✨│ai-help

[viscid moss]

I'll check

maybe colab has updated some shit

[quiet ginkgo]

Speaking of It should I keep on using It for covers

The one made by Tiger

[low shard]

https://github.com/Tiger14n/RVC-GUI this one? outdated

[quiet ginkgo]

Yeah this one

[low shard]

do you got a good pc?

[quiet ginkgo]

It's pretty decent

[low shard]

wait share that google colab link

[quiet ginkgo]

https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS

[low shard]

training would require atleast an rtx 2060

[quiet ginkgo]

I've got a Radeon rx 6600 but It's not supported

That's why I'm training using my CPU

[low shard]

hell nah, training on CPU would take years basically, lemme check the google colab and see if i get that error too

alright the colab has an error

@rigid mango rvc disconnected has errors rn

[quiet ginkgo]

Makes sense

[low shard]

lemme see if i can fix it

[violet heron]

Doesn’t have RMVPE so good luck with raspy (no rmvpe) or breath protection

[quiet ginkgo]

Oooh so It was the GUIs fault

Is there anything as easy that's not outdated

[low shard]

@viscid moss seems like all colabs are fucked about pip with fairseq

ALL

holy shit

[viscid moss]

holy shit

[tranquil sinew]

What do you think of my redesigned aicovergen?
Here, the song does not split itself into vocals and instrumentals, you only need to upload vocals, but there is a tab for combining vocals with instrumentals:🙂
https://colab.research.google.com/drive/1HzuPgICRrjqUWQWb5Zn-l07m099-n-Nr

[rare goblet]

Ayo? @tranquil sinew level 1 !!!

[low shard]

i think downgrading to pip==23.1 should just be the fix

aicovergen worked all fine like that

[low shard]

@rigid mango this should fix it:

#@title Install Dependencies
import subprocess

packages = ['build-essential', 'python3-dev', 'ffmpeg', 'aria2']
pip_packages = ['setuptools', 'wheel', 'faiss-gpu', 'fairseq', 'ffmpeg', 'ffmpeg-python', 'praat-parselmouth', 'pyworld', 'numpy==1.23.5', 'numba==0.56.4', 'librosa==0.9.2']

print("Updating and installing system packages...")
for package in packages:
  print(f"Installing {package}...")
  subprocess.check_call(['apt-get', 'install', '-qq', '-y', package])

print("Updating and installing pip packages...")
# Install pip packages individually to identify problematic ones

# downgrade pip
!pip install pip==23.1
# install packages
for package in pip_packages:
  print(f"Installing {package}...")
  subprocess.check_call(['pip', 'install', '--upgrade', package])

print('Packages up to date.')
firsttry = True

[rigid mango]

Already on it, just making sure the installer doesn't crap itself during install

[low shard]

was just suggesting as the fix didn't give any errors for the setup

[rigid mango]

gotcha, just wanted to see for myself

@low shard @tranquil trail Pushed to the main notebook.

[low shard]

alright

[narrow nova]

it works again thank you very much 🫡 🙌

[lilac canopy]

I currentrly facing a very annoying problem, if soemone could help i would be gladly appreciate, i have a 4080 RTX, Can't open any .bat files, and if i try to open them as ADM the CMD just opens ups as closes, any ideas?

[gleaming viper]

what's the best pre-trained model for female voice, I have more than 4 hours dataset

[lean mist]

hiya - how do we add a custom trained model from #1235952130855010365 once we download it what folder do we put it in? thx in advance

[proper shale]

just use original

[proper shale]

you don't need to put it in a specific folder, but for the sake of convenience I'd recommend putting it on the pretrained_v2 folder

[gleaming viper]

thank you buddy, and what about Ov2Super ? can i use it on a large dataset ?

[proper shale]

yes, but it's much more unstable compared to normal pretrains.

hence why I say, just use original

less chance of getting it wrong

[gleaming viper]

gotcha

[rare goblet]

Ayo? @gleaming viper level 1 !!!

[thin stump]

@vagrant marsh Is this correct? https://rentry.co/RVC-Mainline-Kaggle#ngrok-token

I forgot the basics so i'm suprised you figured it out

[vagrant marsh]

Yep, correctly

Lol

[thin stump]

oh my bad

[vagrant marsh]

Ig you've forgotten to add the basic, like add ngrok token

[vagrant marsh]

Np

Also, just to suggest, explain how to test the audio

[thin stump]

kk, an inference section. I will have to try that out later 🤔

[vagrant marsh]

👍

[analog obsidian]

what i do is inferencing multiple songs and test the model in realtime, if i like both results i keep the model

inferencing songs helps u knowing the voice range of your model
and realtime tests your model generalization ability

generalization means creating new audios that weren’t in the dataset

[vagrant marsh]

I think realtime rvc is bugged imo

I've used it before

[analog obsidian]

rvc realtime is actually better than wokada

[vagrant marsh]

When I joined some VC here

[vagrant marsh]

I'll think about it

[analog obsidian]

yea just a suggestion anyways, inference alone in rvc already tells u how well your model is performing

[vagrant marsh]

I get it

[valid ivy]

everytime I talk, my ai voice hears itself (I think) and causes it to say some things again at a higher pitch. idk how to fix this

[lilac canopy]

i'm receveing this error now, any fixes?

[soft tiger]

i need a radeon okada

can someone send me it please and thank you

[rare goblet]

Arcane Commands

The prefix for AI HUB is /

Useful Resources

Command List
Dashboard
Support Server
Arcane Status
Premium

[soft tiger]

@tired kraken

?

@odd shale

[violet heron]

@soft tiger

?

@soft tiger

[soft tiger]

.

[violet heron]

https://github.com/deiteris/voice-changer

[tired kraken]

Thanks bro

[lilac canopy]

help please

all antivirus off

python installed

etc

[alpine lotus]

depends on the frequency cutoff of the dataset

[violet heron]

Is it installed on the drive your booting off of?

[rare goblet]

Arcane Commands

The prefix for AI HUB is /

Useful Resources

Command List
Dashboard
Support Server
Arcane Status
Premium

[violet heron]

@copper flax do you need help?

[marsh drum]

how do you install RVC?

[knotty moth]

either 40k or 32k, again depending on the cutoff

[craggy patrol]

guys its okay if i rename the index files right

[royal olive]

I think there used to be more RVC models, but they disappeared?

[glass ember]

does not adding the index file make the model sound any different? worse or better

[analog obsidian]

old server got deleted so most models of that era were deleted, tho some survived and are in https://weights.gg/

[analog obsidian]

index files handles the way the model pronounces things, for realtime its advised to not use index files in realtime as they, most of the time, decrease the quality of the model

for local purposes (rvc) they're fine and helps the model pronounce things like the original voice does

start with a value of 0.5 and increase max to 0.75

index value too high might cause artifacts (making the model sound worse)

index is wrongly called the accent of the model, is not really the accent, the accent is already in the model itself and you can't remove it/increase it

[serene hull]

I dont really want to disturb you, but I have a problem with easyGUI📱- 4/20/24, /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 8, in <module>
from infer.modules.vc.modules import VC
File "/content/RVC/infer/modules/vc/modules.py", line 11, in <module>
from infer.lib.audio import load_audio, wav2
File "/content/RVC/infer/lib/audio.py", line 2, in <module>
import ffmpeg
ModuleNotFoundError: No module named 'ffmpeg'

it seems like ffmpeg didnt install but if i`ll try to install what he request in the end it will show me that module fairseq is missed, and I cant install it

[analog obsidian]

is this locally or collab?

[serene hull]

its collab

[analog obsidian]

-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
• EasyGUI, by rejects 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

[analog obsidian]

i recommend you to use between Applio, rvc disconnected, and rvc mainline

easyGUI is very outdated

[serene hull]

thanks, I`ll try

[hollow thunder]

Wait Colab doesn't work anymore?

[analog obsidian]

colab works, he was using an outdated one

[hollow thunder]

Cause rn I'm having issues which I never did before

[analog obsidian]

which colab? btw the only person who can fix a colab related issue is their creator

today rvc disconnected was having issues but got fixed now

[hollow thunder]

Hina_Mod_AICoverGen_colab

[analog obsidian]

when a colab is having issues the only thing u can do is using another one until the error gets fixed

[rare goblet]

Ayo? @analog obsidian level 36 !!!

[hollow thunder]

I have no idea how to use Applio, I'm used to using Hina

[serene hull]

i`ll try to use applio now

its not so hard as you think

like, they all the same at list

[hollow thunder]

Hina has a setup that basically makes the whole cover for you

I just have it make the cover, separate the ai vocal track and music track, fix the ai vocals as much as possible, then finish

[knotty moth]

if you only want inference, it's enough to use local applio even in cpu mode, or try this https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18

[serene hull]

@analog obsidian thank you, taking off my hat

[hollow thunder]

I'm doing song covers

[knotty moth]

yeah for covers as well, for bgm separation use UVR https://huggingface.co/spaces/TheStinger/UVR5_UI

[hollow thunder]

You have a link to Applio?

[spice blaze]

ik, dumb question, but is it bad to upload 3 hours of sound files for a voice model?

am new

cause the google collab thing isn't working and I don't know if that's the issue

[analog obsidian]

hina said he fixed his aicover colab, try it if its working now

[hollow thunder]

Aight

Was sending the error I was having

[analog obsidian]

remember to exit your current session

and reload the site

[hollow thunder]

I'll try it, if it doesn't work still I'll send what I'm seeing and tell you the end result I'm getting

[hollow thunder]

Ik

[serene hull]

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

[hollow thunder]

@analog obsidian

Yes I use a phone, I don't have a computer

[analog obsidian]

@glad zealot

nothing i can do here, i didnt made the colab srry

[hollow thunder]

It's alright dw

[glad zealot]

just refresh

fixed a couple mins ago

[hollow thunder]

I did, multiple times

And disconnected the runtime

To start a new one

I still get the issue

[digital pulsar]

guys whats the song name that sounds like this

[glad zealot]

weird

[cloud finch]

Is there a way to get APPLIO to display more emotion in the generation? In some clips i input their angry but the ai will make them sound normal

[hollow thunder]

Idk what's going on with it

[rare goblet]

Ayo? @hollow thunder level 6 !!!

[glad zealot]

ye imma check again

[analog obsidian]

this depends in the model and how they were trained, so if your model was trained using regular speech is not gonna be able to express anger

[cloud finch]

I gotcha

[hollow thunder]

Okay

[glad zealot]

nope seems to work just fine

[hollow thunder]

Why am I still getting the errors then?

[glad zealot]

a wait it was covergen

i havent used that think since forever lemme update

[hollow thunder]

Yea it's covergen

Hina_Mod_AICoverGen

That's what I'm using

[glad zealot]

yeye lemme test it again

[hollow thunder]

Okay

[glad zealot]

ye no clue why, its the install stuff thats not working properly when its should work

[hollow thunder]

What do I do?

[glad zealot]

you can try using something like ilaria rvc to do it instead

but you need to clean your audio first

[hollow thunder]

So Hina is outta the question for the time being?

[hollow thunder]

Any other colabs?

[glad zealot]

found a fix

works now

[vague hinge]

How do I open the ai voice changer because it won’t open

[gray crow]

hey appreciate you help. May I ask depending on what cutoff? Or a source where i can read about this

[vague hinge]

Can some body help with this issue i try to open the voice changer and it doesn’t open ???

[gray crow]

we need more information. What happens after opening? Any error? What system are you using?

Has it ever worked before?

Fresh installation?

[thin stump]

here's a 15k cutoff so that would be 32k as your target sample rate

on the spectrogram

https://rentry.co/RVC-dataset-RX11#trying-to-explain-a-spectrogram

[gray crow]

Awesome, now I understand, thank you so much! ❤️

[thin stump]

and they were right, not 48k so you can choose between 40k or 32k

32k just has less harsher sibilances and plosives

[vague hinge]

im on windows this is my first time and this is what pops up

<@&1159293204038955078> pls help it will just pop up with this???

[gray crow]

You are in Select mode. Please press enter on the commandline and show the output again

[hot violet]

what kind of fix on that colab? is it the downgrade colab cuda?

[vague hinge]

It dissapears when i clcik enter

[gray crow]

okay than start it again, and try not to click on the cmd

[vague hinge]

if i dont click on it it just pops up for a milisecond then dissapears]

[gray crow]

then edit the batch file and put a pause at the end

[vague hinge]

what? bro ims

im stupid

i dont know how

[gray crow]

youre good bro. Just right click and edit, go to the last line and type pause

[vague hinge]

if i click any button it dissapears

[gray crow]

Yeah thats how batch files usually work when it errors

[vague hinge]

so how am i supposed to edit it

[gray crow]

1. right click the start_http.bat
2. edit
3. go to last line (or create a new one)
4. type "pause"

I dont know the tool youre using but even if somebody is here who exactly knows, they would need some kind of error message to know whats going on.

[vague hinge]

good

[rare goblet]

Ayo? @vague hinge level 2 !!!

[gray crow]

put it in a new line

[gray crow]

after doing that, the cmd should not instantly close, but instead show the error message

(in the best case because i dont exactly know how the batch file is built)

looks good, try running it again

[vague hinge]

still dissapears unless i click on it

[gray crow]

what tool you use exactly

[vague hinge]

what do u use cuz im tryna get the vc changers

[gray crow]

I dont use realtime vc

-vc

[vague hinge]

idk i searched a tutorial they said download vb cord and what ever this is then to start_.http https://huggingface.co/wok000/vcclient000/tree/main

[gray crow]

one sek

[rare goblet]

Ayo? @gray crow level 6 !!!

[gray crow]

which file did you download? Whats your gpu?

[vague hinge]

this one MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.15.zip

[gray crow]

Do you have NVIDIA gpu?

[vague hinge]

yeah NVIDIA GeForce GTX 1070 Ti

[gray crow]

try MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.17b.zip
or MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a.zip

[vague hinge]

all it does is pop up for slighty longer time

[vague hinge]

above

[gray crow]

im currently at work, but (if nobody else can help you)
I will try downloading the tool at home and get back to you

[vague hinge]

wait one last thing this is what it shows

[gray crow]

From what they say here (exactly same problem) its your outdated windwos version:
https://github.com/w-okada/voice-changer/issues/1288

And this allso adds up to your message Access is denied:

If the cmd closes right away after you run it that means that there is some issue with your windows installation. Either your firewall or windows is blocking the file, or you are missing something important on your end. You should try to update windows to the latest version. Update net framework if possible. Check that you have webview installed. Have you removed any important feature from windows in an attempt to uninstall unnecessary things? Have you tried disabling any security software like Windows security, firewall, antivirus, anti malware, et? What about disabling UAC?

So you could try disabling you av

to double check

[vague hinge]

what so i need a new windows??

[knotty moth]

GTX 10-series can work but it may have performance/some other issues

[vague hinge]

yeah well it isnt working

[gray crow]

1070 shouldnt be a problem IMO

[vague hinge]

it stays but i cant click enter anymore

[gray crow]

ive ran it with 1660

[vague hinge]

nah but like ik it isnt worjing but how do i fix it

[rare goblet]

Ayo? @vague hinge level 3 !!!

[gray crow]

first try disabling you anti virus.
If its still not working (i guess) you need an windows update

[vague hinge]

i dont have one

[gray crow]

by default windows defender should be active on windows

[vague hinge]

what so i need to turn that off that dosent seem very safe or nessacary]

[gray crow]

Its just to double check if something is blocking the voice changer my guy

Disable it, try the tool, if still not working its not the AV and maybe an outdated win ver

but however, if you dont feel safe disabling it you could try updating your win version

And you should also try to get help here #🔍│help-w-okada as this channel is specificly for the tool youre using

[vague hinge]

yeah well nobody speaks english in there my man

[gray crow]

let me check if this error is known

https://github.com/w-okada/voice-changer/issues/950

Caused by renaming folder, or abruptly pausing unzipping. I fixed by downloading again and unzipping without changing folder name

[vague hinge]

i didnt change any names man @gray crow

[gray crow]

The first comment is the guy who developed the tool @vernal lily

[vague hinge]

i did extract it

[gray crow]

The 3rd comment also suggest a problem when unzipping the files

At this point its hard for me to help you.
As i said, i can try it when im at home and get back to you. sorry pal

Some other chinese guy suggested to use ver MMVCServerSIO_win_onnxgpu-cuda_v.1.5.2.3
however i couldnt find it in any hf repo @vague hinge

[knotty moth]

fyi 18a is the latest and stable, but you can try this fork for better performance https://github.com/deiteris/voice-changer

[knotty moth]

and I don't recommend the alpha version yet

[vague hinge]

@gray crow its the same now but it dosent dissapear and i can click buttons

wait something is happening

[gray crow]

👀

[vague hinge]

all these bars started loading and when they where done it dissapeared

wait so wgar do i do from here

[knotty moth]

till you try another version like I suggested above

[rare goblet]

Ayo? @knotty moth level 20 !!!

[vague hinge]

tf no he was giving me instructions i already tried thoes and he told me this one and it looks like its working but idk

[knotty moth]

ok, so?

[vague hinge]

@gray crow turns out i just had to let it do its thing thank YOU SOO MUCH for helping me i tried getting this done for so long so many different times

[knotty moth]

you just need to set everything up and change settings according to the guide, should you have some more questions just ask
https://rentry.co/VoiceChangerGuide

[gray crow]

Awesome, good to hear! Have fun 🙂

[gray crow]

Thats normal. When you initially run the .bat file it will install everything.
On the second start it starts the software!

[vague hinge]

ok yeah one question how do i make vb cable input to my mic?

[knotty moth]

I'd recommend VAC as said in the guide https://software.muzychenko.net/freeware/vac470lite.zip

[vague hinge]

bro i dont need any more trouble

[pastel oak]

input: your mic
output: vb cable

on discord and apps
input: vb cable
output: headphones

[knotty moth]

use it as audio output, but I'd recommend so to avoid troubles when using in discord/games

[vague hinge]

where on the ai?

[vague hinge]

k

but then what do i put input for voice mod tho

[pastel oak]

can you explain what youre trying to do

youre trying to use voicemod and the ai voice changer?

[vague hinge]

make it so i can hear myself with voicemod?

[rare goblet]

Ayo? @vague hinge level 4 !!!

[pastel oak]

you can hear yourself by setting "monitor"to your headphones, no need for voicemod in that case

[vague hinge]

okie

tf the voice changer stopped working man

the code keeps dissapearing again

nvm just i cant hear myself still like why

[pastel oak]

No clue

[vague hinge]

thank you man this helped

like litteraly no sarcasim it was just cuz my headset was dying

[pastel oak]

ohh lol

[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
• EasyGUI, by rejects 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

[lilac canopy]

i fixed it, but i can only open .bat files trought a command prompt or terminal

it was a pain in the a$$$

14 hours to fix the problem

at least i can use it now

XD

[crimson depot]

@viscid moss Check this out
Traceback (most recent call last):
File "/content/AICoverGen-NO-UI-es/src/main.py", line 11, in <module>
import gradio as gr
ModuleNotFoundError: No module named 'gradio'

1

[crimson depot]

This occurs in SP and EN

[lilac canopy]

is there a guide on how to clone the voice you want and download it?

[crimson depot]

In fact all the collabs has problems LOL

[glad zealot]

upadated pip

[lilac canopy]

I want to train a voice of a person, for a text to speech, what are the best programs to do that and how i do it it the best results possible? i have a 4080 RTX

[hot violet]

how did you do that?

[rare goblet]

Ayo? @hot violet level 23 !!!

[rough lark]

I need someone to dm me so I show them a video of the problem am having should be a quick fix.

[pastel oak]

Just post it in here

[rough lark]

Well idk if this channle is for this specificly my audio lets say i talk on discord the sound is fine but sometimes the voice just cuts out like am lagging or somthing when am not but the voice just does it randomy applys for all voice modles so idk how to fix that

[pastel oak]

Whats your GPU, whats your chunk and extra

Sounds like your gpu cant handle the chunk you set it to, would need to raise it to a higher value

[rough lark]

Can I just show you an example on dms?
Chuck:192 (512.0ms,24576)
Extra: 8192

[rare goblet]

Ayo? @rough lark level 4 !!!

[pastel oak]

I dont need an example

I already understand what the problem is lol

and whats your GPU

[rough lark]

Alr

Navidia geforce gtx 1660 ti

[pastel oak]

for gtx cards i would recommend the modified wokada version that runs much better without problems

https://github.com/deiteris/voice-changer/releases

Download voice-changer-windows-amd64-cuda.zip.001 AND voice-changer-windows-amd64-cuda.zip.002

---

Make sure you have 7zip or winrar (download if not: https://www.7-zip.org/)

• Right-click the voice-changer-windows-amd64-cuda.zip.001 ZIP file. In the opened action menu select 7-Zip > Extract to "voice-changer-windows-amd64-cuda". This will unpack both files, no need to unpack them separately.
• Open the extracted folder (voice-changer-windows-amd64-cuda) > MMVCServerSIO.
• Run MMVCServerSIO.exe.

[rough lark]

Okay no need to change my Chuck and extra after that? And one more question how did I before hear a replay of the voice tru my headphones but I dont hear it anymore?

[pastel oak]

monitor = your headphones

[rough lark]

Unzip and extract all the files you named into the folder with the voice changer correct?

Second to last step

[pastel oak]

if you use 7zip and you do what i wrote, itll unzip both files automatically

if you use winrar im not sure if it does that, but yes into the same folder

[rough lark]

Okay thank you man

[low osprey]

Can someone help with settings for vcclient 2.0.45 gpu asus rog strix rx 580

[knotty moth]

I'd rather recommend the fork version https://github.com/deiteris/voice-changer/releases

[low osprey]

What version should I install?

[pastel oak]

Follow the AMD steps
I just did this guide 2 minutes ago, its not fully finished but should be useful

https://rentry.co/bettervoicechangerguide

[low osprey]

Ty

[rare goblet]

Ayo? @low osprey level 1 !!!

[crimson depot]

Your collab has the same problem as Not Eddy's

[lilac canopy]

which number it peaks?

44100 Hz

can someone help out?

If the frequencies don't reach the top of the spectrogram, see at which number peaks & multiply it by 2.

[low osprey]

What about settings?

[pastel oak]

f0 det rmvpe
chunk start with 150 and test. if its laggy, raise it, if its not laggy, you can go lower (thats the delay)
extra 2.7s

[lilac canopy]

@pastel oak

can u help out bro

sorry to bother u

i was reading the documentation u sent me

[lilac canopy]

these were my results

[pastel oak]

you cant see the sample rate on that screenshot

it should be on the left of it

[pastel oak]

not that

is there nothing else

on the diagram

[lilac canopy]

just this bro

from the official documentation

this is my results

for my audio

[pastel oak]

Uhhh weird that thing doesnt actually show the correct spectogram

i use this https://www.spek.cc/

if you could use that instead

[lilac canopy]

alright

just upload the audio that i saved from audacity

right?

[pastel oak]

yes

[lilac canopy]

do i need to download

this

program?

[pastel oak]

yea

[lilac canopy]

ok

gimme a mommentum

@pastel oak

[pastel oak]

wherever you downloaded your audio from has very bad sample rate

this is like under 20khz

[lilac canopy]

i used

videos

extracted the audio

from the videos

and used capcut

[pastel oak]

what videos from where

[lilac canopy]

youtube

[pastel oak]

well cant promise the quality will be good but train with 32k

[lilac canopy]

i did this > 3 videos from youtube > extracted the audio from these videos with capcut, and used their pro option of removing music and noise, just the person talking, and addded it again to audacity and followed the documentation

24min total

of the audio

[pastel oak]

oke

[lilac canopy]

is that ok?

can i start training?

[pastel oak]

yea

[lilac canopy]

alright

i will need help with that too

dsaojsdajpsfajfd

sorry i'm stupid

[pastel oak]

idk if theres anything else you need to do with training TTS models

i cant help with that

its somethign with gpt-sovits idk lol

[lilac canopy]

i dont hat is that

;-;

i dont know what is that

[pastel oak]

https://docs.aihub.wtf/tts/gpt-sovits/

it has training here

[lilac canopy]

so

i wont

need

to use

the

RVC?

[pastel oak]

i suppose not

[lilac canopy]

i spent 12 hours trying to fix it

yesterday

'-'

but ok

i hope this program works fine

thats actually the program i need

[rare goblet]

Ayo? @lilac canopy level 6 !!!

[lilac canopy]

lol

ty so much @pastel oak

cause i just need to train a voice

and make it speak whatever i want

did u get it?

XD

[pastel oak]

ye

[lilac canopy]

thats actually all i need

should i delete RVC them?

or nah

i spent 12 hours trying to fix it

lmao

i wont delete it

i was going insane at 4 am

[knotty moth]

such a lofi quality, even trying to reconstruct using AudioSR won't still be good for training

[lilac canopy]

what should i do

hotaru

i want the perfect audio

i'm cell

i'm perfect.

could you assist me?

pleaserino

[pastel oak]

use better video downloader if youre going for youtube

cobalt.tools

y2down.cc

[lilac canopy]

i used

y2mate

;-;

[pastel oak]

then idk if your capcut ruined it

[lilac canopy]

problably

haha

no worries

i ill do again

i just need to download the

audio

right

[pastel oak]

check the sample rate of before you put into capcut

[lilac canopy]

from the youtube

[pastel oak]

ya

[lilac canopy]

alright

i will do another one

no worries

i want it to be realistic perfect

that will chock people

and make them pass out

[lilac canopy]

i will get good results with this?

[rough lark]

where do i find voice-changer-windows-amd64-cuda

[pastel oak]

youll train fast yea

[proven hill]

for?

[lilac canopy]

tts

and training

[proven hill]

hell yea

[lilac canopy]

HELL YEAH

[pastel oak]

on the first link in releases

[lilac canopy]

I NEED TO LEARN HOW TO MAKE A VERY GOOD AUDIO QUALITY

LIKE A VERY GOOD ONE

[lilac canopy]

i'll try the suggestions of @pastel oak and @knotty moth

[rough lark]

i downloaded thoes 2 now idk where to extract them

[pastel oak]

put them in the same folder
extract 001, it will automatically extract 002 aswell

dont open and drag and drop, you have to use the extract to .. function

[rough lark]

a random folder i create in the other voice changer folders?

[pastel oak]

what other voice changer folders?

Just any folder you want to save the voice changer in

you put those 2 zip files in there then extract

[rough lark]

so in what folder of there i put the files in am sorry am confused

[pastel oak]

the 2 things i sent you have nothing to do with the rest of the stuff

ok let me simplify this even more

put 001 and 002 in "ai-voice-changer"

extract

it will make new folder

done

[rough lark]

okay thanks lol

have to explain it like a kid to me (:

[lilac canopy]

anyone knows why i can't open .bat files on my pc? like any, i cant open any, it says windows couldn't find the path to that archive, the .bat path is on my ambient variables

[rough lark]

file replace yes to all right?

[pastel oak]

there should be nothing to replace but i gueeess yes? lol

[pastel oak]

yes

[rough lark]

yes or yes to all lolll

[pastel oak]

ig yes to all

[lilac canopy]

shad

i have a major problem

with my pc

i need to solve

i cannot open any .bat files with double clicking on it

it needs to be on a terminal

i've tried all fixes on the internet

literally

[rough lark]

@pastel oak well it extracted and there is no new folder,so what now

[pastel oak]

Idk google

[lilac canopy]

registry

stuff

chatgpt is teaching me

[pastel oak]

idk what you did ngl

[lilac canopy]

i love AI

[pastel oak]

im not sure if thats what you need for something simple as that..

but ok

[rough lark]

well i downloaded thoes 2 clicked on the first one extracted it to the folder where all my ai stuff is clicked all and yeah thats that lol

[pastel oak]

did you do it with 7zip or winrar

[rough lark]

7

[pastel oak]

if you did everythign i said there should be a folder called voice-changer-windows-amd64-cuda

look in your downloads folder too

[rough lark]

in downloads are only the 2 things you told me to downaload

[pastel oak]

You did something wrong then

I cant tell what

[rough lark]

1.download 2.go to downloads and right click 7zip extract to ai folder and click yes to replace.

[pastel oak]

or open mmvcserversio

maybe its in there

yes as i said the replace thing is not normal

maybe it copied everything into the mmvcserversio folder

itd have been easier if you just used an empty folder to put it in

maybe try again on an empty folder

[rough lark]

emty folder that is in ai voice changer folder?

[pastel oak]

whole new folder

[rough lark]

what am i looking for in here

[pastel oak]

just for testing

[pastel oak]

Desktop make a folder put it in there

or new folder in ai voice changer folder yes

and then put 001 and 002 inside that new folder

[rough lark]

i just need to extract 001 there u said that 002 will automaticly?

[pastel oak]

yea

[rough lark]

this time in the new folder it dident ask to replace anything obviusly

[rough lark]

now i have the same file in the new empty folder as is here that i had on top "MMVCServerSIO"

[vague hinge]

Anybody know why the Elliot gtag voice changer doesn’t work

[mystic fern]

anyone know why the application keeps shutting down when i turn on live mode

[lilac canopy]

bro

its cause its annoying

[lilac canopy]

why my pc isn't opening .bat by double clicking it them?

and windows saying it couldnt find the file

lol

i've tried everything

even with chat gpt

[pastel oak]

i dont know

[lilac canopy]

alright

i will fix that later at night

now i need to train my model

which websites u said again

to download youtube videos

and do stuff with the audio on audacity?

[rough lark]

@pastel oak

[pastel oak]

does it look like this

except without pretrain and model_dir u probably dont have that yet

[lilac canopy]

i need the voice of the same person talking for 1-30 min right?

i want the best

how long should be the video?

you recommned

[pastel oak]

yes everything is correct

now start the mmvcserversio file

[pastel oak]

15 mins should be fine

[lilac canopy]

alright

[rough lark]

i did now what

[viscid moss]

I'll update it in a bit

i just woken up

[pastel oak]

its installing files

thats all

then you have a good voice chnger working

[viscid moss]

https://tenor.com/view/woahm-anime-girl-anime-girl-anime-girl-sleepy-gif-1859288045321179131

[rough lark]

its just black lol

[rare goblet]

Ayo? @rough lark level 5 !!!

[lilac canopy]

is 13 ok?

;-;

[pastel oak]

ye

[rough lark]

oh acuually now its working

[lilac canopy]

alr

[rough lark]

thanks man! @pastel oak

do i need to start it ever time i use the ai voice changer or this once is good?

[pastel oak]

which gpu did you have again i forgot

[lilac canopy]

used cobalt @pastel oak now i will go to audacity correct?

[pastel oak]

you always start MMVCServerSIO.exe

[pastel oak]

ye

[rough lark]

.

[pastel oak]

use like 120 ms chunk and 2.7s extra

f0 det use rmvpe and try around

[lilac canopy]

why my noise gate plugin is different

[rough lark]

2.7extra?

[pastel oak]

2.7s yes

thats equal to the max setting

[lilac canopy]

@pastel oak

[pastel oak]

i dont know

[lilac canopy]

do you have the correct noise gate plugin

?

so i can replace it

[pastel oak]

im using stock

[lilac canopy]

or it doesnt matter

what's stock

[pastel oak]

the normal one

i dont know why yours is different

[lilac canopy]

i downloaded it from a different

thing

cause mine

didnt came with it

so i had do install the noise gate plugin

[rough lark]

wait now i have this on google i dont need the app no more?

[grim tangle]

Hey, is there a website yall do these AI voices or do u do it through this server? I want to have good quality Star Wars AI voice generator but I havent found any.

[pastel oak]

It automatically opens with browser always, its more effective

theres no app function in this

[proven hill]

https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit

[rough lark]

ah okay so what u reccomend for s.thresh and rmvpe normal or rmvpe_onnx

[proud anvil]

Guys who don't have a working neural network https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw#scrollTo=DZDKirCM0F9g

[grim tangle]

Thanks, it looks complicated 😄

[rare goblet]

Ayo? @grim tangle level 1 !!!

[pastel oak]

use normal rmvpe

s.thresh depends on your mic if it picks up a lot or not

[rough lark]

anything on advance settings or all all good with default

[proven hill]

thats why theres a tutorial :)

[pastel oak]

should all be fine

[grim tangle]

How long does it take to make a voice line?

[proven hill]

8 seconds

[rare goblet]

Ayo? @proven hill level 105 !!!

[grim tangle]

Nice, can I generate Star Wars voices? Darth Vader, Obi Wan, Clones, Ahsoka. Basically any character from star wars and is it good quality? Ive tested many AI voice generators and they sound terrible

[proven hill]

#1175430844685484042

[crimson depot]

Take your time. I would appreciate as always if you ping me when you repair it

[lilac canopy]

yo guys

i used audacity ok

but theres a music

on the background

of the people that is speaking

should i remove that on capcut

or you guys recommend another?

[proven hill]

uvr

[lilac canopy]

whats uvr

[proven hill]

-uvr

[azure marsh]

Suggestions for @lilac canopy

Ultimate Vocal Remover

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

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

[lilac canopy]

ty

you're literally saving my life

i didn't even knew these kinda of free softwares existed...

[proven hill]

:3

[graceful pike]

hi guys I trained with applio, but only pth was created without an index file. What should I do in this case?

[lilac canopy]

@proven hill tell me the best config please

my pc can handle it

[azure marsh]

Ultimate Vocal Remover

This guide shows how to separate the vocals and the instrumental of songs using UVR locally
Link: https://docs.aihub.wtf/vocal-cloning-guide/vocal-and-music-separation/uvr

Credits: Nick088

Ilaria UVR

Link: https://docs.google.com/document/d/1M3rEj5RPSE60f17RdM9MaIKKkS2bLrj-dwwuCUOdRaY/edit?usp=drivesdk

Credits: Angetyde

[lilac canopy]

lol

[uncut solstice]

can someone please help me

idk what im doing 😭

[proven hill]

uh idk, im not staff anymore

[thorn socket]

Why vol. 0?

[subtle phoenix]

if I have a large dataset and I preprocessed and extracted features for it, can I copy preprocessed files and use them in a new experiment. does the sample rate o f the old experiment matter?

[rare goblet]

Ayo? @subtle phoenix level 3 !!!

[low shard]

link has been changed to https://docs.aihub.wtf/rvc/resources/vocal-isolation/

[low shard]

@quasi dagger u may need to update https://docs.aihub.wtf/vocal-cloning-guide/vocal-and-music-separation/uvr to https://docs.aihub.wtf/rvc/resources/vocal-isolation/

[uncut solstice]

my mic is picking up all the background sounds and it sounds really funky

[rare goblet]

Ayo? @uncut solstice level 1 !!!

[jaunty bloom]

Having a problem installing dependencies on Google colab. Anybody know how to fix it?

[low shard]

what google colab link?

[jaunty bloom]

Rvc disconnect v2 I think its called

[low shard]

this https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS ?

show a screenshot of the error

[jaunty bloom]

I tried to take the best picture I could. I'm not near my pc at the moment.

[lilac canopy]

@low shard is this correct?

i want only the voice

of the person

of this video

[jaunty bloom]

It was working for me but now it gives me this error every time

[low shard]

the name of the google colab looks a bit different, u sure this is the right link https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS ?

[lilac canopy]

Nick, on my audio file, there's a background music, to train my model i need to remove the music

is this correct?

[low shard]

if you got a good gpu, id sugges to use mdx23c instead

[lilac canopy]

tell me

which configs

so i will save it

pls

[low shard]

alright thats good

one sec

[lilac canopy]

i need the best one

[jaunty bloom]

I'll try once I get home

Thr link looks mostly the same, but I do see a slight difference from the one I was using

[low shard]

use this mdxnet model

[lilac canopy]

i dont have it

[rare goblet]

Ayo? @lilac canopy level 7 !!!

[lilac canopy]

oh

download more models

ok

[low shard]

Click on the Settings Button (wrench icon to the left of Start Processing) and go to the Download Center tab

after click mdx-net and u can select the mdx23c model and download it

also id leave the overlap as default

[lilac canopy]

alright

downloading the model

basically is a voice of a woman talking, with a music background, i want only her talking

[low shard]

alright

[low shard]

yours looks like its from github, looks like a fork

also send me where u got that link

[low shard]

the guide for that is https://docs.aihub.wtf/rvc/cloud/rvc-disconnected/

[jaunty bloom]

I'll send it to you when I get back home

[lilac canopy]

@low shard

what do i do now

[low shard]

now you should be able to see the mdx23c in your model list

[lilac canopy]

is that correct?

i want the best possible result

ops

vocals only

right?

@low shard

[low shard]

Well you don't need the instrumentals so yea

[lilac canopy]

well it worked

but

i'm hearing some noises

can i send u the file

so u can hear it

my gpu worked fast af

[low shard]

Use the vocals as an input, and after use the de noise model

[lilac canopy]

whats the de noise model?

the best one

[lilac canopy]

is that correct?

[low shard]

Yea

[lilac canopy]

the output

i should let it

flac

or

wav

[low shard]

They are both quality lossless, the only difference is that FLAC is smaller file size

[lilac canopy]

IT WORKED

LIKE

A CHARM

I LEARNED THE PROGRAM

THANK YOU SO MUCH

I LOVE YOU

I HOPE YOU HAVE A LONG LIFE AND A GREAT DAY

❤️

is this good enought?

[pastel oak]

it seems better than the one before, itd be ideal if it reached up to 16khz at least but not much u can do

[lilac canopy]

i'm sad

i'm new to this

i will try with this one

which value should i use? @pastel oak

[pastel oak]

32k

[lilac canopy]

ok

ty guys

i'm confused

[knotty moth]

that seems more like lowpass "phone call" effect

[lilac canopy]

what is that?

is it that bad?

i will use it for tts

😦

its very confusing for me

https://docs.aihub.wtf/tts/gpt-sovits/

i'm trying to follow this

can u help me?

i already have the audio

[knotty moth]

I can dm you the example of such that effect

[lilac canopy]

i mean

for me its good enought

i will improve it later

i just need to learn the tool

and how to config the folders

can you assist me?

i'm at this part

already opened locally the .bat

mine is in portuguese

[knotty moth]

lmao you still want that lofi recording thing as dataset?

[lilac canopy]

BRO

i just want

to learn

how to use it

i can train another model later

its my first time using these tools

[knotty moth]

you can find some vtuber streams, which mostly are 16 khz cutout and good for making 32k datasets

[lilac canopy]

hotaru

this is my model

a specifc

person

that i want to clone the voice

and use as a tts

to impersonate the person

did you understand whats my goal now?

@knotty moth

[knotty moth]

bruh

even lofi audios aren't good for tts as well

[lilac canopy]

i will show you my model

[knotty moth]

and if you specifically want to make models for tts, there's GPT sovits or this
https://github.com/fishaudio/fish-speech

[lilac canopy]

YES

I'M USING

GPT SOVITS

XD

[knotty moth]

you gotta find some guides, as I dont have much knowledge about it

[lilac canopy]

I'M GOING TO SHOW YOU THE AUDIO I WILL BE USING

check dms

[viscid moss]

CoverGen en and es working again

[crimson depot]

You’re the bee’s knees!

Thanks

[jaunty bloom]

https://colab.research.google.com/github/MedicDoesStuff/RVC-Disconnected/blob/main/RVC_Disconnected_(V2).ipynb

[rare goblet]

Ayo? @jaunty bloom level 2 !!!

[low shard]

yep, outdated fork, use the original one i sent you

[jaunty bloom]

Thank you!

[lilac canopy]

@low shard

can u help me out?

pls

i'm getting errors

its gpt-sovits

idk why

the audio is the one i showed you

[low shard]

Sorry but I never used gpt so vits and can't use it locally

[lilac canopy]

you cant use it locally?

or you mean you dont know how to help me?

[low shard]

yep cus bad pc, and i dont know how to help you because i never used it

[lilac canopy]

can u recommned

soemone

to assist me pls

[low shard]

maybe @proper shale

[lilac canopy]

mj can u assist me pls

[low shard]

Not sure what specific helper knows about gpt so vits, hes very active so i @ him as he might

[lilac canopy]

alright

@proper shale can u assist me please

[proper shale]

dawg i dont know

sorry

[lilac canopy]

:/

[low shard]

is there any helper who knows about gpt so vits here ?

cus i honestly dont remember

[proper shale]

well

im

not sure

[knotty moth]

delik

[low shard]

he left this server months ago, hes in sanctuary

[knotty moth]

what I know is @karmic heron

[lilac canopy]

it doesn't support portuguese?

cause the voice i'm trying to clone is in portuguese

maybe thats the error?

[low shard]

gpt so vits is only for english, chinese & japanese

[lilac canopy]

which one

supports

portuguese

?

thats why i was getting errors

bruh

[knotty moth]

that's why RVC is more versatile as STS

[lilac canopy]

does it has text to speech

and voice cloning

?

'-'

[proven hill]

no

[lilac canopy]

which one supports portuguese-brazilian please

for tts

voice cloning

[proven hill]

maybe fish speech

[neat pilot]

yo what mean this RVC 2 is new version or something i dont get it?

[low shard]

gpt so vits doesn't idk if there's a TTS that supports that language but you could Try RVC which is (STS, Speech to Speech) as Text To Speech
RVC forks such as ilaria rvc mainline & applio have built in tts (using Microsoft Edge TTS to make a generated tts audio, which i suggest you to choose a tts model that is the same gender and language of the rvc model you wanna use, and then convert it with rvc)

If you wanna do tts locally (if you got a good pc):

• You can find Ilaria RVC Mainline here (no guide as of right now)
• While Applio in our docs

If you don't got a good pc you can do it online:

• Ilaria RVC Zero (Running on A100 GPU, free fasted rvc on cloud) and the guide

• Use Applio UI Colab (with google colab T4 free daily limit gpu)

• if you don't wanna use edge tts, you could try another tts ai from [our tts index](https://docs.google.com/document/d/12hCYJqNCFl6jWKoVvCxtwt2V6nSoilgi5La8dkZa1KY/edit] and use the output as an input in rvc

[karmic heron]

who summoned me

am I here to help

[odd monolith]

how do you make your own ai model?

[karmic heron]

I know

im not helper but i know

[karmic heron]

Oi

Deixa eu ver

hummmm

tu pode me mostrar oq tem dentro da pasta de modelos? @lilac canopy

[low shard]

the guy wanted help for gpt so vits but changed idea as gpt so vits doesn't support portuguese

[lilac canopy]

ta em portugues

[karmic heron]

oh

[lilac canopy]

eles n aceitam portugues

😦

queria clonar uma voz em portugues

e usar como tts

[feral bough]

Hello!

[karmic heron]

no gpt-sovits ainda não tem infelizmente

[lilac canopy]

Posso te chamar no privado rc4?

[low shard]

make an application lol

[karmic heron]

i will

[lilac canopy]

to passando dor de cabeça

ja q tu manja

pode dar uma ajuda?

[karmic heron]

Sim

[lilac canopy]

posso chamar dm?

[karmic heron]

sim

[feral bough]

Im new to ai and want yo make a project sekai ai cover. Can someone help me?

[low shard]

check https://docs.aihub.wtf/essentials/how-to-make-ai-cover/

[feral bough]

Oh thanks! Do I just click on the link

?

[rare goblet]

Ayo? @feral bough level 1 !!!

[low shard]

thats the docs link, the guide, do you got a good pc?

[feral bough]

Uhh, do I have to? Im on my iPad 💀

Maybe that’s impossible then

[low shard]

you can use the cloud ways

use Ilaria RVC Zero

Which is the fastes free cloud wayt using zerogpu (A100)

you could also use the google colabs in the docs https://docs.aihub.wtf, but google colab gpu has a daily limit of 4 hours and T4 is slower so id recommend Ilaria RVC Zero

[proven hill]

second that

use me

[feral bough]

Thanks, I’ll try

Nope, that won’t work. I have like 2GB free space

But thanks anyways! I’ll probably give up making an ai cover on my iPad, I should get a pic first.

PC*

[undone kindle]

Hello, I just downloaded this but i cant find where to add models

[karmic heron]

click on edit

[undone kindle]

Lol im dumb thankyou

[karmic heron]

youre not dumb, you just didnt know and yw

[undone kindle]

and the input only has system, file and none

Oh nvm, it i refreshed it and it gave me the other options

[timid crow]

Does anyone know where I could download instagram stories into a wav file?

[undone kindle]

I tried it but it says there is something wrong and to check the code but im not sure what im looking for

Im thinking that the model im using isnt the correct file type but i dont know how to get them

[wispy lodge]

It's correct, but this error appears if you use onnx model that was exported with old version

Try uploading pth model

[undone kindle]

Hmm, where do you find those? I followed a tutorial how to get them on the okada version. is it the same thing?

[wispy lodge]

You can get them from #1175430844685484042 or weights.gg

But which model did you download?

[undone kindle]

Hmm i can find it real quick i think

[undone kindle]

this one

i tried turning it into a onnx file cause i saw that it might help solve delay issues

[wispy lodge]

Ah, then you need to upload it and turn it into onnx again

[undone kindle]

I have it like this right now

[wispy lodge]

Replace it with pth model, then export to onnx

[wispy lodge]

And choose your gpu

[undone kindle]

Is the pth model like the original download?

[wispy lodge]

Yeah

[undone kindle]

Ahh much better, thankyou! Hate to keep bugging but is there a way to lower the sensitivity cause it seems like the voice is trying to react to even like the sound of my typing lol

[wispy lodge]

You can use sup2 for noise suppression, or increase S.Threshold that controls the voice activation limit

[undone kindle]

Ahh so much fun~~

[rare goblet]

Ayo? @undone kindle level 3 !!!

[worthy crest]

i need help

the console of voice changer keeps saying

[Voice Changer] Waiting generate pipeline...
[Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline...
[Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline...
[Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline...

[undone kindle]

Is there a setting to use to make it so you dont hear the feedback?

[pastel oak]

install failed

delete "pretrain" and "model_dir" and then start the app again

also make sure not to have antivirus program active, it could be deleting files

[pastel oak]

set monitor to none if you dont want to hear yourself

[undone kindle]

It is at none but also, how can i test if someone can hear me online

[pastel oak]

you can use discord mic check to make sure the thing is working, and then just assume its working

[undone kindle]

I dont know why its not working

[wispy lodge]

Did you also set VB-Audio cable as output in voice changer?

[undone kindle]

i think so

[wispy lodge]

Hmm, I think someone also had similar problem. Could you try changing audio to server?

[wispy lodge]

That should work, yeah

[undone kindle]

Hmm its not making any sound at all now

[wispy lodge]

Strange. Does it write anything in command line window?

[analog obsidian]

vb cable is really a pain in the ass, download virtual audio cable instead tbh

[undone kindle]

I can try that, is there a link?

[analog obsidian]

https://software.muzychenko.net/freeware/vac470lite.zip

[fallow nebula]

hey i just downloaded the voice changer but the gpu is not showing up. Right now its using my cpu and i canr

cannot figure out the problem

[pastel oak]

Whats your GPU name and does it say DirectML or onnx cuda on the top right

[fallow nebula]

onnxgpu-cuda

i have a radeon rx 6950 xt

[rare goblet]

Ayo? @fallow nebula level 1 !!!

[pastel oak]

you downloaded the wrong version of wokada then

[fallow nebula]

damn

[pastel oak]

however since ur on AMD i would recommend you actually download the modified version which runs really well on AMD gpus

https://rentry.co/bettervoicechangerguide

follow amd steps

and step 6 virtual cable

[fallow nebula]

ok thank u for help 🙂

[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
• EasyGUI, by rejects 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

[undone kindle]

I tried restarting my computer and the app didnt save my models which is fine but now it wont let me pick my gpu

[flint vapor]

Which gpu do you have? and which version did you downloaded?

[undone kindle]

the browser version

[proud anvil]

Help

[pastel oak]

thedy were asking if you downloaded DML or cuda version

[flint vapor]

So are you using on google colab?

[pastel oak]

No theyre using the modified wokada version

[undone kindle]

I think so. Using it cause my computer wasnt handling the app that well

[flint vapor]

Oh ok ok😁

[flint vapor]

This modified version have the same options like the classic okada? with the download for amd versiona and cuda?

[wispy lodge]

Wdym it didn't save models? 🤔 Are you sure you're running the same version?

[flint vapor]

I don't remember well

[flint vapor]

Because if like this maybe you downloaded the cuda version

and you have amd

[pastel oak]

the modified version shows your GPU name even if you have the directML version 😄

[flint vapor]

Ok 😄 thank you

[undone kindle]

Like i had a couple downloaded, restarted my computer and it's not there now