✨│ai-help

[uneven parrot]

oa aq

az mı

[elder cloak]

Oh Okay

[wispy burrow]

Remove as much silence as possible and any unwanted noises.
And if you have many short files, make sure none of them are shorter than 1 second.

Yes. Model fusion is possible.

[vapid gust]

Is there any software to remove noise?

and silence

[wispy burrow]

UVR (Ultimate Vocal Remover) can remove background noise, echoes and reverb.
Audacity to remove silence.

That's what I use.

[rare goblet]

@brittle wing has reached level 1.

[long rose]

whres the link to hugging face mp3 to model

the easygui

[brittle wing]

So I'm trying to train this model but when I click on Train, it dosn't complete the proccessing but shows me this 训练结束, 您可查看控制台训练日志或实验文件夹下的train.log anyway

Nothing in my weights folder after that

[wispy burrow]

Check the cmd.

[brittle wing]

What for in the cmd?

[rare goblet]

@brittle wing has reached level 1.

[wispy burrow]

Command Line or Command Prompt or Terminal.
That black box with text.

[brittle wing]

It showed this:

runtime\python.exe train_nsf_sim_cache_sid_load_pretrain.py -e Riddle -sr 40k -f0 1 -bs 1 -te 20 -se 10 -pg pretrained_v2/f0G40k.pth -pd pretrained_v2/f0D40k.pth -l 1 -c 1 -sw 1 -v v2 -li 105

[wispy burrow]

that doesn't tell me much...
What are you using?

[brittle wing]

Wait using in what?-

I'm so sorry I'm new to this and don't know the specifics 😭

[loud berry]

hey does anyone know how to remove noise from a recording?

[wispy burrow]

What "program" are you using to train a Model?
RVC? Mangio-RVC? Applio-RVC?
Or is it a Colab? Like RVC Disconnected?

[brittle wing]

Mangio-RVC-v23.7.0_INFER_TRAIN.bat

[wanton swan]

Hello

[wispy burrow]

UVR (Ultimate Vocal Remover).

[wanton swan]

Does anyone know how to use Bark TTS on applio without waiting too long? While also having an audio render?

[wispy burrow]

Ah Mangio, ok.

[wispy burrow]

Normally a lot more should be displayed when training...

[wanton swan]

?

[proper shale]

apparently uh

bark

does not work

so you'd have to go with ElevenLabs' TTS which is reeeally good

or go with edge which is already ok

[loud berry]

didn't work. Should I dm you to show the case?

[wanton swan]

How do you make edge tts, sound natural like bark ai?

[rare goblet]

@wanton swan has reached level 3.

[wispy burrow]

What Model did you use?

[proper shale]

Honestly I'm not sure, Edge sounded natural whenever I tried it so

¯_(ツ)_/¯

[wanton swan]

LOL

[loud berry]

Can someone do it for me? its short audio

[spring turtle]

is this the best settings for mdx23c

also another question, whats the best possible ensemble you can use with UVR 5.6 rn

[wispy burrow]

Render it as wav, not mp3.

[spring turtle]

is there a huge difference in quality?

i just do it cuz mp3s are smaller

[vapid gust]

unless u want bits of quality most nesscary

but there’s like no difference

[proper shale]

for dataset shit

it does

make a big difference

[vapid gust]

I doubt it

[wispy burrow]

Is it for a Dataset or an audio to infer?
If it's the latter mp3 is fine.

[proper shale]

Hearing wise, you can't almost tell the difference between a 320 kbps mp3 and a wav. But RVC can. And RVC will apply that slight compression noise to all final results

And it will make your model worse

[spring turtle]

its for infer but if wav is better for infer then ill use wav

[loud berry]

pleaseeeeeeeeeee

[proper shale]

Ig that's it then

[vapid gust]

What happens if I do not remove silence ina audio clip in a dataset

What will happen

[wispy burrow]

You can get certain errors when trying to during Pitch Extraction.
But it'll affect your training as well.

[noble basin]

Does this stuff work on iPads?

[dense crane]

No

[vapid gust]

once I didn’t remove silence

on like 30mins dataset

[noble basin]

Ok 😦

[vapid gust]

and nothing happened

[noble basin]

: (*

[vapid gust]

it was fine

[dense crane]

It doesn't run on iPads and barely runs on Macs

[wispy burrow]

Fine. But it's very likely it would have turned out even better if you did.

[timid echo]

use the adobe https://rb.gy/qm9tj

is everyone doing cloing not having to pay? (like for colab)...it's been a minute since I been on here and the old links don't work

[wispy burrow]

#1163883017446629508

[timid echo]

you have to pay for all of them?

[rare goblet]

@timid echo has reached level 1.

[wispy burrow]

All should be free.

[flat jungle]

nevermind i found out where it is

the model i mean

[wispy burrow]

If the Model can't be found in #1159289700490694666 try searching on https://www.weights.gg/

[flat jungle]

thanks

[dusky plaza]

ayo the fuck is this

[proper shale]

itll make your model bad

g will screw up and generate silence and... it wont

it wont trian

[autumn hill]

what is blud on about (it didnt install shit)

nvm fixed it lols

[weary fable]

Okay so I am basically using Mangio-RVC-v23.7.0_INFER_TRAIN to train a model but for some reason.. It's not giving me the output information... I also cant send in pngs cause idk whats going on my end

[sand iris]

downloading model files

but colab is skill issuing

[ancient latch]

Can't get past "AttributeError: 'NoneType' object has no attribute 'tobytes'".

Apparently it is caused by Gradio. Is there any other GUI available?

[brittle wing]

If i just want to use models other people created how should I go about it? I want to make text to speech. For example if I download Matpat's voice in the model's here how do I turn it to t2s using something like kaggle/colab?

[wanton swan]

How come I see models with 1,000 Epochs? Does having a high amount make it better?

[wispy burrow]

You can post screenshots here once you're level 1.
For Mangio, (and other versions of RVC) check the cmd.
It usually gives more info.

[wanton swan]

?

[wispy burrow]

RVC Voice Models and TTS Voice Models are not the same.
You can first generate a TTS voice and then infer that with RVC voice.

[wispy burrow]

Not necessarily, no.

[wanton swan]

Those are just the amount of training triles

[wispy burrow]

And if it's trained too much the Model gets worse.

[brittle wing]

Okay then how would I go about infering (if that's the right word idk) an audio from a model without installing on my machine?

Thank you btw

[wanton swan]

So If I have 4 mins of a characters dialouge and train it on 250 epochs is that okay?

[wispy burrow]

It should be ok.
You might even go for more Epochs with a Dataset that small.

[wispy burrow]

To change one voice to another you'll want Colab for "RVC".
Not RVC Disconnected as that's for training only.

[brittle wing]

please can i get help for running RVC in Mac i am complete newbie

[rare goblet]

@brittle wing has reached level 1.

[proper shale]

check guides

and if theres an error please specify

[slow hornet]

hello ive downloaded the AMD version now the promt says GPU (dml) and 5 selections starting from cpu to gpu0 - 3 now they are all grey im trying to click gpu0 but it wont turn red

its using gpu1 now which is my apu i want it to use gpu0 which is my dgpu

[chrome solar]

I'm trying to download Applio and I followed the instructions and everything but once I get to the actual installation of Applio I get this:
Installing dependencies...
Invoke-WebRequest : Forbidden
You don't have permission to access this resource.
Apache/2.4.38 (Debian) Server at frippery.org Port 443
At line:1 char:1

• Invoke-WebRequest -Uri https://frippery.org/files/busybox/busybox.exe ...

•   + CategoryInfo          : InvalidOperation: (System.Net.HttpWebRequest:HttpWebRequest) [Invoke-WebRequest], WebExc 
   eption
    + FullyQualifiedErrorId : WebCmdletWebResponseException,Microsoft.PowerShell.Commands.InvokeWebRequestCommand
  

'busybox.exe' is not recognized as an internal or external command,
operable program or batch file.

Extracting the runtime.zip file...
Exception calling "ExtractToDirectory" with "2" argument(s): "Could not find file
'C:\Users\josep\Desktop\AICovers\Applio-RVC-Fork\runtime.zip'."
At line:1 char:63

• ... ileSystem ; [System.IO.Compression.ZipFile]::ExtractToDirectory('runt ...

•             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  

  • CategoryInfo : NotSpecified: (:) [], MethodInvocationException
  • FullyQualifiedErrorId : FileNotFoundException

Could Not Find C:\Users\josep\Desktop\AICovers\Applio-RVC-Fork\runtime.zip

Applio has been successfully downloaded, run the file go-applio.bat to run the web interface!

Batch process completed with exit code 0

What do

[fathom remnant]

Try running go-applio.bat if you have not already

[chrome solar]

I opened it and put Nvidia ("1") and got this:
Starting Applio with runtime for Nvidia support (you must have it installed)...
The system cannot find the path specified.
Press any key to continue . . .

[fathom remnant]

Yep thats where im at

[rare goblet]

@fathom remnant has reached level 1.

[autumn hill]

whar

[chrome solar]

So we both beat then lol

[fathom remnant]

Yep

How do you install Nvidia support runtime?

[autumn hill]

its been so long since ive used applio lol now my shit isnt working

[fathom remnant]

So we are all stuck

[chrome solar]

I think there website maybe

[rare goblet]

@chrome solar has reached level 1.

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇫🇷 France

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

🇩🇪 Germany

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

🇮🇹 Italy

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

[chrome solar]

I've been using replay since it was so nice and then it broke so I'm trying to get applio back and of course it doesn't wanna work now :/

[fathom remnant]

I looked but ill look again rn

[fathom remnant]

I think its this

https://developer.nvidia.com/nvidia-container-runtime

it is

tensor is one of the applications listed as using it on the website

nope

idk

[fathom remnant]

Its the applio manager

[rare goblet]

@fathom remnant has reached level 2.

[solar geode]

hii, i use rvc disconnected to make a model but i have only the index and not the .pth😭

[fathom remnant]

\Applio-RVC-Fork\Applio-RVC-Fork\go-applio-manager.bat

then option 3

yep

worked for me

[autumn hill]

T_T

[glad zealot]

Did you download the pth from the weights folder?

[solar geode]

I only see the model folder and inside there is only the index and another zip file that I checked but nothing important appears

[glad zealot]

No weights? There should be a weights folder somewhere that has the pth filez

[solar geode]

idk i just follow this doc https://docs.google.com/document/d/1XuxQYiqEhYrdYeCZRRLrmV_ciMKo0bV-jTCGHu_-5Cc/edit#heading=h.hg5mjyg07cdn

[rare goblet]

@solar geode has reached level 1.

[fathom remnant]

Is mangio-crepe Crepe-full ?

becasue the faq says Crepe and Crepe-full

[solar geode]

Yes I think so

[fathom remnant]

Ok

[autumn hill]

man.

ive been sitting here for like 30 mins TT

what do you want from me...

i have given up,.

[chrome solar]

🙏 TY

[autumn hill]

back to mangio i go

[vestal vault]

Applio keeps erroring saying I don't have Torch

what's Torch?

[brittle wing]

i can't render ANYTHING in applio without this error...

TTS seems to work, but converting actual vocals gives me this error, is there any fix to this?

oh NOW it wants to work.... crisis averted ig

[fathom remnant]

What save type should I use?

for the model?

[rare goblet]

@safe crow has reached level 3.

[violet heron]

You can always use colab

[safe crow]

People in this channel said it work only for inferences not training

Apparently you can't train with AMD at all

[safe crow]

I will have to learn how it works indeed

Sorry I tried to modify my original message on my phone and it got deleted oopsy

[fathom remnant]

What model save type should I use?

On Applio

[safe crow]

What do you mean with this?

Oh yes. I have linux I will try tomorrow with it thanks for bringing a lil hope in my heart

[lament oriole]

Does training rvc need a high spec pc or just need the cpu?

[proper shale]

Needs a recent NVIDIA GPU

[violet heron]

High spec

[proper shale]

3060 12G will do

[violet heron]

You can always use colab

[proper shale]

Yea, like I do

[lament oriole]

Oh thank you

[rare goblet]

@lament oriole has reached level 1.

[autumn hill]

im going crazy

[half cove]

Crazy?

I was crazy once

[autumn hill]

poopmaster..........

yeah thats what im doin

dunno why its refusing to work when its worked before

sick

[sand iris]

[bright moat]

I have a model downloaded how do I use it?

[fathom remnant]

#1159949278270193734

[modest spruce]

Is this happening to anyone else?

[fathom remnant]

What are you using?

[modest spruce]

Applio

[fathom remnant]

Yea not with me bc im training with applio rn

[modest spruce]

I have all the shit in correctly

Is there any other way I can make a cover?

[fathom remnant]

You using rvc v2?

[rare goblet]

@fathom remnant has reached level 3.

[modest spruce]

yes

[fathom remnant]

Could use aicovergen with easy gui

Its what I used before I got my new pc

It worked most times

but now I go into makeing so I use Applio

[autumn hill]

retried applio

[burnt ermine]

i keep getting an error saying there is not enough gpu memory available

specifically it says "RuntimeError: Could not allocate tensor with 223323132 bytes. There is not enough GPU video memory available!"

or something along those lines with a different amount of bytes

how do i lower it? do i just use smaller audio files?

[rare goblet]

@burnt ermine has reached level 1.

[burnt ermine]

good to know

[uncut vessel]

Hello guys i want to train voice with applio but when i launch it (Nvidia) it say to me i dont have the module called "torch" i installed it via commands , and every module that apllio console says i installed it , but im stopped to the module "fairseq" because the python console say to me " error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/" , i installed it but it dont work , i want to know why i every time i want to start it i have to install the module by my self and it dont go automatically , sorry for my bad english guys , if someone want to help me , u can chat in private too with me , Thanks all , PS: I tryed to use the RVC Train Models but its the same error

[burnt ermine]

well unfortunately for me i happen to be an idiot. sorry lmao

does this also get used for making covers? im not trying to train one

sorry if im asking the stupidest questions known to man im not that bright

sounds good

[autumn hill]

got applio to work. forgot where u put audio files

[atomic otter]

Does anyone know why RVC v2 keeps loading infinitely when I click on any option, whether to convert or import the audio?

[fathom remnant]

Do you have more info like are you using aicovergen and other things like that

[atomic otter]

local host

[rare goblet]

@atomic otter has reached level 1.

[fathom remnant]

What are you using?

[atomic otter]

I'm using run_easiergui.bat locally on my computer, when I try to do anything it keeps loading infinitely for seconds

Mangio-RVC-v23.7.0_INFER

[fathom remnant]

I dont know about mangio

[viscid moss]

For inference u can use AICoverGen, Hugging Face and WordBand bot

[brittle wing]

i need some help

it's not giving me my index file when i use rvc disconnected v2

i ran feature extraction already

[viscid moss]

dataset was imported successfully?

[brittle wing]

yeah i have the pth file

[rare goblet]

@brittle wing has reached level 2.

[viscid moss]

How long is the dataset?

[brittle wing]

5 minutes

[viscid moss]

hmmmmm

[brittle wing]

i loaded it into the rvc gui

it seems to work

i'll convert something

[viscid moss]

If the model is already trained, you can restart the environment and generate only the index this time

[brittle wing]

alright cool i'll do that

[vocal fern]

Can somebody help me? When I run cmd to open the cover I get this.

DEPRECATION: omegaconf 2.0.6 has a non-standard dependency specifier PyYAML>=5.1.*. pip 24.0 will enforce this behaviour change. A possible replacement is to upgrade to a newer version of omegaconf or contact the author to suggest that they release a version with a conforming dependency specifiers. Discussion can be found at https://github.com/pypa/pip/issues/12063

[merry sand]

@tranquil cliff

[tepid nimbus]

Could someone send the colab link?

I found it, nevermind.

[tranquil cliff]

Training?

[tranquil cliff]

send a photo of everything so I can see well....

[tepid nimbus]

No, just the covers.

[tranquil cliff]

I recommend ilaria rvc v2 which can be found on huggingface.co .

I'm sending a guide now

[tepid nimbus]

Thank you! ^^

[tranquil cliff]

in English it is not available...only in Italian

but ilaria rvc is very easy to understand

[merry sand]

How about this?

[rare goblet]

@merry sand has reached level 2.

[tranquil cliff]

I mean the code block where you put the zip file

[tepid nimbus]

I'll give it a shot, and I could use translation as well. ChatGPT's is pretty good as far as I remember.

[tranquil cliff]

Okay

https://docs.google.com/document/d/1S2X-IRDZZVL6_7xBD-0jx1PEXKuY3yV3y1lkBn16gWc/edit?usp=sharing

here is the guide

[tranquil cliff]

anyway I think you didn't put the file correctly...from what I see

[merry sand]

It's on "Load Dataset".

[tranquil cliff]

I know...I wanted to understand if you had placed the zip file correctly

The error says the file does not exist or was not found

[merry sand]

Is it the dataset path I was suppose to put it in?

[tranquil cliff]

Yes

I want to see if it's right

[merry sand]

On Set Training Valuables. Does the path have to be from File Explorer or from the Drive folder file?

[tepid nimbus]

https://docs.google.com/document/d/19zRi_7dVIexe5ExiKD1OmkNsZOG-ACdaPHPrSVCDonI/edit?usp=drivesdk

@tranquil cliff
Here's the English translated, in case anyone needs it.

[tranquil cliff]

on your left there is the file browser where you have to press drive-mydrive-Rvc disconnected and you load the zip file there

Thank you

I'll give you the credits

[tepid nimbus]

No problem. :)

[tranquil cliff]

Can you give me access so I can fix something?

[tepid nimbus]

I did, just got the email.

[rare goblet]

@tepid nimbus has reached level 2.

[tranquil cliff]

Thank you

[merry sand]

Well I did load it, but while on Set Training, what Path to Training folder do I have to link?

[tranquil cliff]

you send the photo and I'll tell you the steps

[tranquil cliff]

nothing here

leave it as it is

[merry sand]

Oh.

[lusty willow]

Colab appears to be working for me today. 👍 However, they've added a "ngrok_authtoken"? Weird. Appreciate the help anyway. 😁

[blazing willow]

Hi There
I Need Voice Models
Can Anyone Help Me

[low shard]

#1159289700490694666 you could search them in this channel, or in https://weights.gg

[placid light]

Are there any disconnect prevention methods out there?

It said they're not hard to come by, but I don't seem to find one.

Dead chat

[cobalt oyster]

@astral jungle How r u

[placid light]

Wtf is this? lol

[placid light]

Is this what I should do to prevent being disconnected? Spam this box here?

[indigo sand]

hey
need help
on RVC v2 Disconnected, how do you download the model once it's trained?
where does it save?

[indigo sand]

you'll get a capcha box

usually happens around 175 epochs for me

based on about 30 minutes of vox data

there are ways, but they often get patched

[astral jungle]

I'm Fine

[sturdy yacht]

ive updated my gpu driver now no ai program wants to utilize cuda my cuda toolkjiut or pytorch is nuked or something

[hard gust]

Hi guys, what's up
is it possible that after 200 epochs every epoch takes 5 minutes when before 200 epochs it was like 1 to 2 minutes?
I mean it's not a short dataset but why did it change 🤔

[proper shale]

GPU usage

if you're on local this happens

[hard gust]

Is there a way to make it better again?

Cuz nothing is opened

it shouldn't be that slower

[placid light]

Hey, err, if I want to save the unfinished model to google drive for now and resume the training later, should I use this option?

That's why I got confused because no one seems to be responding immediately.

[proper shale]

Yes

[placid light]

Is it necessary to manually enter the step and epoch count?

[proper shale]

No

[placid light]

So that means I should leave everything on that cell alone.

[proper shale]

Yeah

Just run it

And it'll save your stuff

[placid light]

Thank you. I wouldn't make my own model without your help. ^_^

[proper shale]

Awww, no problem! I'm glad to have helped you. :)

[placid light]

Does this graph here indicate how bad my model is?

[proper shale]

Check g/total

And no - lower loss is apparently better

[placid light]

Am I doing it right? (I'm at about 100 epoch btw)

[rare goblet]

@placid light has reached level 8.

[placid light]

At this point, should I leave the colab page alone or should I keep on touching stuff to prevent being disconnected?

[stoic cloud]

Seems so

It's going down

That's good

[placid light]

I'm at 175 epoch now, and it seems like I have 1 hour of training left.

[astral jungle]

I'm fine

[placid light]

I'm at 200 epoch now and it seems like they're looking good.

Am I correct?

This is just my first time training a model.

[glad zealot]

yes

[forest zenith]

whats the formula based on loss graphs and epochs to know when i need to stop training

[rare goblet]

@forest zenith has reached level 1.

[forest zenith]

i dont remember

[proper shale]

Go on scalars tab

Btw what Colab is that, TensorBoard hasn't been working for a while

[placid light]

It's RVC Disconnected, running on Opera GX.

[spiral finch]

guys

have any reference for the Gradio got Banned by google collab ?

[brittle wing]

hello, is there still any guides for vocal and instrumental isolation in UVR? in the last server it was pinned on the help channel and it was very helpful

[copper pebble]

https://discord.com/channels/1159260121998827560/1159588254220288121

this is a good one

[brittle wing]

thank you

[copper pebble]

-colab

[azure marsh]

Colab suggestions for @spiral finch

RVC

• AI-Cover-Gen-No-UI (inference only)
• RVC v2 Disconnected (training only)

🌐 Translated colabs

• PT-BR: -colabs_br

Audio Separation/Isolation

• MDX Net Colab
• UVR arch colab 5.0.2/c1.04
• KaraFan colab

[copper pebble]

these are the new colabs

[spiral finch]

thanks

[brittle wing]

ah unfortunately this isnt what I'm looking for because its for low-end rigs
im trying to do it locally with a decent GPU

[copper pebble]

gotcha

[copper pebble]

https://discord.com/channels/1159260121998827560/1159636677178953802

heres a complete guide and there is a local guide aswell

dunno if theres what you looking for tho

[spiral finch]

uhh how to use ?

[copper pebble]

wt?

[brittle wing]

i just checked it doesnt have what I'm looking for :<

oh well

I screenshotted the pinned message in the old server but I think its already outdated now

[copper pebble]

mhhh thats outdated i think

[meager mulch]

where i can download voice changer?

[copper pebble]

not sure tho

[copper pebble]

https://discord.com/channels/1159260121998827560/1159636677178953802

complete guides here

[copper pebble]

what?

[spiral finch]

this

[copper pebble]

oh yeah sorry

[copper pebble]

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

heres a guide for RVC disconnected

[copper pebble]

for more guides here

[spare solstice]

sorry off topic but what happened to the og ai hub server

[copper pebble]

some assholes reported some audios that were copyrighted and discord banned it

this is what i understood

[spare solstice]

uh man really sad i just hope they have backup of everything

@copper pebble can u help with something

[copper pebble]

i hope so

@spare solstice

[spare solstice]

bro i am getting the error in which it doesn't register the model

[rare goblet]

@spare solstice has reached level 1.

[spare solstice]

btw i am using kind old collab of rvc

can i post links here?

[copper pebble]

yeah

[spare solstice]

https://www.youtube.com/watch?v=bP8AMf20MAY

from this video

i mean the collab in this one

it worked fine with almost all models except the one i am trying right now

so anyway my question is, is the up to date collab important?

[copper pebble]

pretty old

-colab

[azure marsh]

Google Colabs

RVC

• AI-Cover-Gen-No-UI (inference only)
• RVC v2 Disconnected (training only)

🌐 Translated colabs

• PT-BR: -colabs_br

Audio Separation/Isolation

• MDX Net Colab
• UVR arch colab 5.0.2/c1.04
• KaraFan colab

[copper pebble]

heres the new one

(inference)

[copper pebble]

@spare solstice

[tiny granite]

every time i click on the post in #🔍│find-models it just tells me you dont have acess to this link

[spare solstice]

okay lemme try this then i will tell you

[copper pebble]

which one?

[tiny granite]

AI HUB (RIP) POST

[rare goblet]

@tiny granite has reached level 1.

[spare solstice]

ayo @copper pebble why is this one so scary looking 😭

any tutorial or something ?

[tiny granite]

i need the link from the post to put it in the colab

[copper pebble]

yeah i think lemme look for a mom

[copper pebble]

https://youtu.be/3Wa5DsJAv3c

theres a tut

[spare solstice]

bro

i think this one is for training

not conversion

[copper pebble]

jesus ahahha i got the wrong link ahah

[spare solstice]

xd

[copper pebble]

ok im here

i usually use ilaria rvc 2.0

not colab

and thats pretty good

[spare solstice]

the collab gives me mr robot feeling

but anyway whats your solution

[copper pebble]

https://docs.google.com/document/d/19zRi_7dVIexe5ExiKD1OmkNsZOG-ACdaPHPrSVCDonI/edit?usp=drivesdk

this a guide of ilaria rvc

did a lot of covers with it

[spare solstice]

the ilaria is really in love with you

[rare goblet]

@spare solstice has reached level 2.

[spare solstice]

with me too apprantly

[copper pebble]

lmao

[spare solstice]

aight rick lemme see

but please if theres a tutorial for that collab let me know

[copper pebble]

https://huggingface.co/spaces/TheStinger/Ilaria_RVC_2.0 idk if theres one in the guide but this is the link

[copper pebble]

yeah sorry as soon as i can

[spare solstice]

aight see ya rick

take care

ilaria

there we go another reactio

xd

[copper pebble]

is the bot tho

[stoic cloud]

Hi

[copper pebble]

hey

do we have a guide for ai-cover gen?

[stoic cloud]

Don't think so

But it's not that hard

[copper pebble]

yeah i know its for ali

going to workout rn so ill be afk for a while

[stoic cloud]

Alrighty

[meager mulch]

why my rvc is not responding after i start

[rare goblet]

@meager mulch has reached level 1.

[lament eagle]

Is rvcDisconnected based on Applio?

[meager mulch]

wdym i dont now what is that

[lament eagle]

training fork

It's for training model

[meager mulch]

i use default model

[meager mulch]

help me sir

[lament eagle]

Sorry I don't use that conversion software

I was asking about different question

[tall grove]

i cant hear myself

how can i fix

[sand iris]

#🔍│help-w-okada

set "CHUNK" to something bigger

and screenshot cmd window if that doesnt fix

[sand iris]

Should be working now, pinned messages in #🔍│help-w-okada have instructions on how to setup settings

[high night]

@sand iris btw mate does bigger chunk mean less delay?

[sand iris]

Nah, opposite

[high night]

So less = good

[sand iris]

yesn't

[high night]

Wdym exactly

[sand iris]

Voice changer kind of brute forces RVC

Highest CHUNK setting is 5s

[high night]

So what should be the ideal chunk

[rare goblet]

@high night has reached level 24.

[sand iris]

That's what RVC works on, 5s chunks. Anything lower just makes it infer on overlapping audio segments

I think quality suffers from lower chunk, but it's not quite as drastic unless you get under the 100ms mark. And that's settings that become wayyy too intensive

[tall grove]

it still dosent work, dosent register my mic ig

[high night]

What chunk do you set

[high night]

Do you have virtual cable downloaded

[tall grove]

i have it

\but i cant hear myself on the app

[sand iris]

170ms - 220ms, my system can't go any faster without struggling

[high night]

then it's preety weird

[sand iris]

what are the settings used for audio devices

[tall grove]

this?

[spare solstice]

ayo i need help

[sand iris]

Which GPU does your computer have ?

[tall grove]

shitty nvidia GPU

[sand iris]

too old

Won't work unfortunately

[spare solstice]

@sand iris

[sand iris]

what do you need ?

[tall grove]

and if i use CPU

[tall grove]

?

[spare solstice]

this is that rvc collab one

[sand iris]

Delay will be high, but it could work. Not a really good experience

[spare solstice]

so do i run that mdx net cell?

[sand iris]

Haven't used RVC disconnected myself

[spare solstice]

also in url i have to put the model link?
and what to put in dir

[sand iris]

Not sure here

[spare solstice]

...

[sand iris]

There's this but they import the model differently

https://docs.google.com/document/d/1XuxQYiqEhYrdYeCZRRLrmV_ciMKo0bV-jTCGHu_-5Cc/edit#heading=h.hg5mjyg07cdn

URL should be a link to a .zip file for a model

dir_name is the name you want to give the model

[spare solstice]

is this about cloning voice ?

[sand iris]

yes

[spare solstice]

i wanna convert

like existing audio to model's voice

[sand iris]

Yeah. Just gotta supply it a model

[spare solstice]

listen dude

this is for training voice?

oh it is i read it

anyway i wanna make a cover

[sand iris]

Oh my bad

I'm stupid lol I sent the wrong thing

[spiral finch]

is that have any gradioGUI on google collab backup version right ?

[tiny granite]

-colab

[azure marsh]

Google Colabs

RVC

• AI-Cover-Gen-No-UI (inference only)
• RVC v2 Disconnected (training only)

🌐 Translated colabs

• PT-BR: -colabs_br

Audio Separation/Isolation

• MDX Net Colab
• UVR arch colab 5.0.2/c1.04
• KaraFan colab

[forest zenith]

what smoothness set on tensorboard? it changing lowest point of graph with increasing smoothness to the brim

[vapid gust]

0.99

Check the outliners off

[red kayak]

dont listen to these guys

u need outliners on

otherwise u wont see shit

[vapid gust]

or the other way around I forgot

[forest zenith]

anyway trust 0.999?

[small zodiac]

Can anyone get in a gc and help me how to do this RVC thing bru

😔

[chrome moat]

haloo ive got an issue with mine all of the sudden, for some reason I can hear it and it sounds perfect but it wont play through the cable output so noone else can hear it

[spiral osprey]

How do I make the voice more faithful with almost no flaws or few flaws? Do you have to change the pitch to do this? If so, what value should I use?

[wispy burrow]

#🔍│help-w-okada

[chrome moat]

oh thank u

[wispy burrow]

Which version are you using?

[winter stirrup]

hey, is there more recent / better collab based interface for RVC than this one? Rejekt's EasyGUI v2 (adapted from Mangio-RVC-Fork 💻)

[wispy burrow]

WebGUI's are banned on Colab.

[winter stirrup]

iim on colab plus

and i';m using it but i dpnt know where to put my older models to get them to appear in the ui

[wispy burrow]

I know most people seem to use RVC Disconnected.

[spiral osprey]

How do I make the voice more faithful with almost no flaws or few flaws? Do you have to change the pitch to do this? If so, what value should I use?

[winter stirrup]

thanks, looks like a step backwards in terms of usability though.

[rare goblet]

@winter stirrup has reached level 1.

[wispy burrow]

RVC Disconnected is only for training but I jrst realised you probably wish to infer audio.

[winter stirrup]

yes, and this has both: https://colab.research.google.com/drive/1bjSxP8tjergb7IzEOwmgQWZSeHseziAn however i dont know where to put the models i created earlier

[spiral finch]

but the other version is shitty

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇫🇷 France

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

🇩🇪 Germany

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

🇮🇹 Italy

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

[small zodiac]

Mangio RVC 2

[wispy burrow]

But those with webGUI's no longer work.

[wispy burrow]

My favourite.

[spiral finch]

have any backup version of GUI ?

[small zodiac]

help me.

[wispy burrow]

Where are you stuck at?

[small zodiac]

Bro

how much ever i try

it keep syaing error

😔

[wispy burrow]

What error? What are you trying to do?

[small zodiac]

mind gettin in vc br9o

[wispy burrow]

No thank you.

[small zodiac]

u can mute

😔

jus help me wit this monstrosity

[wispy burrow]

What was the error that was shown in the cmd?
What were you trying to do?

[brittle wing]

Does anyone have a list of the best and most well trained models? I’d love to try a bunch of the one that normal users make but just wanted to see if there was a list of well known models done professionally

[small zodiac]

i run juice wrld vocals wit X AI and it keeps showin this

[wispy burrow]

There's no such list as far as I know.
Allen Poe...That name sounds familiar.

[forest zenith]

dunno? also i must use trained or added index in logs?

[wispy burrow]

added_IVF*¨.index.

[proper shale]

audios weren't selected properly

[forest zenith]

hm, thanks

[tranquil cliff]

Hi

ask me questions so I can get to lvl 15 🫠

[proper shale]

How to make AI cover

(detailed)

[tranquil cliff]

Being Italian I use Ilaria_Rvc_v2 but it also exists ai cover gen which is very useful for simplifying the whole process

[frigid wasp]

Is anyone else having issues downloading datasets onto the RVC huggingface?

All I see is this: "There's been an error."

[tranquil cliff]

I'll explain Ilaria_Rvc_v2....go to the ilaria rvc website which I'll send in the next message...then download a song from YouTube and go either to ilaria uvr (I'll send everything later) or uvr locally then upload the song. Top on the left, press download model and paste the link of the model to use.... after uploading the song and updating the model page, select the model and start everything

ps: Ilaria uvr is used to divide the base from the singer

[proper shale]

Ok but how to make AI cover on 2012 13" MacBook Pro with Intel® Core™ i7-3520M, 8GB RAM, 1280x800 TN screen, 240GB SSD and Intel® HD Graphics™ 4000 with 1536 MB of VRAM?

[proper shale]

RVC in Huggingface is not intended to be used for training.

[tranquil cliff]

Hey hey, calm down....I'm a new helper, I'll try to answer anyway

[proper shale]

Ok but how

I need ai cover. Pls

[tranquil cliff]

should I do it? jjaja lol

[proper shale]

No i wajt

[tranquil cliff]

you're playing the part very well

[proper shale]

Part??? where?? weights..g?? colab??? what do i do...

[tranquil cliff]

Ai cover gen

[proper shale]

Ok how to run ai cover gem

[copper pebble]

lmfao

[tranquil cliff]

Shhhh sono sotto prova

[tranquil cliff]

enter here https://colab.research.google.com/github/ardha27/AICoverGen-NoUI-Colab/blob/main/CoverGen_No_UI.ipynb#scrollTo=B3BqnUoePVGd

[proper shale]

ok

how to know when step done?

[copper pebble]

(meanwhile me that i didnt ever use aicovergen )

[tranquil cliff]

a ✅️ appears next to the code

[proper shale]

ok

no I cover generate

it say cant find

what i do wrong

[tranquil cliff]

start the first 3 codes

after code 4, paste the huggingface.co model link and start

[rare goblet]

Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote

[proper shale]

it say runtime disconnected.

pls give working colab.

(I'm being serious here, AICG-NO-WEBUI does not work)

[tranquil cliff]

sorry I didn't know anything

[proper shale]

Honestly, you're a good helper

You found out immediately why the path wasn't found

👏

time for more errors

Ok so

RVC quiz!

https://media.discordapp.net/attachments/1159290139609137264/1167121721338363945/image.png?ex=654cf9ce&is=653a84ce&hm=dbf248e4e0001ae06a6a73924c247dc22f5a6b5c83cdee0a0d6e0ea10bf9cbe1&

[tranquil cliff]

are you serious?

[proper shale]

What does the latest error mean? What have I done wrong?

[proper shale]

About what

About you being a good helper? Yes, I'm being serious

[tranquil cliff]

This

[proper shale]

Yeah

100% serious

Very agile thinking

[tranquil cliff]

so thank you

[low shard]

same

[astral jungle]

Heloo

[low shard]

hii

[winter stirrup]

Any idea why noone has created a easy to use cloud based RVC client in the style of Run Diffusion fr Stable Diffusion?

[astral jungle]

How are you?

[low shard]

im good,wbu?

is there smt you need help with?

[astral jungle]

No, I just came up to say hello to everyone

[gusty basalt]

hello why i cant find models older than 3 weeks?

[sand iris]

I mean there's paid platforms out there that do just that

[winter stirrup]

such as?

[sand iris]

kits.ai, voicemy.ai, among others

But it's great that stuff is open so that we have local options available

(or (ab)use cloud platforms to run them)

[winter stirrup]

cool thanks, hopefully it lets me run and train my own models

[sand iris]

-colab

[azure marsh]

Google Colabs

RVC

• AI-Cover-Gen-No-UI (inference only)
• RVC v2 Disconnected (training only)

🌐 Translated colabs

• PT-BR: -colabs_br

Audio Separation/Isolation

• MDX Net Colab
• UVR arch colab 5.0.2/c1.04
• KaraFan colab

[sand iris]

There's also those which aren't paid

and/or the hugging face pages that some people have

[winter stirrup]

im cool with paid, thanks!

[vapid gust]

can someone help me

with isolation vocals

can someone do it for me im having trouble

[mystic hull]

can anyone help me, i got this error while running go-web.bat on RVC1006Nvidia

Traceback (most recent call last):
File "G:\KenForce Data\RVC1006Nvidia\infer-web.py", line 7, in <module>
from infer.modules.vc.modules import VC
File "G:\KenForce Data\RVC1006Nvidia\infer\modules\vc\modules.py", line 8, in <module>
import torch
File "G:\KenForce Data\RVC1006Nvidia\runtime\lib\site-packages\torch_init_.py", line 122, in <module>
raise err
OSError: [WinError 193] %1 is not a valid Win32 application. Error loading "G:\KenForce Data\RVC1006Nvidia\runtime\lib\site-packages\torch\lib\cublas64_11.dll" or one of its dependencies.

Thank you

[sand iris]

space in folder name

"KenForce Data"

[sand iris]

put an underscore there instead of the space

[vapid gust]

can someone help me

for isolating vocals

[mystic hull]

thank you so much, i will try to rename them

[rare goblet]

@mystic hull has reached level 1.

[copper pebble]

watchu need

https://discord.com/channels/1159260121998827560/1159588254220288121

Theres a guide for UVR

[floral chasm]

ok so

[vapid gust]

could u do one thing for me

one audio file

bea cause it says that my cpu is not strong enough or something about ram

[floral chasm]

it doesnt like do anything

its open but it doesnt change my voice

[copper pebble]

nvm i cant

i almost forgot

[floral chasm]

help pls

like is there something im missing?

[copper pebble]

sorry bro im not in voice changing stuff but i can send you someone

[copper pebble]

not right now

[rare goblet]

@copper pebble has reached level 15.

[copper pebble]

#🔍│help-w-okada there should be someone

[mystic hull]

does anyone know why i always get this error while extracting RVC
or is this normal?

[teal glacier]

if applio doesn't work for me and the colab doesn't work what can I use to make models?

[mystic hull]

i still got the same error after deleting the space, is there any solution?

[safe crow]

hey im trying to run RVC real time on linux .. and I got this

any idea ?

[winter apex]

hello o/ I'm new to this and decided to create my own little ai v singer, but interestingly after 9 clear singing demo's and 350 epochs it weirdly doesn't sound exactly like me [but it's cute]

I render him through Yuma samples for my Infer base [ since I own yuma and our singing range is similar]

[bright moat]

why do all of the generators make me wait 22 hours

to refresh models

even when I duplicate the space

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇫🇷 France

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

🇩🇪 Germany

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

🇮🇹 Italy

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

[visual jungle]

trying to process data but im getting this :runtime\python.exe trainset_preprocess_pipeline_print.py "D:\AI RVC TRAINING\Mangio-RVC-v23.7.0\datasets\squid" 40000 2 D:\AI RVC TRAINING\Mangio-RVC-v23.7.0/logs/squid False
Traceback (most recent call last):
File "D:\AI RVC TRAINING\Mangio-RVC-v23.7.0\trainset_preprocess_pipeline_print.py", line 25, in <module>
f = open("%s/preprocess.log" % exp_dir, "a+")
FileNotFoundError: [Errno 2] No such file or directory: 'D:\AI/preprocess.log'

[rancid knoll]

did you preprocess the dataset?

[visual jungle]

oh

no

[flat zealot]

why does my rvc just do rvc>pause every time i launch realtime

it worked once

then just stopped working

[visual jungle]

so wait how do you preprocess the dataset this is like my first time using this

[rare goblet]

@visual jungle has reached level 1.

[rancid knoll]

you can just hit this big blue button right here 🙂

just be sure to enter the dataset path on the left

[rancid knoll]

is there no error output at all?

[flat zealot]

nope

[rancid knoll]

hmm thats weird

[flat zealot]

this has happened before

[visual jungle]

the error literaly shows up from that

[flat zealot]

and i knew why

but now i dont know

because its for a different reason

[rancid knoll]

oh

[rancid knoll]

uhm ive never seen that

[flat zealot]

not even kalomaze could figure it out last time

[rare goblet]

@flat zealot has reached level 3.

[flat zealot]

so i have no idea

[rancid knoll]

also move your RVC folder to a path that doesnt have any spaces in it

[flat zealot]

whats causing it

[rancid knoll]

that's whats causing it

[visual jungle]

like the whole root folder n stuff?

[rancid knoll]

well then im out of ideas as well 😄

[rancid knoll]

yeah

its in "AI RVC TRAINING" and it just sees the "AI" part

to a path that doesnt exist

[visual jungle]

OOOOH

[rancid knoll]

you can just pop it into your desktop

if you have any other questions i can try to help lol

[visual jungle]

we'll see

ayyy it works

[rancid knoll]

lets gooo

[visual jungle]

this is how it should look like right?: runtime\python.exe trainset_preprocess_pipeline_print.py "D:\123RVC\Mangio-RVC-v23.7.0\datasets\squid" 40000 2 D:\123RVC\Mangio-RVC-v23.7.0/logs/squid False
start preprocess
['trainset_preprocess_pipeline_print.py', 'D:\123RVC\Mangio-RVC-v23.7.0\datasets\squid', '40000', '2', 'D:\123RVC\Mangio-RVC-v23.7.0/logs/squid', 'False']
thread:0: 100%|██████████████████████████████████████████████████████████████████████████| 5/5 [00:04<00:00, 1.11it/s]
thread:1: 100%|██████████████████████████████████████████████████████████████████████████| 5/5 [00:04<00:00, 1.05it/s]
end preprocess|██████████████████████████████████████████████████████████████████████████| 5/5 [00:04<00:00, 1.49it/s]
start preprocess
['trainset_preprocess_pipeline_print.py', 'D:\123RVC\Mangio-RVC-v23.7.0\datasets\squid', '40000', '2', 'D:\123RVC\Mangio-RVC-v23.7.0/logs/squid', 'False']
end preprocess

[rancid knoll]

yep exactly how it should look like

[visual jungle]

aight good

what is the recomended pitch extraction algorithm?

[rancid knoll]

rmvpe all the way

[visual jungle]

aight

[sonic tapir]

anyone know if there is any rvc colab that still work? for coverting and not training

[rare goblet]

@sonic tapir has reached level 2.

[rancid knoll]

check the pinned message

i have no idea if it works tho

[visual jungle]

does feature extraction take forever?

[rancid knoll]

uh it shouldnt

[visual jungle]

its still at 0%

[rancid knoll]

hmmm

[visual jungle]

checking task manager i dont see any load on cpu and gpu

[rancid knoll]

it normally takes a minute at most

uhh maybe try restarting?

[visual jungle]

aight

ok it works now

[rare goblet]

@visual jungle has reached level 2.

[rancid knoll]

ah good

[visual jungle]

so i just train the model and later the index right?

[rancid knoll]

you can train the index at any time iirc

just not during the model training 😄

[visual jungle]

ah

[rancid knoll]

it takes ~5 seconds on my pc so i just hit it before i hit start training

[visual jungle]

where do i find the trained models?

[rancid knoll]

in the "weights" folder

with the filename format "{name}_e{epoch count}_s{step count}.pth"

those numbers are important so they are included in the filename

make sure this is on tho

[rare goblet]

@rancid knoll has reached level 7.

[visual jungle]

yeah its on

does it keep training until i press stop or will it end automaticly?

[rancid knoll]

it will keep training until it hits the epoch count specified by this

you can see the current epoch count in the console window

also i recommend that you install tensorboard

these files are handy to keep in your folder

[summer glacier]

I used applio and its telling me to run a file called go-applio.bat, i dont know where its located

[rare goblet]

@summer glacier has reached level 1.

[rancid knoll]

it should be in the applio root folder

[summer glacier]

which root folder

it installed off a .bat

[rancid knoll]

ah damn i forgot the folder name

it should be in the same folder as that .bat file

[summer glacier]

oh

i didnt put it in a folder

[rancid knoll]

did you run it off the desktop?

[summer glacier]

from downloads

[rancid knoll]

it should be in downloads then

[civic goblet]

how do i put my .wav files in the colab

i got them all but how

[rancid knoll]

ah damn ive never used the colab before

sorry mate

[summer glacier]

i dont see it in my downloads

[rancid knoll]

huh thats weird

[civic goblet]

dang what else can i use

not local

[rancid knoll]

no idea

[summer glacier]

so i put it a folder and ran it, theres still nothing

[civic goblet]

wait i got it figured

[rancid knoll]

is there no folder that has a name like "applio-rvc" or something?

[civic goblet]

said os is not defined.

[rancid knoll]

uhm what

[summer glacier]

theres one called Applio-RVC-Forkdatasets

i made that

all it has it my mp3

[rancid knoll]

does your windows username

contain any whitespace?

[summer glacier]

what?

it has no spaces

[rancid knoll]

hmmm

where the hell did the install go

[summer glacier]

brb

[rancid knoll]

i recommend mangio if you cant find applio btw

[visual jungle]

trainings going fine rn theres visible load on the gpu and the console shows progress

[pallid igloo]

so is there a way to make it so it doesn't auto download the new ai voice track and lets u listen to it first and then decide to save or keep it

[rare goblet]

@pallid igloo has reached level 1.

[naive sandal]

maybe someone can help a newbie, i, tryng to use a model, i downloaded it, how should i intall it?

[rare goblet]

@naive sandal has reached level 1.

[pallid igloo]

so do u have winrar

[naive sandal]

7zip

[pallid igloo]

oh ok thats fine so did u run the .bat file for the rvc

[naive sandal]

no, how do i do it?

[pallid igloo]

did u download rvc

[naive sandal]

oh for the web ui?

[pallid igloo]

oh sorry i have no idea

[naive sandal]

i managed to install it'

the webui is working

[pallid igloo]

oh good sorry i didn't help much

[naive sandal]

i can create music with it, but i want to do the speech to speech

u dont know how to install models?

[pallid igloo]

no i use the download version

[naive sandal]

what is the download version?

[pallid igloo]

just the one where u run a .bat file it like an app on ur desktop

[naive sandal]

u download just one file? or copied the repo?

cus i also just run a bat file for it

[pallid igloo]

oh well i'm really unsure tbh I am totally new and found out how to do it through youtube

[naive sandal]

okay thanks for trying, good luck

[pallid igloo]

thank ya!

[summer glacier]

ok

[rare goblet]

@summer glacier has reached level 2.

[summer glacier]

about my issue

it dosent show where my applio installation is

nvm i got it working

yay

[merry sand]

Okay, I think I got it now. So the last thing I want to know is about the epochs. So like the original RVC and Kits, you think 250 to 300 epochs is recommended to make the voices sound good? If I ask @tranquil cliff?

[summer glacier]

uh

im clicking update list in appilo and my folder dosent show, its called Applio-RVC-Forkdatasets

nothing is showing in the dataset thing

[dreamy yoke]

guys

how do i download python

or pytorch

or wtv ist called

[dreamy yoke]

do yk

[dreamy yoke]

or u

[summer glacier]

what version are you trying to get

[summer glacier]

but yes. ik how https://www.python.org/downloads/

[dreamy yoke]

but is it necessary

the python

[summer glacier]

yeah

i sent the download link

[dreamy yoke]

ok is there anything else i need to get

[summer glacier]

i used applio

[dreamy yoke]

yup

[summer glacier]

there might be a certian version you have to get

[merry sand]

I'm using the 300 Epochs.

[brittle wing]

Hi I just wanna say something here

[dreamy yoke]

why does it say failed

and it dont open

[sand iris]

remove the pretrain and model_dir folders

Where you unzipped

then relaunch and wait

[summer glacier]

anyways anyone know why im clicking update list in appilo and my folder dosent show, its called Applio-RVC-Forkdatasets
nothing is showing in the dataset thing

[dreamy yoke]

wait vc me

help me

idk how to

where is pretrain

[sand iris]

you unzipped a folder

Which is called MMVCServerSIO. Under that there's the two folders I talked about

[dreamy yoke]

oh i fgound it

ok now its updating again

or downl9oading

idk

[summer glacier]

nothing is showing where when i click update list

[brittle wing]

funny text haha

[high night]

@brittle wing damn

[brittle wing]

Hey why does he get AI member?

[rare goblet]

@brittle wing has reached level 1.

[high night]

@brittle wing you got it too

Level 1 = ai member role

[sand iris]

-AI Hub moment

[brittle wing]

Oh I only saw the ai noob

[high night]

@half cove

[summer glacier]

i need help

[brittle wing]

@summer glacier

[high night]

@brittle wing

[summer glacier]

...

im being ignored

[brittle wing]

shoulda been red, not blue

I was joking blue

[half cove]

Ok stop with big text it's annoying now

[brittle wing]

Ok

[high night]

Why are you using applio

Buggy asf

[summer glacier]

only one i can get working

ill try another one

how do i fix this

[proper shale]

you uh

did not link to google drive

or the google drive folder isnt there

[dreamy yoke]

guys

why cant i hear it in obs

[proper shale]

do you have the virtual audio cable installed?

[placid light]

Oh, and not to mention only chads show their model names in screenshots, as censoring it before sending is a waste of time.

[dreamy yoke]

whats that

help pls

[proper shale]

you need that so it puts the audio anywhere else outside of the program you're using

gotta download it tho

its easy

[dreamy yoke]

show

pls

how to install

[agile scarab]

Can someone explain to me in detail how I can use rvc gui with macbook m1?

[dreamy yoke]

??

[dreamy yoke]

?

[placid light]

I just finished training the model at 300 epochs. Is this okay to you or did I overtrain it?

Btw, only noobs blur out the model names as it doesn't contain any sensitive information, until there's a valid reason on why I should.

[orchid ibex]

Can someone help me, how do I use the voice model I downloaded?

What app do I use to make it work?

[placid light]

Haven't you checked the guide?

[orchid ibex]

I'm sorry, no, what channel can I watch that on?

[placid light]

https://discord.com/channels/1159260121998827560/1160164377006194728

[digital blade]

trying to get TTS using RVC working, have tried installing rvc-tts-webui to no avail, and am currently trying RVC-Studio which seems easier but am still getting stuck, is there any advice/help I could get? I've looked over the guides but there doesn't seem to be an application there that suits my needs, as I'd prefer to run everything locally

it's a shame uberduck had all its models taken off of it since it was the easiest solution, though running locally is better in a lot of ways at least :)

sorry in advance if there's a simple guide I've completely missed, I tried to search what I was looking for but couldn't find a solution in the message history

[rare goblet]

@digital blade has reached level 1.

[runic thunder]

help?

[vapid gust]

Go to scalars

Then check loss graph

Turn smoothing up to 0.999

Then send another pic

[placid light]

I closed the Tensorboard when I finished the training eventually.

[vapid gust]

u won’t know if it overtrained or not

[placid light]

But I tried it, and it trends up a little when it reached 290 or smth epochs.

[vapid gust]

then use 290

[placid light]

Does that mean that if the loss graph starts to trend up, it means I am overtraining the model?

[rare goblet]

@placid light has reached level 9.

[vapid gust]

Yes

[spiral osprey]

How do I make a cover more faithful? Is it changing pitch or something? How can I make it more like the original?

[torn pond]

are there any forks that allow you to change hop length with rvmpe?

[cloud pasture]

how do I make it so I dont hear the ai voice when using it?

[nocturne python]

when i launch the https_start file it will briefly open consol and then disappear how i get the rvc to actually launch. (dont tell me to reinstall ive done it 8 times)

[digital summit]

I'm getting the following error after installing on arch:
Traceback (most recent call last): File "/home/keon/Documents/Software/Applio-RVC-Fork/infer-web.py", line 55, in <module> from lib.infer.modules.vc.modules import VC File "/home/keon/Documents/Software/Applio-RVC-Fork/lib/infer/modules/vc/modules.py", line 21, in <module> from lib.infer.modules.vc.utils import * File "/home/keon/Documents/Software/Applio-RVC-Fork/lib/infer/modules/vc/utils.py", line 3, in <module> from fairseq import checkpoint_utils ImportError: cannot import name 'checkpoint_utils' from 'fairseq' (unknown location) Press Enter to continue...

Any suggestions?

[nocturne python]

bruh this channel might be useless nobody has answered min the past 4 times i sent it

[tranquil cliff]

It depends on the length of your audio....I use 300 epoch for 6/7 minute audio

[shrewd spruce]

could anyone help me with this error i'm getting? i downloaded RVC1006Nvidia and i'm trying to run go-web.bat but this happens every time i run it. i made sure the file path has no spaces, i've updated to python 3.12, i added python to PATH, and i haven't done anything to the files in the RVC folder (aside from adding and unblocking run_easiergui.bat and EasierGUI.py). i followed the tutorial word for word and somehow i've still messed up

J:\RVC1006Nvidia>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897 Traceback (most recent call last): File "J:\RVC1006Nvidia\infer-web.py", line 7, in <module> from infer.modules.vc.modules import VC File "J:\RVC1006Nvidia\infer\modules\vc\modules.py", line 11, in <module> from infer.lib.audio import load_audio, wav2 File "J:\RVC1006Nvidia\infer\lib\audio.py", line 4, in <module> import librosa File "J:\RVC1006Nvidia\runtime\lib\site-packages\librosa\__init__.py", line 209, in <module> from . import core File "J:\RVC1006Nvidia\runtime\lib\site-packages\librosa\core\__init__.py", line 5, in <module> from .convert import * # pylint: disable=wildcard-import File "J:\RVC1006Nvidia\runtime\lib\site-packages\librosa\core\convert.py", line 7, in <module> from . import notation File "J:\RVC1006Nvidia\runtime\lib\site-packages\librosa\core\notation.py", line 8, in <module> from ..util.exceptions import ParameterError File "J:\RVC1006Nvidia\runtime\lib\site-packages\librosa\util\__init__.py", line 77, in <module> from .utils import * # pylint: disable=wildcard-import File "J:\RVC1006Nvidia\runtime\lib\site-packages\librosa\util\utils.py", line 9, in <module> import numba File "J:\RVC1006Nvidia\runtime\lib\site-packages\numba\__init__.py", line 42, in <module> from numba.np.ufunc import (vectorize, guvectorize, threading_layer, File "J:\RVC1006Nvidia\runtime\lib\site-packages\numba\np\ufunc\__init__.py", line 3, in <module> from numba.np.ufunc.decorators import Vectorize, GUVectorize, vectorize, guvectorize File "J:\RVC1006Nvidia\runtime\lib\site-packages\numba\np\ufunc\decorators.py", line 3, in <module> from numba.np.ufunc import _internal SystemError: initialization of _internal failed without raising an exception

[teal glacier]

--rvc

[hard gust]

UVR's VR models won't work in Applio, someone knows something about this maybe?

[late canyon]

why am i getting error in rvc v2 hugging face

i did everything correct

Traceback (most recent call last):
File "/home/user/app/app.py", line 367, in vc_multi
paths = [os.path.join(dir_path, name) for name in os.listdir(dir_path)]
FileNotFoundError: [Errno 2] No such file or directory: 'E:\codes\py39\test-20230416b\todo-songs'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/home/user/app/app.py", line 372, in vc_multi
paths = [path.name for path in paths]
TypeError: 'NoneType' object is not iterable

[brittle wing]

I have launch the file and i don't know where is the file

[sturdy yacht]

i have no clue why

[rare goblet]

Ayo? @sturdy yacht level 1 !!!

[sturdy yacht]

but ever since updating my driver

now pytorch wont or cant find or use my gpu ?

how to fix this ?

[jaunty nimbus]

-colab

[azure marsh]

Google Colabs

RVC

• AI-Cover-Gen-No-UI (inference only)
• RVC v2 Disconnected (training only)

🌐 Translated colabs

• PT-BR: -colabs_br

Audio Separation/Isolation

• MDX Net Colab
• UVR arch colab 5.0.2/c1.04
• KaraFan colab

[rare goblet]

Ayo? @jaunty nimbus level 1 !!!

[jaunty nimbus]

is there a colab with ui?

[pulsar snow]

@glad zealot Hi, sorry for the tag, but since it's your colab, I figured I'd ask you here. What's "Pitch Change" option that you added?

[glad zealot]

its basically the same as the voice changer tune setting

[weak palm]

Why aint index training working 😭

[rare goblet]

Ayo? @weak palm level 1 !!!

[lament eagle]

should I continue

[rare goblet]

Ayo? @lament eagle level 10 !!!

[harsh laurel]

I'm training RVC models on Kaggle with Tesla T4 x2. Currently in my notebook my batch is 7. What's the maximum one I can use?

[tranquil cliff]

Hi

[brittle wing]

how do i create an ai model??

[high night]

https://discord.com/channels/1159260121998827560/1159618293594861579

Go for this

I see you got a good pc

[brittle wing]

thanks alot man

[high night]

Basically first you create dataset @brittle wing

Collect audio files of the target person

Then clean it using uvr

[tranquil cliff]

here too if you want https://discord.com/channels/1159260121998827560/1166983964657975376

[high night]

Then open up applio

[brittle wing]

how do i make it really good tho?

because some of them are really shitty

[high night]

Make better dataset

Clean dataset

Without any music and artificacts

[brittle wing]

so like alot of audio training probs?

[high night]

Wdym by that

Generally, 8 minutes of audio is enough for a model

[brittle wing]

damn,really??

[high night]

I have trained models with 7 seconds of data too

Yeah

If your dataset is small

You want to go for more epochs

[brittle wing]

so if more epochs then better?

[high night]

Yeah

But you always wanna look for over training

@brittle wing I'm going to eat, you can dm me if you need more help later

[brittle wing]

how long will it take to create hm... 1000 epochs?

[brittle wing]

sure,thanks alot

[high night]

You don't need that much

[visual jungle]

anybody knows the default settings kits ai uses for training? trying to match the quality of my new one with the old one

[brittle wing]

so like 300?

400?

[high night]

@brittle wing dm me

[loud berry]

help wtf does this mean😭

[rare goblet]

Ayo? @loud berry level 4 !!!

[loud berry]

pleaseeeeeeeeeee

[brittle wing]

hey guys,so im doing my first model