✨│ai-help

[red kayak]

https://www.googleadservices.com/pagead/aclk?sa=L&ai=DChcSEwinjpbE_-6FAxV7OQYAHaUlA5QYABABGgJ3cw&ase=2&gclid=CjwKCAjw88yxBhBWEiwA7cm6pQal3bcpcBKrq8slADtARpNP5iFhRRb0QJhfMFyevTOQuwatg5G_WhoCdaUQAvD_BwE&ohost=www.google.com&cid=CAESVeD2RuagC5H40kwKxDgOX1rTeh4WLtTwbRtc0YJnQQKKT5DIMFhbhLp08WyllgSFnwZp5yOvPsNr7kaUs32SMEEkaPDiT7M6dPd55_UVLBt2Px_WIP0&sig=AOD64_0LaC-GeEl3pRIXZG4iVzp3CpWgdg&ctype=5&q=&nis=4&ved=2ahUKEwj6_4zE_-6FAxXLBNsEHUhQC0EQ9A4oAnoECA4QCA&adurl=

6 gigs can do stuff so

[fiery raptor]

what do I do with the passowrd IP?

[verbal quiver]

[red kayak]

[fiery raptor]

i found out at the end

i said

i found out

its okay

lmao

[blissful scaffold]

is there something that can isolate different type of instruments to get stems from a song?

[brittle wing]

can someone send me the direct gradio link

for ai music

[proper shale]

gradio links vary from colab to colab so

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

try this out, see if it works

[tough fiber]

bro please can someone tell me solution

[proper shale]

make sure you're using a path to a folder with audios, not an audio itself or a zip

[tough fiber]

so my data here

im pasting /notebooks/ixtivhdata > this

folder path not data path

i did 10 times and it work so idk whats wrong

[proper shale]

frankly idk why that's happening

i say, move ixtivhdata to inside the mangio fork folder...

[finite galleon]

try /notebooks/ixtivhdata/cyberprincessdata.wav

Oh wait this is training

My bad

[tough fiber]

yea

[rare goblet]

Ayo? @tough fiber level 10 !!!

[finite galleon]

My guess is that either the path is somehow invalid or the audio file is corrupted

[lost relic]

what does resample mean(model inference interface)

[wraith coral]

RVC doesnt respond when i click on change voice

[violet heron]

-upload

[azure marsh]

Suggestions for @brittle wing

huggingface.co upload Guides

English guide

Link: https://rentry.org/fdg_guide_newer
Credits: FDG

Spanish Guide

Link: https://rentry.org/AI_Hub_guide_esp
Credits: Julia (ailen2091)

Hindi guide

Link: https://rentry.org/AI_hub_guide_hi
Credits: Enes

Turkish guide

Link: https://rentry.org/AI_hub_guide_tr
Credits: Enes

[wispy burrow]

If you mean uploading your Model to HuggingFace
https://docs.aihub.wtf/essentials/voice-models/#uploading-to-hugging-face

[violet heron]

Translate it into your language

[wraith coral]

Which one am i supposed to click on to open the RVC, go-realtime-gui.bat?

[violet heron]

no the go web bat

Might not be there

Put this in the main folder and run it if it’s not there

[wraith coral]

there is one yeah

[violet heron]

(If you don’t want to do that open the infer web py file with the python installation it comes with)

[fluid sage]

I can't find Turkish voices

[rare goblet]

Ayo? @fluid sage level 1 !!!

[fluid sage]

I don't understand

[wraith coral]

Uh it says error when i try to voice change

[fluid sage]

how to level up

[wispy burrow]

What is the error?
Look in the CMD

[grizzled zealot]

hello, i kind of need a second opinion since i think i'm blind
https://files.catbox.moe/qvt0ym.png < is this any sign of overtraining or...

[rare goblet]

Ayo? @grizzled zealot level 1 !!!

[wraith coral]

Traceback (most recent call last):
File "C:\Users\kkkkkkk\Pictures\RVC1006Nvidia\runtime\lib\site-packages\gradio\routes.py", line 321, in run_predict
output = await app.blocks.process_api(
File "C:\Users\kkkkkkk\Pictures\RVC1006Nvidia\runtime\lib\site-packages\gradio\blocks.py", line 1007, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "C:\Users\kkkkkkk\Pictures\RVC1006Nvidia\runtime\lib\site-packages\gradio\blocks.py", line 953, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "C:\Users\kkkkkkk\Pictures\RVC1006Nvidia\runtime\lib\site-packages\gradio\components.py", line 2076, in postprocess
processing_utils.audio_to_file(sample_rate, data, file.name)
File "C:\Users\kkkkkkk\Pictures\RVC1006Nvidia\runtime\lib\site-packages\gradio\processing_utils.py", line 206, in audio_to_file
data = convert_to_16_bit_wav(data)
File "C:\Users\kkkkkkk\Pictures\RVC1006Nvidia\runtime\lib\site-packages\gradio\processing_utils.py", line 219, in convert_to_16_bit_wav
if data.dtype in [np.float64, np.float32, np.float16]:
AttributeError: 'NoneType' object has no attribute 'dtype'

[civic meteor]

uhhh anyone knows how to fix this

[wispy burrow]

I assume your path to the audio file is wrong.

[wraith coral]

nope its correct

[wispy burrow]

Can we get a screenshot of RVC?

[wispy burrow]

Did the link end with ?download=true?

[violet heron]

You didn’t put the path to your audio

[wraith coral]

Why do i need an audio? Don't i need a model

[wispy burrow]

The language for the demo should not mattter.

[violet heron]

Maybe but I would stay on the safe side

[wispy burrow]

Seriously?

[wraith coral]

.

[wispy burrow]

That makes no sense if you ask me...

[wraith coral]

i need it to convert the voice when i speak not sing it tho

[wispy burrow]

Oh...in realtime...

[wraith coral]

yep

yeah but it doesnt respond when i click on convert

no in the go-realtimegui.bat

[wispy burrow]

Try asking in #🔍│help-w-okada
That's the channel for realtime voice conversion

[wraith coral]

thats web.bat

not the realtime

[brittle wing]

Language doesn't matter WTF

[wispy burrow]

What is the song you used?
I recognise it, but don't know the song.
I'm trying to gather some audios to test my models.

[violet heron]

Shouldn’t but accents mix depending on the index ratio

tell that to the QCs

[brittle wing]

The same person called waste of rvc zero also rejected my model once with the statement that it's an Acapella not a raw rvc output...like it was an Applio output Just because it contained "Denoise, dry, de-reverb" in the file name

They prob thought I did it with cover gen webui and then extracted the lv with xminus or something

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 Spain

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

🇫🇷 France

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

🇩🇪 Germany

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

🇮🇹 Italy

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

[brittle wing]

Seriously

[proud elbow]

huh?

[brittle wing]

Like this person made me nervous but then thanks God Litsa came and told me my model was noisy

Probably

Why is this person even a QC

[jagged flax]

Whats the simpliest way for rvc? I want mega easy, easygui v2 on google colab good?

[wispy burrow]

Thank you! Much appreciated.

[worthy bridge]

what is the best uvr model for instrumental? I want to make clean versions of music for youtube (yes they get like HELLA views)

[proud elbow]

ensemble with BS roformer and Inst HQ4, personally when trying to make some retro song covers

[jagged flax]

?

[proper shale]

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

this is it, pal

[jagged flax]

pal?

[rare goblet]

Ayo? @jagged flax level 1 !!!

[proper shale]

?

just a way to mention somebody

[jagged flax]

okay

[proud elbow]

wait, u mean instrumental stem for covers/karaoke, not specific instrumental stem (drum, bass, etc)?

[worthy bridge]

yes just instrumental

[violet heron]

Waste of RVC zero more like waste of a qc

[brittle wing]

Correct.

[pale brook]

help it sounds like i have a major speech impediment and it cuts off the end of my sentences

[proper shale]

try another model and check the settings for ur gpu listed in

https://rentry.co/VoiceChangerGuide

[mystic marsh]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[rare goblet]

Ayo? @mystic marsh level 2 !!!

[tough fiber]

how can i fix this. 10x

[rotund prawn]

I got this error on rvcDisconnected

how can i fix this?

[odd shale]

Are you using Mangio locally or on Paperspace?

[proper shale]

list index out of range happens when ur audio is screwed up from my experience...

[rotund prawn]

🙀

[proper shale]

what file format did u use

for ur dataset

[rare gale]

Tell me, is there an RVC with text? Like...to voice the text, not the audio. on site 2 only RVC - Appilo and Ilaria

[rotund prawn]

wav

and it's 20 seconds long

[tough fiber]

on paperspace

[velvet flower]

im trying to download it in the cmd thing and it keeps coming with with this warning "The system cannot find the path specified."

[odd shale]

That's pretty weird, i never got that error before on the time i used Mangio on paperspace.

[proper shale]

check if there's any spaces in your folder names

[velvet flower]

wdym?

what folder names

[proper shale]

the folder you placed RVC in

[proper shale]

huh... weird

how did preprocess/feature extraction go?

[rotund prawn]

well

without error

[proper shale]

send a screenshot

[velvet flower]

it happens when i try and instal AICoverGen

[rare goblet]

Ayo? @proper shale level 98 !!!

Ayo? @velvet flower level 1 !!!

[proper shale]

tbh i think you should use normal RVC

[velvet flower]

wdym

[proper shale]

its not gonna give u everything mixed down but u have more control over everything

u can use whatever filtering you want

[velvet flower]

i see

[velvet flower]

Where do i get the normal one

[proper shale]

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

[rotund prawn]

I closed the tap. but I can run colab again

[proper shale]

wait no

[velvet flower]

hm

[rotund prawn]

I will have the same error already

[proper shale]

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

well that looks fine

[rotund prawn]

okeyyy

I will click the training button

omg

it worked

I have tried it on 3 differnt accounts

they were failed

THANK YOU SENİOR SUKUNA MJ ❤️

[proper shale]

yaaay

[proper shale]

HAHAHAHAHAHA you're welcome ;)

[rare goblet]

Ayo? @brittle wing level 6 !!!

[proper shale]

pretrain_type in Set Training Variables should have some of the main pretrains used out there

otherwise... i think maybe you'd have to tweak the code

[velvet flower]

how do i import my own rvc

[proper shale]

put the .pth on the weights folder

[velvet flower]

huh

[proper shale]

index on logs folder

dmr seems too short with too few voices to be stable enough imo

but titan eh

it seems to be good

[velvet flower]

wheres the weights folders

[proper shale]

somewhere inside the rvc files

i cant recall correctly which one it was

[velvet flower]

is there a tutorial on how to use this thing bc it look hella confusing compared to the other one

[proper shale]

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

this should have it

[marsh vapor]

how can i make a model on #1159514067187277865 ????

[wispy burrow]

I think it's in the assets folder.

[proper shale]

use QCBot to upload ur model

[marsh vapor]

but what type of commands do i need to use, i mean, what should i paste on "link" and "Number of epochs"?

[proper shale]

link should be a download link for your model, number of epochs is pretty self explanatory

unless you've trained it on kits or something

[marsh vapor]

so, download link? is like i made the link manually?

[rare goblet]

Ayo? @marsh vapor level 1 !!!

[proper shale]

again, a download link for your model

you must upload it on Huggingface or somewhere similar

[marsh vapor]

ohhh, srry

ok ok

[proper shale]

https://docs.aihub.wtf/essentials/voice-models/#uploading-to-hugging-face heres how to upload to Huggingface in case you wanna know

[marsh vapor]

ty

[proper shale]

yw :)

[haughty saddle]

why wont it let me upload?

[iron laurel]

whats ModuleNotFoundError: No module named 'soundfile'

[pastel oak]

what program are you using

[odd shale]

You must put the .pth and index fies of your model in a folder, then .zip up that folder.

You did that?

Also, do you know why the QCBot told you your model had artifacting?

You didn't clean it properly?

[frank tiger]

can someone help me

[thin stump]

Sounds great on my headphones 🤔. But some QC's only look at the spectogram and not the sample itself so they may be right

[mellow token]

what’s the issue

[odd shale]

1SKI

[mellow token]

LEO

[odd shale]

1SKI

[mellow token]

How’s you been

[odd shale]

Doing good, recollecting stuff.

[mellow token]

That’s good that’s good

Glad to hear you’re doing great

It’s a shame active people aren’t here as often anymore

[odd shale]

Welp, i'll stay on the server no matter what.

Ain't going to leave.

Altho i suggest you all to wait till summer.

[full torrent]

hi guys i'm trying to make my own model off my own voice how can I do that is there a tutorial somewhere?

[odd shale]

Maybe the activity will increase for obvious reasons haha. (Let's just stay optimist anyway)

[odd shale]

Read our docs, don't watch video tutorials since these are outdated.

-rvc

[azure marsh]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[warm depot]

i cant hear myself on rvc

[odd shale]

Ask for help in the #🔍│help-w-okada channel

[frank tiger]

idk how to hear in discord

my voice

[warm depot]

im using rvc not okada

[odd shale]

Oh, okay then what do you mean with "i can't hear myself on rvc"?

You're using RVC's built in voicechanger?

[warm depot]

everytime i click the image it wont come up

[odd shale]

You're using w-okada, right?

it seems like you're not using RVC.

[warm depot]

whats the difference

[frank tiger]

i dont hear myself in discord

[rare goblet]

Ayo? @frank tiger level 1 !!!

[frank tiger]

idk how to setup for discord

[violet heron]

Download VAC and set the output on Okada to Line 1 and on discord select it as the input

https://rentry.co/VoiceChangerGuide#virtual-audio-cables-mandatory

[warm depot]

can anyone help me with "is not accepted orgin" how do i fix this

[frank tiger]

what is okada

[warm depot]

how do i send a photo here

[violet heron]

Are you using RVC or realtime voice changer client demo

Either way you have to do it to use the voice changers on discord

[violet heron]

That is w-okada

[violet heron]

Read here

Also I would consider checking other parts of the guide for recommended settings and such

[frank tiger]

k

thanks

[gusty kestrel]

@loud flint, I have found 2 results that match your search!

waltman13

Thomas the Tank Engine (Thomas & Friends/Magic Railroad)

• Uploaded: <t:1692748800:d>
• Server: AI Hub
• Likes: 0

Epochs

200

Technology

RVC

Algorithm

Rmvpe

Requested by rings2006wilson · May 3, 2024, 12:23 AM

[warm depot]

how do i fix this the model is not getting loaded

[rare goblet]

Ayo? @warm depot level 2 !!!

[hollow raven]

can i turn on that i hear my voice ? myself?

Like this Monitoring in W-okada , there you can Input, Output and Monitoring

[balmy prawn]

**RVC1006Nvidia\RVC1006Nvidia>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897
'runtime\python.exe' is not recognized as an internal or external command,
operable program or batch file.
** Slightly confused, is there an issue with my Python install ? If so any idea how to fix it ?

[violet heron]

Did you unzip everything

[balmy prawn]

Update : I was just tired and confused some folders, all good

[rare goblet]

Ayo? @balmy prawn level 4 !!!

[fallow nexus]

hey all i downloaded mainline

and im trying to use the ui change

but i dont know where to extract the zip file

or maybe its right but the how do i run it'

[violet heron]

Go-web bat file

If it’s not there put it in your RVC folder and run it

[fallow nexus]

ok i have this blue box with some chinese and other stuff

where do i put my taining data

[rare goblet]

Ayo? @fallow nexus level 1 !!!

[fallow nexus]

👀

[violet heron]

Wait till a gradio ui opens

[cobalt cradle]

when i go to voice settings on discord (on input device), it doesn’t even hear me speak. i’m not sure how to get it to work?

[violet heron]

In your browser

[violet heron]

Is your output on your voice changer set to the audio cable

[fallow nexus]

yeah i have something called gui

[cobalt cradle]

there’s one that has “default” and one that says “communications” and “microphone array” at the beginning of it but they both have “audio cable” at the end

[fallow nexus]

its got a pth file and a index file

my training data is one 9 minute long mp3

do i just path to the mp3?

[violet heron]

Yes

Wait

Are you trying to make a cover or model

[violet heron]

test both

[fallow nexus]

i want to have a model

[violet heron]

Make sure you have VAC Lite or other audio cable software installed

[violet heron]

have like use or make a new one

[fallow nexus]

make a new one

model is the baseline for voice covers right

and then covers use models and other audio sources together?

[balmy prawn]

Did a training via this : https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS Any clue why index file isn't working ?

[fallow nexus]

ive used models on the web version i just wanted to get my feet wet training

[violet heron]

That is not your index file

The index file should start with added

[violet heron]

Oh k

[balmy prawn]

https://drive.google.com/drive/folders/1wet4AHPgKgyicDUH9cgyQdHLSSfGnZo8?usp=sharing iI don't see it

[violet heron]

On the training section enter the path to your dataset, epoch count, change anything else if needed then train

[violet heron]

Did you run train index?

[proud elbow]

does your model name has some unicode asian/special characters?

[fallow nexus]

WAIT

i dont think im in the training gui

i think im in the cover gui

[balmy prawn]

none I didn't even use a space, just underscore + lowercap

[fallow nexus]

this is for covering using other models right

i wanted to make my own

[balmy prawn]

I think that I went through all the steps

[violet heron]

No that’s the real-time ui

[fallow nexus]

shit

[violet heron]

Try running train index again

[balmy prawn]

FUCK

Maybe not

I'll give it another shot/

[violet heron]

The go-web bat file is the right one

[fallow nexus]

i literally dont have one

[violet heron]

Now you do

Put it in the same location as the other bat file

[fallow nexus]

ok i see

i was just seeing html didnt realize this was a file lmao

[rare goblet]

Ayo? @fallow nexus level 2 !!!

[balmy prawn]

Yep, mistake's on my end, forgot to train the index, thanks for the help folks.

There's an installer for Mangio ?

[violet heron]

Use mainline/original RVC instead

Mangio is too bloated

https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/tree/main

[balmy prawn]

That's also what I've been rolling with

[rare goblet]

Ayo? @brittle wing level 1 !!!

[balmy prawn]

If you can run Discord, you can run it

[violet heron]

For covers no

Should be faster then Mangio

Should be similar to Mangio

[fallow nexus]

ok im on feature extraction

when is it done and i can click the train index button?

[violet heron]

Yes then click train model

[fallow nexus]

sorry im asking when will it be done

How will i know its done

[violet heron]

When it shows a output instead of the little gif

Or all feature done

[fallow nexus]

so i have now 334 all 264 wav

so i can train index now?

IT SAID DONE

[violet heron]

No it should say something like all feature done iirc

[fallow nexus]

SORRY

YAY

train index started

ok more questions

i clicked the train model button, and it ran some processes but it hasnt updated since. the last line talks about buckets being rebuilt idk

am i going to be idling my pc for days?

[violet heron]

What’s your GPU

[fallow nexus]

im not sure if it was using my gpu

[violet heron]

Sorry I didn’t ask this earlier

[fallow nexus]

i hnave a radeon 540?

nothing crazy

radeon 590 my bad

ope

just not enough horsepower?

[violet heron]

Wait

Yeah your gonna have to use colab

[fallow nexus]

my gpu wasnt being stressed at all when the training was happening

[violet heron]

Because it’s a AMD GPU

[fallow nexus]

it said amd/intel 😿

[violet heron]

RVC training depends on Cuda which is NVidia only

[fallow nexus]

i thougt it would work

[violet heron]

For covers sadly

[fallow nexus]

FUUUCCCCCKKKKKKk

ok so learn how to collab

[violet heron]

-colab

[azure marsh]

Suggestions for @fallow nexus

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[violet heron]

RVC Disconnected is for training has no ui

Ilaria RVC and Applio can also train

Simply run each cell and follow the directions

[fallow nexus]

ok thanks

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 Spain

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

🇫🇷 France

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

🇩🇪 Germany

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

🇮🇹 Italy

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

[gaunt whale]

-overtrain

[azure marsh]

How do I know if my model is overtaining?

All-In-One Guide on how to make a good model

This guide explains how the D and G files works and much more: https://rentry.org/RVC_making-models

Credits: LUSBERT

Automated Overtraining Detection (AOD)

Will be available soon in #1159513888199540817

Credits: grvyscale

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[thin stump]

@brittle wing Just include these two, that's it

No Ds store or mac OS folder

[crystal vector]

Hello, can someone explain to me where the problem is coming from? When I modify my LRDecay in my config file, the change doesn't seem to take effect correctly. During training, I want to go from "LRa" to "LRb" over a set of "y" epochs. I adjust my LRDecay accordingly, but when I monitor the training in my TensorBoard, I never reach the target "LRb"... Is there another parameter that could be affecting this LRDecay? 🤔

[tight kindle]

Anyone know how to optimize VAC (or other) settings for output through voice over network in a video game? RVC realtime output sounds fine when I monitor VAC, but in game over VON it sounds ass compared to normal voice via direct mic input. It's choppy, cuts out, doesn't pick up low volume etc. Any help appreciated! (sorry if cross-posting in Okada forum, but I use both.) - I'm using full version Muzychenko VAC, specs: rtx 4090, 13900k, 64GB RAM

[brittle wing]

is WASAPI better for rvc compared to ASIO

[pastel oak]

both sound nearly the same on rvc but asio is bettah

[turbid meteor]

what index rate does?

[pastel oak]

It forces the accent of the voice model (like if theyre australian then it will try to sound a bit more australian)

Setting it too high (over 0.7-ish) will cause autotune.. also if you say long-stretched syllables ("whaaaaaaaaaaat"), then it will bug out a bit aswell

[turbid meteor]

ohh i see

any tips for reducing cpu usage? or resources?
im using i5 10400 and rx5600xt

[fallow nexus]

good morning. i thought i had a counter timeout running while i fell asleep

my training finished but i was timed out

i can still see the g and d numbers

is there any recovery i can do? it was 500 epochs :/

im using disconnected collab

[pastel oak]

Im not sure your gpu can handle 0.20 sample length but i might be wrong. 50% cpu usage sounds alright imo for AMD users

You could try to increase sample length to something like 0.25, see if it reduces cpu usage. If it does, I can show you how to reduce delay naturally with wasapi to get back to the delay you were using before. Lmk

also put fade length to max, have you messed around with the extra inference time? And did you try to see how the usage is if you use rmvpe - fcpe?

[turbid meteor]

ive tried rmvpe but this error popped out and says file doesnt exist? i use crepe/fcpe atm. i dont see any usage difference between pitch alg. sample length is really affect my cpu usage if i set below 0.20 so i set it to 0.50 but yes, more delay

[pastel oak]

Lol this is funny because fcpe is the one that doesnt work for me xd

did you manually install rvc with the pip requirements and all or downloaded binary release package

[sacred hinge]

why doesnt it work when i try to use voice i downloaded from here

[turbid meteor]

i manually install rvc with downloaded binary release package. i remember last time you gave me link to download zip files for my amd system, but when i tried to start audio conversion, it crashed. i think its because i select the wrong audio device

so then i try downloading rvc again with this archived repo https://github.com/Blanc-dot/RVC-test and downgraded my python to 3.10.11 and here im now

here is when i use released rvc downloaded from hugging face

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 Spain

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

🇫🇷 France

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

🇩🇪 Germany

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

🇮🇹 Italy

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

[stiff pollen]

Hello everybody! I installed the Voice changed client demo program today. But for some reason my sound comes with a large delay and lags ms..
Maybe I did something wrong? :(

[fallow nexus]

Where does the final file go?

[rare goblet]

Ayo? @fallow nexus level 3 !!!

[fallow nexus]

Like it said training completed

And I have a log file in my google drive folder

I was also auto saving every 10 epochs so I’m not sure what you mean by “doesn’t have auto save”

I just reread the page. “Unlike the original RVC notebook, this notebook does not have auto save functionality”

Sad. Thanks for trying to help

[proud elbow]

click the folder icon in left pane, then expand to "Mangio-RVC\weights" then drag some pth files to your Drive (faster than downloading directly, but make sure you have enough Drive space)

[tame mural]

Hello
So I'm using the KafkaEN model from this server on the RVC gui app. I have a 3060 gpu, and followed this guide
https://rentry.co/RVCRealtimeGuide#virtual-audio-cable-mandatory
I did everything the guide said, but the changed voice sounds really choppy, robotic, and weird!
I wanted to include a pic or a voice record but I'm unable to do either.
What could be the issue?

I hope that this is the right channel for my query.

#🔍│find-models message
This is the model.

Could the model itself be the issue? or is it my headphones mic?

[rare goblet]

Ayo? @tame mural level 1 !!!

[tame mural]

Appreciate the response. to help mitigate the issue, would you kindly recommend a model you know for a fact sounds and performs good enough? preferably a free one.

I'm afraid my knowledge, nor my time allows me to do that. I'd rather use a pre-existing one at this point in time.

I truly appreciate your helpful advice. I'll look into it. Thanks a bunch.

And to whomever is reading this, I'd also appreciate a good free model suggestion, to help make reaching my goal easier.
Much appreciated.

[brittle wing]

try making the sample length 0.35

works for my rx 580 and intel i3 10105

removes the choppiness

[turbid meteor]

yep, thank you

[pastel oak]

Go to sound devices of your pc

Go to the "recording" tab
Look for your virtual cable. right click - properties
On the 2nd tab i think not sure what its called you can enable to play it back. cant control the volume tho

[fiery raptor]

is gradio still having problems?

[hollow raven]

Do anyone know what that means?

If i start Audio Conversion the Dev version says: "Invalid index. You MUST use added_XXXX.index but not trained_xxxx.index!"

In the normal version it works

[dusty cedar]

on this colab link https://colab.research.google.com/drive/1Gj6UTf2gicndUW_tVheVhTXIIYpFTYc7?usp=sharing#scrollTo=OVQoLQJXS7WX

i keep seeing jsondecodeerror

pls help

[fallow nexus]

alright yall i just made my first model and im pretty happy with it. i now want to change my mic in real time to use it so i can record, but theres an awful delay when i use the mainline

is there a better way to use my model to get faster conversion?

[azure marsh]

🎙️ Realtime Voice Changer (Local)

Guide for W-Okada's RealTimeVoiceChangerClient

Description: This one has more options and is somewhat more convenient to use than Real-time RVC.
Link: Rentry
Credits: ravencutie21

How to use real time conversion with RVC (voice changer)

Description: More accurate than W-Okada's, but requires a decent GPU.
Link: Google Docs
Credits: teas.hi

RVC TTS

Description: Text to speech with RVC
Link: Rentry

[fallow nexus]

nevermind

[hollow raven]

If i use RVC ingame its lagging it is normal?

[proud elbow]

yeah you must use added_...index file like this

[fleet sleet]

Hello, do you know how to contact the RVC's creators?

I have some questions

[proud elbow]

for some support and problems just ask here

[fleet sleet]

Well, I would like to remove accent from a model, the model is in English(voice) and when I speak my language it has American accent.

I don't know how to remove the accent

[fleet sleet]

Do you know how to do that?

[hollow raven]

I dont think that you can remove the accent

[modern gale]

i have a bunch of models i used to use that dont work in ilaria and i suspect it might be the type of file (.pth) is this a known issue?

[rare goblet]

Ayo? @modern gale level 2 !!!

[fleet sleet]

Really?

[hollow raven]

It is but then cames a error from Python-CFFI

[violet heron]

RVC uses pth

should work if it’s a RVC model

[fleet sleet]

Do you know who created the RVC?

[modern gale]

can u try to replicate it?? :https://huggingface.co/Shadow-AI/Juice_WRLD_Raspy_WOD_Era_RVC_V2_250_Epochs/resolve/main/JuiceWRLDRaspyWODEra.zip

takes bout 10 secs before breaking

ive even reuped it to google drive and it doesnt work

[violet heron]

Did you duplicate the space

[modern gale]

no

[violet heron]

do it, 3 dots, Duplicate This Space

[fallow nexus]

chat help me

[modern gale]

im using google colab

[fallow nexus]

im trying to use the thing recommended for amd computers for real time changing

[violet heron]

Oh

[modern gale]

lol?

can this be a reason?

[violet heron]

Try a diff colab

[fallow nexus]

it tells me that i need to get this

but i dont understand git

im manually clickingf all bat files to launch stuf blah blah

how do i clone a repo?

[modern gale]

on huggingface it still also says the model was downloaded but doesnt pop up after refresh

just ps

[hollow raven]

Which python version do i need for this? ">> pip install torch==1.11.0+cu113 torchvision==0.12.0+cu113 torchaudio==0.11.0 --extra-index-url https://download.pytorch.org/whl/cu113"

because he said my version does not satisfies the requirement

In a game

in billiard

[rare goblet]

Ayo? @hollow raven level 6 !!!

[hollow raven]

Pool Nation FX

[fleet sleet]

Do you know how to remove accent?

[proud elbow]

what the hell is "Python-CFFI"? some fork?

[fleet sleet]

Anyone in here knows who created RVC??

[violet heron]

Click refresh

[hollow raven]

I can`t send pictures so i cant show you the error

[violet heron]

Don’t use the .index file

[fleet sleet]

Unfortunately it still has accent

[hollow raven]

I cant download the requirement file because the Error that i cant install this pytorch version

[fleet sleet]

I want to dub a English video to Portuguese(translate and dub)

Anyone in here knows the creators of RVC?

[proud elbow]

which RVC you're actually using?

[hollow raven]

In normal RVC its working, but in the Dev version it gives me a error

[fleet sleet]

When I speak in Portuguese with the RVC(the voice trained) it has accent(American accent)

[proud elbow]

haven't seen the dev version, or prob outdated by the latest RVC1006Nvidia

[fleet sleet]

That's my goal to remove the accent to sound like a native speaker

[hollow raven]

I have the RVC1006Nvidia

[fleet sleet]

I'm Portuguese native speaker

[hollow raven]

I tried this https://rentry.co/Realtime-less-delay#wasapi-on-rvc-realtime

and for it i need the dev version

[modern gale]

is it possible to push more cpu / gpu power into applio on huggingface?? take like 400 secs every time

[proud elbow]

I should prob summon @pastel oak

[violet heron]

You can buy a GPU on HF Spaces or alternatively you can use the applio colab

[modern gale]

whats perfered ilaria or applio??

[fleet sleet]

Do you know how can I do that?

[violet heron]

People had issues with Ilaria lately

Try both and see what you like more

[fallow nexus]

can someone help me with my git questions 👉 👈

[proven hill]

lmao no, if people dont know what to do its not a fork issue

[modern gale]

idk why u throwin shade im just having problems with models loading

they say importing and the links seem to be p fine

but dont load

[fleet sleet]

Ok, I'll try that

Thanks for your help 👍

[pastel oak]

you should be able to send screenshots. can you send one of your gui again and the error code

rvc1006 build and the dev version i linked work together wirhout issues so lets see what else is missing

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 Spain

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

🇫🇷 France

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

🇩🇪 Germany

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

🇮🇹 Italy

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

[scarlet wedge]

It's saying that the page doesn't exist.

[proper shale]

Brazil page doesn't work anymore

[scarlet wedge]

yeah, but in the other pages it's the same

[night tree]

hello everyone once i trained my model on this colab https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw#scrollTo=FFfC9x239kC1
where can I take my index & pth file?

I can't find it

[proper shale]

they should be saved somewhere on ur google drive

[night tree]

I just checked the logs folder in RVC and there's a log called pulla.pth (pulla is the name of my model) and a added and a trained .index

which should I get the added or the trained?

[wispy burrow]

Added.

[proper shale]

added

[night tree]

ok and then in the weights folder I take only the named one without any number

right?

[proper shale]

that's your final model yeah

[night tree]

alright thanks a lot

[proper shale]

yw :)

if you have any more questions ask em here or #📑│making-models

[fallow nexus]

I have a question

[proper shale]

?

[fallow nexus]

im trying to use someones updated mainline ui

i want real time changing and the delay on mainline is atrocious

this is in the readme for the update

im not sure what the "initial rvc folder" is and when i try to launch the bat file in the folder it fials

[proper shale]

did you try okada already?

[fallow nexus]

idk what that is

i can abandon this im not attached to it

[ebon maple]

Can I learn how to adjust the sounds in rvc gui? The sounds are very robotic. ???

[proper shale]

https://rentry.co/VoiceChangerGuide

[fallow nexus]

yes

[proper shale]

easy to use GUI

but tbf, if it lags on mainline idk what will happen on it

[fallow nexus]

this is what im trying to do

[proper shale]

also same person who made the mainline fork wrote the guide

[ebon maple]

bro help

[proper shale]

mess around with the settings, test and see

[ebon maple]

I tried but it has no effect

[proper shale]

so it's either your audio or the model

[fallow nexus]

get a new model :3

[ebon maple]

Does sound affect the model?

[rare goblet]

Ayo? @ebon maple level 1 !!!

[fallow nexus]

there are two things that effect the outcome of a models work. the input and the model itself

[proper shale]

an audio can affect the quality of the final result

so y

[fallow nexus]

ideally you have super good audio going in

[proud elbow]

RVC-gui is outdated, try the original one instead
https://docs.aihub.wtf/rvc/local/mainline/

[proper shale]

replay/applio ftw in this case

[ebon maple]

Is this a more updated version?

[proud elbow]

it has pitch extraction rmvpe which is more recent and currently state-of-the-art

[ebon maple]

Can it replicate the sound quality exactly?

[pulsar tartan]

hello
how can i make my own custom voices using rvc models?

[proud elbow]

generally better

[pulsar tartan]

i want to use cypher valorant

[ebon maple]

I trust you I will try

How easy is the usage logic?

[proud elbow]

if not sure, leave the default settings except transpose (pitch shift) depend on the kind of input sample voice and the model being used

[fallow nexus]

do you have a decent graphics card?

[rare goblet]

Ayo? @fallow nexus level 4 !!!

[fallow nexus]

if you dont you will need to use a collab to train your own voice model

[pulsar tartan]

nah

[pulsar tartan]

i want the cypher voice
i dont wanna train my own

[fallow nexus]

ok you can search in #1175430844685484042 and see if someone has made one

[azure marsh]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[marble forum]

link to RvC i can use to train my own model? help please

[violet heron]

What’s your GPU

[marble forum]

RTX 3050 4GB

[violet heron]

-local

[azure marsh]

Suggestions for @marble forum

🖥️ 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

Need some help? 🤔

You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.

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

[violet heron]

Original RVC or applio

[marble forum]

please which is better out of the two?

and where can i geta tutorial on how to use them?

and what about RVC WEBUI? i just saw i have that on my system already

[violet heron]

That is original RVC

I would go with that

[marble forum]

wow. that's great.

[marble forum]

any video tutorial on how to use it. i'm trying to train my own voice to use in RVC real time and W-OkADA

[violet heron]

-docs

[azure marsh]

Suggestion for @marble forum

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[violet heron]

Should be here

[marble forum]

okay thanks will check it out

[fathom crater]

yo

im getting this error

ImportError: cannot import name 'CSVutil' from 'my_utils' (D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\my_utils.py)

while trying to start the easier gui

what do i do

???????

fixed it

[tame mural]

An update on this, I've tried two models with relatively high likes/reactions (Shylily) and (Tiny noise The owl house) and the issue still persists = both voices don't sound as good as their posted examples/demos.
Despite doing and picking everything the guide for RVC has recommended.

[tame mural]

This was the original issue I'm still facing, just to clear things up.

Did that issue happen to anyone before? searching doesn't find anything sadly.

[terse turret]

YOOO MAN

ITS BEEN SO LONG

SINCE YOU HELPED ME MAKING THAT MODEL

[tame mural]

It actually seems that I'm now able to attach pictures, when I couldn't before. Here are my settings

[rare goblet]

Ayo? @tame mural level 2 !!!

[tame mural]

Would someone point me in the right direction, as to what am I doing wrong?

[lime otter]

!colab

[fathom crater]

bro

im gettings ths error now

when im trying to

train

DEBUG:matplotlib:interactive is False
DEBUG:matplotlib:platform is win32
INFO:torch.nn.parallel.distributed:Reducer buckets have been rebuilt in this iteration.
max value is tensor(1.0702)
Process Process-1:
Traceback (most recent call last):
File "multiprocessing\process.py", line 315, in _bootstrap
File "multiprocessing\process.py", line 108, in run
File "D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 225, in run
train_and_evaluate(
File "D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 579, in train_and_evaluate
cursor.execute("SELECT stop FROM stop_train LIMIT 1")
sqlite3.OperationalError: no such table: stop_train

[lime otter]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[fathom crater]

?

[jolly parrot]

I'm trying to understand what kind of specs you need to run RVC locally. I read that for training models GPU and Vram are most important. But what if I just want to use existing models? What specs of my system are key? Could it run on an older PC?

[fathom crater]

?

me or who

can someone help me

[terse turret]

uhh why does this error happens?

[covert turret]

I've started to train using the Applio NoUI colab and so far seems good but I guess the tensorboard did not load correctly.
Is there a way to try to get the tensorboard up and running or is that normal, what should I do to get tensorboard essentially as I want to track the training process in order not to overtrain.

[rare goblet]

Ayo? @covert turret level 6 !!!

[terse turret]

hey can you help me on mine¡

[covert turret]

uhhh, my cookies are cleared (my browser deletes cookies when closed, and I just opened it, so, I guess that's not the issue?)

lol yeah ig i'll have to check on that lmao

[terse turret]

youtube link

oh wait

wdym

ooh youtube link

yeah youtube link

the boiled one

https://discord.com/channels/1159260121998827560/1235491881694003262

sooo..

so i have to use another model?

alr

thanks

please give this guy a helper role!

[vale dune]

trained a model with mangio but the index file seems to be missing 😭

am i supposed to press train feature index

[thin stump]

Sometimes I forget too so I have to run through all the steps again just to get the index (dependencies, training variable, processing feature, then train index)

[vale dune]

ahhh

pressing that seemed to do it

what's the difference between added and trained index? (im sure this question has been asked a billion times 😭)

i think added is the one i use right

[thin stump]

I had to step away for a moment, yes

[vale dune]

thank you bestie

[rare goblet]

Ayo? @vale dune level 1 !!!

[vale dune]

mods kill him

[hollow raven]

Can i do something to deaktivate the lag in games with RVC? Because they says: "Better for games"

Because with RVC the voice in discord is clean, but the game lags
and w-okada laggs in discord but the game is clean

[pastel oak]

what kind of lag

static or actual syllables repeating

[hollow raven]

it is not a static lag i reach my FPS but if i move its repeating always

[pastel oak]

i mean if its a very high end game then your gpu is running into 100% and cant properly work

so depends

[hollow raven]

My GPU is on 40% and CPU on 37%

Its Raft

But in Pool Nation FX which is billiard its lagging too

my FPS are still on 144 i dont know

and the ping is still good, my voice is clear

[pastel oak]

if you have index rate at 0.04 you might aswell put it at 0

try it

[hollow raven]

it is a different?

i dont know what is Index Rate

[hollow raven]

Makes no different ingame

[pastel oak]

try this

open task manager hit details

right click audiodg.exe and set Priority to HIGH

right click it again and press set affinity and only have an EVEN number CPU core set, 2 is what I use. AKA everything unchecked except CPU 2

and/or try to reduce extra a little bit ; increase sample length

[hollow raven]

makes no differnt

voice is perfectly clear but game is lagging

[dusty rampart]

use okada, realtime is not updated

[hollow raven]

Okada is weird, than the voice is lagging if i am in a game (the res:ms raise to 24.000ms)

[dusty rampart]

that is called bad GPU

[hollow raven]

This is called RTX 4080 Super

[dusty rampart]

and what resolution/settings are you playing?

you do know realtime eats almost half your GPU right?

[hollow raven]

1080p 60fps capped now

[hollow raven]

No, 18% 😅

[dusty rampart]

CUDA cores

rip those

best way to avoid lagging when gaming is either lowering game settings or using a separate PC so it doesnt affect

what game are you trying to play it with?

[dusty rampart]

i`ve seen people making it run with Squad, VALORANT, CS2 and others

[hollow raven]

Raft

[dusty rampart]

tf

[hollow raven]

yes 😅

[dusty rampart]

well, not sure if it might be due to the game, but something like that should run fine

if okada doesnt run, realtime won't either since they are basically the same, okada is just more updated

[pastel oak]

tbh it seems like somethings up wtih your hardware

[hot citrus]

anybody got examples of training graphs? ive been seeing graphs having a big dip untill it flattens out but mine seems to not have such a drastic dip and flattens out pretty quick

[rare goblet]

Ayo? @hot citrus level 1 !!!

[proper shale]

that could be the result of selecting the wrong batch_size for ur model

also, not setting it to scale too

[hot citrus]

how do i set it to scale or how do i know how to set it to scale

[proper shale]

if u press the second button below the graphs it should put things in scale as they should be

[hot citrus]

how does having the wrong batchsize effect the training

[proper shale]

it can lead to the graph flatlining too soon

[hot citrus]

this is basically how my graph looks rn and had my batch size to 4

[proper shale]

that seems fairly normal

it's not that biiiig dip but it's ok

have you already set it to scale?

[hot citrus]

this is if its set to scale

[proper shale]

yeah feels pretty standard to me

[hot citrus]

most of the examples i guess that ive seen have been looking like this:

[proper shale]

good ol "point of overtraining" graph on kalo's easygui

(which actually wasn't right. sad)

[hot citrus]

haha but yeah seeing the big dip got me thinking if mines is normal or not

[hot citrus]

ayyo not right like how?

[proper shale]

That's a good concern tbf, but like... if you have a good model... it doesn't matter

[proper shale]

Because g/total is the average model.

[hot citrus]

ill say my models are pretty decent but i thought i could maybe improve it a bit more with different/ cleaner datasets

[rare goblet]

Ayo? @hot citrus level 2 !!!

[proper shale]

And most of the time, g/mel, kl and fm are still going down (fm has issues with me though, it rises for some reason lol), your model is still good and can keep training for longer

[hot citrus]

ayy thanks for clearing things out about my model's training

[proper shale]

You're welcome

If you have more questions feel free to ask!

[hot citrus]

will do!

[grand mist]

Having problems with the app booting up... This is what it states.

File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named '_distutils_hack'

Remainder of file ignored
Error processing line 1 of C:\Users\Downloads\RVC-GUI-pkg-20220525-mp3fix\RVC-GUI-pkg\runtime\lib\site-packages\google_auth-2.16.2-py3.9-nspkg.pth:

Traceback (most recent call last):
File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
File "<frozen importlib._bootstrap>", line 562, in module_from_spec
AttributeError: 'NoneType' object has no attribute 'loader'

Remainder of file ignored
Error processing line 1 of C:\Users\USER\Downloads\RVC-GUI-pkg-20220525-mp3fix\RVC-GUI-pkg\runtime\lib\site-packages\matplotlib-3.6.2-py3.9-nspkg.pth:

Traceback (most recent call last):
File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
File "<frozen importlib._bootstrap>", line 562, in module_from_spec
AttributeError: 'NoneType' object has no attribute 'loader'

Remainder of file ignored
Error processing line 7 of C:\Users\USER\Downloads\RVC-GUI-pkg-20220525-mp3fix\RVC-GUI-pkg\runtime\lib\site-packages\pywin32.pth:

Traceback (most recent call last):
File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'pywin32_bootstrap'

Remainder of file ignored
Traceback (most recent call last):
File "C:\Users\USER\Downloads\RVC-GUI-pkg-20220525-mp3fix\RVC-GUI-pkg\rvcgui.py", line 3, in <module>
from tkinter import filedialog
ModuleNotFoundError: No module named 'tkinter'
Press any key to continue . . .

Any help??

[proper shale]

that's... extremely outdated

like 2 years old atp

[grand mist]

oooooo

OOf

Thanks for letting me know XDD

[proper shale]

id recommend mainline for just the basic rvc experience, also applio or replay if you want ease of use

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

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

guide for applio too

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 Spain

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

🇫🇷 France

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

🇩🇪 Germany

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

🇮🇹 Italy

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

[wraith swan]

my voice is not picking up ion know what to do

it works on passthru

[violet heron]

What’s your GPU

[wraith swan]

rtx 3050

it used to work

but now i don’t know what happ

happened

[proper shale]

check command prompt for any tracebacks

[wraith swan]

it works on obs but ion hear it

[rare goblet]

Ayo? @wraith swan level 1 !!!

[wraith swan]

lemme test on a call

[wraith swan]

works fine i just don’t hear it

[ornate gulch]

hi

[proper shale]

oh just set monitor to your headphones if you havent yet

[ornate gulch]

stupid question , but how can i run uvr on google colab

[proper shale]

are ya using this?

https://colab.research.google.com/github/Eddycrack864/UVR5-NO-UI/blob/main/UVR5_NO_UI.ipynb

[ornate gulch]

thanks so much

[proper shale]

yw :)

[ornate gulch]

hey

i get this issue when i try to run mdx23c : No such file or directory: 'audio-separator'

[warm hound]

i need help

[proper shale]

hm... might not have that directory wherever you inserted it

tbh i havent really used that so im not entirely sure

[ornate gulch]

is there are any uvr ui in colab

i get the same error with every other model

[proper shale]

don't think there is

See if you're doing everything right with this guide:
https://github.com/Eddycrack864/UVR5-NO-UI#how-to-use

If you are, and nothing has helped, I suggest opening an issue in GitHub:
https://github.com/Eddycrack864/UVR5-NO-UI/issues

[ornate gulch]

now i know what was wrong

no

[proper shale]

oh, that explains it

[night moss]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[mortal lodge]

once ive made a model whats the collab link to run it with? ive been trying to find it on the site or here in the server but haven't had any luck

[proud elbow]

144 fps avg but stuttering? how is 1% low on some fps counter? if neither gpu nor cpu usage nor VRAM is maxed, prob the game itself

[humble quail]

I ask if this software can practice the model by itself, that is, pass the model by itself.

[finite galleon]

Ig what you can do is train a model with your dataset, use that model to create another dataset and with that dataset you can train another model

[stark tendon]

I need help, i've downloaded the GPU version and doesn't let me choose the GPU, only the CPU and it consumes 100% of my cpu when i have it activated

How do i make it take % from the gpu instead of CPU, i'm seein in alot of videos that people uses the GPU but it doesnt let me set it to gpu, only gpu

[median monolith]

what should i use to download a youtube video audio without loss in audio quality

[terse turret]

i have an nvidia drive

[viscid moss]

#1208578250943504465 message
Try this

Or use another acc

Maybe you ran out of GPU usage time

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 Spain

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

🇫🇷 France

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

🇩🇪 Germany

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

🇯🇵 Japan

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

🇰🇷 Korea

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

🇵🇱 Poland

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

🇷🇺 Russia

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

🇮🇹 Italy

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

[terse turret]

it still have left

[viscid moss]

That's the Disk usage

[terse turret]

ooh

alr ima create another account

[viscid moss]

Did you try this?

[viscid moss]

Voice changer? If so please go to #🔍│help-w-okada

[terse turret]

i had it at tgu

[viscid moss]

U need to choose GPU (T4)

[terse turret]

yeah thats the one i had

[terse turret]

yeah thats the one im talking about

i had it at that

and the error started accuring

i also deleted all my data and restarted my runtime

i already created another account

[viscid moss]

Try:

runtime > disconnect and delete runtime > reloading the page (F5)

[terse turret]

after reloading the page i press connect?

[viscid moss]

Yep.

[terse turret]

alr

[viscid moss]

Like the video.

[turbid meteor]

rvc for linux>

?

https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/docs/en/README.en.md#linuxmacos-one-click-dependency-installation--startup-script

kinda found it

[terse turret]

does it works?

[turbid meteor]

im still reading

[terse turret]

send a test

[turbid meteor]

how do you fix this error?

[rare goblet]

Ayo? @turbid meteor level 9 !!!

[fathom crater]

Can someone help me

Im trying to train a model localy

[brittle wing]

everytime i use it and run start

i cant even hear myself at all

[fathom crater]

????

[brittle wing]

rvc

i press start

i cant hear myself

[fathom crater]

DEBUG:matplotlib:interactive is False
DEBUG:matplotlib:platform is win32
INFO:torch.nn.parallel.distributed:Reducer buckets have been rebuilt in this iteration.
max value is tensor(1.0702)
Process Process-1:
Traceback (most recent call last):
File "multiprocessing\process.py", line 315, in _bootstrap
File "multiprocessing\process.py", line 108, in run
File "D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 225, in run
train_and_evaluate(
File "D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 579, in train_and_evaluate
cursor.execute("SELECT stop FROM stop_train LIMIT 1")
sqlite3.OperationalError: no such table: stop_train

[brittle wing]

does anyone know why the mf voice changer wont work

[rare goblet]

Ayo? @brittle wing level 1 !!!

[brittle wing]

everytime i press start

i cant hear nothing

[turbid meteor]

check your output device

[turbid meteor]

have you tried messing around with the chunk?

[brittle wing]

i press start and nothing happens at all

[turbid meteor]

change your chunk size to 192

[turbid meteor]

how about using one of the default voice models?

[brittle wing]

the thing is

i click it

it takes forever just for it to actually use the voice model

[turbid meteor]

you cant hear yourself even with default voice model?

[brittle wing]

nope

[turbid meteor]

shit idk then

[brittle wing]

even with monitor 💀

[rare goblet]

Ayo? @brittle wing level 2 !!!

[fathom crater]

What about me bro

[rare goblet]

Ayo? @fathom crater level 4 !!!

[fathom crater]

I need help too

Wtf

[turbid meteor]

i cant read lmao
i dont know what exactly your problem

[fathom crater]

so

DEBUG:matplotlib:interactive is False
DEBUG:matplotlib:platform is win32
INFO:torch.nn.parallel.distributed:Reducer buckets have been rebuilt in this iteration.
max value is tensor(1.0702)
Process Process-1:
Traceback (most recent call last):
File "multiprocessing\process.py", line 315, in _bootstrap
File "multiprocessing\process.py", line 108, in run
File "D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 225, in run
train_and_evaluate(
File "D:\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 579, in train_and_evaluate
cursor.execute("SELECT stop FROM stop_train LIMIT 1")
sqlite3.OperationalError: no such table: stop_train

this

i am getting this error

while trying to train the voice

[fathom crater]

?

bruhhhhhhhhhhh

[turbid meteor]

yea i cant read that

[fathom crater]

well what am i suppost to do now

xd

[turbid meteor]

i never use mangio and train a voice model

[fathom crater]

what do u use then

[turbid meteor]

just for a conversion from my voice

[fathom crater]

well

what do u use

[turbid meteor]

rvc

[fathom crater]

can u do it localy

if u can, can you link it to me?

[fathom crater]

?

[turbid meteor]

im doing what this repo said https://github.com/Blanc-dot/RVC-test

[fathom crater]

well what do i choose

[turbid meteor]

official repo

choose directml if you use amd. other use nvidia

[fathom crater]

ok

wait

do i have to pay

[turbid meteor]

no

just fill with whatever

[fathom crater]

ok

[fathom crater]

do i need all MSVC VS 2022

[turbid meteor]

not all

just make sure you select the windows 11 sdk and the other one

[turbid meteor]

that as well

[fathom crater]

got it

[brittle wing]

What version of RVC u got

[fathom crater]

thats visual studio installer

bro

[turbid meteor]

it is yeah

somehow it need some components from visual studio

[fathom crater]

wait whats the folder called that the bat cloned

[pastel oak]

@fathom crater i didnt follow chat, what are you trying to do

[fathom crater]

i acutally tried to fix a issue with mangio

but exkoi said i should try rvc

to train it

[pastel oak]

training voice models?

[fathom crater]

yea

[pastel oak]

you can just download rvc mainline prepackage and youre done

[fathom crater]

fr

[pastel oak]

no need to install additional stuff

[fathom crater]

can u send me the link

[turbid meteor]

shad come to the rescue

[pastel oak]

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

theres download links, explanation on what everything does aswell

assuming you use nvidia ofc cause amd cant train afaik

[fathom crater]

yea

i use nvidia

1060 XD

[pastel oak]

very ambitious to train with a 1060

should be possible i think but no clue to what extent

else theres colab alternatives

[fathom crater]

i mean it is possible

it has cuda cores

and 6gb of vram

so it should be totally enough for the minimum

[wispy lodge]

You need to delete pretrain folder and restart voice changer to redownload the files

Because pretrain/content_vec_500.onnx failed to load

[fathom crater]

what the fuck

bro my gaming laptop is currently lagging tf out

with 20gb ddr4]

[pastel oak]

issa lot dw

[fathom crater]

k

ik

xd

[pastel oak]

but 9.6gb

[fathom crater]

its stuck on the same thing

[pastel oak]

make sure u didnt accidently delete old models you trained or downloaded before

[rare goblet]

Ayo? @fathom crater level 5 !!!

[fathom crater]

i couldnt even train models

[pastel oak]

wait it out

[fathom crater]

it would always error out

it is done

bruuh

my laptop is being bottlenecked by my sata ssd

holy

these sandisk ssds are not good

my old samsung 860 evo was better

tbh

but it was only 500 gigs

and i got this ssd for 50 bucks

[fathom crater]

it is installed

what do i choose

wtf\

found it

[pastel oak]

remove the space between "AICOVER" just do it one word

[fathom crater]

is that good

[pastel oak]

then try again

if it still doesnt work, open the folder "runtime" in your rvc1006nvidia folder and check if python.exe is in there or not

[pastel oak]

ye

[fathom crater]

ok

so it started training

shouldnt it show like the epochs

oh there we go

it started

[rare goblet]

Ayo? @brittle wing level 3 !!!

[tame mural]

I truly appreciate the response.

[fathom crater]

Wow it took two minutes to make one epoch

[pastel oak]

open the folder "runtime" in your rvc1006nvidia folder and check if python.exe is in there or not

[fathom crater]

How long do you guess its gonna take to get like 400 epochs

[tame mural]

My issue (weird choppy changed voices even after following guides) is still present in RVC itself, as seen in the screenshot.
Unless I'm mixing things up?

So far I've tried 3 voices, to the same end result.

[pastel oak]

you shuold be using Tensorboard to track your graphs to determine when the voice model is ready instead of assuming based on epoch numbers

[fathom crater]

Ok

[tame mural]

I'll try to upload a voice sample momentarily.

[fathom crater]

How can view it

[pastel oak]

u can still install tensorboard while the training is ongoing

on the guide i linked you there was a tensorboard section somewhere on the top of training i think

[radiant pivot]

https://cdn.discordapp.com/attachments/837901953999831124/1236237628324708484/image.png?ex=66374782&is=6635f602&hm=18a3c0ec1fd434887efa4517cd3e3203579c1ab3fde0dd258c22c79861c8d2b3&

[pastel oak]

ye send it

[radiant pivot]

im having an issue trying to runV2

[pastel oak]

we generally use rvc mainline instead of mangio here. theres probabl someone that knows more about this but would just use mainline for now if i was you

[radiant pivot]

mainline?

[rare goblet]

Ayo? @radiant pivot level 1 !!!

[radiant pivot]

like this?

[pastel oak]

ye mainline looks like that aswell. but its working now for you or not

[radiant pivot]

The Version 2 isnt

i just noticed there was a version 2

and it dosnt launch

[pastel oak]

dont know anything about a version2

[pastel oak]

is wykaz zabezpieczen a "executable file" ?

[pastel oak]

well that still doesnt change the fact im hearing about that for the first time lol

read up on github what it is ig

i dont think theres an updated rvc, id know 😭

[radiant pivot]

this looks better version than mine

and i was confused

[pastel oak]

thats mainline

bro just translate what it means to me xd

[pastel oak]

https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/RVC1006Nvidia.7z

[radiant pivot]

but mine dosnt have the cpu thing on it and some of it isnt translated

[pastel oak]

the cpu thing is harvest

its not needed anyway since people use rmvpe/fcpe

[radiant pivot]

ahh

[pastel oak]

ok i think your pc deleted python when you extracted the file out of the zip

check if the python.exe file is still in the ZIP folder you downloaded
if its there, turn off antivirus and stuff and extract again

if you dont see it in there, then redownload the zip file and try again

[tame mural]

Thanks for the response. Here's the recorded voice, and the link to the voice model itself. Mine isn't even clearly audible, let alone on the same tune of the posted voice.
#1212177800455266446 message

[fathom crater]

Shad

What do i do now

I installed it

[pastel oak]

it sounds alright

put respnse threshold to -60, on rvc its pretty shit
you can try raising extra to max to see if it makes ht clearer

anything else will depend on your mic and the voice model itself i think

[fathom crater]

So i installed tensorboard

What should I monitor

And how do i know if the voice is finished

[tame mural]

Will do and get back to you, thanks a bunch.

[pastel oak]

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

if youre new then just follow g/total graph, thats the best average model

so once that hits the lowest point you use the version of before the lowest point in steps

to know if its the lowest point you have to keep training the model until youre sure

[fathom crater]

Ok

[tame mural]

2nd test ... that means it's a mic issue, correct?
I'm using the mic attached to a headphone, specifically LENOVO H402 USB RGB Gaming Headset - 7.1 Surround.

I don't know if that makes a difference, but these are the specs of the headphone.

[tame mural]

Just to confirm, the Tl;dr of the issue is that my voices don't sound as good as their posted samples.