#🧬│ai-chat
1 messages · Page 223 of 1
dont forget* ai art doesnt have copyright
if the owner post it in a website that asks for payment, then that's where I understand
^
the input, yes it has, the output doesn't because it's a new thing that never existed
are artists supposed to paylock everything because stealing is magically okay when it's pubically viewable lmao
lmaoo
lagger why are you arguing in AI HUB
Ayo? @upbeat trout level 6 !!! 
was just trying to find a way to depoison and it became this
depoison? nani?
I can stop it here
never heard of nightshade?
uh... no?
leave poisioned images alone lmao you're not in the right here for going directly against the artist wishes
Ayo? @uncut olive level 2 !!! 
nightshade is a program some artists use to 'poision' images so they cannot be used to train AI. this dude wants to undo that so he can train AI with images he obviously doesent have permission to do so with lmao
Nighshade is the anti-stable diffusion. it poisons data sets to corrupt outputs made by SD
I'm with lagger on this one
hey, the artist is posting images for free to be seen for free an downloaded for free.
If you're a truly talented artist then ai won't hurt you
If your work speaks to people they'll always gravitate towards it regardless
but I can end the arguin here
If someone wants to use it then so be it
please, encourage everyone on the internet to paywall everything. that would be totally epic and fun dude!
so BASICALLY what you guys are saying is that people are now embedding malware so to speak into images that are meant to be corruptive to AI training, as a countermeasure to people using images from the internet to train their models
did i get that right?
yes, nightshade is malware
a malware designed to be a counter against people training with certain images that contain the code
I saw someone compare it to "putting hot sauce in your lunch at work so no one else eats it"
Thought it was apt
nah
bc if someone else eats your lunch, you don't have lunch
There's a difference between copying and theft
Yea I'm in agreement with you
Well time to make a Latina e-girl model
Point is that legally you're in the wrong if you do that to your food and the other person gets sick
bro, that itd be like if i download a video from youtube in order to use in an editing software to make a new video (Which I do often btw) and all of a sudden it corrupts my save file because how dare you use my video
Just because someone does something to you doesn't make it okay for you to destroy their stuff or get them sick
whaaa
@pale basin @spiral junco
anyway
yes
Ayo? @spiral junco level 6 !!! 
whats a tortoisett
it's ai tts
is it edible?
can i eat it?
basically, marcimus, it's already implemented on yotube called a strike, age restriction, shadow-ban, etc
nope
Ayo? @north peak level 1 !!! 
The signal is shit here so I can't look it up to confirm
or if people start using your art for purposes you clearly don't want them to, usually against copyright, you're able to protect yourself
my brother in christ
first of all
dont watch while signed in
second
youtube downloaders dont care
copyright booo
but whatever lmao if you want to label it as the artist being the bad guy here for wanting their stuff to be used non-maliciously sure. Just dont download malware in this analogy
got a better solution: Remove the ability to right-click-save, and inject a cookie/cache to prevent computers and phones from taking screenshots.
that's straight up what some mobile apps already do with sensitive information LMAO i'm not sure what your point is here
you can't do that everywhere
and you'd just want to bypass it already, not hard
if you're trying to un-do work artists did specifically against the thing you're doing, disabling right-click save won't work
oh, forgot to mention, only in sites like Patreon.
As long as an image is posted in a site accessible for free, we can yoink it for free
True, but would it still be traceable to that one certain image from the final prompted result? if the model isnt intended on emulating the asthetic of the artist in question, then id say: probably not.
I got shit to do too, ciao!
it doesent matter though!! it's still using someone's art that they obviously don't approve of lmao
Good luck!
It's possible that both of you have valid points
mariopilled
my ass is NOT nintendo 😭
I think someone reported me to the gym security since I was in the shower for nearly 4 hours lol
I make poor life choices
wow, you have a gym shower where everyone can't see your balls hanging out?
nintendo is a multi-billion dollar corperation lmao steal from them all you want. small artists are pretty cool though
Don't you?
not typically in america
Damn lol every gym I've been to here in the good ole USA has normal curtains
Have you tried planet fitness?
nah, I haven't been to many gyms. I don't like to exercise around other people
Yea I don't like to either, that's why I usually go late at night when only a few people are there
nice
Although I couldn't work out today since my arms in a sling
rip
btw, do you know if "pause" and "deactivate" are built-in command-line commands?
I'm trying to read the error messages I get when installing TortoiseTTS, but the command prompt exits whenever it finishes
Pause is Ctrl+c or Ctrl+s
Ayo? @north peak level 2 !!! 
Screen record it
i dont have OBS
git submodule init
git submodule update --remote
python -m venv venv
call .\venv\Scripts\activate.bat
python -m pip install --upgrade pip
python -m pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
python -m pip install -r .\modules\tortoise-tts\requirements.txt
python -m pip install -e .\modules\tortoise-tts
python -m pip install -r .\modules\dlas\requirements.txt
python -m pip install -e .\modules\dlas
python -m pip install -r .\requirements.txt
setup BnB
.\setup-cuda-bnb.bat
del *.sh
pause
deactivate
I just tried OBS out the other day how funny
I'm wondering if I can just remove those last two lines to read the command line
I think it's worth a try
I'm just worried it'll fuck up something that im too smooth-brained to fix
When I find myself worrying about that I just make a backup copy of whatever I'm working on
That way I can always go back when I inevitably fuck it up
nah I mean like, maybe it'll mess up my installation of Python
or venv, whatever that is
I think I fucked up my venv. Now I just avoid anything that mentions it
Oh the argument's over sick
I saw the same people typing and assumed you guys had been arguing for like half an hour
idk i just joined a few minutes ago
I love when people argue. It's fun to see things from the Pov of multiple people
It's help me to expand my own perspective of the world
I enjoy arguing because I watch other people argue way too often
Well I tried running as admin, but now I'm getting completely different errors from before
Are you able to at least see this error?
yes
ERROR: Could not open requirements file: [Errno 2] No such file or directory: '.\modules\tortoise-tts\requirements.txt'
LIES
oh no
running as admin
If you are using a virtual environment just use the following line.
pip freeze > requirements.txt
automatically starts the .bat file in system 32
Google told me that
So I have no clue if it'll help
Yea running command prompt as admin does that
is there no way to edit a .bat file as admin?
You can open the bat in notepad
For mangio rvc local is dataset just a folder i make to put all the audio file inside?
the command line is auto-exiting right after that ".\setup-cuda-bnb.bat" line. It's always after bnb runs but before del *.sh
it makes no sense
You're right. It does not make any sense. I'm as lost as you are right now
I'm trying running the .bat file with a command prompt already in administrator
it's taking a long time, despite everything being already downloaded
so that's a good sign
alright, now start.bat doesn't immediately die
Ayo? @north peak level 3 !!! 
Yea slow mode sucks, once you hit level 5 you can join us in the verified chat where slow mode doesn't exist
Aw man verified chat used to be Poppin off back in the day. You joined back in November so a little after I made my first departure but man was that the best back in September/October
Anyways back to helping you
Nostalgia got me getting off topci
still downloading
im on university wifi so
...
64%
67%
69%
aaaand now I have to download something else
Oh shit you're in uni? What's your major?
take a wild guess xdd
well it's running now
but the first thing it did was
make me download more things
Computer science
I'm in my first year, but I have tons of credits from high school
so I can load up on CS courses in the upcoming semesters
I may even get to 400 levels next year
https://www.youtube.com/watch?v=b7kmP1fsGg8 versus https://www.youtube.com/watch?v=-Pf97N9PnzM
We need to find that owner of that AI generated music, tell him/her to release the full thing and say that AI > Taylor Swift 
Guess you can say, we shall do a Forthnight challenge - where we find the owner within 2 weeks to the day
Listen to “Fortnight (feat. Post Malone)” by Taylor Swift, from ‘The Tortured Poets Department’.
►Subscribe to Taylor Swift on YouTube: https://ts.lnk.to/subscribe
►Shop Merch: http://taylorswift.lnk.to/store
►Follow Taylor Swift Online:
TikTok: http://tiktok.com/@taylorswift
Instagram: http://instagram.com/taylorswift
Twitter: http://twit...
I don't have the right or own this music
#taylorswift #postmalone #fortnight #lyrics #fyp #ai
I figured out how to wait faster
While I download to the C drive in 1 window
I move games from my C drive in another
so many downloads
fuck it im going to sleep
hopefully this finishes by morning lul
Man that's fuckin awesome! Keep it up you're gonna go far kid
thx xd
that dont help like do u know a spesific one
Does anyone know like a really good model maker or model that’s already made im trying to egirl voice troll in dahood and I was tryna sound like how duckus did in his video
i got some
oh word
check my kofi
is it in ur bio
ye
Does anyone know how I can generate MCQs from a pdf?
hi
Mcq?
Multiple choice questions
Ayo? @green yew level 1 !!! 
a pdf? what is it like?
give chatgpt a try ig
You can upload files to gpt?
I don't see the promt to
GPT-4, or some other LLMs
Large Language Model
Basically chatbots ais in simple terms
Idrk one that can transform pdf into a MCQs tbh
Maybe https://www.quizwhiz.ai/ ? Dont know much tho
I am pretty illiterate when it comes to stuff like this sorry. But thank you
Dw yw
what is the pdf content like? if just plain text, copy-paste the text
hi
Hi
Hi
Hi
i need help with smth
im trying to make a ai voice for smth
i have a clip of the voice i want to make
so how do i make a model
Check this out https://docs.aihub.wtf/essentials/how-to-make-voice-models/, if you got any questions or errors ask in help channels
can you make a off topic screensahre channel?
Promo only in #1159290752195633273
oops sorry
Dw
what does the gpu section mean which gpu do i put it on 0 1 2 or 3
Open your task manager, it'll tell you which GPU is which. Then just select the number that correlates to the best GPU
On my PC my Nvidia GPU is 0 and the Intel integrated card is 1
where do i go in task manager
also lowk is it hard to make voices because i was thinking about trying the voices dont sound bad but in the end the egirl voices sound too robotic
My guy sheeesh 🔥
Performance
not sure what exactly im looking for in performace
Ayo? @hollow pewter level 2 !!! 
the GPU number
it doesnt say that
Also do you actually have an AMD gpu?
i got a 4070ti
Then you got the wrong version of wokada anyway
wait wdym
how do i have the wrong one
the one with DML you downloaded is specifically for AMD users
yea just delete the entire folder
aight
also do u know how to get the crepe setting
Please stop watching the tutorial that you are watching
Its all outdated xd
For your gpu youre gonna use RMVPE instead of crepe, and like 80-96 chunk and 16384 extra
why is the 4070ti bad?
I never said that ?
oh well i js asked becauase i thoufgt 80-90 chunk was low
The lower the better
oh
It controls the delay
80 should be good on 4070ti
Yeah just dont use yt tutorials lmao
16k extra, don't trust the outdated and misleading yt tutorials
not only the suggested chunk and extra, read the pinned guide carefully in #🔍│help-w-okada
Ayo? @coarse vessel level 33 !!! 
hello, how to make a custom voice?
ok now im having a issue where the voices sound like gibberish whenever i talk
thanks 
yeah now it sounds like gibberish whenever i speak how do i fix that
I dont know what to understand under "gibberish" do you have an example
ok so whenever i speak into it the AI voice sounds like im speaking gibberish and not actual words
when before i didnt have that issue but now it seems like i do
i mean i am testing the default models but i didnt have that issue with the default models before
send a screenshot of wokada in #🔍│help-w-okada and also advanced settings on the bottom
sent
anyone know realistic egirl ones?
Hello !
hi how are you ?
Hi
shit bro im tryna find them 2
anyone know if i can make a voice feel more energised some way?
Ayo? @frigid dawn level 5 !!! 
hi
looking for someone who can create rvc french voice of a girl (will pay)
Is ollama good?
You mean as in speaking or singing in an energized tone?
speaking
Oh, I see.
Have you tried to use inference audio with an energetic voice tone?
well yeah but the person im trying is like overly energetic
wish there was a advanced setting for this ngl
merhaba
ello
hi
Hi
hi
Is this server for all kinds of AI or does it only focus on voice models?
It focuses on various sorts of ai, but mostly voice models.
#1163571683848900629 updated to 2.5.2
[2;31m[2;40m[2;40m[1;31m[4;31m[4;31m[0;31m[0;31mHello, guys
Check on https://weights.gg
Create with our AI tools for free. Generate AI voice covers, text-to-speech, and more. Join our community of creators sharing RVC and AI voice models.
Models are safe
Thanks

Hello
Ayo? @elder willow level 1 !!! 
👍
Hello
This technology is freaking me out what software
i am dumb i uninstalled the program and now i dont know the software anymore haha
what was the name of the software
Ayo? @honest temple level 1 !!! 
viggle?
what software
like... the realtime voice changer?
yeah thx
i forget the name haha
W-Okada
ignore the rickroll 
I left my PC training for 10,000 epocs... Took quite some time. Turns out it stopped at 9001, and the voice model sounded exactly like Goku.

never seen somebody go that far
how long was the dataset?
Hehe, that was a joke 😉
Bro had us in the first half
Serious question though, at what stage do you find it doesn't improve much?
With tensorboard you can monitor the point of overtraining
_ _
ehhhhh
It really really varies
the other thing I want to do, is merge mutliple voice types... seems to work well, surprisngly
but from what I've heard, after like 2k epochs on a 10m+ dataset its improvements will probably be neglegible...
yeah it does
e.g, data sets quite different
but might kill range
Yeah, true
C

DM for cheap commission voice models 
hi
no need to post this here lol
who has valorant ai voicelines
if u want models, go to #1175430844685484042
or weights.gg
Dm for nothing 🔥
hey guys! new here, how do I apply the voices I download?
@stray merlin test
what do you wanna do with the voices
just the hf link
can someoen give me illaria's discord?
use it as a realtime converter
realtime conversion orrr
EOL - No further Updates
Github - Blanc-dot
Discord - Blancdot
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful Links - You need to follow this if you are on an AMD or INTEL ARC graphics card
Frequently Asked Questions
Jump To The Important Sections
Other Links
Downloa...
@hollow tangle
should help
thank u dude
ty
Ayo? @weak thunder level 2 !!! 
yw, let me know how it goes
in case u have any issues head to #🔍│help-w-okada
where did he say he wanted to use realtime?
i mean they clearly said "use it as a realtime converter"
so idk
nah its ok
Hello all, Chayanne model is here?
Hello
not sure, but you can search for it
ello!
hello
i cant hear the models when i speak
check command prompt
also, move to #🔍│help-w-okada
h
Ayo? @pulsar rampart level 3 !!! 
Can anyone help me make a voice model that fits my voice perfectly and doesnt lag and sound robot
what are some things yall recommend i do to clean up audio after separating stems with BS-Roformer?
like, my first idea was maybe normalize it in audacity or something
check this out https://docs.aihub.wtf/essentials/how-to-make-voice-models/
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.
the options are endless tbh
with uvr denoise.. dereverb.. with rx spectral denoise, declick.. then do the audacity stuff from the guide
what guide?
the one that Nick linked. On the dataset section

да
Rafadan tayfa voice no
hello
ello
can anyone point me to a channel or something with directions on how to use the ai voices
nvm im slow sorry
it's ok dw
Last update: Mar 10, 2024
thank you
where do i find the
250 Epochs.
RVC V2 - Ov2 Super.
Targets 40000kbps.
Ayo? @rich iris level 2 !!! 
this is so confusing to me, idk if im doing something wrong but nothing is working
Friends, I just want to change the lyrics of a song. Is there such an artificial intelligence?
remake it i guess, or leave whatever lyrics you want and replace others
Ayo? @mossy geyser level 1 !!! 
is there like a youtube tutorial for this somewhere
is the rtx 3070 good for most ai stuff?
an official one
@lusty whale hi

guys what does index do lmfao
hi
hi
how are u guys
what
yeah no sadly
sadly no
Heyy
hello
hey is it true that the better mic u have the better output you'll get?
ds
ds?
That's how it is with general chat. But to answer your question, yes. #🔍│help-w-okada
level nasıl atlanır
beyler ses modelleri nerede hangi kanalda yeni girdim
is there a way to create a replicate project that runs rvc as an api?
Ayo? @rancid sentinel level 1 !!! 
Hello
hi i'm gura hasband
WASSUP GUYS
hi
hey guys
hello
Is there a jake future Ai
hi
😔
Hello fellow touhou fans! So is this a good place to start here ?
My online friend never played Touhou before
@lethal karma, I have found 1 results that match your search!
anyone here ever used w-okada on collab?
Hey does anyone know an AI capable of generating videos in this style?
Idk if I consider myself a touhou fan, I like the games and stuff, but I've only played Embodiment of The Scarlet Devil
He should definitely play that, touhou 6
yeah why?
I believe no AI is capable of doing that right now
Good evening good people, could someone give me a good video tutorial on how to do a cover with AI?
I mean
technically yes?? if you have such a bad mic that it's not able to detects inputs clearly that will be worse but generally speaking it should sound the same for 100$ mic and a 500$ mic
Hey y'all my friends!
Hey friend! How are you doing on this beautiful Saturday afternoon
how do i convert an audio clip to ai
Is replay safe lol?
is there a jin woo?
It's fun for my online friend, but no translation on steam deck 😢.
you mean how to make a voice model? check https://docs.aihub.wtf/essentials/how-to-make-voice-models/
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.
ye
ey
Ayo? @dapper aspen level 2 !!! 
Out on all platforms :
https://open.spotify.com/artist/39Z1xNqwFbj2qr236jA5My
Best Phonk Playlist :
https://open.spotify.com/playlist/16XQKRJQQh9Op0aDgsoaTX?si=c591cea8e7ba45d5
Let's Go!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Time bomb ⬇
🔴 Spotify : https://open.spotify.com/artis...
how do i turn google drive link to hugging face?
Why is Replay so slow brah
hey y'all
boutta install linux on a small 300gb hdd i had laying around which one would be a good option
also what are some important things to install
i think i could do it on my other 1tb one but it only has like 300gb available as well so
i guess i could get like 400 or something if i uninstalled a couple of games but i'm not sure if it would be worth it since i only plan on running some ai stuff
Ayo? @real estuary level 1 !!! 
bro where is turkish voice models
go for linux mint maybe... not really a linux user so 
search for em on https://weights.gg
Create with our AI tools for free. Generate AI voice covers, text-to-speech, and more. Join our community of creators sharing RVC and AI voice models.
turkish?
like from turkish folks?
asking because as far as i'm aware of all voice models can do their thing regardless of the language the original audio or text is in
Python, pytorch, cuda
oh i'm only installing linux because i have an amd card so i guess it would be rocm in this case
Is that their cuda equivalent? Recently haven't kept up with Nvidia cards recently
i believe so
Oh cool, what card do you have?
Oh shit that's one sick card. Must get some great performance from it. Especially gaming
I have a GTX 1650 laptop card
Not great but it gets the job done
Alguém poderia fazer uma voz do Castiel de supernatural pra mim por favor
yeah i mostly use it for gaming but i probably should've gone for a 4090
since i wasn't interested in messing around with this stuff before i didn't even consider it
What’s the best rvc training for quality
-rvc
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
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
Aw that sucks, but it looks like the 4090 costs twice as much as the one you got so that's always a plus
-collab
You could always try and sell it and buy a 4090
-colab
- 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
- 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
nah i don't have any reason to do it right now since i'm already installing linux
Ayo? @real estuary level 2 !!! 
Shadow has a great 2017 die lit carti model
deep voice tho?
batch sizes r all tiny im guessing given the lack of actual studio quality acapellas
but im jw
Nah not deep voice, it's classic carti before he started doing the baby voice and the deep voice
Shadow also has one for the MUSIC era carti which I assume is the deep voice?
im asking bout deepvoice specifically
yea, all the recent leaks/tracks
carnival, 2024, eviljordan, fein, etc
Yea it's made from evilJordan and backrooms.
whole model is trained on just those 2? thats like 6min of data 😂
Yea but that doesn't mean it's not good,btw it's 300 epochs
300 epochs isnt necessarily a sign of quality tho
can still have a bad model at 1000 epoch
or a great one at 100 depending on ur data
Absolutely! But shadow is a great model maker so it'll probably be good. They also have an older carti deep voice model from 7 months ago
how can i get file perms?
pic perms etc
Pretty sure you gotta reach level 5
-level
I think that's when I got them
Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote
✨ Tip: Use /card to modify your rank card
soon inshallah
Yeah you'll get there in no time, the help channel is another good place to level up quick
no bullying in the halls
why are most models here 300-400 epochs max? do you guys train your models on your local computers?
It's hard to find good audio, better to make a low epoch model than one using shitty audio
good point batch sizes are all pretty small
Selam
Arkadaşlar benim ysuuf
Merhabalar
Nabionuz chet
Ben iyiyim siz peki
Pubg falan kim oynuyo mesela ben oynuyom
Shhshshshdhdhdhdhhdhdhdhdhdhdbdbbebe da hdhdhdhdhdhhdhdh en dbb

Ayo? @still socket level 1 !!! 
Arcane çok saol
Arkadaşlar siz napiyonuz
Ben iyim peki siz
Jajsjsjsnsnnsnsnsnsnsnsnsnsmsmsksksksksjs
Level 2 olmak için yazı yaziom

anyone vc
En güzel sunucudur agaaa
Türk olan varmı aranızda
Ben türküm peki siz
Level 2
Yes
Olsanaaaaaaaaaa
Hadiiiiiiiiiiiiiiiiiiiii
Olamiyom agaaaaa
Hadi olsannaaaaaaaaa
Hadi
Ya alalah
Hwhehejegejeevietwhwyeye ehehehejeyejevejevejgbeeigeeguejeguejbehebe bu enebbeej
Whehehhwjwhwhehehehehheheheheheheheheehehh
Hwhewhehhehehejejejej
Versene 2 level böyle yazı yok
Jajsjsnsbsbdhjdjdjdjdhe iyi iwyeueyeu3gggehejehjeeehhejehebebejehehejejejejejehehdsbbsbsheheheheheheehehehehehhehehehehehehhehehehehehhehehe
Haddiii
2 level letsgooooo

Agaa
Bsbdhdhdjdndhdhshshshshs bu ayarları yaptıktan SONRA ilk kez bir Mac OS bu ayarları yaptıktan sonra spreyleriniz bu ayarları nasıl yapılır resimli nasıl hazırlanır malzemeleri nelerdir hakkında detaylı bilgi için yazı ve resimlerin üzerine tıklayınız ve shhshshshshsjsjsjsjsjsjsjsjsjsjsnsnsnsjjsj ne kadar önemli olan bu ilk defa böyle bu ilk olarak Allah Rahmet en güzel
Once I moved my voice model, since the gpu was running out on colab... to start the train again do I have to change google account to get the gpu again?
Shjssjjsjsjsjsueueueueueuueueueueueieieuiee
Once I moved my voice model, since the gpu was running out on colab... to start the train again do I have to change google account to get the gpu again?
2 level lets gooo
Hayiirr
Ejjejskekdke shhshshshdhdhdhdhhdhdhdhdhdhdbdbbebe en çok güzel olmuş ellerine emeğine saygı ile anıyoruz bu kadardı en çok izlenen pornolar indir etiketi için içerik sitemizden takip etmeyi bir
+₺++₺+₺sijsjejddjdddjhe
Jwjsjsjsnnsnsnsns
Kkk
Since you're a carti fan I wanted to ask;
- What's your favorite carti era
- What do you think of wlr?
- What do you think of this current era?
how do i make models homie
love this
love this
love this
love THIS
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.
too complciated
LOOVE this!
Yea that's why I don't make them
how do u mute the playback on okada
You can always pay someone to make one for you in #1191429836321849435
#1159289738314919936 be specific on what kind of voice u want
ion got $
wat
No Mac version ?
hello, how do you make and train models
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.
thx
monkeybabies
Well if you don't want to put in the time to learn it and don't want to spend the money to have someone else make it then there really isn't much we can do lol
Maybe I can find you a video tutorial if that'd be easier but those tend to be outdated
No problem
do yk any videos?
Not off the top of my head, I don't know how to make models and I also don't usually watch video guides. But I can look for you
please do
The signal is shitty here at the gym so I'm gonna go hop in the shower real quick and then I'll look after I leave. In the meantime here's a mini guide I found online
tyt, i'm just getting into this stuff so i'm trying to be patient as well
-
Download the RVC version for your GPU and extract
run go-web.bat (your web browser will open the app page) -
Select "training" tab and enter a name for your model
3.Specify your dataset full path (wav or mp3 files are OK, maybe other formats as well)
- Most default settings are fine, set your epochs number
5.Click the "one click training" button bottom right
6. Wait till the end
Once done you can add epochs and train more your model.
You can test your model in inference tab.
Of course training can take time depending on your dataset length, and CPU / GPU.
Can I do this on Mac?
Ayo? @brazen spindle level 1 !!! 
^thats the mini guide. I know basically nothing about making models but I want to learn too at some point. So hopefully I'll find a video that helps us both
which webui/gui do u guys use to run ur RVC?
fam
lol whys this idiot here
this guy is everywhere 😂
I think so
@twin fog get a load of this guy bro
u wouldnt understand 🤫 🧏♂️
I use the WebUi that comes with RVC when I run it locally. I use AiCoverGen Hina mod when I'm using my cloud GPU/collab.
But lots of people swear by applio
u use Vast for ur cloud?
Yes. And I would recommend it over any other
I tried TensorDock, which has good pricing but limited stock and is harder to set up
Vast.ai is the cheapest
my friend used vast we had lots of issues though
very annoying having to set up all ur dependencies
over and over when ur maintenance window runs out
I'm a total idiot who struggles to set all this up but I found vast easy
and u have to swap to a new host
You can run collabs in vast
for inference tho, no? i'm talking about training
I think it should work. I'm gonna try it tonight or tomorrow since I've never trained before
training seems to be considerably more annoying on vast than inference
Ayo? @orchid wigeon level 4 !!! 
inference is lightwork
I got the AiCoverGen collab to work with it, so if you can train on that it should be fine
I've also used TensorDock and still have some credits left on it
not familiar with either of those
Aicovergen is dope. Probably my favorite fork of rvc
on cloud we run 4090s but my local has 1070, im prolly cooked running inference local no?
never ran locally before but wanted to test smt
I run inference on my 1650laptop
wtf is a 1650
it should be enough for local inference
GTX 1650
That's what I use
Scroll down to where it says RVC1006Nvidia.7z that's the newest version
say word
Word
Excel
i was already 1 hour into downloading the wrogn one lol
Shit man that's gotta hurt, which one?
I downloaded beta0717 by mistake and used it for a while before realizing it was old
they change the inference alrogithms no?
im assuming theres a marginal output quality difference
beta0717 that spew "no triton" message but still worked properly, I remember
No clue what changed. But I think I've noticed a quality difference.
still 1006Nvidia is the stable one being using
Your GPU is definitely gonna work. You've got twice as much Vram as I do and your card has nearly double the tflop performance
My main issue with local is I run out of Vram with UVR so I have to use the worse vocal separator
i wouldnt know lol we've never tested local afaik just vast and now AWS
If you want to try TensorDock I still have some credits, but it's harder to set up than Vast and costs slightly more so I would recommend against it. Unless you have a windows 10 license key in which case id recommend using it
was it MDX23C or BS roformer?
Because then you can just RDP into the vm @orchid wigeon
is it rly any different in comp to vast
i alr have credits on vast
rdp as in remote desktop?
Vast has all those amazing templates, more GPU options, and cheaper pricing.
But TensorDock let's you use windows on their machines if you have a license
Yeah
How do I install on Mac
Whichever one HP3 is
can u even use on mac? needs CUDA no?
that's what i'm wondering lol
Yeah
Ayo? @ebon mortar level 1 !!! 
@twin fog look at this guy struggle lol imagine
couldnt b me
Bro thinks were struggling.
not talking u gang u good 💯
Mac supports old versions of cuda
lol I got a vision I just need to put the pieces together
But if you can afford a Mac you can afford a cloud GPU service
What version can I use
And where
https://developer.nvidia.com/nvidia-cuda-toolkit-11_7_0-developer-tools-mac-hosts
https://www.nvidia.com/en-us/drivers/cuda/mac-driver-archive/
Appreciate you
First link is a bunch of up to date info and instructions for cuda on Mac, check that first.
2nd link is the old archived versions for Mac
Thanks
Question
Do I need to download older archive files?
whats the vision
nah it is standalone
I’m so confused😭
I have no idea honestly, I'm pretty sure you'll have to download the latest version that your GPU supports
Ayo? @umbral pendant level 1 !!! 
omg!!!
How do I know what my gpu supports
okay what is going on 😂
What GPU do you have?
nice little link up
get out of here kid ur ruining the veil of secrecy
I’m so slow when it comes to computers, mb bro
do u think @brazen spindle missed us
he's following us everywhere
lowkey miss the guy
Ayo? @jade bronze level 1 !!! 
I’m about to give up💀
@brazen spindle thoughts?
best e girl model?
No that's all good man no need to apologize, everyone has their strengths and weaknesses. I have used a Mac since high school so idk where to find the GPU info
I downloaded
Ayo? @ebon mortar level 2 !!! 
A gpu
You can't download a GPU, the GPU is the graphics card your computer has
I assume you mean that you downloaded cuda?
lmao just imagine downloading a gpu
lol whole gang here
weird yute why r u dming casey
e
ong dont ever do some shit like that again
say wallahi I did
according to casey 🤷♂️
Yes
I don't even have dms with casey
@pale basinyo I added you
if you find any videos dm them to me
say wallah
wallah
catfishing gon go crazy
I got you
who made the voice models for drake
do I choose 48k for training if my dataset is this?
bro is there no obama model in the discord
shows it's hitting max on spek
Hello
hello
yow
@pale basindo yk if the archived versions for mac are supported for apple silicon
Ayo? @brazen spindle level 2 !!! 
Cuda is a Nvidia only thing. Say your chip does not support it
Yo
where do i put my index and my pth?
There is a folder, it's just in a different spot
Let me grab my laptop and I'll lyk where
so pth goes in assets>weights and index goes in logs?
Yeah
Yea the go-web.bat
Does Mac have a version?
nvm it just took bare time to load
No, Mac has METAL which is their version of CUDA
i have to feed it a path to a file?
Ayo? @orchid wigeon level 5 !!! 
windows>mac
The file you want to convert?
Yeah that's it
what do i need to know about the settings
Rmvpe is best

You already know about the search feature ratio?
Also you might wanna start renaming your pth and index files otherwise it gets real confusing real fast
It works best when the pth and index have the same name
So I cannot use bootcamp
?
I'm not familiar with boot camp, what is that?
Ayo? @pale basin level 13 !!! 
Booting up windows on a Mac
Or would that also not work
That'll let you use windows, but CUDA still won't work
What if I download a CUDA
no i dont
idk bout any of them tbh
So I could use a cloud GPU service
do yk any?
@twin fog aint no way
@twin fogmiss you
Search feature controls how much of the singer's accent and shit there is. Making that higher will usually make it sound more like how the singer actually sounds. But sometimes if you have it too high it can cause artifacting
I use vast.ai, it's the cheapest paid service
and the other 4?
should i leave as is
okok
I don't ever really fuck with them
ur just some peter onna laptop doe
||hey buddy ||
😂
you seen the social network?
yes
then you know why I can't tell you
i dont but go off
dont shush me kid
u dont have motion
u baited it brother 💔
we were like a family
rs we rlly were
it was supposed to be @orchid wigeon @jade bronze @twin fog @brazen spindle till the end
now its just @orchid wigeon @jade bronze @twin fog 💔
I envisioned our nightly vc debates would eventually happen face to face
very very unfortunate
when should i transpose @pale basin
@twin fogInshallah we reunite one day
🇼
bro is saying everything but an apology
crazy
Not that deep
Ayo? @brazen spindle level 3 !!! 
but I did text him a couple days ago about it
Hopefully Atticus got promoted
crazy
I keep transpose at 0
always?
if you want to keep it in its original key then it should be 0
Oh I'm slow yeah keep it at 0
Unless you're doing a song with a major difference in pitch
still the wrong one
chat is he ok?
yeah this is how i normally act
i pop in once a week say rando bs for 10 min
leave
yeah i dont go on the server for 1 week
thought u told me to leave
NO FUCKING DOORNOB
lmfaooo
mofo why write it in a diff msg
n00b #pwnd
learn to TALK IN ONE SENTENCE
Mgk is ass
that's what I thought
mgk better cuz he puled megan fox
thought u were gay
yeah now u trippin
:P degen moment
also HAPPY COLUMBINE DAY
24 years ago today
many people survied the worst american tragity
yur
whats the best lunch meat to stick your peepee in?
umm
Ayo? @fair anvil level 2 !!! 
cuz i think it has to be black forest ham
Hot take but suno sounds like trash
Pretty sure he said that as a joke but I wouldn’t be surprised if he was
Who is that is that like some kind of potato
Tater tots
tots-totally
Oh wait it’s machine gun Kelly
mgk and eminem are in love but megan fox is the cuck
also mgk is so a bottom to eminem
.
.
Ayo? @runic bloom level 1 !!! 
Hi
-colab
- 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
- 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
?
If there any Indian(other can too say) can you tell hoe it is
It's first time uploading a video
💀
Hi, do you guys know any movie trailer generator with ai starting from a youtube video link?
hi
Ngl my balls hurt
Ayo? @wanton apex level 2 !!! 

1v1 me

Hi
Ayo? @ionic glen level 2 !!! 
Anyone knows the best ai tts website or program? Where I can very quickly just place a pth file in there and directly type out a tts (with expressions like how "!" or "?" effects the speech tone)
Uhh, im not sure about the last part as I never used it. But we have a few direct simple guides, check out the TTS part on our guide
Real. My online friend thought Inuyasha was a shoujo series for quite some time. https://cdn.discordapp.com/attachments/1115866252729139221/1231027598172225606/RDT_20240420_0043571570079001892354618.jpg?ex=6624534a&is=662301ca&hm=6d1dbfa3d8facbb2a9dbf48ba7cea912ae22261a7de1dfe2eb3f5af7d45a1c3d&
how do i make my own ai voice ?
we are using rvc, go check https://docs.aihub.wtf/ for tutorials
Last update: Mar 10, 2024
thx
Ayo? @lusty violet level 1 !!! 
Hi all, can anyone tell me how to change the lyrics of a song? Like sing it yourself and then process it through rvc?
ye
Well, I'll be motivated to learn to sing xd
bro is back 
hi

what is a reverse trap
ig a female that looks like a male as traps usually are males that looks like females
only in ai hub
😭
or poopmaster
you say
how to create pth file
it should be created in training
i want one for my voice recording
how?? can u dm me and show
have you already made the model?
in voice???
yes
nope
I'll send you a guide
can u show me how to
Ayo? @chilly crow level 1 !!! 
okk thxxx
yw
wer r u gon send