#🧬│ai-chat
1 messages · Page 146 of 1
And that only happens if it fails to download it on first launch for any reason. Delete the pretrain folder & relaunch, it'll redownload
i type
me too
and i talkjand i talk and i type
VB Audio Cable, bcuz VAC is only on Windows, but VAC should be better yeah
d
eez nuts
Yo guys do i upload the pth file to drive to use it
Dude this is crazy
Ayo? @grim phoenix level 2 !!! 
I ran it at 2000 chunks
What are you currently using
sounds like a real person
It does, that's the thing
It's the best voice changer rn
Searching for the perfect RVC e-girl voice model? 
I paid for a $15 voice model and Im like blown away
_ _

Oh niiice
I did this I got rick rolled
Custom models r always great
(It actually shows how to get a good model though iirc)
I was runing it off my 3070 laptop but now Ima run it off my 3090
Ooh the experience will probably be better
My body is ready
192 chunk should be perfect for gaming with a 3090, I believe
Oh yeah, it's...
very demanding
But worth it
Have you paid for any custom
Nah, I make my own models tbh
I noticed it was alot better sounding then #1175430844685484042
And I use it via Colab because I have a 2012 MacBook Pro 💀

Sometimes it's hard to gather dataset and all
a macbook?!?!
Yep
Dual Core i7 from 2012, still runs quite well though
I could upgrade its RAM and SSD if I wanted to
man I love ai
Same
At the same time I think it should be controlled somewhat... It's scary and it could be used badly
So remember to use it with responsibility :)
I still think theres tell signs that the voices are fake
Yeah, not perfect
Depending on the model you can hear a sharp S sound
Im sure that in this discord that there are better voices available or something as you level up by spending time here, so that people looking to just quickly nab something cant access it. Id assume
And the fact that there is no background noise is also a sign
my go xlr has a De-esser
true about the background noise but some people have nice microphones
also I noticed my laugh does NOT translate over to a real laugh
Nope, it's just that most models are kinda meh, mainly because gathering a dataset is a hard task and most model makers don't even clean up their audio with stuff like RX10
Oh yeah, that's one of the limitations too
what is the Model master role with the fire sign next to it?
ai is def potentially dangerous but theres so many good stuff it can do too
basically they are the best model makers
like for disabled people
yeah, definitely
Apple did this actually
I wonder if this could help people with a speech impediment
I saw in accessiblity
Yeah... It's very interesting and could help so many people
was this server deleted before? or was i kicked or what
Once this comes on phones I believe this could help so many people
Yep, this is ai hub 2.0
when did it get deleted and for what
2 months ago
October yes
long story short but DMCA
No
L
Yeah no, you have to grind all over again

Rvc3 when
Hello friends! Does anyone know at least some kind of neural network that could translate video into another language?
🏡
Ayo? @elder willow level 1 !!! 

Hello i try to make a voice from a youtuber. When using his videos as dataset i have first to remove every noise and background music from his voice right?
What does a higher chunk actually do
how do I get it to work on discord?
Basically, means more delay, but it's easier for the GPU
you need a virtual audio cable
I think I have one
Not really... I guess it is a bit better since it takes more time though
Doesn't the oculus rift s come with one?
and am I able to use it
we wont have a friday the 13rth again until 2028 and 2034
Ayo? @glad panther level 2 !!! 
Ehhh idk if it'll work that way
Just download a virtual audio cable, the Virtual Audio Cable by Muzychenko is a good one
because sometimes when I leave discord uninterupted it askes me if I want to switch to virtual audio cable
Ayo? @tacit prairie level 1 !!! 
hmmm
can someone help me about python?
Can anyone tell me where to find the program for making rvc models?
what do you need
-colab
Suggestions for @edgy forge
- 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
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
-local
- Applio-RVC-Fork, 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
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
local
Yeah there it is then
I downloaded Virtual Audio Cable by Muzychenko how do i get it to work?
Hiiii
you said it's where?
Now here are the input/output settings for W-Okada:
Input - Your mic
Output - Line 1
And in Discord or any other application:
Input - Line 1
Output: Your headphones
what
... that's how you get the stuff to work
for me it is a blank message
Ayo? @edgy forge level 1 !!! 
yeah but like where do I find the options to do that
-local
- Applio-RVC-Fork, 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
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
@edgy forge here it is
On the W-Okada GUI, and on the settings of any application that uses your mic
For some reason the message is blank
- Applio-RVC-Fork, 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
NoteExciting update! Introducing the latest, streamlined version of our bat installer, eliminating the need for manual installation. Now, everything is seamlessly handled through Python environment...
A WebUI to create song covers with any RVC v2 trained AI voice from YouTube videos or audio files. - GitHub - SociallyIneptWeeb/AICoverGen: A WebUI to create song covers with any RVC v2 trained AI ...
There we go
yes there it is. now to make my own rvc2 models which one do i download
Applio, Mangio and Original RVC are all great ones for training
Mangio and Original RVC being generally more stable and easier to install, but with less quality of life features than Applio
yap
okay sounds great. for someone new to making ai voice models which one do you reccomend?
Probably Mangio or Applio, they're a bit easier than the original
okay sweet thank you
yap
...what
You're welcome :) Good luck on the model!
thank you
quit yapping
Ayo? @elder willow level 1 !!! 
Hi, my name is Yanis,
nuh uh 
-help
lel
-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
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Ayo? @lyric sonnet level 2 !!! 
whats the best e girl voice model for a deep voice?
Searching for the perfect RVC e-girl voice model? 
this gives rick rolls bruh
like deadass tho can someone give me one
teaches you how to search for models tho...
Where can I find Billie Eilish voice model link to be pasted for google collab
check #1175430844685484042
Ok
Ayo? @soft pecan level 1 !!! 
Why though 😭😭😭
@tardy hare
Hi. I remember there being language specific-channels here. Are they gone?
Nope. Scroll the channel list all the way down
I only see voice chats
hi all
hello
someone can help me with R3gm/SoniTranslate, plz write me back
.
.
why can't I add/drop file on kits ai?
Okay so generally: if I want to change a file path from a .bat set CONDA_ROOT_PREFIX=%UserProfile%\Miniconda3 to the same named folder in directly on my C: drive - how would I write that out? is it simply set CONDA_ROOT_PREFIX=C:\Miniconda3?
hi
bruh my last ai model took 3 days to trian
@waxen grove
Oh boy, I don't do batch scripting
but that could do what you want, yeah
just make sure it actually allows you to modify files there
What do you mean?
you might not be able to create folders directly under C:\ with your user
does anyone have any good voice changers that run well with a GTX 1050
Hellooo I was in this server a couple of months ago but left due to having no time for AI creations, but I'm back and ready to explore!! ❤️
^ Sorry just realised this is gen, the thingy said I should introduce myself lol
Welcome back!
You can try using a Colab for W-Okada, check #🔍│help-w-okada
hello 😄
hello there
where do i find the download links?
nah you don't need to download anything, it's all online
wdym
doesn't run on your machine, it's basically a virtual machine with a really nice GPU so you can run stuff
oh nice
is there a way to use the "applio"rvc fork online? I've been dealing with this for two days and it will not work due to a space in my user directory
YOOO THAT'S CRAZY! please what (would you rather) RPM?
how do i set it up tho?
yo wsg guys
Just run the steps in order, it should be easy
oh kk ty
Ayo? @elder willow level 1 !!! 
you don't need to use the download model step btw, you can do that directly on the GUI
YOOO THAT'S CRAZY! please what (would you rather) RPM?
and btw read what each step says, they explain what u need to do
Sadly not anything free
maradona es mas grande que pele
hey please what (would you rather) RPM?
Ayo? @visual axle level 2 !!! 
3 world cups, Pele revolutionized football
what do you mean by this
on the steps i click the start button thing to do it?
no ai originals channel?
Ayo? @modest ember level 1 !!! 
yeh
alr
i only see covers
what's the rpm of these vids>
no clue
anyone knows what's rpm of (would you rather x or x) pleaase?
so.. I have this "juypter" notebook thing from my school that is also used for AI stuff, is there a way to upload it and use it there?
ooooh I've heard it could work maybe. dunno how you'd make it work tho
hii
hai
which one is the best one in your opinion
most believable
trolling?
ah idk i don't use girl voices
ik someone down bad
Hey, so I can't use applio rvc cuz of a space in my user directory. Won't work, talked to a dev who confirmed.
Is there a way for me to run this on a jupyer notebook? I have access from my school and they also use it for AI stuff.
Can anyone please point me in the right direction? I beg
is that thing running on a remote machine or locally
remote
Honestly unsure, depends on how they have things set up
technically you could create a folder right under C:\ to run RVC locally
it's provided to us for Ai stuff. like literally its main purpose so theres gotta be a way right?
just gotta make sure permissions don't prevent everything from working
I did that, tried, doesn't work cuz miniconda3 is still installed in a user directory. Changed the directory manually in the .bat files but still doesnt work
do you have any youtube video or post you could point me to, to set something similar up on jupyter?
it's pretty much the same as google collab, no?
https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/releases
Alternatively, the original RVC project exists 
Unzip & profit, that's it
but it doesnt have tts as far as I know? or am I falsly informed?
Oh yeah, it doesn't
If you really want to run it on jupyter you gotta write a notebook that does kind of the same thing the colab notebooks do, so :
- install dependencies (if possible)
- get the RVC code
- install python dependencies (again, if possible)
- modify a line or two so that it exposes the instance over the internet
- run
глеб голубин есть?
2023-12-31 00:53:08] connet sid : E89tXhFtsjTiGHRYAAAB
[2023-12-31 00:53:08] connet sid : F4NzwwKPKpYd_6PjAAAD
its stuck on this
you should try deleting pretrain and model_dir folders and restarting W-Okada
i dont know that 😭
you don't know how to delete folders?
I mean i dont know what ur talking about
talkin bout dez nuts in yo mouf
_ _
damn
basically, this means the initial download probably went wrong
oh
Ayo? @rapid reef level 1 !!! 
so you should delete the pretrain and model_dir folders
and restart, so it installs again
okay wait
Booting PHASE :main
PYTHON:3.10.11 (tags/v3.10.11:7d4cc5a, Apr 5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)]
Activating the Voice Changer.
yeah it has to install
Ayo? @sturdy kraken level 1 !!! 
Don't. There are free and unlimited alternatives already
We don't support kits.ai here
Are you looking for training or doing inference?
Doing inference and training both
What's your GPU?
The best way to run RVC is locally
But you'll need decent specs and GPU
I use RVC
what are the best models
Ayo? @atomic quarry level 1 !!! 
for like a fake girl
The ones that are best trained
-egirl
Suggestion for @atomic quarry
Searching for the perfect RVC e-girl voice model? 
when i change another voice, but res ping so high, how can fix it
There are way more in #1175430844685484042 and the website weights.gg
i suggest you to visit weights.gg
its a good site
deep voices
-egirl
Suggestion for @atomic quarry
Searching for the perfect RVC e-girl voice model? 
Use the second one
I'm not looking for AI IMAGINE
Ayo? @willow grotto level 2 !!! 
weights.gg is for storing voice models
No??????
Weights.gg is for voice models
ai hub general chat needs patience
what are you guys doing?
ai hub
i am the least schizophrenic ai hub user
proof?
YAY
@idle light bitch him
WEIGHT.GG BASED ON RVC MODELS, NGL MOST OF RVC MODELS IS ASS PEOPLE DON'T TAKE MAKING MODELS SERIOUSLY, I'M NOT COMPLAINING, I'M LOOKING FOR A BETTER SITE
^^^^^ schizophrenic ai hub user
weights.gg is based on ai hub models
and thats most the models on the internet
so good luck ig 😭
does anyone know where to get pytorch visual models?
from pytorch?
train one yourself
kinda like what weights.gg does
thats what i mean
pytorch does have models
weights.gg has a bunch of community made models
pytorch website doesnt

Smartest person on ai hub:
visual models ?
truly
As in models to recognize stuff on images?
models used for inferencing on images?
bro is visually using rvc 
ye
https://universe.roboflow.com/
There's this
Has lots of models/datasets for various stuff
thx
oh now i understand what he meant ig
Spectrogram AI Cover 
(even models for games like csgo lmfao)
Respect at least
i am very interested...
just for funsies
You ask for voice models, we tell you where are the voice models, you say you don’t want RVC models
Wait till you realize that you can’t use RVC without RVC models
conf is prob too high
So you can’t make ai cover
nah even with low conf threshold it didn't do great
kept jittering
Plus with all the agents & shit, I doubt it can be reliable
iou?
hadn't tweaked that, but isn't it only relevant when there's overlapping detections ?
i think it kinda helps
Ayo? @marsh warren level 5 !!! 
I was talking about a kits.ai alternatives, where people can post their models, and people can use it in a easy way, do you know something like that?, i used most of rvc models, they take a lot of time to post models that I'm interested in.
talkin bout dez nuts in yo mouf
Tf u talking about
talkin bout dez nuts in yo mouf

Yo
😔
I'm sensitive
I hate people hating me
What tf i did to yall i actually have a hard struggle to contact with people
Im i overreacting
Yeah
I suggest logging off of discord for now
And give yourself some time
oh my
I have a question
What is it?
is colab safe like the voice changer
Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb
Yeah why wouldn't it?
idk how to use it 😭
Ah yeah that one is trustworthy
hello
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
Hm wait let me find you the guide
🎙️ Realtime Voice Changer (Online)
A free, online version of W-Okada that allows you to run code on Google's powerful GPUs.
i can't find my model in search models for some reason
although i made it like 3 months ago
wait i can ban people?
yes
Did you search on #1175430844685484042 or weights.gg?
yo guys

what other servers have these
taking over the server
😭
these?
:)))))))))))))
DEEZ NUTS
these are so cool
tysm
😭
yw! Have fun
idk if i can say this
but i got a really convincing girl filter
and got CoD for free
Okay

😭
gg wp
hi
Hai
@pliant elm you have competition >:3
Your doppelganger is here
🤭
Yui vs Yui-Chan
xd
ajdvakbdbsd
you can know rus ?
😭
xD
yui is taking over ai hub
Did you search on #1175430844685484042 or weights.gg already?
Good luck
I've already searched there, there doesn't seem to be anything\
Ayo? @elder willow level 1 !!! 
Check weights.gg
Guys whats the best female model
Searching for the perfect RVC e-girl voice model? 
she looks like she should purposly die in a valorant game
That problem is usually because of a bad dataset
If there is noise, reverb, silences, or overlapping voices, remove them
And then train again
mic sensitivity basically
if u increase less stuff will be ai
@elder willow rep stp
o alright thanks
tf is that
I made a model , how can I go about posting it?
it says i dont have permission
you first need to apply
-upload
English guide
Link: https://rentry.org/fdg_guide_newer
Credits: FDG
Spanish Guide
Link: https://rentry.org/AI_Hub_guide_esp
Credits: Julia (ailen2091)
Hindi guide
Link: https://rentry.org/AI_hub_guide_hi
Credits: Enes
Turkish guide
Link: https://rentry.org/AI_hub_guide_tr
Credits: Enes
Where do I go to apply?
Ayo? @wooden ridge level 1 !!! 
bro didnt read the guide
hi
Hi!
Ayo? @acoustic hatch level 1 !!! 
hai
Hey , I need ai cover Love me like you do by voice Dua Lipa 🙂 someone may have it ??
thak you
how the fuck

ya sure, what for ?
I guess there's #1159289738314919936
Would be best if you didn't, discord is already being pressured with copyright, and the server had issues with it in the past
But do mention you have the audio files available
hey guys can i go a little off topic
it straight up used more than 70gb of page file
what are you doing for that to happen 
what did you set the batch size at
640
Batch size ?
wait idk

It's unless that, or it's on the default automatic setting, and if you happen to be on CPU, it might just eat up everything it can
i just used mode.train(data = r'data.yaml', workers = 1, epochs = 1)
It might be training on CPU then, I guess
How the fuck is it using so much memory
might want to spin up a colab notebook to train on tbh
Is there an inference colab other than ai over gen?
easyGUI can infer too
without WebUI
-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
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
it was originally at 16gb page size
Tysm
dont want the server to end up having 18,819 files uploaded to it
I’m on Mac
You spent all money on Mac
unplug your pc and throw i t in a tank of water
So you have no money for ai cover
😡
the arm based apple devices are straight up cheating
ain't a bad thing
but holy shit they went nuts with the memory
is that to make AI voice?
yes
I’ll give you robux if you can show me how to use AI cover maker
plss
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
I already have enough

Buy kits.ai
Ask @wintry sinew for the paperwork
goes to a bakery
"please teach me how to make bread"
okk
can someone help me to get this work on discord i live stream
The input output settings should be like this in W-Okada:
Input - Your mic
Output - CABLE Input
And in Discord:
Input - CABLE Output
Output: Your headphones
Install the virtual audio cable and set it as your output on Okada
And your input on your streaming program
And you need the virtual audio cable yeah
take a model or make a model and use RVC/applio locally or use colabs/websites bro
i have all of those
sorry I was just joking around 
friends, where i can get voice model "shadowkekw"? https://www.twitch.tv/shadowkekw
hi @urban crane how to make Michael Jackson ai cover model
Ayo? @forest quarry level 64 !!! 
search on #1175430844685484042 or weights.gg
download SVC, then find Micheal Jackson #1175430844685484042
-svc
Yeah I still use so-vits-SVC
ok givme download @urban crane
me wondering if I actually go look for the SVC link or send a meme one
Guys does anyone used the deepfake face fusion
why arent we using rvc v3 yet its been out nearly the whole year
has any1 use this on valorant?
do yall not remember when rvc v3 came out and everyone was hyping it up saying they just had to add it to the gui or something to get it to work
and u can already train models on it and use the voice changer
?
there's exactly 4 kind of RVC models at the moment
v1 with and without pitch
v2 with and without pitch
it does bruh the github link was put in the announcemenets channel on the old server and everything and kalomaze and all were hyping it up
am i tweaking rn am i the only one who remembers this
i think you tripping lol
i found the link
Installing RVC v3 (nadare fork) shitty guide
Clone or download the github repository: https://github.com/nadare881/rvc-webui/
Now, unzip it and go to the train.py file in rvc-webui/lib/rvc_v3 and edit it (you can do this with notepad++) see this and go to the line 419 in your train.py file: https://github.com/ddPn08/rvc-webui/pull/54/files if you know how github works, simply replace the old commit code with the new one.
Now download the pretraining models here: https://huggingface.co/datasets/nadare/rvc-v3j/tree/main You must place them in any folder, it doesnt matter much but I leave them in rvc-webui/models/pretrained/v3 (you need to download the files f0D48kv3.pth, f0G48kv3.pth, v3_speaker_info.npy and v3_speaker_info.npy)
Run the webui clicking the webui-user.bat or running the webui.sh file
Follow the settings in this screenshot and click on train: https://huggingface.co/datasets/nadare/rvc-v3j/blob/main/v3_setting.png
this guide really sucks because this is very experimental, if you cant wait for a final release, you can TRY to follow this guide
for now you can NOT make inferences as such, only use the nadare fork for voice-changer.
thats unofficial
i was not tripping
well it was posted in announcements and hyped up so much
for like 2 weeks then everyone forgot about it
"While we wait for RVC v3 to get a proper, usable release, as it is brand new and using state of the art audio reconstruction and isn't yet available as a finished build; RVC v2 added 48khz as a training mode. It seems like the 48khz trains might actually be somewhat better. Potentially even 20-25% gains in the right circumstance. Now is the time to test this if you have datasets you can compare that you've trained before.
RVC v3 is different from v2 in that it's a huge refactor of RVC in various ways. The main thing being, more accurate audio reconstruction. " - kalomaze https://cdn.discordapp.com/attachments/743824122223788052/1120175222474158080/image.png?ex=659b5af8&is=6588e5f8&hm=8cf15caa772259d0ff6ad41b2a520832b77e9ff8bbd8dad533be77661c5c0a62&
how to use
have a look at one of the free models - Monotone Mommy - Voice Model (RVC V2)
monotone mommy 😭
Find a model, 3 dots, download
tho yeah wtf happened to this
Drag the .pth in the weights folder and .index in logs
These folders should be in assets on most modern forks of RVC
try asking @velvet hill about this
apparently the original RVC project isn't interested in adding new features or tweaks to the original algorithms
So this might have stayed an unofficial spin, whose creator lost interest
they were saying they barely had anything left to do on it, all they had to do was fix inference on the webui
Kalo kalomazed away
He is know into LLMs
a little birdie told me someone is working on an update to rvc..
If that’s it do it then
i cba
im fr
should i train an rvc v3 model and upload it in the voice models forum so everyone can be like huh how u got rvc v3
cause nobody seems to know about this
Yeah
First v3 model 
its not gunna change the pretrains or anything just make it way faster
if your talking about what i think you are
talkin bout dez nuts in yo mouf
someone in the old server trained kanye on it im pretty sure and they were using the voice changer with it on vc
idk, but i think theres some new features
Old server
oooooooooooooo
And I don’t think it was uploaded to voice models
😈
Just try it
why has v3 been memoryholed
I want to see the QCs reactions
im downloading the files now
don't think voice-changer 's got support for it though
yes lmfao
Wth is this 
Is this official and legit?
not official
probably that prototype thing
How can I train a RVC v3 model?
just use
Wait
rvc2
You need to provide a sample
im pretty damn sure its official cause nobody said it was unofficial but i'd say since the moderators here say it isnt it mightnt be
in the old server it was touted as official by everyone
Interesting, me and some of my colleagues developing another big pretrain, for this time it'll be multi lingo
Rip


Oh??
It is only one way
in the old server it was posted on announcements and hyped up by everyone for like 2 weeks then everyone forgot about it like nothing happened
Ye, my previous "Russian" pretrain works good
So now we are trying to do stuff more interesting, with multiple languages
Also the dataset is a big think, and I think training will be hella long
Is it works?
what you using to train it 
yea it works, u can train models and use them in the voice changer
Ayo? @elder willow level 4 !!! 
but u cant infer on the webui
yes its supposedly wayyy better than v2
it only uses dio and harvest 
HI
hi
harvest is decent tho
ai hub devs were working on adding it to the webui and adding proper inference but they never finished it
rmvpe ftw
I've met an edge case where harvest inferred better than RVMPE

nuh uh
Yuh huh
Previous pretrain was trained on mangio (RMVPE) (40 and 48) versiona.
Little 21 hrs dataset
probably as good as Dio
Rmvpe work better with noisy inputs

it's somewhat official
on what gpu 
wha

nadare is one of the main RVC contributors IIRC
rvmpe is best for most things
but. BUT. there are cases where it screws up
then use crepe
crepe also screws up on this audio btw
4080, 4070 ti and 4070
(Rip 4080 melted connector)
only harvest did it decently

bozo audio
write filelist done
use gpus:
runtime\python.exe train_nsf_sim_cache_sid_load_pretrain.py -e Plaguedocplays -sr 40k -f0 1 -bs 1 -te 50 -se 25 -pg pretrained_v2/f0G40k.pth -pd pretrained_v2/f0D40k.pth -l 1 -c 1 -sw 1 -v v2 -li 32
can anyone tell me how do i fix it
i turn on train then this just shows up and there is no pth file in weights
nadare is one of the devs of rvc
they'll need to also download nadare's fork
v3 uses VORAs tech, not supported NOR WILL BE in the mainline
not on rvc
Hello btw
Check out LIke a DRAGON : ISHIN https://www.twitch.tv/rukachan54

Harvest can be good
whos VORA

It can be it’s just slow since it’s mainly runs of your CPU
interesting
this is such a deep rabbit hole
i cant believe im the only one who remembered rvc 3 hype
s<s
What did i make 💀 https://www.youtube.com/watch?v=RXYxVw7wADU Marge simpson sings godzilla
Can someone explain to me why my voice doesn't work on the app?
Which app?
voice changer
“Real Time Voice Changer Client Demo”?
Do you have a NVidia or AMD gpu?
yes
Ayo? @leaden shore level 1 !!! 
I do not know
Check task manager
What’s not working
Like does the ui load and stuff
hello
cpu
so, i am a trans woman i was wondering if there are any good voice models someone could recommend me for a more feminine voice?
My voice tends to be a bit monotone or raspy due to having throat cancer in the past.
GPU
Any should work if you configure the settings right
Ayo? @elder willow level 1 !!! 
when I click on start I don't get any voice feedback
i plan on using it for VR gaming but ty
Ayo? @empty mirage level 1 !!! 
task manager
does cmd say anything?
yeah, it has a tab where u can see all ur components n stuff
yes
so is the more trained a model is the better?
generally, yes
yes
good question, i havent used windows in age so idk
anything involving "Pipeline is not loaded"?
ages*
that's probably not ur gpu then
does cmd say "wait web server"? or something
oooh
Sometimes but not always
it's a possibility
onnxgpu-cuda is the NVIDIA one
onnxdirectML-cuda is the AMD one
just answer this
and maybe I can figure this out
okay what does it say then
the CMD tab
Scroll down it tells you how to solve your issue
^
No it’s a bugs and fixes thing
You scroll down untill you find your issue
And it’ll tel you how to fix it
where do i get amazon discount code from?
command line, terminal, black screen with text, whatever you wanna call it
How do I run the program after extraction?
there should be a .bat file
Run the .bat file
does it say anything like, wait web server?
Ayo? @elder willow level 2 !!! 
There isn't
What files are there
_ _
Wait
yea just wait
takes like 3-4GBs if I remember right
pyache
Infer
Infer_pack
Models
Output
Runtime
There's one application called ffmpeg and one windows batch called RVC-GUI All the other files are PY files
Run RVC-GUI
That is the batch file
Thanks!
It's so simple i don't know why I struggled lol
I did
Ayo? @ruby lichen level 1 !!! 
What?
Do I have to open the windows batch everytime I want to use the app? Is there a shortcut
Create a shortcut of that bat file.
Yeah, you can make a shortcut by right clicking and making a shortcut
monitor option or set the output to your speakers
i think u should read the guide tbh
<

I assume I just drag the model I downloaded in the "models" folder?
Then you wouldn't know how to do the thing u just asked about ¯_(ツ)_/¯
The guide holds your hand like a baby.
Ayo? @elder willow level 1 !!! 
I'm just saying... If you took the time to read it
You wouldn't have so many issues
I'm not gonna spoonfeed you then
I'll just ask you to read this
https://rentry.co/VoiceChangerGuide
CTRL + F to find whatever you want I guess
This guide no longer has an author, Raven (me), has decided to stop working on this for various reasons, I am no longer in AI Hub, I have given the Edit Code to a few people, I also just want to step away from everything, and live my own life. Do not seek me out, when I want to reappear I will. B...
your out of luck
nobody can read
ai hub moment
I just submitted my first model of Lex Lightning, so what exactly is the process for this model to be reviewed and checked off as high quality?
,
Our QCs will test that model and if it's good they'll approve you?
how to use these models? is there any tool recommed
But it seems you already have the model maker role so
You can use them locally or online. I'd recommend using a Huggingface Space or a Colab (both online solutions).
-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
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
It says "not responding" when I convert. Is that normal?
You can check #1159513888199540817 for Ilaria RVC, strongly recommend using that one fist
first*
Like, the Windows "not responding" thing?
Yeah
Oh I think it's done
I think all's good
Thank you
You're welcome :)
Hi
Does this normally take long
Ayo? @ruby lichen level 2 !!! 
Accidentally deleted the message but I was asking about the converting talking long
Taking
Depending on what you're using yeah
It's been aprox. 15 mins
oh damn what are you using?
Laptop
there's various rvc programs, those r called forks
I downloaded the latest version of RVC GUI and I'm converting a 5 mins video using the "harvest" method
oh.
All other things are default
rvc gui is kinda outdated, id recommend using mangio or something
has faster pitch extractors
i think the 15 min wait is probably cuz of harvest
Is there an C.A.S.S.I.E ai ?
It says the Dio and pm are the lowest quality
I'm using this for a shitpost so I don't need it to be the highest quality I'm just following a guide and they did this
I just submitted my first model of Lex Lightning, so what exactly is the process for this model to be reviewed and checked off as high quality?
You already have model maker
You can just upload in #1175430844685484042
I thought those that get uploaded in voice models have to then be assessed as high quality
So we just upload 14 other ones onto that section w/out needing that assessment?
Yeah
I would love to see Bowsette have a voice. There's not one on here.
With what data???
I’m saying there is lack of data to make a model of it
She lacks a voice actor
oh i cant do that myself. i suck at this stuff
What I mean is that voices have to be off already existing audio
I watched a youtube video and there are like a few out there. I had to look but people have done them on YT. Idk where to find them at
Check weights.gg
Not one on that site
Ayo? @wary oak level 1 !!! 
You should find a fan dub and look up their socials. 50/50 they'll find out their voices have been uploaded to weights
For anyone who has made a voice model, how long would you estimate someone new to learn how to make one?
Yes, but the model there was not good and could only speak English.
It was not in the chat channel #1175430844685484042
Ayo? @tired scaffold level 1 !!! 
No results found for the search ataturk...
**Method 1 🔎: ** The search accuracy is 100% 💕
Upload the <t:1691333123>
Server uploaded: AI HUB (RIP)
Channel uploaded: AI HUB (RIP) Post
Profile Owner: upfox
iceFox
upfox
3485
EN 🇺🇸
N/A
68 MB
N/A
1000
0
sent you a dm rq
How wdid my voice changer get bad overtime
THE FIRST TIME I USED IT, it was so good
now it became bad
might sound stupid but is there a ai that can filter out motor noise
elo
-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
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Sorry
Have you tried UVR denoise?
or maybe just any vocal seperation model
rx10 spectral denoise prob could
Hi, whats the best female vc changer here?
Searching for the perfect RVC e-girl voice model? 
guess il try that thank u
😭
Ayo? @delicate shuttle level 1 !!! 
i tried it aswell
the e girl voice that the bot linked
Searching for the perfect RVC e-girl voice model? 
shut up
dosent seem to get alot of it though
can i hab audio to test somthin
or nah
here uh
Hmm.
I tried, but it seems it's only using my CPU no matter what I select
How to make an ai cover on cell phone ?
Ayo? @shy flax level 1 !!! 
Ayo? @elder willow level 1 !!! 
does voxmaker have chris griffin
What is voxmaker
ai voice gen
-upload
English guide
Link: https://rentry.org/fdg_guide_newer
Credits: FDG
Spanish Guide
Link: https://rentry.org/AI_Hub_guide_esp
Credits: Julia (ailen2091)
Hindi guide
Link: https://rentry.org/AI_hub_guide_hi
Credits: Enes
Turkish guide
Link: https://rentry.org/AI_hub_guide_tr
Credits: Enes
do you know how i can make a chris griffin ai voice since its just to hard to get my hands on normally rn
Ayo? @elder willow level 1 !!! 
Just use RVC
do you know how i can make one since its rare
Did you check weights.gg
its not even on fakeyou
nor other websites so the ones who have it have it for personal use
is there a bot that does it for u
????
anyway thank u all big juicy synced to 10.9m streams on spotify
did you try uberduck
@thezoenicole ur literally my bestie tysm for making a ayesha erotica one
the ayesha song?
i found it on weights yeah
yes lmfao
i own the 351k profile
it synced with the ai version somehow
YOU BROUGHT THE IMPERSONATOR PROFILE?
probably bc i used isrc and upc LOL
no i made it girl..
eye-
ew….
I'm a minor
ew
Ayo? @elder willow level 1 !!! 
stfu



