✨│ai-help

[brittle wing]

wym can yu?

[wispy burrow]

Locally probably not.
But using it on a Colab or Huggingface Space should be possible.

[silver bobcat]

dunno

[brittle wing]

i wanna do huggingface space

i just dont know how to do all of dis

[waxen cliff]

does anyone know how to train without gpu?

[silver bobcat]

[wispy burrow]

You can't.

[waxen cliff]

theres a colab that lets you do it

does that use gpu

cuz mine just isnt good enough

[wispy burrow]

Ah yes, Colabs uses Google's GPU's.

[waxen cliff]

ohhh

[rare goblet]

Ayo? @waxen cliff level 1 !!!

[brittle wing]

if yu can screen record cam yu show me how to make a ai cover

[silver bobcat]

is there any way i can fix error: gpu card in aigencover?

[waxen cliff]

use replay or hugging face

[brittle wing]

how

[waxen cliff]

what

[tame mica]

what is this channel

[wispy burrow]

I never used Colabs or Huggingface Spaces...
I've only used several RVC's locally.

[brittle wing]

oh

[silver bobcat]

[violet heron]

Colab

[waxen cliff]

ok

[violet heron]

Are you on colab?

Or local

[silver bobcat]

local

[brittle wing]

screen record how to do it

[rare goblet]

Ayo? @brittle wing level 2 !!!

[lunar rapids]

hello
I just tried it, is it normal for 2 hours stuck in cmd?

[rare goblet]

Ayo? @lunar rapids level 1 !!!

[wispy burrow]

Stuck at what step? What were you doing?

[wispy burrow]

#🔍│help-w-okada
It's the realtime VC channel.

[lunar rapids]

ahh ok, thank you 🙏

[subtle rover]

what am i doing wronh

why is it not training

[wispy burrow]

Ah, original RVC.
Did you do all the steps prior to training?
Is that all the cmd shows?

[subtle rover]

yup:(

[wispy burrow]

How long is it like that?

[subtle rover]

like over 10 mins

[wispy burrow]

Something should have happened by then, normally.
Maybe the program just froze.

[subtle rover]

tried multiple times and it stops during training everytime

[silver bobcat]

a

[subtle rover]

sigh

it works fine when i do one click training

[wispy burrow]

So it actually trained a Model?

[subtle rover]

yep but no file in the weights folder

so i decided to try "train model"

[wispy burrow]

Then it didn't train the Model either, just the Index.

[subtle rover]

right

[wispy burrow]

Can you show your settings for all the steps?

[wanton swan]

Should you put the save frequency at 50 or 100? I am able to put it all the way to 100.

[wanton swan]

@violet heron

[wispy burrow]

Ah, I see that your GPU is not supported.

[subtle rover]

yep

:/

so im guessing it wont work then?

[wispy burrow]

I'd advice to save often.

[violet heron]

How much epochs are you doing

[wispy burrow]

Not really.
Or it is going to take a veeeeery long time if it does.

[violet heron]

I ideally do 50

[wanton swan]

I am doing 250

[wispy burrow]

I also notice you're using an old version of RVC.

[subtle rover]

okay.. will wait for an hour or two and see if something happens

yup this is the onlyonei could fie

find

[violet heron]

50

[wanton swan]

50 Epochs is good?

Should I put the save frequency on 100?

[wispy burrow]

https://huggingface.co/lj1995/VoiceConversionWebUI/tree/main
There are several here.

[violet heron]

No what I meant is have the save frequency at 50

[wanton swan]

Do I keep improvement smoothness at 975?

[violet heron]

I usually don’t touch it

[wanton swan]

The option for stop early if there is no improvement

[agile glade]

hello can you give me a tuturial on how to install rvc online in browser

[subtle rover]

is it going to make a difference though? i still dont have a compatible GPU

[wanton swan]

Do I save all or voice?

???

[wispy burrow]

There is a version for amd/intel apparently.
Assuming your GPU isn't an NVIDIA.

[wanton swan]

Hello???

[subtle rover]

ah i see

[rare goblet]

Ayo? @subtle rover level 4 !!!

[subtle rover]

thanks

[wispy burrow]

If you think you may need or want to train more later, Save all.
If not, Save voice.

[proud lily]

grazie

[wanton swan]

Thanks I saved Voice

[sand iris]

Not for training unfortunately

[barren swift]

how can I stop the rvc v2 disconnected colab from closing the runtime automatically

from inactivity

[wispy burrow]

Not even Intel? I thought that one could?

[sand iris]

The amd/intel build can't

But intel can train assuming you have everything setup yourself

Not sure about OS support but I've heard windows should work, but it might require some tinkering

[gaunt ruin]

I have the same question @tranquil cliff

[barren swift]

I opened a second display and put an autoclicker on it while i watch youtube on the main display
seems like it is working thus far but i'll update you if it fails

[wanton swan]

THANKS MAN, IT DOESN'T SOUND ROBOTICY! IT SOUNDS ON POINT!

[rare goblet]

Ayo? @wanton swan level 5 !!!

[tranquil cliff]

okey

[gaunt ruin]

thanks

[gaunt ruin]

bro 💀

[tranquil cliff]

aaaa sorry....-I read it wrong

[tame mica]

the answer is you don't

if you don't have colab pro you may get disconnected from the runtime quickly

[tranquil cliff]

It is not possible to prevent disconnection after 4 hours or after inactivity

[tame mica]

also, you don't even need to train for that long so it shouldn't matter

[round stirrup]

some9one help me pls

[gaunt ruin]

damn

[rare goblet]

Ayo? @gaunt ruin level 6 !!!

[gaunt ruin]

fck off

[fickle mason]

https://colab.research.google.com/drive/1TU-kkQWVf-PLO_hSa2QCMZS1XF5xVHqs?authuser=1

what happens if i use this^

[fiery shadow]

Which colab?

[violet heron]

Notebook loading error

[fickle mason]

why does my 250 epoch say done even tho its all at 0%

=> Epoch: 249 [2023-11-04 21:13:28] | (0:00:01.313936)
INFO:experiment_name:Train Epoch: 250 [0%]
INFO:experiment_name:Saving model and optimizer state at epoch 250 to ./logs/experiment_name/G_2333333.pth
INFO:experiment_name:Saving model and optimizer state at epoch 250 to ./logs/experiment_name/D_2333333.pth
INFO:experiment_name:saving ckpt experiment_name_e250_Success.
INFO:experiment_name:====> Epoch: 250 [2023-11-04 21:13:35] | (0:00:06.583571)
INFO:experiment_name:Training is done. The program is closed.
INFO:experiment_name:Saving final ckpt: Success.
/usr/lib/python3.10/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 20 leaked semaphore objects to clean up at shutdown
warnings.warn('resource_tracker: There appear to be %d '

[violet heron]

Why colab is it

[tame mica]

what

[violet heron]

It shows an error

[violet heron]

That colab is broken

Don’t use it

You will get banned from runtime

[inland hill]

Anybody else having worse quality in discord calls when compared to everywhere else?

[rare goblet]

Ayo? @inland hill level 1 !!!

[barren swift]

do you mean the realtime rvc thing

[rare goblet]

Ayo? @barren swift level 1 !!!

[inland hill]

yeah

[barren swift]

yeah I get that too sometimes

restarting your computer should fix

[fickle mason]

i used rvc disconnected though.... help

how can i fix it so it actually runs the epochs

[inland hill]

any other methods i can try? i just resarted the computer so i can try that now

[barren swift]

idk any other methods

restarting always worked for me

[violet heron]

I don’t think so

That is not RVC disconnected

That is the very old, outdated, banned, colab

[tame mica]

afaik it doesn't matter that much tbh

im pretty sure you can run it just fine

[fickle mason]

https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS?authuser=1#scrollTo=Yj-_npuA_HRB i used this link daddy and its said that error

[fickle mason]

wut, but its at 0%

[violet heron]

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

[fickle mason]

its the same thing? i dont get it

[violet heron]

Use that one instead

This shouldn’t give error

Remove your runtime on the other colab first

And use a alt

[proper shale]

timbo... thats the same one

as they're using rn

[proper shale]

Should be done. That's just a progress bar whenever the epoch isn't finished yet.

[violet heron]

Wait I’m dumb

[fickle mason]

so i just did 250 epochs within 5 minutes on my dishwasher pc?

[fickle mason]

its okay buddy u tried ur best. not all heros wear capes

[proper shale]

...it's a virtual machine

so uh

your pc did nothing

[fickle mason]

oh.. thats kinda mean

idk last collab it took me 1 day

thanks brazillian bbl chan

[tame mica]

timbo helping fail

[violet heron]

I got confused

[violet heron]

Because they linked this

[fickle mason]

damn no reply

sad life

[tame mica]

💀

[subtle rover]

not training on the intel version either. would anyone be kind enough to do it for me on their end?? i'll be so grateful! 🥲

[tame mica]

-colab

[azure marsh]

Suggestions for @subtle rover

All working colabs / spaces ☁️

• AICoverGen-NoWebUI, by Ardha Google Colab
• RVC Disconnected, Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• RVC-HF, by r3gf Huggingface Spaces
• AICoverGen, by r3gf Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces
• Advanced RVC Inference, by neuclya Huggingface Spaces

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, -local, -overtrain, -realtime, -rvc, -help

[tame mica]

use rvc disconnected for training and ai covergen for inference

[subtle rover]

k lemme try

[tame mica]

rvc disconnected is for training

[brittle wing]

Everytime I try to train a model in RVC V2, it gives me this error:
RuntimeError: The expanded size of the tensor (12800) must match the existing size (0) at non-singleton dimension 1.

[rare goblet]

Ayo? @brittle wing level 1 !!!

[brittle wing]

nvm fixed it

[brave eagle]

Hello! I was wondering if someone could help me, I'm trying to use Easy GUI v2 for RVC2, but when I'm uploading an audio it doesn't show up in the dropdown menu even after refreshing. Any help would be much appreciated!

[solid pawn]

whats up, Im making a passion project, a remix of Cherry Lady but with ai generated vocals of a city legend. anybody wants to use his expertise to help me?

[violet heron]

The GUI is broken

On free tiers

[brave eagle]

So what can i do? Run it locally?

[violet heron]

Yes

Or

Use HuggingFace spaces

But it’s a bit slow

https://huggingface.co/spaces/TheStinger/Ilaria_RVC

[molten pecan]

What do you need help with?

[brave eagle]

Thanks, I really didn't want to run it locally, so I will check the HuggingFace spaces out

[rare goblet]

Ayo? @brave eagle level 1 !!!

[subtle rover]

is this an error

there are 207 wav files in my dataset

i used 5 songs

too much?

[solid pawn]

the voice model

[rare goblet]

Ayo? @solid pawn level 1 !!!

[molten pecan]

What about it?

Wanna search one?

[solid pawn]

making it i dont know how to start as im a begginer in this field of ai

[molten pecan]

There are a few guides here

-rvc

[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

[molten pecan]

Also on #1159513888199540817

[solid pawn]

thanks you polish cuz i can explain you what it is im making

my bad

[molten pecan]

@half cove I'm polish from now on

[solid pawn]

just saw the argentina in the nick

[molten pecan]

Yeah 😭

[half cove]

💅

[molten pecan]

That's what I thought too

[half cove]

polish is polish Polish is 🇵🇱

[solid pawn]

yeah but even so in my small town theres this maybe alcoholic 17 year old guy my homie and for the last month every time we turns up he sings chery lady but he sing cherry cherry beer and just keeps saying beer beer and i just need to make a remix of that as it is such a stupid song but so catchy and when you drunk and singing that shit POLAND

[half cove]

😭

[molten pecan]

Most coherent sentence in ai hub

[molten pecan]

I'll help you then

[half cove]

na i understand it

[molten pecan]

Search for the voice model you want to use in #1159289700490694666 or weights.gg

[subtle rover]

contiue anyway?

[molten pecan]

@violet heron contiue anyway?

Or ransomware?

[subtle rover]

this is on google colab

[half cove]

which one?

[solid pawn]

can i make a custom one

?

[solid pawn]

or i need to use arleady made ones

[molten pecan]

Of course

[half cove]

is it webui or no?

[molten pecan]

You can do both

[subtle rover]

yup its google

[half cove]

send link

rq

[molten pecan]

Create one you want or use one that already exists

[solid pawn]

but i need to add it to #1159289738314919936 first?

[molten pecan]

Add what?

[subtle rover]

https://colab.research.google.com/drive/1mrHh7uqWp43F5jJp6MOXPvumMho3Lj4X?authuser=1#scrollTo=F6zbsuNs6xZF

[solid pawn]

cuz to make a model on #1159289700490694666 i need permision

[rare goblet]

Ayo? @solid pawn level 2 !!!

[molten pecan]

Ah yeah

[half cove]

hmm

[molten pecan]

I know nothing about making models from scratch so I can't help you with that

I think you should start by requesting the model maker role there

[half cove]

it uses gui but on ngrok

let me try this

[subtle rover]

no idea what that means

but thx!!

lol

[half cove]

so like

[tame mica]

🔥

[half cove]

colab banned webuis for free users

but ngrok works somtimes

like for tenserboard

nope it dont work

f

use

-colab

[azure marsh]

All working colabs / spaces ☁️

• AICoverGen-NoWebUI, by Ardha Google Colab
• RVC Disconnected, Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• RVC-HF, by r3gf Huggingface Spaces
• AICoverGen, by r3gf Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces
• Advanced RVC Inference, by neuclya Huggingface Spaces

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, -local, -overtrain, -realtime, -rvc, -help

[half cove]

🙏

[tame mica]

such a fire explanation Mr poopmaster

🔥

[subtle rover]

thank u

❤️

[half cove]

use one of these

[subtle rover]

i triedthe regular one but i dont have a compatible gui

thats why i decidd to use colab

decided*

[half cove]

oh shit there is a huggingface spaces aicovergen now?

[subtle rover]

is this code normal?

[gaunt ruin]

so correct me if I'm wrong, but basically, thanks to the google colab fuckup, the online alternatives are:

RVC Disconnected = training
Ilaria = model usage

[subtle rover]

can i use these if i dont have a compatible gui?

[half cove]

yep

[gaunt ruin]

thanks

[half cove]

there are some others but they are basically the same

[half cove]

which one is this?

[subtle rover]

this is colab lol

[gaunt ruin]

do you recomend those to me? or are there better alternatives?

[half cove]

hmm

[subtle rover]

nvm u were right

[gaunt ruin]

F

[subtle rover]

ill just use an alternative

sigh

[half cove]

https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw this might be easier for some people for training

[half cove]

which one was this?

[subtle rover]

so i can use rvc disconnected if i dont have a compatible gui?

[subtle rover]

colab

[half cove]

send link

[subtle rover]

its the same one i sent earlier

u said only paid users can do it

[half cove]

i said that one doesnt work 😭

[gaunt ruin]

do you have a tutorial for it? I've never used easygui without a user interface before

[subtle rover]

yeah loll i was just trying 😭
\

[half cove]

#1159513888199540817

let me find the other one rq

[gaunt ruin]

for model usage?

[half cove]

this guide for this onehttps://docs.aihub.wtf/guide-to-create-a-model/model-training-colab

[gaunt ruin]

thanks, and also, do you suggest ilaria is actually the best at model usage?

[half cove]

i mean any of them work

[gaunt ruin]

I'm saying cus the other day, ilaria took like 10 minutes to process a single audio

[gaunt ruin]

which others are there?

[half cove]

you could try ai cover gen ui but i dont know if that still works

#✨│ai-help message

[gaunt ruin]

I'll give it a try later on, thanks

[subtle rover]

rvc disconected is also a colab link
are you sureitll work?

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

[half cove]

it doesnt have webui

so it will work

colab only banned webui for free users

[subtle rover]

even with an incompatible gui?

sorry if this is a dumb qn lol

[half cove]

it doesnt use gui

[subtle rover]

im ne to this

okay

thanks

ill give it a try

[half cove]

that ones for training btw

are you making a model or making a cover?

[subtle rover]

making a model first

[half cove]

then thats the right one

[subtle rover]

what to use after this one

to make the cover?

[half cove]

you can use huggingface spaces

[subtle rover]

thank u poopbert ur a legend 🙏

[half cove]

you need to upload to huggingface to use

-upload

[azure marsh]

huggingface.co upload Guides

🇺🇸 English guide

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

🇮🇳 Hindi guide

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

🇹🇷 Turkish guide

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

[half cove]

here is the guide for that

[subtle rover]

okay
thank u!!!!🙏

[rare goblet]

Ayo? @subtle rover level 5 !!!

[subtle rover]

what does this ean

mean

[half cove]

hmm?

did you connect to drive?

[half cove]

why was the message deleted?

[eternal galleon]

Running go-web.bat

C:\RVC-beta0717>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897
Traceback (most recent call last):
  File "C:\RVC-beta0717\infer-web.py", line 23, in <module>
    import gradio as gr
  File "C:\RVC-beta0717\runtime\lib\site-packages\gradio\__init__.py", line 3, in <module>
    import gradio.components as components
  File "C:\RVC-beta0717\runtime\lib\site-packages\gradio\components.py", line 24, in <module>
    import matplotlib.figure
  File "C:\Users\Owner\AppData\Roaming\Python\Python39\site-packages\matplotlib\figure.py", line 43, in <module>
    from matplotlib import _blocking_input, backend_bases, _docstring, projections
  File "C:\Users\Owner\AppData\Roaming\Python\Python39\site-packages\matplotlib\projections\__init__.py", line 58, in <module>
    from mpl_toolkits.mplot3d import Axes3D
  File "C:\RVC-beta0717\runtime\lib\site-packages\mpl_toolkits\mplot3d\__init__.py", line 1, in <module>
    from .axes3d import Axes3D
  File "C:\RVC-beta0717\runtime\lib\site-packages\mpl_toolkits\mplot3d\axes3d.py", line 45, in <module>
    class Axes3D(Axes):
  File "C:\RVC-beta0717\runtime\lib\site-packages\mpl_toolkits\mplot3d\axes3d.py", line 718, in Axes3D
    set_zticklabels = _axis_method_wrapper(
  File "C:\Users\Owner\AppData\Roaming\Python\Python39\site-packages\matplotlib\axes\_base.py", line 56, in __init__
    doc = inspect.getdoc(getattr(maxis.Axis, method_name))
AttributeError: type object 'Axis' has no attribute '_set_ticklabels'

C:\RVC-beta0717>pause
Press any key to continue . . .```

[subtle rover]

i dont wanna bother yall too much lol

trying to figure it out on my own

[eternal galleon]

Similar error in any other batch files I try
go-realtime-gui_new.bat

C:\RVC-beta0717>.\runtime\python.exe    .\gui_v2.py
1.0
2023-11-04 19:31:59 | INFO | fairseq.tasks.text_to_speech | Please install tensorboardX: pip install tensorboardX
Traceback (most recent call last):
  File "<frozen __main__>", line 3, in <module>
  File "<frozen temp>", line 41, in <module>
  File "gui_package.py", line 24, in init gui_package
  File "C:\RVC-beta0717\rvc_for_realtime.py", line 3, in <module>
    from lib.infer_pack.models import (
ModuleNotFoundError: No module named 'lib.infer_pack'

tf is wrong here

[rare goblet]

Ayo? @eternal galleon level 1 !!!

[eternal galleon]

[subtle rover]

can the files be mp3 in rvc disconnected?

[half cove]

yes

[subtle rover]

thx

[vocal cipher]

i can make rvc talk but i cant make it sing

[copper pebble]

waw its betta

[eternal galleon]

how might i get around this?

[radiant mesa]

i use a rvc gui to make covers

very ez

[eternal galleon]

the rvc gui isn't opening, thats the point dude

https://tenor.com/bR6o2.gif

skeleton embed

[half cove]

#🔍│help-w-okada

[eternal galleon]

??

what am i doing in that channel?

[neat spade]

hello guys, I have this error when I try to start audio
RVC object has noattribute 'tgt_sr'

[rare goblet]

Ayo? @neat spade level 1 !!!

[neat spade]

what is the problem ?

[neat meteor]

I have a question, maybe a stupid one but I'm new to the topic, there is a free program that will turn longer text (several pages) into speech using my own voice model (I already have the model ready)

[eternal galleon]

no i'm trying to use the webgui

but both error

[half cove]

hmmmm

i might be dumb

[eternal galleon]

if i can make the realtime work then great i guess

[neat spade]

I had an error earlier that came from my settings (I hadn't checked harvest or crepe). Maybe this one comes from the settings again?

Do you know if there's an updated tutorial, as the one I followed seems to work on the old gui

[agile ridge]

i purchased colab pro
and using rvc disconnected colab. does it keep all my training work even i get close my browser? or i need to stay it on?

is it possible to save all my training checkpoitns to my drive automatically ?

this notebook does not have autosave functionality due to the massive amount of underlying stress that RVC's autosave features places on Colab to Drive communication

nvm found it

thanks!

[azure marsh]

All working colabs / spaces ☁️

• AICoverGen-NoWebUI, by Ardha Google Colab
• RVC Disconnected, Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• RVC-HF, by r3gf Huggingface Spaces
• AICoverGen, by r3gf Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces
• Advanced RVC Inference, by neuclya Huggingface Spaces

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, -local, -overtrain, -realtime, -rvc, -help

[honest pond]

probaby if you have the colab pro, the appilo stel work good for you or magnio

[subtle rover]

what to do after this step is done?

[tame mica]

scroll down, click export and also train index

[subtle rover]

thx!

[half cove]

Export after training index

@karmic oxide

[silver snow]

@karmic oxide

[tame mica]

raven fr

[daring raft]

Ayo i have a question
In April i had better accent results than nowdays, why?
I am a Serbian so my English accent is shit when singing, but in April I used RVC 2 some models and it was perfect but nowdays, with Replay app, it sounds more like mine accent which is not so great

[lean agate]

someone knows why this program exactly

goes well while I'm playing overwatch

and randomly it starts going really lag

use

nono

there you go what

I tried everything

even If I close the game

it still going lag

but I restart and it goes well, maybe it start lagin 5 mins after

it's totaly random

listen

it goes well

sometimes nothing happens but this days I'm having so many issues

if you have anything helpfull to say instead of limit fps, go ahead

[rare goblet]

Ayo? @lean agate level 1 !!!

[lean agate]

omg

really?

with that one I can't use it

it goes to 23k of delay

with this one it goes well

yes?

okey I will dowload it

[subtle rover]

its been training for an hour and 44 mins.. how much time does it typically take?

[wide verge]

I know it's a stupid question, but where's a button to record my microphone and put the voice through AI? 💀

[subtle rover]

total epochs is 500. is that too much?

138 files

[wispy burrow]

Nowhere.
That RVC does not have that.

[subtle rover]

20 mins

[barren swift]

there is a real-time RVC client you can use

which changes your voice in real time with a few seconds of delay

or you can use something like audacity to record your microphone and use that as the "vocals"

also

while i'm here

on the local installation of RVC which file do I open to start it

I have it installed but I've forgotten which file I had to open to use it

[austere echo]

is there any online website to train a model because my gpu isn't strong enough

[barren swift]

yes

[subtle rover]

rvc disconnected

[barren swift]

there's a google colab rvc disconnected

[austere echo]

oh ok i'll look for it thanks

[brittle wing]

помогите Help

[austere echo]

ok i need help with rvc disconnected

[barren swift]

when it says to "run" something click the white arrow button on the left side

run the dependencies first, wait for it to finish and it should ask for access to your google drive files

give it access

[austere echo]

hhuh

[barren swift]

you're also going to need to put your dataset into a zip file

[austere echo]

wait wait

i runned the cell where it says run me first

[barren swift]

okay

[rare goblet]

Ayo? @barren swift level 2 !!!

[barren swift]

so just wait for that to finish

[austere echo]

it's done

[barren swift]

.

[austere echo]

so like

the audio file?

[rare goblet]

Ayo? @austere echo level 3 !!!

[barren swift]

make a folder on the root of your google drive called "rvcDisconnected"

[barren swift]

yes, all of them

[austere echo]

oh uh

[barren swift]

put them in a zip and name it something

[austere echo]

i only have a 10minute single audio file

[barren swift]

i'm not sure if that will work any differently than separate audio files

but I would recommend cutting it into smaller files first

just in case

[austere echo]

how many?

wait like

i have a 30 minutes audio files with many songs

[barren swift]

alright

[austere echo]

i can cut every songs

[barren swift]

did you separate the vocals first

[austere echo]

yes....

[barren swift]

if you're training off songs you should separate the vocals before you train a model with them

alright

[austere echo]

it took 13 hours

[barren swift]

put that audio file into a zip

make a folder on the root of your google drive called "rvcDisconnected"

and put the zip file in there

[austere echo]

ok so

ok wait i'll do that

[brittle wing]

hey which library could i use in python to generate tts with a .index and .pth file?

[austere echo]

should i spearate every songs in multiple audio files?

[barren swift]

nah

[austere echo]

oh btw is it bad if they are like

backing vocals too?

[barren swift]

that might have an effect on it yeah

[austere echo]

i mean

would it like tune it

[barren swift]

it'll most likely try and fuse the main vocals with the background vocals

assuming both are the same person

and that could really throw off your model

[austere echo]

oh

well

uhhhh

[barren swift]

try and cut them out as best you can

[austere echo]

i'm not waiting another 13 hours

[barren swift]

you don't need to

[austere echo]

wait

[barren swift]

open the file up in something like audacity

[austere echo]

is there any ways to like

take only the main vocals

i can only do backing vocals on audacity

[barren swift]

i don't really know

but if there was, it'd require you to separate the vocals again

[austere echo]

i mean does backing vocals really matter?

even on the 10 minutes file

[barren swift]

yeah

[austere echo]

well

uuuuuhhh

[barren swift]

you can manually cut out the background vocals

[austere echo]

how

[barren swift]

by doing it manually

[austere echo]

...

how

[barren swift]

in audacity

find any part of the file where there is background singers

but not a main singer

highlight it

and replace with silence

[austere echo]

what about the main voice

ooh

[barren swift]

that's what you want to keep

[austere echo]

ok

ok

so like

when ever there is backing vocals

i delete it

[barren swift]

highlight it and press generate silence

it's in the dropdown menus at the top of the window

under generate

[austere echo]

ok ok

but like there are backing vocals and main vocals at the same time

and i really dont want to wait 13 hours

[barren swift]

there's not much that can be done about that

but rvc can usually weed out stuff like that

in training

not if the background vocals are just on their own though

[austere echo]

it's like

backing vocals and main vocals at the same time

harmony

[barren swift]

i'm not really sure

[austere echo]

i mean

it wont matter than much

[barren swift]

see if you can find other sources of that person's voice

[austere echo]

maybe

well

it's kanade yoisaki

[barren swift]

who is it anyway

mk

[austere echo]

from the game project sekai

and there is already a model with her

but i'm sure that the voice used is her talking voice

i need the singing voice

[north halo]

Okay I downloaded the rvc thing now what? This guy said to drag it to my the main root of my hardrive?

[barren swift]

training a model on a talking voice doesn't mean it isn't going to be able to sing

try using that model and see how it turns out

[austere echo]

i know

but it doesn"t sound like her singing

i want to make a realistic thing

but wait

i'll just separate each songs

and remove the backing vocals

it may take the whole night but

doesn"t matter

she's my fav character

[barren swift]

alright man

[solid sand]

How do download audio from AICoverGen UI? It's just giving me errors

[hushed quartz]

for models without .index files where is the default .index file located for rvc v2?

or can i not run the model without the original index file it should come with

aight thanks

[wind crow]

can i use the voice changer on discord

[subtle rover]

how many epochs should i set it to?

its taking way too long w 500

[barren swift]

trained a model earlier today, imported it back into colab (rvc disconnected) to generate the index file and got this error

[rare goblet]

Ayo? @barren swift level 3 !!!

[barren swift]

i ran the dependencies first, then set training info, imported the model with the number on the D/G file and all of that worked

I ran feature extraction first because it told me to do that before I tried to generate an index

I plan to use this in the realtime rvc thing so I need an index and the index for the previous version of this model I made didn't work very well with it

so i'm stuck on being able to get an index

[wary crystal]

Might be a very dumb question, why downloading model on RVC hugging face said error occur

[radiant mesa]

Gah

Don’t know where else to post this

But how long does this usually take

[rare goblet]

Ayo? @radiant mesa level 5 !!!

[radiant mesa]

alright whats a easier ai model trainer that you could use locally

mangio dosent like my computer as it seems

[north halo]

Are you supposed to train the Ai models before you use them?

[radiant mesa]

yeah

i wanted to know if theres a easy local training thing like rvcv2 disconnected

because i Cannot connect to GPU backend'd all on the online okada

[north halo]

I meant like how those models that are posted here. Am I supposed to take those and "train" them to fit my voice or something? Or do I just use them how they come?

[brittle wing]

im training with colab and the google doc im using to help says

In Google Drive, locate "rvcDisconnected* -> "ModelName" folder.
Download important files:
".pth"
"added.index"
Move downloaded files to your local machine, and place them in one folder.
zip the files.

But in drive theres no stuff like that, idk if im being dumb or smth, never really used drive before and this is my first time training in colab

[radiant mesa]

i think you create the foldier

[wary crystal]

So it's not like the old AI HUB, you supposed to train it first before using it?

[rare goblet]

Ayo? @wary crystal level 1 !!!

[brittle wing]

sorry i didnt give more info but ive already trained it if that makes sense, using this doc if that helps https://docs.google.com/document/d/1XuxQYiqEhYrdYeCZRRLrmV_ciMKo0bV-jTCGHu_-5Cc/edit?pli=1

[radiant mesa]

also dolab

[brittle wing]

dolab

[wary crystal]

will this cause a problem

[radiant mesa]

rvcdisconnected?

[wary crystal]

Maybe I've done something wrong?

[rare goblet]

Ayo? @raven elbow level 1 !!!

[brittle wing]

Fix:
Upgrade torch and torchaudio to 2.1.0

pip3.8 install --upgrade torch==2.1.0 torchaudio==2.1.0

and run server using

PYTORCH_ENABLE_MPS_FALLBACK=1 PYTORCH_MPS_HIGH_WATERMARK_RATIO=0.0 python3.8 infer-web.py --pycmd python3.8

[pine mango]

@radiant mesa i got this dataset in my drive rn

after i connect drive

[radiant mesa]

so for rvcdisconnected i think you put the dataset in the folder

the rvcdisconnected folder

[pine mango]

should i put the path of it here?

[radiant mesa]

yeah

well

only the name

of the dataset

and .zip

[pine mango]

so i gotta make a rvc folder?

for the ai to search in?

[radiant mesa]

yeah

[pine mango]

this good?

[radiant mesa]

yeah

[pine mango]

and upload the dataset zip file?

this is the data set

[radiant mesa]

yeah upload it in there

[pine mango]

alr its done uploading

now im connect colab to drive

expriment name can be anything right?

[radiant mesa]

yeah

[pine mango]

apox how long can it take?

ahhhhh

[wispy burrow]

I'd say avoid any spaces and special characters.
Not sure if it'll cause issues on Colab, but it does locally.

[pine mango]

alr bet tysm

well it did not work

it just cant find the file

but i did put it in there

@radiant mesa did this happen to you?

[radiant mesa]

if it does finish with no errors

refresh the google drive page

[pine mango]

im redo and see if anything new happens

[wispy burrow]

There is a guide for RVCDisconnected.
#1160926856548200531
Just in case you haven't read it yet.

[pine mango]

omg bro

[rare goblet]

Ayo? @pine mango level 5 !!!

[pine mango]

its there

bruh

im read it

[dire dagger]

Hello, with my hyperx mic, people hear themselves in games as ai because of an echo, and when i use RVC, i have that robotic tone that my friend doesn't have, is it because i need to change my voice meeter parameters ?

[pine mango]

python trainset_preprocess_pipeline_print.py "/content/dataset/labara" 40000 2 "/content/Mangio-RVC-Fork/logs/labara" 1
python3: can't open file '/content/trainset_preprocess_pipeline_print.py': [Errno 2] No such file or directory error 1

python extract_f0_print.py "/content/Mangio-RVC-Fork/logs/labara" 2 rmvpe 64
python3: can't open file '/content/extract_f0_print.py': [Errno 2] No such file or directory
python extract_feature_print.py device 1 0 0 "/content/Mangio-RVC-Fork/logs/labara" v2
python3: can't open file '/content/extract_feature_print.py': [Errno 2] No such file or directory error 2

[mystic ferry]

Why does it just not do anything? (Running locally)

oh I can't send imgs

But I swtiched to v2 and then it just does nothing, have been waiting for 400sec twice

In addition, how could I host a public link?

[sand iris]

The process for that is a little funny

[mystic ferry]

I already found it from your previous msgs lol

am doing that rn

but the other thing tho

[sand iris]

Not sure, training is... Unreliable at times

[mystic ferry]

Like it's been 800 secs and not doing anything

[sand iris]

Which GPU and how long is the dataset?

[mystic ferry]

No I mean I think it's just not communicating between web and app

dataset is like 8 clips of around 15 sec

[sand iris]

Oh, yeah, check command line logs

The web page doesn't display the progress

[mystic ferry]

it stops at "To create a public link, set 'share=True' in 'launch()'."

[rare goblet]

Ayo? @mystic ferry level 1 !!!

[sand iris]

It should be displaying other stuff though

[mystic ferry]

oh now it lets me send screenshot

[midnight needle]

Hey guys what app do I use to make ai voice say something bizarre :))

Already got the model downloaded

I keep finding voice changers :((

[brazen schooner]

I am doing rvmpe today. And I'm currently on the feature extraction. It's taking too long. There is no movement. I don't understand why this is happening

[mystic ferry]

Oh is it meant to show me these logs?

[brazen schooner]

I also attempted doing mangio crepe. It's also been stuck like this for a while

I don't know why it takes forever to do the feature extraction. Please someone help. My gpu is 12 vram. Nvidia Geforce RTX 3060.

[mystic ferry]

Hey uh I don't know if it's related but same GPU, absolutely no response

[brazen schooner]

I can't pass this to actually start the actual training process. It's taking too long. It's been hours

[brazen schooner]

Idk why we are struggling

[sand iris]

There's multiple software for that

But most people use Applio-RVC-Fork

Really not sure what's up with it though

wait @brazen schooner does your user name happen to contain spaces ?

If yes, it should've errored by now so I suppose it's not the case

[mystic ferry]

Alright got this to work

But uh what do I do about it not responding?

[midnight needle]

Is it beginner friendly

[brazen schooner]

But they say applio is not good and isn't successful

[brazen schooner]

I don't think the username has a problem

[mystic ferry]

If you don't mind me asking, are you using the mangio rvc fork?

[brazen schooner]

Yes I am

İt's what they told me to use

[mystic ferry]

How do you start it?

Like this is asking for myself

because maybe I start it wrong?

Do I just run go-web.bat?

[sand iris]

Wait did you get the training archive or the infer one

[mystic ferry]

training

[sand iris]

I don't get why it isn't outputting anything to cmd though

[ornate tartan]

guys who's very good at settings I kinda need some help and some explaining

[proper shale]

What can we help you with?

[ornate tartan]

I wanna know how to sync my voice to the AI, Mine sometimes glitches idk why

[proper shale]

Right now it isn't possible to use this in sync

Because it's a heavy program to run

[ornate tartan]

I mean how to set the best settings

Is there a guide to adjust the settings

[proper shale]

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

should help you

[ornate tartan]

Thanks a lot

[proper shale]

np ;)

it has specific settings for series of cards which is really nice

[mystic ferry]

Maybe I'm opening the wrong file?

Am I meant to open go-web.bat?

[sand iris]

yeah

[mystic ferry]

hm

[sand iris]

I really don't know what's up with it, never saw it do that

[mystic ferry]

this is all I get in the cmd prompt

[rare goblet]

Ayo? @mystic ferry level 2 !!!

[mystic ferry]

maybe because windows 11?

[sand iris]

Should work even then

[mystic ferry]

Yeah I don't know what else it could be

[sand iris]

I guess you could try Retrieval-Based-Voice-Conversion-WebUI (RVC-WebUI) or Applio

[mystic ferry]

Hm

[sand iris]

It's always like that lol, switch projects until one works

[mystic ferry]

I did have some errors during installing and had to manually extract some files

Is compile.bat meant to do something?

Because I thought maybe open that up and realised I don't have a python compiler or anything

not sure if that's needed

Oh I got something just now

[sand iris]

Great we went from nothing to... One of the obscure errors

[mystic ferry]

Right

I'm just reading it and don't understand anything lmao

line 347

line 220

[sand iris]

Yeah you aren't supposed to delve into the code, it should work, but sometimes it just doesn't

[mystic ferry]

oh

[sand iris]

I'm sorry but I don't think I can be helpful there

[mystic ferry]

Alright, thanks

ill try just reinstalling this one file

since it seems to ve just it

oh and one other

IT WORKS!

If anyone else has this issue, you don't need to reinstall the whole thing, I just reextracted ...\Mangio-RVC-v23.7.0\runtime\Lib\site-packages

[sand iris]

It didn't extra properly?

[mystic ferry]

guess so

into another error

How am I meant to have the datasets folder?

[rare goblet]

Ayo? @mystic ferry level 3 !!!

[mystic ferry]

like

[sand iris]

Bunch of audio files inside of it

[mystic ferry]

do I place a zip of audio files in there or audio files

[sand iris]

Format shouldn't matter

Oh yeah, you shouldn't have them zipped on local

[mystic ferry]

oh nvm mine went datasets/insert/insert

when I unzipped

lol

now it seems to work

is this good?

[sand iris]

Been a while since I used mangio, so someone correct me if I'm wrong.

It should display the path of every audio file of the dataset followed by "Success"

[mystic ferry]

I can quickly try on collab

brb

it shows this though

collab still loading

[copper pebble]

what colab r u running?

[mystic ferry]

rvc crepe v2

[austere echo]

i need help with rvc v2 disconnected

[copper pebble]

link pls

[copper pebble]

yh

[mystic ferry]

https://colab.research.google.com/drive/1TU-kkQWVf-PLO_hSa2QCMZS1XF5xVHqs?usp=sharing

[austere echo]

i got an error

[mystic ferry]

looks the same

[copper pebble]

thats an old one

use rvc disconnected

[mystic ferry]

I mean I got a local one now

[copper pebble]

https://discord.com/channels/1159260121998827560/1160926856548200531

[copper pebble]

u just started the "depencies cell right?"

[copper pebble]

if you didnt i suggest to use our guide its pretty useful

[patent shore]

Friends, tell me what to do if, while talking in the background, a woman says “Trial”

[austere echo]

yes it's on the 2nd cell of depencies

[lean agate]

Somebody knows why mangio rvc or all the ai programs start lagging while I play a game even If I lock fps, put lower graphics or close the game? when all the components are going well? it looks like a problem of the program because it goes well but then randomly start having issues

[copper pebble]

start all depencies at the same time

[mystic ferry]

How do I start up tensor board on local?

[austere echo]

did it

but i get the same error*

[mystic ferry]

if I launch the .exe or .bat it just closes

[copper pebble]

like this??

[austere echo]

yes

maybe it's because of the zip file

[copper pebble]

then disconnect the runtime and create a new one

[copper pebble]

gpu??

[lean agate]

rtx 3050 ti

[copper pebble]

mhh that should work

[lean agate]

ik

and it does

[copper pebble]

i mean recommended its 3060 but 3050 its goos aswell

[lean agate]

but for no reason, the program start failing

[copper pebble]

it doenst change that much

[lean agate]

yeah

the problem is

it works well

but randomly start going bad

even if the fps of the game are the same

the gpu temp is the same

[copper pebble]

try without playing

[lean agate]

then it goes ok

[copper pebble]

thats the only way i think ://

[lean agate]

but

playing also goes ok

no problems no delay

[copper pebble]

Suddenly it fails

[lean agate]

also when it starts lagging, when I close the game continue lagging

[copper pebble]

.

if you didnt tried without playing you should

[lean agate]

but I wanna use it playing too

[lean agate]

I did

but shouldn't be a problem

if I can do both things without problem

sometimes it doesn't lag

but sometimes yes

when everything is the same

so is not a problem of the pc or the game

it's the program and I wanna know if there is any solucion

[copper pebble]

mhhh im not very familiar with local programs and this looks like a strange problem so

@zinc anchor scusa per il tag non è che sai qualcosa su questo problema??

@austere echo fixed?

[austere echo]

yes thank you

[rare goblet]

Ayo? @austere echo level 4 !!!

[tranquil cliff]

hiiii

[tranquil cliff]

aiuto?

[copper pebble]

vado afk vedi se riesci tu

[tranquil cliff]

ok amore

coetaneo mio

[tranquil cliff]

what's the problem?

[honest pond]

you probably have a bottleneck, what processor do you have?

or you have the amd version instead of nvidia if you talk about w-koda

[sand iris]

the "amd" version should GPU accelerate on any GPU

Either way

[plain drift]

is there an ai where we can train our custom language for translation...not looking for heygen or elevellabs..Im trying to train a custom language ?

[surreal summit]

Im just looking for a TTS site that can use RVC V2 models

[subtle rover]

what should i select as the epoch limit and how long should my dataset be? i dont want to wait for too long... yesterday it was at 500 epochs, my data set was 20 mins long and it took more than 8 hours and then froze

rvc disconnected ^

[copper pebble]

rvc cant train long datasets's model

but you can try 400 epochs for 20 minutes dataset

heres a tab btw

[subtle rover]

can i do 200?

[subtle rover]

or is it too little

thanks!!!

will it turn out good if my dataset is only 5 mins long?

[copper pebble]

yeah

it also depends on how good are the audios of the dataset

[subtle rover]

okay

thx!!!

[copper pebble]

ping me for any problems or if you will need someone to test the model once finished

[hot lagoon]

hose dose uvr5 Ensemble mode work??

[subtle rover]

are these settings fine?

should i switch to harvest?

[hot violet]

it's good thought

[half cove]

Rmvpe ftw

[fickle mason]

Do i do anything here "Set Training Variables" https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS

nvm thats my same question

[austere echo]

i have a question about rvc disconnected

with the pth file

which one should i use

[copper pebble]

.pth and .index (only the added one)

the last 2

[austere echo]

ok thank you

[copper pebble]

yw

[astral jungle]

Yoo

[fickle mason]

how did u get them to download. when i download it from files in rvc disconnected its not downloading

[fickle mason]

herp me

i cant donwload the D_233333. pth and G_ one

[half cove]

It saves to Google drive

[fickle mason]

its not

i did it four times and its not saving anywhere

it shows it on my logs - experiemnt_name

also i dont gotta a index file

[copper pebble]

did u run the last two cells??

if u did and the problem persist

[fickle mason]

uh..no i didnt

ur smart

[copper pebble]

hahah

[fickle mason]

thanks helper

[copper pebble]

yw

[fickle mason]

i ran feature extraction and then index training, it says no features exist for this model yet. I need for da index right

[copper pebble]

yh

did u follow our guide?

[fickle mason]

no..

can u link it lol

[copper pebble]

lemme link it for ya

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

this will be helpful

[fickle mason]

thanks so much

[copper pebble]

ping me for any problems

[fickle mason]

i have many.. index training keep giving me red.

[copper pebble]

r u on pc??

[fickle mason]

yuh

[copper pebble]

mhh

[fickle mason]

In Google Drive, locate "rvcDisconnected* -> "ModelName" folder. I dont go this i got my model as a zip

[copper pebble]

yeah thats what it means

[fickle mason]

yuh when i open it theres an error in it

[copper pebble]

can you send me some screens pls??

[fickle mason]

i wish... but discord isnt letting me for sum reason

when i click print screen nothing happens lol

i also tried snip it

im cursed

aye it worked

nbm

[copper pebble]

maybe the better thing to do its restart all

disconnect the runtime and redo all

make sure to save like this next time

[fickle mason]

kay https://cdn.discordapp.com/emojis/1125604910113759353.webp?size=48&name=chorei&quality=lossless 4th times a charm ig

[copper pebble]

embed failed lmao

[fickle mason]

YES fail

i hate it here

HAHA typing failed

[copper pebble]

follow every guide's step

[copper pebble]

ahah

[copper pebble]

maybe you want to change ur account

idk how many hours of runtime you have left

[fickle mason]

eh

so basically im redoing the pth files too

this is normal right @copper pebble

or should i nuke the file

[copper pebble]

that usually happens when you dont disconnect the runtime

make sure to do it

"runtime" button above

"disconnect and delete data"

or something like that

[fickle mason]

ohhh that makes more sense. I used to just refresh it

https://cdn.discordapp.com/emojis/633986571191844884.webp?size=48&name=thumbsupkurisu&quality=lossless

[copper pebble]

whats wrong with your embeds

[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

[fickle mason]

i blame vancord

im still getting the same ugly ass error

[copper pebble]

wtf

wth does this ugly ass bastard pops up

[fickle mason]

idk https://cdn.discordapp.com/emojis/1159569588435046512.gif?size=48&name=lfgweird&quality=lossless

idc if its failed

[copper pebble]

if u followed the guide step by step this shouldnt happen

[fickle mason]

are u accusing me?

[copper pebble]

nah

[fickle mason]

am i missing a file

specifically this one 3_feature768

cuz thats what its looking for right

[copper pebble]

dunno what that shit is

[fickle mason]

its what the logs says

[copper pebble]

can you send some screen of the colab

[fickle mason]

okay

[copper pebble]

from the 1st to the last cell

[fickle mason]

wait i kinda just realize i ran this cell " Download Pretrained Models"

[copper pebble]

oh bro

[hearty jasper]

Which is the best f0 method beside rvmpe?

[copper pebble]

thats for resumes

[fickle mason]

oh

[copper pebble]

maybe its that the problem

:((

[mint umbra]

What's this guys "OSError: [Errno 38] Function not implemented"

[fickle mason]

ill just rredo it lol 5th times a charm

[copper pebble]

please make sure to read the cell

[fickle mason]

kay

u were right to accuse me

[finite galleon]

For training or infer

[copper pebble]

admiring your patience

[fickle mason]

i forogt about the doc until last minute

[fickle mason]

[hearty jasper]

I don't really know but I wanted to make a cover

[rare goblet]

Ayo? @hearty jasper level 1 !!!

[copper pebble]

mangio-crepe

[finite galleon]

Mangio-crepe or fcpe

[hearty jasper]

Ah I don't have that

[finite galleon]

Just use rmvpe

[hearty jasper]

I'm using rvc gui

[finite galleon]

It aint that different

Once its stacked with the instrumental its going to not matter if its robotic or not

Unless its some robotic screech

[hearty jasper]

Should I use crepe or crepe-tiny?

[finite galleon]

Crepe

Crepe tiny is for faster speed and crepe is for more quality

[hearty jasper]

I see, thanks

[copper pebble]

its a lil it old

did u tried ilaria rvc?