✨│ai-help

[verbal dome]

I'm here in the video, but I' m not sure what hes doing with the link.

https://www.youtube.com/watch?v=GckqyWYSOls&t=2m9s

it looks like he's downloading something but I don't know what it is.

Okay, when i try opening up this file, it says "the contents of tensorvenv.bat.exe cannot be extracted with the unarchiver"

Oh its becuase I downloaded the Windows version

[rare goblet]

Ayo? @verbal dome level 3 !!!

[west lark]

anyone knows how to make it work?

its from #📰│dev-updates

[verbal dome]

it says "traceback (most recent call last):
File "/Users/shawn/Downloads/AI Voice/Applio-3.1.0/app.py", line 10, in <module>
from tabs.inference.inference import inference_tab
File "/Users/shawn/Downloads/AI Voice/Applio-3.1.0/tabs/inference/inference.py", line 3, in <module>
import regex as re
ModuleNotFoundError: No module named 'regex'
(.venv) (base) shawn@Shawns-MacBook-Pro-4 Applio-3.1.0 %" after typing shrun-applio.sh

[brittle wing]

why is showing this?

[odd shale]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

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

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

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

__**BE PRODUCTIVE.**__ 🤝

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

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

[west lark]

do you think rvc v2 is good?

[low shard]

yes, rvc is the best for speech to speech

[west lark]

thanks

[verbal dome]

I left my mac for an hour to do something while it was downloading in the terminal and I guess it stopped at 55% . What do I do to make it continue the download?

No supported Nvidia GPU found
Using FP32 config instead of FP16 due to GPU compatibility (None)

Downloading https://huggingface.co/IAHispano/Applio/resolve/main/Resources/hubert_base.pt to hubert_base.pt...
100% [......................................................................] 189507909 / 189507909
Downloading https://huggingface.co/IAHispano/Applio/resolve/main/Resources/rmvpe.pt to rmvpe.pt...
100% [......................................................................] 181184272 / 181184272
Downloading https://huggingface.co/IAHispano/Applio/resolve/main/Resources/fcpe.pt to fcpe.pt...
100% [........................................................................] 69005189 / 69005189
Downloading https://huggingface.co/IAHispano/Applio/resolve/main/Resources/pretrained_v2/D32k.pth to rvc/pretraineds/pretrained_v2/D32k.pth...
100% [......................................................................] 142875703 / 142875703
Downloading https://huggingface.co/IAHispano/Applio/resolve/main/Resources/pretrained_v2/D40k.pth to rvc/pretraineds/pretrained_v2/D40k.pth...
55% [.....................................

5

This is what it was doing

[verbal dome]

nvm

when i open up applio in the browser every time i click on something it says "Error. Connection Errored Out."

[sharp sparrow]

So i'm hoping this is the correct place to ask this ^^'

but does anyone have any experience with wanting to run an external GPU just for your voice?

i use it while streaming (and make no secret of this) but since i run my model, voice, hands tracker and game all on my 3070 ti , this can be somewhat limiting in terms of games i can play/stream, due to my voice taking up an amazing amount of computing power.

Hence the idea came to run a spare 1080 on an E-GPU case, with thunderbolt.
As it may also be the cheaper and more compact option to getting an entire streaming PC on the side.

Question is, am i overthinking this? Or just entirely wrong for wanting to try this out?
Or have people in this lovely community already tried to do this? 🙂

I hope to hear back from you guys ^^

[wild bobcat]

i let my colab run for too long and my time is up

is there any other alternative solution to making a voice locally

[undone verge]

hi help i need instalacion pls

algguien que hable español porfavor

@dusty rampart

[dusty rampart]

I speak portuguese not spanish

[undone verge]

okok

[dusty rampart]

Use DeepL to translate the guides

[undone verge]

Olá, preciso do programa para mudar a voz, a minha não funciona por favor

[rare goblet]

Ayo? @undone verge level 1 !!!

[undone verge]

@dusty rampart

[dusty rampart]

Não dou suporte ao voice changer

-realtime

[azure marsh]

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

[undone verge]

nao tendo bro

[dusty rampart]

https://rentry.co/VoiceChangerGuide

[undone verge]

link

download

pls

:c

[dusty rampart]

read.

https://www.deepl.com/en/translator

[undone verge]

@brittle wing Hi, I need to download the program, I can't find the link.

[glad zealot]

Not sure but as long as your pc can detect the gpu, you should be able to use it as normal. Idk how to tho

[low shard]

yes, kaggle and lightning, kaggle is better tho

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
  • Applio (ui)
  • Mainline (UI)
  • RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide as of right now)
• Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
• Mainline (UI, No guide as of right now)
• Applio (UI, No guide as of right now)

[warm oriole]

hey guys i downloaded applio on my pc locally the tts sound robotic no mater what ive use lots of model but they do not get good like the online service provider ones like they do at fish audio for example im using default nural models for tts voices i geuss that is the problem are there any good tts voice model that does not come defaultly ?

help me out guys i got 4090 as my gpu so pc isnt a problem i guess

[low shard]

Rvc is natively speech to speech, so its not as good for tts, applio tts just uses microsoft edge tts (usually multilingual and good quality but emotionless) api to make a tts audio that gets used as an input for the model

you could either play around with the settings and models

or either try actual tts programs such as fishspeech (0 shot which means no training needed, good for being fast, theres also a local version dw) or gpt so vits (the best for emotions and quality, requires training a model, and ofc is open source too)

[tough mist]

hello, do u guys know how to uninstall everything that applio comes with? like what are all the other programs/files that come with applio?

[warm oriole]

@low shard Thanks bro appreciate the help will try it

[rare goblet]

Ayo? @warm oriole level 1 !!!

[low shard]

your welcome

[agile gate]

i dwonloaded RVC GUI and extracted it, but the .bat thing isnt there

i dont know if i download it again or what

[knotty moth]

-gui

[azure marsh]

https://cdn.discordapp.com/attachments/1122285248844144733/1203460490475343953/caption.gif?ex=65d12cec&is=65beb7ec&hm=bd2fb8d010006dd7c6e3c1c67d3ae846fd1478e1a3124c544c31b43086fe54aa&

[knotty moth]

I'd recommend this #🧬│ai-chat message

[low shard]

RVC GUI is outdated

What's ur PC gpu?

[plush ocean]

Havent used colab in ages so please forgive me if this is simple. Step 5 returns this

ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-7-0e44d4d1db8f> in <cell line: 4>()
      2 import threading
      3 import time
----> 4 from utils import backups
      5 
      6 #@markdown #Click here to run the GUI

ModuleNotFoundError: No module named 'utils'```

[knotty moth]

it is no longer updated and conflicts with the latest pip 24.x in the installation step

[plush ocean]

That would explain it then haha.

[rare goblet]

Ayo? @plush ocean level 2 !!!

[plush ocean]

Yeah sorry I haven't used RVC in ages, not sure if this was a me issue or a dated issue. Thanks 🙂

[knotty moth]

again I'd recommend Ilaria RVC or Applio for just inference
https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit

[plush ocean]

God damn that's much easier

You're a legend. Thanks!

[mild sluice]

Who helps me create ai cove Google colab

[rare goblet]

Ayo? @mild sluice level 3 !!!

[mild sluice]

Who helps me create ai cove Google Cobal

[turbid wolf]

Which ais that most ppl use for l4d2 voice mods? I wanna make mine i dont know which one actually allows you to upload almost 2.9k voice files all in one..

[sudden socket]

can i have the link to download rvc?

i cant seem to find it

[lone lagoon]

I wanted to download a voice changer to talk through it on discord. Where can I find the link?

[rare goblet]

Ayo? @lone lagoon level 1 !!!

[low shard]

What's ur PC gpu?

[low shard]

What's ur PC gpu?

[lone lagoon]

кеч3060

rtx3060

[low shard]

Alr that's very good

-rt

[azure marsh]

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

[low shard]

1st link, it's a wokada fork which is optimized

[low shard]

Tbh I'd suggest you to use Ilaria RVC zero, it's easier to use

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[low shard]

If u really wanna do it on Google colab, go for applio

[sudden socket]

i did download the cuda one after watching a youtube video btw the screens stuck on command prompt of https so what should i do

[low shard]

Never watch YouTube videos

99% of them are outdated

[sudden socket]

oh its started

[low shard]

What's ur PC gpu?

[sudden socket]

i see i see

its 1650 ti

[low shard]

I'd personally suggest u to use our wokada fork written guide, the fork is optimized with better performance

-rt

[azure marsh]

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

[low shard]

The 1st link

[sudden socket]

ok ok

ill try that one

[elder halo]

-rt

[azure marsh]

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

[limpid trout]

@low shard alright it's just my audio from a Video in YouTube
But when I upload it in The web the first audio seems like it's empty

[low shard]

Yea the problem is with your file, you forgot the '.mp3' at the end

i added it, u could retry with this now

[limpid trout]

Ok

Dang it I am being more stupid

[azure marsh]

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[limpid trout]

Well even after doing that,
The first audio is completely empty
And after clicking on convert output once showed error
And I again converted
It just shows music logo but it's not clickable

[limpid trout]

Help

Well imma try this enxt day
I am going slepe

[still kite]

where i can download rvc

[low shard]

with first audio do you mean the one u put or the one after u click convert?

U might need to re-download your file as it seemed u downloaded it not correctly

[low shard]

whats ur pc gpu and what do u want to do

[still kite]

gtx 1660, change voice for speak in games

[low shard]

-rt

[azure marsh]

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

[low shard]

1st link which is a wokada fork, so an optimized version

[rare goblet]

Ayo? @still kite level 1 !!!

[still kite]

i already downloaded one, can i send link here? i just want you to check is this link safe(sorry bad english)

[low shard]

If you are using one from a youtube video, its 99% outdated and not performance optimized

its way better you do it only from that guide i sent you

[bright zodiac]

installed latest applio

get error when selecting model

Running on local URL:  http://127.0.0.1:6969

To create a public link, set `share=True` in `launch()`.
ERROR:    Exception in ASGI application
Traceback (most recent call last):
  File "C:\Users\Benni\Desktop\Applio\Applio-3.2.5\env\lib\site-packages\pydantic\type_adapter.py", line 270, in _init_core_attrs
    self._core_schema = _getattr_no_parents(self._type, '__pydantic_core_schema__')
  File "C:\Users\Benni\Desktop\Applio\Applio-3.2.5\env\lib\site-packages\pydantic\type_adapter.py", line 112, in _getattr_no_parents
    raise AttributeError(attribute)
AttributeError: __pydantic_core_schema__```

[still kite]

https://github.com/w-okada/voice-changer/blob/v.2/README_en.md is this good rvc?

[low shard]

wokada is a program that uses RVC models in realtime, and yes thats the official wokada

If u want better performance, i suggest using its fork

which i sent u the guide

[still kite]

ok thank you ❤️

[low shard]

yw

[still kite]

theres vcclient_win_cuda_2.0.61-alpha.zip which is 3gb and vcclient_win_std_2.0.61-alpha.zip which is 600mb, what to download?

[low shard]

u are downloading this one from the guide right?

[still kite]

no, i want to download wokada (https://huggingface.co/wok000/vcclient000/tree/main)

[low shard]

I would highly suggest you to use the one of the work which has better performance

If u really dont want to, use https://rentry.co/VoiceChangerGuide#download-nvidia instead which is the guide of the normal wokada

[odd shale]

Buddy, download the Deiteris fork instead.

We suggest you this one.

[bleak basin]

Tyler derden voice

[agile gate]

whats the new?

[mighty mirage]

I can't seem to get the deiteris fork to work, I just keep getting errors

[odd shale]

Use Applio Colab or download RVC locally/use Mainline Kaggle colab.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[odd shale]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

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

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

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

__**BE PRODUCTIVE.**__ 🤝

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

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

[low shard]

Depends, whats ur pc gpu?

[agile gate]

ill check

[rare goblet]

Ayo? @agile gate level 1 !!!

[low shard]

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[agile gate]

its a NVIDIA GeForce GTX 1060 6GB

[low shard]

You can only inference (use models) locally, but not train (make models). What are u looking for?

[agile gate]

i want to use models

[low shard]

Local ways (on ur pc, no limits but will be slow bc of ur gpu, and needs installation):

• Applio: A fork of RVC with some extra features like Applio TTS, same quality tho
• Mainline: The original RVC

Cloud ways (remote good pc, runs faster but its kinda limited, but its easier and wont need any installation:

• Ilaria RVC Zero Fastest on cloud using A100 free gpu
• Applio Colab Slower as its using T4 gpu, but still faster than urs locally

[agile gate]

thanks

[low shard]

your welcome

[fleet grove]

Can anyone tell me all the prerequisites for deiteris W-otaka fork? I wanna make sure I haven't missed anything.

[odd shale]

There you go.

[fleet grove]

sorry for not saying it correctly, i meant what apps are necessary, things like python etc. thanks for trying to help me though, i really appreciate it. 👍

[rare goblet]

Ayo? @fleet grove level 1 !!!

[odd shale]

The guide doesn't say anything about what apps are necessary.

But you need to install VB-Cable beforehand

[fleet grove]

Thanks a ton!

[odd shale]

🐢 👍

[fleet grove]

Do you need pytorch for AMD fork? I only see cuda in the torch downloads. Or would I just select CPU?

[stable kindle]

Hello everyone, I am new here and also a newbie in the AI world. Can I use the RVC voice models for commercial purposes? Since the models are based on real person voices I doubt that I will be able to use them for that. Are there also original AI-based royalty-free models available?

[odd shale]

The only way to have/create RVC royalty-free model is by mixing 2 or more voices in a same dataset and then training it.

[nocturne mural]

download the file, run it in the applio folder and your error will be fixed.

[odd shale]

Holiwi Vidal.

[stable kindle]

Thanks

[stable kindle]

In TTS can we specify certain emotions like anger, joy, etc to produce voice with that emotion?

[regal scarab]

what do i do if the voice is being weird?

like i saw some people using it and for them it worked fine but when i try the voice its being weird

any tips?

[odd shale]

No, you can't.

At least on Applio's TTS.

[odd shale]

Actually if your voice is laggy (If you're using W-Okada) then you should play around with the settings.

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[regal scarab]

Voice isn’t laggy it’s just the voice sounds weird af

[low shard]

Not on the TTS built in RVC forks, I think you should be able to with gpt so vits but it's different models and only japanese, Chinese and english

[limpid trout]

haha it worked

[rare goblet]

Arcane Commands

The prefix for AI HUB is /

Useful Resources

Command List
Dashboard
Support Server
Arcane Status
Premium

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[knotty moth]

for that purpose, use elevenlabs or GPT-sovits (few shots)

[steep trout]

this is hell

I can't even post a picture here

My first try I got something almost usable, so I added more data to the dataset, and then it never worked again. Even reverting to only the original one file I can't replicate success and only get garbage

[rare goblet]

Ayo? @steep trout level 1 !!!

[steep trout]

oh heck ya

Top is my good try and all the lower ones just don't work right

[echo gull]

Hello everyone! How are you doing? :)

I'm new to using AI voice changers and I'm currently testing RVC. To my surprise the quality is above what I expected. I'm using it for realtime voice to voice usage in order to use it for DnD when speaking in character of tones I can't do on my own. A problem I seen though is that the voices seem to be quite rigid even if you try to really act in it. You can slightly hint at annoyance maybe but it's hard to convey anger, fear, pain or any important emotion really.

Would anyone know if there would be something I could improve in the software settings? Or if there would be a better AI for my usage?

Thanks for taking the time to read me :)

[odd shale]

RVC and W-Okada/realtime work differently.

RVC works like speech to speech conversion.

If you're using a model on realtime and it doesn't work properly, maybe it's your settings.

And no, you can't do realistic emotions on W-Okada/realtime due to limitations.

Not even laughing loud or coughing.

[echo gull]

I'm using RVC and I'm guessing the model works properly, it's just a bit limiting on acting range, I mean a bit more than what I expected. And for noises yeah I'm not expecting to be able to do those at all

[rare goblet]

Ayo? @echo gull level 1 !!!

[odd shale]

And that's fine.

[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, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab
• 🆕 FaceFusion UI, by Nick088 Google Colab
• 🆕 FaceFusion NO UI, by Nick088 Google Colab

ℹ️ Note

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

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

[dire shadow]

hi can someone help? i didnt use voice changer for a while and tried to start it today and this showed up.

[low shard]

which google colab are you using?

send the link

[jagged summit]

Can u help me in #🔍│help-w-okada ?

I need help rq idk if I can turn off the pc

[low shard]

I don't understand ur issue in here, u should need to run the install cell

[jagged summit]

The server link doesn't show up

That is the issue

Like this

That is on the example photo

[low shard]

IM not really a wokada helper,
But what shows in the output when u run it?

[jagged summit]

i cant find [2]

this is the issue

[low shard]

You can't find the whole 2nd cell?

Did you delete it?

[jagged summit]

no it just apperead idk how

[low shard]

apperead or disappeared?

Ig u accidentally deleted it, just refresh the page

[jagged summit]

apperead

lol sry

[low shard]

Oh lol

[jagged summit]

🙏

[low shard]

😭

[jagged summit]

sry again

[low shard]

Nahh dw

[jagged summit]

which VAC or VB u suggest for windows?

[low shard]

I'd suggest the vac also given in the guide, i remember using it sometimes

[jagged summit]

k

thx

all working

now ima go download th villager voice

lol

test

[low shard]

yw lol

[dire shadow]

this: https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb#scrollTo=lLWQuUd7WW9U

[low shard]

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

the guide is the first link under this

-rt

[azure marsh]

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

[low shard]

(so the guide is https://docs.google.com/document/d/e/2PACX-1vTIceEcBfS6Zqolv_QEysrFfI_EJikPxozWptP_EjkpLVl-l-gdo-ijBonQMTviAHEYm5emmd9k9TdC/pub)

[elder folio]

gang my mangio crashing

[dusky lion]

I am running a private duplicate of Ilaria RVC, does anyone know how to remove the 500mb model download limit?

[tough sequoia]

how to clone my own voice?

[young halo]

What does that mean?

[glossy mulch]

Hi everyone i want to find a rvc detector

[low shard]

there isn’t an rvc detector

And even if there was, ai detectors are not reliable

[low shard]

@proven hill may know
However, duplicating ilaria rvc makes so that you wont use the zerogpu paid by ilaria, so it will use cpu and be alot slower

[low shard]

its better to not use mangio, also whats ur pc gpu

[low shard]

Be sure u uploaded the added.index not trained.index

[dusky lion]

Yeah I just kept getting rate limited on the public one so I duplicated

[young halo]

Idk, i'm using a model from Huggingface

[low shard]

i think u mean the retry again issue, yea thats because zerogpu is not unlimited
Im just telling u that it will be alott slower on ur private one as its on cpu

[low shard]

could u share the link and also tell me what u using

[dusky lion]

I’m mean I’ve been using it with no problem on the private one, no noticeable dip in performance, maybe an extra few seconds here and there

I found a ton of models that are around 1gb each tho that I currently can’t use 😦

[young halo]

https://huggingface.co/RavenCH1902/Christina-Aguilera/resolve/main/RVCChristinaAguilera800.zip?download=true

And using the Hina Mod AI Cover gen

[low shard]

Yea ofc theres no problem, but im guessing u are running it on cpu right? Cus thats slower unless ur paying for ur own gpu

anyways should work, was just warning

[dusky lion]

I paid for one of the cheap plans that was per hour, don’t remember which it was tho, most likely CPU tho

[low shard]

alright

maybe ilaria knows how to remove that limit

[low shard]

yea it has useless files inside of it

[young halo]

Oh, I see

[low shard]

If u can, download it urself, unzip it and upload only the pth and added.index inside of hina mod ai cover gen

or else u need to re zip it and upload it to hf with credits publicly

[young halo]

Tysm!

[low shard]

Yw

[oak schooner]

Yow.
Is there any topics that I can read about fine-tuning voice model?

[limpid trout]

Hello nick,
I was messing around with the illria VC . It works good now
But one problem is that The character I use for Voice changer sounds kinda different. Like it's kinda pitched out or doesn't really sound like the character. How to fix it?

[rare goblet]

Ayo? @limpid trout level 3 !!!

[limpid trout]

Also I have problems to check the settings section which you can change pitch or envelope
I have no clue how those works

[low shard]

in the Inference tab, before the Convert button, there should be a Settings button where you can change the pitch

lower = deeper voice, male
higher pitch = higherr, female

you need to play around with it till the voice sounds good, dependds by case and model

[low shard]

fine-tuning model or pre-train?

[oak schooner]

I trained a model on pre-train. Is it possible for fine-tune after it's trained already or...
I want to know what is fine-tune at least haha.

[low shard]

fine-tuning is when you take an ai model as a base, and train making adjustments to it

You should be able to continue training, what did u use ?

[oak schooner]

I used SnowieV3.1_32k.

[low shard]

I mean the program

[oak schooner]

Ops...
Applio.

[low shard]

locally (on ur pc) or on kaggle or on colab?

[oak schooner]

Locally (that was kinda funny how you decided to clarify haha)
Yeah I know that I can continue tho.
But I'm not sure should I use fine-tune or no, so I'm mostly asking about technical (?) information. And when it can help to improve a model.

[low shard]

well, depends by the tensorboard, u need to always check whats the lowest point, it could help but also depends specifically on your dataset (it needs to be super cleaned)

how much epochs did you train it before ?

[oak schooner]

It's 200 epochs. At least this lowest point. Hm... is there any «right» way to check it's quality?

Also about dataset. Any inhales also shouldn't be in it right?

[oak schooner]

I forgot to answer mb

[low shard]

inhales could count as noises for rvc

maybe its better to re-clean the whole dataset and be sure theres no sound and re-train instead of continuing training,
Also usually let it run for like 500 epochs and see which is the lowest point

[oak schooner]

So the most important thing is dataset?
5-10 mins will be enough with SnowieV3.1, is it correct?

[rare goblet]

Ayo? @oak schooner level 2 !!!

[oak schooner]

Oh by the way. I have one more question if you don't mind.
Do embedders help somehow?

[low shard]

yea and also checking the tensorboard

[limpid trout]

Well I could do that
But the problem is that I am not sure how long Illria VC will stay in free section,
Is it limited? Like only 4 hours for 1 day?

[low shard]

well they are neural network models that convert the raw audio input to an high-dimensional vector representations
But its usually better to leave the default one

[low shard]

Well u cant really check that, it depends by zerogpu quota per account which cant be seen

[oak schooner]

Okay got you.
Thanks for your help.

[limpid trout]

Huh? Ok?

(didn't understand anything)

[low shard]

ilaria rvc zero is an huggingface space, to be fast it uses ZeroGPU (huggingface gpu service) paid by ilaria, it works on each huggingface account have a quota of time to run it, but as of right now it cant be checked

[low shard]

Yw

[limpid trout]

Hmmmm, ok
I

[limpid trout]

Also uh nick,
Kinda of out question. How much you get paid for answering our stupid questions

[low shard]

Depends

[oak schooner]

Is it okay to get the same values over and over again?
The further training is going the more epochs needed to change «lowest_value»

[fiery raptor]

the go-webui.bat doesnt work for me

[turbid wolf]

which one of the huggingface spaces lets me upload more than 1 audios ?

[real nymph]

hello, I have a question, how can I use the voice models from #1175430844685484042 with python? are there any libraries for that?

[burnt dune]

hello guys

when using the program, the sound vibrates, how can I solve it, it happens on the way to the other party

[low shard]

https://huggingface.co/spaces/r3gm/rvc_zero

[low shard]

whats ur pc gpu and what are u using

[low shard]

u could either try https://pypi.org/project/rvc/ (made by an official dev of rvc) or https://pypi.org/project/rvc-python/

[real nymph]

thanks

[turbid wolf]

does it works if i put all on rar ?

[low shard]

I dont think so, ig u can click the button to upload files, then select one, then ctrl+click every other files

[turbid wolf]

alright thank you for your help hopefully i can manage to do it im planning to make l4d2 character voice mods

[low shard]

alright yw

[dawn bronze]

How do I create a ai model using a mp3 and convert it to huggingface

[rare goblet]

Ayo? @dawn bronze level 2 !!!

[dawn bronze]

@low shard do you know?🤔

[low shard]

what's your pc gpu?

also its better to use .wav not mp3, and it doesnt really convert to huggingface, u just upload it
But i think u will understand better with the guides

[dawn bronze]

I'm using a android Samsung tablet

[low shard]

You can't do it locally (on ur pc as u dont have one), but you can use cloud (remote good pc), All cloud ways are:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
  • Applio (ui)
  • Mainline (UI)
  • RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide as of right now)
• Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
• Mainline (UI, No guide as of right now)
• Applio (UI, No guide as of right now)

The easiest one is one of the Google Colabs, but has low gpu time, the best one is one of the Kaggles which give more gpu time but requires phone number and its kinda harder

[oak junco]

hello, I am just starting my journey into this and was wondering what the best software for this was (preferably with lowest latency) wanted to use it with my TTRPG table as i think it could bring a lot of fun having so many different voices. I am especially concerned considering the announcemnet that one of them was hacked.

[odd shale]

Maybe you could try by installing Deiteris' W-Okada fork.

-rt

[azure marsh]

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

[odd shale]

Read the deiteris guide.

[oak junco]

ah ok thanks

[knotty moth]

the fork wokada is the optimized one, but the delay on switching to another model on the fly seems impractical for your RP-ing purpose

[oak junco]

does it have like a delay when switching to other voices even if i have all of them added beforehand?

[rare goblet]

Ayo? @oak junco level 1 !!!

[oak junco]

i have an rtx 4070 if that helps and using Foundry VTT which is a pretty light RPG system

[fiery raptor]

3070

[rare goblet]

Ayo? @fiery raptor level 7 !!!

[fiery raptor]

and by using do you mean like what gpt sovits am i using?

[low shard]

Oh ur using gpt so vits? could try following https://docs.aihub.wtf/tts/gpt-sovits/

But i dont think i would be able to help you with that

[knotty moth]

you can try if it works

[fiery raptor]

what did you think I meant?

[maiden remnant]

when i use the rvc my 3060 ti is using like 80/90 percent and if i run a game its kinda laggy how do i limit rvc to 60% or 50% usage?

[honest shoal]

I'm new to this stuff why can't I get it to sound how it always sounds in the test clip

[low shard]

RVC, like most people here

[viscid tartan]

Where should I get the .PTH file of my model? I can't find any information about that in #1159513888199540817

[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, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab
• 🆕 FaceFusion UI, by Nick088 Google Colab
• 🆕 FaceFusion NO UI, by Nick088 Google Colab

ℹ️ Note

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

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

[nocturne oak]

does someone know why my voice lags so much when using the voicve changer

[azure yoke]

how do i change the epoch in this rvc ?

[odd shale]

That's not called RVC.

It's W-Okada.

And what do you mean by changing epochs?

[azure yoke]

every model has its own epoch

what does that mean

[odd shale]

First of all, epoch is not related to realtime, that's related to RVC.

[azure yoke]

rip

[odd shale]

Epoch is like, how many times it was trained on RVC (not W-okada)

And no, you can't change it.

[azure yoke]

ok thank u for telling me but do you know how to use rvc?

is there a tutorial?

[odd shale]

RVC is not the same as W-Okada.

Since RVC is for speech to speech conversion (audio)

And W-Okada is realtime.

[azure yoke]

ohh

okay but like some voices in okada doesnt sound the same as

the previewed samples

of it

why

[odd shale]

Because these samples were made on RVC.

RVC works pretty different than W-Okada.

[azure yoke]

oh ok

[odd shale]

And keep in mind some models on the #1175430844685484042 won't work properly on w-okada (either for dataset lack, undertraining or the voice simply doesn't work for realtime)

[azure yoke]

ohh okay thank u for helping
\

[odd shale]

You're welcome.

[odd shale]

After finishing training of your model on RVC, you can get your model's usable pth on the "weights" folder

[flint hornet]

hello guys. how can I open a model in new rvc ?

[odd shale]

Open a model?

[small linden]

hi

[odd shale]

Or do you mean how to use a model?

[flint hornet]

how to import to rvc idk, i cant upload image

i dont have a npy file

[odd shale]

You don't need .npy files.

[odd shale]

How to import models to RVC?

So, you want to use models?

[flint hornet]

yea, i dm'ed you

[rare goblet]

Ayo? @flint hornet level 1 !!!

[odd shale]

You're using W-Okada voice changer, not RVC.

RVC is a different thing.

[flint hornet]

so, where i can find rvc?

[odd shale]

Do you want to use models on realtime or test models with audios on RVC?

[flint hornet]

i wanna to use on realtime

[odd shale]

I would suggest you to read the Deiteris' fork guide.

-rt

[azure marsh]

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

[odd shale]

Also, i recommend using Deiteris' w-okada fork

[flint hornet]

damn, ty !!!

[odd shale]

You're welcome.

[west ether]

hey im getting this error, thought it was a problem with my files so i fresh installed and im still getting this, running on an amd 7900xtx rn

current settings

[brittle wing]

which is the best for amd board?

[dense marsh]

-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, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab
• 🆕 FaceFusion UI, by Nick088 Google Colab
• 🆕 FaceFusion NO UI, by Nick088 Google Colab

ℹ️ Note

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

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

[loud zodiac]

Hey. I am pretty new to this. Is there any way for me to add my own model to UVR5

[frigid verge]

Hello everyone, have a nice day.
Today I tried using RVC GUI to convert a singer's voice to a model (Ayaka) but when I finished converting, I got a wav file without hearing anything in it. Does anyone know how to solve this problem? Please help me

[low shard]

Are you using https://github.com/Tiger14n/RVC-GUI ? its outdated. Dont use it. Whats ur pc gpu?

[low shard]

for amd gpu? You are on windows?

[brisk quarry]

what uvr5 model is reccommended for stuff with background music

local

[knotty moth]

-gui

[azure marsh]

https://cdn.discordapp.com/attachments/1122285248844144733/1203460490475343953/caption.gif?ex=65d12cec&is=65beb7ec&hm=bd2fb8d010006dd7c6e3c1c67d3ae846fd1478e1a3124c544c31b43086fe54aa&

[brisk quarry]

what gui do people even use

i use applio cuz i got nothing else

[low shard]

Some use https://github.com/Tiger14n/RVC-GUI which is literally archived Js bc of yt tuts on it

[brisk quarry]

never seen someone use that before

[low shard]

https://youtu.be/5i_Pyw0gH-M?si=QvQoevwXt3hXuHtq it's VERY old

[craggy bane]

I've been out of the loop since January. Latest model was Ov2Super40k, which was pretty good. What is the latest/best model and any new GUI developments? I was using it locally, not collab. Thanks for any answers in advance.

[brisk quarry]

give me a more modern one duh

[low shard]

I think you are talking about pertains, yes u could check https://discord.com/channels/1159260121998827560/1233407331405004954

[low shard]

I was just telling u that one is outdated as we were on topic

Applio is good to use

Else, if u really wanna use another one, there's RVC mainline (the original rvc)

[brisk quarry]

me: downloads ilaria on windows
ilaria: .sh

[brisk quarry]

what

[craggy bane]

Thank you for that link.

[low shard]

U can just use the latest release

[brisk quarry]

what even is the difference between applio and ilaria

its both the same thing isnt it

[low shard]

It's said in Ilaria rvc mainline

[low shard]

Well the ui

And I forgot

If there's any other major differences

The ui, some info tab, and some pretrains built in

[brisk quarry]

but the performance is the same

[low shard]

Yep, every RVC fork has the same performance, it's Applio performance is the same as mainline

Quality doesn't change

[brisk quarry]

waht about speed

[low shard]

Should be the same

I don't really have tests to prove that

[brisk quarry]

2it/s is pretty slow isnt it

[low shard]

But as they are all forks usually the main difference is just the ui and things like built in tts

[low shard]

I forgot how much colab does but shouldn't be that bad

[brisk quarry]

its local

[low shard]

I mean if u want to do a test to see which is faster inferencing the same audio on same model with applio and Ilaria rvc mainline u can

[low shard]

But there shouldn't be anything much different

[low shard]

Ye i was comparing it to google colab

As u have an rtx 4080 should be pretty good

Lucky to have that gpu 😭

[brisk quarry]

who said i have that

i have a 4070 super

[low shard]

Oh well my bad

Still way better than my integrated graphics for sure lol

[brisk quarry]

rvc is hella unoptimized

[low shard]

From what I hear by most devs, it has kinda spaghetti code

Which means not really ordered well

[brisk quarry]

is it open source?

[low shard]

Yes RVC is open source

[low shard]

If u are talking about spaghetti code, it's just a term to say this

[brisk quarry]

i want to fix it.

[rare goblet]

Ayo? @brisk quarry level 30 !!!

[brisk quarry]

i can fix it.

with chatgpts help

where do i get the code

@low shard

[low shard]

That won't help much 😭

[brisk quarry]

give code

[low shard]

https://github.com/IAHispano/Applio/ for applio or https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI for mainline

[brisk quarry]

is there even a difference between the latest version and the one before it

[low shard]

Of what ?

[brisk quarry]

APPLIO

i have 3.2.4

[low shard]

I didn't really check their changelog sorry

[brisk quarry]

not much

[low shard]

Wtf they used groq API for

It's used mostly for Chatbots, speech to text, or vision chatbots

https://github.com/IAHispano/Applio/blob/a0cc4100229c9231771bb5288e0b9ca2bd75a3c4/.github/workflows/release.yml#L86 oh

[brisk quarry]

applio so broken i change language and error
Connection errored out

something with schema = self._generate_schema_inner(obj)
File "C:\Users\Lukag\Desktop\Applio-3.2.5\env\lib\site-packages\pydantic_internal_generate_schema.py", line 929, in _generate_schema_inner
return self.match_type(obj)
File "C:\Users\Lukag\Desktop\Applio-3.2.5\env\lib\site-packages\pydantic_internal_generate_schema.py", line 1038, in match_type
return self._unknown_type_schema(obj)
File "C:\Users\Lukag\Desktop\Applio-3.2.5\env\lib\site-packages\pydantic_internal_generate_schema.py", line 558, in _unknown_type_schema
raise PydanticSchemaGenerationError(
pydantic.errors.PydanticSchemaGenerationError: Unable to generate pydantic-core schema for <class 'starlette.requests.Request'>. Set arbitrary_types_allowed=True in the model_config to ignore this error or implement __get_pydantic_core_schema__ on your type to fully support it.

If you got this error by calling handler(<some type>) within __get_pydantic_core_schema__ then you likely need to call handler.generate_schema(<some type>) since we do not call __get_pydantic_core_schema__ on <some type> otherwise to avoid infinite recursion.

For further information visit https://errors.pydantic.dev/2.9/u/schema-for-unknown-type

i better not use latest

[low shard]

Sorry but idk what's that 😭

[brisk quarry]

literally just language change

[low shard]

I meant like idk how to fix that

[brisk quarry]

i change language and the whole thing crash

[low shard]

Tbh always used English on Applio

[brisk quarry]

i wanted to change to english..

[low shard]

Oh

On Google colab it's set to English by default

[brisk quarry]

who tf

ong

@noble crater fix yo stuff

applio crash when changing language in latest

[brittle wing]

i need asome help everytime i use the real time voice changer and say like how are you doing and then say something else after that it will say how are you doing again

does someone know how to fix that

[burnt agate]

@low shard for some reason the format of the doc is causing commands to be copied incorrectly, for instance
git clone https://github.com/MFDGaming/ubuntu-in-termux.git

was copied as

git clone
https://github.com/MFDGaming/ubuntu-in-termux.git

causing git clone to error out with no specified repo

think this might happen on all commands that have a line break in them on the doc

[brittle wing]

nick could you help me?

[low shard]

Wtf, I + unitedshoes honestly copied all commands but it didn't copy it like this

Could you try to re-copy it ? Or try other commands

[burnt agate]

if I copy it directly from the doc it has a line break on my phone for some reason lol

if I copy it from anywhere else its fine

[low shard]

What issue

[low shard]

What about other commands?

[brittle wing]

dont worry someone helped me in this server thank you tho 🙏

[rare goblet]

Ayo? @brittle wing level 1 !!!

[burnt agate]

all of the commands that have multiple lines on the doc copy with a line break in them on my phone

idk why tbh

[gusty mortar]

hey pls whats the link of RVC doc ?

[rare goblet]

Ayo? @gusty mortar level 1 !!!

[burnt agate]

let me open the pdf in another app and try it again

[gusty mortar]

@low shard pls what is the link for RVC doc ?

[low shard]

Alr

[low shard]

I just tried removing the formatting on that command, could u try now?

[low shard]

What's ur PC gpu and what u looking for

[gusty mortar]

MSI Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz - Looking for record my voice then clone it with celebrity voices

[brisk quarry]

msi intel fr

[low shard]

That's CPU not GPU

[burnt agate]

same thing

[low shard]

Wtf I think the problem is either ur phone or google doc

Cus as u seen it's without the extension

[burnt agate]

let me try a different app

[low shard]

It's plain text

[low shard]

Wdym

[burnt agate]

to open the pdf

[low shard]

Don't you have the google doc app?

[burnt agate]

I do yeah

[low shard]

Ye try with that instead of opening it somewhere else

[gusty mortar]

sorry mb, its nvidia geforce gtx 1050

[brisk quarry]

forget it already

[snow arch]

hello why i cant seen voice model ?

[low shard]

U can only inference locally on that (basically convert ur voice to the model on ur PC) also are you looking for realtime for calls or pre recorded audios

[low shard]

Be specific

[gusty mortar]

I don't know anything about PCs, I just want to clone my voice with celebrity voices as a hobby

[snow arch]

i cant seen a model of voice in channel

[brisk quarry]

thats like wanting microscopy as a hobby but you dont know anything about it

[gusty mortar]

no just record my voice and convert to famous voice

[brisk quarry]

thats not how it works

[burnt agate]

I'm going to jump out of a window, I updated the app and now its fine

[low shard]

You can search rvc ai voice model at:

• #1175430844685484042
• #🔍│find-models
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://applio.org/models
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• #1159289738314919936
• #1191429836321849435
• make it yourself with our docs guides https://docs.aihub.wtf/essentials/how-to-make-voice-models/

[low shard]

Lmfao

I should prolly add this issue to the FAQ

[snow arch]

thx

[burnt agate]

well if someone has this issue in the future we have a answer lol

[low shard]

So, just record your voice, and use the recorded audio file converted to the person voice model right?

[low shard]

U just updated the google doc app right?

[burnt agate]

yup

[gusty mortar]

Yes perfect but how to do it? is there a tutorial or a link?

[low shard]

You could do it locally choosing between:

• Applio: A fork of RVC with some extra features like Applio TTS, same quality tho
• Mainline: The original RVC

But you won't be able to make models, just use them, and also it will be slow bc of your GPU

I suggest you to use Ilaria rvc zero that is faster

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[burnt agate]

to be fair I haven't used google docs on this phone since I reloaded the stock rom on it so it was prob like a version from 2018/2019

[gusty mortar]

thanks

[low shard]

[low shard]

Yw

@burnt agate btw guessing u wont install anything else on termux for now, after u run it sucessfully, pls go in app settings and tell me how much storage it took so i can be more accurate in the guide

I cant on my own bc i have other things installed

[burnt agate]

sure will let ya know as soon as its done

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[sterile bronze]

Shi, wrong chanel

[brisk quarry]

how do i continue training cuz it keeps making a new model

[dawn tendon]

I did convert but The song result says it's 0:00

For 3 minutes now...

So what do I do?

[low shard]

did u duplicate ilaria rvc ?

[dawn tendon]

Nevermind it did work

It needed to load

[low shard]

Alr

[grizzled cove]

Which one of these pitch algorithms is the best? The current one messes the pitch up a lot.

[tender ridge]

how do i train my own model

[quartz iris]

where do i actually find the voice changer application?

[low shard]

leave it at rmvpe+

[low shard]

whats ur pc gpu

[low shard]

realtime for calls? whats ur pc gpu

[quartz iris]

nvidia

[low shard]

full name

[quartz iris]

nvdia geforce rtx 4070

[rare goblet]

Ayo? @quartz iris level 1 !!!

[low shard]

alright

-rt

[azure marsh]

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

[low shard]

1st link

[quartz iris]

does that guide tell me exactly how to do it

step by step?

[low shard]

yes

[quartz iris]

nick it isnt working

[low shard]

im not really a wokada specific helper, but be specific

there could be alot of reasons why it isn't working lol

[amber grotto]

To create a public link, set share=True in launch().

Traceback (most recent call last):
  File "C:\Applio-3.2.5\env\lib\site-packages\pydantic\type_adapter.py", line 270, in _init_core_attrs
    self._core_schema = _getattr_no_parents(self._type, '__pydantic_core_schema__')
  File "C:\Applio-3.2.5\env\lib\site-packages\pydantic\type_adapter.py", line 112, in _getattr_no_parents
    raise AttributeError(attribute)
AttributeError: __pydantic_core_schema__```

[rare goblet]

Ayo? @amber grotto level 1 !!!

[amber grotto]

Recently I have some problems running applio

[azure marsh]

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[azure marsh]

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[rotund zephyr]

hey does anyone have the link to the updated colab rvc?

[lyric berry]

-rt

[azure marsh]

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

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[regal thistle]

noob here, what software is in the rvc tts picture

[azure marsh]

🍏 Applio-RVC-Fork

Enhanced RVC Variant: Optimized Performance Through Modifications, Built upon Mangio-RVC-Fork.

[novel vessel]

when i am pressing the button "Start" and i say something, it say it on the other voice after 2 seconds. how i can remove the delay?

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab
• 🆕 FaceFusion UI, by Nick088 Google Colab
• 🆕 FaceFusion NO UI, by Nick088 Google Colab

ℹ️ Note

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

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

[pastel oak]

Whats your gpu

[novel vessel]

NVIDIA GeForce GTX 1650 SUPER

[pastel oak]

You can use chunk 128 or 112 if it doesnt lag, or even lower. F0 det: rmvpe

Else, I recommend the fork wokada for gtx gpus, it reaches lower delay
https://rentry.co/Forkvoicechangerguide#download-nvidia

[distant geode]

I open the https start batch file and it opens command prompt and closes immediately. My GPU is RTX 4050 laptop. I don't know how to fix this issue

I got version MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.18a

[wild bobcat]

how many epochs would i need for a dataset that is 52 seconds long

[pastel oak]

https://cdn.discordapp.com/attachments/1159290161683767298/1159607580184821760/issue.mp4?ex=66fa7377&is=66f921f7&hm=fd618432a31ca57b1ef5d8530741f66a68a2f87ea66f0dd0762172f7d40ba38d&

[pastel oak]

You track the ideal trained point with tensorboard

[wild bobcat]

oh

okay

[pastel oak]

https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/

[wild bobcat]

thank you

[latent kettle]

Hey. Please tell me rtx msi rtx 4060 8gb oc edition is good for model training and for real time voice changing ??

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[patent pasture]

I had never trained model with turned off pitch guidence. If I turn it off, will it put pitch automatically or how does it work?

[pastel oak]

Should be fine for both

[pastel oak]

From what I remember, if it goes too far out of range, it will "break" and sound similar to falsetto voice. I dont remember if this was only in v1, or also in v2

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[opal kelp]

How do I clone an instrument, in this case the drums? Is it the same thing as cloning a voice or is it different? because I tried to make one and it didn't turn out as I expected

[lilac carbon]

hi, i need help creating an ai model

[low shard]

what's your pc gpu?

[lilac carbon]

how do i check, i forget

[low shard]

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[low shard]

I honestly never tried with instruments but ig its the same
Maybe @broken prairie knows more

[low shard]

tell me the full name btw

[lilac carbon]

amd radeon RX 7600

[rare goblet]

Ayo? @lilac carbon level 1 !!!

[low shard]

You could be able to do it locally, follow https://docs.applio.org/getting-started/installation#amd-gpu-support-windows

[lilac carbon]

is this safe?

[low shard]

Yes

[lilac carbon]

ok, doubble checking

[low shard]

We don't share malicious things, its all safe

Its open source which means you can even check the code yourself

[lilac carbon]

ok good

is this normal????

[lilac carbon]

ok its working, im wondering how do i get a certain characters voice and what if they have overlapped voices in the audios

[low shard]

well, you mean making or using a model?

[lilac carbon]

making a model

[low shard]

if they overlapped voices, you could try to separate them with the 6_HP-Karaoke-UVR model

[low shard]

you need to gather samples, like on youtube or whatsoever, clean the audio with UVR and then train using the tensorboard

[lilac carbon]

whats that

[low shard]

Ultimate Vocal Remover, the AMD Download is https://github.com/Anjok07/ultimatevocalremovergui/releases/download/v5.6/UVR_v5.6.0_setup_opencl.exe

the guide is https://docs.google.com/document/d/1_T9d4pmNg0iS7yzbYdLc0AjDG03cNPACzC0tXT6COUg/edit

[brittle wing]

Does somebody know why Weights isn't working?

[rare goblet]

Ayo? @brittle wing level 1 !!!

[broken prairie]

Like you do with voice models

The caveat is, the training audio must not be polyphonic or too low

[dusky wind]

Can anyone tell me if there are current disconnect prevention methods to keep Colab from timing out?

Also how do you know which pretrained model to choose from, and should I use one or should I train from scratch?

I'm new to this

Also there is something I don't get it. Is the X axis in this graph the steps or epochs? It says in the guide it is the steps, but mine shows the epochs. I have 6 audio files in my dataset, so shouldn't it show 6x500 which is 3000?

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for AMD GPUs

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

F0 Det.: rmvpe_onnx (suggested for all series)

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

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

[brave swallow]

are the pth filles work in rvc only ?

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab
• 🆕 FaceFusion UI, by Nick088 Google Colab
• 🆕 FaceFusion NO UI, by Nick088 Google Colab
• 🆕 EasyGUI, by Rejekts Google Colab

ℹ️ Note

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

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

[brave swallow]

do u know ?

[brittle wing]

know what

i just wanted to get the colab links

[brave swallow]

if pth are for rvc only

[brittle wing]

no idea

[brave swallow]

hmm

if i want to share model

[brittle wing]

i know nothing about the technical stuff

[rare goblet]

Ayo? @brittle wing level 4 !!!

[brave swallow]

where is the steps

[brittle wing]

i just make personal models and make em sing lol

[brave swallow]

gotcha

[snow hazel]

How do I seperate harmonies?

[solid grail]

the program can't detect any sound from my microphone even though i can get audacity to hear my microphone, i use an audio interface with an xlr mic which might be the issue

[odd shale]

Use Mel Karaoke model on X Minus or MVSEP.

[snow hazel]

I see Mel-RoFormer on X minus

[odd shale]

Click on "keep backing vocals" option

And you'll see a Mel Roformer Kar model, that one is Mel Karaoke

[snow hazel]

OMG THAT'S PERFECT!

[distant geode]

'MMVCServerSIO.exe' is not recognized as an internal or external command it says

[snow sinew]

-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, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab
• 🆕 FaceFusion UI, by Nick088 Google Colab
• 🆕 FaceFusion NO UI, by Nick088 Google Colab
• 🆕 EasyGUI, by Rejekts Google Colab

ℹ️ Note

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

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

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for AMD GPUs

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

F0 Det.: rmvpe_onnx (suggested for all series)

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

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

RVC Colabs and Spaces

⠀

Google Colabs

⠀
AICoverGen-WebUI
Useful for making quick covers, by Hina.

AICoverGen-NoWebUI
Useful for making covers, doesn't include a UI, by Ardha, by Eddy, Hina and Gdr.

RVC Disconnected
To train new voice models, by Kit Lemonfoot.

EasyGUI
The OG interface, by Rejects.
⠀

Made with love by Ilaria 💖

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[dense musk]

do i download realtime voicechanger

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

[sonic agate]

what exactly D and G means?

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for AMD GPUs

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

F0 Det.: rmvpe_onnx (suggested for all series)

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

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[low crag]

what to do after install

[brave garnet]

RVC Colabs and Spaces

⠀

Google Colabs

⠀
AICoverGen-WebUI
Useful for making quick covers, by Hina.

AICoverGen-NoWebUI
Useful for making covers, doesn't include a UI, by Ardha, by Eddy, Hina and Gdr.

RVC Disconnected
To train new voice models, by Kit Lemonfoot.

EasyGUI
The OG interface, by Rejects.
⠀

Made with love by Ilaria 💖

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

[distant geode]

😭

I'm just gonna reinstall it\

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[brittle wing]

-local

[azure marsh]

🖥️ Local stuff

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

Credits to Faze Masta and Antasma for compiling these links.

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

[tame mica]

oh hey i think i know u

[brittle wing]

you do?

[tame mica]

i think ive seen some of your comments on rincon agasddfaf

[brittle wing]

oh lol

[tame mica]

ive also seen a bit of your videos before

[brave garnet]

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs

F0 Det.: rmvpe (suggested for all series)

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

Advanced Settings

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

⠀

Made with love by Ilaria 💖

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[brittle wing]

Hey guys, quick question, if i run into trouble after installing this is there a way to uninstall it so i can make changes?

Also i'm only on an Nvidea 3070 can that run it?

[pastel oak]

can be anything, if you have antivirus program it deleted files, python cant find the file (if its on an external drive so just try putting the whole thing in a different drive and try again, also different locations), also no spacebars in the url path

[charred beacon]

heyy , i want hindi models where can i find it ?

[odd shale]

Hi Shady!!!!!!!!!!

I missed u a lot.

[grizzled jungle]

hello, i have a question about the real time voice changer where is it

[rare goblet]

Ayo? @grizzled jungle level 1 !!!

[low shard]

whats ur pc gpu

[grizzled jungle]

6700 xt

[low shard]

-rt

[azure marsh]

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

[low shard]

1st link

its the wokada fork, optimized performance

[dusky wind]

Update: I setted up my dataset wrong, and it didn't parse correctly, it wasn't actually training anything, that's why the epochs were the same as the steps

Also new question in RVC 2 disconnected, doesn't this section of tensorboard work to test the audio?

[pastel oak]

uwu

[shrewd bolt]

what extra does

[zealous mirage]

Hey, everybody, a quick question. Are you supposed to have gaps between phrases in the dataset? Mine are minimal at the moment, so that the voice doesn't overlap itself

[round fog]

The difference is negligible

It could help the dataset and if it does it would be hard to tell a difference

so if you have extra time do it if you want

[zealous mirage]

Thanks!

[mental lantern]

hello

how to do for mac os guys?? plz

[tribal apex]

in trying to use the voice changer AI i get this issue

"The error message indicates that the setSinkId function is being called, but it is not available or supported in the current context."

it seems my browser is not working with my microphone. i am not sure how to fix this

chat GPT isnt helping me much with this issue

maybe all i need to do is try another web browser. waterfox must be too secure for this app

i will try firefox

[young halo]

What's the model on UVR to remove backing vocals?

[tribal apex]

The error message you are encountering relates to multiple issues with the AI voice changer's pipeline, specifically issues with loading models and using ONNX and PyTorch. Here's a breakdown of the causes and potential solutions:
Key Problems:

PytorchStreamReader failed reading zip archive: This indicates a problem with loading the PyTorch model. It suggests that the model file (probably a .pt or .zip file) is corrupted or incomplete.
INVALID_PROTOBUF: This occurs when the ONNX model (content_vec_500.onnx) fails to load, likely due to a corrupted or improperly downloaded model file.
Torch Serialization Error: This indicates an issue when trying to deserialize (load) the model from disk using PyTorch.

[rare goblet]

Ayo? @tribal apex level 1 !!!

[crude bolt]

Can you tell me what software you installed , i am not familiar with the term "voice changer AI".
What did you install, where did you get it from ?

before we can help you - lets avoid assuming things and verify we talk about the same thing.

[tribal apex]

MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.15 this is what i downloaded from hugging face

[crude bolt]

fair enough. What graphics card do you have ?

[tribal apex]

it runs but my mic isnt working. i thought it was an issue with my browser

[crude bolt]

is your mic working on discord ?

[tribal apex]

yes

[crude bolt]

alright do you feel comfortable to discuss your problem with me on voicechannel or do you rather wanna keep typing ?

[tribal apex]

i can get it to work with the program but not to change my voice.

sure a sec

i am in VC

[crude bolt]

lol now i got technical issues 🤦‍♀️

[tribal apex]

i can wait, 😄

[Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline... i get this error

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[azure marsh]

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[brave garnet]

RVC Colabs and Spaces

⠀

Local Forks 🖥️

⠀
Mainline RVC
Original project, suggested for advanced users,
by the RVC-Project team.

Applio
Simplified, suggested for all, by the Applio team.

RVC Studio
Simplified, suggested for all, by SayanoAI.

Mangio-RVC
Simplified, may not be supported anymore, by Mangio621.

AICoverGen
Simple yet great way to make covers, by SociallyIneptWeeb.

Replay
From the greators of weights.gg, excellent product for everyone.
⠀

Made with love by Ilaria 💖

[knotty moth]

keep it as is, then apply labeling in audacity

[azure marsh]

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[ashen vine]

hello how to troubleshoot error when downloading model, everything was working fine but now i cant convert or download voice models onto huggingface

[hearty peak]

-rt

[azure marsh]

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

[low shard]

what are u using and what download link

[ashen vine]

rvc v2

https://huggingface.co/Goldeath/Gow/resolve/main/KratosOG.zip?download=true

[low shard]

yea but like which

local mainline? A google colab link?

[low shard]

the model files seems fine

[next wharf]

"Realtime_Voice_Changer_on_Colab" a question about the appearance of the error "ERROR: Cannot install -r requirements.txt (line 19) and fairseq because these package versions have conflicting dependencies.
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts" and "
ModuleNotFoundError: No module named 'pyngrok'"
can help please.

[ashen vine]

https://huggingface.co/spaces/Clebersla/RVC_V2_Huggingface_Version

[rare goblet]

Ayo? @ashen vine level 1 !!!

[ashen vine]

i just go there and type in the download link

[low shard]

dont use that, not updated and its on cpu not zerogpu

use ilaria rvc zero which uses A100 gpu for free

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[low shard]

what google colab link u using

[next wharf]

https://colab.research.google.com/github/Misadov/voice-changer-colab-fix/blob/master/Realtime_Voice_Changer_on_Colab.ipynb#scrollTo=lLWQuUd7WW9U

[low shard]

where did you find that? Dont use that

whats your pc gpu btw?

You could do it locallly if its good enough

[next wharf]

gtx1650

[rare goblet]

Ayo? @next wharf level 1 !!!

[low shard]

oh thats good enough to do it

-rt

[azure marsh]

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

[low shard]

1st link which is optimized for low gous

its a wokada fork

if u really dont wanna do it locally i can give you the cloud options

For Realtime Voice Changing for Calls online (for who doesn't have a good pc, YOU CANT DO THIS ON MOBILE):

• Google Colabs (4 hours daily of free T4 gpu, easy to use, require only a google account) :
• How to use Original W-Okada's Voice Changer Google Colab
• Modified W-Okada's Voice Changer Google Colab
• Kaggles (30 hours weekly of better GPUs, T4x2 & P100, harder to use, requires an account and a phone number)
• Original W-Okada's Voice Changer Kaggle
• Modified W-Okada's Voice Changer Kaggle

but i would suggest to do it locally as you got no limits

and got a good enough gpu

[azure marsh]

Suggestion for @bold saddle

💻 (Local) Realtime Voice Changer

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[azure marsh]

☁ (Online) W-Okada - Colab Version

• How to use Original W-Okada's Voice Changer Google Colab
• Modified W-Okada's Voice Changer Google Colab
• Original W-Okada's Voice Changer Kaggle
• Modified W-Okada's Voice Changer Kaggle

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[autumn vine]

@azure marsh помоги

[azure marsh]

Wassup I'm Automaze!

• My prefix in this server is -
• Interested in how I'm built? I'm actually open source!
• Forgot a specific command? Try /help or -help

[brittle wing]

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[brittle wing]

-hf

[azure marsh]

<:huggingface:1179800228946268270> Hugginface Spaces

• UVR5 UI, by Eddy and Ilaria Huggingface Spaces
• Ilaria RVC Zero, by thestingerx Huggingface Spaces
• RVC⚡ZERO, by r3gm Huggingface Spaces
• Applio, by IA Hispano Huggingface Spaces
• 🆕 FaceFusion UI, by Nick088 Huggingface Spaces

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

[flint solar]

yo how do i pin the graphs

[idle osprey]

my rvc does not work how can i fix

only the preset something called beatrice jvs speaker

[rare goblet]

Ayo? @idle osprey level 2 !!!

[low shard]

are u using wokada for realtime voice changing?

Are yo following a youtube tut?

And also whats ur pc gpu?

[idle osprey]

4060 ti

[idle osprey]

yes

[low shard]

that's good

[low shard]

never follow youtube tuts

[idle osprey]

idk uhh

whats that

[low shard]

they are outdated

[idle osprey]

i know

[low shard]

-rt

[azure marsh]

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

[idle osprey]

i didnt do anything of them

[low shard]

Use the first link

its a wokada fork

it has optimized performance

[idle osprey]

i dont know what half of those means

but

alright

i already have a version installed

not from the youtube tutorials

[low shard]

wokada is a program specifically for using RVC (retrieval-based-voice-conversion) models (the one used also for ai covers yk)

Fork = modified version

optimized = better

I would suggest you to get the fork so you have a more optimized version personally

[idle osprey]

no

i just want the rvc thing to work

do i need fork thing

for that

[low shard]

it would help you to get better performance, especially if u are gonna run it on a game (but in general it helps)

if u really dont want to i mean u can not use it

[idle osprey]

no

its a mining pc

ill just try it and then leave it

[low shard]

u mean those gpus used for mining bitcoin before?

[idle osprey]

what does that man

[idle osprey]

yes

[low shard]

if you want, you can use the fork that can help you

[idle osprey]

what is it gonig to help me with

[low shard]

What and where did you get whatever program you are using ?

[idle osprey]

uhh

howaksknah

howa

idk

the ai thing

[low shard]

better performance, runs basically better

[idle osprey]

webiste

[low shard]

?

website?

could u send the link?

[idle osprey]

hold on

[low shard]

just to check

[idle osprey]

https://huggingface.co/wok000/vcclient000/tree/main

[low shard]

oh yea

u got the normal wokada

also be sure u got the onnxgpu cuda version as you got an rtx 4060

[low shard]

be sure u got a VAC https://rentry.co/VoiceChangerGuide#virtual-audio-cables-mandatory

[remote steeple]

i just got muted

wtf is that

oh yeah

the stupid thing

yeah i got that

but it appears that its not about audio and stuff

i just cant use the rvc audios

@low shardwhat do i do

[rare goblet]

Ayo? @remote steeple level 1 !!!

[idle osprey]

yeah

[low shard]

it seemed to be because of sending messages too fast

[idle osprey]

yeah its fine i got unmuted

[low shard]

are you tryna use it for prerecorded audios or for realtime in calls?

[idle osprey]

realtime

[idle osprey]

realtime

idk if rvc even works

cuz i just tried 2 markiplier rvcs

and they do no sound at all

i tested it in call too

[low shard]

did you setup correctly the audio inputs and outputs? https://rentry.co/VoiceChangerGuide#audio-setup

[idle osprey]

yeah

i already did that

[low shard]

did u try also with other voice models ?

[idle osprey]

uhh

i tried saying konichiwa and stuff with the japanese bundles i think

they did nothing too

just the female and male thing works

[pastel oak]

only beatrice model working is a failed install

need to delete pretrain folder and model_dir and let it reinstall those

[idle osprey]

oh

[pastel oak]

or use the fork that should work smoothly

[low shard]

🙏

[idle osprey]

fork

okay

what where i

download

[pastel oak]

https://rentry.co/Forkvoicechangerguide#download-nvidia

[idle osprey]

got it

is this just another voice changer

or the same

[pastel oak]

its a modified version of the one u know

[idle osprey]

i dont want modified version

[pastel oak]

y

[idle osprey]

shit

[pastel oak]

no

XD

[idle osprey]

wheres that

[pastel oak]

in mmvcserversio

[idle osprey]

where do i

[rare goblet]

Ayo? @idle osprey level 3 !!!

[idle osprey]

what is gonna reinstall

[pastel oak]

delete

then start start_http

[idle osprey]

uh

[idle osprey]

oh

okay

uhh

the moment i downloaded it

nevermind i dont wanna mention it

it was too dumb

do i dellete model dir static

[pastel oak]

no

[idle osprey]

ok im runningi t

do i need phyton for this thing

[pastel oak]

its already installed

[idle osprey]

what

[pastel oak]

it already

comes with

python

[idle osprey]

oh

okay

thank you

it worked

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[desert sphinx]

my RVC gui doesnt have mangio crepe

how to add?

nvm, found it

[rare goblet]

Ayo? @desert sphinx level 1 !!!

[brittle wing]

@pastel oak

[unkempt moss]

these voices confuse me! i'm not sure what settings i need to tweak in order to sound more like the model's preview

[pastel oak]

You can see the wave from 0:20 - 1:00 goes up to 20 kHz. You multiply it by 2, just a thing you always do, which is 40 = so 40000 sample rate

However, 1:00 - 1:20 only reaches consistently around 16 kHz (multiplied by 2 = 32000) while having some spikes up to 20k, but its mostly 16k

[brittle wing]

What about, when it's giving me 44k?

[pastel oak]

This is a divided opinion, people will tell you to use 32000 because sections like 1:00 - 1:20 and 2:20 - 3:00 can not reach up to 40000. Others will tell you to use 40000 because it will be just fine

[pastel oak]

Same thing about the divided opinion. I'd use 40k probably

[brittle wing]

Or just cut out the ones that are 40k and keep them and scrap the other

[pastel oak]

Yeah, if those 40k ones are very clear, you might get away with a short dataset

The more clean your dataset is (studio vocals - no background nothing) the better results

[pastel oak]

Voice changer? Mess around with tune/pitch

If male to female voice, usually range of 8 - 13 will do it

[brittle wing]

Thank you, I appreciate the help

[unkempt moss]

For the ones that are in #1175430844685484042 like the celebrity voices? I'm not sure what any of the jargon used there means 😭
i.e Epochs: 1000
RVC v2, RMVPE, bs 8, default pretrain
Sample rate: 40k

Do I need to change anything on the application to reflect this?

[pastel oak]

All of that is not relevant for you, has no correlation to wokada. Only important thing is RVC v2 info

On the application you use the settings recommended for your GPU and then pray the voice model you selected works well on realtime

[shrewd drift]

people how do you separate multiple voices from a track

[low shard]

if they overlapped voices, you could try to separate them with the 6_HP-Karaoke model in UVR

[unkempt moss]

I don't see a tune setting, only a pitch. Are these the same? I'm watching a youtube tut and they have tune but no pitch. I have pitch but no tune

[rare goblet]

Ayo? @unkempt moss level 1 !!!

[pastel oak]

Same thing

[shrewd drift]

where can i download this

[pastel oak]

-uvr

[azure marsh]

Suggestions for @shrewd drift

Ultimate Vocal Remover

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

[shrewd drift]

thanks guys

[rare goblet]

Ayo? @shrewd drift level 1 !!!

[low shard]

also would suggest to check ur pc gpu, if its good enough do it locally or else would suggest cloud uvr

[terse siren]

how do i make a voice model

[low shard]

whats ur pc gpu

[low shard]

yep u cant train locally (on ur pc) with that

u need cloud (remote good pc)

[terse siren]

whats cloud 😭

[low shard]

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
  • Applio (ui)
  • Mainline (UI)
  • RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide as of right now)
• Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
• Mainline (UI, No guide as of right now)
• Applio (UI, No guide as of right now)

those are all cloud ways

you use a remote pc that is better than yours to run the program

the best option is kaggle

the easiest is colab

[terse siren]

thank u

[low shard]

yw

almost each link is a guide

[terse siren]

i dont understand 99% of what its saying to me

i cant even find the mainline 😭

oh crap nvm

[low shard]

its the blue text

its an hyperlink, so a blue text that redirects u to a site when u click it