#🧬│ai-chat
1 messages · Page 154 of 1
Suggestion for @elder willow
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)
i mean ingame realtime mic
what?
can someone make me an ai model, i have the audio and everything
can you run the voice changer in realtime using it as mic
microphone?
yes i want to use the voice changer in game
yeah you can
how?
-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.
Best S.thresh recommendation?
Is Google Colab Pro worth to buy?
alright what happend to the english dub models for konata and yui? they seem to be lost to time
Can you do an AI cover please?
How does epochs work, and how do you know if its a good model
Im busy with one rn
if you're just looking to download a model, listen to the sample and decide if its good or not - or just try it :>
But what about the epochs, does that change anything
Bc some have 200 and some have 900
And idk what those numbers mean
more = better, generally
ai model of a certain type of piano, converting a piano song using that model = the piano song sounding like the certain type of piano model?
would that even work?
and i mean just piano only, nothing other
i saw drumset ai models and im guessing it might work the same for piano? i dont even know
Hi xd, only sending this message for Starting purpose xd
F
Ye it can work if u will use index
-tts
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
why does it say potentially dangerous file
It says they with all files dw
ok
.
Can an official look at me?
Someone knows what is the best Olivia Rodrigo AI in Kits?
Can someone help to me with the realtime ai voice changer? Everything is working fine but I would need to use Virtual Audio Cable to use it on discord, but I am already using it because of another app called voicemeter. Is there any other way to make it work on dc?
how does one even train an ai model i have a couple 4080 16gb's just sitting there
.
anyone know a good kai cenat voice model?
What you can only use it in DC with virtual cable
why custom voice doesnt work
i am looking for nice girl change plzz
what site can i use to make an ai cover? the colab i use is down
but it seems not to work
because I use it for voicemeeter aswell :))
you probably did something wrong
hi have a question how can i change so the program uses my gpu
it only uses the cpu idk why
and how to change it
Can the voice changer support mobile
hello
Eewww voice ai
https://discord.com/channels/1159260121998827560/1175430844685484042
:right_arrow: /search https://discord.com/channels/1159260121998827560/1163592055830880266
or take a look at weights.gg
https://discord.com/channels/1159260121998827560/1191429836321849435
cant find the model you want?
you can commission one from a Model Master 👑 https://discord.com/channels/1159260121998827560/1159289738314919936
1 question, im kind of confused about the number epochs in each voice model threats
can someone make a voice model of this guy
Training cycles; more or less does not mean better
I cant find one
wait wait, if it doesnt mean better what it wont affect anything?
Make it yourself
Just ignore it
Really just depends on the dataset the person used to train it
ah ic ic, tyty
I dontk now how to
I just want someone to make it for me
nah bro you are just lazy
-rvc #1159513888199540817
Suggestion for @tiny anchor
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)
yo guys my friend sent me here to get the voice changer but where tf is it...
im good
So that definitely won’t come out any good
still
so you want the voice model to speak a 2 second audio or you are gonna train it using a 2 second audio?
Ayo? @boreal canyon level 1 !!! 
I want someone to train it using a 2 second audio
just 2 second?
yes
Ayo? @tiny anchor level 3 !!! 
isnt that too short
no
thats what she said
if the roblox pizza audio can be turned into one
Yes
I think a 2 second audio can be turned into one
Yeah and does it sound any decent?
Hello
hello
Model Composer
Terminally online
That just means I’m online a lot
It means you're braindead
hi
lol

Yknow what the model composer role means right?

thats fucking wild coming from u i love that
🤔
yUo kNoW wHaT mOdEl CoMpOsEr rOlE mEAnS rIgHt?!?!? 🤓
🤓 🤓 🤓 🤓 🤓 🤓 🤓 🤓 🤓
do you?
Yeah, it means that im better at making models than you
kekw


My model is super popular on weights.gg right now
Ayo? @limpid pecan level 5 !!! 

hello
hello
how to run real time voice changer on amd 6655xt gpu can any one help me
Get Okada voice changer
hi guys does anyone know how I can have Tommy Shelby's voice in French model?
https://rentry.co/VoiceChangerGuide
check this
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...
Check weights.gg
tells you everything you need to know
Ty
Ty
Ty
You're welcome
hi guys does anyone know how I can have Tommy Shelby's voice in French model?
search it on weights.gg and here
Check weights.gg
if you can't find it, it probably doesn't exist
how to create it
People will search with applio
And then not choose the “technology” thing and set it to RVC
huh
oof
And they are forced to use kits
😭
u can either make it urself or commission it
https://discord.com/channels/1159260121998827560/1175430844685484042
:right_arrow: /search https://discord.com/channels/1159260121998827560/1163592055830880266
or take a look at weights.gg
https://discord.com/channels/1159260121998827560/1191429836321849435
cant find the model you want?
you can commission one from a Model Master 👑 https://discord.com/channels/1159260121998827560/1159289738314919936
-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)
How do I download a voice from a website that is issued by a bot in the voice search channel?
Use weights.gg for models instead
Senky
what to do after download the file
Extract everything and run the start_http bat file
ai ty
esto ke
Who will join my server DM me
no habla espanol
Who will join my server DM me
promoting alert
banned
i don't think you can even promote discord servers in #1159290752195633273 LOL
Who will join this channel go here #1159290752195633273
nobody uses it so who cares lmfao
I am new here
Hello new here
Welcome
Well I wanted to try Corpse Husband's voice but I can not able to do it
Thanks guys
Why?
Well I'll share screenshot
?
Oh ok
Or #🔍│help-w-okada if your using the voice changer
yo @ionic rover did you make this? https://on.soundcloud.com/hmHoA
holy shit lmao
Poopy moment 🚔
w pfp

E-GIRL VOICE MODEL SALES!! 
DM ME FOR DISCOUNT PRICES! 
E-Girl, Mommy, V-Tuber, Anime girl and much more!
Wondering what models?
You find them here: https://ko-fi.com/ilariaowo/shop 
INCLUDING FREE ONES!! 
I also do commissions if you want aspecific voice! 
I accept the following payment methods:
- Ko-Fi

- PayPal

- Steam Gift Cards

- Amazon Gift Cards

- Throne

Grab yours today!!! 

discord is not geting a mic
I have a question: what is the difference between RVC model and voice changer models on random web pages?
I know web pages sounds worst than the rvc but, why?
Why are all the kurt cobain voice models so trash
Did you personally make these with your own voice
also yes
are you a guy
no
then no sale
LMFAOOOOOOO
i only purchase from femboys
Why are all the kurt cobain voice models so badly made

Also rule 5, I may have to warn you
make yours
Idk how
go to guides hahaha
Ok
Ayo? @elder willow level 2 !!! 
Latina Mommy but your are italian zzzz
Ilaria-RVC seems familiar hahaha

wdym?
how do i appeal ban
you will get a notification from sapphire and it will have a link to the appeals server
what if i didnt get notification
i can send u the link in dms
sure
hey im new is there any way to see the top rated voice models
Like in quality or in popularity?
quality
i mean if its popular its probably quality right
ive just been downloading ones with the most reactions but idk how else to get quality ones
No coughs villager coughs
ciaoo
hello
Hello
hiyaa
hai
make your own lmao 💀
what
make your own voice model rather than buying it
people want to buy it because they dont know how to make an HQ one
:>
i just let the thing train overnight
max the epochs
Overtraining moment
well i have to use my 4080 somehow
yea but again, you have to have a good dataset, you cant just put a random number of epoch and call it a day 💀
Overtraining moment
:>
Ayo? @unique gyro level 2 !!! 
itsn not like i play gpu intensive games so
16gb vram on the 4080 tho a second hand 3090 is better for training if you wanna do that
Overtraining moment

This server needs a "creations" area where peoople can share the ai covers they made
im not listening to a robotic 'call me maybe' cover
No
Im killing myself
Training more does not make it less robotic 
I KNOWWWWW
Server went bye bye
😭
Ok then its prob for the best that it doesnt exist
hi
Does anyone know where I can find loli voice models?
What?
define loli
lolipop
Who do you think you are, dream?
shut up timbo
Like soft spoken chibi characters I play VrChat
no we dont have those here
I'm not an adult its not that weird.
timbo instead of using
either shut up or contribute smth
Ayo? @median viper level 1 !!! 
even then
we dont have any
The emoji makes it seem like they think im a pedo
Can an official look at me?
Yeah
Hello, I bought VIP, but it does not sound properly, it gives a very bad sound.
,
is there a way to use Easy GUI on windows, google collab keeps kicking me out every 2 damn second
Google collab is honestly trash
they banned webui
- 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-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

Lol
@hollow tangle ban time
wha the fuck
Account joined discord a few days ago with a absurd username
Why say that in a server for ai hmmmmmm?
i think there is some basement dweller that has the same problem
and he prob managed to fix it
so im looking how to stop the smell
where else do i say this
no where
Ayo? @rain stream level 1 !!! 
is it the boy smell?
u must know this
!ban @rain stream Trolling and talking about nsfw topics in front of kids
yes, my apartment smells like boy too
@scatmanjeffrobuck banned
Reason: Trolling and talking about nsfw topics in front of kids
Duration: Permanent
he was not talking about that
i just joined chat, sure looked like it from a glance
i nust be correct
everyone else is wrong
OH MY FUCKING GAWD I WANTED TO MAKE JOKES BUT WTF I JUST GOT A DM ABOUT WHAT HE WAS TALKING ABOUT
sooooooo howd yall get into AI?
Ayo? @woven gulch level 1 !!! 

depends whos askin
not.....a......United states government official? 😁
i was always interested in using these ai things, but for voice specific stuff it was around april when i say a foxjoy cover of rick astley singing plastic love, using diff-svc
(Please buy that)
a lot of ai chat bots started messaging me on my OnlyFans and so i started looking into how it works
lol I heard a mosquito singing "Cupid"
thats a good song
😭
gf
omg the nonweb GUi also got shutdown for me
yknow the thing is that diff-svc impressed me a lot back then, normally when they meant "cover" of a song by another artist they just copied and pasted some parts of an artist's song and called it a day... but when i heard the vocals i was like "what???"
This is so fricken bonkers 
oof
this world hates me 
generic anime girl as larpimg, claiming not to be from the government, sex joke in bio, well what can i say, i can trust this person
Loll
im just gonna speedrun WebGUI before they turn off my runtime 
and again, why exactly cant i just download easy GUI in windows?
so i can share it with my fellow totally not CIA members 🙂
easygui is pretty much a thing of the past
what r yall using now
omg i followed an outdated guide
I just now realized ive been following an oudating guide
Thank you so much
So do i just download it and run it on my PC and thats it?
Applio has had and is having(?) issues related to installation, so unless you absolutely need things it offers (I don't even know the extra features it has lmao), prolly stick to the original
refer to this for the full dl links
but yeah, download the thing, extract it using 7zip, and then it'll just work™️
(anything other than 7zip doesn't decompress it right for some reason)
Thanks for helping i appreciate it lol
Ayo? @woven gulch level 2 !!! 
Cant believe i followed an outdated guide
time to compile 15 mins of bandit heeler's voice and make a Sophisticated AI model
just make sure to have a recent-ish nvidia gpu if you want to make models
Rtx 3070 good
yeah great
bet, i figured it would most likely require Tensor cores
Can i train a already existing voice model to develope on that?
(and credit it ofc)
@mellow crag why you lefttt
you should be able to, but you'd need more files than what people here submit if I remember right
so unless you have all the files related to the model, sadly no
darn, well ok
i ma join in a bit
sure
im looking at a directory called "RVC10006Nvidia"
hi
does anyone know where to get the stampy longnose ai voice model cuz i cant find it anywhere
double click go web
.
thats literally what i did tho
it did nothing
it opened up a CMD
and thats it
tbf those are nvidias ai training cores
you gotta wait awhile
Tensor cores work for other workloads, RIFE (an interpolation AI) uses tensor cores
yeah
aswell as some pytorch .pth upscaling AI models
so thats why i thought because its trainging an AI model, that they would also use Tensor Cores
training*
how do i install and use the voice models?
its pretty vram intensive too
i changed gpu0 to gpu1 and now i program doesnt work and when i click gpu0 it doesnt change
i should get a higher VRAM invidia card, 3070 with 8gb vram not enough smh
eh 8 is alr
a second hand 3090 is pretty ideal
what's displayed on it ?
or better yet get a 4090 lol
8gb G6 its alr
floopie
yeah
how do i download and use the model?
wokada voice changer
coz im tryna do it and i cant find out how to do it
i got the voice changer
and it works im just tryan add the ones on here
Ayo? @halcyon flower level 1 !!! 
its working now, how do i put a voice model lol, RVC webGUI is on but where is input for voicemodel
is there any yt vids u can send or sum
for .pt models it's under assets/weights/ if I remember correctly
and for .index it's under logs/
for accent and stuff
if youre using a voice changer its not reccomended
hi
!ban 656491204938039326 Nitro boost spam
@rinor_z banned
Reason: Nitro boost spam
Duration: Permanent
LOL
Ayo? @woven gulch level 3 !!! 
whats a good song to test an ai
Ayo? @rustic swan level 2 !!! 
hi is there a way to put voices into chat gpt ?
@ionic pumice
i have streamerbot
I've been making daily videos on my channel with the villager voice AI, but for some reason it just doesn't sound as good now.
how do i put the pile into streamer bot to play the custom voice ?
Ayo? @rustic parcel level 1 !!! 
Ballin lol
file
bet
does anyone have a video on how to
It always makes me laugh when some random character is singing it for the examples
Ayo? @proud oasis level 1 !!! 
I'll be honest, idk what kind of voice I'm looking for.
why when i laugh it doesnt laugh like the voice i want
Ayo? @quasi vapor level 2 !!! 
Cause its not trained on laughing.
are there any trained on laughing?
which fearless one sounds the best
cuz like i download them and they sound scuffed asf
I actually don't know, mabye?
someone got saiko voice?
Not always
An epoch is basically how many times your dataset has been trained fully.
E.g. 200 epochs
Means the dataset has been trained fully 200 times.
And technically, more epochs mean the dataset is closer to the dataset. Too much, however, and you'll get an overtrained model.
Basically, every x epochs (the number you select), the model saves
whats an overtrained model
someone have plk voice here ?
Training a model is like baking a cake - it'd be like burning the cake
Because you cooked for too long
when the ai try to mimick the dataset too much and it breaks the knowledge
🌝
Ayo? @unique gyro level 3 !!! 
soooo. more epochs could or could not potentially make my model sound better
LOL
The scout model by radical_larry in this server is an exception because it's a 48k model. It's good stuff especially his sample

Too much and it might screw up
How do I transform text into character voices?
but too little and it wont sound as natural, right?
Welp, you don't
You just check the TensorBoard graphs
how does one find out what the ideal amount of epochs to train is
yeah too undertrained
Guys does the voice changer work for mobile
no
That'll tell you if the model is undertrained or overtrained
also the more mins of audio i have the better itll sound? especially with more syllables and clear pronounciation right?
And you train it for long
More mins of audios could be better but keep in mind they must be similar in quality
Is there any way to generate voice through texts with character models?
it needs to be constant talking right? So i can grab a bunch of clips of a character and just mash em together end to end and itll be fine?
homie i used to have da same exact pfp
yuhhhhhhhhh
No, but you can generate speech samples in other places using text and then converting with RVC
yeah, no silences or anything
lmao cuz 1000 epochs takes roughly an hour for me to train so why not
GPU/CPU?
gpu
what GPU/CPU tho
4080/13600KF
HOLY COW
4080 oc'ed to 3000mhz 13600KF@6ghz
MY i7 9700f and RTX 3070 gonna cry then
gaming system used to train AI 💀
hello
Ayo? @split pasture level 1 !!! 
Hello
Im tryna get either a 4070 ti or a 4080 super and i WAS going to go 13600kf but thats a dead platform so AM5 it is
on a non overtrained model doubt that
how can i do this
Was gonna get a Z790 AORUS ELITE AX
decided not to cuz of dead platform
i might wait for 15th gen tho
im waiting for nvidia 5000 and amd medusa to drop or intel lunar lake
Intel catching up, AMD sucks for compute work and NVIDIA 5000 might not be worth it
ur gonna have to plug that 5080 directly to your wall socket
amd's getting there
i have a commitment to buy the 6090 tho
too bad they have their head up their ass when it comes to consumer hardware support for the compute stack
#1159289738314919936 Make a nice dataset so someone could train or it, or cough up the money. It's a niche request
its not just that amd sucks for compute, its that NVIDIA's tensor cores are mostly used for compute and their good as heck at it, so for stuff such as container jobs, RIFE interpolation (or FLAVR and DAIN), Upscaling, anything else sucks because amd only supports NCNN and not pytorch
dont like how the 4070ti and above uses the stupid 12vhpwr which will prolly be used for like 1 generation
Explodes in fire
yeah i held off on buying a 90
NVIDIA focuses too much on AI and AMD too much on gamers
and there wernt any white 4090's on launch day so
ive held of buying AMD just because of the amount of compute work i do
import torch
device = torch.device("cuda:0")
device_name = torch.cuda.get_device_name(device)
print(device_name)
$ python test.py
AMD Radeon Graphics

love me ROCm
though i may just buy a 7900xtx when its cheaper and use my RTX 3070 for compute work and my 7900XTX for gaming
though amd has trouble is some games like fortnite cuz their engine is sh*t
Pretty much 1:1 support when compared to the CUDA backend for pytorch, minus the few bugs that may happen
the xtx realistically has 20gb of vram
still better than a 4080 doe
XTX is only better at raster
hello
HEYY
Hi everyone
Hai
Idk if it’ll work but I’m going to try the RVC GUI using wine on Linux
1st off
RVC GUI is outdated
2nd off
Just run it natively
how do I run RVC on Linux
@waxen grove teach this man RVC on Linux
@elder willow https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/docs/en/README.en.md you might have to compile it yourself tho
You would run the sh file to open it
DO YOU KNOW WHO I AM? 😡
enna
do you know WHO I AM?
excellence
DO YOU KNOW who i am?
Or Estllee
Emi
WRONG
WRONG
WRONG
barf
Ayo? @elder willow level 1 !!! 
🇷🇺
Russia
correct
edit pings dont ping
silly
hello

ehhhh the sh file isn't all that great, and monstly targets macos
even on macos if python 3.11 pr 3.12 is default it just dies lmfao
🤓
Ayo? @neat crescent level 1 !!! 
what software is it to actually make a model?
RVC
🤓
Ayo? @storm meadow level 88 !!! 
actually to make a model you just need computer science
funny
Ayo? @viscid viper level 1 !!! 
What's your GPU ?
Smartest person on ai hub:
im not wrong tho
unzip a file & be done with it 
real
btw is applio still the best thing to use or did someone make smth even better
ew lol player
is it not possible to inference on CPU?
Ayo? @elder willow level 2 !!! 
Oh it is
I did on windows....
possible but slow
It's just that on linux pretty much all recent GPUs are supported lol
that's fine
And the instructions differ slightly for the different vendors
btw is it true that all linux users are femboys
Assuming the GPU/iGPU has over 2GB of VRAM, it can infer as far as I'm aware
At least the realtime stuff does work, I've yet to try inference
all i wanna know is how to install and get RVC running on my system
..how to install dez nuts in yo mouf?* :ayo:
I have a script to get it working on AMD cards, but the instructions on how to get it working is something along the lines of :
- get python 3.8, 3.9 or 3.10, not sure what mint defaults to nowadays, but it's one of those three I think
🤓
💀
🤣
What are u on bro
no speak ingles pls no toxic
😭
Bro wat
it came with Python 3.10.12 automatically installed
bro my fish started climbing the stairs so i had to put it in its cage 😭
- get the
python-venvpackage, as well asbuild-essentialandaria2(required to download models) - Setup a venv via
python -m venv ~/RVC(or anywhere else) - Activate the env via
source /PATH/TO/VENV/bin/activate - Clone the code via
git clone https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI.git - install torch according to what's needed (here it's
pip3 install torch torchaudio --index-url https://download.pytorch.org/whl/cpufor cpu only) - edit
requirements.txtin the repo, and remove the lines withonnxruntime-gpuandtorchandtorchaudioif present - navigate to the RVC cloned repo and
pip install -r requirements.txt - from the rvc repo, run
./tool/dlmodels.sh python infer-web.pyand voila, it'll work
so real my dog started meowing 😭 💔
You know they shut everyone off when it's a subscription model. Use weights.gg instead
alr thanks
You don’t
Use weights.gg
Bro is just mad he couldn’t make ai cover
where should the venv folder be
Anything you can remember
Anywhere you want it to be
Also once you put it somewhere, it's best not to move it, even though it can be done, it's just a little tedious
yeah you tell him
Ayo? @neat crescent level 2 !!! 
watafak bash: /home/reu/Documents/VENV/bin/activate: No such file or directory
did it successfully create the VENV folder ?
I don't remember if it creates one if it doesn't exist or silently fail
well I do have a RVC folder
but no VENV folder
make one
It should place the venv at the place you told it to via python -m venv
why does useing the voice changer make my 5800x go to 100%
because you're making it run on CPU ?
It's some real intensive stuff
no its on my 6800x
GPUs are much better for this
6800xt
NEVER MIND I FOUND IT (It was in the RVC folder))

Hey guys, i need a little help with making ai cover
first program to make my pc crash hard in 3 years
https://rentry.co/VoiceChangerGuide#uploading-models-fixing-amdarc-res-issues
refer to this, there's an extra step to get GPU acceleration on AMD cards
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...
that means its a good program
How u all doing? 🙂
Im ok i love voice ai
Ayo? @fluid falcon level 1 !!! 
Hope you’ll love prison
??
i am extremely new but I was wondering how I would even go about doing a text to speech with a characters voice (some pony from MLP)
what xd
Voice ai sucks
Use Okada or RVC
Okada for real-time and RVC for covers respectively
im a newby to it
im super new too hahah
There is guides for a reason
where can i find okada
i just tried making ai cover but i got an error
and i dony know how to deal with it
What’s the error and how did you get it
ok so im looking for specific models
Does the vocal file have any spaces in its filename?
Look on weights.gg
Should have all the models you need
i tried running it through coogle colab easy gui
Ayo? @lilac meteor level 1 !!! 
Link?
Super outdated
-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-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
is a 2-hour dataset suitable for a rvc model?
Use RVC Disconnected for training
Use AICoverGen for covers
fak it didnt format
More then enough
Yup
won't it cause any issues?
Oh actually it might cause issues
if it complains when running, try pip install onnxruntime
is it easy for the first time?
Check weights.gg
Should have it
Eh
I would recommend using a HuggingFace space over it
for example?
But how do i know if its that exact one
Oh I wasn't replying to you, sorry
Try it and see
k. ill try, thanks
as for your 2 hour dataset, the index file might have issues getting generated
no problem
https://rentry.org/ilarvc_inf_guide here’s the guide for the Ilaria RVC HuggingFace Space
and what if i use a 1 hour dataset?
RVC does weird shit to it when it decides it's too big, so you might have to reduce the audio length
Not sure about the minimum affected duration
but you might want to train the model with the full two hours (if they are good quality without random noises), and the index with as much dataset as possible without it getting trimmed
and is it better to set a similar length for every part of the dataset?
doesn't matter really
at the end it doesn't differentiate between all the different files
okay, thank you a lot
(RVC) reu@reu-Latitude-E5440:~/Retrieval-based-Voice-Conversion-WebUI$ ./tool/dlmodels.sh
bash: ./tool/dlmodels.sh: No such file or directory
uhm what is the best app/site to use to make ai singing
it will take a lot to train
Ayo? @wind stratus level 4 !!! 
RVC
Cc je suis nouveau ici
It’s good man 👍 just downloading the models now, almost done
if i want to do male voice but female voice rvc what pitch is best, (AI Song wise)
Searching for the perfect RVC e-girl voice model? 
Nice ! Going to sleep right now, I'll be back in a few hours
feel free to ping so I'll see 'em later
femboy model
Hello
weights.gg tengo
NO WAY IT ACTUALLY WORKED

Hej 💪👋
Hai
Ok so my last question is how do I get my own custom models loaded onto RVC
.pth in the weights folder
.index in the logs folder
Got it thank you
hi
How do you increase the output of the audio/
Ayo? @woven gulch level 4 !!! 
ok so for the next time I want to run RVC, what do I do now that I have everything installed
Ayo? @elder willow level 3 !!! 
should i do 40khz or 48khz
You ned to check using a spectrogram but it's probably 40hz
for the target sample rate?
Yea
ok
Do you not have uhh 32?
For most models 32 is best
Most datasets aren't higher
For most I've seen
You can check on a spectrogram like in spec or audacity
How do I use the models on my phone?
im tryna use this voice changer it it sounds choppy
Covers or realtime
Thx, do you have a link

Idc
Ayo? @prime kayak level 1 !!! 

-links
Suggestions for @prime kayak
- 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.
Ok so you would want to use ilaira Rvc

@wintry sinew person is on phone
Oh
I asked if they wanna do it on pc 
Hi
Hai
Hai :3
female
Ayo? @bold herald level 1 !!! 
@bold herald Searching for the perfect RVC e-girl voice model? 
Wat
Literally any other AI tool is barely functioning lol
What?
How do I avoid copyright when posting my AI covers?
Don't use instrumentals
Ok
You don’t
you cover songs that arent copyrighted or make your own instrumentals
hi
can i donwload the voices on @search-models ?
🗿
Don’t use kits ai
Find models using weights.gg instead
can someone help me
Here kitty you can has cheeseburger
yo can u help me by any chance
w this
Idk it depends on the issue
its the voice changer ojoin max 2
Sorry man idk anything about that maybe go on #✨│ai-help
it isnt a glitch its just
yo guys
Ayo? @elder willow level 1 !!! 
what is the rvc link to get the thing
IM SO CONFUSED
Ayo? @grim dagger level 1 !!! 
-local
Suggestions for @elder willow
- 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.
So I've been gone for a bit has there been any improvement on ai voices?
why are people saying "epochs" what does that mean?
hello
Training cycles, more or less does not mean better
wait so what are training cycles 😭
Ayo? @long gazelle level 1 !!! 
When it goes one time over the dataset (audio it's trained on)
hi
Ewwww voice ai
It basically means a full cycle that the AI goes through The entire training data
Like for example if you set it to 10 epochs it will go through the entire training data 10 times
how do you set a number of epochs?
Where exactly? In rvcDisconnected?
Mainline? Mangio?
Assuming you're using the Google colab You can find it in "total_epochs" under training
nah I just looked up a video on this voice changer cilent demo and i saw people saying "epochs" in the downloads and I have no clue what it means
Ah I see. You determine the epochs amount before even creating the model. It's not something you can change after creating one or in the voice changer app
You don’t set the epochs
The model creators do
OHHHHHH
Epochs is just how much training cycles the model had
that makes sense now thank you
More/less does not mean better
thank you
hi
i esspecially dont know why im currently making a video totorial for kits.ai when it suck so much
Do not post that
🔫
RVC is better in every possible way
yes but kits is like the easiest way to make covers
Ayo? @stray mortar level 4 !!! 
you dont even need to download any software
and i cant bother explaining how to use rvc
idk myself how to use it
hi
Same with RVC
you can use a online method like colab or hf spaces
thats still too complicated
And it’s open source and FREE
i couldnt figure it out
Try wordband or Astra labs then
It’s as simple as a discord command
Because that’s what it is
hi
sup
i made hundreds of thousands of views from my ai cover vidos
but made 0 dollars because of copyright
🗿
Kits is also the lowest quality method
No it’s not that
It’s because it’s using a outdated version of RVC
And you get 0 training settings
oh wow i didnt know that
and here they charge you at least 10 dollars a month for basic things you can do in other training softwares
shut up
I like this post a lot
If youre gonna pay just get professional software. Rvc is free
kid named yen
$50 for the best one
no way u mean i cant just take people's work and make money from it
how dare they
LOL
Hello
Meanwhile vocaloid producers are smart and share their stems because they know the more people cover their work makes them more popular
Would it be possible to make an easy to use gui for rvc for mobile
?
Ayo? @round python level 2 !!! 
Ilaria rvc
Link
i barely know what kits is, but i irrationally hate it for no reason
a website with a lot of unnecesary restrictions and annoying features
also, aren't ai covers basically dead now anyway
views started tanking on them instantly
its not gonna do well if its only 1 person singing the song
Paywalled tool that’s worse then RVC in every way for making ai cover
what i do is make multiple people sing the song and put in things to increase attention span
i'm only going to occasionally make them now, with videos that actually have effort put into them being my first priority
subway surfers moment
question im new how do you upload the custom voice files to work in tts ???
i use to do that, but all of those videos were forced to be taken down because of a little known voice actor
who you might have heard in open fortress
Generate tts then convert that with RVC
the guy got super pissed, joined my discord server, and told me to take down the videos 🗿
i know digital circus voic actor does that
Did you?
no wae he went that far?
Because that’s the right thing to do
wow
the longest time their voice had been in a single video was about (i think??) about 12 seconds
yeah, but i put one of them up on niconicodouga out of spite because it's goofy
because he never specified to not reupload them, nor to take them down from non-youtube platforms
i'm pretty sure they're still on odysee
if my voice ever gets used in ai im actually gonna laugh and promote that
it's already happened to me lol



