🧬│ai-chat

[storm meadow]

You can download it locally

[loud oak]

just tried a few but the quality is not good

[coarse vessel]

yea most of them are not even as good as streamers'. you got to remove crowds (try MDX crowd in UVR though not sure 100% works) and many other noises. at least not in MP3 compression.

[half horizon]

what best agressive settings on deecho dereverb and for denoise if the voice come from a podcast in youtube

as ur experience

[atomic ether]

what this meaan

[patent quest]

AI is smarter than me

In a comedic twist on the trolley problem with celebrities and now crocodiles involved, the scenario could take on an even more absurd and unpredictable turn:

Introduction of the Crocodile Track: As the protagonist contemplates the decision between 20 Jameela Jamils and one Ted Danson, suddenly a third track appears. This track is filled with hungry crocodiles eagerly awaiting their next meal.

Comic Deliberation: The protagonist, now faced with three bizarre options, starts weighing the pros and cons in a comedic fashion:

They might ponder aloud whether the crocodiles would prefer the sophistication of Ted Danson or the charm of Jameela Jamil.
Consider if the crocodiles could potentially be distracted by Jameela Jamil's quick wit or Ted Danson's charismatic smile.
Wonder if throwing a celebrity-endorsed snack onto the crocodile track could buy enough time to figure out a more creative solution.
Unexpected Resolutions: The comedy solution could involve:

The trolley suddenly transforming into a giant disco ball, distracting everyone long enough for an unconventional rescue.
Discovering that the entire scenario was part of a celebrity-themed reality show, where contestants are judged on their ability to navigate outrageous moral quandaries.
The protagonist deciding to take a completely nonsensical action that somehow resolves the dilemma, like convincing all parties involved to join forces in an impromptu dance-off to "save humanity."
In a comedic twist, the addition of crocodiles to the trolley problem with celebrities amplifies the absurdity and opens the door for unexpected and hilarious resolutions that defy traditional logic and ethical considerations. The focus would be on entertaining the audience with imaginative and outlandish scenarios rather than exploring serious ethical dilemmas.

[candid echo]

does 4 batches size really improve results?

[loud oak]

nope

[tepid basin]

It could. Usually small batch size is recommended for very clean data, because smaller batch = smaller samples RVC is looking at = possibility of learning randomness in noise = overtraining

I usually stick to 8 batch for most my models, 4 if I know the dataset is super clean, and 12 if it's a bad dataset

Larger batch size also increases learning rate, meaning you will train steps faster

[wispy thunder]

uh

someone want join but he get banned by no reason / weird thing

[elder willow]

Yo guys

Erect

[lean blaze]

ratio

[elder willow]

No

[lean blaze]

ratio

@grave spire

[grave spire]

so, I just did an AI cover, just in case, Is there a tag to know if a voice is only [Conversation] or [Singing]?

[dusty imp]

Yo guys there are any models trained for spanish accent? Or i have to speak english

[grave spire]

Could I share a creation? like "x sings y"

[young mica]

@hazy gorge can u help me ?

mine isn't working

[hazy gorge]

@young micaI am a Korean channel helper. I am not good at English. I am sorry

[young mica]

alr thank you

[prisma furnace]

@jolly zephyr Done Crediting you. Please Check My Video, Thank You: https://www.youtube.com/watch?v=jcWZISP4xRY

[subtle fox]

https://cdn.discordapp.com/attachments/1217733581749882951/1240919326610034788/fooddupstep.gif

[jolly zephyr]

wow! It really sounds amazing, you've done a good job!^^ Thank you for using my models💓

[silk sequoia]

oops sorry

[prisma furnace]

Your Welcome.

[torn halo]

I hope everyone has a great day/night

[feral halo]

Ayo? @torn halo level 2 !!!

[torn halo]

💙

[elder willow]

okay dude your just asking to be touched

[feral halo]

Ayo? @elder willow level 1 !!!

[elder willow]

same with you

[thorn scroll]

yo

[elder willow]

what you tryna be touched aswell?

[thorn scroll]

so I want to troll my friends as an e girl is there a beliveable ai model

[thorn scroll]

please no

[elder willow]

too bad

i got a believable egirl voice

deadass

[thorn scroll]

uh-huh

[elder willow]

being deadserious

[thorn scroll]

can you help out ?

[elder willow]

sure 🤷‍♂️

[shut schooner]

hello

[thorn scroll]

sup

[shut schooner]

im new

[elder willow]

i finna touch you lil bro

[shut schooner]

how to reply chat?

[elder willow]

say smth like say less or bet

[heavy rampart]

hi, how would i import my pre-made voice file to the app and generate a modified one? cause i only know its base on mic input

[shut schooner]

idk

[heavy rampart]

using rvc

[shut schooner]

i cant english ffffff

[feral halo]

Ayo? @shut schooner level 1 !!!

[thorn scroll]

have you tried importing it?

[feral halo]

Ayo? @thorn scroll level 1 !!!

[elder willow]

@paper blaze you think you could help out parz rq

[shut schooner]

im new dude

[paper blaze]

what app

[heavy rampart]

theres no import option, afaik i cant see

[heavy rampart]

wait i cant send a pic ill send the yt link instead

https://www.youtube.com/watch?v=tWd1SMRXRMw

[thorn scroll]

oh

[shut schooner]

what is rvc?

[elder willow]

erich how old are you

[heavy rampart]

i use this program, already downloaded the model here, and want to apply the modification for the pre-made wav file

[shut schooner]

guess

[elder willow]

prob 9

[shut schooner]

16]

[elder willow]

either 9 or 90 the way you askin so many questions

[thorn scroll]

yo the model u were talking about is it in the voice model channel here?

[elder willow]

i mean my actual voice

[feral halo]

Ayo? @elder willow level 2 !!!

[shut schooner]

lol

[elder willow]

💀

[paper blaze]

if you want to change a file then use other app, search for rvc webui

[thorn scroll]

bruh

[paper blaze]

the youtuber you've watched probably has a tutorial on it or any alternative to it

the app you'd like to use only has realtime voice change capabilities

[thorn scroll]

and I let u touch me and everything #better_thug_it_out

[paper blaze]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[heavy rampart]

oh i see so this one i have rn, isnt capable of doing what i want? right?

[feral halo]

Ayo? @heavy rampart level 1 !!!

[paper blaze]

nuh uh

[heavy rampart]

alright thanks ill look it up

[elder willow]

ill thug it out for you next

[heavy cipher]

Hlo ladies and gentlemen

[elder willow]

uhoh its a mod

[thorn scroll]

#thug_life

[elder willow]

#thugging it out for life

im tryna thug it out for sum girls

[thorn scroll]

ong💯

but I'm tryna become one tho

[elder willow]

bro what 💀

[thorn scroll]

yo mods hook me up with a good e girl ai voice model

[elder willow]

hook up with me frfr

[thorn scroll]

gladly

[elder willow]

chat i have a funny thing i just did

a swedish guy i never once talked to sent me pics of his hot classmate and said help me fck her and sum other stuff and i said alr and he said hes in class with her and he wants to smash then and there and i said bring her to the bathroom and i said make sure u get consent

he def aint bro 💀 💀

[thorn scroll]

bro what

[elder willow]

what did you need?

[thorn scroll]

bro they are probably underage why u helping him out

[elder willow]

cool

[elder willow]

im underage too 💀

[thorn scroll]

yo are you the swedish guy💀

[feral halo]

Ayo? @thorn scroll level 2 !!!

Ayo? @elder willow level 1 !!!

[elder willow]

you tryna be my swedish guy?

[thorn scroll]

seems like lunch tbh

[elder willow]

w take

[thorn scroll]

🇸🇪

[elder willow]

having humor isnt being able to make a joke its being able to take a joke

im creaming on my toes

[stone delta]

hello everybody

[bold saffron]

@bright lily check your dms!

[stone delta]

what`s up

idk. good and bad together. but anyways gooodddd

do you like anime

[lean blaze]

lil bro

[stone delta]

today i watched 2 episodes of jujutsu kaisen

[lean blaze]

ok gangsta

[fallen ruin]

Hot and scary

[olive sinew]

Did anyone know which person's voice would be the best for a sci fi movies, I am making a Movie trailer I want to use some Rvc voice for my movie I think of using Spongebob for the voice. Does anyone know some good voice models else spongebob that will sound good

[stone delta]

wow and funny too 🥲

[stone delta]

wish you all the best for your movie 👍

[urban crane]

what

[wispy widget]

ai hub

[stone delta]

hello. i`m new here. may i know your name ?

[lean blaze]

freaky hub

[stone delta]

nope..... it`s boring!!!!

[ionic pumice]

💀

bru

[stone delta]

yes..... i`m gansta!!!😎

[feral halo]

Ayo? @low sphinx level 1 !!!

[stone delta]

yepp......

[feral halo]

Ayo? @stone delta level 3 !!!

[lavish depot]

hi

[ionic pumice]

2

[bitter moss]

e

[fallen ruin]

😆

[shut schooner]

test

[wispy widget]

erm what the

[tender prairie]

Can anyone Suggest me a attractive female voice model please

[elder willow]

Does anyone have a realistic sounding voices if yes pls tell me dm

[wispy widget]

define attractive female voice

[tender prairie]

I don't get it ?

[wispy widget]

what female voice attract you? not everyone have the* same taste. like, california valley girl set to maximum cringe is attractive for me

[crystal comet]

у

[coarse vessel]

ur mom might be attractive to you

[tender prairie]

Looks like you're projecting some family issues

[coarse vessel]

it doesn't mean I have fucked up relationship with my family. also, there are lot of anime, games, youtubers, etc. you can find out the voices in.

[tender prairie]

What i need is kinda attractive and serious voice UwU voice isn't working with my content

[coarse vessel]

yeah, your definition is still vague. think of like art and fashion styles.

[tiny glade]

--guide

-guide

[rare sorrel]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

[sullen sonnet]

hello

[minor blade]

If you want a model like that, make a post in the #1159289738314919936 channel asking for one.

[minor blade]

Also, please don't be rude towards others bud.

[tender prairie]

sorry but its not me who Started

[minor blade]

Nanashi, please don't start arguing with Toxic.

[minor blade]

Nanashi was?

[tender prairie]

yes

[minor blade]

Anyways, i'll ask you both to stop arguing. okie?

[tender prairie]

iwas just asking for help

[minor blade]

Well, i've already told you what you can do.

[minor blade]

make a free/or paid request on this channel i mentioned. (Keep in mind that using "paid request" tag obviously means you must pay to any model master that agrees making your model)

[tender prairie]

sure

[feral halo]

Ayo? @tender prairie level 2 !!!

[winter viper]

general in chaos over egirl models...

[minor blade]

egirl models are overrated

[winter viper]

yep

not going to see me argue about that xD

[minor blade]

me neither.

Altho there are some awesome girl voices out there on internet.

[winter viper]

dangerous to say that in here

that's like putting out a fishing line in general for people to dm you about it

[minor blade]

XDDD

I mean, there are some nice girl voices out there on internet.

But probably these don't have models done.

[winter viper]

xD

[light bone]

nooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

[minor blade]

yes

[light bone]

why

[quartz turtle]

Has anyone gotten luck with monetizing their AI content? (Covers, etc)

[timid musk]

no

[quartz turtle]

Unfortunate

[gusty sand]

why RVC GUI archived in github

[forest quarry]

Because it's outdated, sadly

[gusty sand]

whats the current tool

[forest quarry]

There's always Applio though

[forest quarry]

There's 2 main ones rn

Applio or the original RVC (known as Mainline)

Mangio can be used too but it's kinda outdated atp

https://docs.aihub.wtf/ for download links/guides btw

[serene lily]

names arsend

[minor blade]

Tell me why egirl voices are not overrated.

[minor blade]

Actually my buddy ProodbyNG managed to at least monetize a bit his AI original songs

We just made a bit of money but our spotify got taken down for some reason

[quartz turtle]

I got hti with Reused Content and was wondering if it was worth fighting at all, maybe trying like some sob story about impressionist & shit

[light bone]

u can be sexy egirl uwu egirl

[minor blade]

something else?

[light bone]

no i told u too many reasons

[gusty sand]

is there a vocal seperator that runs on google workspace like the RVC itself?

[minor blade]

De-Echo-DeReverb model under "Ultimate Vocal Remover HQ" tab in mvsep

[minor blade]

i repeat, something else? bc u didn't XD

[light bone]

lol

[dapper drum]

Hey leo dose useing a vpn for Paperspace work? like if I run out gpus? will it work if have my vpn go into like the uk will the free gpus show?

[minor blade]

u can be sexy egirl doesn't count as a reason haha

[minor blade]

I'm not sure bud

[elder willow]

Edaters

[quartz roost]

Yes

I have legal use

[elder willow]

Legality is a form of slavery

[quartz roost]

I have personal permission from the voice actor

[elder willow]

Of who

[quartz roost]

She's making bank

[reef locust]

teto

[elder willow]

I see

[feral halo]

Ayo? @reef locust level 1 !!!

[elder willow]

With those thighs I’ll be making a sperm bank

[quartz roost]

Commercial voices have existed for 40+ years

[elder willow]

Good for you

Now, I want your consent or I’ll do it without

[quartz roost]

Beg for it

[elder willow]

Oh I love it

[quartz roost]

https://youtu.be/fNZE46C5V78?si=zn4mnqq-lFsDrHyf

[elder willow]

You’re the one who made it?

[quartz roost]

My friend did

[elder willow]

Tell your friend I love him

[quartz roost]

She's the best selling singing ai for a reason

[elder willow]

Off course

I too will love to hear an anime girl sing

[elder willow]

I retract my words

[reef locust]

miku miku beam

[storm meadow]

-colab

[rare sorrel]

Suggestions for @gusty sand

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[storm meadow]

UVR5

[quartz roost]

https://youtu.be/nTYoBcAldL0?si=0u-4DgR1lANW3ri-

That's why I like this one more

[gusty sand]

which one do i use do seperate audio?

[feral halo]

Ayo? @gusty sand level 1 !!!

[elder willow]

I am afraid my ship sails in another sea

[fallow orchid]

Is there any free Audio Super Resolution things anymore?

And how do I get verified to access the verified chat?

[quartz roost]

The real singer is fantastic live

[elder willow]

I see, perhaps we can discuss this matter further in the personalised room and chat

[feral halo]

Ayo? @elder willow level 2 !!!

[quartz roost]

I'd rather show it off to everyone

[elder willow]

I see, it is truly unfortunate but I won’t be here at all times. If you wish to get a drink with me (you pay) I’ll be happy to

[light bone]

[minor blade]

👍

[gusty sand]

guys

[minor blade]

lapajita

[storm meadow]

UVR5

[gusty sand]

thanks

[storm meadow]

And your better off using the MVSEP website and creating a account

[elder willow]

I rather the term “making”

[indigo anchor]

that just seems really scummy... you can't just make money off of using a voice you don't have the rights to (unless it's of yourself or your friends and they gave you the rights to do it)

[quartz turtle]

Would you have the same attitude about impressionists trying to monetize their work?

[indigo anchor]

honestly, yes

why try to impersonate someone else?

[quartz turtle]

Because it's a skill, and there's a demand for it?

[indigo anchor]

a demand for it where?

[quartz roost]

Weird al

[quartz turtle]

^

[indigo anchor]

does weird al do impressions?

[quartz turtle]

In the populace, look at Weird Al, look at Azzers

[quartz roost]

Comedians

But those are parody fair use

[quartz turtle]

https://www.ranker.com/list/list-of-famous-impressionists/reference

[quartz roost]

Impressionist acting =/= copying someone's voice to impersonate someone for personal gain.

They will copy your voice and scam your grandmother for bail money

[ionic pumice]

oh man if i had just leveled up i woulda made the cringiest joke

[quartz turtle]

https://www.youtube.com/@Azerrz

[quartz turtle]

LOL

[indigo anchor]

wouldn't the latter be what monetizing ai covers would be?

[quartz roost]

Just need written permission it's the same rules as voice acting.

[indigo anchor]

you think people making ai covers have that permission?

[quartz roost]

There's a couple rvc voices in here that are indeed legal to use

[quartz turtle]

Arif didn't get written permission when he started

Do you just want written permission for any derivative works?

[quartz roost]

Is he using rvc

[indigo anchor]

derivatives are tranformative. using an rvc to clone someone's voice is not

[quartz turtle]

Irrelevant? The conversation has been about humans impressions immediately

[quartz roost]

Human impressionists are perfectly legal

[quartz turtle]

Would you have the same attitude about impressionists trying to monetize their work?
Jellirby — Today at 10:16 AM
honestly, yes
why try to impersonate someone else?

context for the conversation btw.

[quartz roost]

Apples and oranges

[quartz turtle]

So why hope in responding about oranges with comments about apples?

[quartz turtle]

We stopped talking about rvc's immediately, you said that impressionists monetizing their work is also scummy because why try to impersonate someone else

[indigo anchor]

got it

[quartz turtle]

I understand there's a symmetry breaker with human impressionists versus using a computer to generate the monetization, I'm not convinced of it being significantly different in the eyes of the law however just due to how hard it would be to enforce one or the other as technology continues to evolve

[wooden jolt]

https://tenor.com/view/frog-sitting-meme-funny-gif-15518043535710952457

[quartz roost]

looks at the most successful computerized voice in history

[indigo anchor]

dectalk? /j

[forest quarry]

i think its teto!!!!

[quartz turtle]

MIT <<<<

[quartz roost]

Dectalk was a commercial sold device made with a real voice explicitly for companies to monetize with

[quartz turtle]

I'm assuming the human gave permission for said company to use their voice?

[quartz roost]

This server would not be here if not for the research of Dr. Dennis klatt

[quartz turtle]

Okay?

And the server also wouldn't be here if not for Bob Kahn

But that's not really gonna be an answer to the question about if the human gave permission for their voice to be used in the synthesizer

[quartz roost]

There's a lot of history

[quartz roost]

Steven Hawking used Dr. Klatt's voice

[quartz turtle]

Also not an answer to the question though

Was it Klatt's voice? If so you implicitly answered

[quartz roost]

Klatt made and recorded his voice to be sold

Hatsune miku is the same thing

And it's made millions

[quartz turtle]

Okay, I don't know why you brought that up when we're talking about impressionists that are often using voices that they don't have explicit/written permission to utilize

I mean it's cool, but that's gonna be different versus someone like Azzerz who utilized voices he didn't have explicit permission to use, in order for personal gain (gaining him a platform, and ability to monetize said voices)

[quartz roost]

There is a legal method to doing this. It's really not complicated.

Did azzers use an ai voice made with the recordings of someone without their consent.

For example: the voice of jaxx from the hit show the amazing digital circus was furious with those who made an ai model of his voice and using it for their own gain.

This is clearly illegal.

[ocean wyvern]

.

[quartz turtle]

Already answered with a no, he's a human impressionist

[feral halo]

Ayo? @quartz turtle level 8 !!!

[shy lark]

how can i make art im new

[upbeat rune]

is there a chat for running llm locally?

[quartz turtle]

-collab

[quartz roost]

✏️

[quartz turtle]

? i thought that was a cmd

[quartz turtle]

Actually I reread question, look into LM Studio

[ionic pumice]

guys should i watch stranger thifngs

[quartz turtle]

https://www.youtube.com/watch?v=ax130yILbw0&list=PLVgLT-e3jXPDgeED0pD0BPq8kY1VAZAGa

[quartz turtle]

I liked it, so yes you should 😈

[quartz roost]

So many cool digital art tools and you just want to automatically generate some images? That's boring.

[quartz turtle]

Buying a used tablet for like 20% of the price from ebay is always the route

[indigo anchor]

instant gratification my detested

[forest quarry]

one by wacom frfr

(i dont know how to draw)

[indigo anchor]

luckily it's never been easier to learn how

[quartz roost]

Automatic generated Ai art and music are quickly falling off in popularity as predicted.

[upbeat rune]

more specifically im looking to either upgrade my current system or invest in a new system. im looking to find somewhere to look to get an idea of what I can run with certian hardware.

[quartz turtle]

(I watched a couple of speedpaints when they were huge about 8 years ago, yea I think I can draw)

[quartz roost]

Speed paints are becoming popular again because it proves the artist made it themselves.

And they're fun to watch

[upbeat rune]

id also like more information of fine-tuning models.

[forest quarry]

Mm, interesting

I'm not really sure about LLM stuff, but hearing from some people, you need a good GPU like a 3080 and like 16/32GB RAM

[quartz turtle]

AI Music seems to be having an uptrend

[quartz roost]

Ai voice covers have flopped. Vocaloid originals are skyrocketing.

[quartz turtle]

Speed paints were the I'm gonna watch while I eat OG videos for me

[quartz turtle]

Seems like both are having an uptick, it's not about which is more popular, it's about whether AI music is falling out of popularity or pop culture or not

(I don't know which is which btw, I only can read English, Spanish & French)

[quartz turtle]

Learn ML through Kaggle and you'll have a pretty good picture

[quartz roost]

That's the fastest growing vocalsynth right now

[ionic pumice]

this is the last day Shrek is going to be in Netflix :spunchbob:

[quartz turtle]

Gonna watch it today then

Unless it's moving to Peacock, then I'm still able to watch

[quartz roost]

These producers are at the cutting edge are using similar tools to build entirely new sounds and super creative songs. It's not the same as a massive flood of copied ai generated stuff like suno

"Mesmerizer" just broke records I haven't seen any ai music do

They did use an "ai tool" But the song is still composed by a human.

[quartz turtle]

Just sounds like two different goals then, but AI tools used by a composing human doesn't sound too innovative, but then again I haven't actually heard it so I'm not sure how informed my opinion on it is

https://www.youtube.com/watch?v=tZVw06KceXA&pp=ygUObHVwZSBmaWFzY28gYWk%3D this is my fav showcase of AI as a tool

[random chasm]

Yo wsp

[quartz roost]

Using ai to generate a bunch of random melodies and samples but still a human is the one to pick from that noise to actually compose it to sound good.

Same thing vocolo producers are doing.

@forest quarry

[forest quarry]

Gone, reduced to atoms

nuh uh he gone

[ionic pumice]

modded client?

[plush jay]

yo wsgg

[molten gulch]

Is weights slow for anyone else or just me?

[ionic pumice]

whar

[molten gulch]

I only use it to find and DL models

[forest quarry]

it did quite well for me idk

[ionic pumice]

bro what did they do to the gingerbread man in Shrek

[signal wave]

Hey fam, we are hand picking AI creators to showcase your songs on a new AI Music platform, think of it as Soundcloud for AI content (backed by Grammy award artists). Drop a DM and i’ll send you details. 🤘

[grizzled flume]

a

[forest quarry]

I'd rather you send that on #1159290752195633273, ty

[junior lark]

why i download the voice model and the voice.ai is not detect the zip file (i tried to unzip it also)

[dire ocean]

can somebody link me to the best current voice changer to use

nevermind looks like its Mangio-RVC

[umbral pier]

do you have any specific link for this?

[dire ocean]

https://discord.com/channels/1159260121998827560/1159620272316502096

[feral halo]

Ayo? @dire ocean level 1 !!!

[forest quarry]

voice.ai doesn't support RVC models

[forest quarry]

realtime orrr?

[junior lark]

then how im gonna download a voice model to use it

im kinda stupid

[forest quarry]

use this instead (linked below is a guide for W-Okada):

https://rentry.co/VoiceChangerGuide

Hm... I'm not sure, but it might be ElevenLabs

Mmmn

[covert lake]

kinda weird as 11labs is one of the best closed sourced ones

also GPT So vits is prolly used

[feral halo]

Ayo? @elder willow level 2 !!!

[covert lake]

Tacotron 2 is old asf

if u want an open source good tts there is gpt so vits

but i think they support only english and chinese voices

https://docs.aihub.wtf/tts/gpt-sovits/ as of now there is only the guide for the local one

but there are also google colabs if u want, js no guide about them yet

Yea, you can do that, but as rvc is made for speech to speech, it might not sound that good, ofc u can try first and play a bit with the pitch but from my experience its not that good for tts

[jagged laurel]

could someone please help me set up this up for discord?

[quartz turtle]

#🔍│help-w-okada

[chilly juniper]

Hi, I am joiner

[slender finch]

Hello everyone

Doing some research on the AI, is it possible to train my own voice for a cover?

[plush barn]

Yes

You create a dataset of your voice so record it like maybe the first 5-10 minutes of data of your voice

Then render it and trim out the empty parts of the recording. Then zip the folder were it has your dataset. The files themselves can be named anything along as it has no spaces

[candid echo]

im a genius

[feral halo]

Ayo? @candid echo level 5 !!!

[candid echo]

I got it to work goddamnit

[plush barn]

Then upload it to your Google drive. And go to a Google Colab link to RVC Disconnected. Follow the steps it tells you.

[candid echo]

EYO AND LVL 5 god bless

I canceled out the phase of the instrumentals in audacity using the instrumental file and the original because no vocals only file exists and was able to erase them like 90%

[plush barn]

Place the dataset zip file to the RVC Disconnected folder. Then once your Google Drive is connected to Google Drive, label anything that you need to add like your name along as it has no spaces.

After that, continue following the steps until you train the index. You have a option to include the cache of the pth files but that's if you have a huge dataset so have it on if you want to continue training it for more than 1 session same thing when saving the model which is the last step but aside from all of that then start training.

[candid echo]

its perfect

[plush barn]

Then after you trained the model with the amount of epochs you needed which I usually stick with 300 epochs if it has enough amount of dataset typically 10 minutes.

So that's how I make my RVC models through my own experience.

[warm sky]

☺️☺️

[gentle canopy]

hi

[fallen ruin]

pretty decent idea

I feel like I should do that. think about the amount of sweat that comes regardless

[feral tundra]

bbb

[tranquil juniper]

hello

[elder willow]

hi

[forest quarry]

hello

[ionic pumice]

hellaur

[tranquil juniper]

Can you recommend a $31 microphone?

[dusk cradle]

I've been wondering if it matters when I should start training the index model, like at the start middle or end. Does it change anything?

[feral halo]

Ayo? @dusk cradle level 1 !!!

[forest quarry]

I'd normally recommend doing it before training - just to make sure everything is alright before training

If index fails, you know something's up

[dusk cradle]

Thanks!

[forest quarry]

Yw :)

[forest quarry]

Sadly no, but research and take your own conclusions... always important to do so

[sacred vale]

Hi

Wqsup

[forest quarry]

Ello there

[tranquil juniper]

I researched, but one channel says it's good, the other says it's bad, so I'm undecided.

[feral halo]

Ayo? @tranquil juniper level 1 !!!

[forest quarry]

See what you resonate with more... check mic tests and stuff

[sacred vale]

How do I get the caseoh ai voice

[forest quarry]

#1175430844685484042 or on weights.gg

[rare shore]

guys

Can someone recommend me a good male voice model? I will use it for my own voice.

[forest quarry]

Go for something like the Stanley Parable narrator...

[rare shore]

thanks i will try

[wind stratus]

video reviews with audio samples should give you an idea on how the mic will sound, it has to sound good to you, i know people who record with their phone mic and the result isnt bad

[hazy zodiac]

does anyone have a website to seperate songs

[feral halo]

Ayo? @hazy zodiac level 3 !!!

[hazy zodiac]

x-minus dont work

[forest quarry]

https://mvsep.com

create an account tho

[hazy zodiac]

Ty

[forest quarry]

Yw :)

[crystal topaz]

does anyone know how to make the voices sound more realistic? for me when i'm using them they sound nothing like the original character

[forest quarry]

if you're using W-Okada, mess with the tune and index sliders, that should help with the pitch and accent part of stuff

[crystal topaz]

aight tysm

[forest quarry]

yw :)

[stiff heart]

I made a little tutorial on voice changing/face swapping video

https://streamable.com/05gk1u

[ionic pumice]

a

[fallen ruin]

dude went apeshit. My cat person friend's not sure which one of the chems he'd choose to be addicted

if this was classic fallout then maybe nuka cola

[ionic pumice]

𝓌𝓱𝓸𝓼 𝓯𝓻𝓮𝓪𝓴𝓎 𝓻𝓷?

[quartz turtle]

I thought it'd be good to have empty parts of the recording, so the model can sound like it's speaking sentences, and not abruptly ending?

[plush barn]

that makes a little more sense if it was TTS

but this is RVC which doesn't need longer empty parts and I remember some used a tool to splice the audio into pieces but I usually used audacity to trim out the silence parts

unless if you want to include the breathing

but a bit silence is fine

[daring frigate]

Hey everyone 🙂

[devout wagon]

you do want to remove low end noises so you need to noise gate and either truncate or do audio labeling

[plush barn]

of course

or a ai denoiser

which that's been discussed alot

[devout wagon]

The problem with audacity is that you can't see it. So some silent parts of the audio can have these noise bed unless you're manually silencing it. It's not necessary to do that because of truncate ofc

[spring sable]

Didn’t they make a discord bot to make ai covers? Or was I dreaming

[plush barn]

I think they had

[tired obsidian]

Can someone show me how to make AI cover

Can someone tell me

[soft gust]

Hello everyone, how can i create AI Covers? All the websites ask me to pay but I can't. How can i create some Covers?

[civic basalt]

Holaaaaauaua

[shrewd flint]

!

[trim holly]

Hey, anyone know how to get a microphone output as an option in the client?
I can get it to my speakers but not to an out.

[tulip oak]

does anyone have the squidward model thats trained with like yeat or smth like that

i found one before but lost it

i cant remember exactly what it was but it was smth to do with yeat and squidward

[wind stratus]

weights.gg if youre in a rush

[sonic tundra]

so chicky nerds

👑

[forest quarry]

Do ya have a virtual audio cable yet?

[trim holly]

like voice meeter?

[forest quarry]

sadly no, but... i think there's good squidward models out there

[forest quarry]

mhhh voice meeter kinda does that but I'd rather use an actual dedicated virtual audio cable

[trim holly]

Any suggestions for which to pick up?

[tulip oak]

and do you have a link for the colab im struggling to find it

[sonic tundra]

oh, pick me guy

[tulip oak]

the pink one

[forest quarry]

VAC (the lite version) by Muzychenko

[forest quarry]

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

[sonic tundra]

chicky!!!!

[forest quarry]

if u wanna know how to use it: https://docs.aihub.wtf/rvc/cloud/applio-colab/

[sonic tundra]

I'M GOING TO CRAZINESS WITH CHICKIES

CHICKY BOMB!!

💣💣💣💣💣💣💣💣💣💣💣
CHICKY OH MY GOOOOOD!!!
💣💣💣💣💣💣💣💣💣💣💣

💣💣💣💣💣💣💣💣💣💣💣💣💣
CHICKINESS IS GOING GLOBAL 💀
💣💣💣💣💣💣💣💣💣💣💣💣💣

[trim holly]

mj, any reason I'd get VAC over voicemeeter?

[forest quarry]

tbh ik this sounds stupid but

i kinda don't know how to use VoiceMeeter, it's kinda complicated (to me) and not really worth it unless you are going to add effects or something

vac lite ftw, if you wanna do anything else

[trim holly]

all i want to do is have an output for the program xD

[feral halo]

Ayo? @trim holly level 1 !!!

[forest quarry]

yeah then vac lite gets the job done

[trim holly]

that one has like, a voice every 30 min yeah?

[forest quarry]

the normal vac (trial one) does

vac lite doesn't

[trim holly]

aaaah, would you mind linking pls

[forest quarry]

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

[wraith vault]

hi

[spice wasp]

Hey guys, is applio with amd support available?

[forest quarry]

Not really, no RVC fork rn really supports AMD stuff

[soft gust]

It doesn't work

[trim holly]

rvc worked like a charm ty ty mj

now anyone have any tips for making sure the voice doesn't cut in/out

[forest quarry]

?

might be the s thresh (if you set it too high)

or, silencefront (in advanced settings) is set to on, that can happen too

also, wherever you're using it, make sure it ain't suppressing anything and that it sends through every part of the audio

like on discord, set that sensitivity slider all the way to the left

[bright flare]

does anyone knows a good free AI pic

[tulip merlin]

Has @keen loom forgiven me?

[feral halo]

Ayo? @tulip merlin level 1 !!!

[bright flare]

help

[forest quarry]

Maybe Stable Diffusion? Check the #1159513888199540817, there should be a SD tutorial

[elder willow]

Why does every women voice sound like a guy when i use it

[rancid atlas]

Hello chat

[elder willow]

Is it cause of the autotune? lmao

[fresh crystal]

good morning night or evening whatever it is for you

[forest quarry]

have you messed with the tune option? that should help with that

[forest quarry]

Ello

[elder willow]

I am on 0

And it sounds like a guy

And when i go higher with the pitch its offkey

[forest quarry]

+12

will be an octave higher

[magic dew]

How to make most realistic male to female RVC output?

[forest quarry]

change the pitch to +12 or something like that

[zenith hamlet]

"when i change keys it becomes offkey"

[tacit willow]

anyone knows why the output sound of my AI Voice changer sounds cracky when I‘m on discord

[tired obsidian]

How do you make ai songs

[feral halo]

Ayo? @tired obsidian level 1 !!!

[fallen ruin]

kasane tetor is the secondary tomboy Vocaloid⁉️❓❔🤯😱

[forest quarry]

could be the virtual audio cable you're using, i suggest using vac lite:

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

this is from Muzychenko's website, if u want u can go and visit the site urself

[fallen ruin]

https://tenor.com/view/vflower-vocaloid-v4flower-flower-v4-flower-gif-7280547662636171319

Flower best tomboy vocaloid

[worthy comet]

guys i downloaded the program for my amd gpu but its not using my gpu at all help plz

[feral halo]

Ayo? @worthy comet level 1 !!!

[tacit willow]

thx bro

[quartz roost]

https://tenor.com/view/vocaloid-dance-kasane-teto-teto-territory-cute-gif-17052544

[light hatch]

does anyone in here know some stuff about case fans for gaming pc's?

[forest quarry]

convert downloaded models to onnx, read https://rentry.co/VoiceChangerGuide

RVC

You can use that via Colab or locally

https://docs.aihub.wtf/ explains everything you need to know

[zenith hamlet]

Or weights.gg

[worthy comet]

thanks i did this but i still have a high cpu usage and low gpu usage i have a 5600x and a 6800xt

[forest quarry]

yeah that is sadly what happens

on AMD version

[worthy comet]

damn is there any other alternitives for amd on windows

[forest quarry]

not much that can be done abt it

[forest quarry]

i mean there's RVC realtime, seems to work better:

https://rentry.co/RVCRealtimeGuide

[worthy comet]

alright ill try that thanks

[forest quarry]

yw, hoping it works!!

[worthy comet]

it works better thanks man

[stiff heart]

Hey guys here's a little tutorial on how to use face swap and audio inferring together

https://streamable.com/05gk1u

With a final result for example

[whole parrot]

how come nobody has done voice model for ian

[forest quarry]

woo, glad i could help

[half horizon]

when u resume training are they start from 1 epoch ? again i pdid all corectly like on doc im confused they dont start from 155 epoch the start from 1 i put in stepcount 2333333

is all normal ?

[forest quarry]

maybe it screwed up on saving the d/g files...

if you stopped while it was saving them

[half horizon]

so when u resum it should start from last ckp

?

this how its looks like

[forest quarry]

mhm

unless you preprocessed and did feature extraction again...

[forest quarry]

actually, from what I'm seeing, it didn't even see the TensorBoard files nor the d/g files

did you export model once you finished training?

[half horizon]

yes

i pressed stop

training then i exported

to drive

[half horizon]

what should i do now ?i feel dumb itrough i did all correct

[forest quarry]

there's not a lot you can do ig

i think you'll have to keep training as is

[silk lintel]

yooo

[forest quarry]

ello

[half horizon]

so yu mean i start from scratch ?

[forest quarry]

yeah, i suppose

[halcyon matrix]

why cant i stream in the voice channels

[half horizon]

and then its not gona finish again so how sould i export it
to resume when i got new GPU on collab

when i save not finished model because session collab to resume it another day what sould i put on this case here

[forest quarry]

keep as it is

[half horizon]

thats was idid

[forest quarry]

you stop training now, export model, then when you come back set the training variables the same as b4, dont preprocess or feature extract again, then load the stuff needed and train

[half horizon]

wait if i changed the gmail and i transfer my file to new one for new session collab i put same names is that can interope resuming because i want resume in new gmail

[feral halo]

Ayo? @half horizon level 10 !!!

[forest quarry]

mhhh maybe

[half horizon]

what mean b4 ? . i didint run prepross and feature extracion

i need run Save preprocessed dataset files to Google Drive ?

[forest quarry]

"before"

[forest quarry]

thats when you begin training for the first time

[half horizon]

yeah i know imean for resuming i understand idont need run prepress and extract

but there is othere one called save under it

this what i was confused if i need run it or no

Save preprocessed dataset files to Google Drive

[forest quarry]

you did need to run this when you were training for the first time

it is not needed for resuming

[half horizon]

yeah got it

and for exporting i nned export before they told me u out of Gpu ?

because as i saw when it close to 40 min they told me its out and then i exported after see that

[halcyon matrix]

why cant i screen share in Unlimited

[half horizon]

may this is the problemthat i eport before ?

[forest quarry]

yh

[forest quarry]

normally its like 3-4 hours

but i normally go training for around an hour

[forest quarry]

probably because you don't have perms

[half horizon]

yetrday itrained it for 2.30 hour before exporting

[forest quarry]

mm

[half horizon]

i have 35 min dataset

[forest quarry]

mhm

[half horizon]

so to understand the idea when they sai u out of runing and i diint export yet that mean i lost my file 2333333

and i cant resume with it ?

[forest quarry]

yeah

you lose your d/g files, and cant resume

[half horizon]

so next time wil expor when i see i have 1 hour more

[forest quarry]

mhm

but like

be careful

when the logs say they're saving the d/g files, keep training for an extra epoch and then stop

just to be on the safe side of things

[half horizon]

u mean if my save rate is 5

and my im on epoch 175 i need wait until the epoch on on 176 or 178 to be sure 175 saved corectly ?

is that the idea ?

[forest quarry]

yeye

[half horizon]

i should run index training ?

i stoped and exported the model and im trying to resume now to see if it work now

[half horizon]

this maybe was the probleme because i remmber that was on save moment when i stoped it i wish that was the prbleme

[reef pewter]

u think a 2060 super is good?

[forest quarry]

mmmm might've been that then

[forest quarry]

yeh

[forest quarry]

it is decent for realtime stuff

[reef pewter]

how far can it go

[forest quarry]

it could probably do some amount of training

but in like, batch 4 probably

would be slow but decent ig

[half horizon]

ty bro

[forest quarry]

yw :)

[half horizon]

it worked resuming test from 55 i see that the graph alitle bit difrent looks like it have more crashes but its good btw its startts from 55

[forest quarry]

yh its working then

[half horizon]

yeah so the prblem was in saving moment as u said

goodl luck bro that was very helpful to understand the thing

[forest quarry]

🙏 tyty

[zenith rivet]

hi

[solemn vapor]

hee hee

[half horizon]

wich ai cover dont need nvidia or cpu in collab ??

[steel scroll]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[half horizon]

i got this 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

[forest quarry]

you could use any of em without GPU, but RVC will be slow

[half horizon]

yeah i tryied before is worked but now idont understand why there is this expetions

[steel scroll]

why did applio randomly stop working?

[forest quarry]

AICoverGen does that now, ig

[forest quarry]

hm?

[steel scroll]

its not loading gradio

and tensorboard is inactive

[half horizon]

u mean its not working for now ?

[forest quarry]

mmm, i believe so

[half horizon]

so what i should use wich ones are good that have youtube link

[forest quarry]

hm. im not sure what could be happening

[half horizon]

somthing on clouds

[forest quarry]

iirc the option for the "share link" should've been there if you ticked an option

my memory aint workin rn

i shall eep

[half horizon]

its ok bro wil keep trying ty so much

[steel scroll]

i tried the share tunnel and it still isnt working

[livid oyster]

hi im rue

[rain rose]

Hey

[tough jacinth]

is there any new real time guis?

[rain rose]

Ey guys

i was just having a break down when thinking about AI and how it will turn out

[zenith rivet]

i accidently left.

😭 🙏

[rain rose]

I had visions on how the Ai endgame will look like while smoking and i can tell you it is over for us

[feral halo]

Ayo? @rain rose level 1 !!!

[elder willow]

is rvc disconnected fixed?

[wind stratus]

i use it a few hours ago, it should be good

[elder willow]

the only reason im asking is because when i tried making a model it was saying i was missing a filelist.txt and it was making some weird folder named %s

[wind stratus]

yes it happened to me when i enabled the autosave option

[elder willow]

so just dont enable autosave?

[coarse vessel]

never tried autosave yet

[elder willow]

im making a juice wrld model

using sessions

[wind stratus]

yeah its still a experimental option, didnt worked for me

[elder willow]

yeah maybe it will get fixed

[wind stratus]

i manually stop the cell to save

[elder willow]

i do 200 epochs at a time

[wind stratus]

the training cell shouldnt be running for more than 3 hours, beyond that i get disconnected

[elder willow]

thats why i doo 200 epochs at a time

then retrain

another 200

[coarse vessel]

dont need to stop, just move some pth files in weights folder to your Drive (in the folder icon in left pane)

though depending on how you predict the training time, can usually do every 100 or 200 epochs

[elder willow]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[wind stratus]

i do it to backup the g_ and d_ files

[rocky light]

yo can who help me?

[half horizon]

WHAT IS DIFRENCE betwen higgingface and google collab cloud

[storm meadow]

Google colab can use a GPU and has limited time

While HF Spaces is unlimited but uses CPU

[pale umbra]

I want to do rvc, thank you in advance.

[coarse vessel]

wonder if you're still there, but DP 2.1 is still in early adoption with some cable support issues esp for UHBR20 (80 Gbps) like this
https://www.tomshardware.com/monitors/displayport-21-has-a-serious-issue-with-uhbr-certified-cables-perhaps-thats-why-nvidia-opted-to-stick-with-dp14-on-the-rtx-40-series
currently RTX 4090 supports DP 1.4a whereas RX 7900 XT & XTX supports DP 2.1 (54 Gbps, or prob 80 Gbps for some professional cards).
let's wait for more confirmations & announcement about 5090 if it supports DP 2.1 80 Gbps, otherwise pass. grab 4090 instead, esp once it drops back to MSRP/below.

read also more about Gigabyte FO32U2P and some important comments
https://www.tomshardware.com/monitors/gaming-monitors/gigabytes-new-trio-of-aorus-qd-oled-gaming-monitors-ship-in-april-1440p-or-4k-at-up-to-360-hz

[feral halo]

Ayo? @coarse vessel level 39 !!!

[coarse vessel]

huggingface spaces use CPU and even a bit slower than average laptop CPU

[half horizon]

so there is no time ? i can train model for hours non stop i dont care if slow my laptop slower than it

[rocky light]

can who help me

[half horizon]

so in this case is better choose for me

[storm meadow]

You can’t train with hf

[half horizon]

why not ?

at least it worked for me now for aicover

[rocky light]

can you help me @half horizon

[still wren]

Oh nice the server is back I remember it got deleted

[half horizon]

what is ur question ? im newbie but i learned somthing if i can ido

[coarse vessel]

still slower than a laptop with 11th gen intel CPU

[rocky light]

i dont know hot to setto voice changer i got setings but i dont know how to set

[half horizon]

oh so better training on vollab then other things on huggineface

is this the genral idea

i know its relative

[rocky light]

wow

@half horizon watch dm

[half horizon]

idint understand ur qution

send a screen

[rocky light]

watch dm

[torn flame]

Hi how to make it talk without my voice?

i mean

[rocky light]

watch dm @half horizon

[feral halo]

Ayo? @rocky light level 2 !!!

[half horizon]

is the result quality from hugging face aicover or from coollab are the same ?

[flint galleon]

can i run okada voice changer on my mac m1 and get good results (less lag) or will i need to run it on a GPU?

[coarse vessel]

M1 is supported, and try recommended setting in the pinned guide in #🔍│help-w-okada

[sinful orchid]

https://youtu.be/TuV1rBp6KmM?si=YS8RC3N0XKJ5Zj8z

[flint galleon]

will it run like it does on the free colab version?

[half horizon]

yeah saw when i put index rate at 1 the voice is more closer to my model is that true ?

iwant leanr how to know which parametre sould u test for ur cover

i cant gamble with alot of parametres at least i want know wich parametre that should not change and wich i can ttest difrent value and see result to find my target

[flint galleon]

basically i wanted to use it for discord using vb cable

[sinful orchid]

https://youtu.be/TuV1rBp6KmM?si=YS8RC3N0XKJ5Zj8z

[half horizon]

my target i clone the voice to looks like my model talk wich parametre u sugget to me change and atest as priority for example for now i think index is a priority to make it closer to 1 then what second place i cant test with it from ur experience

[torn crow]

I there any model maker who's willing to train for free?

[grave crane]

hey

[elder willow]

has anyone ever heard of the fatz gerinomo model?

[lean blaze]

no

[lean blaze]

.wk

[indigo pike]

novagang in AI HUB

👑  ⁦exan⁩ - 1194 plays
 2.  ⁦kinja⁩ - 609 plays
 3.  ⁦yam⁩ - 574 plays
 4.  ⁦rode⁩ - 393 plays
 5.  ⁦mox⁩ - 352 plays
 6.  ⁦gram⁩ - 311 plays
 7.  ⁦crows⁩ - 311 plays
 8.  ⁦Moa⁩ - 295 plays
** 9.**  ⁦gilbert⁩ - 289 plays
 10. ⁦lizardhorse⁩ - 272 plays
 11.  ⁦decaln⁩ - 272 plays
 12.  ⁦luca⁩ - 259 plays
 13.  ⁦delta⁩ - 256 plays
 14.  ⁦remy⁩ - 248 plays

Crown claimed by exan!

glitchcore - hyperpop - plugg Artist - 265 listeners - 12464 plays - 47 avg

[elder willow]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[sonic bobcat]

can someone train voice models for free?

[opaque wave]

hello

[golden vapor]

hello

[sonic bobcat]

dm me if you can

how

[storm meadow]

-docs

[rare sorrel]

Suggestion for @sonic bobcat

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

[storm meadow]

Check if model already exists #1175430844685484042

[sonic bobcat]

it dosent

[storm meadow]

Did you check weights.gg

[sonic bobcat]

yes

HOW DO I USE KAGGLE

[storm meadow]

What’s the voice?

[sonic bobcat]

Sheldon Cooper from young sheldon

[coarse vessel]

-colab

[rare sorrel]

Suggestions for @sonic bobcat

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[thorn scroll]

yo any good e girl ai model ? ,idc if it is paid

[storm meadow]

Bro did not check

[sonic bobcat]

they don't work

i already tried

yes

[storm meadow]

https://discord.com/channels/1159260121998827560/1181319399936893029

[storm meadow]

What doesn’t work about them

Is there a error

[sonic bobcat]

they don't load in weights.gg

[storm meadow]

Why don’t you use the models on weights gg

[sonic bobcat]

there is none

[thorn scroll]

thx

[storm meadow]

Request for the model in #1159289738314919936

Or make it yourself

[sonic bobcat]

why does it say 19.5GB

[feral halo]

Ayo? @sonic bobcat level 6 !!!

[storm meadow]

-docs

[rare sorrel]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

[sonic bobcat]

nvm

[storm meadow]

That’s the storage on the virtual kaggle machine

Iirc

[sonic bobcat]

how do i see my model

and can you just make it for me?

[storm meadow]

Just use these guides #🧬│ai-chat message

[sonic bobcat]

if you know how to do it and i don't

im too dumb

[storm meadow]

Or request in #1159289738314919936

[sonic bobcat]

here's the dataset

[storm meadow]

Delete it

You can’t post copyrighted datasets in this server

And I’m not training it for you

[sonic bobcat]

who will than

[storm meadow]

Once again, request the model in #1159289738314919936 or check our documentation for tutorials #🧬│ai-chat message

[coarse vessel]

-colab

[rare sorrel]

Suggestions for @sonic bobcat

☁️ Google Colabs

• Applio, by IA Hispano 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
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -local, -overtrain, -realtime, -rvc, -help

[coarse vessel]

if you dont have decent Nvidia GPU, use colab link above (esp RVC disconnected)

[sonic bobcat]

i don't have a nvidia gpu at all

i have amd laptop graphics

[storm meadow]

Then use a colab then

[sonic bobcat]

ok

[storm meadow]

Applio has a ui, or you can use RVC Disconnected

[smoky reef]

anyone know a good voice changing program for amd cpu's and gpu's? ive been using voice ai but they're always having problems

[storm meadow]

Use W-Okada

Completely free and open source and can use models from #1175430844685484042

https://rentry.co/VoiceChangerGuide guide looks hard but it’s really not

[smoky reef]

can u send me the link for that?

ah thx

[storm meadow]

Dunno

I’m not helper anymore anyways

[sonic bobcat]

its stuck

[storm meadow]

Wait

[sonic bobcat]

nvm

[smoky reef]

should I do the alt ver?

[maiden gale]

how do i use this

[lean blaze]

.wk

[indigo pike]

Cutthroat Mafia in AI HUB

👑  ⁦ina⁩ - 63 plays
 2.  ⁦soto⁩ - 34 plays
 3.  ⁦ashy⁩ - 20 plays
 4.  ⁦Korozelly⁩ - 16 plays
 5.  ⁦proandcon⁩ - 16 plays
 6.  ⁦chillidog⁩ - 10 plays
 7.  ⁦bulxcy⁩ - 7 plays
 8.  ⁦jxyytf⁩ - 5 plays
 9.  ⁦bae⁩ - 4 plays
 10. ⁦dis app fed⁩ - 4 plays
 11.  ⁦chopp⁩ - 3 plays
 12.  ⁦00.q⁩ - 2 plays
 13.  ⁦Korskyy⁩ - 2 plays
 14.  ⁦b00p⁩ - 2 plays
** 34.  ⁦gilbert⁩ - 0 plays**

Crown claimed by ina!

Artist - 33 listeners - 209 plays - 6 avg

[lean blaze]

.wk

[indigo pike]

Duster in AI HUB

👑  ⁦hat man⁩ - 8344 plays
 2.  ⁦cassidy⁩ - 7754 plays
 3.  ⁦5b⁩ - 1968 plays
 4.  ⁦squirt⁩ - 1958 plays
 5.  ⁦carry⁩ - 1870 plays
 6.  ⁦f⁩ - 1713 plays
 7.  ⁦notdylanlmao⁩ - 1671 plays
 8.  ⁦Übel⁩ - 1661 plays
 9.  ⁦angelito⁩ - 1545 plays
 10. ⁦Dj Clamor⁩ - 1343 plays
 11.  ⁦Wacc⁩ - 1286 plays
 12.  ⁦tee⁩ - 1191 plays
 13.  ⁦swimp⁩ - 1158 plays
 14.  ⁦yokitsu⁩ - 1083 plays
** 1122.  ⁦gilbert⁩ - 1 play**

Crown claimed by hat man!

corecore - slowcore Artist - 1122 listeners - 90202 plays - 80 avg

[lean blaze]

.wk

[indigo pike]

Tenkay in AI HUB

👑  ⁦yeresin⁩ - 806 plays
 2.  ⁦Frankie⁩ - 768 plays
** 3.**  ⁦gilbert⁩ - 482 plays
 4.  ⁦k0ne⁩ - 435 plays
 5.  ⁦angelito⁩ - 337 plays
 6.  ⁦sqwab⁩ - 267 plays
 7.  ⁦pharma⁩ - 261 plays
 8.  ⁦mox⁩ - 218 plays
 9.  ⁦shawn⁩ - 212 plays
 10. ⁦uh1c⁩ - 195 plays
 11.  ⁦zishan⁩ - 178 plays
 12.  ⁦lame⁩ - 169 plays
 13.  ⁦eugy⁩ - 166 plays
 14.  ⁦libann⁩ - 117 plays

Crown stolen by yeresin with 806 plays!
Previous owner: Frankie with 768 plays.

sigilkore Artist - 128 listeners - 6711 plays - 52 avg

[indigo pike]

Crown for Tenkay

Current crown holder

<t:1716010229:D> to <t:1716010229:D> — yeresin — 806 to 806 plays

Crown history

<t:1709995015:D> to <t:1716010229:D> — Frankie — 768 to 768 plays
<t:1706350010:D> to <t:1709995015:D> — yeresin — 596 to 646 plays

[lean blaze]

.wk xaviersobased

[indigo pike]

xaviersobased in AI HUB

👑  ⁦Adiiose⁩ - 4860 plays
 2.  ⁦remy⁩ - 3999 plays
 3.  ⁦exan⁩ - 3143 plays
 4.  ⁦luminal⁩ - 2823 plays
 5.  ⁦yeresin⁩ - 2302 plays
** 6.**  ⁦gilbert⁩ - 1361 plays
 7.  ⁦sqwab⁩ - 1045 plays
 8.  ⁦uh1c⁩ - 1033 plays
 9.  ⁦mox⁩ - 862 plays
 10. ⁦osc⁩ - 821 plays
 11.  ⁦gtemq⁩ - 775 plays
 12.  ⁦buga⁩ - 696 plays
 13.  ⁦Teckmek⁩ - 668 plays
 14.  ⁦cxvz⁩ - 617 plays

cloud rap - dark plugg - sigilkore Artist - 293 listeners - 41151 plays - 140 avg

[sacred oar]

yoyo

[sonic bobcat]

https://weights.gg/download?fileUrl=https://models.weights.gg/clwbot5e303ihveit17yjiyar.zip I DID IT

nvm it sounds terrible

i need someone to do it for me

[rustic horizon]

a

[queen grotto]

hi

[ionic pumice]

whaer

[queen grotto]

i want to ask, why my okada with rx 6600, the voice sounds flicking

[jaunty trench]

Hi yall this is incredibly stupid and probably very niche but does anyone happen to know what AI voice the people who do those tiktoks where its like “little john lives in a 0.5cm^2 apartment and will expand it with galvanized square steel” use? Im trying to make a parody of it and I really want the same text to speech bot for consistency. If anyone happens to know I would love it if you could help me out! Thanks you! 😅

[teal stirrup]

Hi came here from voicechangerguy's channel

[twilit venture]

Recently i bought helldivers 2 as a key and bought the latin american version, i wanted to see if anyone wanted to gift or trade (game key) for the normal game

[chrome zinc]

where i can find voice models

[snow sedge]

Depends on which settings you use. You also need to do convert models to onnx

[queen grotto]

what setting should i use, btw im using rx6600 and 5 5600g, already convert to onnx still laggy

[feral halo]

Ayo? @queen grotto level 1 !!!

[snow sedge]

https://rentry.co/VoiceChangerGuide#gpu-chart-for-known-working-chunkextra

But i'd suggest using my fork that offers better performance and doesn't require onnx conversion for amd
https://github.com/deiteris/voice-changer/releases/tag/b1823

[light hatch]

Ah alright, thanks

[last river]

hello

[tulip merlin]

Can someone tell @keen loom I’m sorry, and please unblock me?

[arctic night]

what app yall using ?

[keen mason]

wshhh les frr bien ?

[violet acorn]

hello

[arctic night]

hmdlah est toi ?

[light hatch]

Ok so that screen doesn't support G-sync though, is that alright? And even if I get dp 1.4 from the 4090, should I use that or hdmi?

If thats the case for dp, do I need to use that dp 1.4 (out) slot?

[keen mason]

al hamdoulilah akhy

ca raconte quoi ?

[arctic night]

oh rian

[coarse vessel]

some FreeSync monitors prob have chance to be G-sync compatible, might have to find out somewhere...

[light hatch]

but I think the other one (the dell alienware one I had) is better, as it had gsync and also it was curved

curved is a little nicer for the big screens as otherwise u need to look further which can be annoying

and the 4090 only supports up to dp 1.4 and hdmi 2.1 anyways

so the screen u picked wont change much

[coarse vessel]

yeah curved gives immersion value for 3D games and videos, but it defeats purpose of viewing angle and distance. also I doubt if it will be good for most 2D games and normal uses compared to flat monitors.

[light hatch]

yeah that is true though

[feral halo]

Ayo? @light hatch level 15 !!!

[light hatch]

idk what to pick no more ;-;

[coarse vessel]

still it may not matter too much compared to other factors

[light hatch]

so should I go with flat or curved?

[coarse vessel]

you just aim for best value for almost all your needs

[light hatch]

and this ^

[pure dawn]

https://cloud.mail.ru/public/hbQr/5sZYgotsS

[light hatch]

I think I'll go with a flat one then, the one u recommended above

but the 4090 only supports dp 1.4 so do I use hdmi or the 1.4 (out) slot on the monitor

[indigo coral]

Salut

[coarse vessel]

DP 1.4 is still fine for 4K@120/144 hz with DSC

[tacit willow]

My voice is working almost fine but sometimes its kinda laggy for example Im talking normally and after a few seconds Im muted and then I can talk again? Is there any fix for that?

or what could be the issue

[forest quarry]

thats... kinda weird, not sure why it's happening

[light hatch]

dsc?

and should I use that DP 1.4 (out) port on the screen u sent then? or is the out port smth else?

[cinder kestrel]

Does any one know how to extract frames from a video using ffmpeg and make those video frames back into a video using ffmpeg after upscaling the frames into 4k quality??

[simple tusk]

[coarse vessel]

that jargon should have been mentioned in the links I gave before
https://www.tomshardware.com/monitors/displayport-21-has-a-serious-issue-with-uhbr-certified-cables-perhaps-thats-why-nvidia-opted-to-stick-with-dp14-on-the-rtx-40-series
https://www.tomshardware.com/monitors/gaming-monitors/gigabytes-new-trio-of-aorus-qd-oled-gaming-monitors-ship-in-april-1440p-or-4k-at-up-to-360-hz

[fallen ruin]

Now this is the pinnacle of Sesshomaru content and I love the attention to detail on the belt

[light hatch]

ah alright

thats probably why the 40 series doesnt support it yeah

so if I wanna buy the AORUS FO32U2P, should I use dp or hdmi?

and would that mean I need to use the 3rd one?

[abstract ridge]

I sharted

[forest quarry]

noob master

[abstract ridge]

amongus

[forest quarry]

fallguys

[elder willow]

OIL UP CHAT

[forest quarry]

dont kill my man michael

[forest quarry]

wha

[elder willow]

OIL UP LIL BRO

[forest quarry]

HELL NAW

[fallen ruin]

https://youtu.be/i3ZqA4tQJWw?feature=shared

[frail ocean]

https://tenor.com/view/reiner-braun-reiner-hello-chat-hello-char-gif-24668821

[fallen ruin]

https://tenor.com/view/reiner-braun-reiner-hello-chat-hello-char-gif-24668821

[elder willow]

i watched this video two hours ago before this guy sent it and i have to ask…

HOW THE FUCK DOES YOUR CARPET CATCH ON FIRE?

[feral halo]

Ayo? @elder willow level 1 !!!

[agile walrus]

henlo

[final sonnet]

hi

[spice wasp]

Can someone tell me what exactly differs from the original one?

XTyde2 is a new modified ui for xtts2 and ported to colab, github (and coming soon to local) and it clones voices in a short time (0 shot) with a small dataset.
It can clone voices for normal text to speech use fast
For now there are about 5/6 bugs including:
the malfunction of a language
Reload the audio every time you need to redo a conversion.
This is the github.
there is space and google colab and a demo, if you like space please like ❤️```

[odd crag]

hello

[cyan oyster]

Generally, the longer the dataset the better right? Is around 30 mins too long?

[viscid wadi]

hi

[ionic pumice]

relatable

ai noov

[fickle estuary]

i have both voicemod and voice.ai which is better? i have HyperX Quadcast S mic. and how do i put the zip files in voice.ai? it says that i "cant upload files of this type"

[forest quarry]

If it has more variations it can get better.

30 mins is okay, tbh

[lean blaze]

cornball

[ionic pumice]

https://www.youtube.com/watch?v=M2x9PgNKbN4 have you guys ever heard of xp's installer iso in viewable video format

[snow zephyr]

#f1

/imf

[potent vapor]

boop

[ionic pumice]

import raw data to audacity duh

[jade girder]

for God’s sake can someone tell me how to make free ai cover?? all the sites i used to use are down

[forest quarry]

You can use Applio: https://colab.research.google.com/github/iahispano/applio/blob/master/assets/Applio.ipynb

Guide: https://docs.aihub.wtf/rvc/cloud/applio-colab/

[honest mantle]

Love it

[dry blade]

amazing ¡¡

[dusk cipher]

I want to know RAG(Retrieval Augmented Generation) capabilities like I want a model that can be trained on information the user provides as input (like a text, PDF, photo etc.), then answer questions and generate outputs related to that input information For example - A model that can take Text or diagrams from the user as input (photo, PDF, etc.), find those same Texts/diagrams in an existing book or data source that the model has access to, and highlight/display those identified lines/diagrams from the book to the user.

[forest quarry]

Mmm, never heard of that

[elfin hemlock]

can someone help me with autotune

[forest quarry]

?

whats up

[elfin hemlock]

what programm/site im can use for it

bc bandlab sound bad

[forest quarry]

mmmmm...

mac only

lol

[elfin hemlock]

hear, its sound bad

im dont need that autotune

https://youtu.be/IIgWfz6FTSM?si=GO_lG5TSQobR-dvd im need like here

[dusk cipher]

never heard of what, RAG?

[forest quarry]

well that would require a decent DAW

and melodyne

[forest quarry]

yeah

[elfin hemlock]

is that free app or no?

[forest quarry]

digital audio workstations aren't cheap

and melodyne too

so yeah

[elfin hemlock]

ohh

[elfin hemlock]

maybe crack?

[dusk cipher]

umm okay , basically nothing just model is trained on the input the user provides

[forest quarry]

wouldnt recommend that

[elfin hemlock]

hear from 2:15

im need autotune like here

[elfin hemlock]

is that name app?

[odd osprey]

why cant i find the voice models channel

[snow sedge]

RAG is not training. RAG is about generating embedding for your data, storing it in the vector database, and then retrieveing it based on the user input

It's basically as you described. You can look into privateGPT or llamaindex that implement RAG pipeline

[dusk cipher]

i dont think i need Retrieval Augmented Generation i just need Retrieval for my project wdyt

[feral halo]

Ayo? @dusk cipher level 1 !!!

[light bone]

ye its app and vst for daw

[ionic pumice]

my wifi so bad i finna be a caveman

[short patio]

say drake i heard you like em young

[feral halo]

Ayo? @short patio level 1 !!!

[short patio]

casca enjoyed it

[thorny raven]

NAH

Griffith fans Are Crazy

[snow sedge]

What do you mean by "Retrieval"? You still need to perform all what I mentioned because you need to perform similarity search and find information relevant to user input

For that you need to generate and store the embeddings

And as a result, your prompt should be augmented with the information from the search. It's RAG anyway

[dusk cipher]

@snow sedge that's not my concern, my main concern is generating an output and processing pdfs

[reef pendant]

any1 have realistic voice for a 12 year old girl? (I mean voice model #1175430844685484042 )

[lean blaze]

bru

[wind marlin]

subscribe to my youtube @zeke9862