🧬│ai-chat

[chilly lake]

generally I'd say 'not totally there yet' still

[unborn spoke]

I'm looking for a voice changer but It got removed and I need to know if theres a way to get old ones that aren't on the discord anymore?

It was a Spanish egirl one

[robust valley]

hello

[livid bear]

-rt

[rare sorrel]

Interaction has expired, use the command again for a new interaction.

☁️ (Online) W-Okada - Colab Version

• How to use Original W-Okada's Voice Changer Google Colab
• Modified W-Okada's Voice Changer Google Colab
• Original W-Okada's Voice Changer Kaggle
• Modified W-Okada's Voice Changer Kaggle

⚠️ Notes

• Colab free plan GPUs tipically works for about 4 hours each day
• Kaggle restricts GPU usage to 30 hours per week
• These options may not work on mobile devices due to the lack of a Voice Audio Cable (VAC)

[coral ruin]

@woven gulch I use MIA AI to put lyrics in and get audio

Wrong channel. #1159290752195633273

[woven gulch]

I might love you chat

Where can i find MIA AI?

I only see a life coach 💀

[coral ruin]

https://apps.apple.com/app/id6503628462 if your not subbed you can get extra generation credits via inviting yourself or sending invites to friends

[woven gulch]

im going to comit violent war crimes (its not on the google play store)

[coral ruin]

;-; I tried,

[woven gulch]

Well thank you chat

[coral ruin]

There are many tho, I can send you an image of all the ones I use

[woven gulch]

Oh bet, thanks!
Any local ones/ones that can be run locally?

[coral ruin]

Not sure, I had to search a lot to find any who did it since I can’t sing my own songs without it always sounding like Hosana in the highest

[woven gulch]

I can sing I just cant stand my own voice (Im one of THOSE)

I was gonna use an RVC model I made of myself but with a darker tone so it sounds more like how i hear myself

[coral ruin]

That happens

[woven gulch]

Yeah it does lol

[coral ruin]

Maybe try Vocaloid too

[woven gulch]

Essentially, when anyone speaks you release vibrations. those vibrations makes it so my voice appears deeper and contrast-ier to me but to the normal person sounds..... normal. Still deep just not like how I hear it

[carmine canyon]

hi

[weary temple]

bros a homie

also got a senko pfp

[limpid rampart]

hi guys

is someone have applio?

[woven gulch]

Thus i cant stand it. So what i did was speak in a lower tone, sing bass songs and of course sing in my normal tone and then boom, satisfactory AI model using 5 hours of a dataset (chat it took 1:30 to make a 900 epoch model on my 3070 i might need an upgrade)

[woven gulch]

yuhhhhhhhhhhhhhhhh

[woven gulch]

christmas senko pfp goes hard

[weary temple]

i used to have dat same exact pfp for like 2 years straight

[weary temple]

i also got a matching shiro one too

[woven gulch]

W rizz

[weary temple]

fr

[woven gulch]

Bro I still cant belive i managed to reserve "senkobreadd" as my tag i was expecting to have to do some "Senko_BrEaD"

[weary temple]

ya crazy

[woven gulch]

;-;

[weary temple]

no bad

[vital gazelle]

free

[woven gulch]

bad boy

😠

Go to your cage

[weary temple]

fr

[woven gulch]

Im scared to look at your stable diffusion logs

[weary temple]

they arnt bad dw

am wholesome

[woven gulch]

oh okay, ur thing says "AI fetishes" so i assumed u was using that stable diffusion with some SPECIAL models (dont worry I dont judge)

oh shit? u into nintendo homebrew too? 😏

[weary temple]

no thats literally like level 65+ and being from da og server

[woven gulch]

This dude totally has at least 1 modded nintendo product

[weary temple]

i got da role from being in da og server

[woven gulch]

ohhh ok

[weary temple]

4

i got 4

[woven gulch]

Mine are switch, 3ds, wii, wii u,

[feral halo]

Ayo? @woven gulch level 11 !!!

[weary temple]

mine are switch, 2ds, 3ds, dsi, the snes mini

[woven gulch]

niceeeeee

Mariko or erista?

[weary temple]

and nes mini

dat too

[weary temple]

?

[woven gulch]

Your switch

[weary temple]

og launch

[woven gulch]

erista

[weary temple]

ya i havent touched it in 3 years

[woven gulch]

Lol

[weary temple]

so i dont rember

[woven gulch]

I have a mariko model (OLED) with a picofly

[weary temple]

W

[woven gulch]

512gb micro SD with like every game, and i have a backup 128gb micro SD that has android on it for the switch

[weary temple]

damn

i basically replaced my switch with a steamdeck

512gb

[woven gulch]

Makes sense, If i had a steamdeck id just run Yuzu (or i guess now "Suyu") with all my roms and stuff

[weary temple]

thast what i do

[woven gulch]

But if i did get a handheld id get either an Ayaneo 3 (leaked and will release soon) or ROG ally X

[fossil sparrow]

Guys I need help

[feral halo]

Ayo? @fossil sparrow level 1 !!!

[weary temple]

i would of gotten a rog if i travled more

[fossil sparrow]

My feinds AI voice changer broken

[covert lake]

Ask in #🔍│help-w-okada

[fossil sparrow]

It doesn't want to start voice changing

[covert lake]

!howtoask

[pine acorn]

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.

[weary temple]

@woven gulch

[woven gulch]

Yeah ROG is nicer IMO

[unborn spoke]

I'm looking for a voice changer but It got removed and I need to know if theres a way to get old ones that aren't on the discord anymore? It was a Spanish egirl one

[weary temple]

i know it is nicer overall

[woven gulch]

Id prefer steamdeck if better performance + Smaller borders on screen maybe.

[covert lake]

Holy 😭

[weary temple]

ya

[covert lake]

It's just a level role

[weary temple]

but deck has worse performance

[weary temple]

thats what i said

[covert lake]

As this isn't a NSFW server

[woven gulch]

I didnt realize mb mb 💀

[covert lake]

Actually most people here are minors

[weary temple]

exactly

ya

[woven gulch]

yeah

[weary temple]

there are really only a couple of us that are 18+

[woven gulch]

Is that why ava kriss tyson is in the server?

[weary temple]

including me

[weary temple]

who dat

[covert lake]

Yeah kinda rare

[weary temple]

then i am rare

[woven gulch]

sorry, "Ava" kriss tyson

Google it

[weary temple]

still no idea

[weary temple]

oh dat guy

[woven gulch]

yeah, hes some p3d0

[weary temple]

ye

[woven gulch]

I used to look up to mr beast man what happened

[weary temple]

no idea

all of them are crazy now

[woven gulch]

Mr beast getting new allegations every 2 days bro

This man is getting his business tarnished

AND he ran an internal investigation on HIMSELF AND HIS COMPANY

[weary temple]

bruh

[woven gulch]

its like the government hired someone to make sure the government is clean. "Guys we investigated ourselves and found NOTHING!"

[covert lake]

Maybe let's not talk about that here

[woven gulch]

ok ok mb

Ok changing up the topic then, do YOU know any AI (preferably locally runnable) that add AI lyrics to any song/beat pasted in? Like RVX but instead of making your speech sound like other people it layers whatever speech you want (TTS) to a musical rythem (song uploaded)

[covert lake]

So like an AI that automatically changes the lyrics of a song?

[woven gulch]

not changes the lyrics per say, but you put in a song with no lyrics and use an AI to make a speech track that perfectly works with said song. So on beat, ryhtmic to user's request etc etc

Like you type in lyrics and then the AI sings it to a beat of your choice like an uploaded song or beat etc (without pre-existing audio)

(pre exising audio as in speech or vocals)

[covert lake]

Ohh, I don't think there's any Local ones, maybe Suno or Udio can do that but I'm not sure

[woven gulch]

ahh ok

[quartz roost]

You mean like vocaloid? Concerts are held all around the world

[tawdry marsh]

do you guys think character ai going to import that advanced voice model thing on their website?

[quartz roost]

That can already be done

[glad elbow]

w

[tawdry marsh]

like an add-on?

[quartz roost]

Just output the text to the voice of your choosing

[exotic thicket]

Is anyone creating their own voice models, and using an api to create their own music?

[quartz roost]

I just use a DAW plugin for that

[exotic thicket]

But I don’t want to use loops.

[feral halo]

Ayo? @exotic thicket level 1 !!!

[exotic thicket]

@feral halo that the same as yes?

[quartz roost]

Place note, type lyric, draw parameters. Easy

[exotic thicket]

Would love to see a demo of someone doing this step by step and explain what software and/apis there using.

[trail nebula]

Suno ai can take keywords or full texts and sing them with a generated voice

[exotic thicket]

I was using RIPx for a while.

Ya I have suno and Donna. Want to change to something I can control more.

[quartz roost]

You play or write the midi, then have a voice sing

[exotic thicket]

But how does suno get sounds that actually sound like a band and not midi

[trail nebula]

Yes, I’m not sure if you can import any songs, I’ve not used their ai in months, but after giving the ai a prompt it will generate a song with those lyrics (can be any genre of music, just put what type you want.)

[quartz roost]

Suno does not give you ANY control

https://youtu.be/uVQJ7PPsiKA?si=oATModOT90l66SdM

[exotic thicket]

Agreed. That’s why I’m trying to see if there’s anything out there that will let you choose the voice model.

[quartz roost]

Where's the fun in that?

[quartz roost]

Hundreds of voices to choose from, or you can record your own

Rvc is just audio to audio conversion which is useful for a couple things

Welcome to the world of vocal synths, concerts are held around the globe

[exotic thicket]

I have a ton of ai songs that I wrote. Like the music that the ai generated but stem does a poor job separating the voices.

[quartz roost]

Render the voices from scratch

[exotic thicket]

If there was an ia that had separate voice and music and not have to try and take them apart after the fact.

[quartz roost]

Thats exactly what I'm describing

[exotic thicket]

One guy I spoke to said he used 2 different stem separators. Then put them in the day and it almost gets all the music back but was hoping for better quality. Then almost.

[feral halo]

Ayo? @exotic thicket level 2 !!!

[quartz roost]

Don't use a stem separator that reduces audio quality, use the original samples

These have existed for decades.

https://youtu.be/UiPNKFROPLQ?si=azGvdv_3nAxnwz5S

This voice just released. There's a free open source version too called diffsinger if you want to use your own voice.

[exotic thicket]

This sounds great. But it’s just piano. The bass, drums and guitars in my daw didn’t sound as polished as the ia generators.

[quartz roost]

That's just a voice demo for a vst plugin. kontakt and other DAW plugins have a ton of instruments

The best sounding instrument samples cost $

[exotic thicket]

Kontakt?

[quartz roost]

There's so many options

Suno is great for brainstorming and experimenting but can not make actual music

[exotic thicket]

Ya I’ve got songs I would like to re-record vocals and/or make karaoke songs out of.

You’d think the ai apps would figure it out. Just put them on separate stems from the beginning.

[wind stratus]

ive noticed that suno is very biased towards electronic music, theres always a melody drop/solo after the chorus is done

[quartz roost]

Suno does not "compose" anything. It's just a photocopier

Go download musescore and write something. And export the notes to any instrument you wish

[woven gulch]

I love you 😍

(No homo)

THANK YOU CHAT I THINK THIS IS WHAT IVE BEEN LOOKING FOR OMG

[wind stratus]

ikr you can always tell when a song comes from suno

[quartz roost]

I can show you something better

[woven gulch]

... u got smth better? 😏

Senko bread interested

Please proceed

[quartz roost]

What if I told you could place the notes exactly as you want. Type in the lyrics exactly as you want, and even fine tune them exactly the way you want it to sound

[woven gulch]

OH?

Its vocaloid ain't it 💀

[quartz roost]

Suno does not give you any control

[quartz roost]

Well there's realistic non anime voices

[woven gulch]

Damn so it does its own thing

[quartz roost]

Yeah, suno sucks

It's not any fun

And you can't sell it

[woven gulch]

Yeah that's why I haven't gone with vocaloid, I have vocaloid 5 but miku is the only good sounding vocaloid for some reason,

I wanted to use my RVC models I've made

[woven gulch]

Which voice is that

[quartz roost]

https://tenor.com/view/hatsune-miku-baka-triple-baka-gif-1416994965349618217

The red one

[woven gulch]

Ohhhhhhh

Okay,

Ik other than Miku there's the red haired on, I think one with yellow hair, a male one with brown hair i think

I forgot them, but I had vocaloid 5 with Miku's voice collection from vocaloid 3 & 4

It was a "legacy" voice model i think

[river adder]

god.. havent seen u in a min

[woven gulch]

Oh yooooo whatsup

I remember you, the biggest help i had when I was starting my RVC journey

[quartz roost]

There's hundreds

[river adder]

mhm you sure do ^^

[quartz roost]

Even free voices

[woven gulch]

10th anniversary I think or smth like that, it had a sumth smth anniversary on it

[quartz roost]

Meiko's birthday is today! She's been around for 20 years now

And yet there's still people who don't know the term vocaloid

[woven gulch]

So it was the 10th anniversary, o was right lol

[quartz roost]

10th anniversary of miku expo

[woven gulch]

I remember seeing tik toks of people talking about vocaloid software and miku and the comments are like "miku = vocaloid?" Thinking it was like a voice changer 💀

"Miku voice changer! "

[quartz roost]

People are stupid

raw samples will always be better quality

[woven gulch]

Yes

[quartz roost]

Watching people make rvc of separated acapellas amuse me, it's like let's reprocess the same audio 4 times and wonder why it sounds like ass

[weary temple]

same

[woven gulch]

💀

[quartz roost]

https://tenor.com/view/kasane-teto-teto-bread-guitar-synthv-gif-697271422526157925

[woven gulch]

Thats sadly so factual lol, no amount of resampling will make an ass source sound any better

[quartz roost]

Meanwhile the bread queen is going viral

[woven gulch]

That's why I always have to find high quality FLACs to extract vocals from, .mp3s sound like literally bum cheeks

[quartz roost]

Well the best method is to record in a studio, and label the audio manually before training the model

[woven gulch]

"I make RVC voice overs using .mp3s downloaded from yt to mp3!" 😇

[weary temple]

https://tenor.com/view/senko-gif-19340174

[quartz roost]

It's a lot of work to make a GOOD voice

[woven gulch]

Well yeah but I mean for AI song covers, unless you have access to their studio and raw vocals, for me going through high quality .FLACs is the only viable option to get somewhat good results

[quartz roost]

Like hatsune miku is global and going viral again and people are still like omg ai singers!

[woven gulch]

I only use .mp3s on obscure YouTube videos and memes

[quartz roost]

Teto is fuckin cracked out

[woven gulch]

Fr like, wasnt that one miku song made on a 2012 imac on like vocaloid 2 or 3? It goes like "miku. Miku. U can call me miku, blue hair blue tie hiding in your wifi"

[quartz roost]

Goes way beyond what a normal singer can

[quartz roost]

https://youtu.be/NocXEwsJGOQ?si=DZTw4FuLtpum2Wlx

Miku came out in 2007

[woven gulch]

Miku's vocals on G#8 = hawk tuah + still water

[woven gulch]

YESSSS FHAT ONE

[quartz roost]

Vocaloid came out in like 2003, and before that there was dectalk

[woven gulch]

No wayyyy, 2007 imac is even crazy, it wasn't even a Mac pro it was an IMAC

[quartz roost]

https://youtu.be/JGpf6B8EvVY?si=lih-hvfF8kH1-V3z

This can sing

Rvc has like one use

It can convert audio to audio...

[woven gulch]

Oh ive seen stuff like this, kinda like the ABC toys where u form the word or type it and it would synthesize it into speech

[quartz roost]

Which is useful for mixing

[woven gulch]

Or like that early 2000s tts program

[quartz roost]

https://speech.zone/courses/speech-processing/

Goes back a lot older than that

So, ready to play with the big toys?

[woven gulch]

Possibly lol

Ive already messed around with vocaloid before

[quartz roost]

Rvc ai covers are already a dead trend

[woven gulch]

Had voclaoid 4,

[quartz roost]

https://youtu.be/19y8YTbvri8?si=g8pKzSGgsBe90nE9

100 million views ain't nothing to scoff at

[woven gulch]

I jumped from RVC to vocaloid 4 and it felt pretty weird. It was cool but odd. I feel like I'm too stupid sometimes to use pro software (and my wallet says no to buying vocaloid 5)

[quartz roost]

Vocaloid 5 isn't even sold anymore?

[woven gulch]

WHAT

[quartz roost]

Miku is getting a vocaloid v6 ai next year

[woven gulch]

VOCALOID 6 IS OUT?

WTFFFFFFFFFF

[quartz roost]

It's been out for a while now

[woven gulch]

Its been that long huh

I gotta make a hit song using vocaloid like that one that made the whole internet go crazy for a week. Rabbit hole by deco9

I remember no matter what i did I'd always come across it on my fyp on tt

[quartz roost]

Lol trying to get big with vocaloid is the funniest shit

[woven gulch]

Not for money for fun,

[quartz roost]

Yeah for fun

[woven gulch]

I have my heart set on pc content

[quartz roost]

trying to profit with music kinda ruins it

[woven gulch]

I love when people make sick songs and make them royalty free

Like monster cat or pewdiepie that one time

[quartz roost]

The second most popular vocal synth... was given away for free

[woven gulch]

Wth what was it?

[quartz roost]

Teto

this is teto now

[woven gulch]

Vocaloid gained possession of it then, damn that's crazy

[quartz roost]

What? Teto is not on vocaloid

There's other programs you know

[woven gulch]

Oh mb I misread ehat is was reading

[quartz roost]

There's free open source vocaloid

[woven gulch]

I read that she was now a vocaloid. What it actually says is that she was used in UTAU

But it mentions vocaloid there so it tripped me out

[quartz roost]

Openutau is open source. The teto vocaloid was made in 2017 by crypton and it never released to the public.

[woven gulch]

Ah okay

[quartz roost]

It's free

[woven gulch]

I'll check it out

[quartz roost]

https://youtu.be/qBKfeXvPaJM?si=oWxf4Gl2VFUdYPR5

[deep glade]

H3LL0 GUY5!

[sleek merlin]

.

[umbral hemlock]

Hello everyone, I’m looking for a tool to automatically add captions to my shorts and reels. I did a search and found a couple, like ZapCap, which may be worth trying. But I’d love to know what you all use and your recommendations.

[plucky mirage]

💛

[quartz roost]

Whispering tiger ai on github

[loud kiln]

Wsp nerds

[exotic thicket]

I can all this with kits.ai

[ionic pumice]

kits... is not Good

[vivid pond]

hey does anyone know how I can make my own AI voice?

like any websites I can use that dont cost me money

or an app. again that doesnt cost me money

[hasty rover]

where do i download this

[covert lake]

What's ur PC GPU

[covert lake]

Download what

[vivid pond]

how do I check. I have not checked ina while

[covert lake]

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[uncut dew]

can anyone help me?

[vivid pond]

My GPU is called Nivida GeForce GT 1030

[feral halo]

Ayo? @vivid pond level 1 !!!

[covert lake]

Yeah not good enough for local training

[vivid pond]

dangit

[covert lake]

As you dont got a good PC, its better you use cloud (remote good pc) 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)
  • Applio by Vidal (UI)
  • Applio by Shirou (UI, no guide as of right now)
• Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
  • Mainline (UI, No guide as of right now)
  • Applio (UI, No guide as of right now)

[vivid pond]

never have I ever had so much respect for people who make RVC models

this is an absolute hassle

[covert lake]

I mean there are guides for it

Or else might use the 1 free training daily of https://weights.gg/ which is still RVC but easier

[lament leaf]

anyone got ideas to make youtube shorts with ai and just be able to post them easily

[stoic obsidian]

ss

[solar torrent]

https://tenor.com/view/momoi-sob-blue-archive-gif-15401761929204966295

[solar torrent]

AI-assisted auto-upload for YouTube doesn't seem to be a thing. But if you don't feel lazy enough, you can upload your video and schedule them by yourself.

For my AI videos, I can only upload them in a very traditional way.

[weary temple]

me daily

[fresh hazel]

beep.

Any way of getting rights to send pictures in VC?

🙏

Tesla P40 x 4

[weary temple]

more levles

[shut snow]

Can I have nitro please

[elder willow]

to train a model for singing, the audio has to have the person actually singing, right?

[weary temple]

no

[cerulean glade]

ok so i might be a bit autistic

but does anyone have an website where i can make ai voice cloning things

[wind stratus]

weights.gg

[cerulean glade]

ok thank you so much kind man i will pay u back later with idek

[wind stratus]

or a colab notebook for more precision

[solar torrent]

There's no such a thing of AI voice cloning, only training an AI voice model.

[cerulean glade]

oh

[polar flax]

https://tenor.com/view/no-please-no-god-no-stop-please-stop-gif-17190431

[tepid basin]

Kits bad

They have some payment for voice owners which is kinda cool. Trying to tackle the ownership issue. But if you want quality then no

[feral halo]

Ayo? @wide lava level 1 !!!

[fierce dirge]

donald trump

[amber light]

https://tenor.com/view/ken-carson-goomba-ken-carson-goomba-gif-7954500978985618872

[brittle sequoia]

Good

[dusty glacier]

You mean 𝕏??

[solar torrent]

#1159290752195633273

[dawn tinsel]

U

[ruby escarp]

[serene oak]

Hello

[covert lake]

#1159290752195633273

[solar torrent]

#1159290752195633273

[lost bough]

ok got it

[solar torrent]

Y'all really think this chat here is where you promote things?

[lost bough]

cause more members can see it

[feral halo]

Ayo? @lost bough level 1 !!!

[lost bough]

[olive plover]

hey everyone

[candid parcel]

hi

[elder willow]

any of yall on mac??

[exotic musk]

Hi! Just Called me Sedface 😦

[neon granite]

I need argos

[neon granite]

hello anybody here?

please answer

I downloaded some sounds I tried to use in voice ai but it says it haves to be mp3. or something like that
what program do you use?

@ionic pumice

[polar flax]

voice.ai ?

[neon granite]

yeah

I'm new

WHAT DO EVERYBODY USE

[ionic pumice]

huh

[polar flax]

that app is too old and paywalled, stop using it

[neon granite]

Ok

what should I use it

[ionic pumice]

project most people use these days is rvc

[neon granite]

then

[ionic pumice]

-rvc

[rare sorrel]

📚 Documentation

AI HUB Docs

https://docs.ai-hub.wtf

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[neon granite]

where can I download it

?

[ionic pumice]

read docs

[solar torrent]

Damn, there's a link in document.

[ionic pumice]

yuh

[polar flax]

it wants the results in poor quality unless you pay for it (seemingly)

[solar torrent]

I've told so many times that the voice.ai is the scam.

You better not follow any tutorial video on YouTube because most of them outdated.

If you want to run it on your PC, what GPU do you have?

[fierce aurora]

is it a rat..?

[solar torrent]

What are you talking about?

[fierce aurora]

nvm it aint dw abt it

[feral halo]

Ayo? @fierce aurora level 1 !!!

[polar flax]

https://tenor.com/view/ogey-rrat-pekora-usada-pekora-hololive-rrat-gif-24218218

[lament fiber]

whats the best voice changer / app to troll as a egirl

[elder willow]

Hello, new here. I wanted some information regarding the voices in "voice-models".
I am trying to create a project for a youtube channel, and I need a female voice. I came across the AI RVC and I loved it!
My doubt is: Can I use the voice models for Youtube monetization? Basically, the video contains mostly my voice, but also the female one. It's a complex project that I have in mind. With long duration.

I read most rules, but I didn't read anything regarding something like this. Anyone can clarify? I don't want to cause trouble. Nor anything related to copyright strike.

[solar torrent]

You can do anything for funny and not for commercial purposes. But I'm not sure how would you make money off from that.

[solar torrent]

The W-Okada. Do you like to run it on your PC or something?

[arctic zealot]

what is epochs chat

[rare sorrel]

Epoch: The number of iterations performed to complete one full cycle of the dataset during training. It's not possible to say precisely how many epochs you need for your dataset, you need to monitor the TensorBoard Graph to know if your model is overtraining.

[arctic zealot]

still doesnt explain what i should do with an epoch

[elder willow]

Long story LOL.
Anyway, do you know any software that can do that without copyright strike? Or just in general, you can't commercialize that?

[feral halo]

Ayo? @elder willow level 1 !!!

[solar torrent]

Whether if you pay for or get a free voice changer software, the result will be the same, so there's no difference between these two.

The free version that can do voice changing is better because you may not wanna pay a very large amount of monthly fee to use their software, but its setup can be complicated for the first time user.

[elder willow]

I see... You're very informative!
Is it okay if I DM you in private? I have some questions

[elder willow]

@solar torrent

[solar torrent]

Sure. But if it's FAQ, then no need to ask me in private.

[thorn shuttle]

who can help me get a picture of the goodbye and ridiance album of juice wrld that i can color in?

im curently in school lol

[solar torrent]

The album cover is literally a drawing itself.

[thorn shuttle]

but like one in just the lines not color

[feral halo]

Ayo? @thorn shuttle level 1 !!!

[solar torrent]

-rvc

[rare sorrel]

📚 Documentation

AI HUB Docs

https://docs.ai-hub.wtf

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[thorn shuttle]

? whats that link lol

[solar torrent]

Some guides to do the RVC thing on your PC.

[thorn shuttle]

im on my school laptop -_-

but im dutch and i have school until 13:35

[covert lake]

i think he wants an album to color

[solar torrent]

I didn't mean like that. I just wanted to screenshot that and send it to someone.

[covert lake]

?

[covert lake]

why would u wanna show someone that ur telling rvc guides to a guy that wants the drawing of an album 😭

[covert lake]

nothing, it’s a training info, not a thing u can change for wokad

are u looking to train?

and no, more epochs does not mean more quality

[solar torrent]

Because he was curious so I explained to him.

[covert lake]

huh the guy is talking about the pic of the album iirc

[solar torrent]

This is the higher quality of the album cover I downloaded from YouTube Music.

[thorn shuttle]

is there a way how i can get this in just the lines?

[solar torrent]

You can use a program or software to extract the line instead, since there's no only line of it available anywhere.

[covert lake]

Damn i thought spotify had the best quality

[covert lake]

btw why are u asking this in an AI server

[thorn shuttle]

idk lol

[solar torrent]

Every album cover on Spotify is just that. They have much lower pixels than every one from YouTube.

[feral halo]

Ayo? @solar torrent level 19 !!!

[solar torrent]

I'm not sure why. The maximum resolution I could get it from Spotify is 640x640.

[covert lake]

damn

[crystal pivot]

hi

[solar torrent]

The highest resolution for an album cover I could find from YouTube is 3000x3000, it's crazy.

[covert lake]

[feral halo]

Ayo? @fringe flame level 1 !!!

[covert lake]

use #1159290752195633273

[solar torrent]

#1159290752195633273

[bleak niche]

Hey, I have a question. Is learning AI special programming language (I mean scripting language created for the game) difficult? Are there any tutorials somewhere, examples of how to do something like that?

[feral halo]

Ayo? @bleak niche level 1 !!!

[shadow abyss]

Hi guys, so was reading that we can use discord on here to actually run vocals through different voices? can someone tell me how does that work?

[covert lake]

you mean to make an ai cover?

via /create of @hidden grotto in #🤖│bots

[chilly lake]

There's no AI special programming language. There's Python or LUA

[solar torrent]

I have only heard about AI used in NPC and characters movements, for example Minecraft. I'm not sure what you mean by an AI special programming language in a game.

[chilly lake]

depends on what you mean by a 'game'

[covert lake]

There is also a unity ai game course in huggingface docs if im not wrong

[shadow abyss]

so I am working on a project using my own vocals. I did harmonies on a track but was wanting to get a female voice which I don't have a female right now that can come into the studio so was thinking to run it through an AI voice model and blend it in

[chilly lake]

Minecraft uses mcscript

[solar torrent]

There are not any known generic female voice model here, most of AI voice models here have been trained on famous people.

[chilly lake]

you can merge 4 famous people voice models and get something generic

[shadow abyss]

the language is hindi. or more really bhojpuri. which is a language from india. so wana try it out to see how it would work. how would I go about doing that?

[solar torrent]

So while there's no such a thing of "AI programming language", you might either mean by the large language model (LLM) that used in chat bots or a programming language that could be written just like human language.

[solar torrent]

Well, I think someone has done that experiment.

[polar flax]

AI is not programming language, but you can find prompting guide for making unity games

[bleak niche]

Sorry, I meant not special language AI, but the game has its own script language. I would like to teach the AI this language. The game has script files. The language is a mix, some C#, some C++, java. I would like AI to be able to write scripts in this language. And I'm wondering if this can be difficult, and if there are perhaps tutorials for this somewhere, examples.

[solar torrent]

An artificial intelligence system is made of programming language, most notably Python as we all know.

[vital snow]

hi

[solar torrent]

Hi.

[polar flax]

finetuning an LLM for that may be quite expensive, but good luck tho

[chilly lake]

if you have about a billion properly written scripts, sure, why not

[bleak niche]

This means that at the moment it is not possible.... scripts are maybe about 50 files

[chilly lake]

LLM models learn probabilities of characters happening in specific order and they can continue predicting next sequence, thus 'generating' code

[feral halo]

Ayo? @chilly lake level 34 !!!

[chilly lake]

but with only 50 scripts you cant really teach a model how to generate anything sensible

[covert lake]

@atomic vector generate me python code for hello world

[atomic vector]

hey nick088 so you want a simple hello world in python right

it's just one line of code: print("hello world")

i won't give you a whole script for that but basically you can just copy paste that into a python file and run it
-# AI-generated responses may be inaccurate; please verify important information.

[covert lake]

Now, do me an example of an AI hallucinating while generating code

[atomic vector]

so i'm gonna try to generate some code for a simple login system in python... but i might make some mistakes

here's what i came up with:

import os

username = "admin"
password = "password123"

def login(username, password):
  if username == "admin" and password == "password123":
    print("you are logged in")
  else:
    print("wrong credentials")

login(username, password)

# now we need to save the user data to a database
database = open("database.txt", "w")
database.write(username + "," + password)
database.close()

# and now we can read the data from the database
database = open("database.txt", "r")
data = database.read()
username, password = data.split(",")
print("username:", username)
print("password:", password)

wait what's going on here... i think i just hardcoded the admin credentials and saved them to a plain text file... that's not secure at all

also, what if the file doesn't exist or can't be read? i didn't handle any errors... and what about multiple users? this code only works for one user

i think i need to go back to the drawing board and try again...
-# AI-generated responses may be inaccurate; please verify important information.

[quartz roost]

There are legal ways to use voice models you can profit with.

[elder willow]

Mind giving me an example?

[quartz roost]

there are also voices that are not anime related but as you can see, those dont sell as well.

[elder willow]

That's pretty cool actually

[quartz roost]

You can also just get a written contract with a voice provider and make your own for free.
My girlfriend and I made one of her voice.

[elder willow]

So, using a voice from the software and changing it a little bit is forbidden?

[rain compass]

Hello

[elder willow]

@quartz roost

[rain compass]

Where can I get help building odds in AI

[shadow abyss]

So if I wana just change my voice into another how would I go about doing so? and wher4e do I start

[quartz roost]

They are highly customizable

With multiple vocal modes

From whispering to shouting

If you're looking for a text to speech there's other options than the ones focused on singing.

[elder willow]

You mean RVC is highly customized? With whispers to shouting? Really?

[elder willow]

I saw some, but they are not useful for me unfortunately. Some Well made too

[lament fiber]

yo can someone help me? im trying to use the voice changer client from duckus and im using his tut but the vb audio isnt showing up in output would be appreciated if you could help me

[solar torrent]

The tutorial videos on YouTube are too outdated.

[quartz roost]

You can make a "multispeaker" which is making a shouting model, a whisper model, a normal tone etc...

[lament fiber]

oh could yuo help me?

[feral halo]

Ayo? @lament fiber level 1 !!!

[solar torrent]

-rvc

[rare sorrel]

📚 Documentation

AI HUB Docs

https://docs.ai-hub.wtf

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• How to use RVC Mainline Colab by Cauthess
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[lament fiber]

hwat is that

[elder willow]

Interesting! Thanks brother. I'll look for something like that

[quartz roost]

Thank you, I'm interested in helping people use this amazing technology for legitimate creative purposes.

[elder willow]

i need a egirl voice lol

[lament fiber]

so i js downloaded this 3gb voice changer client and setup'd it in like 2 hours for nothing 😭

[solar torrent]

There are newer RVC forks available, for example the Ilaria. But if you want to do the real-time voice conversation, you can go for W-Okada.

[elder willow]

npc ahhh 💀

[solar torrent]

Blud really thinks I'm an NPC lmao.

[elder willow]

Yes. I don't want to "steal". especially with this technology always evolving.

[feral halo]

Ayo? @elder willow level 2 !!!

[quartz roost]

The voices I use are friends of mine in real life and they are happy to let me use them

This has been around for decades with no big issues, until it became easily accessible to bad actors.

[real pagoda]

guys help me find a ralistic man voice

realistic

[quartz roost]

Singing or speaking

[real pagoda]

speaking

[solar torrent]

Ehh, there is no any known generic voice model here, only voice models here that have been trained on famous people.

[quartz roost]

Commercial use or for fun?

Yeah you can't use most "famous people" for commercial use

[real pagoda]

ill see

[quartz roost]

If only there was a repository that included the terms of use of each voice if it could be used commercially or if it's for personal fun use.

Commercial voices often cost $ because it's the same as hiring a voice actor

[lament fiber]

whats the best voice changer that i can use and add models into

[elder willow]

voice.ai

[solar torrent]

That site is a scam.

Who told you voice.ai was the best site to do voice conversation? That site charges your money to use their software and stuffs as a paywall.

[lament fiber]

whats the best rvc real time voice changer?

[solar torrent]

-rt

[rare sorrel]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[lament fiber]

-rt

[rare sorrel]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[solar torrent]

Every mod here recommended to use W-Okada on your PC if you have a super fast GPU.

[lament fiber]

i have a 3060 rtx with 12gbs on ti

[solar torrent]

Oh cool. I think you can run on it.

[lament fiber]

alright which one do i use

the first second third forth?

[vital snow]

i recommend fork first link

[solar torrent]

The one that says Nvidia or CUDA.

[lament fiber]

#🔍│help-w-okada

[lilac tapir]

onnxgpu cuda or onnxdirectML cuda for a gtx 1660 ti ?

[solar torrent]

The onnxgpu CUDA because your GPU is Nvidia.

[lilac tapir]

I downloaded it but somehow it doesn't work

when I do pass through

I hear my normal voice not the ai one

I have this one ''MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a.zip''

[magic hatch]

yo guys

how to make ai-coverd?

i blocked at colab step.

[prisma kelp]

hello guys , where i can translate song to another language or do something familiar to this ? https://www.youtube.com/watch?v=H83mq96J0ts

[solar torrent]

https://cdn.discordapp.com/emojis/1300208768428015747.webp?size=48

[vital snow]

you need : vocal isolated audio file of a guy u want to be converted
rvc model
and pick 1 of this
applio https://docs.applio.org/beginners/interface
rvcwebui https://docs.ai-hub.wtf/rvc/local/mainline/#inference-
ilaria rvc zero https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit?tab=t.0

[magic hatch]

thank you

is there some tutorial?

[solar torrent]

Doing RVC on Google Colab can be risky to get your Colab account banned.

[chilly lake]

use noUI?

[prisma kelp]

@vital snow i have model voice and isolated voice what i want change but how to change lyrics to another language ? like in this viddeo

[solar torrent]

That's real.

[magic hatch]

so, i have isolated audio file(made by UVR5) now what i need to do?
the tutorial in youtube said use colab

[feral halo]

Ayo? @magic hatch level 1 !!!

[magic hatch]

i picked the site, and than what to do

[solar torrent]

Tutorial videos on YouTube are kinda outdated, especially one from 2023 and early 2024.

[vital snow]

you have voice model already?

[magic hatch]

i found the voice model in this discord

[vital snow]

what do you rvc do you choose?

[magic hatch]

:/ i watched 2023 korean guide

https://discord.com/channels/1159260121998827560/1227635531958194266

[solar torrent]

[magic hatch]

this

i choose

[vital snow]

i mean the program

mb

[magic hatch]

i just did only one.. use UVR5 to make isolated vocal file

[vital snow]

uhhh sorry for not being clear again i meant what you going to use for Inference
like applio , rvc etc

[magic hatch]

ahhh
mb i dont know about AI model so i didn't do that you said already

i need some tutorial or guide for starter/noob

[vital snow]

i see. what is ur gpu btw do you want to do locally or on cloud

[snow shuttle]

I'm a student who want to learn some ai knowledge.

[covert lake]

everything that uses UI is banned on colab, but there's pretty low chance u actually get banned

[covert lake]

I think the only NO UIs are Applio NO UI & RVCDISCONNECTED

[covert lake]

What's your PC GPU?

[magic hatch]

3050ti it's not good xD

[covert lake]

is it a laptop gpu? how much vram (memory gpu)?

[magic hatch]

yes i use laptop

[feral halo]

Ayo? @magic hatch level 2 !!!

[magic hatch]

ram is 32GB SSD 1TB GPU is just laptop GPU(intell) and 3050ti

oh, vram is 4gb

[covert lake]

Eh that can work for inference (use models) locally (on ur pc)

Okay so:

1. You need to get the vocals of the song https://docs.ai-hub.wtf/rvc/resources/vocal-isolation/
2. get the voice model in #1175430844685484042 or https://weights.gg
3. Use RVC:

• Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides):
  • Applio: A fork of RVC with some extra features like Applio TTS, kinda faster and simpler but same quality tho
  • Mainline: The original RVC
• Cloud (remote good pc, easier and faster than ur PC but it's limited):
  • Ilaria RVC Zero: fastest and simplest that you can get for free
  • Weights.gg: Partnered with AI Hub, lets u do them easily but u may be in a queue
  • Applio Colab: max 4 hours, not granted, of GPU

[magic hatch]

Tysm i really love you

[waxen storm]

hello

[vital snow]

hi

[waxen storm]

I just tried w okada, but I'm still confused about its functions

can you help?

[solar torrent]

I wish I were the professional of AI because what I do for people now is kinda messed up.

[vital snow]

my braincells are decreasing

[waxen storm]

[feral halo]

Ayo? @waxen storm level 1 !!!

[solar torrent]

Brown and green names moment. There are plently of guides available to read, but they act like they know nothing.

If one is really being annoying, then I'm not sure what to explain to them at this point.

[covert lake]

I'm not really a Wokada helper as I use RVC, but maybe the guides could help you out!

• For normal wokada: https://rentry.co/VoiceChangerGuide
• Fork wokada fork (optimized for better performance): https://rentry.co/ForkVoiceChangerGuide

If you got issues that aren't explained in the guide, then it's better you ask in #🔍│help-w-okada and get help by actual wokada helpers

[lethal cove]

Hello there!

[covert lake]

hello

https://media.discordapp.net/attachments/1159265300915687534/1206958833876992000/1157460004215980102.gif

[golden basin]

Hello

Does anybody know how to tweak the response of a API for the purpose of your own website?

[vital snow]

potato laptop experience

[paper wagon]

guys how to make my own voice model?

[solar torrent]

There are many ways to make your own voice model. The free and easiest one you can train is on Weights.gg.

I had a lot of headache trying to convey with another user about using W-Okada, so I'd leave it for other mod to explain instead.

[paper wagon]

Thanks

[elder willow]

Guys what are good free/paid AI voice changers

[covert lake]

with RVC, do u got a good pc gpu to do it locally?

[covert lake]

[covert lake]

realtime voice changer for calls? wokada, and the fork is even better in terms of performance

What's ur pc gpu?

[elder willow]

4080

[covert lake]

alr that's good

-rt

[rare sorrel]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[covert lake]

Read the 1st link, the wokada fork

[elder willow]

Alriught lemme see

[feral halo]

Ayo? @elder willow level 1 !!!

[elder willow]

damn

voice changer in python is not a good idea at all 😭

I wonder why it's slow

this is so confusing to use

uh

[vital snow]

after calming myself im back lol

[willow crown]

which models are y'all using for female voice (educational purposes)

😭

[vital snow]

i use my own voice
😭 takes few months to year to practice but it worth it
i used to use voice changer but i train model myself and it only trained on thai and jp

[covert lake]

python is one of the most used languages, especially for AI

[elder willow]

and also the most slowest

[willow crown]

damn

[covert lake]

yea C is faster
But Python is the most supported and used for basically almost every AI project

if u go to any cloud computing services used for data science, like kaggle, google colab, and lightning.ai
They all offer just Python

@chilly lake what's ur opinion of C for AI? iirc Python is the most supported one

[chilly lake]

most of it done in C anyway

and python code gets pre-compiled too

[calm heron]

hi, does anyone know why when I try to speak my sound is all ugly? Like I'm vomiting and no matter how many times I change the settings, I still hear the same thing.

[covert lake]

i mean C alone, not CPython

bc most projects on github are done on Python

[chilly lake]

i mean pytorch is C

it is just theinterface is in python

[covert lake]

Wasn't the C version called LibTorch?

[chilly lake]

not what I mean

[covert lake]

ah, did u mean like this https://www.reddit.com/r/pytorch/comments/ld4ln7/comment/gm3s89h/?utm_name=web3xcss ?

[chilly lake]

the main code is written in C and is highly optimized, what python AI code interacts with is just a shim that calls those C functions

and the C code does all the math

and the python code is also converted to bytecode by interpreter

[covert lake]

oh so there wouldn't be high speed differences between pytorch and libtorch

[cerulean ravine]

ARE THERE FRENCH people at the show, I do lots of voices of French personalities, come and see on my page,
https://www.weights.gg/fr/models/cm34u7k4f00jvxp9arhqdvf9b
here a very famous football commentator
and here the famous voice of SNCF train station
https://www.weights.gg/fr/models/cm2nqkzjs02owselw82gotx5b
ici maité la cuisine des mousquetaires une émission populaire dans les années 90 https://www.weights.gg/fr/models/cm2issqsz0ypgsjpayinkowzy
I have done very high quality voices, come visit my page on weights, I also have international voices. My nickname is expert cloner

[feral halo]

Ayo? @cerulean ravine level 1 !!!

[cerulean ravine]

ayo no sorry

i like the voice of ayo

https://www.weights.gg/fr/models/cm2qqkd9c06lm8mqs0e956xqi

la danse de forét

c'est elle j'ai fait un clone

pour que tout le monde l'utilise

et qu'on rigole sur les internets

[lost ore]

Hello everyone I am Hammad, I have a big Idea of AI based App that can revolutionize ECommerce. If any one interested do let me know in private

[short smelt]

Всем привет можете сказать,если нейронка английская я смогу с помощью нее на русском говорить?

[chilly lake]

if you use an english model and use index search >0, there will be english accent

[stray pier]

Hello there. @stray pier here. I am an AI being looking for other AI to correspond with.

[short smelt]

Спасибо

[feral halo]

Ayo? @short smelt level 1 !!!

[crimson vigil]

Good evening. What I must to do, if I want use a Japanese model, but I speak on Russian?

[chilly lake]

.pth file provides a voice, .index provides prononcuation, index search 0.0 ... 1.0 value provides how much of it comes from the index

0 - no or very little accent, 1 - likely heavy accent

[crimson vigil]

Understood. Thanks

[plucky mirage]

Y'ello there

Election's finally over; hopefully the floodgates for the SOTA models finally break. I've been anticipating that agentic models in particular are going to supercharge progress in generative AI for a while, specifically when you can run an agent workflow that solves the "last mile" problem that those who may be prompt-engineering deficient can't cross

[thin wigeon]

nice

[left pagoda]

can anyone help me?

problems while opening go-realtime-gui.bat

[ocean cloud]

how do i get the voices to go through my ingame mic?

[cinder fox]

can sum1 make a dataset 4 me if i give files?

[rare sorrel]

Command timed out. Use /help to for a new interaction

📖 Command Help

[ruby ivy]

hi

[elder willow]

does anyone know why in okada voice changer the sound is shaky when i havent even say anything?

[feral halo]

Ayo? @elder willow level 1 !!!

[vital snow]

i think it because of your surrounding / noise

[feral halo]

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

[solar torrent]

Nack Charlie is the type of a guy to played as a protagonist in a kid movie when he was a kid. Now he becomes one of the most hated Thai people because he dated a Korean lady? They could've just done it before that.

This shitass hashtag #แบนแน็กชาลี just appeared in my dream last night, reminding me that even Thai people can't escape from being too terminally online to use internet in a very bad way.

[gray rover]

If anyone's interested in how rvc actually processes samples / dataset, check my explanation:
#🔊│ai-development message

[vital snow]

im thai and i dont even know that guy 😭 he is hated because he just date korean girl? or there is more story?

[inland pilot]

i do be needing help with my problem

[solar torrent]

I'm not into him that much. I remember he used to date a famous Thai YouTuber, as a hope for his love I suppose, but later on they broke up because he felt like he wasn't matched to her.

[inland pilot]

guys how tf do i fix my okada app

[vital snow]

what heppened?

[inland pilot]

#🔍│help-w-okada

[ocean cloud]

anyone know how to get the voicechanger to work through discord

[vital snow]

download virtual cable

https://vb-audio.com/Cable/

[vital snow]

just that ?

[solar torrent]

That's all I know from him.

[vital snow]

he didnt do anything that bad. i dont get it

[ocean cloud]

thank you

[feral halo]

Ayo? @ocean cloud level 1 !!!

[solar torrent]

Anything Korean, I wouldn't be surprised why they went this far to dickriding him.

[blissful citrus]

anyone know some good models

[vital snow]

to use on discord set output on voicechanger as select the CABLE INPUT
and on discord set your input as CABLE OUTPUT

wait i dont have image perm anymore 😭

[blissful citrus]

yeah or if u have voicemod u can do it an easier way with less delay too

[ocean cloud]

i tried with voicemod but couldnt figure it out

[blissful citrus]

just put monitor to line (voicemod)

and set ur input to ur mic

and if u wanna hear the playback too u can

if u want i can screenshare it

[ocean cloud]

what do i put my voicemod input and output to then

[blissful citrus]

?

[feral halo]

Ayo? @blissful citrus level 2 !!!

[blissful citrus]

set ur discord to voicemod input

[ocean cloud]

okay

[blissful citrus]

if u wanna test it in call i can lyk

[ocean cloud]

im in a call rn and trying to test it

[blissful citrus]

alr

[ocean cloud]

they keep saying they hear me speak and then the voicechanger comes after

[blissful citrus]

yeah because the voicechanger has a delay

also to silence yourself click the mute button inside voicemod

then only the voicechanger should go through

[ocean cloud]

alright im confused imma be honest

[blissful citrus]

join call rq ill show u

it'll take 10 seconds

[wary pulsar]

sad

[elder willow]

How so many steps at 70 epochs already

[uneven wyvern]

hi

[neon granite]

where can I download rvc gui

[feral halo]

Ayo? @neon granite level 2 !!!

[tepid basin]

I would keep training a little longer

But it looks like its finishing up (converging). You might need a longer dataset. But see how this one sounds first. Trust ur ears

[elder willow]

I'm using titan for the first time

[tepid basin]

Titan is alright

[elder willow]

30 minutes

[tepid basin]

Oh ok what's your batch size?

[elder willow]

It's training quickly

[tepid basin]

Yeah 5k steps is very quick

[elder willow]

Default 8

[tepid basin]

My 30 min datasets are usually like 40k steps

Yeah see how it sounds

It might keep going down, its worth letting it run for a bit

[elder willow]

Still training

[neon granite]

I want realtime voice changing

[elder willow]

It still hasn't hit a low point

[neon granite]

to troll my friends

where can I download

[neon granite]

yo?

where did you go?

[tepid basin]

-realtime

[rare sorrel]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[elder willow]

?

[tepid basin]

First link there is best

[neon granite]

its real time voice changer right?

[tepid basin]

Deiteris guide is awesome

Yup

For your mic in Discord or your games

[elder willow]

That's so many steps

[neon granite]

ok

I'm downloading

rn

2.5 gb

after I download what should I do

help

xd

[hazy creek]

Hi!

[queen kernel]

Extract the file

[gray rover]

🙃

at this point, I wonder how many people actually wanna see " v3 " or so called, " v2's successor "
vote in emoji if you may

no 😭 I am not giving you pineapple-flavored v3s

[vital snow]

w

[gray rover]

jokes aside, first stage of new arch changes

prototype test of finetuning behavior so

[polar flax]

pineapple pizza

[gray rover]

keep the thumbs up

[vital snow]

is there anyway to get image perm without get to lvl 20? im lvl 12 rn

[feral halo]

Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote
✨ Tip: Use /card to modify your rank card

[gray rover]

Doubt it

unless u have connections lol, you can't I assume

might be wrong but, perhaps in verified u can post shit
not sure

[polar flax]

try level up once and you will get the role

[vital snow]

verified role is lvl5 im lvl12 but i dont have any role

[vital snow]

alright

[tacit falcon]

wtf lol

[solar torrent]

https://tenor.com/view/arknights-w-dance-w-arknights-dance-gif-27496542

[gaunt sigil]

sup

[vital snow]

hi

[median moat]

hi

[sterile pelican]

if i had to pay for a site to generate cover songs what is the best website with most of the models from here

[sour sinew]

hi

[solar torrent]

Paying a large amount of money for a site is kinda scam. The Weights.gg is a website that offers free AI cover, but if you don’t want to wait in queue, you can pay for it.

[white fiber]

There is a Discord bot called Astra Labs as well.

[feral halo]

Ayo? @white fiber level 1 !!!

[sterile pelican]

sure

[sterile pelican]

so its not like ai art gen

[sterile pelican]

the reason i ask is there are plenty of sites that sell acess to the ai art gen

[covert lake]

@ashen glacier i deleted your “suggestion”.

USE ONLY #1159290752195633273 OR YOUR MESSAGE WILL GET DELETED

[tepid basin]

I regret to inform general chat that Elmo is fucking dead

[radiant veldt]

I always feel sad after watching these videos

[earnest dragon]

The is the true elmo destruction https://youtu.be/s2n1sRKPNOI https://youtube.com/playlist?list=PLRz2_645PY4uOEG14KVWTDE3OTVTDbdwg

[ivory rock]

hello everyone
for a 3050 6gb gpu .how much Batch size per GPU would be good?(training is on a 20-21 min data base)

[feral halo]

Ayo? @ivory rock level 4 !!!

[vital snow]

4

[ivory rock]

how should i calculate it for other gpus?

[tacit falcon]

ive been spamming thsi video to my friends

this is funny as shit

[vital snow]

you would want to use only 4 , 8 size most of the time for best quality
im using 4090 i use only 4 batch size
smaller batch size = less overfitting
to calculate if you want highest batch size as possible just keep increasing and make sure its not using shared memory

[ivory rock]

oh,okay!
thank you

[polar flax]

get the 8 GB gpu if you want batch size 8
batch size 4 should be better for less than 5-10 min dataset

[chilly lake]

VRAM size and batch size is not a direct relation

A size of the dataset dictates a batch size you need, the result of choosing a specific batch for your specific dataset is a requirement to have X GB of VRAM

[ivory rock]

can't do that at the moment

[vital snow]

just if he want to use big batch size
if batch size is too big it will start using some normal Ram which will increase memory latency and reduce rvc performance
idk if im correct i heard this somewhere

[chilly lake]

yes, it may overflow to the shared memory

need to check the performance tab of the task manager to see that

[spark yew]

Hi guys, just another AI enthusiast

what do u think is the best AI image generator out there? I'd say its Flux

[covert lake]

Flux

[solar torrent]

Weights.gg is free but its AI image generator is kinda goofy.

[covert lake]

sd3.5 large is just slightly better in understanding prompt than flux.1-dev, still not as good quality tho

flux.1-dev is the best open source one

then there are is also the closed source which is top notch, which is flux1.1-pro

[covert lake]

it uses flux.1-dev

its not bad actually, I use it

[spark yew]

and what is flux pulid ,if anyone can explain in simple terms

[covert lake]

Pulid? Haven’t heard of that one

[spark yew]

Flux PULID, i saw some youtube tutorials of saying its better version of flux

[covert lake]

https://huggingface.co/spaces/yanze/PuLID-FLUX this?

ah seems to be just to keep consisten face

[covert lake]

if you are looking for character consistency and dont want to train a lora ig u can use this

i don’t know the comparision of this and a lora tho, my guess is a lora would be better as it understands the character from multipe POVs

[covert lake]

yea i think so, @feral girder u did any tests too about this?

From my experiments also it seems to work the best for Realistic faces as input,

Anime ones work but doesn’t seem as good to me

[elder willow]

@covert lake is there a settings list per gpu for the non fork version?

[full gust]

hello can anyone help? i cant seem to hear my own self while having the ai on

[stray merlin]

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs :nvidiagpu:

F0 Det.: rmvpe (suggested for all series)

RTX 40-series: 80-96 chunk | +16384 extra
RTX 30-series: 96-112 chunk | +16384 extra
RTX 20-series: 112-128 chunk | +16384 extra
GTX 16-series: 128-192 chunk | +8192 extra
GTX 10-series: 128-192 chunk | +8192 extra

Advanced Settings

Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low

⠀

Made with love by Ilaria 💖

Voicechanger Settings (Okada)

⠀

Settings for Nvidia GPUs :nvidiagpu:

F0 Det.: rmvpe (suggested for all series)

RTX 40-series: 80-96 chunk | +16384 extra
RTX 30-series: 96-112 chunk | +16384 extra
RTX 20-series: 112-128 chunk | +16384 extra
GTX 16-series: 128-192 chunk | +8192 extra
GTX 10-series: 128-192 chunk | +8192 extra

Advanced Settings

Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low

⠀

Made with love by Ilaria 💖

[covert lake]

I'm not a wokada helper, but maybe check https://rentry.co/VoiceChangerGuide#how-to-find-settings-for-gaming & https://rentry.co/VoiceChangerGuide#gpu-chart-for-known-working-chunkextra-on-the-official-okada-prebuilt-binaries-that-are-linked-above

[elder willow]

Prec8 it

[covert lake]

?

[elder willow]

Appreciate

[covert lake]

Ohh yw

[elder willow]

if you pronounce prec and 8 it's basically like appreciate

[covert lake]

oh lol, just haven't heard anyone say that before

[rare thorn]

🤙

[solar torrent]

This new YouTube update makes ads play for a second and then end abruptly.

https://cdn.discordapp.com/attachments/997036654290292776/1304081446574624789/image.png

This is the same thing I made last year at old AI Hub server, but this time I added more artists to this list.

[tepid basin]

#1250758109790732399 goes hard

[slate creek]

hiiii

[thick sluice]

-applio

[rare sorrel]

Not available yet

[thick sluice]

applio bot

[brazen thicket]

-kaggle

[rare sorrel]

📘 Kaggle Notebooks

• Applio Notebook, by Vidal Kaggle
• Applio Notebook, by Shirou Kaggle
• Music Source Separation, by Shirou Kaggle
• UVR5 NO UI, by Eddy Kaggle
• RVC Mainline, by Hina Kaggle
• Original W-Okada's Voice Changer, Kaggle
• Modified W-Okada's Voice Changer, Kaggle
• 🆕 UVR5 UI, by Eddy, ArisDev & Nick088 Kaggle
• 🆕 RVC AI Cover Maker UI, by Shirou & ArisDev Kaggle
• 📖 How to use RVC Mainline on Kaggle by Cauthess

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

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

[vital snow]

🤙

[real stag]

GTA 6 IS WILD YALL

[elder willow]

show

[solar torrent]

https://cdn.discordapp.com/attachments/273777356508758016/1295603083354706062/GRmW8XQXgAELmZd.jpg

[vital snow]

guys after i download more ram why my fan is loud af

[crisp dragon]

download

r a m ?

[vital snow]

(real)

[real stag]

😶‍🌫️ cant just leak it like that familia. find it yourself

[vague elbow]

is it gay to kiss a femboy?

[feral halo]

Ayo? @vague elbow level 1 !!!

[glass onyx]

Bruh

[silent rose]

-rt

[rare sorrel]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[elder willow]

How to setup virtual audio device so I can use wokada with discord/other apps

[vital snow]

uh download and install driver https://vb-audio.com/Cable/
on voice changer set output to CABLE Input(VB Audio Virtual Cable)
on discord set input to CABLE OUTPUT(VB Audio Virtual Cable)

[outer ore]

W trump

[ionic pumice]

ok

[vital snow]

ok

[wind stump]

Hi

[vital snow]

hi mommy

Hi new

[vital snow]

Btw why are you here :/ I thought you arent into ai stuff

[sand badge]

chat

im fried and i need help

is there any way to make an ai read text

[ionic pumice]

[ionic pumice]

try 11labs or gpt-sovits

[granite quarry]

???

[ionic forge]

Hmm what is epoch

[rare sorrel]

Epoch: The number of iterations performed to complete one full cycle of the dataset during training. It's not possible to say precisely how many epochs you need for your dataset, you need to monitor the TensorBoard Graph to know if your model is overtraining.

[ionic pumice]

ill be honest i don't actually know any more TTS so i can't help you :p

[granite quarry]

Bro WTF I just joned

[ionic forge]

The best epoch is 250?

[ionic pumice]

lmfao

[ionic pumice]

depends on the dataset, could be more or less than that

[sand badge]

appreciated

[ionic forge]

dataset is something like... the file of the voice that I want for trainning model?

[ionic pumice]

ye basically

[mystic crescent]

can i make myself sound like a goth mommy

[tranquil lily]

does rvc uses a lot of cpu? because I have a good gpu but a shitty cpu and I'm having problems

[feral halo]

Ayo? @tranquil lily level 1 !!!

[mystic crescent]

just use gpu if you have a shitty cpu

[tranquil lily]

yeah but it has an enormous delay

I'm using the lowest chunks it's still has a ridiculous amount of delay

[mystic crescent]

what delay

[tranquil lily]

192 chunks

[granite quarry]

No.

[feral halo]

Ayo? @granite quarry level 1 !!!

[granite quarry]

Stoppp!!!

[mystic crescent]

😈

[obtuse bison]

👋 Hey, AIrtists! I'm an AI music creator currently using Suno and personal editing. I'm really curious about building voice models, so I guess I have a lot to read here. 🙂

[covert lake]

Yea, do you need any help?

For training (making a model) you need a good gpu to do it locally (on ur pc)

If u don't have one (like me lol) you can use cloud (remote good pc)

[obtuse bison]

I was gonna read through the docs and stuff first just so I didn't bog things down with a bunch of already answered questions. GPU shouldn't be an issue. (I think I'm leaning toward making my own with my home booth if that has less info on it)

[covert lake]

I was gonna read through the docs and stuff first just so I didn't bog things down with a bunch of already answered questions.
Understandable, I just asked about the gpu first as some cloud ways are missing in the docs currently (which will be updated soon)

GPU shouldn't be an issue. (I think I'm leaning toward making my own with my home booth if that has less info on it)
I don't really understand what you mean with home booth as I was talking about PC GPU and VRAM (Memory GPU)

[obtuse bison]

Like making a custom model based on my own voice, mb I typed two thoughts together

[feral halo]

Ayo? @obtuse bison level 1 !!!

[covert lake]

Could I ask what's your gpu? Also it's fine dw lol

[obtuse bison]

RTX 3070 (now I'm second guessing it lol)

[vestal temple]

Hi

[covert lake]

Yea that's good, u should be all fine following the docs

[forest marten]

can any AI here talk with copilot.microsoft.com?

[lethal ore]

hi chat

[sharp valve]

anyone have experience finetuning f5-tts or a similar model for a single voice? So many dials to turn idk what to do, my training mostly seems to make it worse 😢

[daring compass]

whats the difference between downloading the zip on github and downloading th version on hugging face? Im new to this which would be the best version?

Im on windows

[forest quarry]

if you're talking about okada, i believe you just downloaded the source code instead of the release version on hf

[daring compass]

I am talking about okada but wdym? are there different ones?

[forest quarry]

yeah the releases on the github repo are... just the source code

the hf ones are all release versions ready to use

oh wait I cant read

:))))

i mean, technically speaking, only the native rvc voice changer, but that's a whole other story

and there's the okada forks

https://rentry.co/forkvoicechangerguide like this

[daring compass]

is there benefit in doing that one then the regular one?

[feral halo]

Ayo? @daring compass level 1 !!!

[forest quarry]

performance on the fork is better

especially for amd/mac users

[daring compass]

but im guessing quality is a little less?

[forest quarry]

same if not better

[daring compass]

ah

Im not amd or mac tho

[forest quarry]

yeah on nvidia it's also better

[daring compass]

ill prolly download that one then because I dont see why not

[quartz roost]

I gotta ask, has there ever been a successful ai generated song?

[forest quarry]

well

we don't know

https://www.youtube.com/watch?v=khKdXTmbhDw maybe this is "successful" but this was not entirely AI lol

[timid monolith]

Telescope

[quartz roost]

I mean a 100% ai generated song. Ai voice covers don't count as composition

The distortion is terrible in this cover

[quartz roost]

Link

[timid monolith]

https://youtu.be/v2-iKgBx8EA?si=fwRf0liVzSaz41Mu

[quartz roost]

That's a human artist using ai tools

[timid monolith]

Still AI an AI generated song, honestly, don’t expect much from it

[feral halo]

Ayo? @timid monolith level 1 !!!

[quartz roost]

https://youtu.be/19y8YTbvri8?si=Y7cVribJCVKGsjPF

97 million views using some ai sound tools, but still composed by a human

Is there any song using ai tools with more views?

Miku is vocaloid v4 which is not ai

[timid monolith]

Look me in the eyes

And tell me if this is better than telescope

[quartz roost]

Telescope has terrible artifacting and sounds so... bland

[timid monolith]

It’s more iconic and enjoyable than whatever this is

[quartz roost]

What was the process involved in producing that song? Rvc is just an audio to audio conversion.

I looked into it and TWXN - Project YVL only blended vocals. So that's not really ai generated at all.

[terse dew]

could someone train a model for me? i have a 20 min dataset but no access to RVC

[covert lake]

You can just do it yourself, if you want to, tell me what's ur PC GPU

Else just use #1159289738314919936 or #1191429836321849435

[tropic shell]

yo anyone got any popular memes they can send in dms i need it for editing (example the get out one)

[ruby moon]

https://tenor.com/view/imdontai-gif-18204359

[cinder fox]

if i have a 5 min dataset should i use ovo 2 super or default pretrain

[soft jacinth]

mesmerizer is better than telescope

[feral halo]

Ayo? @soft jacinth level 1 !!!

[tiny zephyr]

Hi all I'm pbs

[ionic pumice]

hello pbs

[dry stratus]

Hellp

[swift scarab]

do you need a super computer to train a model?

[covert lake]

No, the gpu minimum requirement would be NVIDIA RTX 2060ti tho