]

of sure

[rare goblet]

Ayo? @slate tinsel level 1 !!!

[slate tinsel]

[vale osprey]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[pastel oak]

are you talking about realtime voice changer?

[slate tinsel]

yes

[pastel oak]

Download modified wokada. Amd on original is not great. It runs great with amd on modified

https://rentry.co/ForkVoiceChangerGuide

[slate tinsel]

tks

[oak grotto]

how do i create my own models? i can download and use pre-trained models from this server but i dont know how to create my own models

[violet heron]

-docs

[azure marsh]

Suggestions for @oak grotto

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[timid valve]

how can you check what pretrain you used to train a model

im assuming it should say it somewhere in the logs

but i cant seem to find where

nvm found it in the train.log file 👍

[limber onyx]

can i use an entire audio?

imma js use an entire audio and uh

one click training i think it is

[thin stump]

If its one whole file then just upload that with a folder so you can just type dataset/charactername

[thin stump]

I don't know what one click training is

just use the regular 'train'

[limber onyx]

uhh i kidna already clicked it

[rare goblet]

Ayo? @limber onyx level 2 !!!

[limber onyx]

Step 3a: Model training started

this is all it says

i mean it did uhh

pitch extraction thing

and other stuff

idk what to do now

thsi is my first time doing this after uhh

months? years?

what do i do after finish training

can i go to sleep and it will still be running

[thin stump]

whats your save frequency?

kaggle has a limited drive storage so you have to delete epochs often

if you did 1 for example

[limber onyx]

10

[thin stump]

the training should be started in the tensorboard URL but mostly monitor for the lowest g/total and download that before you sleep

but if you don't want to train now, then

stop the final cell from running and turn off the session

[limber onyx]

im already training

so uhh

do i just wait

where can i find the epochs ns tuff

[thin stump]

https://rentry.co/RVC-Mainline-Kaggle#monitoring-epochs-and-graphs

[limber onyx]

its on the 66th epoch

so can i leave my pc turned on

and it will be running

[thin stump]

ok but you have monitor this

[limber onyx]

oh

how can i make it so it haves to save

more epoch

how cna i increase

[thin stump]

delete epochs but I just told you how to stop the training

if you think thats going to interrupt your sleep

[limber onyx]

hmmmmmmmm

so can i just delete epochs or i have to stop the training first

[thin stump]

You can stop the training if you want

your processing and feature extraction file are already saverd in the logs folderhttps://rentry.co/RVC-Mainline-Kaggle#resuming-your-training-session

[limber onyx]

okok last question

if i stop training can i somehow put it so it saves at like 30 epochs

or 50

and then resume

[thin stump]

Ok, so it saves at a save frequency of 10. What you have here is what you get. It resumes from the latest cpkt

[limber onyx]

oh

[thin stump]

you didn't skip this right, because that would be crazy

[limber onyx]

ok so to turn off you have to stop running the ngrok cell, and ican remove the epochs while its running

[limber onyx]

thankfully no

i mean

i did what the uhh

picture had

[thin stump]

yeah you can delete epochs while the final cell is running

but if its off, then imjoy is off

[limber onyx]

alllll of them or only a few

[thin stump]

so go to sleep 😵 :

• stop the final cell from running
• stop the interactive session from running

[limber onyx]

One-click training: step2b, model training and feature index training all at once.

took this from the internet

so i guess its both

[thin stump]

oh, and one more thing. When you resume tommorow, you just have to rerun cells 1-5 again. It's all there in your logs so dw

[limber onyx]

all the cells again??

[thin stump]

including any past models you've done

lol yeah

[limber onyx]

even the uhh pretrain model and shit

lkike

the ones u manually install

[thin stump]

the orginal is the best pretrain by the way

[limber onyx]

!wget https://huggingface.co/blaise-tk/TITAN/resolve/main/models/medium/32k/pretrained/G-f032k-TITAN-Medium.pth

[thin stump]

so next time, you don't have use wget

[limber onyx]

the one thats in the kaggle right

[thin stump]

you can run the pretrain cell just to pass requirements

yeah

[limber onyx]

theres one taht says it install pretrain for internet

[thin stump]

its already loaded in as the default

[limber onyx]

from*

[thin stump]

Because if you don't run any pretrain cell, it hits you with "run cell above" or some shizz

[limber onyx]

ohh yeah that

[rare goblet]

Ayo? @limber onyx level 3 !!!

[limber onyx]

imma delet esome epochs

do i delete them all

or do i delete a few

[thin stump]

ctrl + a to delete them all. The first 100-200 epochs doesn't matter much. Whenever you delete them, the drive storage amount is updated in realtime

I don't remember if I told you about g/total https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#usage-guide

T4 X2 GPU is faster than colab so you'll be done with training within like 2 hours mostly

[limber onyx]

hooray

after finish if im sleeping nothing happen right

[thin stump]

Just make sure interactive session is off before you leave the page and you're good

It's the geese icon to your bottom left corner

[limber onyx]

okok so

interactive session off before i go to sleep? (i will keep my windows on

[thin stump]

yup, so that way it won't consume GPU if you're not doing anything. It's the same thing as turning off the session

[limber onyx]

and it wont cancel ze training

[thin stump]

because you have to stop the final cell from running first

to stop the RVC GUI and tensorboard from running

[mystic tangle]

what does index ratio mean?? in dummy terms

im using ilaria rvc to make a ai cover

[thin stump]

drag the index ratio to 1 and you'll notice tons of noise in return for accent accuracy. I like to keep it at 0.50 to 0.60

[mystic tangle]

appreciateit\

[thin stump]

@eternal turtle Just stop the final cell from running and re-run it again

[eternal turtle]

alright

also there's a guy sayin the n word in #🧬│ai-chat

[thin stump]

im not a mod. general chat is the lower realm plebs

[eternal turtle]

it's still showing up as a white screen

[thin stump]

idk then. Let me check to see if its just you. Give me 10 minutes to load hubert base and everything

[eternal turtle]

alright

[thin stump]

this is after you pressed training or after you ran the fifth and final cell?

[eternal turtle]

i'm using (this)[https://www.kaggle.com/code/hinabl/mainline}]

[thin stump]

and I wrote the guide https://rentry.co/RVC-Mainline-Kaggle so i'm just wondering where it went wrong

[eternal turtle]

it's when i try to get the urls

[thin stump]

kk, let me see

[eternal turtle]

https://98d4-34-66-36-189.ngrok-free.app

[thin stump]

Alright, it's up and running. It's your ngrok code

Recheck here

[eternal turtle]

guess i'll make a new account

alright

[thin stump]

nah just visit this page

https://dashboard.ngrok.com/get-started/your-authtoken

[eternal turtle]

yeah i've done that

still white screen

[analog obsidian]

use chrome

imjoy doesnt work on firefox

[eternal turtle]

god damn it

really??????

[analog obsidian]

yea blame imjoy lol

[eternal turtle]

bruh

[thin stump]

lol, never knew that

[analog obsidian]

I just downloaded chrome for that reason

[eternal turtle]

can't believe it

[thin stump]

adding that in

[eternal turtle]

installed ungoogled chromium

works

[analog obsidian]

[eternal turtle]

mfw

Thanks ImJoy, very cool!

[eternal turtle]

alright where's my .index file

nvm

[gray igloo]

thank you very much !!!

[brittle wing]

f >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness. What does this do?

[soft depot]

is there any other way to use the rvc on discord without downloadning exess cable?

[odd shale]

Do you mean W-Okada?

Because RVC is for audio conversion, not realtime.

[soft depot]

yeah

[odd shale]

And nope, you must install VB Cable.

-realtime

[azure marsh]

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

[soft depot]

ah ok, is there any way to fix buffering?

[odd shale]

I'm not sure, but you can read the W-Okada guides.

https://rentry.co/ForkVoiceChangerGuide

[soft depot]

only buffering in dc tho

[rare goblet]

Ayo? @soft depot level 1 !!!

[odd shale]

I don't really know how to fix buffering, sorry.

[soft depot]

alr ok

[odd shale]

You can ask it to one of our helpers.

@pastel oak

[knotty moth]

if the voice doesn't cut when testing in the voice changer, that's a network latency issue

[soft depot]

and how is that fixable :>

[lavish sonnet]

i have a problem with the pytorch installation:

import torch
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "C:\Users\user\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\torch_init_.py", line 148, in <module>
raise err

[signal oracle]

hi! why i can't open RVC gui?

i have winerror 5 access is denied

[tame mica]

hey um so i don't think this is the config.py 😍

[pastel oak]

VB Cable is bad, use VAC Lite (download in the guide under Virtual audio cable)

On discord turn off echo cancellation and use Audio Subsystem: Legacy, this could perhaps help

If those didnt help, send screenshot of voice changer

And what gpu do you have

[brittle wing]

Hi guys, I'm following this guide https://docs.google.com/document/d/1haQAWn4Hnh3Aq8SSGX0tBSY3rDzjYJAcczrUy63oTTs/edit
But I see it is from more than 1y ago and I don't know if there are better options now, but can't find recent guides

[slender blaze]

did you solve this ?

[supple silo]

whats this?

[pastel oak]

Use this
https://rentry.co/ForkVoiceChangerGuide

[odd shale]

Just replace your config.py for this one:

[knotty moth]

double click or right click open the file(s)

[wide kernel]

how to use

😭

[odd shale]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[wide kernel]

how to run

to get gradio

[odd shale]

Read the docs buddy.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[wide kernel]

dont do this to me 😭

tell me what button to press

[rare goblet]

Ayo? @wide kernel level 1 !!!

[wide kernel]

to get gradio

I don't see a button tos tart

all i see is trian

create

[odd shale]

Gradio?

[wide kernel]

share model

[wide kernel]

yea how do i get the link from collab

to gradio

where i can input things to get output

[odd shale]

Which colab are you using?

[wide kernel]

https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw#scrollTo=Sb5fzhzEXK8X

[odd shale]

Ehhh.. That colab is outdated.

Use others instead.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[wide kernel]

send

link

pls

[odd shale]

Please, read the guides mate.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[odd shale]

There you got guides you can read that have also the links.

[wide kernel]

where do i put

rcs

of artist

in collab

[odd shale]

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

If you want, just use Applio Colab.

First, run the Install Applio cell

[wide kernel]

https://huggingface.co/Sztef/Teemo_Omega_Squad1/resolve/main/CartiDeepBySztef.zip?download=true

where do i put it

[odd shale]

[soft depot]

still very glitchy

[wide kernel]

sorry

[wide kernel]

i am using it

rn

mainline collab

[odd shale]

Are you using Mainline or Applio?

[wide kernel]

appolio

now

[odd shale]

First, run Install Applio cell

wait till it finishes installing

[wide kernel]

done

where do i put

https://huggingface.co/Sztef/Teemo_Omega_Squad1/resolve/main/CartiDeepBySztef.zip?download=true

[odd shale]

wait till it installs everything.

[wide kernel]

this

[odd shale]

No need to spam model links.

[wide kernel]

my bad

[rare goblet]

Ayo? @wide kernel level 2 !!!

[odd shale]

wait till Install Applio cell finishes installing everything

then run the "Start Applio" cell

[wide kernel]

okay and

wait

[fluid zephyr]

Wait can RVC use NPU???

[odd shale]

I don't think so.

[eternal turtle]

what should i do if the model dataset has the person severely stuttering

[fluid zephyr]

Weird question but is it possible to run RVC based AI voice model trainer on a NPU? I would imagine it's faster for training on a GPU since NPU is literally a "nerual processing unit"

[wide kernel]

i cliked on start

after it was done

now waht

[odd shale]

Now wait till a Public URL appears and click on it.

[fluid zephyr]

Sorry hold on I should clarify. Not on RVC for nvidia that's for nvidia GPUs only I know... but u mean in the future is it possible to see them being trained on NPUs?

[odd shale]

The dataset should consist only of either a person talking or singing with no noise/instrumental on background

[wide kernel]

im in

the gradio

now where

[odd shale]

Since i'm not an RVC dev, i can't give a clear response to that question, sorry.

[gloomy basin]

hi guys, can you tell me where I can make my voice in RVC? I noob sorry

[fluid zephyr]

Training tab

[odd shale]

You can start reading the docs buddy.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[gloomy basin]

thx

[fluid zephyr]

I figured I'd get that answer lol. We'll see in the future I guess lol

[eternal turtle]

alright, cool

[wide kernel]

where do i put the ai models link in gradio

[odd shale]

There should be a "Download" tab you must get in and paste your model link on it's respective slot

The same GUI tells you where to put the model link once you click on the Download tab.

[odd shale]

Go to the Download tab.

[wide kernel]

yea done

but when im truing to download output

inturrupted forbbidden

[odd shale]

Download output?

[wide kernel]

it isn't downloaded

{"detail":"File not allowed: /content/program_ml/assets/audios/Hanumankind_Big_Dawgs_Ft._Kalmi_Official_Music_Video_Def_Jam_India_1_vocals_output.wav."}

[wispy lodge]

I think none of the existing apps support NPUs atm. It may work in general, but I don't think anyone has hardware to test and add support

[odd shale]

I think you should give a shorter name to the audio you want to use.

[fluid zephyr]

NPUs are becoming mainstream because of copilot AI being local

[wide kernel]

oh oka

[fluid zephyr]

For example, amd ryzen 8000 amd ryzen AI mobile chips come with an NPU

My laptop has a 32TOP Npu. AMD Ryzen 9 8945hs. I see some thread of people talking about implementing it into RVC

[wide kernel]

{"detail":"File not allowed: /content/program_ml/assets/audios/vocals_output.wav."}

[fluid zephyr]

We'll see in the future

[wide kernel]

still isn't working

[odd shale]

Have you downloaded the model you want to use?

[pine dagger]

hey im having issues with a specific rvc notebook that im trying to use,
the problem is just as i added the audios in the dataset folder and it still does this issue.

which it said something like:
** ModuleNotFoundError: No module named 'parselmouth'
Traceback (most recent call last):
File "/content/RVC/infer/modules/train/extract_feature_print.py", line 21, in <module>
import fairseq
ModuleNotFoundError: No module named 'fairseq'
Error preprocessing data... Make sure your data was preprocessed.

ModuleNotFoundError Traceback (most recent call last)
<ipython-input-3-16cc06630afc> in <cell line: 69>()
67 print("Error preprocessing data... Make sure your data was preprocessed.")
68 import numpy as np
---> 69 import faiss
70 get_ipython().run_line_magic('cd', '/content/RVC')
71 def train_index(exp_dir1, version19):

ModuleNotFoundError: No module named 'faiss'**

is there a way to fix this issue?

[wide kernel]

yup

it said download successfully

[pine dagger]

if so let me know.

[wide kernel]

I did

[odd shale]

Also remember you gotta press the "refresh" button and select the model

(I mistakenly deleted the wrong message)

[wispy lodge]

Afaik, pytorch doesn't work with AMD's NPUs (and even Intel NPUs) yet, so I can wish good luck those people. onnxruntime can work though and it's not that hard to make it work for inference. But I don't think training will be easy to do

[fluid zephyr]

Yeah, I figured that much. Pytorch is mainly Nvidia

[limber onyx]

i kinda uhh

turned off my pc while training

nothing happen right

[limber onyx]

right guys

[hoary spear]

“You are using gradio version 3.48.0, however version 4.29.0 is available, please upgrade.”

Anyone have the link for 4.29.0?

[limber onyx]

it was done training tho

[pastel oak]

!pip install gradio --upgrade should work I think

However, using an older gradio version is NOT an error

I am not sure if a newer gradio version will work with whatever you are using. It could break

You probably should not update it but can be a learning experience xd

[odd shale]

By experienced, i learned to think 2 times before updating Gradio.

I wouldn't suggest updating.

[pastel oak]

@hoary spear you should trust Leo 🐢

[hoary spear]

Alright, thanks

[rare goblet]

Ayo? @hoary spear level 1 !!!

[hoary spear]

How do I upgrade anyways?

[odd shale]

Updating Gradio isn't really mandatory.

[hoary spear]

Okay

[limber onyx]

turned off my pc while training
nothing happen right

it was finished tho

[pastel oak]

If it was finished its fine

If it was on colab probably not

[limber onyx]

its kaggle

mainline kaggle

[odd shale]

Probably if you set Persistence to "Variables and Files" it's likely your pth and D and G files are saved

[limber onyx]

oh yeah it is

where can i find my pth and index

[odd shale]

Check Imjoy's file explorer whenever you get the link when executing the last cell on the notebook and go to the "training" folder

assets/weights for your pth's

logs/yourmodelname for your D and G files

[limber onyx]

can i get it fro mthe kaggle

since im having issues running some cells

There will be no output on what is going on, but something is going on. Just wait till this cell finishes loading.
This is done to reduce the output spam on other places this notebook can be used on.
Using Python 3.10.13 interpreter at: /opt/conda/bin/python3
Creating virtualenv at: .venv
Activate with: source .venv/bin/activate

[gleaming scarab]

C:\Users\myname\OneDrive\Bureau\vcclient_win_std_2.0.58-alpha>main.exe cui --https true --no_cui True

It show that when I click on start https

do you know why

[odd shale]

Yep, i was talking about kaggle.

[pastel oak]

Voice Changer doesnt work inside OneDrive. You need to move it somewhere else
I also would not recommend using the new alpha. It is unstable right now
Do you want a link for a different one?

[gleaming scarab]

yes I would like please

[rare goblet]

Ayo? @gleaming scarab level 1 !!!

[pastel oak]

https://rentry.co/ForkVoiceChangerGuide

[gleaming scarab]

What I have to download for change ?

[pastel oak]

Everything you need is explained there I dont understand the question

[scarlet wedge]

Hello. I already know how to make voice models and use tensorboard. But now, i wanna learn how to make a overtrain voice model like the villager model

can someone help me?

[gleaming scarab]

for replace Onedrive

[pastel oak]

You put it anynwhere but OneDrive. For example on C: drive

[gleaming scarab]

ohh okayy

thank you

[odd shale]

Just put an epoch limit of 1000 or more epochs and ignore tensorboard when you already see the lowest point.

[scarlet wedge]

what?

[odd shale]

Yep, that's the way you can overtrain a model.

Seriously.

[scarlet wedge]

lol

so what about the dataset? i need to do something different?

[odd shale]

Or just put a epoch limit of 500-600-700 epochs and ignore the lowest point on graph.

[odd shale]

Nope.

Nothing.

[scarlet wedge]

thanks, i will try it.

[gleaming scarab]

D:\vcclient_win_std_2.0.58-alpha\dist>main.exe cui --https false --no_cui True

it continues

[pastel oak]

You should run http, without the s

[gleaming scarab]

yes I ran start http

[tacit bramble]

Hey so I wanted to know if anyone had a solution to this every time I go to use a voice and it's instabilizing my voice AI will crash and then send me A bug splat report it only happens with certain voices. Some voices I can use like the default one and one of my custom voices but everything else was instantly crash. If anyone has a solution please let me know. I have tried uninstalling, updating my drivers, using a different account, buying the premium on a different account, and resetting my computer.

[vale osprey]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[pastel oak]

Which software

[tacit bramble]

Voice.ai I have tried Wokada but the voices are just too scratchy on there

[brittle wing]

Please advise me a utility in which you can dub the text, so that the voice sounds like a real one

[pastel oak]

then you either used wrong settings or models that dont work on realtime, cause voice.ai is worse and behind paywall xd

[pastel oak]

try this fork wokada and the voice models in the extras section that i linked
https://rentry.co/ForkVoiceChangerGuide

[brittle wing]

in elevenlab i can voicies text so good? for example mrBeast voice

[rare goblet]

Ayo? @brittle wing level 1 !!!

[tacit bramble]

oh, is there a guide to best settings for a rx7800xt? Idk if its even using my gpu even though I put it at gpu0

[pastel oak]

On the original wokada it required you to convert models to onnx, did you do that?

On this fork i linked there is no need for that anymore, its optional

[tacit bramble]

oh, how do you do that?

[gleaming scarab]

pleaase heeelp

[pastel oak]

Welp now we know why they were choppy haha

get the fork i linked, its the best choice for amd gpus:)

[pastel oak]

As I said the alpha is unstable

people here havent used the alpha, i dont recommend it, we cant help on errors

[tacit bramble]

okay, is it a download or just a tutorial? c: I clicked on it and it seeems to just be a tutuorial

[rare goblet]

Ayo? @tacit bramble level 3 !!!

[pastel oak]

Tutorial with downloads, go to download section and so on

[brittle wing]

what about my request?

[pastel oak]

elevenlabs is good with TTS

[tacit bramble]

ah, thank youI am assuming this is what I download for my rx7800 gpu?

[brittle wing]

TTS - what is it? (sorry i’m noob in this theme)

[pastel oak]

Yes

[pastel oak]

text to speech

[tacit bramble]

awesome! Thanks

[pastel oak]

theres a cool new feature to figure out what chunk to use, if you get any problems read "i am lagging, what to do" part on the guide

[brittle wing]

hm, here i can tts with voice mrBeast so good?

i can send link here?

(youtube)

[odd shale]

He wants to use MrBeast's voice on a tts.

[tacit bramble]

okay, I mainly just use these for my dnd to be able to sound scary when I need ot but I still wanna be hearable

[odd shale]

Idk if Elevenlabs still allows you to make TTS models of anyone.

[brittle wing]

elevenlab have voice cloning

[odd shale]

But it's frikking paywalled.

[brittle wing]

money is not really an issue, it's more a question of quality, will there be such a good voice on the dubbing there

[gleaming scarab]

oh okay sorry I have difficulty understanding English

[tacit bramble]

sorry if I sound stupid but on the website when it says 50-60 ms is that talking about chunk?

[odd shale]

To be honest i'm not sure.

[vale osprey]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[pastel oak]

Yes

Howveer, those settings are how low it could go if you didnt run a game and so on. When gaming you might have to raise it

the perf number will indicate what chunk to use for all cases

for example got one that can run 30ms on 3070 but when gaming has to raise to 80, depends on the game too

[limber onyx]

wha do i downlaod after training

i download g and d like peope tlel

[rare goblet]

Ayo? @limber onyx level 4 !!!

[limber onyx]

idk what to do

[hallow valley]

Hello, Im having an issue that when I do a batch conversion all of my .wav files Im converting are being named like this .wav.wav Does anyone have a solution so i dont have to go in and individually rename 600 different files?

[mortal stirrup]

i am using mangio i trained my model upto 700 epochs and mangio gives me error while converting any fix?

[gleaming scarab]

D:\T2L2CHARGEMENT\MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a (1)\MMVCServerSIO>MMVCServerSIO.exe -p 18888 --https false --content_vec_500 pretrain/checkpoint_best_legacy_500.pt --content_vec_500_onnx pretrain/content_vec_500.onnx --content_vec_500_onnx_on true --hubert_base pretrain/hubert_base.pt --hubert_base_jp pretrain/rinna_hubert_base_jp.pt --hubert_soft pretrain/hubert/hubert-soft-0d54a1f4.pt --nsf_hifigan pretrain/nsf_hifigan/model --crepe_onnx_full pretrain/crepe_onnx_full.onnx --crepe_onnx_tiny pretrain/crepe_onnx_tiny.onnx --rmvpe pretrain/rmvpe.pt --model_dir model_dir --samples samples.json

[rare goblet]

Ayo? @gleaming scarab level 2 !!!

[gleaming scarab]

okay now it show that

but it doesn't download

[odd shale]

Donwload the .index and the final pth/or the pth that got the lowest point on the tensorboard's g/total graph

[wheat bane]

hello, I would like to ask for help in making ai vocal models, it would be very important, please help a friend

[limber onyx]

1k epochs and the thing is shitty

sad

[thin stump]

Did I hand you that thing? I forgot

[noble heath]

im getting an error that says no module named mega found

[limber onyx]

wat

idk i use kaggle mainline

and i made 1k epochs and it kidna shit

[thin stump]

Already dm'd. It's just a case of not looking at the tensorboard

[brittle wing]

Thanks! It has nice explanations of stuff, but I still don't know what the advanced settings do

Is the info about that anywhere?

[odd shale]

You can start by reading the docs buddy.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[odd shale]

Probably that was due to overtraining or the dataset was poorly made.

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• 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
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[jade root]

Cloning into 'Mangio-RVC-Tweaks'...
remote: Internal Server Error
fatal: unable to access 'https://github.com/alexlnkp/Mangio-RVC-Tweaks.git/': The requested URL returned error: 500

FileNotFoundError Traceback (most recent call last)
<ipython-input-2-66c1853df4d6> in <cell line: 11>()
15 get_ipython().system('git clone -b pr-optimization --single-branch https://github.com/alexlnkp/Mangio-RVC-Tweaks.git')
16 #Rename to keep backwards compatibility with old variants of Disconnected
---> 17 os.rename("/content/Mangio-RVC-Tweaks", "/content/Mangio-RVC-Fork")
18 get_ipython().system('git clone https://github.com/maxrmorrison/torchcrepe.git')
19 get_ipython().system('mv torchcrepe/torchcrepe Mangio-RVC-Fork/')

FileNotFoundError: [Errno 2] No such file or directory: '/content/Mangio-RVC-Tweaks' -> '/content/Mangio-RVC-Fork'

why error?

[hot lagoon]

See that the docs are not working

[jade root]

What should I do?

github is down

[rare goblet]

Ayo? @jade root level 1 !!!

[jade root]

oh...

[devout wraith]

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[coral roost]

my rvc sounds nothing like the singer i trained it with how do i fix that?

[thin stump]

• get the voice as natural as possible for rvc (no noise or weird vocalizations)
• distribution of accent (reason why era models exist)
• batch sizes
  https://docs.aihub.wtf/essentials/how-to-make-voice-models/

[coral roost]

I only used one song with the instrumental taken out about 3 minutes is that not enough?

[rare goblet]

Ayo? @coral roost level 1 !!!

[thin stump]

for 3 minutes, you can use a batch size of 4. But ideally, you want at least 10 minutes of audio

[coral roost]

Okay that’s what I did it kinda sounds like her but has a lot of robot artifacts

[thin stump]

also i'll dm you something if you dont mind

[coral roost]

Yeah I don’t mind

[thin stump]

I don't expect anyone to understand the guide 100%. I'm just speedrunning your knowledge here

[coral roost]

I just started today hahah 🤣

I have zero knowledge

[thin stump]

I am not the best teacher for this lol. People just expect everything to happen in a single click = high quality model. It's least I/we can do

[knotty moth]

it's not so easy for beginners to learn how to improve the quality, but you can take a look at this guide https://rentry.co/RVC-dataset-RX11

[patent pasture]

does it seem it's overtrain point?

[knotty moth]

there are many mode collapses, so it has probably overtrained earlier

[patent pasture]

ok, thanks

[silent flame]

can u do it for me too pls?

[thin stump]

check your dms

[snow hinge]

opa

[pastel oak]

I can add info to them, most things should be left by default. Some of the fp stuff and jit are only relevant for very low end gpus
export to onnx button is for directml

[snow hinge]

I try to start the program but nothing appears, it just stays on a white screen, can someone help me?

[rare goblet]

Ayo? @snow hinge level 1 !!!

[final sentinel]

Can you please explain one by one the items of setting in the inference of Ilaliarvc?

[mystic tangle]

how much epoch is overkill>

?*

[potent helm]

Does any have the link for the other HIFI GAN BWE colab? I have the original one but there was another one where you can change the model and it kinda had a UI

[rare goblet]

Ayo? @potent helm level 1 !!!

[brittle imp]

i open start_http for install, but nothing happens

[low shard]

there isnt a right amount, look at the tensorboard https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/

[low shard]

be sure you aren't using the google colab easygui or the old ilaria rvc (mangio), share more info about the issue

[gleaming scarab]

Same

It shows the good message but the download doesn't start

[plush compass]

When creating an instrument AI voice, if there is not enough data set, is it okay to add the data set by converting the pitch of the data set from -12 to + 12?

[silent flame]

any help

[spice vine]

anyone here can help me create a custom ai voice?

please?

[ashen pendant]

does the tutorial of making my own voice model start here?

[odd shale]

You can read the docs or make a request on the #1159289738314919936 (You can select if it's free and/or paid) or you can go to the #1191429836321849435 and DM any model master.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[spice vine]

thank you

[odd shale]

Yes it is, tho you must click on "How To Make Voice Models"

[low shard]

Yes that's the guide to make a model with the applio colab ui

Be sure u also follow the guide to make the dataset

[low shard]

Btw solved in https://discord.com/channels/1159260121998827560/1273596379989807205 so other helpers know it's fixed

[agile gale]

@pastel oak jow do i like train a ai to get a voice i want

@low shard

[low shard]

whats your PC GPU?

if u dunno, u can check that via task manager by ctrl+shift+esc then performance tab and gpu

[agile gale]

forty sixty

[low shard]

rtx 4060?

[agile gale]

yes

[rare goblet]

Ayo? @agile gale level 3 !!!

[low shard]

As you got a good PC, you can use RVC locally, you can choose between:

• Applio: A fork of RVC with some extra features like Applio TTS which uses Microsoft Edge API
• Mainline: The original RVC

[agile gale]

can you explain for me do i like put in raw voice of someone then it does it?

[low shard]

nah, it doesn't work like that, you have to first install the program (choose any of the links i just sent before) on your PC and then make a clean dataset to train the RVC AI Voice Model using the Tensorboard

you just need to follow the docs guides

be sure to only read the docs and no yt tuts, all youtube tutorials are outdated

[agile gale]

ims orry but i still dont undertsand dp o have to like get raw mmp3 of them

then use it

then itll automatic

train?

[low shard]

Not mp3, use either .wav or .flac as explained in https://docs.aihub.wtf/extra/glossary/#lossless-formats

[agile gale]

but samew rthign right

it trains on there voice

[low shard]

first of all, read #✨│ai-help message and choose to download either applio or mainline, its blue hyperlinks that redirect u to the guide

[agile gale]

alright which is better applio or mainline

[low shard]

mp3 is different than flac and wav

[agile gale]

or same thing?

[agile gale]

i understand but it uses the same process like it uses the persons voice to train it

correct?

[low shard]

in terms of performance its the same, only thing is applio has some extra thing like TTS, but its really your choice

[red kayak]

they are just different codecs, mp3 is just low quality

[agile gale]

i udnerstand but what im asking is this like uses their raw voice to train off their voice

am i correct?

[red kayak]

yeah, typically u should provide clean raw vocals for best settings

[agile gale]

im kinda confused on what to do im on the docs

[red kayak]

well u wwant to train no?

[agile gale]

yes i do want to train

@red kayak @low shard what now excatly?

[low shard]

make a dataset and read mainline training

[red kayak]

go to the train tab

give your model a name

select 32k

then take your dataset

your mp3 file or whatever file you have and make a folder on your desktop and put it in there

right click on the folder and copy its path

C:\Users\Ζαρκάδας Ηρακλής\Desktop\chasee

ur path should look something like this

[agile gale]

@red kayak question

could i use like a 10 second raw wav and lopp it to be 20mins

would that work?

[odd shale]

Nope, that's not gonna work.

[agile gale]

how would i for examplem get 20mimns raw vocals on someone

[odd shale]

I don't know. Look at the web

[red kayak]

search the web

and look for leaks

[somber valve]

I need help

this not workinnnnnnnnn

[odd shale]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[agile gale]

audacity helps out alot correct?

for removing background noises