✨│ai-help

[delicate wharf]

ok thank you

[covert turret]

Thank you, it's my first try at making a model and I kinda like how it turned out, I might try to improve it later as I figure this model-making thing down the road

[proper shale]

Yay, glad I could help :)

It sounds pretty nice for a first model. I hope you can improve and become a great Model Maker :)

[unkempt shoal]

I finally got it working

Looks like my Focusrite Solo was the problem

[proper shale]

Oh nice

[proper shale]

_ _

[unkempt shoal]

I'm scared to plug it in again

[proper shale]

Wow, didn't expect that to be the problem

Huhhh that was weird

[unkempt shoal]

Yh

Ima try to fix it

Lemme plug in it

Oh nice it works

[rare goblet]

Ayo? @unkempt shoal level 2 !!!

[proper shale]

Wow

...that was weird

But I'll recommend people to check their interfaces (is that how you even call those? I'm just going by their portuguese name, sorry) and stuff

If they face that BSOD

[unkempt shoal]

Yep interfaces

Alright, glad I could provide some info

[proper shale]

Thank you 🙏

[spare shale]

@proper shale https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI Will this work for me? 😅

[proper shale]

If you have a good GPU, it'll work for training and inference.

If you don't, it can do inferencing.

[spare shale]

I'm not sure but I can just show you what and how I do

What kind of laptop do I have and Other

[proper shale]

Do you have a good GPU

[pseudo fox]

Hello, can you help me? I am downloading the file but the application does not appear on my screen. what should I do.

[proper shale]

What application specifically?

And do you have any more details?

[pseudo fox]

RVC GUİ

[rare goblet]

Ayo? @spare shale level 23 !!!

[spare shale]

If I say I don't have a GPU

[pseudo fox]

Can I take a photo for the step?

[proper shale]

You shouldn't be using that nowadays, try other local stuff

-local

[azure marsh]

All Working Local Forks 🖥️

• Applio-RVC-Fork, 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

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

[proper shale]

Or maybe Colab

-colab

[azure marsh]

☁️ Google Colabs

• 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab

🤗 Hugginface 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

[proper shale]

Colab is online tho

[proper shale]

Then you shouldn't use the local forks

[spare shale]

T-T

As you say

[pseudo fox]

Which app should I download?

[rare goblet]

Ayo? @pseudo fox level 1 !!!

[pseudo fox]

I don't understand these things at all

:/

[proper shale]

You should probably stick with online stuff if you don't understand a lot of the technology then

[shy marsh]

Does anyone have a marin RVC MODEL>?

[proper shale]

Use Ilaria RVC, the one in here, @pseudo fox

Here's the guide for it too: https://www.rentry.org/ilarvc_inf_guide

[pseudo fox]

I'm not good with artificial intelligence, that's all.

[proper shale]

Check weights.gg and #1175430844685484042

[spare shale]

GUI:go-realtime-gui Won't Work Too?

[proper shale]

If you don't have a (dedicated) GPU yeah no

[shy marsh]

ty

how about Rezero Rem RVC?

model

I couldnt find it

[spare shale]

@proper shale

[proper shale]

:sob;

😭

Integrated GPU moment

[proper shale]

If ya couldn't find it, probably doesn't exist

[spare shale]

So... I still have a Bad laptop, right? 😅

[proper shale]

...yeah

😭

i have one too dw

[spare shale]

https://media.discordapp.net/attachments/1167948353833816154/1194044120142983168/gif_make_mikus_11.gif?ex=65aeeb3e&is=659c763e&hm=3e54d24e909dd4a017ec5430f7b981d7c5bda045311e0b703288660c1e3621b5&

[humble monolith]

I'm trying the RVC onnxgpu-cuda client and my audio is easily coming through, but none of the models have any effect, regardless of tune or index. How do i change that?

[stable flint]

sorry guys its been a while, i forget whats the best option to separate vocals using Ultimate Vocal Remover?

[proper shale]

Are you on passthru

[humble monolith]

ah looks like it; my tab was too small i couldnt see it

thank you

[proper shale]

you're welcome :)

[proper shale]

-audios

woops

-audio

[azure marsh]

📚 Audio Guides

• Perfecting Audio Isolation on Low-End Rigs: A Practical Guide, by Litsa The Dancer and Faze Masta Google Docs
• Gathering and Isolating Audio, by SCRFilms Google Docs
• How to make a good model All-In-One guide, by LUSBERT Rentry
• Creating Datasets for RVC using iZotope RX, by LUSBERT Rentry
• Vocal Mixing Tutorial, by Roomie YouTube

🛠️ Tools

Youtube Downloaders

• https://github.com/yt-dlp/yt-dlp
• https://github.com/dsymbol/yt-dlp-gui

Audio Separation/Isolation

• https://mvsep.com/
• https://multimedia.easeus.com/vocal-remover/

More commands: -colab, -uvr, -karafan, -overtrain, -help

[proper shale]

I guess MDX23C stuff should be good for all vocals, for backings use BVE

(And for leads too)

[desert niche]

what folder do I put my models in using local applio?

nvm

[half cove]

[desert niche]

thanks anyways

[desert niche]

sowwy

[half cove]

[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

[tulip pasture]

Can I use this for vrc? and if so how?

[proper shale]

You can, just change the input to the CABLE Output

It'll be laggier than normal tho

[tulip pasture]

so?

[proper shale]

...so what?

[tulip pasture]

so like

when I download a voice

what do I do with the file

Coz it'll just be like this

[rare goblet]

Ayo? @tulip pasture level 2 !!!

[proper shale]

You do the uploading process in W-okada, read the guide: https://rentry.co/VoiceChangerGuide

[tulip pasture]

tysm

[proper shale]

You're welcome :)

[brittle wing]

Does anyone know how to lower the ms?

quality is more important，if your dataset quality is high，the longer the better

[proper shale]

Yeah, should be

[errant dagger]

whats going on? 😭

[stiff basin]

the same happened to me, no idea tbh haha

[proper shale]

what colab is that

[hushed flame]

my ai voice is cutting out when i talk can i fix this

[errant dagger]

https://colab.research.google.com/github/iahispano/applio/blob/master/assets/Applio.ipynb?authuser=2

[proper shale]

...yeah, use RVC Disconnected

[proper shale]

check https://rentry.co/W-Okada-FAQ

[errant dagger]

I'm trying to use it for interference tho

[proper shale]

you might be doing something wrong on the GUI then

send a screenshot

[errant dagger]

the thing is

it has nothing to do with the gui

its when I run the collab cell to start the gui up is when the issue starts

[proper shale]

honestly im not even sure

for now, just... dont even use applio

[stiff basin]

there are only 2 buttons, i have the same problem

[proper shale]

_ _

-colab

[azure marsh]

☁️ Google Colabs

• 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab

🤗 Hugginface 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

[hushed flame]

didnt help

[proper shale]

Alright, let's get specific here: What's your GPU? And does CMD have any Tracebacks or anything?

[hushed flame]

gtx 1660 and idk what a traceback is

[rare goblet]

Ayo? @hushed flame level 1 !!!

[modest trout]

how i add my dataset in applio v3?

[proper shale]

Can you at least send screenshots of CMD?

[hushed flame]

??

[proper shale]

Probably in the "Extra" tab, but haven't used Applio V3 yet.

You can try going on https://docs.applio.org/ and seeing if they got anythint there

[hushed flame]

this?

[proper shale]

Yeah this.

Hmmm nothing wrong as far as I can see

Maybe you got your chunks too low?

[hushed flame]

its on 192

i put it on the max and the same thing happens

[proper shale]

What F0 are you using

[hushed flame]

crepe_tiny

my thing looks like this

[proper shale]

...yeah you got the wrong version for your GPU.

[hushed flame]

oh

[proper shale]

https://huggingface.co/wok000/vcclient000/resolve/main/MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.15.zip

Here's the right one

[hushed flame]

thanks

[proper shale]

Always remember onnxgpu-cuda is the NVIDIA version

onnxdirectML-cuda is the AMD/other video cards version

[hushed flame]

okay

[proper shale]

You're welcome, if ya have any more issues report in #🔍│help-w-okada

[hushed flame]

can you link my friend to the amd version

[errant dagger]

I come back to say attempt signing out and back in

[silver mist]

what do i plug my .pth into? i was using a colab but it didnt have the option to change voice models srry im rlly new

[proper shale]

they should check the https://rentry.co/VoiceChangerGuide too, links for all of em here

[proper shale]

nope the applio colab is just broken

everybody's having the issue

[proper shale]

Try something like Ilaria RVC

-colab

[azure marsh]

☁️ Google Colabs

• 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab

🤗 Hugginface 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

[proper shale]

The guide: https://www.rentry.org/ilarvc_inf_guide

[errant dagger]

funny
because I did that and I got in

[proper shale]

😭

shits confusing today

first somebody having kernel errors with W-okada and now this

[errant dagger]

very weird I must admit

[dusty python]

guys I have quick question, I use Ilaria firstly I convert the voice from one of the text-to-speech applications and then upload it to Ilaria to convert the voice into the voice of the person I want, but this is a bit laborious. Does anyone know a shorter way?

[proper shale]

there's a TTS part inside the GUI, but your mileage may vary with it

[hushed flame]

the version you sent me does this whenever i open it

[dusty python]

I'm a beginner I dont know what is TTS or GUI, im sorry 😦 I don't understand what are you mention

[proper shale]

TTS - text to speech, GUI, graphice user interface

When I mean GUI i mean the Ilaria RVC GUI

[proper shale]

... I don't know what occurred there

[half cove]

the gui is like the page you use to do it like this

[proper shale]

Anything on CMD this time?

[proper shale]

oh i see the two problems

first, move the MMVCServerSIO folder out of OneDrive

and secondly never put spaces on the folder above it, causes stuff to break.

[hushed flame]

oh

[rare goblet]

Ayo? @hushed flame level 2 !!!

[hot lagoon]

@odd shale How did you rub Mainline RVC on paperspace?

[abstract flower]

Hello,
I would like to have your opinion on whether I should change this to get a better result, please? Thanks for those who respond

[odd shale]

I had to git clone an repository.

https://github.com/Mojobones/RVC-Paperspace

This one

[hot lagoon]

thank you for this

[silver mist]

does Ilaria usually take a long time?

[half cove]

it may take a few minutes

the time taken depends on audio

[silver mist]

<3

[rare goblet]

Ayo? @silver mist level 1 !!!

[silver mist]

the last one i tried took almost 10 minutes before i gave up on it

[half cove]

if it takes too long you can try the aicovergen or applio colab

if you use ila rvc make sure you duplicate space

[silver mist]

i did yeh

thank ya it worked !

[half cove]

[viscid moss]

The default configurations are the best, but if you want to customize it here you can find the information on all the parameters:
https://github.com/SociallyIneptWeeb/AICoverGen#running-the-pipeline

[gloomy dew]

I just posted a model on #1175430844685484042 but it hasn't automatically posted on weights .gg

what am I doing wrong?

my discord account is already linked to weights

[half cove]

it takes some time

how long ago did you post?

[gloomy dew]

like 20 minutes ago

[half cove]

id wait a bit longer

[gloomy dew]

but I saw other ones, like all of them that appeared an weights after a minute

of being posted here : s

that's what I'm thinking I'm doing something wrong xd

[half cove]

i dont really know at what timings weights checks the models

but

it does take time

[gloomy dew]

Ok, it may could be bc it's my first model

and being reviewed(?)

okok

ty

[glacial veldt]

i got the voice changer and vb cable and voice meter and i copied the right settings but its still not working idk why i cant even hear my own voice

[proper shale]

check cmd

[glacial veldt]

alr i opened it

[rare goblet]

Ayo? @glacial veldt level 1 !!!

[proper shale]

see if there's any Tracebacks, anything useful

[glacial veldt]

WAIT

I THINK IK WHY

is it because i got clown fish?

[proper shale]

_ _

... maybe

also now you can send screenshots

so uh when stuff appears in cmd once ya click start send the screencap

[glacial veldt]

just this

[half cove]

did you just open a random cmd?

[glacial veldt]

or this?

[half cove]

yea thats it

[glacial veldt]

alr ima just wait for them to respond

[sand iris]

[glacial veldt]

also these the settings

[sand iris]

It has started to do that for some reason

hang on a sec

#🔍│help-w-okada message

^try this

[glacial veldt]

alr thanks

[sand iris]

note : the default models will still not work

you will have to import voice models from elsewhere

[glacial veldt]

ohhhh thats probaly why alr

[last dagger]

what's the best conversion method?

I've always used harvest but apparently that's bad now

[glacial veldt]

alr i rested my pc

[proper shale]

pitch extractor... it's RVMPE

harvest is ""good"" but it's aged

[hidden barn]

anyone have the okada with like the youtubers and streamers pre installed??

[glacial veldt]

alr so i got a kai cenat one and i still cant even hear my voice and i dont think its working

[rare goblet]

Ayo? @glacial veldt level 2 !!!

[sand iris]

did you follow the message I linked ?

this : #🔍│help-w-okada message

[desert gale]

when i try to download the software, this shows up in a loop, (wait web server) im on mac

[proper shale]

Try deleting stored_setting.json inside the W-Okada folder and restart

[glacial veldt]

yes im trying

[toxic temple]

I don’t know😭

[desert gale]

it worked, thx

[proper shale]

Yayy, glad I could help.

:)

[drowsy anvil]

Hello, the bot to search for models is failing me, could someone give me the link to the Kageyama Tobio model?

[rare goblet]

Ayo? @drowsy anvil level 1 !!!

[proper shale]

check https://www.weights.gg/

[glacial veldt]

so should I use client or server?

I tryed both and i dont hear anythign idk what im doing wrong this sucks bruh

[drowsy anvil]

I found the model but not the link, and I need the link:(

[ember coral]

So i have voice recordings for about 15 minutes what epock should I use?

[molten pecan]

upload it yourself and copy the link

-upload

[azure marsh]

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

[molten pecan]

follow the first two steps only

[molten pecan]

the appropiate amount of epochs depends on the length, quality and size of the dataset, so there isn't a way to know beforehand

you should instead aim towards using the tensorboard

you can know exactly when to stop training with it

[ember coral]

Okay thank you ☺️

[hot elbow]

I have a pretty shitty pc build, anyone know how I can make the voice changer have priority over my game @molten pecan

[brittle ibex]

Hello, I'm trying to find wav audio files to train a voice model, where should I start?

[ember coral]

One more question how much is enough to make it sound good like voice recordings if the quality is like pretty good?

[molten pecan]

how much what?

[molten pecan]

I don't use voice changer so idk

ask in #🔍│help-w-okada

[ember coral]

Like voice recording the length

[molten pecan]

well first, WAV is not mandatory, you can use FLAC if you wish.
Both are lossless formats, except FLAC uses an algorithm that compresses the data without losing quality, making it have a smaller file size than WAV

both work the same, but flac is more space-efficient

[hot elbow]

I did but nobody cares 😭

[glacial veldt]

does this matter? cause i changed it to crepe

[molten pecan]

10 mins should be enough. go for 20 mins if you want it really high quality

[ember coral]

Okay nice thank you so much 😊

[molten pecan]

crepe is more prone to artifacting (glitchy voice), go at your own will.
in my experience it's the best sounding one but RMVPE is more convenient

so yeah it has its pros and cons

[half cove]

rmvpe faster

[molten pecan]

yw!

[glacial veldt]

alr thanks i was changing it cause i was just trying to get my audio to work i still cant even hear my own voice

idk what i did wrong

no sound is coming thru

[molten pecan]

try making a post in #1192011222023950368

[molten pecan]

yeah the f0 are the algorithms for processing the input audio

are you using a virtual cable?

[half cove]

screenshot ur entire thing

[glacial veldt]

ye

[half cove]

it should be workin

[glacial veldt]

ye idk why i dont hear anything

bruh this sucks

[glacial veldt]

wait can u send me the exact one ur using

i probaly downloaded a wrong one

[half cove]

that should be the right one

try ask in #🔍│help-w-okada or #1192011222023950368 i dont use voice changer much

[glacial veldt]

alr

[brittle ibex]

thanks

[sullen plover]

Hi i need help i can't hear myself

[rare goblet]

Ayo? @sullen plover level 1 !!!

[sullen plover]

i mean it goes super late

i probably need to up the chunk but idk if my gtx 1660 super is good enough for it

also idk why there is not my gc

[half cove]

thats the amd/intel one lol

[sullen plover]

ah

so i donwoladed the wrong one

which one is the nvidia?

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 17b cuda

Download for AMD GPUs
Version 17b directml

Download for Intel GPUs
Version 17b directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[sullen plover]

alright thank you

[bright umbra]

How do I get my voice changer to stop sounding like a robot on discord? I have a 3070 ti so I think this shouldn't be happening

[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

[bright umbra]

?

[hot lagoon]

OSError: [Errno 122] Disk quota exceeded

[silver anvil]

hello this is a question

[bright umbra]

Can anyone help me make my voice changer less choppy on discord? sounds fine when it's going thru my headphones

[rare goblet]

Ayo? @bright umbra level 1 !!!

[hot lagoon]

Sometimes when I go to train a modle at some point it stops and say this :
During handling of the above exception, another exception occurred:

During handling of the above exception, another exception occurred
Traceback (most recent call last):
File "/usr/local/lib/python3.9/dist-packages/gradio/routes.py", line 437, in run_predict
output = await app.get_blocks().process_api(
File "/usr/local/lib/python3.9/dist-packages/gradio/blocks.py", line 1346, in process_api
result = await self.call_function(
File "/usr/local/lib/python3.9/dist-packages/gradio/blocks.py", line 1074, in call_function
prediction = await anyio.to_thread.run_sync(
File "/usr/local/lib/python3.9/dist-packages/anyio/to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "/usr/local/lib/python3.9/dist-packages/anyio/_backends/_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "/usr/local/lib/python3.9/dist-packages/anyio/_backends/_asyncio.py", line 867, in run
result = context.run(func, *args)
File "/notebooks/RVC-Paperspace/infer-web.py", line 529, in click_train
f.write("\n".join(opt))
OSError: [Errno 122] Disk quota exceeded
2024-01-24 00:12:01 | INFO | httpx | HTTP Request: POST http://127.0.0.1:7860/api/predict "HTTP/1.1 500 Internal Server Error"
2024-01-24 00:12:01 | INFO | httpx | HTTP Request: POST http://127.0.0.1:7860/reset "HTTP/1.1 200 OK"

[proper shale]

This could be related to the virtual audio cable you're using. I've heard some peeps have problems with VB Audio Cable, so try the VAC by Muzychenko and see if it works.

[bright umbra]

Im using VB

[proper shale]

yeah exactly

[bright umbra]

Thank you! Sounds so much better

[proper shale]

Glad I could help :)

[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

[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

[cedar orbit]

/guides realtime

[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)

📚 Audio Guides

• Perfecting Audio Isolation on Low-End Rigs: A Practical Guide, by Litsa The Dancer and Faze Masta Google Docs
• Gathering and Isolating Audio, by SCRFilms Google Docs
• How to make a good model All-In-One guide, by LUSBERT Rentry
• Creating Datasets for RVC using iZotope RX, by LUSBERT Rentry
• Vocal Mixing Tutorial, by Roomie YouTube

🛠️ Tools

Youtube Downloaders

• https://github.com/yt-dlp/yt-dlp
• https://github.com/dsymbol/yt-dlp-gui

Audio Separation/Isolation

• https://mvsep.com/
• https://multimedia.easeus.com/vocal-remover/

More commands: -colab, -uvr, -karafan, -overtrain, -help

[foggy basin]

does anyone have a tutorial for that real time ai voice changer saw there was a section for it in the old server

[violet heron]

-realtime

[azure marsh]

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

[violet heron]

First one

[warm falcon]

I made the inference UI pretty bland, hopefully people cant mess THAT up too often

[glad zealot]

add dark theme :D

[violet heron]

People will still mess that up somehow

[warm falcon]

alwaysss, that is just the huggingface screenshot

[glad zealot]

me frfr

[warm falcon]

that is literally stuff I coded

everything actually

[nimble rock]

The tts in the new applio isn't working properly at least on my end. First it doesn't make the tts_rvc_output file, I have to make that myself. That and it doesn't use the model I select to change the audio. It's always just the tts base voice. Is anyone else having this issue or is it something I'm doing on my end

[violet heron]

Doesn’t applio have it in a separate tab?

Not in a drop-down

Send the space link

[warm falcon]

ooo idk let me check it out

https://huggingface.co/spaces/thejenja/Applio-V3

i just looked up applio on huggingface and that is where i got the screenshot from

i assumed it was applio ^

[violet heron]

Not real

Looks like it was forked off Ilaria RVC

[warm falcon]

damn so even her fork is getting ripped off

she actually gives credit

[violet heron]

coughs applio

[warm falcon]

ima run applio rn see what their UI is like

https://colab.research.google.com/github/iahispano/applio/blob/master/assets/Applio.ipynb#scrollTo=ymhGfgFSR17k

from this colab link

[viscid moss]

https://huggingface.co/spaces/PlayerBPlaytime/Applio-V3

[violet heron]

Wait

I made a fork of ilaira RVC to make a fake applio leak

Never meant supposed to be for the public

[warm falcon]

ohh niice

that is actually a different UI

[violet heron]

(Space was set to public because I’m on iOS and doesn’t work on private)

So bro probably forked that

[warm falcon]

lolll

okay this one is good

they actually changed it up a bit

back then it was just a rip off of my gui

[violet heron]

(The reason it doesn’t say forked off my name is because I deleted the fork when I was done creating the “leak”)

[zinc ore]

what should settings for morgasn freeman be???

[glad zealot]

depends on your voice

[zinc ore]

oh yeah lol

[abstract flower]

hello, how to train a voice on rvc on pinokio? Thanks

[brittle wing]

how do i use it on discord

[frank spire]

Hi, help me find tyler durden in Russian voice acting

[delicate wharf]

I'm using easyGUI by rejekts and always get this error though I put the input with only just 1 .mp3 file, anyone has any idea?

[deep bear]

anyone have Hiroi Kikuri model?

[rare goblet]

Ayo? @deep bear level 1 !!!

[brittle wing]

What do you guys use on UVR to separate main vocals from background vocals + instrumental? And vice versa

[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

[torn halo]

Guys, I did trained my model, but my pc crashed after the Epoch finished, any way I can train that model anymore?

[rare goblet]

Ayo? @torn halo level 1 !!!

[delicate wharf]

I found this

and this

but still it's a hard task

[hearty folio]

guys is it possible to train a model with an AMD GPU (mine is rx5500 xt)?

[brittle wing]

Thank you!!

But the thing is, is I'm trying to get the main vocals not the background vocals

It is just isolating all of the vocals

Together

[delicate wharf]

I'm testing too

but I tried another thing

uhm so you have the vocal and inst isolated right

[brittle wing]

Yeah?

[delicate wharf]

I used BVE model on MVSEP

but the quality of course isn't too great

it has too many layers of AI that makes it sounds robotic when you used it for inference

so I'm testing UVR Karaoke-2

[brittle wing]

Alright :D

[modern surge]

How do I add my dataset in Applio Colab?

[gloomy dew]

-karafan

[azure marsh]

KaraFan

🥇 Awarded to 1st place by the community as the Best software in the World, for having the less vocal residues in Music track!

Credits to Eddy, Leo_Frixi and Gdr for fixing CUDA problems on the google colab notebook!

[rare goblet]

Ayo? @delicate wharf level 4 !!!

[delicate wharf]

so uh, yes Karaoke-2 has the problem you asked

it isolated all the vocals

[brittle wing]

Dang

[delicate wharf]

while with BVE it's best to split the vocal

so go with BVE

[brittle wing]

Alright thank you!

[upper furnace]

sorry, the dq model meker role was removed because I made an ugly model but which one did I not understand? Thank you

[crimson mango]

HI guys. how to clone uploaded audio file

[brave hearth]

what do i do when it says extension of file should be the following, pth, onnx?

[keen silo]

Hey yall, what does cache all training sets mean

[forest sluice]

Don't need to enable it

It's just speeding the training process for short datasets, far slower on longer ones

[brittle wing]

what crossfade setting should i be using?
start 0.2 end 0.8 or start 0.1 end 1.0?

the setup guide and bot are conflicting each other :x

not hearing any difference between the two

[tame finch]

can someone help me with rvcwebui?

[zealous shuttle]

Uhh....Excuse me, how can this happened> my model is already zip format

[forest sluice]

Remove the ?download=true portion in your link

[dull shard]

Why when i search the model it doesnt respond

[lost crescent]

Hi guys, there a single command I can run in RVC mangio fork to download all models in one go?

[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

[regal hill]

has rvc always had a pretrain

cuz i dont want pretrain at all but that option is gone in v0.25

[hushed knot]

how to add voice models?

[brittle wing]

why dont the models sound like they should for me

[forest sluice]

new pretrains

[oblique venture]

i tried to run MDX-Net_Colab but it give me the error IndexError: too many indices for array: array is 1-dimensional, but 2 were indexed

[elder iron]

how to make the google collab saved to google drive (so it will be loaded from there and not downloading again and again)

[regal hill]

yea ik but was it using pretrains 1 month ago

like without the option to deselect pretrains

[deep hare]

how do i make a model ? 😭

nvm

[glad zealot]

it always had a pretrain yes, and people are currently interested in making new and better pretrains, hence why you see 0v2 and rin_e3

[glad zealot]

voice model training guis always had a pretrain it comes with by default

[regal hill]

Ohhhh ok

[glad zealot]

basically from my understanding how many times the program trained the model using the dataset

[proper shale]

An epoch is basically how many times your dataset has been trained fully.

E.g. 200 epochs
Means the dataset has been trained fully 200 times.

[rough tinsel]

https://github.com/SociallyIneptWeeb/AICoverGen/tree/main?tab=readme-ov-file

i wanted to run locally on my windows pc
i had python 3.11 so i uninstall and reinstalled 3.9
i cloned the repo
i created a virtual env with python and tried to:
pip install -r requirements.txt

even tried running ipynb locally idk i should or not but heres the...:

[sinful plover]

hello

can someone help me find a kris jenner ai

[copper pebble]

#1175430844685484042
#🔍│find-models (/search command)
weights.gg website

[rough tinsel]

no help for me? lol 😆 looks like i encountered something new

[proper shale]

yea honestly i got nothing

[copper pebble]

same

[proper shale]

@sand iris help me out on this one 😭

[rough tinsel]

you seriously did the same steps?

[proper shale]

nah I don't use locally

and, don't have experience on AICoverGen local issues

i can figure out what it is (sometimes) on other forks

[rough tinsel]

ok in requirements.txt its 4.65.0 but in that link its 4.64

[copper pebble]

mj doesnt us locally ??

wha-?

[proper shale]

yeah I don't

i have a MacBook

a 2012 MacBook

[rough tinsel]

ic

[copper pebble]

YO BRO SAME SITUATIONS

[rough tinsel]

so macbook not so powerful to run this?

[copper pebble]

i mean potatoes pcs

[rough tinsel]

ic

but i wanted to run locally

i tried the same with linux

[rare goblet]

Ayo? @rough tinsel level 2 !!!

[rough tinsel]

and same error

i mean i had py 3.11 there

and i dont want to mess up with python there

because many other may be dependent on 3.11

so i came back to windows and tried and this now

[small umbra]

you came?

[rough tinsel]

i mean i first tried on linux

then i switched to windows

[proper shale]

😭

[rough tinsel]

can i get someone who run locally

[proper shale]

that's why i pinged him

hoping he comes on here soon

[rough tinsel]

oh ic

me now installing all the packages one my one manually ehh

[ember coral]

Okay so i have a question can I train a model to a different language?

[proper shale]

yeah

[sand iris]

What's going on ?

are y'all still on fairseq ?

[rough tinsel]

means

huh

no i ran pip install -r requirements.txt

[sand iris]

Yeah

[rough tinsel]

and ended up here

[sand iris]

Ended on this ?

[rough tinsel]

yea

[sand iris]

Also before we continue : Which OS do you want to do it on ?

[rough tinsel]

windows

[sand iris]

aight

[rough tinsel]

also tried on linux

kali

[sand iris]

it gave you a link on the error message

[rough tinsel]

yea

[sand iris]

you'll need the "Microsoft Visual C++ Dev tools". If you hadn't installed that

[rough tinsel]

in windows

oh ic

[sand iris]

If you did, mind going back to it's installer ?

[rough tinsel]

no i didnt

[sand iris]

there's like two optional components needed to have it work and they aren't selected by default if I remember right

screenshot once you get to the "component selection" part

[rough tinsel]

which to install

if you could tell

[sand iris]

can I get a screenshot pls, things aren't fresh in my mind

[rough tinsel]

sure

[ember coral]

Okay so i have a question can I train a model to a different language?

[rough tinsel]

this when i upgraded pip and tried

[proper shale]

Yes. You can.

[sand iris]

see the microsoft link it gave you ?

[ember coral]

How?

[proper shale]

If ya have the person speaking on the other language, just train that

[rough tinsel]

https://visualstudio.microsoft.com/visual-cpp-build-tools/

this

[sand iris]

yes

[proper shale]

Simple as tht

[rough tinsel]

i came here

[small umbra]

you came?

[sand iris]

great

[rough tinsel]

i mean this ss

[ember coral]

Oh okay it’s that easy thank you ☺️

[sand iris]

aight, download 'em now

[proper shale]

You're welcome :) Lmk if you have any issues

[rough tinsel]

should i choose python?

if i choose python i landed here

[rare goblet]

Ayo? @rough tinsel level 3 !!!

[sand iris]

the download link is up on the top

top left

[rough tinsel]

which one

after i click on no 6 i am here

[sand iris]

https://aka.ms/vs/17/release/vs_BuildTools.exe

[rough tinsel]

and sedly heres no link

suspicious

shall i download?

[sand iris]

if you really want to make sure, it's the link I got from there

[rough tinsel]

oh like this

[sand iris]

ya

[rough tinsel]

downloaded

[sand iris]

great

beware : it's gonna be like 2-3GBs of download on the next step

just so you know

[rough tinsel]

oh hell

which one to download now

[sand iris]

"Desktop development with C++"

And take a good look at the optional components

[rough tinsel]

welp

[sand iris]

select those two :

• MSVC v143 (or v140 something)
• Windows 11 SDK (even on windows 10 systems)

the rest is not needed

once that's done all that's left is to install, and re-attempt to install the python packages it previously failed on

[rough tinsel]

i selected win 11 sdk and which one the bottom most 3

[rough tinsel]

shall i uncheck all other?

[sand iris]

yeah

[rough tinsel]

ok there are 2 win 11 sdk

[sand iris]

keep the one with the bigger number

[rough tinsel]

one at 2nd and other bottom 7

[sand iris]

the one that was selected before is good

[rough tinsel]

ok top one

these two?

5.39GB??? woah

[sand iris]

MSVC v143

[rough tinsel]

ohkk top 2

[sand iris]

older ones will also work but stick to newer tbh

[rough tinsel]

can i install them on other drive? like D drive?

by default it's

[sand iris]

Not sure

[rough tinsel]

ok then lemme give 5.63 GB

;-;

installation started

how long will it take

25% downloaded and 10% installing

50% ; 35%

[little meteor]

like please help me, goofy 1 step has me like:

[proper shale]

oh

you didn't run Set Training Variables

[little meteor]

aaaah bruh

there is like a button next to it

i just completed the information in it

😂 😂 😂 😂 😂 😂 😂 😂 😂 😂 😂

[proper shale]

this happens

dw

[little meteor]

hahaha now it works

thank you!

[proper shale]

you're welcome :)

[little meteor]

never thought the king of curses would ever help me out

[proper shale]

[little meteor]

cursed energy can be used for good

[proper shale]

frfr

[wraith shale]

GUYS

i cant open start_http

[proper shale]

check #🔍│help-w-okada pins

[brazen merlin]

can anyone help me Im trying to make a rai voice for Bloom from winx club but idk how to do it. I want to make it so i can type something and it says it in her voice but I want it to be super accurate any tips?

[proper shale]

just train with a good dataset, then use a TTS like ElevenLabs and infer with Ilaria RVC

check #1159513888199540817 on how to train models and make good datasets

[brittle wing]

please?

[molten pecan]

@brittle wing no music is allowed

next time send a youtube link instead

or a google drive file link

[brittle wing]

Oh

[brittle wing]

I will send it by google drive

https://drive.google.com/file/d/1-30SosklyPMOYlQ9SZQXJQf8RTzV3BP3/view?usp=sharing

here better

Can you upscale this song please?

[little meteor]

why i see 3, my project name is me but the me one has the lowers disk space compared to the others so which one is my voice model

[half cove]

me.pth

[little meteor]

and the other 2 ones what are?

[rare goblet]

Ayo? @little meteor level 3 !!!

[brittle wing]

https://drive.google.com/file/d/1-30SosklyPMOYlQ9SZQXJQf8RTzV3BP3/view?usp=sharing Please Upscale this song? @half cove

[half cove]

Training

G and D are training files

[proper shale]

^^^

[half cove]

On mobile rn

Also I don't know how to upscale songs lol

[pastel hatch]

I just trained a model and whenever I try to use it, I get an Error:
Anyone ablet to help me?

Traceback (most recent call last): File "C:\Users\gbl\Desktop\Music\AI_Music\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\routes.py", line 437, in run_predict output = await app.get_blocks().process_api( File "C:\Users\gbl\Desktop\Music\AI_Music\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\blocks.py", line 1349, in process_api data = self.postprocess_data(fn_index, result["prediction"], state) File "C:\Users\gbl\Desktop\Music\AI_Music\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\blocks.py", line 1283, in postprocess_data prediction_value = block.postprocess(prediction_value) File "C:\Users\gbl\Desktop\Music\AI_Music\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\components.py", line 2586, in postprocess file_path = self.audio_to_temp_file( File "C:\Users\gbl\Desktop\Music\AI_Music\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\components.py", line 360, in audio_to_temp_file temp_dir = Path(dir) / self.hash_bytes(data.tobytes()) AttributeError: 'NoneType' object has no attribute 'tobytes'

[proper shale]

You didn't select something properly

Make sure you've selected the audio correctly/put the right path on the audio file path place

[pastel hatch]

When I change the model it works fine

[maiden mulch]

Guys, can u help me with voice changer? First question. When I say something, the sound begins to repeat itself many times and becomes, as it were, an echo (repeated and becomes higher). Second question. Is it possible to somehow reduce the sensitivity of the microphone so that it does not hear any noise?

[proper shale]

So it's probably the model then

[pastel hatch]

So I probably gotta retrain it?

[rare goblet]

Ayo? @pastel hatch level 1 !!!

[proper shale]

1. Activate Sup2 and increase S thresh

2. S Thresh is that

[proper shale]

Nah, try using it somewhere else, like in a Colab

And see if it's the model that's broken

[pastel hatch]

Okay, so if it doesnt work in colab too, its probably broken?

[proper shale]

yeah

[pastel hatch]

Okay, thanks!

[proper shale]

you're welcome :)

lmk if anything happens

[pastel hatch]

okay 🙂

[maiden mulch]

Oh thanks, that helped. But a few more appeared - how to reduce the res column at the top left, what does the extra item do and how to make the voice smoother and clearer? Otherwise it sometimes gets interrupted and the note changes abruptly

[proper shale]

You can't reduce res, technically - it's the time the GPU takes to process stuff. If you're on AMD however, and that res is way too big, you can export to onnx and solve it being laggy.

Extra uses more of your CPU to make the result better

The last question is kinda... impossible to solve, as you can't really do much other than play around with s thresh and gain sliders

[maiden mulch]

uh, sounds sad. Thank u 🙂

[rare goblet]

Ayo? @maiden mulch level 1 !!!

[delicate wharf]

uhm

[proper shale]

You're welcome :)

[delicate wharf]

since now this is the active place can I ask for opinion about a sample?

[proper shale]

sure (as long as it ain't copyrighted)

[delicate wharf]

thoughts on this?

(it's Italian)

[proper shale]

does have artifacting

[delicate wharf]

so the model is not good?

should be trained again I guess

[proper shale]

i mean... it sounds okay i guess

just artifacting

how many epochs is this?

[delicate wharf]

400

[proper shale]

oh, it's probably the dataset then

[pastel hatch]

How many minutes of audio is good for training?

[proper shale]

but it sounds okay

[proper shale]

go for 10-20

[delicate wharf]

idk, the dataset is clean and high quality (collected from wiki), but it seems the character's Italian is so heavy

for short samples it sounds quite alright

but when I put it into covers it does sound more artifacting

what a headache

[ember coral]

okay so how do i keep training my model later?

[proper shale]

#📑│making-models message

[delicate wharf]

so what from the dataset could be causing the model to artifacting? the dataset is low quality?

I'm currently unhappy with any result that comes out (too much artifacting) so I'm redoing everything...

[rare goblet]

Ayo? @delicate wharf level 5 !!!

[proper shale]

You can use any extraction type in any model

Also, that's the AMD version

[small umbra]

..how to install dez nuts in yo mouf?*

[rare goblet]

Ayo? @sly summit level 1 !!!

[azure marsh]

RVC Guides (How to Make AI Cover)

Documentation

🇺🇸 English (main)

https://docs.aihub.wtf/

​

Translation by country

🇧🇷 Brasil (PT-BR)

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

🇪🇸 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

[bitter wave]

Do you think it's possible to use other TTS Method than Edge-tts or google-tts ? like TorToise, AudioDLM or Bark ?

[rare goblet]

Ayo? @bitter wave level 1 !!!

[proper shale]

Here's the guide on it: https://rentry.co/VoiceChangerGuide

[rare thorn]

on CoverGen_NO_UI_v2_en.ipynb why do all my models sound shit on a lot of stuff

its NOT a model issue

[fickle dust]

Hi guys how do I make the real time rvc voice appear to others I’ve tried a lot and it doesn’t work

[waxen lintel]

yo can i use this shit in fl studio

[rare goblet]

Ayo? @waxen lintel level 1 !!!

[proper shale]

You need to have a Virtual Audio Cable

[proper shale]

covergen has some issues in quality ngl, try other stuff

[proper shale]

i mean, probably (if you're talking realtime voice changing)

it'll probably be off sync tho

[waxen lintel]

how do i make it work on discord cause it doesnt seem to work

[proper shale]

ya need a virtual audio cable

[waxen lintel]

i have it downloaded

[proper shale]

now just set the input/output stuff correctly

so,

W-Okada:

Input - Mic
Output - CABLE Input (if using VB Cable)
Monitor (optional) - headphones

Discord:

Input - CABLE Output
Output - Headphones

[rare thorn]

ah

[subtle island]

/search naruto

[gusty kestrel]

I have found 7 results for the search naruto...

Naruto Uzumaki

Owner: texx18
Uploaded: <t:1692316800:R>

Epochs

200

Technology

RVC

Algorithm

Crepe

Requested by cxsho · Jan 24, 2024, 7:10 PM

[sullen plover]

Hi how can i put my ai voice on discord i dont know how to do it i already put my micro on input but it did nothing

[proper shale]

You need a virtual audio cable, check https://rentry.co/VoiceChangerGuide

[serene leaf]

Hey can someone help me with this

im having a problem with the start up

whenever it opens its just white

and it just closes after a few

[subtle island]

perfect settings for naruto?

[proper shale]

What does CMD say?

[serene leaf]

where can I send a screenshot

[proper shale]

do a post on #1192011222023950368

[serene leaf]

there

[errant scroll]

hey guys...we will organize an event for a poetry organization, but unfortunately we have a friend who passed away. i want to read a poem with his voice recording. is this possible?

[proper shale]

yea, you'll need to train a voice model tho

check #1159513888199540817 for more info

[errant scroll]

thanks 🙏🏼

[hoary zodiac]

i wanted to make my own voice model but it's just stuck since i clicked connect with google drive and nothing happened, any help?

https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS#scrollTo=GnZdaTi6FgHf on this site

i have voice prepared in .mp3 and .wav but i cant even login to google drive because when i click connect with google drive nothing happens

[proper shale]

maybe opened an extra window on ur browser

right click on the icon on the taskbar and check the browser windows

[hoary zodiac]

i dont have any windows opened

[rare goblet]

Ayo? @hoary zodiac level 1 !!!

[hoary zodiac]

just youtube, this rvc v2 site and my rvcDisconnected folder in google drive

[proper shale]

huh

did you even load the preprocessed stuff back in the first for resuming step?

try running that again

[proper shale]

u prolly disabled pop-ups then

disable for colab only

[little meteor]

all of them are green this time, only index training isnt

[proper shale]

meant enable it again lol

[proper shale]

you didn't need to preprocess/feature extract again btw

also maybe this could be because... dataset wasn't formatted properly

[little meteor]

rip

i will go from scratch if i cant fix it, buy before, i see this error now

[proper shale]

oh yeah ok this is fine

[little meteor]

like for sure i fucked this up because i pressed stuff in bad order

but maybe is fixable

[proper shale]

if ya wanna start from scratch delete the "me" folder inside rvcDisconnected

on google drive

[earnest bobcat]

When I request the voice file of any character or person from the bot, it gives a link? @molten pecan

[little meteor]

yeaaaah, but what about you saying is fine? can i fix it?

[proper shale]

not the index training part unless you uh

delete the current dataset .zip

[molten pecan]

that link contains the files of the voice model. you download the files and put them inside the software

[proper shale]

and you'll have to zip the dataset folder again, following certain things

[molten pecan]

i'm busy rn so ask MJ

[little meteor]

aah

alright

[proper shale]

noooo 😭

[little meteor]

thank you

[proper shale]

dont overwhelm me

[proper shale]

you're welcome, wait ill link you to the dataset formatting thing, said this earlier btw

[molten pecan]

@flint meadow ask MJ how to make AI cover

🤭

[proper shale]

#📑│making-models message

[little meteor]

how do i reset stuff? cuz when i reentered some stuff was already green, so i couldnt do it in order like for the first time

[earnest bobcat]

Can I do anything I want after clicking the link from the bot and downloading the audio file? @proper shale

[proper shale]

on the top: runtime --> disconnect and delete runtime

[little meteor]

alright

but first

[proper shale]

...yeah?

[little meteor]

that format thing

[earnest bobcat]

that`s it?

[proper shale]

...what do you want me to say

i mean, you'll need a place to do the inferencing (voice converting)

[proper shale]

#📑│making-models message

im lazy to retype

[earnest bobcat]

oh okay thanks

[proper shale]

you're welcome :)

[sullen plover]

ty

[rare goblet]

Ayo? @sullen plover level 2 !!!

[proper shale]

You're welcome!

[little meteor]

bro wait

i looked in drive now

and i found a new zip thing and in it has the stuff the error said couldnt find, so ig now i have to link this path rather than my last zip? i mean i will try see if it works 😂

said something before that it couldnt find that 0_gt_wavs

idk i will try

[proper shale]

whats weird is

the folder in the 2nd shot is just named "experiment_name"

😭

[little meteor]

they shouldnt let noobs like me cook with AI 😂 😂 😂 😂 😂 😂

[rare goblet]

Ayo? @little meteor level 4 !!!

[proper shale]

my first training session was on a phone and I was so dumb that i screwed up big time and it trained an epoch every second

so idk... seems like ur doing better

[little meteor]

that sounds like the pc should burn 😂 😂 😂

[proper shale]

naaah it just... didn't do the right steps

and didn't train properly

[little meteor]

how long are you into this?

since when

[sullen plover]

can someone help with my voice changer idk why it don't sounds well, i mean it work but the ai voice is not what i excepted, can someone send me his settings audio please?

[proper shale]

man, training since july last year

[little meteor]

respect, seems like it went well for you, you got good at this, niice, im curious what you cook with ai, if you just use it for dumb stuff like me 😆 making covers and stuff, or something actually useful for humanity

[proper shale]

i just make models of artists i like, and covers yeah

[little meteor]

niiiice

[proper shale]

there's not a lot you can do about audio settings other than noise suppresion n stuff

[little meteor]

but now i got what you said

with folder is named experiments_name

and in error is searching for folder me

[proper shale]

yeah

[little meteor]

so i should rename the folder

[proper shale]

thats why i think you should delete the me folder entirely

😭

because probably ur dataset is not formatted properly too

and then u can fix that

and it should work fine

[little meteor]

shit is so funny 🤣🤣 🤣 🤣 🤣 🤣 like idk how i got the system so confused, i remember putting me as name but got saved with different name, idk man 🤣 🤣 🤣

[proper shale]

im also confused lol

[little meteor]

now training works but

it starts epochs from 1

not from 500

is this normal?

[hoary zodiac]

i cant do it, now it says that it cant find the file in rvcDisconnected

[hoary zodiac]

help pls

[sullen plover]

idk when i turn off the ai it sounds good

wdym by that?

[proper shale]

yea then ig there's not a lot that can be done

[sullen plover]

or maybe i have a bad mic idk

[proper shale]

hhmmmm this happens when d/g files corrupt/screw up on saving

[little meteor]

i think i just made new training, cuz now is a different folder, experiment_name and the me folder is untouched, i mean is fine as long as it works, but when it ends i will ask you where to press to continue resuming and not press by my own and screw it again

[proper shale]

yeah sure

oh that explains it lol

[little meteor]

😂 😂 😂 😂 😂

[proper shale]

if u didnt connect to drive that might happen yeah

[little meteor]

i need a lower role, the default noob is too nice for my skill

[hoary zodiac]

but i logged in

google drive

[proper shale]

then check if the folder is named correctly

[hoary zodiac]

its "rvcDisconnected"

[proper shale]

try restarting runtime and connecting to drive again, that's weird

[hoary zodiac]

i tried again now its says this

wait

[proper shale]

that's... probably why

[hoary zodiac]

i logged in google drive again and this happened

so idk

im not good at coding

not even understanding

[proper shale]

idk either bro, try a diff google account maybe

on everything

[little meteor]

what program to use to split voice from song, vocalremover on google did great but it limits you to one song per ip, i can just vpn but is annoying, tried some other sites and they couldnt even detect the voice good, ig going local is the way, but i want to be sure there is a tool that is worth to install and does a good job, otherwise is pain and i choose the vpn pain cuz that service did split it fine

[proper shale]

-audio

[azure marsh]

Suggestion for @little meteor

📚 Audio Guides

• Perfecting Audio Isolation on Low-End Rigs: A Practical Guide, by Litsa The Dancer and Faze Masta Google Docs
• Gathering and Isolating Audio, by SCRFilms Google Docs
• How to make a good model All-In-One guide, by LUSBERT Rentry
• Creating Datasets for RVC using iZotope RX, by LUSBERT Rentry
• Vocal Mixing Tutorial, by Roomie YouTube

🛠️ Tools

Youtube Downloaders

• https://github.com/yt-dlp/yt-dlp
• https://github.com/dsymbol/yt-dlp-gui

Audio Separation/Isolation

• https://mvsep.com/
• https://multimedia.easeus.com/vocal-remover/

More commands: -colab, -uvr, -karafan, -overtrain, -help

[proper shale]

mvsep might be a good one

not local tho

-uvr

[azure marsh]

Ultimate Vocal Remover

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

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

[little meteor]

uvr good?

[proper shale]

this one is local

[proper shale]

yeah, but may take long if you have an iGPU only

[half cove]

You can use uvr on cpu it just slow

[proper shale]

haven't used it tho

[little meteor]

i have dedicated but only 4gb, like i could wait like half and hour 1 hour for a song if the result is good, but is not worth if i have to wait half a day for a song or if my bad components influence the quality of calculations, so it doesnt do a great split, how long would it take? long as voice cloning?

[half cove]

With gpu it should take less than that

[proper shale]

^

probably like 5-10 mins idk

[little meteor]

then i will try, thank you

those sound good

[proper shale]

hope it works

[little meteor]

even longer, just not longer than 1 hour, i tried cloning voice locally with only 4gbs, goofy, waited 3 hours for 30 epochs, when can i wait that on colab for 500 echos or even more, and no loud fans

[proper shale]

yeah

batch size 2 is unbearable probably

[little meteor]

did 5 😂 😂 😂 😂 😂 😂 😂 😂

was shiiit

tried more and memory limit error

rip

😿

[proper shale]

oh damn

[ripe yoke]

So how do I request a voice to be made?

[proper shale]

that's rare for 4 gig cards

[proper shale]

#1159289738314919936

[little meteor]

i mean probably poor settings made by me aswell and bad program used, i followed a youtube video, but it can be aswell my card, pretty old so rip

now life is easy after i joined this helpful community

respect to you all

[proper shale]

🙏

[little meteor]

what s the deal with index files, you get 2 files .pth and one .index, and i see if i just put the .pth i can still use the model, how does this work?

[proper shale]

actually the right .pth is the one with your model's name - the other (G, D) are training

and yes u can just use without index

[little meteor]

but in general, what the use of the index file

cuz in rvc it shows box to link to that aswell, if i dont link it will sound worse?

[proper shale]

basically, all mannerisms (accent, how they say certain things) in a file that modifies outputs

[little meteor]

aaaah

so i should use it?

if i want better results

[proper shale]

yeah, it's good for accents

[little meteor]

thank you

[proper shale]

yw :)

[little meteor]

it finished, now what is the first thing to press to resume?

[rare goblet]

Ayo? @little meteor level 5 !!!

[errant brook]

edp model???

[molten pecan]

Why 😭

[errant brook]

dw ❤️

[half cove]

[errant brook]

found one

[half cove]

😭

Did you install the precompiled?

[brittle wing]

no I did the normal install with run-install.sh

since im on linux

[half cove]

Oh i dont know about how to install on Linux

[small umbra]

..how to install dez nuts in yo mouf?*

[half cove]

I wish i could help more

[molten pecan]

@viscid moss can help you :3

[brittle wing]

Oh ok thank you

@viscid moss do you mind if I dm you for help or would you rather do it here?

[viscid moss]

Let's do it here

[brittle wing]

Traceback (most recent call last):
File "/home/reu/Documents/Applio-3.0.4/app.py", line 12, in <module>
from tabs.inference.inference import inference_tab
File "/home/reu/Documents/Applio-3.0.4/tabs/inference/inference.py", line 3, in <module>
import regex as re
ModuleNotFoundError: No module named 'regex'

Did install regex via pip, still the same error

[rare goblet]

Ayo? @brittle wing level 5 !!!

[half cove]

Verified chat moment

[pearl tangle]

is an rtx 3060 good enough to run this

[molten pecan]

Yes :3

[viscid moss]

You used this method, right?

[brittle wing]

Yes

[viscid moss]

Can you try the makefile method?

[brittle wing]

make run-applio?

[viscid moss]

yep

[brittle wing]

python app.py
make: python: No such file or directory
make: *** [Makefile:20: run-applio] Error 127
reu@reu-Latitude-E5440:~/Documents/Applio-3.0.4$ make run-applio.sh
make: Nothing to be done for 'run-applio.sh'.

[viscid moss]

After cloning the repository, you didn't install the dependencies or did you?
with the:
pip install -r requirements.txt

[brittle wing]

i ran run-install.sh, then it automatically created the venv environment and installed the dependencies

[viscid moss]

and now u have that missed dependence hmmmmmmmmm

[proper shale]

might be late but, export model and once you come back you'll do every step before load dataset zip. then skip that, preprocessing and feature extraction until u see the steps for resuming. run those, skip train index and run training step

make sure you got the same training variables

and same batch size

[viscid moss]

lemme ask applio staff rq.

[wary mantle]

Why won't my voice changer client work? Been triying for awhile.

[rare goblet]

Ayo? @wary mantle level 1 !!!

[viscid moss]

https://github.com/IAHispano/Applio/issues/200
Check this issue

[lavish viper]

I would like to ask for a little help
does not start
I received a link the other day from which I had to download and copy something to make it work, if I remember correctly
but I don't know what and where I need it :c

[little meteor]

I did something and it started fine now, from 500, is at 700 now, i think is fine, but what im curious now is how to understand that data people talk about, there is a table at the bottom with graphs and idk at what i should look, to know when to stop. I put to make 2000 epochs, so in total 2500

[proper shale]

-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

[proper shale]

there r guides here

this'll help you comprehend TensorBoard

[brittle wing]

oh wait the makefile method is doing something

its downloading a bunch of packages

[sand iris]

That's why I prefer manual installs, you can actually manage them

[brittle wing]

mine was a manual install, i did run-install.sh

then did run-applio.sh then got an error

[sand iris]

Yeah it sets up most of the things for you though

Actually I haven't taken a loon at the newest iteration of the script yet

[brittle wing]

i think i find applio as one of the best RVC forks ive ever used

[little meteor]

https://tenor.com/view/swap-jujutsu-kaisen-jujustu-tech-high-noroi-curses-gif-20946635

[sand iris]

Oh noes

Wait a sec do they tell you to install python 3.11 ?

[brittle wing]

on what

[sand iris]

lol, lmao even

The install script will not give you a working install depending on your distribution

e.g. if your distribution ships python 3.11 or 3.12 as the default

The error message that pops up when it's not installed also tell you to install 3.11 which will break

[proper shale]

best friend moment

[brittle wing]

@viscid moss ISSUE SOLVED

[half cove]

[viscid moss]

[brittle wing]

I had to change import regex as re to import re

[viscid moss]

oh, as the issue said

[brittle wing]

yes

I did that and it worked

[brittle wing]

NOOOOOOOOOO

Downloading https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/pretrained_v2/f0G40k.pth to rvc/pretraineds/pretrained_v2/f0G40k.pth...
100% [........................................................................] 73106273 / 73106273
Downloading https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/pretrained_v2/f0G48k.pth to rvc/pretraineds/pretrained_v2/f0G48k.pth...
100% [........................................................................] 75465569 / 75465569Traceback (most recent call last):
File "/home/reu/Documents/Applio-3.0.4/app.py", line 17, in <module>
from tabs.tts.tts import tts_tab
File "/home/reu/Documents/Applio-3.0.4/tabs/tts/tts.py", line 3, in <module>
import regex as re
ModuleNotFoundError: No module named 'regex'

[sand iris]

isn't regex built-in?

[brittle wing]

i did pip install regex, installed the package but still the same error

[sand iris]

did you do so in the venv ?

[little meteor]

bro my training is stuck at 660 epochs bcs i got limited by no subscription, how do i save these 160 more epochs, cuz i checked drive and there is no file that saved at this current hour so i dont think any of these echos are saved, last .pth is saved one hour ago at 500

[brittle wing]

yes

[untold dome]

hi.. how to get the download link for the models on huggingface itself? https://huggingface.co/sail-rvc/SpongeBob_SquarePants__RVC_v2_ this one for example? 🙏

[sand iris]

I think right click > copy link address on the download button works

[untold dome]

cant see any download button, probs cuz no account made on huggingface?

[sand iris]

You should still be able to download it

[sand iris]

Just to make sure, your shell starts with (.venv) right ?

[untold dome]

ser wtf.. no download button : (

[brittle wing]

applio on linux should automatically start the venv

ive never had this problem before

[sand iris]

Yeah but when you installed regex

I'm just checking if you did so in the venv

[brittle wing]

IT WORKED

[half cove]

[wary mantle]

Anyone know why I can't hear my voice changer and it isn't working? Please I've neen trying for A WHILE.

[untold dome]

which one, ze model.pth?

[rare goblet]

Ayo? @untold dome level 1 !!!

[sand iris]

If it asks for a model that should probably be it? Depends on tbe the tool you are using

[half cove]

They didn't format it correctly

[sand iris]

format as in ?