✨│ai-help

[odd shale]

Maybe 25-30 mins or a bit more.

Or whatever amount you're able to gather/use

[analog obsidian]

the longer the dataset the more natural its going to sound
quality itself is only related to the dataset, so if you have like 5 minutes of audio but the quality is high, the model is going to sound high quality, tho is going to lack the natural sound of a longer dataset

[odd shale]

WTF?

💀

[analog obsidian]

[low shard]

isn't there a limit to discord connections

ngl i keep thinking about pewdiepie when i see those instead of riot

[odd shale]

I'm not sure.

[fading lodge]

i like throwing punches with my profile

and i like valorant

[warm sandal]

Same

[rare goblet]

Ayo? @warm sandal level 1 !!!

[warm sandal]

Except I’m dog

😔

[low shard]

did u do it only for the discord connection

[fading lodge]

um

yes

[fading lodge]

dw im harstuck b2

😭

[warm sandal]

How come it won’t let me download the Google drive

[low shard]

thats so many accounts

[warm sandal]

I’m not even ranked yet

[fading lodge]

its only like 45

[warm sandal]

😭

[fading lodge]

i was gonna add like 70

[low shard]

u will have no issues on colab if u finish your daily gpu quota 😭

at first i thought it was the discord connection rare name exploit

[fading lodge]

nah

i just made alot of accounts and connected all of them

will 34 quick sounding audios work to make the voice half decent? its not over 20mins of audio but it is around 10-15min i belive

thats in total for all of the audios

[odd shale]

If it's clean enough it will sound good.

[fading lodge]

alright

perfect

so everythings going well so far but

not sure how to do this

i did audiacity effects n everything

@odd shale

[odd shale]

How to do what?

[fading lodge]

idk how to start datasets

its confusing

[odd shale]

Nope, it's not.

Do you at least have audacity on your PC?

[fading lodge]

yes

i did the effects and everythhing

[odd shale]

What effects?

[fading lodge]

effects -> noise removal/repair -> noise gate

[odd shale]

Alright, that cleaned audio you want to use, you can apply audio labeling to it.

[odd shale]

You're already making a dataset.

The dataset is the audio you'll use.

[fading lodge]

oh

sorry

didnt know

what i am stuck on tho

is this

the thing circled

it tells me to put that in RVC

idk where

[thin stump]

@low shard lol rip all my imgur images were taken down on the guides. I have to reupload this really quick

[low shard]

shit

[odd shale]

First of all, which is the length of the audio you'll use? You can apply audio labeling by going to Analyze -> Label Sounds, then put these same values:

[low shard]

imgur takes them down?

idk if u can use data uri in rentry

[thin stump]

might have something to do with me making a new post. I'm pretty sure im shadowbanned on that account

[odd shale]

Just look at the numbers.

After putting the same numbers as in this config, hit "apply"

And then go to Export -> Export Multiple

[fading lodge]

well im not sure what im doing really

[odd shale]

@analog obsidian Muy bien mijo ahora si puedes proceder.

[fading lodge]

is the length like how long the audio is

[odd shale]

Yes.

[fading lodge]

about 1-2 seconds

for each

or 3-4

one i got noqw

is like

2 seconds

[analog obsidian]

usted es el helper

[odd shale]

Jaja si cierto XD.

Me se olvidó.

[analog obsidian]

[odd shale]

Solo que parece que STAR no sabe que es lo que hace.

[low shard]

i just tested and image uri seems to work all fine on rentry, maybe use that instead

[odd shale]

I'm confused.

[thin stump]

alright i'll check it out

[analog obsidian]

https://discord.com/channels/1159260121998827560/1203830360878743612 aqui hay una explicacion en ingles

[odd shale]

Gracias Lyery.

[low shard]

example:

![test](data:image/png;base64,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)

I remember using some websites like https://ezgif.com/image-to-datauri

[odd shale]

I'm not understanding.. But there you got a explanation about how to use Audio Labeling

#1203830360878743612 message

[fading lodge]

Alright

[still stone]

how can i set applio up so it has a share link?

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

like.... how do I do that

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

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

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

[still stone]

I'm running it locally though, so is the thing my cmd displaying just not a thing?

[thin stump]

I found a permanent hosting service. Apparently it's not just me because Niah's gif was deleted too

[low shard]

oh is it free?

[thin stump]

https://postimages.org/ looks like it's just shared to you only which is way better

[honest cobalt]

-realtime

[azure marsh]

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

[rare goblet]

Ayo? @honest cobalt level 2 !!!

[honest cobalt]

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[marble rivet]

yo star

@fading lodge what ever you gave me bro my pc is getting hacked bro

[fading lodge]

wdym??

[rare goblet]

Ayo? @fading lodge level 7 !!!

[fading lodge]

you sent me some wierd stuff

bro

[marble rivet]

Ik i got hacked

And i just lost 1000$

[fading lodge]

bro thats your problem not mine

idk why ur coming to me for that

dm me

for this

not here

[marble rivet]

Ok

[woeful cave]

i got this in the applio colab, is it not working anymore😭

[fading lodge]

bro got scammed 😭

idk why he dmed me about that on here

[knotty moth]

wat 💀

[fading lodge]

huh

i was copying what u siad lol

after u deleted it

[knotty moth]

u look sus, why my message was red on you? ||(vencord?)||

[fading lodge]

yes

i used vencord

[magic badge]

Applio interface not works

[thin stump]

the new kaggle mainline guide. Now with 20% larger images https://rentry.co/RVC-Mainline-Kaggle

imgur sucks

[fading lodge]

what is this for? just asking

im curious

is it a fork for rvc?

[thin stump]

training voice models with 8 hours of GPU daily. It's the mainline fork by Hina/niah

[fading lodge]

ohhh

[thin stump]

a better alternative compared to colab so I just try to recommend that instead of telling people to use rvc disconnected

[shadow sparrow]

is applio colab down right now? links arent showing when I start it up

[knotty moth]

it is 12 hour daily session and 30 hour weekly GPU quota

and the weekly quota reset is at Saturday 12am GMT+0

[lethal basin]

@acoustic scarab PROVIDE help pls

@acoustic scarab

pls

[sleek grail]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

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

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

[spare owl]

yess help meeee!!!

[mellow kernel]

I DONT UNDERSTAND

[carmine hearth]

Tell us more about what you need help with

[mellow kernel]

do you have enhypen sunoo voice ai

[carmine hearth]

I found the model, but the URL is not uploading for some reason, I'll send it to you in a DM.

[brittle wing]

when I open MMVCServerSIO.exe

Exceptions.PretrainDownloadException: "Failed to download pretrain models."

Press Enter to continue...

[pastel oak]

Is this original wokada or deiteris-Fork?

[brittle wing]

Fork

[rare goblet]

Ayo? @brittle wing level 2 !!!

[brittle wing]

But I fixed by installing Python ^^

3.11.9

[knotty moth]

delete stored_setting.json and pretrain folder, then try again

[pastel oak]

Interesting, there is usually no need to additionally install Python on these releases...

[brittle wing]

Why

[pastel oak]

the fork can utilize 3.10+

Original wokada v1 might have issues with anything above 3.10 though that might be true, but this is the Fork anyway

[pastel oak]

Just in case: If you do run into the issue with: Failed to download or verify: ... which I first thought was the case, then make sure to check out this section of the guide:
https://rentry.co/forkvoicechangerguide#press-enter-to-continue-failed-to-download-or-verify-files

If not, then ignore

[knotty moth]

well, that might be for the original one/older

[pastel oak]

Yup that is true

[mellow kernel]

yall how do i text to speech

guys help

[mellow kernel]

what next guyus

helpppppppppppppppppp

meeeeeeeeeeeeeeeee

[tame mica]

just unzip 😭😭

if you're fine with cloud you can use ilaria rvc if you just want tts/inference

[odd shale]

If you want, just use Ilaria RVC on huggingface.

[vale latch]

i've been trying to use ilaria rvc for 15 mins now and there's no gpu :')

[low shard]

does it say something about retrying later?

or does it error out in 60 seconds

[vale latch]

error out in 60 seconds, i keep refreshing and stuff. i assume there's no gpu available which is fine but idk if theres much point waiting

[low shard]

that's an issue of ZeroGPU, seems like there are alot of ZeroGPU spaces being used rn as seen by the top 5 ones

could u retry now?

[vale latch]

sadly i've now reached the gpu quota, so i have to wait :') ty for trying to help tho

[rare goblet]

Ayo? @vale latch level 1 !!!

[fringe summit]

yo can somebody help me

[serene hull]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• 🆕 Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

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

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

[rare goblet]

Ayo? @serene hull level 4 !!!

[low shard]

yw

[low shard]

whats the issue and what u using

[pastel oak]

Already helped nokiakiller all g

[low shard]

alr

[serene hull]

smth wrong with Applio today

[knotty moth]

yea dont use the colab

[low shard]

what happened?

[low shard]

something happened to the colab?

[knotty moth]

I suppose the package conflict issue that the pip install 23 might prob solve it, havent tried it tho

[low shard]

it doesnt install?

lemme test it rq

@nocturne mural

[serene hull]

can somone recomend me smt similar to applio?

[low shard]

for training (make models) or inference (use models)?

[serene hull]

inference

pls

[low shard]

For inference, you can use Ilaria RVC Zero which is the fastest one running on a better gpu (A100)

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[serene hull]

thank you so much

[low shard]

your welcome

[low shard]

just tried, doesn't work

I think it's because it uses a prezip rather than pip installing them, but idk much cus the encryption makes it hard

[dreamy kindle]

umm.. i'm trying to reach to the ko-fi link of this (ASMR E-Girl Whispering - RVC V2 Ilaria) but it doesn't work at all

[low shard]

what link ?

[dreamy kindle]

the download link to this model

[tame mica]

paid model perhaps

[dreamy kindle]

i thought so, but where could i find it i tried to look for it on ilaria shop but it's not there

[rare goblet]

Ayo? @dreamy kindle level 1 !!!

[low shard]

what download link? are u putting a ko fi link in rvc?

[dreamy kindle]

no i do not, the ko fi link just directs me to ilaria profile

[low shard]

where did u find the kofi link tho?

[dreamy kindle]

on voice-models channel

[low shard]

could u share the link?

u can do it via the button next to 'follow' in the model post

nvm found it https://discord.com/channels/1159260121998827560/1194034332143190086

@proven hill seems like that model kofi link doesn't work

[dreamy kindle]

great i was having a trouble lol

[low shard]

no i meant i found the model post, so i told ilaria that the link doesnt work so she can tell if she deleted it or will fix it

[dreamy kindle]

yeah.. how could i find it

[dreamy kindle]

okay ^^, thank you so much

[teal dome]

is something wrong with applio

it just stays stuck on an infinite loop

it was working fine 2 days ago

[dull orchid]

is there anything like Udio or Suno but than local on your own pc ?

some sort of DAW or program

[teal dome]

like, running currently? just audacity

i don’t use any other ai thing

i want to try “npm audit fix—force” but idk how to execute that or anything lol

[low shard]

Applio is broken at the moment

are you looking for inference (use models) or train (Make models)?

[teal dome]

ah alright

[teal dome]

just inference

[low shard]

you can use Ilaria RVC Zero for that, which runs on zerogpu (A100, better than colab t4)

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[fringe cradle]

my model is not working properly. Very late voice, inaccurate and doesnt work much

[pastel oak]

What program are you using?

[fringe cradle]

MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.16a i downloaded from github or hugging face dont remember

[pastel oak]

Alright, and what GPU are you using?

[fringe cradle]

wdym?

[pastel oak]

What is the name of your GPU?

[fringe cradle]

im on laptop. I don't really know

[rare goblet]

Ayo? @fringe cradle level 1 !!!

[pastel oak]

Task Manager > Performance

Check if there is a GPU1, tell me its name. If there isnt, tell me GPU0 name

[fringe cradle]

i have both of them

[fringe cradle]

GPU0: NVIDIA GeForce GTX 1650 Ti
GPU1: AMD Radeon(TM) Graphics

[karmic linden]

how to clone voice man google collab got banned

[pastel oak]

Oh lol yours is switched, glad I asked for both then:)

You downloaded the wrong version of the voice changer.

I will also suggest you use the modified wokada version instead of the original, which works great with GTX graphic cards to reach lower delay and better performance. Follow this guide, download the NVIDIA one
https://rentry.co/ForkVoiceChangerGuide

[fringe cradle]

thats the same program as i had?

[pastel oak]

Not 100%, the one you had was the AMD/Intel version of original wokada made by wok. The one I linked is an "optimized" version of wokada, made by a developer in this server called deiteris/Emojikage

[fringe cradle]

how do i download english version on that website?

[pastel oak]

https://huggingface.co/Shadicti/deiteris-Fork/blob/main/Windows-Nvidia-b2300.zip

[fringe cradle]

thats this?

[pastel oak]

Yes

[fringe cradle]

alright

[pastel oak]

In this, you run "MMVCServerSIO.exe" file to run the voice changer

For the ideal settings for 1650, you can use

F0 det: rmvpe or fcpe (try out both, fcpe can reach sligthly less delay)

Chunk: check what number you get on the "perf", its a colored number on the top left of the image. If you get for example 100, set chunk to 120-130, always a little bit above the perf number
Extra: 2.7s

Do this first before you hit "start"

Audio: select SERVER
Use the prefix [Windows Wasapi] on all 3 options where it says [none].

And then use your mic as your input, and your virtual audio cable as output. You can use monitor as your headphones to listen to yourself

Using windows wasapi reduces some more delay, if you want to know why I am suggesting it

[magic pendant]

Hi i don't know why i can't send screenshots

but I have a problem with just the sound

[sudden sigil]

what settings do i put for Voice Changer AI i thought there was a command to see what settings i can use

[pastel oak]

Send a screenshot, I can tell you what to set it to

[pastel oak]

You need level 1 on the server for screenshots. Can you describe:

What program you are using, what your GPU is, and then your issue

[sudden sigil]

Oh its cool

[pastel oak]

Oh, unfortunately I dont for the new alpha yet. I havent benchmarked it, dont think anyone else specialized in voice changer helping has used it all that much either, sorry

[sudden sigil]

@pastel oak oh okay ill screenshot these tho this is the one i orginally used but i forgot what settings are good

[teal dome]

wow that was fast as hell thank you (and thank you @proven hill)

[pastel oak]

F0 det: rmvpe would be better and faster. Rest seems fine, though 4060 can probably run 80 chunk aswell.

Also, using the Wasapi guide you can reduce some more natural delay of 80ms.
https://rentry.co/lessdelaywasapi

[magic pendant]

The problem is: I watch a video from a youtuber called "Duckus" and finished all the step (I now have the app) and he says to download VB audio what I did but even the speaker doesn't work, i normally use in input my microphone and output headphone but I changed the output to what is called cable input from VB audio (Weird because there is no cable output in output section) and even following the exact compostion of the youtuber still doesn't work fro me

I have a laptop with a rtx 4080 and i9 14 gen

[pastel oak]

duckus video is bad

dont recommend any yt videos

Also vb-cable is bad

[magic pendant]

oh ok

[pastel oak]

You probably also downloaded the wrong version of wokada I bet 😭

Here is what you can do:

[magic pendant]

maybe😭

[rare goblet]

Ayo? @magic pendant level 1 !!!

[pastel oak]

Original wokada guide (download Nvidia, download VAC Lite from the "Virtual Audio Cable" step)
https://rentry.co/VoiceChangerGuide

Modified wokada guide (this has some better performance, it can reach lower delay aswell than the original)
https://rentry.co/Forkvoicechangerguide

Not to influence you, but theres a benchmark test i wanted to see on the modified wokada and i was looking for someone with a 4080, if you would want to help id appreciate it 😂

[magic pendant]

yeah no problem

[fringe cradle]

after running mmvcserversio.exe and installing everything i got into a website. is it normal?

[pastel oak]

yes it runs on the webui. Its less bloated and runs better

[fringe cradle]

welp i cant choose model

[pastel oak]

You need to upload one first

[fringe cradle]

oh thats how it works. no free models ready?

[pastel oak]

On the guide i linked you can go to "voice models to try out" and download one of those to try

[pastel oak]

That was also removed cause its bloating up the program yea

[fringe cradle]

nah i already have one model

[pastel oak]

We out here for the efficiency 💯

[fringe cradle]

wdym virtual audio cable

[rare goblet]

Ayo? @fringe cradle level 2 !!!

[pastel oak]

You need a virtual cable to use the program on discord and games. The guide has a download link included

[fringe cradle]

VBcable?

[pastel oak]

vb cable works if you have it already

[fringe cradle]

i downloaded it like 30 mins ago

[pastel oak]

We do suggest the other one because vb cable can have random bugs, but if theres nothing wrong with it after you use it then it can stay

[fringe cradle]

what settings do you recommend me for the best accurate?

[pastel oak]

Start with

F0 det: rmvpe
Chunk: 140
Extra: 2.7s

And do
Audio: Server
input: [Windows WASAPI] your microphone
output: [Windows WASAPI] vb cable
monitor (optional) also wasapi and your headphones

[fringe cradle]

its hard to change chunk

[pastel oak]

Pitch is the.. pitch. if youre a man and using female model, set this to around 12

[pastel oak]

Doesnt have to be exactly 140 just roughly tgat number

[fringe cradle]

149.3 is good?

[pastel oak]

Dont forget to select your GPU instead of using cpu aswell

Yes

[fringe cradle]

i did

[pastel oak]

Now start, talk a bit then tell me what the colored perf number in the top left of the pic is

[fringe cradle]

any changes here?

[pastel oak]

Ah and do S.R. 48000

I also see already that your perf is 84

so Reduce the chunk to 110

Reducing chunk means less delay, but it has to be above the perf number and green for stability. For future reference, everytime you run a game aswell the number will go up because of resources used, so you have to adjust
Reducing ingame quality and capping fps can help

[fringe cradle]

i can only 104 or 120

[pastel oak]

You can move with your arrows once you clicked on the chunk slider first time

Use 120 then

[fringe cradle]

ok done now what?

[pastel oak]

Settings seem good, youre done

[fringe cradle]

it looks good but kinda easy to recognize its voice changer in my observation (in english)

[low shard]

yw

[pastel oak]

This might be voice model dependant. It can help to imitate the voices way of speaking more, else not every voice model is 100% suitable for realtime

[fringe cradle]

Well It's good if i speak english efficiently but in my language (polish) its different. its good in my language

[pastel oak]

Other languages can sound a bit unnatural sometimes, sounds may sound different which the AI either knows or it doesnt. Depends if the model you are using was trained on those sounds or not

[fringe cradle]

i used trained one

[pastel oak]

That is not the correct index file, the added is correct. Index forces the voice models accent more, you can increase index usage to 0.0 - 0.5 for a realistic result, but this is optional

[fringe cradle]

ill try added

1 min

[wise crater]

So I need a dataset for Tom Hanks's Woody. Does anyone have one, if so. Feel free to DM me. I am trying to create a Woody voice on FakeYou. So if anyone has a dataset, feel free to contact me. I know this probably isn't the right place for this. But I just want help with this.

[fringe cradle]

it sounds better

[pastel oak]

I dont think theres a point in asking for a dataset, because someone who created a dataset most likely also created the voice already

You can search rvc ai voice model:

• #1175430844685484042
• #🔍│find-models
• https://weights.gg/
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://applio.org/models
• https://voice-models.com/

if there isnt one, you can:

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

[fringe cradle]

idk if its only when monitoring but it kinda sounds kinda worse. is it just on self hearing and it will be better on talking with someone?

[pastel oak]

What do you mean with "sounds worse" exactly? Any recording or comparisons

[fringe cradle]

the voice sounds a bit worse now.

[fringe cradle]

idk if its me but when i test it now it kinda sounds worse than previously

[rare goblet]

Ayo? @fringe cradle level 3 !!!

[fringe cradle]

@pastel oak do you recommend any good virtual audio?

[pastel oak]

https://software.muzychenko.net/freeware/vac470lite.zip

• Run setup64 , not 64a, after extracting the zip to a new folder
• After installing the Virtual Cable, it changes your default audio system which is not something you want. Click on Yes when it asks you to open the audio device settings (or if you closed the window already, press WIN+R, type "mmsys.cpl"), and change your Recording and Playback devices back to your normal microphone and normal heapdhones. Make sure to do this for communications device aswell (right click your normal mic or headphones -> set as default communication device)

[fringe cradle]

which one is the lite?

[pastel oak]

Line 1

[teal dome]

i got a “no gpu is available for you for 60s”?

[pastel oak]

But, you are not putting it in the right slot

[teal dome]

i’m trying to get a take where the index is at 0 cause some words get pronounced better, and i cut those into the normal index take

it’s not working this time though

[pastel oak]

It does run on zero-gpu publicly so you may lose GPU access for a short time i guess

[fringe cradle]

oh i see

[teal dome]

oh that’s what it is

how do i bypass that

login with a different account?

[pastel oak]

Having an account doesnt make a difference I think, at least I am never logged in when i open ilarias rvc to test stuff

I dunno, maybe refresh or different browser but maybe someone else has a direct answer

[teal dome]

well i just tried with the index back at 0.7 and it worked fine so it looks like it’s related to the index being low

[fringe cradle]

my mic does not work with line1 as input device

[teal dome]

which is kind of an issue cause i need the index low sometimes to get some words pronounced correctly

[fringe cradle]

@pastel oak how to fix this?

[low shard]

it means that there are many people using zerogpu spaces rn, you can just retry dw

[low shard]

zerogpu is being destroyed by flux lol, it depends by how many people are using zerogpu spaces generally, as its basically shared gpus for all the spaces using the zerogpu

[fringe cradle]

can anyone help?

[pastel oak]

You mean you selected Line 1 on discord, NOT on voice changer right

[fringe cradle]

yes on discord

[low shard]

nope that doesnt work, when that thing happens it means that they are having problems giving u a gpu bc all are being used, you just need to retry till they give u one

[pastel oak]

Try this:

select Windows Start ,then select Settings > Privacy > Microphone . Select Change, then turn on Allow apps to access your microphone.

[low shard]

its not related to the index or any settings, its related to some moments the zerogpus are full, and some others are avaible

[fringe cradle]

its on

but doesnt work

[teal dome]

ah, i see

[pastel oak]

Try to switch to Audio: Client to see if that works at all

[low shard]

yea just retry till it works lol

[teal dome]

that makes sense, i can deal with that fs

it’s a lot better than not knowing the problem

thanks x2

[fringe cradle]

now it works on client

[low shard]

yw

[teal dome]

ugh… no gpus the second i choose 0 index

[pastel oak]

Not sure why it doesnt work on wasapi right now then. Maybe try to make sure your sample rates match, to see if this is the issue:

https://rentry.co/LessDelayWasapi

The sample rate of the virtual cable might not be the same as the mic

[fringe cradle]

is this how to make it sound better or smh?

[pastel oak]

Wasapi is to reduce delay

Not quality

"Extra" controls quality. 2.7s is usually the max but you can go higher to see if it helps

[teal dome]

will applio be fixed eventually cause this no gpu thing is getting pretty annoying

[low shard]

@nocturne mural should take care of that soon (sorry for ping man)

[teal dome]

sweet

[proven hill]

@low shard is rvc disconnected broken?

[low shard]

I haven't tested

it doesnt work for you?

[proven hill]

not me personally but a friend

[low shard]

does she get a specific error?

[proven hill]

@little scroll

[proven hill]

---

NameError Traceback (most recent call last)

<ipython-input-21-fd87a7218589> in <cell line: 6>()
4 import subprocess
5
----> 6 assert cpu_threads>0, "CPU threads not allocated correctly."
7
8 sr = int(target_sample_rate.rstrip('k'))*1000

NameError: name 'cpu_threads' is not defined

[low shard]

weird never seen that, seems that the error is by the cell Preprocessing and Feature Extraction, and that variable is defined in Set Training Variables (by looking at the code)

i will give it a test run with a random dataset rq

[proven hill]

lmk

[low shard]

i just uploaded a model zip and did feature extraction, and it works all fine

its better she tries to disconnect and delete the runtime and retry, there might be have been a problem in the initial setup or setting the training variables for her

[proven hill]

@little scroll

[little scroll]

oh ok will try

do i need to redo everything?

[low shard]

btw did u check the ping of before?

[proven hill]

no

[low shard]

yea you need to re setup the colab, go Runtime -> Disconnect and Delete, and re run all cells, be sure to not miss any

[low shard]

#✨│ai-help message

[little scroll]

ci provo

[rare goblet]

Ayo? @little scroll level 1 !!!

[proven hill]

the model was originally deleted because people didnt know how to use it

ill reup

[low shard]

lmao what

[low shard]

alright, just told u as someone could find it in #1175430844685484042

[little scroll]

this is so weird

it keeps saying the same problem

[low shard]

can u show a screenshot of your 'Set training variables'?

[little scroll]

can i show it in dm?

[low shard]

you have image perms here as you are level 1

[little scroll]

this is the error

i did it, i am the dumb one

my apologies

[low shard]

you sure you ran that?

[little scroll]

exactly what happened

that's *

[low shard]

the error was u forgot to run it right?

[little scroll]

i didn't run it

yes

i didn't

i'm dumb af

qwq

[low shard]

Oh lol, well its all fixed now, goodluck with training

[low monolith]

im new and dont understand why my voice changer wont produce sound

help pls

[proven hill]

enable sup2

[low monolith]

still nothing :/

[proven hill]

oh sorry i read it like it was producing random sounds

sens okada screenshot

[low monolith]

okada ?

[proven hill]

what are you using

[low monolith]

the voice changer

[rare goblet]

Ayo? @low monolith level 1 !!!

[proven hill]

send the screenshor of the voice changer

most likely its okada

[low monolith]

three secs having technical deficulties

[proven hill]

yea its okada

to hear yourself set monitor as your headset

also theres a lot of wrong configurations, what gpu do you have?

[low monolith]

idk

[proven hill]

😭

[low monolith]

ik im eally bad with tech

how do u check ?

[proven hill]

open task manager

then click on processes

and see what you have

[low monolith]

i think it sais gpu 0

what now ?

[proven hill]

can you send a screenshot?

[proven hill]

its an rtx 3060

youre using the wrong version of okada

[low monolith]

ok how do i fix that ?

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[proven hill]

download the first link

[low monolith]

i downloaded it

and it should fix all my problems ?

[proven hill]

kinda

let me know when you open it

[low monolith]

ok thx

[low monolith]

opened

[rare goblet]

Ayo? @low monolith level 2 !!!

[proven hill]

send screenshot

[proven hill]

set gpu as your rtx 3060

rmvpe normal, without onnx

[low monolith]

done

[proven hill]

extra 16k

chunk 192

[low monolith]

done

[proven hill]

now press start and the magic should happen

[cinder wyvern]

im using the applio colab and trying to blend two voice models together but im getting this error:

Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/gradio/queueing.py", line 532, in process_events
response = await route_utils.call_process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/route_utils.py", line 276, in call_process_api
output = await app.get_blocks().process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1923, in process_api
result = await self.call_function(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1509, in call_function
prediction = await anyio.to_thread.run_sync(
File "/usr/local/lib/python3.10/dist-packages/anyio/to_thread.py", line 33, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 877, in run_sync_in_worker_thread
return await future
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 807, in run
result = context.run(func, *args)
File "/usr/local/lib/python3.10/dist-packages/gradio/utils.py", line 832, in wrapper
response = f(*args, **kwargs)
File "/content/program_ml/core.py", line 422, in run_model_blender_script
message, model_blended = model_blender(model_name, pth_path_1, pth_path_2, ratio)
TypeError: cannot unpack non-iterable RuntimeError object

[proven hill]

also you didnt install vac lite

see this: https://rentry.co/VoiceChangerGuide

[low monolith]

i dont understand what u want me to do ?

and plus my voice changer really glitchy

[cinder wyvern]

An error occurred blending the models: [enforce fail at inline_container.cc:135] . file in archive is not in a subdirectory

am i doing something wrong?

[low monolith]

i think i fixed it thanks for the help

[nocturne mural]

now fixed, it seems I compressed the tar wrong.

sorry for not warning you earlier, I was asleep.

[covert hawk]

help

[low shard]

[covert hawk]

what i can do if the Google Colab of AICoverGen-NoWebUI, by Ardha (modified by Eddy) give me this error: "Exception: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx 1"

[low shard]

nah dw, also sorry that i ping you alot, but i dunno who else to ping when problems like this happen

[low shard]

You aren't using the T4 daily free gpu, Runtime -> Change Runtime Type -> T4

[brittle wing]

where's they newest guide for using the google colab one

[low shard]

what are you looking into?

Make models?

AI Covers?

[brittle wing]

use it

[low shard]

use what?

There are different programs

[brittle wing]

rvc google colab one

[low shard]

for inference (use models, like for ai covers) or train (make models) ?

[brittle wing]

use models

[rare goblet]

Ayo? @brittle wing level 1 !!!

[low shard]

for that, its better you use Ilaria RVC Zero, which is an Hugging Face Space that runs on ZeroGPU (dont duplicate the space), A100 which is better than google colab T4

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[brittle wing]

i meant to use it live

[low shard]

thats another thing, its called realtime voice changer for calls

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

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

here are all the options

for help ask in #🔍│help-w-okada or #1192011222023950368

[covert hawk]

okok

[rare goblet]

Ayo? @covert hawk level 1 !!!

[covert hawk]

but i'm using T4 GPU

...

[tame mica]

make an alt or try again tomorrow

[covert hawk]

ok

[magic pendant]

I don't know why but when using the start the voice changer only works on headphone output and when using the passthrough I change as told to line 1 (virtual cable) but i don't even hear my voice ( I can only hear it without a voice changer when going for headphone), I use the web because I don't know how to load the app (P.S: I already check the virtual cable driver and it's ready)

[pastel oak]

Change to "server" mode with the prefix [MME] and try to see if it works there

Else, we can have a deeper look at it tomorrow

[low shard]

you finished the T4 Daily Free GPU

You can either:

• Use an alt google acc
• wait tmr
• Pay for colab

[covert hawk]

TYSM, it work

[covert hawk]

thanks

[glad zealot]

• use kaggle

[low shard]

is there an ai covergen kaggle?

[vale osprey]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

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

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

[covert hawk]

how i get in the project folder if i did press this button

[proper yarrow]

can someone give me a place to use https://drive.google.com/file/d/1-KzrVfLFhjeZrwBIRSgaH2MZtSgma-1H/view this ai on all the hugging face ones say i have to use diffrent ones

[thin stump]

https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit Download the pth file and index file so you can upload a model for inferencing

[carmine hearth]

content

[covert hawk]

nvm i "resolve it"

[mellow kernel]

yall help me how do i download it

what key do i press

help meeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

plsssssssssssssssssssssssssssssssssss

yall help meeeeeeeeeeee

heloooo

helllo

hellllllooooooooooooooooooooooo

[misty elk]

-rx

shoot

[mellow kernel]

help

[rare goblet]

Ayo? @mellow kernel level 3 !!!

[violet heron]

You need to extract it

[misty elk]

-kaggle

[azure marsh]

📘 Kaggle Notebooks

• 🆕 Applio Notebook by Vidal
• 🆕 Applio Notebook by Shirou
• 🆕 Audio Separation by Shirou
• UVR5 NO UI by Eddy
• How to use RVC Mainline Kaggle by Cauthess
• ✨ RVC Mainline by Hina
• Original W-Okada's Voice Changer Kaggle
• Modified W-Okada's Voice Changer Kaggle

Note: Kaggle limits GPU usage to 30 hours per week.

More commands: -colabs, -hf, -local, -rvc, -help

[mellow kernel]

ok how do i put the voice

and what do i do know

ok what next

guys

guys

hello

guys

[misty elk]

use the download feature

wait

do you want to create a voice

or

make your own

[mellow kernel]

um i just wann put the voice

[misty elk]

ok use the download feature and import the model you wan

[mellow kernel]

what download

[misty elk]

it's next to plugins

[mellow kernel]

i dont how

[misty elk]

in the top bar

[misty elk]

yeah so

[mellow kernel]

ok know

[misty elk]

do you have 1 or 2 links

[mellow kernel]

1

[misty elk]

okay import the link and name it something

[mellow kernel]

ok i cant

[glad zealot]

Ye the very old one that I made, not checked tho if it still works

[mellow kernel]

look

@misty elk

[violet heron]

To put the voice you have to put the link in the download section

[violet heron]

Try removing the ?download=true part

[misty elk]

how do i use kaggle for conversion

[rare goblet]

Ayo? @misty elk level 11 !!!

[misty elk]

i mean for like

to make ai covers and shi

[knotty moth]

Ilaria RVC is easier to use https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit

[mellow kernel]

i did sitll doesnt work

[violet heron]

Have you checked that the voice model wasn’t already imported in inference

[knotty moth]

uh nvm, but is the link directly downloadable? and also check if the zip content structure is correct

[knotty moth]

alternatively, you can also add model files manually like this, and then click refresh models in the inference tab

[mellow kernel]

im too confuse

[knotty moth]

yeah, if you're not tech literate enough to know how to manage files, I can't help

https://tenor.com/view/gold-star-succession-gif-27712928

[fickle dust]

gusy ehlp i get this erro No GPU is currently available for you after 60 secs
on illiria zero

[fickle dust]

gusy ehlp i get this erro No GPU is currently available for you after 60 secs
on illiria zero

[late marlin]

guys which one is the best colab?

[random falcon]

Guys i need some help

[rare goblet]

Ayo? @random falcon level 2 !!!

[random falcon]

I hear myself when i speak, and my voice is not been convey to my WhatsApp even though my audio is set at Line 1

[knotty moth]

either Zero GPU was flooded by some flux workloads or you have reached GPU usage limit

[wheat field]

Hello, I was watching a video and when I typed "turk" in the voice-model channel, Turkish voice models appeared. I'm trying right now but I'm getting an error. Can you help me?

[rare goblet]

Ayo? @wheat field level 1 !!!

[wheat field]

Hello, I was watching a video and when I typed "turk" in the voice-model channel, Turkish voice models appeared. I'm trying right now but I'm getting an error. Can you help me?

[knotty moth]

that video must be a year old, and it was the old AI hub server that got mass bombed by RIAA

[wheat field]

Yes, it is a 1 year old video. Are there no Turkish models anymore? Or Where can I find Turkish models?

[wheat field]

I also searched for the name by typing it in Turkish. But I couldn't find it.

[knotty moth]

type some Turkish ppl's name you may know

[wheat field]

I was also able to give their names. There is no sound in Turkish

Many names remain durable

Could the sounds from 1 year ago have been deleted?

[knotty moth]

could you find those names in the video in weights.gg?

https://tenor.com/view/smacking-myself-squidward-tentacles-spongebob-squarepants-hitting-myself-ouch-gif-25583121

[wheat field]

There was no need, I found another discord channel with Turkish stuff.

[rare goblet]

Ayo? @wheat field level 2 !!!

[wheat field]

Thanks

[low shard]

read https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit#heading=h.bhen2m4ubvsz

[low shard]

not the gpu usage limit, but zerogpu being busy https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit#heading=h.bhen2m4ubvsz

[low shard]

there isn't a best one, what are you looking for? inference (use models) or train (make models) ?

[late marlin]

inference

[low shard]

never heard of it

[low shard]

use ilaria rvc zero that is an hugging face space that runs on ZeroGPU (A100, faster than google colab t4)

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[dreamy kindle]

sorry again.. is there any update about this model

[proven hill]

no

why its so important

[low shard]

Should I delete the model post then?

[proven hill]

whats the point

[low shard]

If it's deleted

[proven hill]

i said i will fix it

[low shard]

Alright if you say so

Just said that in case it's left broken making more people find a deleted model

[proven hill]

simply, i cant be 24/7 on discord fixing a free model no one uses anyway, ill do it when im free

[dreamy kindle]

okay sorry for intruppting, thank you anyway 🤍

[proven hill]

no prob!

[covert elm]

anyone have problem like me, my pc monitor black screen during using rvc voice changer, i dont see cursor mouse and cant force shutdown, help pls

[tough fiber]

https://github.com/Anjok07/ultimatevocalremovergui guys anyone know is it good?

[tame mica]

everyone here uses that btw

[tough fiber]

oh rly

[tough fiber]

so do u know settings

[knotty moth]

read the comprehensive guide here https://docs.google.com/document/d/17fjNvJzj8ZGSer7c7OFe_CNfUKbAxEh_OBv94ZdRG5c/edit

[tough fiber]

omg

[oak lily]

i start crsahing when its calibrating

[tough fiber]

bro how can i read

thats whole book

[knotty moth]

what do u want to do with the UVR? like extracting vocals from a song? the model recommendations are there

[pastel oak]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

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

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

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

__**BE PRODUCTIVE.**__ 🤝

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

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

[tough fiber]

i mean theres youtube video one of famous youtuber. she playing games and theres game musics or voices. i want to seperate her voice and game voice. so i need her voice only

lalal.ai is really good

maybe ill buy subscription idk

if this one doesnt work good

i started once and my pc is going to burn lol

[pastel oak]

Did u try mvsep.com before its free and does it too

[tough fiber]

let me try i didnt see this one

[knotty moth]

I'd recommend some vtuber talking (雑談) streams

[pastel oak]

Make account on mvsep.com to skip 97% of the queues

Use BS Roformer -> ver 2024.08

[tough fiber]

ill try is it good btw

[pastel oak]

Why would I recommend it if its bad

[knotty moth]

that seems rather old, probably still stuck on demucs arch, instead of the best roformer models

[pastel oak]

I use it too

[oak lily]

oh mybad

[rare goblet]

Ayo? @oak lily level 1 !!!

[tough fiber]

trying

thanks btw <3

[pastel oak]

Dont forget to create an account to skip most queues

is free

[tough fiber]

well it says 285mb is too much

or should i convert mp3

[pastel oak]

need to chop it up to smaller parts then

[knotty moth]

maximum 10 mins and 100 MB, you could upload it as flac

[tough fiber]

well its 35mins

[oak lily]

whenever i try to use the model i selected (it does the same with every model) it starts calibrating and then it just hard crashes and this screen pops up does anybody know how i can fix this?

[tough fiber]

is there any paid thing

[pastel oak]

Then do 4 files

[knotty moth]

you could always split the audio first (make sure not to cut the voice sentences)

[tough fiber]

so can i upload 4 files in a same time

[pastel oak]

mvsep has paid option

One file at a time for free users

100 minutes + 1GB + 10 files for paid

[knotty moth]

the paid option is just some convenience, and model ensembles (optional because you could do it manually)

[knotty moth]

that's why you should stop using that susware

https://tenor.com/view/wyaking-gif-9712475764034023502

[oak lily]

ohh wait i am in the wrong server

😭

[tame mica]

what were u tryna do ...

[tough fiber]

so can i upload 1hour data lol

and 2 gb file?

[tough fiber]

yea i just saw

it gives credit

i prefer subscription instead credit system

so i can pay lalal.ai for this only 10$ for month

its nothing and its okey for me

[knotty moth]

I'd say not worth paying if you could split the audio first

[tough fiber]

but let see

[tame mica]

no more using lalal

lalal is for tech boomers that don't know a better separator exists

[tough fiber]

i tried lalal on 1min data and its pretty okay for me idk

[tough fiber]

yea better ones getting more money

[knotty moth]

imagine spleeter

[knotty moth]

srsly, is it that voice.ai garbage perhaps?

[ivory solar]

Yall how do I send photos here

[rare goblet]

Ayo? @ivory solar level 1 !!!

[ivory solar]

tf?!?!

Lol

Guys fr

[low shard]

now that you are level 1 u can send pics in help channels only

[ivory solar]

Oh okay

[low shard]

u need level 5 to send pics in #🧬│ai-chat sans

[ivory solar]

So I extracted the file. What do I click?

[ivory solar]

Oh thanks

[low shard]

Are you using https://github.com/Tiger14n/RVC-GUI ? Because its outdated

[knotty moth]

who tf uses that old program?

[ivory solar]

Ill check wait

[knotty moth]

-gui

[azure marsh]

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

[ivory solar]

Bro I dont undertsand😭 Ill just download it

[low shard]

That program you are using is outdated

[ivory solar]

Oh.

ok

[low shard]

tell me what program (like use models, make models, use models in calls) you want and i can help you

[ivory solar]

Okay

[low shard]

Also, other than what program you want, say your pc gpu so i can tell you if you can do it on your pc or have to do it on cloud
(You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU)

[oak lily]

yes it is and it indeed is garbage idk where or how to download okada

[knotty moth]

https://rentry.co/VoiceChangerGuide

[ivory solar]

Amd radeon Vega 8 graphics

[sleek grail]

Does Applio work now?

[knotty moth]

that seems old, but u can run in cpu mode

[ivory solar]

Ill try. Its downloading atm

2 hours. Aw man

[old pulsar]

how

[rare goblet]

Ayo? @old pulsar level 3 !!!

[old pulsar]

how to make ai cover ?

[knotty moth]

btw ilaria rvc is an easy alternative to do https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit

[ivory solar]

ill just use it as an alt. if this doesnt work

Thanks a ton

[wheat light]

whats the best female voice i can use i have a 3090

[umbral iris]

guys

I need help about the app

[ocean thicket]

I need to see rank

[low shard]

use /rank

[rare goblet]

Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote

[ocean thicket]

thank you

[low shard]

yw

[rare goblet]

Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote

[foggy roost]

Need help with Ilaria RVC.

Im doing a project right now and im trying to automate the entire conversion process.
I need help with utilizing ilaria from a seperate python program that I can send messages to be converted into an mp4.
Hoping that Ilaria has some sort of API but i dont know where to start looking.

[low shard]

Nope Ilaria RVC doesn't provide an API

[foggy roost]

Dang, u know of any other RVC that provides TTS and also has a API support?

[low shard]

If what you were looking for is TTS, ilaria rvc uses microsoft edge tts api to produce a tts audio to then convert it to the voice (this means edge tts only makes tts without the rvc models, it uses basic microsoft edge tts models, you can use that for free tho)

[low shard]

afaik nope, things like @proven hill RVC (sorry for ping, its just in case im wrong) and other RVC forks are Open Source projects, they aren't made to be used as an API for other websites

[foggy roost]

Dang, Thanks for the help tho!

[low shard]

yw

[fleet cedar]

hey can you rate my accent and tell me what accent does it sound like?

since im arabic and my accent is arabic strong accent

i changed the index

[low shard]

the voice kinda sounds robotic

[opal kelp]

Guys, I have a dataset of 1 minute and 19 seconds, how many epochs do I train with?

If it's 250 or 300, I tried it, but it felt a little robotic

If you want me to post the results, just let me know

[thin stump]

add more to the dataset if you could. You have to get that 1 minute really clean and train until the best g/total https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#--monitoring

[thin stump]

felt a little robotic
just a problem with your dataset. This is what I had with 31 epochs but the accent is off

[blazing atlas]

Can you help me set this up?

[odd shale]

Gwenpool??????

[thin stump]

https://tenor.com/view/gwenpool-fortnite-emote-gwen-stacy-gif-7200082616495047851

https://discord.com/channels/1159260121998827560/1274196791109222471 already done

I wouldn't have done one if it didn't came from Marvel itself

[ivory solar]

RVC-GUI-pkg-mp3fix.zip Is the new app right?

[rare goblet]

Ayo? @ivory solar level 2 !!!

[ivory solar]

The last version was on 2023/05/25 right?

Why is it so damn confusing? Or am I the idiot

[brittle wing]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

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

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

[brittle wing]

Is there a "Chills" RVC because I cant find it. the "Number 15 burger king foot lettuce" guy

-audio

[azure marsh]

📚 Audio Guides

• Creating Datasets for RVC using iZotope RX11, by Cauthess
• Perfecting Audio Isolation on Low-End Rigs, by Litsa The Dancer and Faze Masta
• Gathering and Isolating Audio, by SCRFilms :snowflake:
• Vocal Mixing Tutorial, by Roomie

🛠️ Tools

Audio Separation/Isolation

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

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

[brittle wing]

-uvr

[azure marsh]

Ultimate Vocal Remover

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

[brittle wing]

-hf

[azure marsh]

<:huggingface:1179800228946268270> Hugginface Spaces

• UVR5 UI, by Eddy and Ilaria Huggingface Spaces
• Ilaria RVC Zero, by Ilaria Huggingface Spaces
• RVC⚡ZERO, by r3gm Huggingface Spaces
• Applio, by IA Hispano Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[brittle wing]

-colabs

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer Google Colab

ℹ️ Note

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

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

[misty rune]

where do u actually download it

[violet heron]

it’s linked

[misty rune]

i need a youtube tutorial 😭

alot going on in there

[low shard]

there isn't any updated ones lol

[misty rune]

dang

why is the entire like

program

foreign

[knotty moth]

that's why many ppl in modern era lack of reading skills

[modern hawk]

does anyone still have this model? https://discord.com/channels/1159260121998827560/1226387758872793139 seems like it got deleted off huggingface

[misty rune]

i dont lack reading skills js finding out where everything is loated

but allg now

[modern hawk]

tbh i think its just normal people finding things too advanced for them and jumping in clueless waiting to be spoonfed

thats why reading skills are going down

[knotty moth]

also youtube contents can't be updated unlike web articles, there's instead a need to make a new video

[cinder shoal]

How do I stop hearing the AI voice in my head?

[crude flame]

https://tenor.com/view/the-voices-cat-mad-going-insane-going-crazy-gif-24274574

[cinder shoal]

nvm I figured it out, it was because the virtual cable I was using was going to my headset

[glad veldt]

what's the best pretrain for a dataset that is under a minute long?

[brittle wing]

i'm trying to reinstall applio but it won't do anything

how do i remove the venv

the folder doesn't even have a .venv i don't think

WARNING: pip is configured with locations that require TLS/SSL, however the ssl module in Python is not available.

installing libssl-dev didn't fix this

[fleet cedar]

what voice is realistic? a male voice

[narrow sentinel]

by any chance is there a rvc with elevenlabs tts

[red cliff]

Can anyone give me some advice about picking the best checkpoint? I'm confused because the lowest raw value is almost at the end of the 300 epochs, you can see it at the bottom right. But there's a few spots around 700-800 steps where you can see the smoothed graph looks like it overtrains a bit

[wicked copper]

where do I need to put the .pth files in applio? I tried searching the documentations and couldn't find the answer
I previously was using the RVC webUI which normally puts in assets/weights

[tight kindle]

What's the best current RVC fork for local real time inferencing? Don't say okada pls.

[random bear]

nvm, it's okada, i'll go re-ask

[low shard]

That's old, it's better you see https://docs.aihub.wtf/

[low shard]

Like you are asking for a voice model?

[ivory solar]

Oh aight thx

[tame mica]

is this a question or an answer

[small sparrow]

What does index do?

[low shard]

controls the voice accent

[fresh temple]

https://www.youtube.com/watch?v=OeG7YQldTbo https://genius.com/The-kid-laroi-love-me-hate-me-lyrics https://www.youtube.com/watch?v=WVY7_HomSsM&pp=ygURaGF0ZSBtZSBraWQgbGFyb2k%3D

could someone remake this song using AI, i tried multiple times but my voice is too bad to do it (theres only a snippet of it) (insturmental, lyrics, and snippet of singing ^)

[fleet cedar]

yes

also what does dataset training mean

cause i always see it in model shops and saying that for sale thingy

[low shard]

You can search rvc ai voice model:

• #1175430844685484042
• #🔍│find-models
• https://weights.gg/
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://applio.org/models
• https://voice-models.com/

if there isnt one, you can:

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

[fleet cedar]

but what is dataset training for?

i see it from sellers or smthing

[low shard]

dataset = the audio file or folder with all audio files containing cleaned audios of the voice

model makers train the voice based on a dataset

[fleet cedar]

do model makers make a personal AI voice model based on how my voice sounds

[low shard]

there are people who can train your voice yea, you can do a free request in #1159289738314919936 and wait or do a paid commission in #1191429836321849435

[fleet cedar]

what person that accepts robux for a model?

[low shard]

i dunno

People usually accept money

[fleet cedar]

i hate that

cause i dont have real money

im too young for a job

[low shard]

I can't do anything about that 😭

You can try to see if any model masters accept robux

or wait for someone to do it for free in #1159289738314919936

or do the model yourself

[fleet cedar]

they ignore free requests except the paid requests

[fleet cedar]

bro i literally dont know how to do it

how do i learn

[low shard]

training takes alot of time (hours), most people rather do it paid

[fleet cedar]

to make a model

[low shard]

https://docs.aihub.wtf/essentials/how-to-make-voice-models/

Also, what's your PC GPU?
(You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU)

[fleet cedar]

my gpu is 4060

rtx 4060

[low shard]

Then you can do it locally

Check the link I sent you

And download either mainline or applio on your pc

[fleet cedar]

i just want a ishowspeed voice and in #1175430844685484042 i used it but its not accurate

cause of my shitty voice

[low shard]

you could try another ishowspeed voice

could depend by the model quality too, or u need to play around with the pitch

are u using it realtime for calls?

[fleet cedar]

yes

[fleet cedar]

bro how do i download voices from weights.gg

it shows me this thing

[pastel oak]

Theres a download model button on the dots

[fleet cedar]

i clicked it

NVMMM

IM SO STUPID

[fleet cedar]

bro? theres no index file or smthing

help

[fleet cedar]

it only shows this

help

[pastel oak]

model.pth is the model file

u can rename it

idk why the index file is missing

[fleet cedar]

nvm it works with only pth

its cause it only makes ur mic weird

[knotty moth]

everything in #1175430844685484042 is ensured to have correct zip contents

[fleet cedar]

i wanna pack ppl like packgod

[frozen iron]

hi, i haven't used AI and RVC for quite some time, is good old RVC still the most reliable to use for music/vocals?

and is it still best to just train a model the good old way locally etc

[low shard]

yes, nothing much has changed other than pretrains in training

nothing really that new other than that

[frozen iron]

god i really have to get people from ai hub poland to finally make some polish pretrains bahahah

[low shard]

there's a pretrain guide https://docs.google.com/document/d/1j9J8A8Oop9bMOHmCs3jDXzPujuD6TQ0Q396rJ0MyuIc/edit

i dont remember if theres a polish pretrain

[frozen iron]

yeah i've checked this one for good measure because last time i was using RVC i've tried Ov2

[low shard]

maybe in #1235952130855010365

[frozen iron]

and yeah i haven't seen any specifically Polish pretrains

i mean russian might work but eh im not risking anything for now

[low shard]

for russian there's Rigel & SnowieV3.1

[frozen iron]

oh and i see that those "new" pretrains dont use that much training info

like 10+ hours is not much compared to for example what rin used

are those actually valid and usable? :p

like is 10 hours really enough to do finetuned pretrained?

if so i might actually want to finally make a polish pretrained model

[frozen iron]

for example are those italian pretraineds any good? i suppose you'll know cause youre italian bahahahah

[rare goblet]

Ayo? @frozen iron level 3 !!!

[low shard]

most pretrains are finetuned rather than from scratch

[low shard]

yea, pretrains mostly help for other languagues as the original rvc pretrain is trained on only english

[frozen iron]

yeah ikik
here im a noob but in fact im the moderator at polish ai hub and made like at least 20 models back in the day 😭 so i kinda know my way around

and polish models on english pretraineds are basically 50/50, even if the model is great

sometimes it messes up our specific consonants

nonetheless thank you so much Nick for info

you da goat 🐐 🗣️

[low shard]

oh lol

[low shard]

pretrains aren't an ''rvc v3" but can help

[low shard]

btw for training pretrains it requires MUCH time and gpu

they are usually done locally

[frozen iron]

i do everything locally since day 1 of my RVC usage because collabs etc never worked for me, so i still stay old-fashioned and do it all on my poor rtx 3060

until im able to use checkpoints i should be fine (i hope)

[odd shale]

To be honest Nick, i got a project in mind.

For sake of fun, i'm compiling lots sessions of various artists on Audacity projects just in case someday a final version of Applio appears with new vocoders implemented.

[low shard]

seems cool

[odd shale]

Yep, and with those sessions i'll plan on making a studio session-based pretrain (or finetune in case a OG Pretrain version trained on a new vocoder like BigVGAN appears) for example, trained on BigVGAN.

I'll prolly work this with someone on the future.

[low shard]

goodluck

[odd shale]

🐢 ❤️

[soft stratus]

Is there a colab like easygui

[crude flame]

10 Hours of hq audio is the minimum for a finetuned. For a from scratch you need at least 50 hours of low to hq audio.

[low shard]

easygui google colab is broken, you can just use an alternative rvc fork #📰│dev-updates message

[fathom sun]

i wanna use rvc locally but dont have a gpu to train models locally theres some way i can train models not locally and use them locally ?

[golden walrus]

guys, do i need to make all of the voice i collect have the same volume in order to train ?

[odd shale]

Welp, yes.

If you're talking about having consistent quality on your dataset

Of course.

[golden walrus]

is there anyway to do it ?

i have no idea how to make these have the same volume

[low shard]

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

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)

Btw if you don't got a good gpu you shouldn't either use them locally (inference), for inference u can use ilaria rvc zero

[brave garnet]

Ilaria RVC: CLICK HERE 🤗

Guide on how to use it: CLICK HERE 📝

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

[odd shale]

Just use the parts where the audio got the same volume.

I'm not sure who you'll make a model of tho.

[golden walrus]

but isn't it make me less data for the bot ?

ye, i'm trying to make Ahri voice

[odd shale]

Bot?

[odd shale]

Quality > Quantity

[golden walrus]

i mean AI models

okay. i got it

[wheat light]

im trying to uninstall W okada but it requires permission to make changes even tho im the damn owner of my pc

[rare goblet]

Ayo? @wheat light level 1 !!!

[brittle wing]

i need help installing applio on linux

i ctrl+c when it was installing and i don't know how to get it to work again

it says .venv exists but it doesn't do anything after i type Y

[tacit arch]

how do i hear myself when i use the realtime voice changer

[unborn crest]

I have a question about merging models. Let's say I got a model with a singing dataset, pretty dynamic, wide vocal range. And another that has a not particularly expressive spoken dataset. If I merge them, will the resulting model be able to sing with decent range or would the speaking model actually ruin it?

I'm trying to figure out if it's worth bothering with the speaking model or if I should try looking for a different voice

[thin stump]

A speaking and singing dataset works. Just make sure the accent is mostly consistent

[thin stump]

I think more vocals on the monotone side may slightly lean towards that for RVC so you want variety too

[unborn crest]

Yeah, I want to make a more "unique" voice by merging two, but the audios I have from the second voice don't have a lot of variety so the model would come out very monotone

[analog obsidian]

merging a singing model and a speech model would not work really well, singing models already aren't good at speech to begin

and u could potentially limit the singing model vocal range

since speaking models reach less higher notes than singing models

what might work is merging the speech one with the singing one, could allow the speech model reach higher notes

but merging by a small amount

like 0.2

(yes, the order of the models in the merge matters)

[unborn crest]

Hmm I guess it's worth a shot, we'll see how it goes 🤔

And I had no idea the order was important!

Thank you for your answer!

[rare plinth]

would anyone mind explaining these sliders? (Respiration median filtering, Envelope ratio, and Consonant breath protection)

[rare plinth]

or at least give me a dumbed down version of it please my audio sounds like shit

[rare goblet]

Ayo? @rare plinth level 1 !!!

[rare plinth]

fuck you

[nocturne marten]

Hey, i've tried using applio but I get this error code in cmd whenever I try to use the TTS function
Traceback (most recent call last):
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\gradio\queueing.py", line 532, in process_events
response = await route_utils.call_process_api(
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\gradio\route_utils.py", line 276, in call_process_api
output = await app.get_blocks().process_api(
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\gradio\blocks.py", line 1923, in process_api
result = await self.call_function(
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\gradio\blocks.py", line 1509, in call_function
prediction = await anyio.to_thread.run_sync(
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\anyio\to_thread.py", line 56, in run_sync
return await get_async_backend().run_sync_in_worker_thread(
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\anyio_backends_asyncio.py", line 2177, in run_sync_in_worker_thread
return await future
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\anyio_backends_asyncio.py", line 859, in run
result = context.run(func, *args)
File "C:\ApplioTTS\Applio-main\env\lib\site-packages\gradio\utils.py", line 832, in wrapper
response = f(*args, **kwargs)
File "C:\ApplioTTS\Applio-main\core.py", line 318, in run_tts_script
infer_pipeline.convert_audio(
TypeError: convert_audio() missing 15 required positional arguments: 'formant_shifting', 'formant_qfrency', 'formant_timbre', 'post_process', 'reverb', 'pitch_shift', 'limiter', 'gain', 'distortion', 'chorus', 'bitcrush', 'clipping', 'compressor', 'delay', and 'sliders'

Anyone knows what should I do about it?

[rare plinth]

im probably wrong because i barely use applio but it looks to me some arguments aren't installed properly, so perhaps you could try doing this?

pip install --upgrade applio-tts

@nocturne marten see if that works

[nocturne marten]

sorry I don't use python often where should I run this command?

[rare plinth]

open a bash terminal with elevated permissions (just run as admin) and enter it

wait

idk if that will work

im stupid lol

give me a moment

[nocturne marten]

sure, thank you for the help :D

[rare goblet]

Ayo? @nocturne marten level 1 !!!

[rare plinth]

Audio processing

faiss-cpu==1.7.3
librosa==0.9.2
pyworld==0.3.4
scipy==1.11.1
soundfile==0.12.1
praat-parselmouth
noisereduce
audio_upscaler==0.1.4
pedalboard

Machine learning

omegaconf==2.0.5; sys_platform == 'darwin'
git+https://github.com/IAHispano/fairseq; sys_platform == 'linux'
fairseq==0.12.2; sys_platform == 'darwin' or sys_platform == 'win32'
numba; sys_platform == 'linux'
numba==0.57.0; sys_platform == 'darwin' or sys_platform == 'win32'
torchaudio==2.1.1
torch==2.1.1
torchcrepe==0.0.23
torchvision==0.16.1
einops
libf0
torchfcpe

Miscellaneous

certifi==2024.7.4; sys_platform == 'darwin'
antlr4-python3-runtime==4.8; sys_platform == 'darwin'
ffmpy==0.3.1
tensorboardX
edge-tts==6.1.9
pypresence
beautifulsoup4
flask
local-attention

i dont know which ones so see if you can update each

[nocturne marten]

sorry for asking a lot, but how can I check which one requires an update?

[rare plinth]

pip install --upgrade faiss-cpu==1.7.3, librosa==0.9.2, pyworld==0.3.4, scipy==1.11.1, ssoundfile==0.12.1, praat-parselmouth, noisereduce, audio_upscaler==0.1.4, pedalboard

[rare plinth]

no no it's fine! :)

pip list --outdated

if it doesn't work i really hope i'm not wasting your time, i use a mac not windows haha

[nocturne marten]

Okay, I get i should run all of these pip installs but what I still don't get is where should I run this command? Just open python.exe and run it there?

[rare plinth]

maybe?

[rare goblet]

Ayo? @rare plinth level 2 !!!

[rare plinth]

i think so

[nocturne marten]

Hmm it's not working, returns NameError: name 'pip' is not defined when I just run pip, otherwise
pip install --upgrade faiss-cpu==1.7.3
^
SyntaxError: invalid syntax

[rare plinth]

oh

that's very weird

[nocturne marten]

Perhaps there is another AI based application that can use w-okada trained voice models on tts?

[rare plinth]

do you prefer locally?

or a website to easily synthesize text?

btw check this source, it could be helpful https://www.alphr.com/install-pip-windows/