#🧬│ai-chat
1 messages · Page 314 of 1
hola
heyo
not even 24 hours into bali i've already seen a handful of foreigners
what do u use to run the voice models
if i want to use call rvc from a python script how would i do that
ah that's what it is
was so confused why they randomly started spewing crypto on a site that is only the landing page
honestly I was making fun of it more than anything
xd
L chat
any of u guys use alltalk tts?
Ayo? @glacial dagger level 1 !!! 
its pretty neat u can feed it a 10sec wav of someone talking and it generates tts and clones it in real time
no model needed
Nah StyleTTS-ZS is the new one I'm looking forward to.
Not sure if anyone's interested, but I modified the RVC v2 model architecture by increasing layer size. The model reached even lower pretrain loss values in under 1/4th of the time. The old .pth file was 53.6 MB, while the new architecture's .pth files are 493 MB, using 18.9 GB VRAM at a batch size of 8. I'm creating a pretrain based on VCTK Corpus + TITAN dataset.
Here's a simple recipe for delicious homemade brownies:
Ingredients:
1/2 cup (115g) unsalted butter
1 cup (200g) granulated sugar
2 large eggs
1 tsp vanilla extract
1/3 cup (40g) unsweetened cocoa powder
1/2 cup (65g) all-purpose flour
1/4 tsp salt
1/4 tsp baking powder
Optional: 1/2 cup (90g) chocolate chips or chopped nuts
Instructions:
Preheat the Oven: Preheat your oven to 350°F (175°C). Grease a 9x9-inch baking pan or line it with parchment paper.
Melt the Butter: In a microwave-safe bowl, melt the butter. Let it cool for a couple of minutes.
Mix Wet Ingredients: Stir in the sugar, eggs, and vanilla extract into the melted butter. Whisk until smooth.
Add Dry Ingredients: Sift in the cocoa powder, flour, salt, and baking powder. Gently fold everything together until fully combined, but don’t overmix.
Add Extras (Optional): If you'd like, fold in chocolate chips or chopped nuts for added texture and flavor.
Bake: Pour the batter into the prepared baking pan. Spread it evenly. Bake for 20-25 minutes, or until a toothpick inserted into the center comes out with a few moist crumbs (you don’t want it to be completely dry).
Cool & Serve: Let the brownies cool in the pan before cutting them into squares. Enjoy!
fish speech is also a good zero-shot voice cloning: https://github.com/fishaudio/fish-speech
how do i use tha app i cant figure it out
Have you heard https://styletts-zs.github.io/
Okay make this public when done someone will probably use it
I'm looking forward to the result quality compared to the normal fp16 and fp32 training
Sure! Will release once it's complete
New pretrain from scratch gonna take a while
that amount of vram is too high tho, even 4080 only has 16 GB
Ayo? @polar flax level 40 !!! 
Hello World
ok i wrote a thing so i have obama and kanye arguing with each other, and i have tts of it in their voices, i can do it for any combination of ppl
If youre going to want to put it in #1235952130855010365 you need model maker, so i suggest you get that.
What is happening on Applio's discord server. They are just removing alot of channels of the models. Have the owner been hacked?
ik
blaise rip
rip
I'm testing using hubert x-large for the embedder model
Ayo? @eager halo level 1 !!! 
if there's an index available for an rvc model in okada, do you guys always use it and crank it up to 1? or is there a downside to having it on max and that we should aim for a lower setting?
I think they hacked ia hispano☠️
and no one will force you, blaise has been under attack lately and that has been the final attack.
I don't know, the last thing he did was to publish the new version of the applio website, went to sleep and started the party.
yupi
I was sleeping but something woke me up.
is still sleeping
where is support help me plz
hi
What happened ??
Why did all of a sudden all of Applio's channels get deleted and someone mentions a token and hacked it?
Blaise got hacked

Fuck 
hi
@signal cairn who asked
fart
man\
ruh roh raggy
wait
tf
i miss iahispano
i did
i just donwloaded from appoloi
🤣
am i cooked
;(
Applio is over
appoloi HAJAHAH
ops
Applio is cooked
thank you for your bitcoin
erm
I guess we are cooked
embed failure
What the sigma
LMAO What happened to Applio
Хуй
I got liek 30 pings from the server
пизда
hacked
Oof
Everyone please avoid the link on their page it is a fake one.
NOONONONONON
Ахуеть
Nice big member drop LOL
What the fuck this woke me up
Ayo? @mortal cobalt level 1 !!! 
NOOOOOO
Everyone ping was not needed at all
When did this crypto scam happen💀💀
disintegrates
big-T of me
Everyone please invest my new crypto called JakeCoin lol
failure
wth is IA hispano

i downloaded it like 5 hours ago? am i safe
Hopefully the hacker is gone
Don’t buy crypto, buy gold instead
bro got hacked
dawg look at the gif
Good question I don’t have an answer
What even is $HISP
hengker tz
I highly doubt it's safe
guys who is he?
it would be time to leave iahispano
Bro just pinged 475,000 people
only
Thanks for pinging me at 3 in the morning 👍
thats what they want you to think, gold is at an all time high rn, so that means markets are more fearful, so play against the market, go greedy
yall are missing the joke 😭
What happend to Ia hispano
No clue
Is applio on pinokio also hacked?
Ayo? @elder willow level 2 !!! 
developer who made an RVC fork, I really liked it. i hope he gets his stuff back
Ayo? @elder willow level 1 !!! 
Is applio on pinokio also hacked?
erm
Ping png
what happend

ping pong
i'm gonna check out the hacked server and try to take them down
why tf they give me android icon

W android
wait is someting isntantly supposed to happen nothings happening to my pc
Ayo? @elder willow level 1 !!! 
I am so confused 😭
save this guy for now
🤣
chat what happened
penis hehehe
server got hacked and everything got leaked
who got hacked, or what got hacked
guys who is the hispano
damn what server
Ia hispiano
get the legit applio asap before it's gone: https://huggingface.co/IAHispano/Applio/tree/main/Compiled/Windows
Which server
DAYM, so i delete all my voice models then?
I'm disconnecting my local version of the applio repo from git just to be sure
Pinging almost half a million people is crazy, let alone at 3 in the morning
im kidding, I dont know what happend I just got here
…
Rip
Applio github might be okay since it has 2FA but to be safe avoid doing any updates
wonder what the motive is
who the heck is al hispano
thank you
i know i know, my git just auto pulls sometimes
i'm so confused on whats going on, if i have the ai rvc from this server am i ok?
I downloaded it like a year ago and never really check this server.
POEPLE DO I UNSISTALL MY MODELS
i got some models and the rvc cuda thing
wrong reply
the guy that has discord pfp (not the owner of this app tho)
Same question
Ayo? @gentle swan level 1 !!! 
Sadly it had to be done, we dont want people to download a trojan and people get scammed, at least people cant say it was an ad lmao
Yes
everything on the hacked server got deleted
mb
got ya, thanks
I'm gonna go click the scam links on purpose brb
Is the applio on pinokio also hacked ???
I feel like they also open sourced all things with malware in it hoping someone would git clone their bot
mb too
Only Applio might be affected, but nothing is confirmed since their github didnt get updated
which one they deleted it
voice models arent affected
Fair enough
on .gg/iahispano
if apilo man isn't a dumbass he should have different passwords across accounts
If anyone gets a possible malware sample from the applio server or cloned page pls send it to me I wanna see what they're doing
Yep now I can't sleep friendly reminder to mute servers you aren't active in
Oooh okay bc im confused idk whats happening😭
can I join these?
Yo is this astra dc server?
yes
Similar to rizz app but astra?
no
but there is nothing
what is this dc server abt
dont click any links there, they are fake websites
💀 people getting hacked is just another level. Just how careless can you be
almost half a million users pinged and I have no fucking clue what an applio or hispano is
Ayo? @spark ferry level 1 !!! 
RVC mostly
this is the GTA online AI RP server
whats that
Real🗣️
wait im confused someone explain i downloaded models from appolio website whats the big deal
everything there got deleted
?
the amount of people that get banned daily for compromised accounts is wild
frfr
@young flume no promoting in general
that's what it is. People gotta learn the security and caution
my dad left for the milk
Wild
tell that to hispano
well, i know cybersec
Did Balsie get hacked?
well i don't even have a product, was looking to just validate the idea between you guys
Google 2faing when
🤣 yeah
Ayo? @glossy flame level 2 !!! 
if anyones got malware samples from hispano's server send it to me
yubikey on top
There's goes my day.
I got an malware.
send it to me
From IA Hispano's hack?
in dms

I want to do some reverse engineering

hydrogenpeaceful.exe
me too lol
heck yeah! great minds think alike
wot
yeah fr
still, I'm glad to hear at least it ain't ransomware bs. if it was djvu people'd be fucked. Basic defender n malwarebytes with rootkit scan on is enough to save poor souls
one of my clients was just hosting malware on their vm 😭
that's an malware app
send it to me in dms pls
this is one of the reasons I had to ping everyone
M
*an is only used for words that start with a vowel, and the correct sentence structure would be “I got malware”, seeming how malware isn’t an object
cant have 300k people getting a ransomware
so its ransomware
Can you email it to me? I sent my email to your dms
why not that'd be funny

can I ping everyone
No
Ayo? @cinder hedge level 1 !!! 
This bot is wild
why
do u have an virtual machine?
(sorry my autocorrect is stupid)
can I embed now?
imma try to get the decryption keys from their c2
yes I sent it in DMs
Yes
damn, lame ass bot won't give me embed perms
either way. All who wanna be sure you're safe, get malwarebytes and run a scan with rootkit enabled ~
take care
is rootkit some gross apple thing
or install linux
yes I do, I do malware analysis as a job :)
🐀
I will email you but my email will be something tho
so I'm gonna do it.
Your email will be what?
server revive strat is wild 
I hope the apilla GitHub becomes an Elon musk crypto scam
That would be something
Ayo? @zinc jacinth level 3 !!! 
That would explain why my fridge was smoking...
Tell the fridge smoking is bad for the lungs
Ayo? @tardy rain level 2 !!! 
lol
Ayo? @marble whale level 3 !!! 
I'm bored back to hibernating
that's not a link bruv
Sophie Philippines_shadow doesn't have a sample from IA Hispano
It's a sample from a random roblox game
that's fake
Still worth a poke, I'm bored so I'll do it
oh
No gif perms


@elder willow
breh
balls
i still have a sfx with python malware posing as a fake fl-studio
Goodmorning @elder willow
wanna try that?
Alright sure
ok imma dm
Do not send any links here or malware related content, do that on DMs.
yeah we know dw
@elder willow @lusty whale the two heros we never asked for 
:3
But don't have a rule for CodPoints

dont try me
Ayo? @signal cairn level 70 !!! 

you're a cod player?
no

against discord tos, and every server must follow tos so adding a additional rule that says the same thing isn't needed
Personally, I do not indetify as a cod player, or a human for that matter . What I identify as cannot be completely described, I guess you could call it the "soul". My physical body is just a vessel for my "soul". My emotions, feelings and thoughts are not who I am, and do not define me but are simply an extension of my physical and mental body, I as the soul simply observe these things without any judgement or resistance. That's what I identify as, I simply observe whatever is going on in my physical and mental body. Therefore, I am not broke, short or ugly, that's just how my body, my physical vessel is. But I am not my body, my body is just a vessel for who I truly am.
bro wtf , codpoints
highlighting the C in cod and P in points you're not slick bro
👍
My bones hurt

hola
yes
holla
hi
you know I've never talked in here, but now that an everyone ping sent me here, I'll just say hi
Hi baby. welcome to AI HUB
do you normally call complete strangers, "baby"?
*discord moment*
´Thanks
do you have software to use and our voices ? (a voice modifier)
Dont.
Do NOT use this channel for promo, if you want to share your content use #1159290752195633273
i cant use this channel for promo?
@limpid citrus
What is that AI Hispano?! 👀
Hispanic community of RVC basically
Has a lot of members
Is it more famous than this...AI Hub?
Ayo? @mint sage level 1 !!! 
no
I see. 🤔
is the app itself hacked or the Discord server?
from the current news only the discord server, but to be safe only use old versions of applio until it is confirmed
I have the new version but I'm not using it yet, should I uninstall it?
i dont have crypto
Ayo? @granite rampart level 1 !!! 
I have the new version but I'm not using it yet, should I uninstall it?
What about the No UI Colab notebook of Applio? Is that safe to use?
colab shouldnt be affected
who ghost pinged me
me
also why is this not getting updated
you need to manually click the refresh button
wait for at least 100 epochs
i have autorefresh too tho
wait you get banned for being racist
ah alr im not familiar with applio
FUCK
did click refresh on the top right of the little UI box
yeah
does anyone have the most realistic girl voice model
Hey does anyone actually follow the guide on mainline rvc and set the epoch to the max and just get whatever is before overtraining
is it possible i got ot already cuz my lowest value has been the 52 epoch ?
check #1175430844685484042
it doesn't matter, but you can stop training whenever it starts overtraining or the colab GPU session is about to end
Sigma
hell with yall
Alright noice, i just noticed that every model has a specific epoch
Hell with ya too
also bc tensorboard doesnt work for me
goodnight, a young brotha went to the club to tonight
Ayo? @lament plank level 1 !!! 
W level up
Eyy
yea it shpuld be safe as it runs on a remote enviroment, unless it adds something to the files but the last commit is of 2 days ago
but who knows
I dunno, it's happened in other servers
Is UVR5 still the best for vocal removing? I was looking up the best karaoke model and couldn't find the one at the top in UVR.
I’m pretty sure it’s just the GitHub and a new phishing site “applios.org”
Ayo? @dull pebble level 2 !!! 
yeah I'm just surprised cloudflare listened to the reports that quickly'
The discord message reports likely won’t do much because they don’t act on deleted messages, this dude keeps reposting
I'm sure the real Hispano's emailed discord support
not like they'd do anything to be fair
@charred sage first of all, i am vegeta, second of all you're not vegeta
Looks like blaise's got his account back
Oh, I didn’t see the message. He announced the account was actually hacked. I thought it was a rouge moderator
mhm yeah i just rejoined to see the message too
more crypto scams. I swear to fuck
It gives an error while searching. I also search for other people but it still gives an error.
can someone help me make a AI voice model for song covers?
i cant find the right website to do it
is google collab still valid?
huh?
whats ur pc gpu? U can do it locally (on ur own pc) if u can, or if u want the easy way or dont have a good pc i can suggest u a cloud way
yes but what is the cloud way, im looking for a solution for months
@covert lakeI go to the find model channel. I use the /search bot but it gives an error. I write the model name that others have found (Katy Perry) but I still get an error. However, someone else finds that search. But it gives me an error.
the applio bot got removed recently because of a problem seen in #✨│announcements
What bot are u using?
whats ur pc gpu?
GTX 1650
@covert lake i got voice so where do i use it?
well, technically you can inference (use model) on locally if you want using Mainline even if it won't be really fast
If you dont want it to download the program or want a faster way u have to do it on cloud (on a remote good pc), which i suggest Ilaria RVC Zero
Last update: Apr 01, 2024
Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Vocal Separator (UVR) Troubleshooting “No gpu is available for you for 60s” Introduction (with website link) Ilaria RVC Zero, is an RVC (Re...
depends, realtime or ai covers and whats ur pc gpu?
I have laptop
Ayo? @mild veldt level 3 !!! 
@covert lakejust /search I write the first one that comes out
- idk :<
i want to train a model not use, i dont have a model yet
the first one that comes out is a bot for youtube songs..
If you are talking about the one in the second pic, ye that's broken, use the /find of the weights.gg bot
@signal cairn btw sorry for ping, but should we remove the modelai bot as its old and doesn't work?
You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU
And do u need it for realtime in calls or ai covers?
@covert lakenot work /find
oh nvm read it wrong lol
Well, nope u can't train locally, all the cloud ways are:
- 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)
I personally suggest mainline kaggle, also about applio, applio MAY not be safe for the recent owner account that got hacked, as its cloud it runs on a remote pc and the code shouldn't have anything bad but just be careful
how do i use the model then?
in that message i told u how to train, for inference way for ai covers instead i told u in #🧬│ai-chat message
ok thanks
your welcome
and do you have a minute to walk me through applio? Nick?
as of right now i would suggest to not use applio (see#✨│announcements ), but almost all the links i sent you are guides (i specified which aren't and which are), ofc u can ask me questions
alright but can you still sacrifise a minute on vc to help me a little?
i can text you there but i can't vc sorry
is there a specific question/issue? Or maybe you just want me to tell you by order what u should follow?
I'm gonna need a step by step tutorial
someone got for dubbing ai site for free?
Ayo? @valid goblet level 1 !!! 
it might be kinda hard as its much steps, but first of all i will help with making a model with the suggested ways:
- Make a good dataset which means gather good quality audio which u can also clean, in the link its also suggested which program to clean the vocals
- Use kaggle mainline which is a guide for training (making) the voice model
- for using the voice model (inference) its better to use Ilaria RVC Zero which is another pretty easy guide
Last update: Mar 8, 2024
This guide for Mainline Kaggle is an alternative option to the Mainline Colab notebook for training voice models
It is complete and should walk you through every step of the way since Kaggle has a difficult learning curve. However, it will be updated constantly to go over parts that need more cla...
Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Vocal Separator (UVR) Troubleshooting “No gpu is available for you for 60s” Introduction (with website link) Ilaria RVC Zero, is an RVC (Re...
is this normal? the lowest value has stayed like this for a whiel
Hi
hAVE Lol Ryze?
judging from what I've searched in the server, increasing index also introduces more delay? 0.3-0.5 from what I've heard is the usual range I'm assuming for using index in realtime?
Hi.
can we still use the applio google colab while this whole situation is going on
It does add up a bit of latency ( Because of faiss ) but can't say how much as hardware differs for everyone naturally but, I'd say you'd typically not notice if your hardware is decent enoigh.
Now whether you should use it or not is up to you. Index is faiss-indexed features of a given voice; Phonetics, patterns, accent and so on.
As to using it vs not using it, it's pretty straightforward. If you can manage to be consistent and somewhat " compatible " with model's own features, go for index at 1, if you can't / have hard times or just want it to be more closer to you ( the voice ), set it to 0.
hello guyz
can we still use the applio google colab while this situation is going on
hello frens
I have this idea to make a dj Aituber, is there a channel here for this?
.wk Född Död
Khatam hora yeh server ab
Ayo? @astral inlet level 1 !!! 
is this OT?
Ayo? @unkempt bluff level 7 !!! 

Who got hacked?
hi
No idea
I kn3ew it I KNEW Applio got hacked
I was like "wait a minute.... who the heck has .org??"
shrug
Ayo? @trail cargo level 2 !!! 
is this fine? or OT?
.
Is there a tutorial on how to install and configure the voices?
hi
Hola
I got a problem with voice changers if I’m in call and someone speaks my voice changer says what the person said how to fix?
Ayo? @worthy ledge level 2 !!! 
What’s AI Hispano and how do I know if I have it
its all good nothing was compromised other than the discord account
Why does the Weights bot say that I am not authorised to use it?
Ayo? @tepid gull level 1 !!! 
where are you using the command?
Im jsut texting the bot directly
not sure if it was supposed to work in dms
Where am i supposed to use the bot?
Its not dms i believe, its the direct chat message option on the right side of the screen where the users are. (following Sikeman tut)
What were they trying to do even?
Make people sign in?
from the looks of it, a criptoscam
and probably credential stealer, idk what you can steal from that tho
Hey ai lover
Ayo? @autumn pelican level 1 !!! 
Why ayo
How can I make AI-Art for a thumbnail for example?
Using this discord or a tool within?
screaming into the void
hiya
Can someone help me please
I got a problem with voice changers if I’m in call and someone speaks my voice changer says what the person said how to fix?
hi
Hello guys where can i ask my questions about using c ai unofficial api usage
Ayo? @quartz notch level 1 !!! 
hi
How much did you invest to make 6k? 🤡🤡🤡
What is ai hispano?
I currently have a voice model with the total voice record duration in dataset being 28 minutes and 36 seconds, how many epochs I need to set in order to have the best quality?
What is the bot for ? I genuinely forgot why the fuck im here
You're asking me?
Anybody but because you are here, why not
Well, I want to let her sing some "cough, cough" communist songs, so that's it
Total duration of my recording is 28 minutes and 36 seconds
Wow impressive who is the singer ?
Ayo? @fallow crest level 1 !!! 
Sakura Yamauchi from the anime I Want to Eat Your Pancreas
Previously I asked on this server if anyone can make a voice model of her, but I received no replies so I decided to take matters into my own hands
Ayo? @mellow monolith level 2 !!! 
Thats nice from you, where can the model be found and used ?
I haven't actually made the model
I have only managed to compile her dialogues
Before I decided to make a voice model for her, I wanted to have the best settings
It's not showing myy gpu
Ayo? @tired widget level 1 !!! 
Thats very kind thanks for your job
So with the total training duration of 28 minutes and 36 seconds
How many epochs would be optimal?
Do I need to set the epochs to 250, or 500, or 2000, or something like that?
And I intend to train her voice model on RVC v2 Disconnected, how long will it take to complete?
Hello
I dowenloaded voice but idk where to use it can pls someon say HOW TO USE THAT VOICE CHANGERS!?

yo I hav a question
If theres a beat that is free for non profit am I able to perform with that beat if there is no money involved?
hello everyone
hello
Hihi
guys what do i do if i have to download a file in training and it gives me this message: Failed to fetch
TypeError: Failed to fetch
^^
Ayo? @gilded edge level 1 !!! 
?
Ayo? @runic bay level 1 !!! 
means u can only post it on nonprofit platform eg soundcloud
not on spotify
+380996095191
Hello someone help with voice changer if im in a call in discord and if the person says something my voice changer says the exact same thing what the person said
How to fix
Yeah because you have the virtual audio cable routing it into the voice-changer
But I need the audio cable to talk no?
And I get lags
Yes but you need it set as output on voice changer and input on discord and then input on voice-changer be mic
I did that
Show a screenshot here #🔍│help-w-okada
What screenshot?
Of w-okada and discord voice and video settings
How do I fix this issue when I try dragging the tune to how I want it won't let me when I look at cmd prompt it keeps showing sum like voice changer rvc updated settings tran-7
I don’t have w okada
What are you using then just send the screenshot of what you are using instead of w-okada here #✨│ai-help
I use voice changer client demo
Send screenshot
Look
hihi
hi
goat pfp
Okay it's w-okada and now send discord voice and video settings and also is your headphones volume to loud that your mic picks it up
I send
Discord input should be virtual audio cable not mic
Yea I know about that one it was always audio cable
But I just did normal one for now
Okay then it's your headphones being to loud that your mic picks up the audio
Oh okay
What about the voice changer saying the same thing of the person what he said when im in a call
Can you also tell me the best settings for my system
It is picking up the audio from your headphones being to loud so the vc is being picked up by it because of that
How to fix
Ayo? @worthy ledge level 3 !!! 
@earnest dragon can u help me with this?
Ayo? @elder willow level 1 !!! 
Did you click the save settings button just to make sure
Well I did now and it still giving me the same thing
Which voice-changer are you using
Wdym
@earnest dragon
?
hi
@earnest dragon can u show me like a example of what u mean bc am confused
yo
W-okada or what
Enable sup1
привет ребят как настроить нейронку
Seend screenshot here #✨│ai-help
What about echo
you can have that if sup1 doesn't do the trick
how to set chunk eztra index
What is AI Hispano
What settings should I use I have a rog ally
Rog ally not good enough you need a dgpu to use good
Applio
hI
I only have that
Then that's something you have to deal with not having a good enough pc for ai
Damn
How many chunks do you recommend
who can call me to test voice?
Ayo? @cosmic patio level 1 !!! 
I can't recommend anything if your pc is not good enough
what i can do if in the games voice changer lagin
ola
I think it should be possible to run fork wokada with just cpu mode (or even try with its AMD igpu) if you're running it under windows
https://rentry.co/forkvoicechangerguide
Guide style is in the same as Blanc_dot's. Thanks Blanc_dot for corrections. Most technical information comes from deiteris.
Last update August 30th, 2024: New b2309 version
Translations added for:
German: https://rentry.co/ForkVoiceChangerGuide_de
Turkish: https://rentry.co/ForkVoiceChangerGuide...
HIHIHIII GUYS
hello guys !
Ayo? @stark gull level 1 !!! 
Do you have the model
That's what he is already doing
https://fish.audio/text-to-speech for zero shot and for trained GPT-SoVITS
Powerful, fast, and customizable text-to-speech solution. Ultra-low latency, rapid voice cloning, and flat-rate pricing for AI voice infrastructure.
thx!
I made a quiz application that quizes you on anything and by anything I MEAN ANYTHING, check it out (no signup required).
https://kuesuto.vercel.app
Give honest feedback on wha needs improvement as its still a prototype
alright thanks a lot! my rig's running a 4050 and I think I don't feel much delay difference with using and not using the index
glad to hear that
-rt first link
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
help plz i need this fast
Not that is free most of the free ones are not that good
is there a guide for the voice to work in discord / in games?
Is anybody here willing to train a voice model from the sample I send and then send the finished voice model to me?
Preferably for voice-to-voice purposes
although I'm not sure if the index is causing this or not but there's a cutoff in both at the start and at the end of the voice sometimes
Ayo? @merry belfry level 2 !!! 
this app isnt working 4 me smh
I’m using quest 3, what do I use to reduce delay? With the voice changer
Does increasing chunk add or reduce delay?
Cxsmo, I have a error while training a gpt so vits model on colav
Hello. I’m new to making AI covers. Can someone explain, or direct me to the proper channel, how this works? Where do you get voices from? How does the ai work? Etc.
-rvc
Suggestions for @warped warren
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
1ski, I have a error while training a GPT So VITS model on colab. Although, I have the audios, but, it saids this. Do you know what I can do to fix it?
I don't do GPT-SoVITS I'm more into this one paid tts or styletts and waiting for the new StyleTTS-ZS to come out which from audio demos sounds really good.
Okay
Hi
hello
bye
Guys, I think I solve the problem. It needs audios to be in the dataset folder. So, I fixed it
hi
Guys, I'm training my first GPT SoVITS model of Conroy Cat (Sa'Marion Holmes)
Why do you keep notifying what models you are making and keep telling people to use them.
I'm saying that I'm trying to make it myself
I just saw the Johnny bravo model that's cloned with GPT So VITS, so, I wanted to make a model myself
let them do them
lmao
bro stop dick riding conroy cat
Hi
Yeah but other people tell him to stop also like nick
It takes time to train the model
What GPU should I get at around 100-150$
rx 580
Do we know why AI Hispano have been hacked?
that's their problem not ours :p
then how are u going to use applio o-o
hi
no ping meh >:O
Hi y'all
guys wth happened w/ all this Hispano hacking? do i like have to be concerned or smth? is weights gg affected or not?
Why do you keep antagonizing everything in chat
everyone who made a rvc conversion or voice model is affected, now you have virus.
Weights gg isnt associated with Applio/Blaise at all, so you're fine
The official Applio website and RVC client are still safe (https://applio.org/)
The only real damage was that Blaise's Discord was hacked. Dont trust any information from their Discord account
they are gonna believe that 😭
oh. alright, thank you so much. i almost lost my shit lol
$ILA coin is out now!
with every conversion you make you get ONE $ILA COIN!!!
NAH CUS I ALMOST DID
pay me pls
No Applio wasn't messed with it was only the ai hispano
you clearly dont know the full story.......
i'll help ya, hop on to dms
No I'm not I was just saying what nick and them have been saying.
hey guys where can i find ai models?
uh you mean in private chat? im from belarus is dont understand very well
yeah
but nice
oh oke
You can search rvc ai voice model at:
- #1175430844685484042
- #🔍│find-models
- https://weights.gg/ (login required)
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- https://applio.org/models
- https://voice-models.com/
- https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)
if there isnt one, you can:
- #1159289738314919936
- #1191429836321849435
- make it yourself with our docs guides https://docs.aihub.wtf/essentials/how-to-make-voice-models/
@covert lake do something ples
That is not promo Nick posts this all the time I got this from Nick
i made that copypaste wdym ?
you clearly dont know the full story
i find it strange because theres links and stuff
does ilaria coin exist?
its a joke about applio getting hacked
give me one dolla and i will do it
Wdym?
i waited shit would be fixed to make it
Balbal 😭
/bal
ty ty
ty too
Is there nsfw?
No that's weird.
is hugging face rvc also safe?
yes, i recommend you get it from the applio website to make sure youre using the right repository. there could be fakes
yes all the links I listed for models is safe
RVC, not models yet
oh, thank you !
Yeah I know
Why are there so many ai‘s now?
ikr? 474,415 of them
Because some ais do better then others and the bad ones stay around.
No but like, as soon as 2024 started so many companies started using AI
It’s unbelievable
Ayo? @uneven plover level 2 !!! 
Some of them Don’t even need to use it but they still somehow do
Yeah that's when it starting becoming a lot better
Yeah from usage, but not for companies
I‘d wonder how an full lengthern AI movie would look like if AI video generating was FAR MORE advanced
Yeah for companies it became better now we have o1 preview and notebooklm which are useful and AI video like Vidu and stuff
kaggle issues be like
Yeah Runway ai partnered with lionsgate so they are training a video gen on there catalog and also we have Adobe's coming out soon
It's true but with a couple ais you can do a short film already with vidu or runway ai and also lipsync and rvc.
nha the Applio website was not compromised, just a fake web called applios was created
is it free to train my own ai
You mean voice.
-local first one
Suggestions for @scenic granite
- 🍏 Applio, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team GitHub
- GPT-SoVITS, by RVC-Boss GitHub
Credits to Faze Masta and Antasma for compiling these links.
is it hard?
you just need to mvsep bsformer or uvr to extract the vocals or speech then grab a bunch of it others can go in more detail.
"🎵 You can't stop the waves as it crash in the ocean! 🎵 " 🔥 🔥 🔥 🔥 🔥 🔥 🔥 🗣️ 🗣️ 🗣️ 🗣️
In the context of RVC, the dataset is an audio file containing the voice the model will replicate. It can be either speaking or singing.
can i send a private link
i think there is a tutorial 5 min
i wanna know if its good or bad
oh why not
they are outdated
it was uploaded 1 month ago
nha
whats the difference
LLM = Large Language Models = Chatbots like chatgpt and llama
RVC = Retrieval-based-Voice-Conversion = Speech To Speech models, used for ai covers yk
here its mostly about rvc and wokada (program to run rvc models in calls)
so rvc better for calls?
LLMs have nothing to do in this,
RVC (program) can be used to make models and use them for pre recorded audios
Wokada can use them in realtime for calls
Ollama is used for llms
If anyone has experience with setting up W-Okada's TTS Client (NOT THE VOICE CHANGER, specifically Text-to-Speech), then any help on my question in #🔍│help-w-okada would be so appreciated!!
is NVIDIA RTX 2060ti more than 1060 ti gtx
yes, it can be used for local training
Yes obviously
so u want to train an llm or rvc model
oof im at 1060 ti but i have all the other requirments will it still work
rvc
nope, the gpu is important, u can only inference (use models) locally with that one,
So its better u just do it on cloud (remote good pc)
oof
As you dont got a good PC, its better you use cloud for training an RVC Voice Model:
- Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
- Applio (ui)
- Mainline (UI)
- RVCDISCONNECTED (no ui)
- Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
- Mainline (UI)
- 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)
alright thanks
I personally suggest Mainline Kaggle as it gives more free gpu time, but its harder, and dw u can still do it
well thanks for ur help
btw
i saw some models when downloading
they have a .pth
no .index thats weird
when downloading what?
if they have only .pth id suggest to find other ones as the index helps with accent
if u could send me the specific posts that have no index could be helpful for me
yea it was only .pth and it sounded weird
in dms?
thats 1
another one
i deleted the other one i had
so thats all for now
oh wait
post link ok
its all fine
ey the applio bot was muted?
why is there some without .index tho?
it do not work in https://discord.com/channels/1159260121998827560/1163592055830880266 also the bot is safe
@night lake is there a reason why those models don't have a .index ?
yep for security reasons before
i have no idea
im asking you why some dont have index
yea im not asking u, im asking the creator of the model
ohhhh thanks
sorry
ima go sleep good night
why does it say invalid username or password when i try and download a voicemodel
it's fine, it work in ai hispano, the real applio website was not touched cuz they cloned/created a fake one (Applios)
Ayo? @ashen vapor level 1 !!! 
is it possible to copy intonation from a voice in another language?
I need to translate voice whilst keeping almost the exact intonation. (I don't mind if 5 AI's are involved in the process or not)
hay, some can do one ai edit for me now? i have 5 seconds clip
Because most (if not all) of my models are made for realtime and index files are actually useless in realtime
@atomic vector hi
Hi! So, it seems like kittyme555 is talking about a situation where someone created a fake website impersonating Applio, the open-source voice cloning ecosystem. It's concerning to hear that someone might be trying to deceive people by creating a fake website.
Just to clarify, the official Applio website is indeed https://applio.org, and if anyone needs help with using Applio or has questions about specific models, they can refer to the official documentation at https://docs.applio.org or explore the models available at https://applio.org/models.
It's always important to be cautious when browsing the internet and make sure you're on the official website of a service or platform to avoid any potential scams or misinformation.
What's your take on this situation, and do you have any questions about Applio or voice cloning in general?
oh never really used wokada
@scenic granite
the vocals and instrumental aren't showing up so i can download them
why does it say invalid username or password when i try and download a voicemodel
Ayo? @mint merlin level 1 !!! 
@atomic vector how to import a voice model for tts
Importing a voice model for Text-to-Speech (TTS) can vary depending on the library or framework you're using. Here are a few common scenarios:
1. Using Python with a library like gTTS or Pyttsx3:
For these libraries, you'll need to download the voice model files (usually in .zip or .tar.gz format) and then import them into your Python script.
For example, with gTTS:
from gtts import gTTS
# Download the voice model file (e.g., 'en-us.tar.gz')
tts = gTTS('Hello, world!', lang='en-us')
tts.save('hello.mp3')
2. Using a cloud-based TTS service like Google Cloud Text-to-Speech or Amazon Polly:
In this case, you'll need to create an account, set up a project, and then use their APIs to import the voice model.
For example, with Google Cloud TTS:
from google.cloud import texttospeech
client = texttospeech.TextToSpeechClient()
# Create a voice configuration with the desired voice model
voice_config = texttospeech.types.VoiceSelectionParams(
language_code='en-US',
name='en-US-Wavenet-A'
)
# Synthesize the text
response = client.synthesize_speech(
input=texttospeech.types.SynthesisInput(text='Hello, world!'),
voice=voice_config,
config=texttospeech.types.AudioConfig(audio_encoding='LINEAR16')
)
# Save the audio to a file
with open('hello.mp3', 'wb') as f:
f.write(response.audio_content)
3. Using a self-hosted solution like Festival or eSpeak:
For these solutions, you'll need to install the voice model files on your server and then configure your application to use them.
For example, with Festival:
# Install the voice model files (e.g., 'cmu_us_slt_arctic_hts' for a US English voice)
sudo apt-get install festival-cmu-us-slt-arctic-hts
# Use the voice model in your application
echo "Hello, world!" | festival --tts --language en-us --voice cmu_us_slt_arctic_hts
Remember to check the documentation for your specific TTS library or service for more detailed instructions on importing voice mod
Ayo? @eager sand level 1 !!! 
els.
If you're looking for a specific voice model, you can check out the Applio models for a wide range of voices and languages!
nevermind i fixed it
El skibidi baño
Ayo? @keen fjord level 2 !!! 
introducing myself 👍
bass teacher
gente no funciona :,v
gonna plonk something in here against my better judgement but i would really appreciate if i could get just a bit of assistance on gathering a dataset because ive been bogged down by college and trying to find a place to download an artist's entire discography let alone sift through uvr isolations for usable audio is gonna take me more time than i can spare to waste
Poo
im sure the request model place scarcely gets looked at so this is like a little last ditch effort before i pass away and die
Wood nut in cider
im not asking for the entire model to be made for me from data to train i can do a good amount of the work
No one is going to do this for free
im just asking for a little bit of assistance so i can stretch the rest of the nine yards myself
UVR or Mvsep and audacity and youtube is all you need.
how would i download an artist's entire discography from youtube in bulk
Ayo? @warped grail level 2 !!! 
Loader.to with a public playlist
well at least that's something
Data also needs to be cleaned up with RX11 if you want a very good model. If you are willing to do all of that I will train it for you when I have the resources available.
that is the one thing i dont need help with
i can train it myself just fine
i have rx 10 in my posession tho
anything significant in 11 ?
RX is not needed if the vocal splitter is good
Oh yeah I forgot to mention that haha. I just always put my stuff in RX11 even if I get a good split
Then yeah the data prep is what model makers would charge for as its the most time consuming part.
Why
I need someone whos good at ai to get something done for me
helps to give more control
Ill pay good if need be
What ai
finetune stuff
video editing
i get it
Clean up clicks, ess'es, etc... depending on where the data came from.
Face or what
Yeah sort of
i can see that being the case, it also takes up disk space too, well thanks for the info anyhow
Nothing weird LOL
Ayo? @hazy palm level 1 !!! 
i DO have money but id rather spend that on sprite artist commissions
Yeah i dont use discord very often
someo ne please help
Honestly if you want to make a lot of models I would learn yourself. Takes time at the start but its worth it.
ive made many models
Okay so what is it if it's Face I can help
Thats my opinion anyways.
Can you dm me?
i was only asking this time around because ive been in worst straits than usual mentally
and i just feel weaker when it comes to tedium in recent times because college has been sapping my work ethic
Why would you need that....
Okay dm'd
I get what you mean.
ah well its fine, i gotta get back into sorts anyway
How do I use Applio v3 in paperspace
i need cute voice changer
you need help on your english most likely instead of a model
Ayo? @graceful forge level 1 !!! 
wtf am I typing
is it possible to copy intonation from a voice in another language?
I need to translate voice whilst keeping almost the exact intonation. (I don't mind if 5 AI's are involved in the process or not)
Whats the best voice changer
Are there any text to RVC models
generating audio from scratch with text
tried running TTS though a model and it still sounds a little iffy
This usually works better than human recorded audio
are there any good ones
11labs
yeah on that rn but I don't think I can download mp3
you can
lmao
I'll try my best to Improve my english
frequent errors occurred. please check if the model of the framework being targeted is loaded
Can anyone help with this one
hello, does anyone know how to use MDX-NET kuielab_a_vocals on UVR 5? or is there other way to separate 2 vocals in one song?
Chat? Anyone knows that website where I can just paste the zip file of the model then it will do the voice changing work???
is any1 here familiar wit koyha
kohya ai
@plain cove
sorry for ping but i was looking in ai art and im wondering if u could help me?
im completly lost trying to use lora and theres not discord servers thats for this type of stuff
hallo
Ayo? @livid heart level 1 !!! 
he dont own the server tho o-o
what a sad person
tea owns the server
not this dude
"Self Important"
warns half million people to not be scammed
Ok
did u time him out?
Wasn't me I literally just opened the app
some people are just sad
wot app o-o
leave them be
i mean ad
Oh nk
hi
Henlo
how to get tts
mami
Hi im mr stanmer, im in love with ai
Hi

AI HUB Docs