✨│ai-help

[rough cloud]

for example jett500Epoch

[rare goblet]

Ayo? @rough cloud level 1 !!!

[rough cloud]

pehh

[copper pebble]

yh that should be good

[obtuse stream]

oke

what does it mean that not all of them have been maximised?

[copper pebble]

mhhh im not sure if i knoe this

btw its late here

so im about to get off

if u keep having problems that make training impossible use rvc disconnected colab

https://discord.com/channels/1159260121998827560/1160926856548200531 heres a guide

@obtuse stream

gn

[uneven charm]

Is it possible to resume training?
I reached 315 epochs before my computer restarted for some reason

[modern oxide]

Yes, you just type in the same name and settings and press every step of training

[uneven charm]

But how do I know it won't over write everything?

And do I have to change something about the G and D path?

[uneven charm]

And will there be any negative affect on the quality, in that case I should likely restart the whole process

[rare goblet]

Ayo? @uneven charm level 5 !!!

[modern oxide]

No

It'll just resume the training from the last saved checkpoint

[uneven charm]

Automatically?
Do I still need to go through the feature extract steps or directly to just training?
I don't think I need to do the index training through right? I just click train model?

[uneven charm]

Or do I click one click training?

[modern oxide]

Yes, you need to go through every step for it to resume the training

[uneven charm]

Alright thanks

[uneven charm]

So previously I set the epoch to 1000, but should I keep that one the same?

[modern oxide]

Yes

[void yoke]

The command line says the file is generated, but the output information does not say anything. Should I still continue?

[rare goblet]

Ayo? @void yoke level 1 !!!

[void yoke]

For Applio

[buoyant charm]

Im lowk confused with all the shit 💀. what link is it where I can convert an audio file into an artist

[molten pecan]

You have quite a few tools for that

[proper shale]

Check #1159949278270193734 and #1159513888199540817

[molten pecan]

Google colab, huggingface spaces, applio, gradio, mangio

[gloomy smelt]

I need help with applio

[proper shale]

?

[gloomy smelt]

So like, i don't know why out of nowhere the model i'm training right now took longer than any of the other ones and the fact that it's literally the same epochs as the ones i made before and they take around 30 minutes. but this time with all the same config it's taking more than one hour

[rare goblet]

Ayo? @gloomy smelt level 1 !!!

[gloomy smelt]

Epoch 1 started at 19:18 and currently it's 21:13 and it's on epoch 134

Should i cancel the model training and start again or is it common

Because it never took this long before

[proper shale]

Are you doing anything else while the model is training

[gloomy smelt]

Hmm

Not really

I mean i had fl studio open but that never did anything?

[proper shale]

Did you change batch_size

[gloomy smelt]

batch_size

What is that?

Everything is translated in spanish to me in applio, sorry

[modern oxide]

What GPU do you have?

[gloomy smelt]

AMD Ryzen 5 3400G with Radeon Vega Graphics 3.70 GHz

And i have a Nvidia graphics card rn

[modern oxide]

I need the Nvidia one

[gloomy smelt]

But not sure if that does anything actually lmfao

[gloomy smelt]

Sure

[proper shale]

It does

[gloomy smelt]

0 NVIDIA GeForce RTX 2070

[proper shale]

It's the one doing all the hard work

[gloomy smelt]

I see

[proper shale]

Wait what are the main settings in training (in Spanish)

[modern oxide]

And how long is your dataset?

[proper shale]

I think I can recognize which one is batch size

[proper shale]

30 mins

[modern oxide]

wut

Where did he mention it

[proper shale]

oh wait...

being dumb, sorry

[modern oxide]

Happens

[hot lagoon]

What to know if Anyone have good setings for uvr5 just for songs

[modern oxide]

So how long is your dataset?

[gloomy smelt]

Hi i'm back

Let me check

[modern oxide]

Mdx23c - deechodereverb - denoiser

[proper shale]

Should've used 2 for CPU processes but sure

[gloomy smelt]

Like, how long are the audios?

Not that heavy

[proper shale]

Oh, tamaño de lote

[modern oxide]

Yes

[gloomy smelt]

Well, i left the one that came by default

[hot lagoon]

witch overlap is good that dose not take long for the mdx part

[proper shale]

Anything else can glitch stuff afaik

[gloomy smelt]

The dataset has to be like no longer than one minute

[proper shale]

So idk

[gloomy smelt]

Which is really weird for me honestly

[proper shale]

Anyways yeah, 4 batch_size and a bigger dataset will be slower

[gloomy smelt]

Yeah dude it's literally 38 seconds long lol

[proper shale]

wha

[modern oxide]

Depends on your GPU, but usually anything >6 should be good in terms of quality

[gloomy smelt]

yeah

[modern oxide]

Try to restart it with the same dataset but putting 8 batch size

[gloomy smelt]

Okay

[rare goblet]

Ayo? @gloomy smelt level 2 !!!

[buoyant charm]

alr thanks

[gloomy smelt]

Tamaño de lote verdad?

also one thing which i think it's common but

My gpu is making a lot of noise

Like the fan is going crazy

[paper wave]

i trained a model but I can't get it to work. epochs is 250 because my dataset is 5 mins. but when i put it to use I hear no change in the voice, like, it's not covering

[gloomy smelt]

well noooot really

But it is kinda hot

Maybe it happends a lot when making models and i never really noticed it

[modern oxide]

Yeah, because it starts to work

It's good if it does that, because it means that it actually works

[hot lagoon]

I mean is there a uvr5 for clobe tho?

[paper wave]

not a beggar but help me please

[gloomy smelt]

i see

[modern oxide]

For what?

[gloomy smelt]

Anyways, i'll restart the training

8 in tamaño de lote right?

[proper shale]

are u using the right files for the model zip

[paper wave]

yes

[proper shale]

the pth from weights and the index from logs/model's folder

[hot lagoon]

google colab*

[modern oxide]

Yes

[paper wave]

yeah i can see the pth file

[proper shale]

just make sure its not the g/d files

[gloomy smelt]

Okay

[proper shale]

those r training files

[modern oxide]

Oh, no idea, I use uvr locally

[paper wave]

what's a g/d file

[paper wave]

o

[hot lagoon]

I there is a google colab for uvr5 but not sure where it is

[paper wave]

do I delete the training files

to get it to change

the voice

[proper shale]

nah nah no nerd

need*

💀

[paper wave]

lol

[modern oxide]

Nerd

[proper shale]

me (im a nerd)

okay uh

any other useful info?

[void yoke]

Is the time for training suppose to go above the processing time?

[proper shale]

because rn I'm not sure what the problem is

[paper wave]

this the model files

[proper shale]

yeah

[void yoke]

So.

[proper shale]

ok so, use starberries.pth

[void yoke]

What time will it be done?

[proper shale]

not G_2333333 and D_2333333

[void yoke]

Or should I manually stop the training.

[proper shale]

idk, it varies a lot

[paper wave]

is it bad that i use webui

[proper shale]

for what

[paper wave]

ai cover

[proper shale]

ai cover gen should be fine

maybe its the song ur pickin

[modern oxide]

No you just wait til it finishes

[proper shale]

heck even the link to the song could be the problem

[paper wave]

i think the problem is the ai chose g/d files insteda of starberries.pth

how do i prevent that

[proper shale]

did u zip the g/d files

[paper wave]

i'll do it now

[proper shale]

nah no need

[paper wave]

then what

[proper shale]

you need to only zip the starberries.pth and the index

and upload it somewhere (maybe Huggingface)

[paper wave]

ok

[proper shale]

-huggingface

[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

[paper wave]

but can i keep it in my laptoop

[proper shale]

theres how to upload ur models

[proper shale]

yea sure

[paper wave]

ok

i put pth and index in a folder of the model zip or another zip inside of the model zip

[gloomy smelt]

Oh yeah i forgot to ask

[paper wave]

do i put the pth and index in a folder or a zip of the model zip inside of the model zip

[gloomy smelt]

@proper shale Okay so uh, does batch_size as 8 affect my pc at all?

Or what does it even do

[proper shale]

just zip the files straight up @paper wave

[paper wave]

alr

[proper shale]

how much VRAM does the 2070 have

[gloomy smelt]

How do i check?

[proper shale]

i mean you could search online or something

[gloomy smelt]

Also i just started the training before sending message lol

let me stop it just in case

[gloomy smelt]

hmm

OH WAIT

You mean this?

@proper shale

[proper shale]

no that's general RAM

im talkin about VRAM

from the GPU

[half cove]

check on task manager

[modern oxide]

Open the task manager and go to the performance tab

[gloomy smelt]

Oh

[gloomy smelt]

Ohhhh okay

gimme a sec

[modern oxide]

There select your Nvidia GPU and send the screenshot

[modern oxide]

8 gb

[glad zealot]

8gb vram

Seems good.. might just be a bit slower than colab but it should run

[modern oxide]

Should be good

Maybe even 12 would be fine

[glad zealot]

Should be half speed of colab since iirc colab t4 have 16gb? But that's just based on basic math... Might not be the same in practice

[gloomy smelt]

So?

[half cove]

Arnt colab t4s shared?

[gloomy smelt]

Do i continue with 8 batch size

like no problem at all

[glad zealot]

Seems not? At least using voice changer with 2 people using it at the same time makes it run worse

[glad zealot]

Yup all good

Pretty sure you can do 10 alao

[proper shale]

yeah sure go for it

[gloomy smelt]

k

Thanks yall, hopefully this goes faster

[agile ridge]

do you guys touch mastering or EQ after ai cover finished? or just add vocal + MR together

[rare goblet]

Ayo? @agile ridge level 1 !!!

[stoic rampart]

i need help making a model i cant seem to get it to sound good

[rare goblet]

Ayo? @stoic rampart level 1 !!!

[glad zealot]

Need a clean dataset, no pause, no background noise, only the voice of the character you want

250 epochs seems to be the safe stop on most models

[stoic rampart]

thanks how much of a sample do i need

[glad zealot]

10mins minimum Ig?

[stoic rampart]

ok tysm

[orchid glade]

what's better rvc gui or w-okada or is it just the same ?

[glad zealot]

Depends on you but I personally use okada

[stoic rampart]

how do i get okada

[glad zealot]

That's for voice changer

[orchid glade]

thanks I'll just use okada then

[stoic rampart]

oh ok

[glad zealot]

Just search w-okada voice changer on google if you want it

[orchid glade]

Is this settings ok ? or should I keep batch size to 6

data set of 24 mins

[glad zealot]

No idea what batch size do but I keep mine on either 8 or 14

[orchid glade]

what gpu do you have

[rare goblet]

Ayo? @orchid glade level 2 !!!

[glad zealot]

I use colab and kaggle

So it's T4

On colab and two T4 on kaggle

[neon mortar]

i still cant get a smooth sound through discord damn

what i found out its like the app problem cuz if i use just steelseries the sound is fine

but steelseries cant control two output

[sly solstice]

sorry im trying to download the italian voice pack of goku but it asks me a password to open the download link. what do i do?

[glad zealot]

Get the password from the person that uploaded the model ig?

[sly solstice]

Information from the original post
Uploaded By: @fickle sun
Uploaded Date: 25 July 2023 20:07

he left?

[glad zealot]

Probably if it doesn't become pingable

[sly solstice]

then its unusable i guess

[neon mortar]

how to use vac470 lite

[stoic rampart]

can someone make a model for me

pls

[arctic solar]

I'm using the Mangio-RVC-Fork you run locally. Where do I put the .pth voice models?

[stoic rampart]

weights

[arctic solar]

There isn't a folder called that

[rare goblet]

Ayo? @arctic solar level 1 !!!

[arctic solar]

Mind if I dm you the folder list?

[stoic rampart]

sure

[arctic solar]

Your dms are off

[stoic rampart]

one sec

[stoic rampart]

try now

[hot lagoon]

shoud i use the 40k or the 48k for this

[glad zealot]

It says so.. so maybe???

[hot lagoon]

cuz I think 20 khz is find cuz 20x2 is 40

[flint summit]

so, fun fact

i accidentally unplugged my external ssd when training a model

how do i continue from a backup?

[iron sandal]

Hello Guys

I only have options for Crepe_Tiny and Rmvpe_onnx?

I was watching a tutorial and the guy had all options, am I doing something wrong 🥺

[glad zealot]

Not sure but rmvpe_onnx is good

[iron sandal]

Okay 🥺

Thanks

[iron sandal]

Are you a mod?

[glad zealot]

Helper but sure

[iron sandal]

I wanted to use a Voice changer platform like RVC for Voiceovers. But, I can’t get the best output out of it. Will it help, with the quality and the accent, if I run my voice to a clear voice same (Lets call it X) and run the X through the Voice I actually want?

But the Voice X is an Elevenlabs Premade voice. And I don’t have the Index or the Pth file for it.

How would you go about this?

[flint summit]

ig i'll just have to retrain the model, because idk how to start back from a checkpoint file

[glad zealot]

If it's voice over I think it would be better to record everything first using your voice then drop it to cover gen

Would be faster on something like mangio then covergen since covergen will try to remove background noise

Will takes 2-5 mins on cover gen.. will take around 10-30sec on mangio

[iron sandal]

I see, thats amazing, I would like to know where I can learn more about Mangio and Covergen

[glad zealot]

There's also an online version of covergen and voice changer

[arctic solar]

My thing just doesn't work at all but it's a little hard to explain since I'm not sure what the problem is, so if anyone is willing to watch me screenshare it, lmk

[glad zealot]

Uhhh... Not me atm me at school

[sinful plover]

hey

[rare goblet]

Ayo? @sinful plover level 1 !!!

[sinful plover]

so i havent done ai since september and where do we do it now?

[glad zealot]

Everything should be in #1159513888199540817

[iron sandal]

Thank you!! So much, I see you are helping everyone! You are a great person :)

[arctic solar]

When I select something for the inferencing voice, the .3 box and one of the other boxes continuously load for some reason

[rare goblet]

Ayo? @arctic solar level 2 !!!

[glad zealot]

Yup they should

But should not load infinitly

[stoic rampart]

can you help me make a model

[glad zealot]

Nope :D

I'm at class atm

[stoic rampart]

oh ok :)

[zenith star]

is rvc avaliable in mac m1?

[green birch]

does it matter?

[glad zealot]

Seems to be

[sinful plover]

i need help

i’m confused

i look in the #1159513888199540817

but i’m confused how to do rvc

i haven’t done it since september when it was the google drive thing

[hot lagoon]

was is the point of overtraining?

[glad zealot]

Nonidea

[hot lagoon]

also is this overtraining?

[glad zealot]

Nope

Keep going

[hot lagoon]

k

[flint summit]

i'm back after two hours to ask again if anybody knows how to start back off at a checkpoint when training a model

[rare goblet]

Ayo? @flint summit level 3 !!!

[flint summit]

because i disconnected my external drive on accident where rvc was running while training a voice model & i don't want to restart the whole process

[flint summit]

https://tenor.com/view/kaguya-touhou-explode-bald-head-gif-26674611

dang, no embed

epic embed fail

[glad zealot]

https://tenor.com/view/jesus-ballin-mars-bars-gif-19910027

[flint summit]

i'll just leave my laptop on over night & wait to get unbanned from ssbu lol

i actually got banned because someone else disconnected in a match

[glad zealot]

Tt

Sounds unfair

[iron sandal]

But mine is Intel UHD 630 and Nvidia GTX 1650 😅

Ahh dammn 😂

Thanks! I will have a look again, but I am pretty sure I downloaded the right one 🤡

[glad zealot]

trust the mother

[paper portal]

@halcyon oxide you resolve? I have same issue

[still crystal]

how can i fix this i ran the feature extraction

[flint summit]

i finally finished the model B)

[hot lagoon]

overtraining?

[modern oxide]

no, it goes down, which means it's not overtraining

[quaint walrus]

Hi guys, I have a quick question: every custom AI model I add is lagging badly. What can I do?

[modern oxide]

in voice changer?

[quaint walrus]

yes

[covert igloo]

Hi internet! I have a question ... would anyone here be willing to help me train a voice? My computer is way to slow.. I would be really happy and thankful!

[modern oxide]

you probably put your setting to high for your system, try to use setting from guides

[proper shale]

You can do it online, just check #1159949278270193734 and guides

[quaint walrus]

Ye, but I put the setting to the absolutely minimum and the RES was by 10000ms

[rare goblet]

Ayo? @quaint walrus level 1 !!!

[modern oxide]

or you can request a model in #1159289738314919936 usually people make paid requests to get their model done and done quickly

[modern oxide]

what gpu do you have?

[covert igloo]

thanks, will check it out!

[quaint walrus]

rx 6700xt

[modern oxide]

ccan you send me a screenshot of your settings?

[modern oxide]

try this

[quasi osprey]

this is directml ver right

[rare goblet]

Ayo? @quasi osprey level 1 !!!

[quasi osprey]

i have same gpu

time to try

[quaint walrus]

the thing its works for the AI models that were already there

but not for the custom from this discord

[quasi osprey]

u see that convert to onnx

u need to do that

[quaint walrus]

its not work

[quasi osprey]

for AMD gpus onnx is recommended otherwise itll be laggy

[quaint walrus]

ok thanks

[quasi osprey]

your whale cum

[quaint walrus]

I'll try it

ok I have the onnx file, what have I do now?

[quasi osprey]

import the file

[quaint walrus]

where?

[quasi osprey]

the way you normally import a .pth

[quaint walrus]

ok

the index also?

[quaint walrus]

Ok its works thank you so much

[quasi osprey]

ya have a good day

no one cares bro 🫡

[gleaming spire]

So I tried Wokada the first time yester. But whatever I do I dont Sound okay-ish. It is always to high or low pitched. Are there other settings that I am missing ?

the slider "gain", "tune" and "index".

Those are the settings that I played around to Sound okay-ish

(Those are not my settings, its a Screenshot cause I am currently not at home)

[proper shale]

You should definitely increase tune if you're a man trying to sound like a woman

[gleaming spire]

I tried that too, maybe I have to experiment a bit further

[rare goblet]

Ayo? @brittle wing level 3 !!!

[proper shale]

#1159949278270193734

[stone dawn]

heyy guyys does it need to have a .wav format of dataset when using rvc disconnect?

for making a model

ping mee thank you!

[sand iris]

Any audio format is fine

[stone dawn]

thank you!!

[sand iris]

They just have to be properly zipped for rvc disconnected to recognize them

It should be written somewhere

[polar mantle]

@sand iris

[sand iris]

Is there a file named run.sh ?

[polar mantle]

yep

[sand iris]

It should be that one then

Oh my bad I hadn't seen the second screenshot x)

it was right there

[polar mantle]

it says this

when i open it

[rich gulch]

The audio path is correct

RuntimeError: You input a wrong audio path that does not exists, please fix it!

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

[sand iris]

Are you on macos ?

[polar mantle]

yeah but i opned it through peazip

should i open it normaly?

[sand iris]

Oh you should unzip it first

It won't be able to run if it's from peazip since it won't see the other files in the archive

[stone dawn]

ooof

[polar mantle]

but i dont think you can extract on mac

[sand iris]

If you can read the contents you should be able to extract

7z is supported on all platforms, if you have a software that can read that

[quartz frost]

guys

how do I ai convert text to speech?

[modern oxide]

https://huggingface.co/spaces/TheStinger/Ilaria_TTS use this to create tts audio and then convert it thrpugh any rvc fork

[quartz frost]

how do I make it convert with a model I already have in huggingface?

[rare goblet]

Ayo? @quartz frost level 2 !!!

[polar mantle]

it didnt work

it extracted 12,000 files seperately into my downloads

my bad 107,000*

[modern oxide]

you can use any fork to inference the tts audio after you make it, or you can use ai covergen

[quartz frost]

I'm a bit confused sorry, what is a fork?

But I'm going to assume it doesnt allow us to insert the models we want, so best is to go there first, use any model that is available, download the line they say and than we convert again by making the ai we want say the line we downloaded?

[modern oxide]

i don't see what exactly you don't understand, first you go to any tts basically (but i recommend ilaria tts) make an audio with any available tts voice, then you open the rvc that you have and inference the audio

[onyx forum]

how do i resume training on the colab

[rare goblet]

Ayo? @onyx forum level 1 !!!

[quartz frost]

whats a tts

[modern oxide]

text to speech

[quartz frost]

OH

mb

alright i got it

so whats the rvc I use?

after

getting an audio voice line from ilaria

[modern oxide]

any that you like, i recommend mangio because it's easy to use

[quartz frost]

link me please

theres like a lot

[modern oxide]

#1159949278270193734

[quartz frost]

this right?

[modern oxide]

yes

the infer one

[quartz frost]

what next?

[modern oxide]

you press this arrow to download it in the previous page

[quartz frost]

ah ok

[vague ermine]

why everytime i try to put the mic or speaker it says

Error
unhandledrejection
no error stack
NotAllowedError: Permission denied by system
no stack

sorry only the mic the speaker is good

[onyx dock]

where can I get expressive RVC models?

[warm dirge]

Hi guys im downloading Mangio-RVC-v23.7.0_INFER.7z
after download this i just open it and can run it and use the voice models here?

[modern oxide]

yes

[warm dirge]

even if these are 7z and not zip? How can i open it just unzipping?

[zinc raft]

not working already :(

https://colab.research.google.com/github/SociallyIneptWeeb/AICoverGen/blob/main/AICoverGen_colab.ipynb#scrollTo=NEglTq6Ya9d0

is there any alternate collab to make a cover?

[hot lagoon]

I know when TensorBoard gose down it means it's not overtraining so how dose overtraining look like

[modern oxide]

Yes

[viscid moss]

#1159949278270193734

[modern oxide]

It goes up for a long period of time and steps

[quartz frost]

what now?

@modern oxide

[void summit]

Can anyone tell me best voice cloning websits plss

[quartz frost]

i was stuck here for a while

[void summit]

Can anyone tell me best voice cloning websits plss

[modern oxide]

Just curious, but have you watched or read any guides?

[void summit]

Plss

[quartz frost]

no, I did have rvc before and vced with helpers BEFORE to help and My laptop cant so thats why ive just been using colabs

I was told to

[modern oxide]

#1159949278270193734

[viscid moss]

Like a website u can use Colab or Hugging Face in #1159949278270193734

[void summit]

Ok thanks

[modern oxide]

Read guides first please, our mods spent a lot of time making them

[hot lagoon]

so this is okay yeah?

[quartz frost]

can you please just tell me what I should press next? since if i remember I have to press web something
I have to do other things its like 9:54pm im multitasking rn

[viscid moss]

it's ok

[hot lagoon]

thanxs

[modern oxide]

Guides brother, guides #1159513888199540817

It takes like 10 minutes to read it and install everything you need

[unique quest]

подскажите что делать?
Не начинаю тренировку эпох и ошибка такая
индекс получился а эпохи нет

скриншот не могу засунуть

[latent lotus]

hi

i wanna ask something

so i alr press download

but there's no file on my computer

can anyone help me?

oh

sory

sorry*

i got it now

[rich gulch]

Anyone know an easy way to merge vocals to a backing track in adobe audition?

[latent lotus]

um how can i change the word or maybe put song so the ai can sing?

[rare goblet]

Ayo? @latent lotus level 2 !!!

[fierce crown]

Can someone help me With the Realtime thingy? When i use it it says a a a a a a aa a a a a a a a a a a a

[glad zealot]

enable sup2 on the voice changer

also #🔍│help-w-okada

[fierce crown]

sup2?

[glad zealot]

ya its n the settings on it

[fierce crown]

could you be more specific i dont see it anywhere

ok 1 question is it because i am using go-realtime or should i be using go-realtime_new?

[fierce crown]

i get an error when using the New one it says...

[glad zealot]

ah that one idk about that

[fierce crown]

so i was using the wrong one?

[glad zealot]

i use okada voice changer

[fierce crown]

what is okada voice changer and what is the diffrence

[glad zealot]

generally easier to use (for me)

[brittle wing]

can somone help me

I need to find link for srk model rvc

[proper shale]

search for it on weights.gg

[copper pebble]

#1159289700490694666 or /search command in #1159290666019471421

[mortal pawn]

I have update Applio and now im getting this error after a Few Uses

[modern oxide]

use mangio or og rvc

[rare goblet]

Ayo? @peak tusk level 6 !!!

[split flax]

i dont understand, is there a google colab link that i can use for make ai covers?

[native relic]

when i try to launch rvc gui this error pops:

[mortal pawn]

is 4GB not enough? Also i was able to Generate 3 before it Started doing that

[copper pebble]

no im afraid

[frozen sandal]

i'm getting this error when I select "split audio"

any idea how to fix it? it was working just fine before

happens to every song

[copper pebble]

what r using??

[frozen sandal]

applio from colab

[copper pebble]

u trying to do inference or training??

[frozen sandal]

inference

[warm dirge]

excuse me guys after i downloaded the Manlio file whats the next step?

[copper pebble]

mhh that works but u can use

ilaria rvc

easier

[frozen sandal]

can you link me? i'm kinda new here

[copper pebble]

ofc

[warm dirge]

Otherwise i downloaded models and trying to use it on Ilaria RVC, since im new i dont know if it could be used text to speech or i need an audio to clone

Can someone help please

[copper pebble]

https://discord.com/channels/1159260121998827560/1167844377046040769 heres a complete guide

[copper pebble]

if u wanna make models say the things that u inputed on the tts

ull need first to download the tts speech and then convert it on ilaria rvc

[frozen sandal]

thanks

[warm dirge]

Bro grazie mille ma anche quando lo scrivo su tts me lo da con la voce classica e non quella del modello

[rare goblet]

Ayo? @warm dirge level 2 !!!

[warm dirge]

e se faccio convert mi da error

[copper pebble]

italiano??

[warm dirge]

si bro

[copper pebble]

downloada quella voce

[warm dirge]

gia fatto

ho il file xpi

[copper pebble]

ok

[native relic]

when i open rvc_gui this happens (dont mind the red)

[warm dirge]

ho scaricato pure manlio ma so negato

[copper pebble]

hai provato a convertire gia con ilaria rvc??

[warm dirge]

si

ma non viene

carico modello e tutto

[copper pebble]

duplica lo spazio

[warm dirge]

poi da text to speech non mi da nulla

come si fa

[copper pebble]

andrà di sicuro

[warm dirge]

grazie mille bro sto provando da stamattina

[copper pebble]

asp ti giro il link

[copper pebble]

oggi sarà lultimo giorno tranquillo

[warm dirge]

di cosa l'ultimo giorno?

Yessir

[copper pebble]

https://huggingface.co/spaces/TheStinger/Ilaria_RVC?duplicate=true

[copper pebble]

hai detto che ci provi da una settimana no??

oggi ti farò avere i risultati

o almeno spero lol

[summer turret]

Help me make a ai song

Please

Guys

[copper pebble]

yh

im here

[summer turret]

How u do it

[warm dirge]

sisi bro mi so preso un colpo quando hai scritto cosi, non stavo capendo grazie lmao

[copper pebble]

haha

hai duplicato btw??

[frozen sandal]

i was checking this out, it doesn't have the "split audio" option, which was making the results sound better and more natural

[rare goblet]

Ayo? @frozen sandal level 1 !!!

[frozen sandal]

is there any other option with that feature?

[copper pebble]

split?? divide??

[warm dirge]

si ma mi dice application started

application startup

[copper pebble]

si sta avviando

[frozen sandal]

in applio there's this option called "split audio", which gives better results

[copper pebble]

wow didnt know abt this at all

[frozen sandal]

yeah, that's the option i was talking about, when I enable it, it gives me an error, it was working just fine

it has been there for a couple weeks now

[copper pebble]

mhh

i see

@warm dirge mi puoi mandare in dm laudio che devi convertire ed il modello che vuoi usare plss??

[warm dirge]

si

[native relic]

guys when i start rvc an error appears

[molten pecan]

Do you watch AbsolBlogsPokemon?

[native relic]

no

[rare goblet]

Ayo? @native relic level 1 !!!

[native relic]

who's that

[stoic rampart]

Can someone make a model for me pls I can't get it to work

[molten pecan]

You can ask on #1159289738314919936

You might have to pay though

[stoic rampart]

Ok thanks

[long gyro]

how do i make it so only the voice comes out of the software and not my mic

[native relic]

guys ineed help there's a error that pops up when i use rvc_gui

[daring dust]

which site did i go to make an ia voice?

[molten pecan]

-rvc

[azure marsh]

Hey, @daring dust!

👇 Here are some useful resources to help you learn how to make ai covers

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

[native relic]

this happens when i star rvc gui

[solar plover]

what is a good cpu for this

[agile quarry]

i know theres a way to download rvc on MAC but like, i need help

i cant find an actual download

but i can find pages

[sand iris]

spaces in folder names leading to RVC cause issues

[native relic]

ight

[iron sandal]

Hello Guys, what’s the best option: Out of RMVPE, Harvest, Dio and Crepe?
To get the best Voiceover for a Youtube video in a different voice?

[prisma thicket]

-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

[rare goblet]

Ayo? @prisma thicket level 3 !!!

[violet heron]

RVMPE

[violet heron]

don’t use cpu, use GPU

[solar plover]

no i know to use a gpu
im just saying what's a good cpu to keep a pc up and running

smh my h

[split valley]

is there a good way to automate applying voices to audio files? pref linux cli or something

[hot lagoon]

How to update my uvr5 to a new visorn to uvr5

[half cove]

like batch upload?

[half cove]

[hot lagoon]

[brittle wing]

Is this overtraining? Since the 100 Epochs the board has not gone up again.

[copper pebble]

how much was ur dataset long

[brittle wing]

15 minutes x 20 batch size

[copper pebble]

why so much batch size??

u could do 300 epochs with 8 batch size

[brittle wing]

The Batch Size issue is very imprecise, since I was recommended to use 20..... But with 8 it would take longer but it would certainly give better results and of course the important thing, more stepts.

mmm... If I have a dataset where the character whispers quite a lot. what agorithm can be useful? Well I made one and I don't know, it sounds very amplified besides making a lot of "s" in the background.

[copper pebble]

thats not bad

but i think u should try to train normally

8 batch size

[brittle wing]

10 Minutes x 20 batch size and 300 Epochs x 5k stetps

[rare goblet]

Ayo? @brittle wing level 3 !!!

[copper pebble]

mhhh

dunno

[brittle wing]

Is the "8" the miracle number for all amounts of datasets? Or has anyone done a comparison of how many batch sizes are needed for a certain amount of datasets?

[copper pebble]

i did a lot of models

using 8 batch size

today i made 18 min / 8 batch size/ 350 epochs

[sharp island]

Or should I just wait

Is there a fix to this

[copper pebble]

cant read a shit sorry

[sharp island]

Zoom in

[copper pebble]

do u mind do "stamp" and paste here

[sharp island]

You can’t currently connect to a GPU due to usage limits in collab

[copper pebble]

i meant to lemme see what the message said

did u ran outta of runtime??

[split valley]

yes but no, I will have continually new files and I want to apply the voice and then send them along. I needs to be automated, server side

[sharp island]

I think

Yeah

[copper pebble]

were u training??

when the runtime finished ofc

[sharp island]

I was gonna…I’ve been trying to make this model but the damn computer usually automatically shuts down before it finished

It’s the epoch section that’s the most problematic

[copper pebble]

how much was ur dataset long?? what settings did u put??

[half cove]

🤷‍♂️

[sharp island]

500 then 1000

[fleet marsh]

google collab rvc its still dead?

[half cove]

you can use disconnected

[fleet marsh]

covers only or creating models too?

[half cove]

colab for creating and huggingface spaces for covers

-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

[fleet marsh]

thankss

[half cove]

[solar cedar]

what does this mean 😭

😭

how do i fix this

[viscid moss]

#🔍│help-w-okada

[hot lagoon]

in UVR5 what is best for songs ( aka just makeing a cover )

[copper pebble]

i usually use VR_Karaoke_5_UVR

[tame sigil]

I’ve been using AICoverGen_collab but all of a sudden today I’m getting this error: anyone know what it could be? I’m on iOS mobile.

Timer: 00:01:30Traceback (most recent call last):
File "/content/HRVC/src/webui.py", line 8, in <module>
import gradio as gr
ModuleNotFoundError: No module named 'gradio'

[rare goblet]

Ayo? @tame sigil 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

[hot lagoon]

Same

[rare goblet]

Ayo? @hot lagoon level 5 !!!

[hot lagoon]

I try to fix it by change up the code to imoprt the gradio tho is not working

[brittle wing]

I have no idea how to work RVC V2 hugging face website

[viscid moss]

no way!

AiCoverGen is down?

I'm going to check

[hot lagoon]

[viscid moss]

https://discord.com/channels/1159260121998827560/1167844377046040769

[dry marlin]

can someone help im trying to open and i get the warning so i press run anyway and it does nothing anyone can help?

[viscid moss]

can u show me?

[dry marlin]

okey

[rare goblet]

Ayo? @dry marlin level 1 !!!

[dry marlin]

where do i send

[viscid moss]

here

[dry marlin]

oh wait i didnt saved

[viscid moss]

@tame sigil @hot lagoon it's working for me.

Try my spanish version of AICoverGen:
https://colab.research.google.com/drive/1qezaBvCn4eSG1U2L4a64tEHTynbIu8eN?usp=sharing

[dry marlin]

it couldnt let me record on pc

[sand iris]

gg

you moved the file out of the folder

Move it back to MMVCServerSIO

and create a shortcut instead

[dry marlin]

ohhhh ty

:Ddd

alr

[potent badger]

?????????

[rare goblet]

Ayo? @potent badger level 18 !!!

[wheat apex]

https://tenor.com/view/papich-папич-arthas-gif-20124926

[potent badger]

.bat should be in the vc folder

[dry marlin]

there wasnt

[rare goblet]

Ayo? @dry marlin level 3 !!!

[dry marlin]

i dont have an app downloaded im using recorder that comes with windows

i turned on

it not gonna do anything but i just turned

[broken sluice]

when i clicked download on a model it gave me a zip file how do i use that

[viscid moss]

extract, pth file into "weights" folder and index file into logs/"folder"(create a folder)/ and paste.

[broken sluice]

whats pth

oh nvm

[dry marlin]

i m trying the voice changer and its like super lagy you cant understand nothing so my gpu to bad?

[toxic snow]

does anyone know how to fix this ?ile "/content/HRVC/src/webui.py", line 8, in <module>
import gradio as gr
ModuleNotFoundError: No module named 'gradio'

[rare goblet]

Ayo? @toxic snow level 1 !!!

[toxic snow]

im on colab

[onyx forum]

anyone know how to resume training on the colab

[toxic snow]

what is dependdency installation block?

i dont get it

like where is it

ok one sec

google

im useing google colab

ok

ok

so what do i do now?

hold on can you dm me what that looks like?

i have no idea what im doing

[viscid moss]

What colab?

[toxic snow]

google colab

[rare goblet]

Ayo? @toxic snow level 2 !!!

[viscid moss]

ya but the title

share me the link

[toxic snow]

https://colab.research.google.com/github/SociallyIneptWeeb/AICoverGen/blob/main/AICoverGen_colab.ipynb#scrollTo=NEglTq6Ya9d0

[viscid moss]

Is that still alive? Now we use a version without GUI

-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

[viscid moss]

The first one

[toxic snow]

ok

[toxic snow]

thats it?

[toxic snow]

THX

hollup

@viscid moss how do i generate ai covers? (ive done it before)

[viscid moss]

just run the cells and replace this

Model link, name, youtube link for cover and name

[eternal turtle]

how could i make colab faster

[rare goblet]

Ayo? @eternal turtle level 3 !!!

[viscid moss]

u can't

[eternal turtle]

sad

[toxic snow]

ok thx

[inland hamlet]

-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

[toxic snow]

@viscid moss where is the finished product?

i did everything its just says # Wait for the process to finish
process.wait()

[unique basalt]

Hi guys . Does anyone know how to cover in RVC without Index?

My friend can but I get this error

https://cdn.discordapp.com/attachments/892851266311573628/1172345109170364416/image.png?ex=655ffa77&is=654d8577&hm=6f83db93753f96755d48bfa95a01ef67a18127f29c8ce47f0296c09effb12c10&

it's applio

[toxic snow]

@viscid moss it doesnt show up the final product like i ran it alot of times and it doesnt show up

what do i do

?

[viscid moss]

search for that route

[toxic snow]

its no whare

i cant find it

[viscid moss]

into the folder

[toxic snow]

ohhhhhh

thank you

wait

what is the name of the folder

[viscid moss]

something like this

a random name

[toxic snow]

ok

[unique basalt]

please?

[little owl]

how would i resume training on applio collab? would i have to preprocess/feature extract before?

[unique basalt]

Nvm , problem solved

[rare goblet]

Ayo? @unique basalt level 1 !!!

[dusk crystal]

Does this work on mac

The voice thing

[viscid moss]

realtime : ✅
rvc : ❓ idk

[dusk crystal]

epic

[agile quarry]

wait but like where do you get the rvc ai cover mac thing

not real time

[tame sigil]

Hi thank you a lot! I just read everything on the new collab and the process! But can this be done on iOS too like the old one and if not is there any running ways to use that same gradio interface without the public queue? So far I’ve been using that same old model but just today it started glitching (https://colab.research.google.com/github/SociallyIneptWeeb/AICoverGen/blob/main/AICoverGen_colab.ipynb#scrollTo=NEglTq6Ya9d0)

[viscid moss]

hmmmm u've this options:

1. Hugging Face Spaces (but it's slow)
2. Colab (u can use my fixed spanish version https://colab.research.google.com/drive/1qezaBvCn4eSG1U2L4a64tEHTynbIu8eN?usp=sharing)
3. WordBand bot on this server (just use /cover and follow the steps)

[tame sigil]

Thank you! Also when using your collab do we have to run all cells every time we input a new song or change pitch, reverb, etc.?

[agile quarry]

wait i swear on the github they said smth about a mac version but i cant find the download

[rare goblet]

Ayo? @agile quarry level 2 !!!

[agile quarry]

dude shut up

[viscid moss]

just the last cell of code using other YT link

changing the pitch, reverb, etc as you want

maybe u need to set another name, i'm not sure about that

[sturdy yacht]

my cuda is not being used somehow i have cuda toolkit installed and have a cuda supported gpu

[mighty parcel]

respect the bot king

[sturdy yacht]

i have tried everything from reinstalling cuda python

anaconda and miniconda 3

reinstalling poytorch to

[sturdy yacht]

but torch.cuda.is_available keeps returning false

[agile quarry]

im just waiting for osmeone to answer my question :(

[sturdy yacht]

since updating my gpu and such everything has crashed

[viscid moss]

hmmmm

[agile quarry]

on the github

in the older versions thing they say they had a mac thing (not tested)

but i cant find it

[viscid moss]

in #🌍│español we have documentation and videos on how to install on Mac but in Spanish

[agile quarry]

ok

[viscid moss]

let me see

[azure marsh]

👇 Suggestions for @agile quarry!

RVC v2 Disconnected

Link: Google Docs

Instalación de Mangio RVC v2 localmente para macOS

Link: Google Docs

Autor: anthonydiazjorges

[azure marsh]

👇 Suggestions for @agile quarry!

🍏 Tutorial de Applio en español

Link: YouTube
Autor: IA Hispano

🍏 Tutorial Applio (M1)(Mac)(M2)

Link: Documentación
Autor: AI HUB FRANCE

Servidor IA Hispano

https://discord.gg/iahispano

[viscid moss]

And Applio

[agile quarry]

well uh i clicked the github in this doc and it dosent exist anymore

[viscid moss]

no way!

[agile quarry]

man this makes me so mad i could just

the only ai thing i can use at the moment before i get rvc mac is... kits.ai!!!!!

[viscid moss]

@waxen wing que paso aquí apa? Porque borraste el fork? La gente no puede instalarlo para Mac

[viscid moss]

and.... if u use a VM?

[waxen wing]

yo no borre nada

XD

de que hablas

[viscid moss]

dice que la repo no existe

[waxen wing]

a ver

[viscid moss]

https://github.com/Anthonyxd22/Mangio-RVC-Fork-m1

[waxen wing]

wtff

te imaginas que me lo tumbaron

dejarme ver

[viscid moss]

revisa

[waxen wing]

se cayo github

no lo boorre

eso paso con repositorio ia hispano

https://github.com/Anthonyxd22/Mangio-Crepe-m1-m2

[waxen wing]

ese es viejo

[waxen wing]

este es en actualizado

[viscid moss]

actualiza en el documento

para que la gente no tenga problemas

[waxen wing]

cual documento

o ya

ya lo areglo

XD

[viscid moss]

D

xd

y con esa repo como se instala? :V

[waxen wing]

https://docs.google.com/document/d/1F7sJ6ZFS5CfQXEZ-I-v9k3TAFTSbCOkPdOyn841UbyI/edit#heading=h.qae8w7o484c5

[waxen wing]

es lo mismo

[viscid moss]

ya vi, olvidalo

ty

[agile quarry]

btw i downloaded this just now and what file is the program actually in?

[viscid moss]

https://docs.google.com/document/d/1F7sJ6ZFS5CfQXEZ-I-v9k3TAFTSbCOkPdOyn841UbyI/edit#heading=h.qae8w7o484c5
Updated guide
@agile quarry

[agile quarry]

cool

i downloaded the github where is the file at

for like actually running the software

[viscid moss]

If you need more help talk to @waxen wing , I don't know anything about Mac 💀

[agile quarry]

yeah

all im asking rn is where the file is located

[waxen wing]

@agile quarry

[agile quarry]

o

[molten pecan]

Morning mr eddy

Looks like you are running late

[viscid moss]

what

[brittle wing]

Does anyone know any good rvc to tts programs? I’ve used things like Tortoise and Applio, but they’ve always seemed to just come out with outputs that don’t really sound great, and not like the thing I’m trying to replicate.

[agile quarry]

not really, but what if you just downloaded tts audio and put it through rvc?

idk

[brittle wing]

That’s not a bad idea. Gonna try it out, needa replicate my voice for my Spanish project, got the model but nothing else 💀

[dusk crystal]

Is there a specific place to downlaod this on ,ac

[rare goblet]

Ayo? @dusk crystal level 1 !!!

[dusk crystal]

mac

[viscid moss]

Applio is the best for TTS

But u can use TTS and then covert using RVC

[austere pivot]

Hola, al querer usar la pestaña de interferencia para hacer un cover, me salio este error. Un pequeño detalle es que el mismo error me sale al querer usar el TTS. Alquien me puede ayudar a solucionarlo, por favor?

Hello, when trying to use the interference tab to make a cover, I got this error. A small detail is that I get the same error when trying to use TTS. Can someone please help me fix this?

[plucky locust]

https://huggingface.co/dacoolkid44/vtuber-rvc/tree/main/gawr gura is not a registered model! Use /list to list all models, or use /register to register a new model.

Community Models

goku (Goku (English Dub))
luffy (Luffy (English Dub))
mickey (Mickey Mouse)
peter (Peter Griffin)
plankton (Plankton)
spongebob (Spongebob)
squidward (Squidward)
waluigi (Waluigi)
... and 200+ more

Check out all of the public community models at https://www.word.band/voice_models

You can use any model using the model id (bolded) in the /cover command. For example:
/cover model:goku youtube:https://www.youtube.com/watch?v=s9FCCiEs84k

To see your private models and register new ones, please start a DM with me!

[rapid spade]

hey is there any way to isolate two voices from each other like say there is someone singing but there is someone in the back ground also singing how can seperate them or this is not the place to do it?

[rare goblet]

Ayo? @rapid spade level 1 !!!

[toxic snow]

uvc

ultimate vocal remover

[plucky locust]

You're in the queue! Your cover will be ready soon. Number of people in front of you: 0 (Status: 🟢)

[rapid spade]

alright i did that but there is still a hum in the back ground of the words is there a solution to that or me screwed

[tame mica]

@toxic snow wordband in botspam or dm

[toxic snow]

my fault lil bro

srry

[toxic snow]

only hum or is there reverb etc there too?

[rapid spade]

only hum i just want to remove that i just want to get the words

i tried all of them

[toxic snow]

ima show a turtorial in yo dms ok

[rare goblet]

Ayo? @toxic snow level 3 !!!

[toxic snow]

@rapid spade i sent you a solution in dms

[rapid spade]

yeah i got it

thanks

[toxic snow]

ok

[toxic snow]

np

[modern oxide]

it probably works check the gpu cuda graph in task manager

[sturdy yacht]

How do I get to that

[hoary zephyr]

hey question Can my pc run this locally? i have a 4 core i5 gtx 950 with 8 gigs of ram. If so can someone help install it, and if i cant run it could someone help with running it non locally Its been a while and I'm confused.

i lost a lot of my knowledge with the deletion of the last server. its just to synthizize voice not making models

[modern oxide]

Task manager - performance - click your GPU

[modern oxide]

Minimum is 1050 ti for training, and 4 cores I think won't be enough

[hoary zephyr]

no not training just to change the voices so my talking can sound like jfk or something like that

[rare goblet]

Ayo? @hoary zephyr level 1 !!!

[hoary zephyr]

if i still cant any way to run it on the web or something i know the colab method got axed

[modern oxide]

Hmm, then I think it might be enough

#1163883017446629508 message

Try mangio fork

[hoary zephyr]

thanks man my brain doesnt think and its been a long while

[austere pivot]

Can someone please help me with this error?

[modern oxide]

There are many useful guides it makes it easier to understand #1159513888199540817

[hoary zephyr]

i may sound old but boy it used to be a few clicks

[modern oxide]

It still is

[hoary zephyr]

its just the whole thing with that colab getting a nerf i used it so i wouldnt have to listen to the same sinatra songs over and over

[sturdy yacht]

wheres the cuda graph

[modern oxide]

Click on video encode and choose cuda

[hoary zephyr]

which might you reccomend from the colabs? like i said its been so long and im a little confused

[modern oxide]

EasyGUI by rejects is awesome, but none of collabs currently have working webui

[hoary zephyr]

for ai covers?

[rare goblet]

Ayo? @hoary zephyr level 2 !!!

[modern oxide]

You still can use them in colab itself

[modern oxide]

Ahh

Ai covers

Wait

[sturdy yacht]

it doesent come up its not on the list

[hoary zephyr]

yee

[modern oxide]

#1167844377046040769 message

[hoary zephyr]

Like i said the same sinatra songs get old so pmj makes the old music so much easier

[modern oxide]

Hmm

[modern oxide]

I had such problem

[rare goblet]

Ayo? @sturdy yacht level 2 !!!

[modern oxide]

Go windows parameters

Display

Graphics

And there make a screenshot

[sturdy yacht]

Windows parameters

[modern oxide]

Nah

No

[sturdy yacht]

oh ok

[modern oxide]

Not that

[sturdy yacht]

how i get to what ur saying

[modern oxide]

Go here

[modern oxide]

Scroll down

[modern oxide]

Graphics settings

[modern oxide]

Disable hardware acceleration and restart the PC

[sturdy yacht]

ok

[modern oxide]

Then cuda should appear in task manager

[sturdy yacht]

ok brb 1 minute

that showed up still rvc does not detect cuda gpu

still

[modern oxide]

Do you have a training on?

[sturdy yacht]

no

i cant even launch rvc

it doesent launch it complains that cuda is not availible and wont let me use the training tab

same with stable diffusion

[modern oxide]

[sturdy yacht]

its happening across all my ai softwares

[modern oxide]

Do you have the latest Nvidia driver?

[sturdy yacht]

yes

[rare goblet]

Ayo? @sturdy yacht level 3 !!!

[sturdy yacht]

correct ever since i updated my driver ive had these issues

i also used DDU

to uninstall my nvidea display driver and reinstalled it

still no change

[modern oxide]

Can you share the screenshot of what it says when you try to launch it?

[sturdy yacht]

this is rvc

this is stable diffusion

may i also mention this was all working before

Stable diffusion same problem as RVC

[modern oxide]

What fork do you use?

[sturdy yacht]

https://huggingface.co/lj1995/VoiceConversionWebUI/tree/main

[modern oxide]

Rvc1006nvidia?

[sturdy yacht]

nope

the one highlighted

[modern oxide]

Try to use rvc1006nvidia

It's an updated version

[sturdy yacht]

ok

i think when i had that downloaded it was complaining abouyt something else for some reason

[modern oxide]

If you have anti virus then just turn it off

[latent thistle]

where is to download the voice changer? any one can help me?

[modern oxide]

#🔍│help-w-okada pinned messages

[sturdy yacht]

question would physical damage or physical interference such as dust could cause this problem

it launched but

still failed to load cuda

M:\RVC1006Nvidia>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897
2023-11-10 19:58:42 | INFO | configs.config | No supported Nvidia GPU found
overwrite preprocess and configs.json
2023-11-10 19:58:42 | INFO | configs.config | Use cpu instead
is_half:False, device:cpu
2023-11-10 19:58:45 | INFO | main | Use Language: en_US
Running on local URL: http://0.0.0.0:7897

[modern oxide]

No idea why it doesn't work

Maybe others will understand

[sturdy yacht]

hmmm

do you think it could be something wrong with my gpu

also there might be some dust buildup physically could that pose an issue

[modern oxide]

No idea honestly, it looks like the program just doesn't see your GPU

[winter seal]

when i train a model tensorboard closes when it reaches an epoch with the number 4

or when it saves

[boreal depot]

guys it says "no module named 'gardio' "

[rare goblet]

Ayo? @boreal depot level 1 !!!

[old hawk]

even tho i put in the vocals (wav file, vocals only), the model and the method as harvest it still says "Please select a model and input audio file" when i try to convert. can anyone help?