#🧬│ai-chat
1 messages · Page 265 of 1
Link?
Ayo? @random crest level 1 !!! 
lol
I just made at least 6 songs. I have a problem.
Don’t we all
I feel nervous before I listen to these for some reason.
Hi
any recommended free voice changers for pc?
Hey Harmony! Can I have a little help?
You get that feeling that the RIAA is watching over your shoulder or what?
W-Okada’s
No, I just get nervous easily.
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
I'm trying to make my duet sing the same lines at the same time but I can't.
Anybody have experience bcs this is hard
is the song normally a duet?
Yea...It's tous les memes by Stromae.
Separating 2 voices has always been very weird and hard
I'm doing my best here...
I think there was a UVR model for separating vocals with 2 singers but I dunno what it was
If it even existed
Basically, I'm trying to make the two singers sing the ending together, but I'm not sure how to.
Do I have to add more sections?
Wym?
Hi
how can i get rid of my voice that speaks initially and then as an echo the voice changer speaks
One more thing, how do I make sure the instrumentals aren't sung by one of the singers?
Ayo? @random crest level 2 !!! 
WHY WONT THE INSTRUMENTAL AND SINGER SEPERATE!?
edit 27.06.24 deton24’s Instrumental and vocal & stems separation & mastering guide (UVR 5 GUI - VR/MDX-Net/MDX23C/Demucs 1-4, BS/Mel-Roformer MVSEP-MDX23-Colab/KaraFan/drumsep/LarsNet/x-minus.pro/Ripple/GSEP/Dango.ai/Audioshake/Music.ai) General reading advice | Discord | Table of content (or ...
Crap now I'm just confused, I don't know where to go.
literally skill issue
Looking for someone who loves singing peep and can articulate well like him . Will pay! Doing a secret project
Ayo? @strange wraith level 1 !!! 
ADORE
PT
BE
LF
ㅤ RE
ADOPT
BE
LF
ㅤ RE
PT
ADOBE
LF
help me pls | Помогите пожалуйста
hi
hi
Hello, is there a free collab that will help me train a model in Mangio-Crepe? because easy gui only trains in rmvpe
Hi
sup
Hello
Hy 😈😈😈
мяу
if you are having issues with the programs, ask in help channels
u need to level up by chatting
hi
alright
Me
someone pmo how use voice models rq
Hi I'm gahy
@orchid slate @tiny gulch
Yes, RVC Disconnected https://docs.aihub.wtf/rvc/cloud/rvc-disconnected/
Last update: Mar 8, 2024
would be better to post that in #1159290752195633273
not a right place for job posting 
.
Ayo? @loud musk level 2 !!! 
Where voice models?
hi
how do I remove backup vocals with UVR?
Use UVR-BVE model, should be in VR Arch. If it ain't there just download it in the Download Center
or whatever the model download section is named 😭 i forget
Should I do it on the acapella I generated with MDX-Net Kim Vocal 2 or just do it to the song directly
It's a michael jackson song there's so many layers of harmony lol
On the acapella itself

What song is it though
Well maybe you can help me do it differently, I do have a much cleaner acapella from youtube but it still has snapping sounds (Billie Jean)
Ah
I can't get rid of the snapping without killing the quality, I tried using the declick in izotope but it sounds worse
Do you have access? It's only a small chunk of the song
You use BSRoformer on the snappy acapella, and it can remove it probably
Unless it's easy to get access myself
hi
hi
You could also use it on https://mvsep.com
what's poppin
Not much just trying to get a Michael Jackson acapella
You're welcome, lmk how it goes :)
nice
Uhhhh hellooooo.....?
Hi
I'm new
What doesn't
Welcome
I just got banned from skrrts now I will go here now(I didn't even do anything in skrrts)
but thanks
Idek what that is
Oop nvm don't worry abt it
@forest quarry It worked kinda, the track without vocals had some snapping and the vocal track had some too but less so.
I'll run it again idk lol
Can anybody do Pokemon Masters EX AI Voices?
Lmao
Ayo? @wild escarp level 1 !!! 
Wut
Some snapping is kinda hard to remove tbf
Oh vocal my friend :<
Yea running it again didn't help but maybe I can put it into izotope afain
I CAN'T STOP THINKING ABT MY FRIENDS IN OTHER DISCORD SERVERRRR
Mmm yeah
Actually there's barely any snapping it's fine
I don't need this for anything important
T-T
just dm them idk
Did you get my question?
I have nothing I can't stop thinking abt my friends
Thanks to the beautiful caseoh Ai cover tutorial, I can now make caseoh sing anything I want😈
welp, idk
ah alr
you're the right age now so
Wordle 1.105 6/6
⬛⬛⬛🟨⬛
🟨⬛🟨⬛⬛
🟨🟨⬛⬛⬛
⬛🟨🟨🟨🟨
⬛🟩⬛🟩🟩
🟩🟩🟩🟩🟩
Ayo? @red rain level 1 !!! 
das crazy brah
code 1 how fix?
give us more details, send screenshots n stuff in #1192011222023950368
console one moment close
й
hello brothers and sisters can some one please help me with w-okadas voice changer version 18a I get 300 to 150 ms response when there is no audio but when i play any audio it gets to 20k ms and stays that will even if the input stops i use 96 chunk and 131k extra i have a rtx 3070ti 16 gb ram and i7 12 gen
what ai model would get me girls
make sure you're using the right version for your GPU
⠀
Download for Nvidia GPUs :nvidiagpu:
Version 18a cuda
Download for AMD GPUs :amdgpu:
Version 18a directml
Download for Intel GPUs :intelgpu:
Version 18a directml
Download for Mac :macgpu:
Version 17b Mac
⠀
whats the hottest ai model voice
i use the cuda version
ah
@forest quarry
can you share other settings in #1192011222023950368 as well? screenshot
bro 😭
ok thank you for helping : )
Ayo? @narrow mountain level 1 !!! 
wait what f0 are you using
i tried 0 and 1 not anyting in between
...the gpu selection buttons?
and can you please tell the use of f0 i cant find it anywhere easily on in a simple way
pitch extractor
i use the gpu there is only gpu and cpu avilable
ah i see
oh thank you
the new 2.02.24 work ok for me but it often crashes
cause if you're using crepe that might be the issue
1 or 0 it does not make any difference
yeah the 2.0 versions r unstable
...what?
it should be like... a dropdown with a few names

wait let me see properly
oh sry, i use crepe full since it is gpu intensive and does not put pressure on cpu (i never fact checked)
Hewo 👋 :3
hello : )
yeah no
use rmvpe
hi
rvmp or rmvpe_onnx
normal rmvpe should work fine
_onnx is DirectML version basically
for some it may work better, even on nvidia stuff
on just rmvpe it is saying frequent errors please check the frame work of the model i think there is some problem with my installation because bout the 18a and 2.0 were not working and i started reinstalling their different version the cuda one worked.
I used the voice changer with no problem before formatting my pc
What would yall say the best ai on here is
rn for voice changing it's still RVC
if you mean model... we don't know
Rvc is the goat
Ayo? @oak jetty level 2 !!! 
You just gotta find a model that fits the singer
fr
Yup
rvc realtime vs w okada?
Ayo? @mossy violet level 1 !!! 
i use realtime sincei ts quicker to change voices but is w okada latest versions better?
Okada is much more convenient
Better model switching, actual decent noise suppression...
Does anybody here play fortnite
Realtime is faster and lighter on resources but lacks noise supression and its harder to set it up
use nvidia broadcast, it helps
i only wish there would be a decent voice changer that you can properly scream
Rvc cant scream nor laugh

Sorry but there's no such thing like a voice changer where you can scream or laugh.
wish
Guys what part of the song would you suggest for the input audio for the best results
vocals only or full song
With Ilaria rvc
Why can't I hear myself at voice changer
Hi, guys
Is there anyone here posting an AI cover on IG ?
And i just need to match the instrumentals to the ai covered vocals
?
Vocals only
Set monitor to your headphones or speakers
you can I did you need a very good rvc2 model and set the chunks to ( 2000 and extra to 131k on a good machine it would cause a 5 to 8 seconds delay) i only use w-okada i cant get rvc to work properly and please don't try to help me trouble shoot rvc i am dumb and would probably waste your time
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
Only if there a was a guide that had all the settings and fixes
Sad times we live in
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
Hi
does anyone have the rio da yung og model cus the link is dead?
Ayo? @sacred sable level 1 !!! 
I tried ChatGPT + and Gemini advanced/1.5 pro but they are both not able to correctly read a book (600 pages).
When I ask for instance: what is the last word of this page ? They answer nonsense.
Do you know any other AI Chatbots that is actually able to do stuff with long PDF ?
The person wants to find out what ai can get the data of a book
ohhhh mb
Guessing your trying to cheat on your homework, I would maybe only give it the data that it needs
i found something called claude.ai and it summarizes books and other things like that
So for example if your question is “who is bob” you would copy the chapter about bob
I wouldn’t recommend random closed source stuff you found online (Claude is fine but I’ve haven’t messed around with it)
Most websites just use a api with a custom prompt and ui
I see, you mean dividing the work in a few more digest part ?
This might work but it's still need the actual data of the book in order to be accurate and not saying nonsense. I wonder if there's way or maybe the technology for retails is not enough powerful yet for this
*If there's a way
Like how
If it’s asking for info from a previous chapter aswell also give it that
ctrl + f on a pdf allows you to find a word anywhere
You can use that if your struggling to find a certain part
If I want the AI to cite authors or scientific paper in the work instead of only at the end, that's where it messed up because it will say something that a guy never said and it will give you the wrong page of the source. So it's still limited. If you want something somewhat accurate it might work but you only do the citations at the end then. Btw I know for CTRL + F, still thanks
Ayo? @elder willow level 1 !!! 
Just manually change it
Are you willing to set up a whole ahh LLM but not change a few words
So far with Gemini 1.5 pro, the result are decent because he actually don't say nonsense when it comes to citation even though he is not able to locate me a specific word.
just ctrl f
look what milk did to him
Because milk ages severely at high temperatures
He was homeless and had no refrigerator to store his bottles of milk.
Hey
wsg chat
Wait a moment
So that's why water is better, because it is good even at high temperatures⁉️❓
Hello Everybody
Guys
hey guysss i keep adding models to huggingface iliria rvc but they keep disappearing once the space restarts. anyone know how to keep them saved and able to be used even when the space restarts?
I was trying to download a ai voice but I can't download it, someone help me
ive tried duplicating the space as well as putting my models into weights/logs folders but they wont save
are you talking about voice.ai? because we dont offer help with that
Ayo? @cold coral level 1 !!! 
Ok thx
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
if you have a good pc do it locally, else use the colab
Are you talking about Ilaria RVC Zero ? because if im not wrong, it makes a temporary directory where it stores the model, so its temporary. @hollow tangle or am i wrong ?
yes, temp dir
yea thats what i thought by looking at the code
i need help asap
bet what do you need

I have opened a help ticket if you could check my error outputs
Ayo? @daring canopy level 7 !!! 
i cant help with applio
yes it does but i suggest you using local instances or rvcdisconnected
let me guess 10391 hours of queue ok got it
just wait their team it cant be that important
Ah sorry but i don't use paperspace. maybe @minor blade could help you
what do you guys use thats fast and reliable?
paperspace is a colab without limits and with a dashboard
my own pc
cloud based
Nah, i don't use Paperspace anymore.
Paperspace remains the best solution
ye ik but i aint paying for it so will never use it
for 8 bucks its a solid alternative
RVC and Applio are broken on paperspace now
i knew it
really? reasoning?
what about the guide made by that user https://docs.google.com/document/d/1ooG2hJrfNNLUln0reTKKIOpNBjp53G0joak50H_sQhE/edit ? is that broken too now ?
RUNNING APPLIO ON PAPERSPACE Step 1: Setting up a Paperspace notebook You will need a Paperspace Pro subscription, otherwise good luck trying to snatch a free GPU. Create a new project and a notebook. Select “Start from Scratch”, select a free available GPU and increase the auto-shutdown tim...
The error you're encountering indicates an issue with the torch library, specifically related to the CUDA dependencies. This is causing the wheel building process to fail. The error message mentions that a symbol (__nvJitLinkAddData_12_1) is undefined in the CUDA library (libcusparse.so.12). This suggests a mismatch or incompatibility between the installed CUDA version and the version expected by PyTorch.
if theres a torch/cuda issue theres nothing a guide can do
ig the paperspace of the docs are outdated, idk abt this one
yes they are thats why im asking the masters here
Does maybe another helper use paperspace? maybe eddy? 
and i have stated many times that docs are outdated
Eddy doesn't use paperspace
i used paperspace
Ah
contact their support
make applio + mangio + ilaria rvc mainline + mainline rvc on paperspace 
Errors when trying to use the run commands and errors related to modules.
mangio, ilaria rvc and ilaria rvc mainline already runs on paperspace
but didnt the one of google docs that i sent you work ?
can i integrate ilaria ui there?
I haven't tred but i don't wanna try anymore.
Oh i remember that you tried it, well nevermind then dw
paperspace is higly unsopported now
reason?
no one use it anymore
i prefer free stuff 
so what do people use that dont have a powerful pc?
colab
that doesnt have a queue
Google Colab or Kaggle
or ilaria rvc zero for inference
I highly suggest you to use Ilaria RVC Zero, Applio Colab or Hina's mainline kaggle notebook
These are the best ones till now
could you provide links to those
Table Of Contents Introduction Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Settings (Inference) Ilaria TTS Introduction Ilaria RVC Zero, is an RVC (Retrieval-based Voice Conversion) Fork made by Ilaria & mikus, running only on Hugging Face Spaces, it’s called this way...
which one has training the second?
again, if youre going to use colab, use rvcdisconnected
i hate when discord sends messages too early
did all this happen because of the updates ? because i didnt have issues on paper for over a year
SPYxANYA: Operation Memories is available now on Nintendo Switch: https://ninten.do/6051Y7SH7
Get ready to enjoy the world of SPYxFAMILY throughout Anya's in-game adventures! Anya Forger has received a new assignment from school: create a photo diary of memories! What kind of photo diary will you help Anya create?
#SPYxFAMILY #SPYxANYA #Ni...
for training rvc models, either use:
I hope someone posts the voicelines of that game soon.
highly suggested to not use UI based ones
will i get banned
If you wanna use an UI notebook, i highly suggest you to use Hina's kaggle mainline
are you talking about google colab detecting ui so disconnecting? i seen that happen on applio rvc idk if it always happens
yeah i need ui and reliance
ui and reliance doesnt go very well
should i switch to kaggle than paperspace? because im seeking for a same experience as paperspace
i need ui's no queues and high speeds even if that means i have to pay
there isnt a paid tier on kaggle, it gives for free 30 hours weekly of T4x2 or P100 GPUs, but its a bit harder to use than colab, that instead gives around 4 hours daily of T4 GPU
if you wanna use kaggle, which may be better for training
check
-kaggles
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
does it have a ui thats all iwant
why is it so important
im not very friendly with a linux command line
kaggle is linux
why is everyone anime / fantasy obsessed here
Hello, does anyone have the voice of Dalas review?
You can search rvc ai voice model:
- #1175430844685484042
- #🔍│find-models
- https://weights.gg/
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- https://applio.org/models
- https://voice-models.com/
if there isnt one, you can:
- #1159289738314919936
- #1191429836321849435
- make it yourself with our docs guides https://docs.aihub.wtf/
after you run everything correctly of the kaggles, you have a ui yes
yeah im completely lost
is there a guide for it
I have no problem with writing up a guide if you can teach him in #✨│ai-help or dms 
i'm lurking for the answers myself, sorry
uhh not really
how are you promoting a method that you havent written a guide for newbies then
that a bit un professional
This server isn’t really supposed to be “professional”
We aren’t a major corpo we are just helping people using FOSS voice cloning tools
great then where did you learn all the rvc kaggle related stuff so i can do it my self i jut need the source of the info
was it trial and error?
Messing around and trial and error
You need to learn yourself in some cases
ok is there a file manager build in kaggle ? that would help a lot
I havent made those kaggle notebooks, it works by clicking the copy&edit button on the link, then click the run cells buttons, then follow the guides for the uis as the ones in our docs https://docs.aihub.wtf, if you really need a guide, use the colabs instead, like i already told you before even telling you about the kaggles
Guides for other tools exist in #1159513888199540817 which may apply in your situation
Here's hoping they dub that and maybe my Aunt will play it.
SPYxANYA Operation Memories Gameplay Walkthrough PS5 Xbox Series X PC No Commentary 2160p 60fps HD let's play playthrough review guide
Showcasing all cutscenes movie edition, all boss fights / boss fights, side missions, upgrades, outfits / costumes, all characters, best moments, final boss and true ending, secret ending.
SPYxANYA Operation Me...
Google/bing/DuckDuckGo/startpage/Escioa/Brave is your friend
do i have ftp access with kaggle?
ok i just noticed input / output labeling
why do they make it so rough just put a manager like ppsc
I literally told you all the google colabs guide here #🧬│ai-chat message, only after that i told you about kaggle, as kaggle as i said is not as easy to use as google colab
i have already appreciated your hgelp a lot dont grab on that comment
even if its true 🌚
the thing is i dont have enough time for trial and error procedures thats why i was hoping for already made guides
impossible
maybe the best solution is to just buy a good pc
btw @daring canopy Kaggle is meant for research, if RVC is banned from the platform you might be unable to use it
I didnt mean to say anything bad ofc, i just told you that if you need much ui & easy to use, its better to use colab as it has more written guides
Sorry we use gradio
oh yea local is way better than cloud
You aren’t gonna get reliance
welcome to opensource software, we do have guides and there's a lot of videos in youtube, but unless you're willing to use a paid platform to do all that you won't have real reliance
i have workde with collab before and usually the 4 hours are not enough to train a model of 500epochs
buy colab pro
if you're looking for a paid structure
its pretty cheap
yeah but wouldnt i be in risk of getting banned ?
also this heavily depends on the dataset and batch size
idk, been using it for a whole year and nothing xd
Every model I made was trained on colab
free tier?
paid
For me free
UIs are even allowed on Colab Pro
Never got banned
in case you wanna use stable diffusion
that's what i was looking for
so the ui limitation is only on free tier accounts
yea
@storm meadow what size were your datasets usually? and what were your batch sizes
12 minutes and like default batch sizes
time / integer vallues
6-12 for batch sizes
and average epochs?
300
heavily depends on pretrain, dataset size and overtraining
average should be between 100-300
i have heard that the main rvc preset is still the best one
wassup chat
and in the near future cause of our trial and errors there will be fixed solutions
have you seen stable diffusion:
i've played around a bit
even programming, art, everything in life works with trial and error
facts
no one will be perfect from the get go
well yeah but in most cases there are some basic templates
unless you want something bland, its not gonna be perfect either
so you're encouraging the use of free tier collab ?
because apparently paperspace is broken
you can use it, but if you don't want any hassles with disconnecting, training large models without issues pay for it
its not even that pricey (coming from a Brazilian lol)
whats your currency there pessos?
Reais
bruh the voice changer keeps humming or some shit
reminds me some people quit hobbies because they expect to be good at it
noise suppression
its whispering demonic shi
well im not that type of guy
istg
thats why im having a convo with you guys rn
fix your mic, probably catching a lot of noise
else i would've given up
what in the Generation Alpha is this
ok i fixed the whispering but idk im tryna use a spongebob model and it doesnt seem audible
copsi or pepke
No one is born perfect, idk why people expect to be good from the start at an hobby
mi amor
hey my favorite italian girl
Monsprite
how is doing my bbg
cold and very anxious
why
is there any voltage consumption table when someone is training on rvc
did 3 interviews today, my hands are literally shaking
on the gpu
any good?
no bc that depends on your GPU
this ain't bitcoin mining, it wont use that much energy
it would be cool if someone made a general map
i really hope so
not sure, waiting for replies
what gpu do you suggest for training in comfort
any RTX with +10GB VRAM
RTX 4070 and up
what about an overvlocked 1080ti
no
not enough VRAM
What is the current most powerful LLM ?
me
cant be measured, but we could say OpenAI considering they fat database
is the 10+vram really that crucial
yes
cause i always trian on 4-8 range batch size
yea unless you wanna risk crashing
or literally leave your PC still with nothing open
and how much will training affect its lifetime?
Is the datase all that matter ? I think the quality of the data is more important, ChatGPT from OpenAI is pretty biased overall
1080ti has 11gb vram though
amigo you here
Ayo? @daring canopy level 8 !!! 
a lot

then its not worth it
.
why is mine so choppy bruvvv
Ayo? @daring urchin level 1 !!! 
today ill drink chocolate liquour
Nah not my aunt I lied
i know this sounds so weird to ask but can you guys help me know what an ai model have been trained on
and know what songs they can sing in
Hallo
Can you clarify what your asking
The voice models can be converted with any audio
But you will need the audio to be clean without any music or sound effects
Though models trained of singing would do better in singing
And would have a greater/more accurate vocal range
@pearl kayak
specify ur issue in #🔍│help-w-okada or #1192011222023950368
Anyone got any tips to be more expressive without the voice quality going all over the place. Checked if W-okada got any updates but doesn't seem like they update it much
Wish there were more local options. The really good programs have limited time which sucks!!!
The really good programs use the same tech most of the time
They just change the settings
Your gonna have to be more expressive
Try to make sure your using RMVPE
Damn, yeah just wish voice changers were a bit more popular
I do comic dubs
and It works but could be better
Its great that it makes it possible for one guy to do mulitple people
You might want to check out RVC, which is what Okada is based off of
It converts audio files instead of realtime
It can get you better quality
Yeah, at the moment I record it in audacity. Remove background noise and make the file as clean as possible
then push it to the program
makes the quality a lot more consistent and better
The problem is some voice models I guess aren't trained in all the ranges. So if I go all out on the voice acting. It can get pretty bad
But I guess thats more of a voice model issue compared to the program itself. But you won't know until you test the model
Ayo? @supple plover level 2 !!! 
someome help me rq #🔍│help-w-okada
i have question plz
I have just installed everything and when my PC will be restarted by a restart or will I have to relaunch the application?
ищу малышку для игр
u have to restart it
I mean i often wanted to try and make an Jungkook (male singer from K-Pop group BTS) AI covers of higher key from female singers or even male singers and I can’t stop finding it hard to find the right pitch and octave changes specifically for his AI voice
So reinstall it?
restart ur pc
That will not sound that squeaky
Are you already pitching it inside RVC
Under pitch or transpose
Yes
Try to see if you can find a different model of that singer that might have a better range
Ok
Thanks
Thoughts on raven team leader?
hello party people
Bro just said that and left
he probably left giggling like he did something hes not supposed to 💀
i dont get how hes going to say that then immediately left, makes no sense to me
anywho how have yall been
hello I am set to rvc but when I use one see it is robotic giga its normal
Hi i am biotech engineer. I focus on manufacturing now, but i m very interested at AI tech.
Ayo? @signal forge level 1 !!! 
any good virtual microphones?
Hey
I'm looking for help with cleaning some audio
https://youtu.be/1-RRrGE3sYs?si=lG2Y91watiMs3wix https://www.youtube.com/watch?v=POXBTv5lEvs
I need help removing the tape noise from these and making them sound better quality
how to verify
Hello guys. I want to explore ai agents, maybe you know current best tools or some thematic discords?
Ayo? @grim rain level 1 !!! 
hi.
Hello
could you use the voice changer to fortnite
what
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
ty
How can i find the ai voice that used most like in every video
What is the difference between MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a.zip and vcclient_win_cuda_2.0.27_1-alpha.zip
the non-MMVCServerSIO is updated 4 hours ago
AI HUB X PORN HUB O
real
Where can i find the almost most used ai voice? It sounds like teds narrator voice from HIMYM
I don't think that it's here
Why am I an android lol
What is the difference between MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a.zip and vcclient_win_cuda_2.0.27_1-alpha.zip
the non-MMVCServerSIO is updated 4 hours ago
MMVCServerSIO_win is the stable version, vcclient_win is the alpha version which is unstable
the alpha version is currently being updated frequently to fix bugs
The alpha vresion is so clean and fast, but the settings are completely different from stable version
So do I get the MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a.zip or MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.18a.zip?
Hi
onnxgpu if you have a nvidia gpu, directml for amd
You are so good omg ty
Is there a proper guide on how to use it and settings?
It's so confusing this program, especially if to use web localhost:8000 or the demo GUI
EOL - No further Updates
Github - Blanc-dot
Discord - Blanc_dot
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful Links - You need...
guide link above ^ explains everything
just use start_http.bat and use GUI for changing settings
(the guide also have good default settings)
The guide does not tell me how to use it properly
Hi
oh i believe you want to see how to use the alpha version, its the same thing
anyways i recommend u to use this #🔍│help-w-okada channel for wokada help
UVR5 on Intel Arc on Linux. It rocks!
Nono, I downloaded stable version like you said
UVR5 on Intel Arc. Just found out why ONNX models were relegated to CPU, it was because onnxruntime-openvino wasn't installed. This changes everything as a result
Elmo's song in FLAC was used for the sample
https://rentry.co/VoiceChangerGuide#opening-on-windows here tells you how to open the software
and below you can see how to choose a model
EOL - No further Updates
Github - Blanc-dot
Discord - Blanc_dot
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful Links - You need...
This is not what I'm saying
Ayo? @elder willow level 4 !!! 
When you start the start_http.bat, you get a popu-up demo gui client
But also on hosted locally on localhost:18888
Which one am I supposed to use?
can u share a screenshot of this in #🔍│help-w-okada ?
The pop up client is a web app of the local host
Same thing
hello are there any colabs like the old illaria(i cant use the new one) that doesn't require index file
HELLO
what should i use on mvsep?
Ayo? @elder willow level 2 !!! 
Ayo? @elder willow level 57 !!! 
Ok
Now, how should the aggressiveness be? I'm doing a BTR cover.
BTR stands for Big Time Rush as an understanding.
0.3?
So, I have to wait in a queue. How long will this take?
Okay.
helo chat
i got someone 2 1 month discord nitros for $5 need to get rid of it.
Hi
Yes
As long as you can change the microphone input
hey i need help
i did change it to cable output, i saw the tut on yt
Hallo
often times i I enter custom instructions for chat gbt and then it will either ignore the custom instructions or make the issue worse. for example I will tell chat gpt to keep its responses to 100 50 words or less And then for its responses it just ends up generating even more text on average for each response. I have no idea why this is happening and I can't find any resources that talk about it online

ChatGPT struggles to follow length instructions overall
If you precise the fonts and the line spacing, it is easier for him I think
is tensorboard available with Google colab or kaggle? and if so how do I access it
wym by precise the fonts?
if it's 3.5, it might have limitation like that, so instead you can try 4o, including @elder willow in #🤖│bots
i use 4o on the web app. what is @elder willow?
basically bot/app of the same 4o
free?
.
there's a way i can take that in put it in a server i create, right?
Just try you'll see
Ayo? @elder willow level 2 !!! 
try what
Brother..
what
click Add App there
Font and line spacing, try to precise it
You cannot change the font or line spacing on GPT
huh? i don't see how font correlates with custom instro, nor do i believe it's even adjustable
The user is imputing text, not a image
And plus if it’s a pdf GPT will grab the text data from the pdf, not look at it as a image
(Unless the pdf is actually a bunch of images)
I’m pretty sure the openai api you can configure max word or token count
It helps him to respect legnth instructions.
The person is not uploading a pdf
And like I said if the pdf has text data it is going to get that
Font size or spacing should have anything to do with it
(In the case of max sentence or word count)
@severe lion instead of using a custom instruction have you considered telling GPT your request as the first question?
does anyone has Amelia Tyler voice model? the creator kiyoshi94 deleted their works on huggingface
Does anyone know how to fix the bugsplat glitch on pc
you just mean at the beginning of the chat, just give the instruction right there and then request it follows that for the rest of the conversation?
How do I make comments in #1175430844685484042 ?
It probably will
Need model maker
it doesn't, usually breaks rule after two responses, and sometimes immediately
Oof
I made one model, is that enough?
Yeah if it’s good #1159514067187277865 /submit
"good" is subjective
Yeah so good luck
@storm meadowwhere are the okada ppl
im literally alone in that channel
more dead than it could ever be
it is 
Does anyone know how to fix the bugsplat glitch on pc
Neat
@ornate jungle lol
Is there any free ai or anything that lets u synchronize ur lips with the audio but really realistic?
If there is no free one what are the paid ones
Ayo? @obtuse salmon level 1 !!! 
anyone!!
Ayo? @elder willow level 1 !!! 
it's ... not an rvc model
im batman
Sometimes it's worth reading on what you download, and what it's purpose / usage, my friend
Guys is there anyone knows pyton codding we have an open source project anyone wanna join ? Its about voice mastering we need pyton llama2 or 3 and RVC knowladge
Hi, i would need some voice actors for my upcoming game, if anyone is interested, then just text me.
hi
What quality is this?
~38-41 khz
compressed 48 khz audio
Clean it up, resample to 44.1 ( preferably in rx or using SoX ) and train 40k model
Clean 48 you'd have when you see actual content
that somewhat matches lower frequencies ( pattern wise), at higher spectrum
rx huh
Naturally
sa
You can resort to a audition too if you wish
Audacity will do as well but yea, I trust rx and other tools' algos
is rx superior for denoise
Yes
did you pay for it
That should be none of your business
Nevertheless, superiority lies in supervision and ability to fine tune the noise profiling / scoping, generally, control
compared to AI approaches or automatic ones
rx11 or rx10
10, 11 ain't worth going for.
all of that " AI " hype in rx is based on outdated tech
it ain't even that great anyways.
All that's crucial, so main tools and components stay the same
de-hum, de-noise tools and so on, that's what you want
for de-reverb, you want vx plugins
Yea what do i get for what i need
cuz ive never touched rx
what even is it, a plugin?
honestly the features in 11 has WAY more details/option than 10
essentially, you get software that's used for mastering, restoration and editing
And that on it's own isn't worth paying the extra
You can get the same job done in 10 so there's no point
essentially:
that's all you'd ever need for rvc purposes
for de-reverb there's clarity vx's de-reverb pro mono
for de-delay? I haven't found neither I know any actual and good tool / ai supported tools which wouldn't damage the spectrum
Hello, anyone know how to find anime character with Japanese voice? Instead of English?
Its an all in 1 program? I dont need nothing else?
Yea
Oh! Okay, thanks!
Hi : D
hi
Hi
can someone tell where can i ask a question about ai covers on yt and attach a video also ?
Hello y'all
hello people! please help me with this https://hoyo.link/fswiFCAL?s_c=NDAQMUQJ9N
what is this 

you got this champ...i know you can do it...
true
It is a link that redirects you to a hoyoverse event for its Zenless Zone Zero game
Can't believe this server had to be remade last year, a lot of lost models resulted from it
they made another game?
yes

Well at least most of them got backed up 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.
is that so
i'd post a gif to express my joy but mfw no img perms

thank you though
You're welcome :)

mhm
HI
hello i made a hitler cover it got hate speech flagged (there was a swastika on his arm) did the ai say something wrong or was it the img ? the video is in #✨│ai-help
guys where to train models , cuz google colab doesn't work
Ayo? @gilded needle level 1 !!! 
obv the image
Cartman pfp?
i did his face on viggle ai and i got flagged
That's awesome.
real 
Nobody has ever made a Bunny Maloney voice model so I made one using weights.gg:
Yes
was it a human or AI that flagged you , because sometimes ai can recongnize the austrian fella.
So there's a way to make different voices 0.0
Using weights.gg, yea. but a bit limited.
ik weight.gg but you have to pay to do private models , so i don't use it
Ooo
thank you bro please see #✨│ai-help
Ima find a pirate one
That's why I used 3 episodes audio of Bunny Maloney
see #✨│ai-help i posted the img there if i make those canges would i be fine ?
i think bot
ai's are sometimes programmed to stop hateful content , they are trained to filter strong topics
it was a cover of the song my way it is a jazz 50s song
Maybe it was because it was HITLER!?
well its the image
yes i think same
did you see yet
yes
ok thank you : )
guys , tell me a way to train an ai model without GPU or colab or a paid website , just free websites , i have a bad pc
weights.gg its free 99% sure
Ayo? @narrow mountain level 5 !!! 
one issues , you have to pay to make it private
you can make an alt acc or maby delete the models after you make them (i have never uploded a model on weights .gg)
i don't wanna get problems because i'm putting political figures on this website
there are many political models there i would recommend an alt account though
i'm gonna try non contreversial topics incase people see it
Why not colab?
time out
Alt accounts
Why, when I use voice, do I get the error “TRY HELL” during a conversation?
hi everyone, just trying to setup the program to use in csgo do i need to use "passthru"
no
you need virtual audio cable its free on internet archive and no piracy as it was uploded by the creators (it is same as paid version)
ah thanks : )
welcome
yo
who is that ?
same 👍 😂
😏
telll fast
bye now i sleep
sup

burger with or without honey mustard 🙀
I've got it working but theres a huge delay
Ayo? @tardy mortar level 1 !!! 
Even though I can hear the voice mod in cs2 its a delay of like 6 seconds
do you use w-okada's voice changer
Turn down your settings on cs2
tell your pc specks
I'm using w-okadas
specks
3070 gpu
@tardy mortar
yes do that
which settings should i turn down
Graphic settings
If Bondrewd from made in abyss wasn't a despicable lying bag of butts and actually took her to the surface like he promised.
Ayo? @fallen ruin level 18 !!! 
Is it only me but if you have ai only and the instrumental of the song you doing it lowkey slow
Please someone say I am not crazy
hi im jo
Hi
Thanks to NordVPN for sponsoring this video!
🌏 Get your exclusive NordVPN deal here ► https://NordVPN.com/foodtheory It’s risk free with Nord’s 30 day money-back guarantee!✌
Dr Pepper has just become the number two best selling soda in the country! After years of climbing the ladder, they're now at the top with behemoths Coca-Cola and Pepsi. ...
Good bye pepsi world's best second cola
Hello dr. pepper world's new best second cola
Hi
Ayo? @gentle canopy level 9 !!! 
Pepsi died after they stopped with Michael Jackson ads
that is an embed
Ayo? @simple gale level 1 !!! 
Hi
@orchid slate What is your model voice and settings plz
Ayo? @timid mulch level 3 !!! 
I was literally drinking Pepsi while reading the chat.
Ayo? @gilded needle level 2 !!! 
Hi
Where do I go for the juice wrld ai bot? Sorry I’m new here
Found the ai link. What do I do with it to make a song? Someone please help I’ll gift nitro if someone can walk me through it
Bet
Do you have the vocals you want to convert?
And would you prefer to do it in the cloud or locally (on device)
when I upload index and model files in Ilaria Rvc it takes too long to upload the files why's that
I don’t have the vocals at the moment I just want to know how to do it myself so when I do have the vocals I know what to do.
is there any progress bar in Ilaria Rvc for uploading
Your upload speeds and the space itself
hugging space
What’s the difference between doing it on iCloud or locally? I’d probably rather do it locally
If you want faster upload you can use a download link instead
It’s not on iCloud
Locally is using your hardware while cloud is via a website
Whichever is easier works with me
Ayo? @elder willow level 1 !!! 
@storm meadow do you have experience with kaggle
Ok, so first get the vocals you want the singer to sing
In wav or mp3 (preferably in wav)
Ok let’s say I have them what’s the next step
Actually yea I do have some vocals I can use, what’s next steps @storm meadow
kedo what's your issue
I’m new to making ai songs and I wanted to make a juice wrld ai song. I see the link in voice models and I want to know what the next steps are to converting it into a song
I’m brand new to this sorry for my ignorance
do you have the model downloaded from weights.gg ?
Are they clean without music
timbo he wants inference not training
No I don’t I have a huggingface link
Either helps
go to a site called weights.gg and search for a Juice Wrld model
once you download the zip go to huggingface.co/spaces/TheStinger/Ilaria_RVC
then in the Model Loader tab upload index file and the model.pth in the correct places
go to Inference Tab upload an audio and press convert
and don't forget to select the model from the drop down menu
Ok I’m doing all that right now I’ll get back to you shortly with any questions I have.
Thanks a lot
ok tag me so I can see it
Can I gift you nitro?
idk what that is
🤨
In model loader tab it says model url where do I get that url ?
not the tab you are looking for
it's the Upload a Model
there you will drop the index and model.pth files as shown
Got it
wait for them to upload and then press upload model
Do anybody know how to remove copyright if i made an aicover on ytb(i m begginer,it s my first video and i paste the original song s link but it still appears copyiright)
you can appeal on YouTube if you actually own rights
How
what ai song did you make
A romanian song with THE WEEKND
My online potato friend loves to mix all the mountain dews at taco bell and call it a Dewicide.
Still have to clean out music
Then go to inference, upload the audio you want to convert and click convert
If you to adjust the strength of the accent go to the advanced options
@elder willow
bro.. did Neymar really die?
what
I recommend MVSEP or Xminus to clear out instrumentals in music
the illumatis finished the menino ney
Hello
the illuminatis do not exist. like fr. trust me.
when i opened a dorito bag i saw a lot of them , they are real
nuh uh
every triangle on earth is an illuminati , beware of the truth.....

illuminati confirmed
they are everywhere.....
they have... evolved....
the earth is.... a velociraptor.....
god dammit
the illuminatis is hiding the truth. the moon landing is fake....
skibidi toilet is in fact.... real
Guys how do you make a cat ai cover
your toilet has a 69.420% chance to be a skibidi
cat voice model than use rvc or appilio or whatever that thing was called drag in the song without instrumental and gg
Ok
and i just found out...... france doesn't exist.
it was ALL a dream.
made up by the illuminatis.
And how do I do that
Ayo? @wild siren level 1 !!! 
Eh?
You making fun of me fam?
i'm sure you are an illuminati arcane.....
spying on us

