i just need settings for my specs
#✨│ai-help
low shard
crossfade: 0.15
extra time: 3
block size: play with it and be sure it's higher than the gpu delay
be sure to also use rmvpe
low shard
idk what to put on block size
there isn't really a good value, it varies alot by hardware and what program you're going to use it with, the only way is by setting it higher than the gpu delay
paper widget
which script ? sorry i am very new to these stuff
low shard
maybe try https://github.com/filipstrand/mflux but depends on ur M processor
low shard
alr lemme try
any updates?
paper widget
also i am trying to train a model but it doesnt save sometimes and i do not know what settings to use
is applio the best way to do this
oak edge
how do I change vocoder
in applio RVC
i'm using google colab public url gradio method
low shard
also i am trying to train a model but it doesnt save sometimes and i do not know...
what's your pc gpu and operating system? which guide link are you using?
paper widget
what's your pc gpu and operating system? which guide link are you using?
gpu 4060 8gb, win11,the on on aihub website
low shard
gpu 4060 8gb, win11,the on on aihub website
be sure to use https://docs.aihub.gg/essentials/how-to-make-voice-models/
paper widget
be sure to use https://docs.aihub.gg/essentials/how-to-make-voice-models/
other than 3rd step all good
uhhh??
low shard
other than 3rd step all good
wdym what's the error? u tried applio
tough fiber
how can extract thi
this* sh*
oh my god this really annoying af
@low shard nick can you help me please
cant extract this
low shard
<@911742715019001897> nick can you help me please
use winrar or 7zip to extract just the .zip file and it will do the rest automatically
don't use windows' defualt extractor
tough fiber
don't use windows' defualt extractor
thanks really <3
urban roost
(js copied and pasted bc my dumbass went to the wrong channel) Hello! beginner here. tried training my own model with rvc and it went on for almost a day before stopping because I ran out of memory 😭 are there any alternatives to making a model (besides me freeing up my memory ig.) unfortunately I only have an NVIDIA Quadro p600
latent kettle
(js copied and pasted bc my dumbass went to the wrong channel) Hello! beginner h...
Just use cloud platforms such as colab or kaggle
urban roost
Just use cloud platforms such as colab or kaggle
thank you! so js to confirm, I can make my model in Kaggle? google collab always cuts off my sessions bc of credit limitations
low shard
(js copied and pasted bc my dumbass went to the wrong channel) Hello! beginner h...
You deffo need to use cloud, here's every option:
Train (make) RVC Models on cloud:
- Prepare the Dataset
- Setup RVC:
Choose a cloud way to use RVC,
- Google Colabs (max 4 hours of daily T4 16gb gpu not granted for free, but easy to use, there's a paid tier):
- Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus, either T4x2 16gb each or P100 16gb, only free):
- Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly, Free Studios run 24/7 but require restart every 4 hours. There's a paid tier):
- Be sure to know about the tensorboard
Google Colab = Easier but risk of getting disconnected
Kaggle = Harder but more gpu time
If you want the easiest way and for free, is using https://weights.com/ which uses RVC
RVC Inference (use models) on pre-recorded audio on Cloud
You can use either:
- Weights.com: Easiest Possible Ever Automatic
- Ilaria RVC Zero: Fast and free on cloud
- Applio UI Colab: RVC Fork with some extra features like TTS
- RVC AI Cover Maker UI: Automatically Separates the Vocals and Instrumentals, converts the voice and mixes them back
For RVC training, Applio Kaggle gives the most free hours which is good for training,
If you prefer the easy automatic 1 click way, Weights does that and gives 1 model training for free weekly iirc
urban roost
tysm! I'll poke around a little and see what works best for me
Also, I have 78 voice samples for the training, is that enough? because I'm not sure if I can get more
low shard
tysm! I'll poke around a little and see what works best for me
You're welcome!
If you're talking about quality, all of them provide the same quality for training btw
Since rvc V2 has been the peak of Speech To Speech ai since 2023, and the original devs kinda left it to rot as it seems to be hard to upgrade it iirc
The forks, modified version, like Applio just have a simplified User Interface with other extra features like TTS and performance improvements
But the quality doesn't change
low shard
Also, I have 78 voice samples for the training, is that enough? because I'm not ...
How long would your dataset be in total?
urban roost
in total it clocks around 9 minutes and 8 seconds
urban roost
The forks, modified version, like Applio just have a simplified User Interface w...
ohhh ok
I'm willing to try/learn anything but easy interface would def help, ty !
low shard
in total it clocks around 9 minutes and 8 seconds
You're good, be sure it is high quality tho
It's better to have a 9 minutes high quality dataset than an 1hour of terrible audio,
Both quality and quantity matter, but quality a bit more
urban roost
You're good, be sure it is high quality tho
great! yeah i'd say the quality is fine lol
urban roost
It's better to have a 9 minutes high quality dataset than an 1hour of terrible a...
ofc ofc
thanks sm again <3
low shard
thanks sm again <3
Yw again, hopefully your mindset isn't confused anymore :trolley: /j
Feel free to ask here again for help, or make your own help post in #1192011222023950368 (both are good, even tho Forum is more organized)
quaint knot
Hi, does anyone know how to find a voice model that isn’t based on a specific person? I’m looking for a young Spanish male voice that isn’t famous, just sounds good
low shard
Yw again, hopefully your *mindset isn't confused anymore* :trolley: /j
Feel fre...
my nitro boohooh
urban roost
Yw again, hopefully your *mindset isn't confused anymore* :trolley: /j
Feel fre...
it's always confused 💔 that's why I'm always asking stupid questions! takes one to know one 💔 ty again!
low shard
Hi, does anyone know how to find a voice model that isn’t based on a specific pe...
Hello, RVC STS AI Voices models need to be trained on something, unless you merge models (basing on trained models, to make a 'unique' one), are you perhaps using any realtime voice changer?
quaint knot
Hello, RVC STS AI Voices models need to be trained on something, unless you merg...
I don't usually use real-time voice changers, but rather I would need to edit the voice in post-production, that is, record everything with my normal voice and in editing change it so that it doesn't sound like mine, but I can't find models.
unique root
I want to make my own AI assistant (eventually to turn into a commercial service). I want the responses to come out as human-like as possible. I will probably need datasets to use (or my own?) for fine-tuning the model and using RAG.
How does one get started with this? I understand I'd need to obtain and use any UGC data legally and all that, I just don't know where to get started and what the best way to gather data (and put into .JSONL file format) is.
Yeah, I wouldn't even know what to search for related to my topic
nevermind, this seems to be voice only or some sht
low shard
I don't usually use real-time voice changers, but rather I would need to edit th...
try applio merging tab maybe?
drifting aurora
Hi I want to know which version should I download for the real time voice changer for windows I want to download the OKADA voice changer
low shard
Hi I want to know which version should I download for the real time voice change...
Hello!
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
- Many users confuse Wokada and RVC, There isn't a program that does everything, there's a program for each thing
- what tutorial link are you using
- a screenshot of the program
simple ore
I don't usually use real-time voice changers, but rather I would need to edit th...
chatterbox tts allows you to create synthetic voice variations based on an input voice
low shard
nevermind, this seems to be voice only or some sht
This was Voice, RVC, focused, but we are slowly shifting to general ai
Maybe this dataset full of conversations used in movies could help for feeling human like https://www.kaggle.com/datasets/rajathmc/cornell-moviedialog-corpus
drifting aurora
- RTX 4070 Super
- Window 10 Pro
- The program that converts my voice to a model in real time either to troll friends or to rp ot whatever
- I am currently here https://huggingface.co/wok000/vcclient000/tree/main
low shard
- RTX 4070 Super
- Window 10 Pro
- The program that converts my voice to a mode...
lemme guess, did you use a youtube tutorial that told you to use that link along with vb audiio cable?
all video tutorials are old, that is a version of original wokada
simple ore
@quaint knot
low shard
and vb audio cable creates issues on windows
delete the folder, zip and uninstall from windows app settings
drifting aurora
lemme guess, did you use a youtube tutorial that told you to use that link along...
no I used to use it before and I formatted my pc I just went to the github but forgot how to download it again
low shard
no I used to use it before and I formatted my pc I just went to the github but f...
Ohh I see, unfortunately that one is original wokada and outdated now
drifting aurora
what should I download now?
low shard
it's not suggested since months
low shard
- RTX 4070 Super
- Window 10 Pro
- The program that converts my voice to a mode...
troll
like spongebob or e girl trolling?
drifting aurora
lol in general not egirl trolling
I used it to record sometimes instead of my voice
low shard
what should I download now?
-realtime
patent trellis
low shard
you can choose either wokada deiteris fork or vonovox
it's better you read each pros&cons
drifting aurora
ooo so the original wokada is not working anymore? and there is a deiteris fork version??
low shard
ooo so the original wokada is not working anymore? and there is a deiteris fork ...
original wokada isn't suggested, because of it's performance, the newer original wokada is just User Interface changes lol
valid axle
Can I talk here instead?
low shard
You can either use wokada deiteris fork (modified version) or vonovox
vonovox is still being worked on and has some potential
drifting aurora
I tried vonovox I couldn't download it
low shard
ooo so the original wokada is not working anymore? and there is a deiteris fork ...
id really suggest you to read both vonovox & deiteris fork pros&cons
low shard
I tried vonovox I couldn't download it
what was the error?
drifting aurora
there was an issue with the bat file like at the end when It was extracting tinkers or whatever it got stuck
low shard
there was an issue with the bat file like at the end when It was extracting tink...
maybe try with the latest version
1.50
drifting aurora
yup that's what I tried
low shard
Can I talk here instead?
#🧬│ai-chat message sure, but it is more chaotic than #1192011222023950368 , would u rather here or in #1192011222023950368 ?
valid axle
https://discord.com/channels/1159260121998827560/1159289354439626772/14049213011...
Forum
pastel oak
lmfao
valid axle
Okie, be sure to elaborate here and we (helpers) will help
Check the forums
low shard
lmfao
im gen shocked, that dude was still asking for catfishing and even admitting it after days of continously asking for "e girl"
😭
low shard
Check the forums
Sure!
pastel oak
im gen shocked, that dude was still asking for catfishing and even admitting it ...
smartest user ever
low shard
smartest user ever
bro made like 2 post forums and 30 messages asking only for egirl "rp" then directly said he says "rp" only bc catfishers get banned
valid axle
im gen shocked, that dude was still asking for catfishing and even admitting it ...
Lmao?
low shard
istg 😭
low shard
yup that's what I tried
weird, if you want you could try again and show me the error for me to help you
drifting aurora
okie I am really confused this happens on every downloader now I am trying with the Deiteris version and this is what it gets stuck on
simple ore
wifi?
drifting aurora
??
on every installer it always somehow stops working for rmvpe and also content vec
I wonder if my network is blocking it or something
pastel oak
I wonder if my network is blocking it or something
It's possible, could also be slow internet losing connection by the time it gets to the end
You can manually download the files from the drive. Instructions and link:
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#how-to-fix-failed-to-download-or-verify
low shard
You can manually download the files from the drive. Instructions and link:
https...
not from drive, that one was deleted by deiteris
I reuploaded it in my github repo release
drifting aurora
where can I find that?
low shard
where can I find that?
drifting aurora
is it the pretrain folder?
pastel oak
Yes
drifting aurora
and what exactly should I do with that folder
pastel oak
You take everything inside the zip and put it into the "pretrain" folder of MMVCServerSIO
drifting aurora
ooo I see a file called pretrain in the voicechanger folder
pastel oak
That's the download process that freezes on your end
Yes
You put the files from the zip inside that
drifting aurora
okie thank you I will try that
oh god it works
thank you
I can't believe I formatted my pc for this
low shard
I can't believe I formatted my pc for this
do you want me to check your settings if u want maybe?
quaint knot
<@1361651985014849637>
What is this
simple ore
What is this
just a solution for your "record everything with my normal voice and in editing change it so that it doesn't sound like mine, but I can't find models."
to make that I used a 30s voice sample + randomizer to make 10 different voices
viral mason
<@1361651985014849637>
Could I have the original audio
quaint knot
just a solution for your "record everything with my normal voice and in editing ...
Okay, I’m so sorry. I’m new and I didn’t know much about this, I just understand it now
earnest forge
Say, what'd realistically be the best case scenario settings for freq/epoch when training off 48000mHz data (RTX 4090)
green elk
how do i download the voice changer?
low shard
how do i download the voice changer?
!howtoask
low shard
i downloaded it, but how do i hear the voice that im using
don't use youtube tutorials, please elaborate as explained above
green elk
don't use youtube tutorials, please elaborate as explained above
Full GPU Name: dont got one
Operating System: (e.g., Windows 10)
Detailed Description: i cant hear the voice that im using
Tutorial Used:https://www.youtube.com/watch?v=LX5en3pZJwM
Screenshot: no error message
low shard
Full GPU Name: dont got one
Operating System: (e.g., Windows 10)
Detailed Descri...
that tutorial is outdated
delete the folder, zip and uninstall from windows app settings
low shard
Full GPU Name: dont got one
Operating System: (e.g., Windows 10)
Detailed Descri...
dont got one
Are you sure?
low shard
idk if i have one or not
that is the most important thing
low shard
idk where to check for that
You can check your pc gpu on Windows via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU
low shard
AMD Radeon RX 7800 XT
oh great, it's a dedicated gpu
green elk
ok so what do i do after i uninstall
low shard
Full GPU Name: dont got one
Operating System: (e.g., Windows 10)
Detailed Descri...
you're just looking for realtime roleplay in vc/games or e girl trolling?
green elk
well im already a girl but im looking for voices to have fun with
low shard
ok so what do i do after i uninstall
-realtime
low shard
1st link, get wokada deiteris fork
green elk
where is it
low shard
where is it
the guide
low shard
im so confused on what im looking at
you need to read the guide and get the wokada deiteris fork amd/cpu/intel version
green elk
is that all i need
green elk
you need to read the guide and get the wokada deiteris fork amd/cpu/intel versio...
but where do i get that
low shard
is that all i need
nope, i was just explaining which version you need
low shard
i need to read all of it?
yeah, you can ignore steps related to other things like linux or nvidia tho
green elk
ok
low shard
it isn't 1 click, be careful, missing a step can literally fuck up your whole audio system
urban ocean
can somone help me find a reinforcement learing model tutorial for python, all tutorials i found are outdated
south verge
So I’m trying to make a 20-30 min long voice clip for a character I want to make an AI song of, but the character always talks with music behind him. There’s hours and hours of audio but it’s all with music or sound effects. How badly will this affect the final result? I’m trying to only use the cleanest clips but even those have weird noises in the background
viral mason
So I’m trying to make a 20-30 min long voice clip for a character I want to make...
use a cleaning tool to remove all music, fx, reverb, ect because any voice model trained must have clean audio
https://huggingface.co/spaces/TheStinger/UVR5_UI
south verge
The link you sent can clean the audio? Thank u!
viral mason
The link you sent can clean the audio? Thank u!
yes!
though there are more options than that, this one I recommend for it being simple and easy to figure out
south verge
thank u! i am so new to this lol
viral mason
thank u! i am so new to this lol
these are the models I suggest for a good cleaning (in order)
Vocals FV4 by gabox
MelBand Roformer | De-Reverb by anvuew
UVR-De-Echo-Normal.pth
Mel-Roformer-Denoise-Aufr33
optional -
(use one or the other)
UVR-MDX-NET_Crowd_HQ_1
Mel-Roformer-Crowd-Aufr33-Viperx
just 4 is needed ^^
south verge
a lot of the music is like, background music so idk how well it will work but i will def give it a shot once im done editing the thing
thank you so so much🙏
viral mason
no problem! if u have any other questions dm me or @ me here!
south verge
Last question. i want to make a full version of this song, so should i add this singing clip and put it through that file too, or should i train the voice model only on speaking clips
@viral mason
sorry for the questions
viral mason
https://youtu.be/hZV5hPmwJ6M
Last question. i want to make a full version of th...
it's best to train it only with talking but I don't follow the rules >:3
south verge
so far it is 10 mins but i have 2 more seasons to add and im going to try to make it as long as i possibly can
viral mason
alrighty ^^
south verge
what length should i aim for?
viral mason
it's best to try for at most 30 minutes at possible
south verge
ok
viral mason
but if u can't get more than 10 minutes that's also ok
south verge
sorry last question, should i be including breathing in the data set? like if the character takes a breath do i keep that in?
viral mason
sorry last question, should i be including breathing in the data set? like if th...
yes breathing helps a lot, although I am not sure about like deep sighs
south verge
ok
rustic swift
hello.
viral mason
south verge
how do i know if im using the vocals isolation right 😩
nvm i think im starting to get it
i just ran out of the credits i think
viral mason
i just ran out of the credits i think
Don't worry it resets after a day
And again there are other options although I'd ask Eddy
Or a mod like Nick
They're both very helpful
native ridge
da fu does ''failed to fetch'' mean lol
nova anvil
I need a smart free ai tool which can copy an edit's visuals and animations and make a same one for the pics I provide on the same beat drop
in short I just wanna swap this pics in a short edit
Can anyone help please?
thorny reef
Does anyone has a AI that can develop whole Programms?
paper widget
wdym what's the error? u tried applio
yes rn what i am doing is first place the audio file in dataset folder, then pre-processing, then extraction of features then creating index and start training
in the whole process i dont understand anything about tensor board thingy also the model is not getting saved till i check the overtraining toggle
visual hull
Hello, I am brand new here and I wanted to know how people take a voice and make it say whatever they want? Is that a specific ai site?
I'm on MacOS
outer wasp
There are setting, training or simple model on Okada to make my voice feel more nature? I sound like my nose is stuffy. When laugh or Yelling it not nature
high basalt
is there a software doing more realistic voice convertion than w okada and maybe faster?
quaint knot
Is it necessary to use models trained in the same language as the audio to obtain a good result or does it not influence?
low shard
Is it necessary to use models trained in the same language as the audio to obtai...
required? Nope, but suggested
you could also lower the index ratio to make sure it uses less of the trained accent
low shard
is there a software doing more realistic voice convertion than w okada and maybe...
what's your pc gpu? I'm guessing you're on windows 11, and what tutorial link did u use?
low shard
Hello, I am brand new here and I wanted to know how people take a voice and make...
Hello, I am brand new here and I wanted to know how people take a voice and make it say whatever they want?
Are you talking about Speech To Speech (STS)
Or Text To Speech (TTS)?
I'm on MacOS
Which version, and M Chip?
low shard
yes rn what i am doing is first place the audio file in dataset folder, then pre...
have you tried checking https://docs.aihub.gg/rvc/resources/training/#epochs--overtraining ?
low shard
Does anyone has a AI that can develop whole Programms?
I mean you can just use any chatbot to develop code
don't expect too much, overall they just do text prediction
low shard
is there a tutorial on how to make a custom rvc voice
Yeah but it depends on your hardware
Could you please tell me ur pc gpu and operating system, so i could guide you in which tool would be good for ya?
brittle wing
ah the bot is taking a long time
my wokada fork worked fine until it didn't and server audio works but client doesn't
whats happening
low shard
my wokada fork worked fine until it didn't and server audio works but client doe...
can you show a screenshot of ur settings?
brittle wing
sure hold up lemme launch it
it works on server but not client
i reallly need the suppression so i cant really switch
simple ore
i reallly need the suppression so i cant really switch
if you want a good noise suppression, use nvidia broadcast app
simple ore
mic-> broadcast app -> voice changer
brittle wing
what browser are u using? u sure u gave it microphone perms?
yeah i checked it all but i just decided to use nvidia broadcasst
thanks tho
orchid spindle
why doesnt the download on applio work
simple ore
why doesnt the download on applio work
what are you trying to download and where?
local or colab?
orchid spindle
local
the download button and then it comes up with the windows save thing
but when i save the file just isnt there lol
simple ore
what are you trying to download/
if you had a model trained, it is in the logs
there's nothing to download
simple ore
im trying to download the converted audio
what browser is that?
orchid spindle
how can i fix that
simple ore
make a ticket for devs
orchid spindle
oh
brittle burrow
why cant i send screenshots in this channel i need help
orchid spindle
with a browser there are no such issues
i can just open it in a browser im pretty sure
simple ore
yeah
simple ore
why cant i send screenshots in this channel i need help
#1192011222023950368 try there
brittle burrow
<#1192011222023950368> try there
i posted a forum
viral mason
is there a tutorial on how to make a custom rvc voice
There aren't any up to date one's on yt don't ever use YouTube for anything rvc or real-time vc related but I have recorded one myself for kaggle applio
lethal iron
Anyone who why my comfyui-zluda gives a cuda error when i try to create a image?
simple ore
Anyone who why my comfyui-zluda gives a cuda error when i try to create a image?
what error, what's your gpu/
lethal iron
what error, what's your gpu/
some cuda malloc error i cant remember 100 percent which cuda error and my gpu is rx 6750xt
visual hull
> Hello, I am brand new here and I wanted to know how people take a voice and ma...
MacOS Sonoma 14.2.1 (23C71)
For the context, I'm starting as a content creator and there's a scene in a movie between 2 characters and I want to make them swap their voices, not their dialogue
low shard
For the context, I'm starting as a content creator and there's a scene in a movi...
it seems like Speech To Speech then, u'd need RVC
what M chip do u have?
visual hull
M1
low shard
M1
For Inference (use models) Mac (which doesn't have great support), You can:
- Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides, probably won't be able to train, make models):
- Cloud (remote good pc, easier and faster than ur PC but limited time):
- Weights.com: Partnered with AI Hub, lets u do them easily but u may be in a queue
- Applio Colab: max 4 hours, not granted, of GPU
- RVC-AI-Cover-Maker-U Colab & Kaggle: Automatically separates the vocals and instrumentals, converts the voice and mix all together back
Easiest possible (automatically separates vocals & instrumentals) : weights.com & rvc-ai-cover-maker-ui colab/kaggle
Easiest cloud: applio colab
Easiest Local: Applio
I'd suggest using cloud, your m chip isn't good
lethal iron
what error, what's your gpu/
I would love to know is it possible to get it work
simple ore
I would love to know is it possible to get it work
i need to see the error
probably running out of vram, especially possible with SDXL's bad VAE file
lethal iron
Oh okay I will take a look what its called a bit later
analog obsidian
probably running out of vram, especially possible with SDXL's bad VAE file
btw do you know which webui people use for sdxl/flux nowadays? 
simple ore
comfy mostly
analog obsidian
simple ore
i use sd.next
analog obsidian
i use sd.next
is still getting updates?
simple ore
a million of them every week
analog obsidian
oh nice, thanks
simple ore
oh nice, thanks
just one week of updates
analog obsidian
just one week of updates
perfect
simple ore
use dev branch to be on the cutting edge
analog obsidian
trim onyx
is there a new model yet?
viral mason
viral mason
idk what you're talking about, elaborate
trim onyx
last time i used the ai voice model was like 8 months ago has gotten any updates or a refresh?
viral mason
what ai voice
viral mason
that's a voice changer
viral mason
there are two versions of the voice changer that are recommended right now, if you use nvidia u can use both
trim onyx
i recently switched to amd
viral mason
ah
viral mason
I am not entirely sure as I use Nvida and more options are available for that lol
viral mason
still u can look here at the options if u like, the first one should have amd compatibility
!realtime
vernal bone
trim onyx
might need Local realtime RVC
viral mason
these are the local options
viral mason
you got this man, btw if u need better help u can ask a mod or one of the online helpers
viral mason
wait but these are mainly for nvidia
vonovox for now is only nvidia but Wokada Deiteris Fork does have amd although I have no info on that
viral mason
is MMVC outdated now?
yes it is, it has older code and is just slower than these new ones
trim onyx
welp time to wip everything of it of this clean pc
you ai guys move too fast
one day on mccv and then on onitrix vi
viral mason
make sure you screenshot which voices u used
trim onyx
is there a reason for? just asking
trim onyx
oh okay thanks yea befiore i just labeled them thing 1 through 6
worked good untill i forogot that i already made a number 5 then label the other 5
work whre can i find the model i need for my vc
viral mason
oh okay thanks yea befiore i just labeled them thing 1 through 6
dang
viral mason
work whre can i find the model i need for my vc
if u still have them downloaded u can just re find them there and make a new folder to put them all in
or redownload them off of either weights or through here in the voice model section
trim onyx
wait it said make sure you hae unistalled all versions of this product
how do i know it wont just blow up
trim onyx
i ahd the same issue
simple ore
you gotta delete that old version and something less moldly
viral mason
if u remember the names of the voices just look here
https://discord.com/channels/1159260121998827560/1175430844685484042
trim onyx
work
\hlp
my ai keeps runniong away
the sofrware
WHERE DO I FIND THE VOICE CHANGER
ALL I SEEE IS THAT AUDIO CABLE MANE
viral mason
I tried..
trim onyx
worm
is the one i should be getting deiteris w okada fork?
<@&1159293204038955078> can one help such as me dearly
viral mason
is the one i should be getting deiteris w okada fork?
yes
low shard
WHERE DO I FIND THE VOICE CHANGER
what's ur pc gpu? i'm guessing ur on windows 11
trim onyx
what's ur pc gpu? i'm guessing ur on windows 11
ojn windows 11 i have arx 9070 xt 16gb
viral mason
copy the settings I have here if you downloaded deiteris as just a starter for testing quality with not too much delay
trim onyx
proccessor is a ryzen 7 7700x
low shard
ojn windows 11 i have arx 9070 xt 16gb
-realtime
low shard
get wokada deiteris fork
low shard
what link do i click on the website
you need to read the documentation, it's more than just clicking the link, be careful and let me know for everything!
that's old original wokada, dont use youtube video tutorials
delete the folder, zip and uninstall from windows app settings
You have been warned. That is catfishing
low shard
It's still catfishing and illegal
trim onyx
i wanna be venom
viral mason
i wanna be venom
finally a person who wants to use the voice changer for a good purpose
which venom btw
quaint knot
is there any way to download a model from elevenlabs?
trim onyx
which venom btw
the movie from alst dance sounds more scary and clearest
low shard
what did I do
you haven't been warned nor anything, just don't help such people tho
trim onyx
yes sir🫡
low shard
is there any way to download a model from elevenlabs?
nope, it uses a completely different technology
11labs is closed source and TTS
RVC is STS
quaint knot
nope, it uses a completely different technology
11labs is closed source and TTS
...
o ok becouse i really liked one model
viral mason
you haven't been warned nor anything, just don't help such people tho
I didn't see the thing he replied to until a little after I sent that message sorry
low shard
I didn't see the thing he replied to until a little after I sent that message so...
It's fine don't worry :D
low shard
o ok becouse i really liked one model
Unfortunately, you can't download it
you can use it in 11labs website though
viral mason
the movie from alst dance sounds more scary and clearest
like I said earlier, voice model section https://discord.com/channels/1159260121998827560/1175430844685484042
low shard
low shard
gosh why was there so many
huh? there are over 10k+ of RVC models iirc
trim onyx
no the venom ones
low shard
didnt know that
venom is a famous marvel character
the "we are venom" is just a line from the movies, we aren't affilitated with marvel
low shard
im so lost
on which rvc venom model to choose ?
low shard
i dont even know where top click to read
ohh the voice changer
trim onyx
yea
low shard
here's the direct link: https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/
you can skip any steps not related to windows amd, like the linux or nvidia ones
be sure to not skip the virtual audio cable, that is for all users
low shard
yea i lready got that downloaded
u downloaded vac lite and run setup64.exe and clicked yes?
trim onyx
yes
low shard
great
low shard
just one issue wheres the download for deiteris
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-amd-intel-and-cpu-on-windows this exact steo is for you!
low shard
theres not one for amd gpus yet ?
there is, the exact link i sent is for all amd or intel or cpu users
that one works for amd gpus
low shard
i think i had this one already
you sure? this one opens in the browser
leaden saffron
im on amd i downloaded the thing from github made sure to download the right one not sure where to go in the files thougjh every exe i run does nothing
wait nvm
got it working i think
low shard
im on amd i downloaded the thing from github made sure to download the right one...
i feel like you're using that over year old version of original wokada with vb audio cable
don't use that
what's ur pc gpu? im guessing ur on windows 11
leaden saffron
windows 11 amd and it's the whichever it's called
valid condor
so I just need help with making an AI, Ive got an idea but I cant get the tools to work with me.
nor do I know what tools I need
low shard
windows 11 amd and it's the whichever it's called
amd and it's the whichever it's called
the gpu is the most crucial point actually, you need to be sure to know the name
You can check your pc gpu on Windows via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU
everything ai related dpeends on your gpu
low shard
https://github.com/w-okada/voice-changer/blob/v.2/docs_i18n/README_en.md
that's original wokada, not suggested anymore
and some users reported issues with vb audio cable, which is why we don't suggest that either anymore
leaden saffron
yeah for amd it's ML
low shard
id suggest you to delete the zip, folder and uninstall vb audio cable from windows app settings
leaden saffron
and some users reported issues with vb audio cable, which is why we don't sugges...
whichever works honestly aslong as it works idm
low shard
yeah for amd it's ML
you need to give the full name of your AMD GPU
to be sure which settings are needed, and if it's good enough for whatever task you want to do
leaden saffron
I'm not completely sure which it is maybe amd radeon
low shard
I'm not completely sure which it is maybe amd radeon
please do this and tell me
leaden saffron
I know that but I'm not at my PC at the moment
which is why I tried to name it off the top of my head
low shard
I know that but I'm not at my PC at the moment
please let me know when you're at your pc, because the settings vary depennding on it, and if it's even good enough
Let me know!
leaden saffron
please let me know when you're at your pc, because the settings vary depennding ...
alright I'll ping you when!
timber merlin
hello i have an isuse, few days ago everything was fine but today i tried to test the voice but its not working, no sound is coming out
low shard
hello i have an isuse, few days ago everything was fine but today i tried to tes...
uncheck sup1, if you got noise isues use sup2
uninstall vb audio cable from windows app settings, many users report it doing randdom issues on windows
timber merlin
uninstall vb audio cable from windows app settings, many users report it doing r...
okay ill try it
timber merlin
uninstall vb audio cable from windows app settings, many users report it doing r...
well i reinstalled it and i still cant hear it
low shard
well i reinstalled it and i still cant hear it
you shouldn't reinstall it
low shard
yeah i was going to say that
low shard
hello i have an isuse, few days ago everything was fine but today i tried to tes...
after you install vac lite
set output to line 1
and monitor to headphones to hear urself optionally
scenic imp
[SIO] rconnection failed Error: xhr poll error ????
timber merlin
I'm using windows 10, gpu 9070 xt (in screenshot) everything was working fine a few days ago but all of the sudden today it would not pick up my voice i think, i tried every input and output option and tried other mics
leaden saffron
please let me know when you're at your pc, because the settings vary depennding ...
AMD RADEON RX 6550M
south verge
which website would you recommend to train an RVC? I am so new to this and im lost in the guide. I have a very good data set, but my computer cannot run a program so it needs to be on the cloud
if theres no good free option, i can probably use someone else's computer but id like to do it on my own if possible :)
lethal iron
probably running out of vram, especially possible with SDXL's bad VAE file
A friend who knows some about these things said that its 99% not that because I streamed that to him when i tried to set it up
simple ore
A friend who knows some about these things said that its 99% not that because I ...
you can only know that by examining task manager/performance tab and seeing how gpu usage does
lethal iron
you can only know that by examining task manager/performance tab and seeing how ...
I remember clearly the gpu was using 9/12gb vram
simple ore
show the error dump
lethal iron
I am so sorry me and My friend restarted and I Have to do The whole thing again by The tutorial and then show you what error i get so i Will show on friday okay
But I am really thankful you want to help me
amber silo
why does it keep saying "trial" in my mic?
simple ore
why does it keep saying "trial" in my mic?
you've downloaded shareware virtual cable
amber silo
?
simple ore
instead of following the guide like a good boy and downloading v4.70 lite
amber silo
ight
prec8 it
theres a bunch of stuff in the extracted file.. i press setup64.exe right?
simple ore
theres a bunch of stuff in the extracted file.. i press setup64.exe right?
yes
follow the guide
timber merlin
hello, i tried to uninstall the old vb and use vb lite as instructed but my issue still presists, i can't hear myself when using it
im using windows 10
it was working fine a few days ago
but today all of the sudden it stopped working
weirdly enough, it works when using server mode, but client doesnt work?
@low shard
jovial shadow
hi guys do u know how i can use the ai voces on discord??? on calls with my friends??
leaden saffron
hello, i tried to uninstall the old vb and use vb lite as instructed but my issu...
have you tried enabling postthru
timber merlin
have you tried enabling postthru
where do i find the option to enable it? (btw it works for server mode, just client mode doesnt work)
well im going to be busy for a while now but if anyone has any suggestions or ideas on how to fix my issue i made a forum about it here, it'd be apperciated ❤️ https://discord.com/channels/1159260121998827560/1405280433360343171
stable perch
everytime i download a model it doesnt work but the original ones work
but models dont
low shard
everytime i download a model it doesnt work but the original ones work
!howtoask
low shard
Please elaborate
leaden saffron
Please elaborate
it's a AMD RADEON RX 6550M
low shard
wokada deiteris fork
leaden saffron
wokada deiteris fork
last time I downloaded the right one it just kept failing right before downloading
low shard
which website would you recommend to train an RVC? I am so new to this and im lo...
what's ur pc gpu and operating system? are u looking for a realtime voice changer or ai covers?
low shard
last time I downloaded the right one it just kept failing right before downloadi...
i'm not sure you used this one, but be sure to have a good connection
leaden saffron
i'm not sure you used this one, but be sure to have a good connection
yeah my connection is very stable
I was so confused
why it kept doin it
low shard
hi guys do u know how i can use the ai voces on discord??? on calls with my frie...
yes, but you gotta elaborate
!howtoask
low shard
why it kept doin it
try now
south verge
what's ur pc gpu and operating system? are u looking for a realtime voice change...
macbook pro M1 chip
no idea what that is for gpu
8 i think
and just looking for something to make song covers with a character from a show that doesnt seem to have any voice models
low shard
and just looking for something to make song covers with a character from a show ...
For Inference (use models) Mac (which doesn't have great support), You can:
- Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides, probably won't be able to train, make models):
- Cloud (remote good pc, easier and faster than ur PC but limited time):
- Weights.com: Partnered with AI Hub, lets u do them easily but u may be in a queue
- Applio Colab: max 4 hours, not granted, of GPU
- RVC-AI-Cover-Maker-U Colab & Kaggle: Automatically separates the vocals and instrumentals, converts the voice and mix all together back
Easiest possible (automatically separates vocals & instrumentals) : weights.com & rvc-ai-cover-maker-ui colab/kaggle
Easiest cloud: applio colab
Easiest Local: Applio
I'd just suggest you to use cloud
south verge
For Inference (use models) Mac (which doesn't have great support), You can:
- L...
weights said my file size was too large
prime tinsel
do any of u know why my rvc comes with extra voice cracks?
low shard
do any of u know why my rvc comes with extra voice cracks?
rvc doesn't mean realtime voice changer
!howtoask
karmic egret
how can i install the vc i have an 2080ti and and win 11
brittle wing
how can i install the vc i have an 2080ti and and win 11
I recommend deiteris wokada fork or vonovox. They have guides in the docs which is really easy to follow, just read the pros and cons for both voice changers to see what you'd like best.
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/
https://docs.aihub.gg/realtime-voice-changer/local/vonovox/
karmic egret
can you look in dm pls
shrewd forum
this isnt help but a question. does FP32 make it sound better? im using the w-okada forked and theres a feature where u can force it. should i enable it?
analog obsidian
this isnt help but a question. does FP32 make it sound better? im using the w-ok...
it's a tricky question, but at least, its needed for fp32 models to work properly otherwise they randomly have glitches, or at least my fp32 models were glitching when i enabled fp16 infer in realtime, tho for some reason seems to works fine in local rvc
shrewd forum
it's a tricky question, but at least, its needed for fp32 models to work properl...
well i just enabled force f32 and i got 0 glitches so far
i tested it by talking for a bit
didnt hear any so far
analog obsidian
yes i mean like, fp32 models glitch when fp16 infer is enabled
shrewd forum
i mean what would u say then? is it genuinley preffered?
analog obsidian
doesnt change that much honestly
shrewd forum
i figured
analog obsidian
but for fp32 models i'd rather have it enabled so they dont glitch
shrewd forum
i mean i just tried it and i hear a good diffrence but idk if im just crazy
like when i talk it sounds more natural
it sounded good before but it sounds better now
but like i said, idk if thats ME or thats actually the voicemodel
analog obsidian
i can't say for sure, all i can say, in regular rvc (not realtime) the results are exactly the same
but realtime pipeline is different so, ye
shrewd forum
alright, thanks then!
analog obsidian
shrewd forum
and also about w-okada updates, did the developers/engineers just stop working on it?
from what i heard AI hub and W-okada its self is literally dead
analog obsidian
and also about w-okada updates, did the developers/engineers just stop working o...
yes, vonovox and training models using the spin embedder is what is best rn
w-okada is something from the past
shrewd forum
w-okada is something from the past
hypathetically if i got people to code w-okada for me is there a way to make it any better then it is now
analog obsidian
afaik, no, because the whole code is a mess
shrewd forum
well IF by a chance theres a way to code it can it be better then it is now?
analog obsidian
for that you have to rewrite the entire program
is using super old outdated software of 2017
shrewd forum
im planning to hire a team to just rewrite it entirley
would i have to ask for permission?
analog obsidian
would i have to ask for permission?
try to get in contact with wok
hes the original author
shrewd forum
is he on discord?
analog obsidian
is he on discord?
wok5681
his disc
he's here
the code is open source
so in theory, is fine? i guess
shrewd forum
welll
id wanna ask some questions
hes the original coder of w-okada?
i thought the original author was gone
😭
like last year i could NEVER find him,
analog obsidian
wokada
wokad
woka
wok
shrewd forum
nobody knew who he was
analog obsidian
he speaks japanese btw
analog obsidian
analog obsidian
no idea lol
smoky arrow
Yeah but it depends on your hardware
Could you please tell me ur pc gpu and oper...
rtx 4080 its on laptop though
smoky arrow
There aren't any up to date one's on yt **don't ever use YouTube for anything rv...
i did try using the mango rvc fork main
@low shard i did try using the mango rvc fork main to train models but not sure if its working or not
viral mason
i did try using the mango rvc fork main
I don't use mangio so I wouldn't know
smoky arrow
what do you use?
viral mason
are u training a model?
smoky arrow
yeah im trying to create my own model
viral mason
I use applio on kaggle, I train with rmvpe
low shard
<@911742715019001897> i did try using the mango rvc fork main to train models bu...
Mangio rvc fork is abandoned since 2023
smoky arrow
Mangio rvc fork is abandoned since 2023
wait wut really?
low shard
rtx 4080 its on laptop though
check https://docs.aihub.gg/ you can do it locally with applio
low shard
wait wut really?
Yes don't use video tutorials
smoky arrow
is it applio?
smoky arrow
Yes don't use video tutorials
so are we gonna get both the index and pth files?
low shard
is it applio?
That is our documentation, you can read here more about how to make models with Applio
smoky arrow
That is our documentation, you can read here more about how to make models with ...
ok2
smoky arrow
That is our documentation, you can read here more about how to make models with ...
do i have to do the cmd thing if i were on 4080? or is it for 5000
low shard
do i have to do the cmd thing if i were on 4080? or is it for 5000
This is only for the Rtz 5000 serie
So nope
Not for your Rtx 4080
smoky arrow
ohh i see
yeah sry its kinda confuse me for a sec
also to train new model, which sampling rate is good? 32k 40k 48k?
low shard
also to train new model, which sampling rate is good? 32k 40k 48k?
You need to train the model depending on the sample rate of your dataset
smoky arrow
ohh i see okay
smoky arrow
You need to train the model depending on the sample rate of your dataset
hi i get abit confuse im on the last step, do i click train model or the index?
smoky arrow
index, then train
okk
finite axle
does anyone know how to install pytorch...
smoky arrow
you can change the sample rate with audacity to match 32k
ohh i see thanks!
south verge
super confused. so i have applio running on a diff computer and i have a dataset but how do i turn that data set into a voice?
the guide doesnt really explain it
viral mason
super confused. so i have applio running on a diff computer and i have a dataset...
are u using it locally or on kaggle?
I got a whole kaggle applio tutorial
exotic ridge
hello! i have a question for Deiteris' W Okada Fork Local Realtime Voice Changer...
I have windows NVIDIA RTX 5060, so would that make me download the NVIDIA RTX 5000-series on Windows version?
If that's the case, is there any step by step guide how to download it? because it said i have to download 3 files, while there's a total of 5 files on the link (https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335) It would help me a lot if there's a procedure or video somewhere
But if it's not recommended for me to download the NVIDIA RTX 5000-series on Windows version, then which version would be the best
south verge
are u using it locally or on kaggle?
locally, and the sample rate doesnt match with the 3 options
my sample rate is 22000 or something and the lowest is 32000 something
did i mess up
viral mason
locally, and the sample rate doesnt match with the 3 options
can't really help with local but for sample rate download audacity#✨│ai-help message
south verge
it also said it failed because there wasnt enough data
The file was WAV and it was 25 mins long
novel lotus
does anyone know how to make a person look really creepy with ai
royal gull
Is anyone proficient with Vonovox? I'm trying to run the program, while also playing Marvel Rivals, and it seems to cut my mic off over and over, or literally any other game, even Peak.
My current specs are 2070 Super, and AMD Ryzen 3700x, I mainly use Vonovox for vtubing. I just want it to work when I'm playing more high intensity games for streaming and offline as well.
south verge
I’m going to just keep trying until I get it I guess 😭
fathom mason
why doesnt the voice changer work in discord
torn junco
Can anyone help me use SDnext properly? My results are really bad I'm trying to upscale + enhance
smoky arrow
can't really help with local but for sample rate download audacityhttps://discor...
hi there, do you know how to make a song but using the voice models as the vocals
knotty moth
Is anyone proficient with Vonovox? I'm trying to run the program, while also pla...
you can try wokada deiteris fork instead if not sure
knotty moth
My current specs are 2070 Super, and AMD Ryzen 3700x, I mainly use Vonovox for v...
marvel rivals may be too demanding for your spec
RTX 3080/4070 is recommended for your typical use case
royal gull
Ohhh okay, usually Rivals runs at least at 90-100 fps usually for me. It's just so weird
knotty moth
Ohhh okay, usually Rivals runs at least at 90-100 fps usually for me. It's just ...
then the voice changer might be underperforming, try adjusting chunk & extra settings
royal gull
then the voice changer might be underperforming, try adjusting chunk & extra set...
Okay I will do that tomorrow, again thank you for the feedback! Also I LOVE your profile picture
leaden saffron
@low shard it still did it it always happens at 170
empty bolt
Hi, I was wondering if anyone could help me to fix my high res, it randomly starts and stops for not apparent reason, sometimes its high, and sometimes its normal.
I downloaded the cuda version since its for AMD if I'm not mistaken, and I have a 7900xtx which should be enough to not have high res.
I use the MaeASMR Model too
simple ore
Hi, I was wondering if anyone could help me to fix my high res, it randomly star...
downloaded the cuda version
of what?
why do you think 7900xtx supports CUDA?
empty bolt
>downloaded the cuda version
of what?
just realized looking at older posts having high res like me, the version I have is outdated
I followed a 2 months old tutorial that talks about old stuff
I can't send images here... can we go into a post so I can send an error I got trying to install the voice changer ?
lol I can't even create a post
empty bolt
Hi, I was wondering if anyone could help me to fix my high res, it randomly star...
and btw it was "ML-Cuda" not cuda alone
same or nah ?
simple ore
-rt
simple ore
read 1st link
empty bolt
read 1st link
ty
empty bolt
read 1st link
alr it fixed the high res, but how to avoid sizzle though ?
any settings you'd recommend ?
dusty mortar
hi, i tried to train a speech model with the freesia speech pretrain with spin embedder and I got only buzzing sound as output. what am i doing wrong?
simple ore
hi, i tried to train a speech model with the freesia speech pretrain with spin e...
you need to use spin embedder for inference
unless you did something even more stupid like using a pretrain as voice model
dusty mortar
unless you did something even more stupid like using a pretrain as voice model
did these settings for training the model https://imgur.com/a/nyJnoMd
empty bolt
@simple ore https://ibb.co/pBdhn9G0 Which settings should I change to make the voice sound better ? I have a good too
low shard
!howtoask
low shard
<@155030383648440320> https://ibb.co/pBdhn9G0 Which settings should I change to ...
!give-media-perms 50m @empty bolt
low shard
<@155030383648440320> https://ibb.co/pBdhn9G0 Which settings should I change to ...
by your model, are you trying to do e girl trolling?
empty bolt
by your model, are you trying to do e girl trolling?
I'm a trans woman....
I'm only on week 3 on estrogen doses so I still have a masculine voice and I'm trying to hide it, is it still catfishing ?
low shard
I'm a trans woman....
Ohh, it's great to help people like you!
f0: rmvpe without onnx
extra: 2.7
uninstall vb audio cable from windows app settings, get vac lite from the same guide you got wokada deiteris fork, then set output to line 1
uncheck sup1, check sup2 if u got noise issues
empty bolt
Okayyy tysm 😄
low shard
Okayyy tysm 😄
also since you're on amd, this triangle will be ur life saver
empty bolt
reallly lol ?
low shard
Okayyy tysm 😄
There's also optional settings, want help with them too?
empty bolt
what does it mean
empty bolt
There's also optional settings, want help with them too?
sure !
low shard
reallly lol ?
hover your mouse over it, it will tell you exactly what you need to do lol
low shard
sure !
on wokada deiteris fork, you can **optionally **use more advanced settings for benefits:
- Advanced Settings -> Force FP32 mode: on (THIS IS OFF BY DEFAULT! Turning this on improves stability. Increases VRAM usage by 200 MB)
- Advanced Settings -> Crossfade Lenght: Controls how smoothly the AI stitches different processed parts "chunks" of your voice back together. 0.1 for fastest voice, 0.15 for improved quality but increases delay by ~50 ms
- Reduce the delay on Windows via the Wasapi / Asio Guide
empty bolt
hover your mouse over it, it will tell you exactly what you need to do lol
ohh indeed that's pretty useful
low shard
that tutorial is outdated, it uses an over year old version of original wokada and vb audio cable
delete the zip and folder of original wokada
uninstall vb audio cable from windows app settings
Are you trying to do e girl trolling like in the video?
empty bolt
on wokada deiteris fork, you can **optionally **use more advanced settings for *...
thank youu that's really helpful !
low shard
thank youu that's really helpful !
you're welcome
, btw do you need any other help? I gave you image perms in case temporarely you need them
empty bolt
you're welcome
, btw do you need any other help? I gave you image perms in case ...
if I do I will let you know 😄
low shard
so delete everything basically ?
yes, forget video tutorials
no i am not
there are different programs based on different things, there isn't a program for everything and it also depends on the AMD version (and if your gpu is actually good enough to do what you need to do, since it's bare minimum), are you trying to do ai covers, or tts. or roleplay in games/vc?
low shard
if I do I will let you know 😄
alright, have a nice day
empty bolt
you're welcome
, btw do you need any other help? I gave you image perms in case ...
oh well I do, for the rmvpe settings without onnx, its grayed out and I can't click on it for some reason
low shard
oh well I do, for the rmvpe settings without onnx, its grayed out and I can't cl...
that was a typo, i meant with onnx since you're on AMD
onnx is meant for amd
what games? i fear you might not be able to bc ur gpu is pretty old
empty bolt
that was a typo, i meant *with* onnx since you're on AMD
okk makes sense, thx
low shard
ehhh, id not count too much on that, your gpu is the minimum
you can try on 1080/720p with 30/60fps cap tho
wanna try?
-realtime
low shard
1st link
wokada deiteris fork, yes
official is the one you're using
show an entire screenshot of your settings
dusty mortar
did these settings for training the model https://imgur.com/a/nyJnoMd
its weird my model works with tts inside the fork but not with vovonox live. i tested a model from this discord and that works with vovonox
low shard
dont crop anything
yeah to check which should be modified and which shouldn't in your case
low shard
did these settings for training the model https://imgur.com/a/nyJnoMd
that doesn't look like applio 3.2.9
also the quality kinda low
!give-media-perms 1h @dusty mortar
dusty mortar
applio didnt work with my 5090 so I used the rvc4 fork
low shard
applio didnt work with my 5090 so I used the rvc4 fork
that looks like codename's rvc4 fork, it's extremely experimental and not suggested for most users
applio does work with the rtx 50 serie
dusty mortar
thank you i will try it again
low shard
its weird my model works with tts inside the fork but not with vovonox live. i t...
this looks like the reason why it's not working
it would work only in codename's fork and applio
dusty mortar
ah i see
low shard
ah i see
you could either:
- use the applio rtx 50 serie fix, and use the spin custom embedder if you want to train spin models
- retry with codename's fork changing that setting, however the whole thing about the fork is being experimental and there is an high chance it can fuck things up for most general users, that's why it's not in the docs anymore and not suggested
dusty mortar
i try the fix first. i did that fix for an old applio version but I forgot about it
low shard
i try the fix first. i did that fix for an old applio version but I forgot about...
it does still work, and iirc in the next major applio release you might even not need to fix it urself as i heard noobies will use the newer pytorch from the start
low shard
i try the fix first. i did that fix for an old applio version but I forgot about...
if u want to train spin models, you can use the "custom" option
dusty mortar
where do i get the custom embedder? is it automatically downloaded?
low shard
where do i get the custom embedder? is it automatically downloaded?
you can either:
- use the rtx 50 serie fix precompiled guide, then download spin manually via selecting custom and getting the file from https://huggingface.co/IAHispano/Applio/tree/main/Resources/embedders/spin
- download applio main branch by git cloning, tho it needs git and would need to skip the precompiled step of the guide and just use the fix part (this will use the main branch which automatically downloads spin, but has some other experimental things and it's not as much stable)
next breach
you can either:
- use the rtx 50 serie fix precompiled guide, then download spin...
Hey why my voice is like lagging so much?
low shard
Hey why my voice is like lagging so much?
the message i sent is related to something else
and it's pretty hard to understand the issue since it's a general ai server
!howtoask
low shard
please provide the info needed for me to help you :D
next breach
Okay
timber merlin
i also need help, same issue as yesterday, my voice isn't being picked up no matter what i do, although yesterday when i switched to "server" it worked but today both client and server dont pick up my voice
everything seems fine, i installed the vb lite and its still not working (aside from the server miracle yesterday)
flint vapor
i also need help, same issue as yesterday, my voice isn't being picked up no mat...
Sometime you should change your default microphone on your browser settings, in your browser settings you have your def microphone?
timber merlin
oh okay
timber merlin
Sometime you should change your default microphone on your browser settings, in ...
yeah its the default i tried that too
i switched from quest's mic to my other mic, both set to default, and still the same
low shard
Okay
let me know!
gpu: rx580
extra: 1.0
be sure the game/discord input is line 1
on wokada deiteris fork, you can **optionally **use more advanced settings for benefits:
- Advanced Settings -> Force FP32 mode: on (THIS IS OFF BY DEFAULT! Turning this on improves stability. Increases VRAM usage by 200 MB)
- Advanced Settings -> Crossfade Lenght: Controls how smoothly the AI stitches different processed parts "chunks" of your voice back together. 0.1 for fastest voice, 0.15 for improved quality but increases delay by ~50 ms
- Reduce the delay on Windows via the Wasapi / Asio Guide
next breach
please provide the info needed for me to help you :D
GPU Name: AMD Radeon RX 5700 XT
CPU Name: Intel Core i5-7600 @ 3.50GHz
Operating System: Windows 10
Detailed Description: I was trying to set up AI voice and followed exactly what was shown in the tutorial video, but the audio still lags, glitches, and is delayed, and it also cuts out.
Tutorial Used: https://www.youtube.com/watch?v=SxdnGxicJOg&t=617s
Screenshot: its just bugging voice
low shard
GPU Name: AMD Radeon RX 5700 XT
CPU Name: Intel Core i5-7600 @ 3.50GHz
Operating...
that tutorial is outdated, it uses an over year old version of original wokada and vb audio cable could caause issues on windows
delete the zip and folder of original wokada
uninstall vb audio cable from windows app settings
Are you trying to do e girl trolling like in the video or are u just trans / doing rp?
next breach
I was try to do like finnish thing in rb
next breach
that tutorial is outdated, it uses an over year old version of original wokada a...
And where i can download newest one?
low shard
I was try to do like finnish thing in rb
ohh, roleplay?
next breach
ohh, roleplay?
yee'
next breach
ohh, roleplay?
Butt where i can download the newest:D
low shard
Butt where i can download the newest:D
-realtime
next breach
is there a like tutorial?
low shard
is there a like tutorial?
if you mean written guide, yeah the hyperlink is up there
if you mean video tutorial? nope, never trust those for realtime voice changers
low shard
ooh
they are always old since 2 years, never use them for rvc nor wokada
If you don't understand a step in the written guide, you can ask me 😁
low shard
gpu: rx580
extra: 1.0
be sure the game/discord input is line 1
on wokada deiter...
@buoyant warren any updates btw?
torn junco
How do I upscale + enhance using SDnext? I cant use sd1111 cuz I got an intel arc card sadly- can someone help me with this?
simple ore
How do I upscale + enhance using SDnext? I cant use sd1111 cuz I got an intel ar...
generate an image, send it to image tab (using image ->) button
on image tab re-select sampler, then expand resize tab and add multilier, and pick the resizing method
torn junco
No like- It's generating higher resolution image but taking away all the details instead of adding
leaden saffron
no mattter what i do i cant get the voice change to work both on presets and customs
simple ore
not 'rescale'
torn junco
what? I'm sorry im new to this
simple ore
torn junco
there is a scale output option in "resize" tab but I thought it's just gonna upscale without any details, that's why I used these
I will try your settings
low shard
no mattter what i do i cant get the voice change to work both on presets and cus...
I told you already to not use that version, delete the zip and folder
And uninstall vb audio cable from windows app settings
Use wokada deiteris fork instead please
leaden saffron
Use wokada deiteris fork instead please
the problem is it wont i ghave stable internet and it faikls at 170
low shard
the problem is it wont i ghave stable internet and it faikls at 170
At 170 what ?
Where are you from btw?
leaden saffron
At 170 what ?
170mb / 274mb
leaden saffron
Where are you from btw?
new york
next breach
If you don't understand a step in the written guide, you can ask me 😁
The latest version as of December 7th 2024 is: dml-b2332 (click here to download)
is this the right one?
leaden saffron
so its definetly not an internet issue
next breach
Okay
low shard
i have the start https but when i click it it doesnt open
Start http is from the old original wokada
Don't use video tutorials for realtime voice changers
low shard
i have the start https but when i click it it doesnt open
Delete the zip, folder
And uninstall vb audio cable from windows app settings
low shard
hey i need help
!howtoask
low shard
Please elaborate
low shard
170mb / 274mb
https://litter.catbox.moe/dry5y5ny486bi8ys.zip try this temporary link maybe?
It's weird that the HugginFace download server link doesn't work for you
But lmk if this works
You're welcome!
leaden saffron
https://litter.catbox.moe/dry5y5ny486bi8ys.zip try this temporary link maybe?
yeah something is defiently weird it still says check internet cinnection near the same mb even though i have stable internet
low shard
yeah something is defiently weird it still says check internet cinnection near t...
It's an issue on your end then, try either:
- use a different browser that is up to date
- use a download manager like FDM
- check if your storage is full, free it up a bit
- use a VPN (maybe?)
Because other people just downloaded it and even the HugginFace link was fine, I even tried both temporary one I just sent and that's fine too
leaden saffron
It's an issue on your end then, try either:
- use a different browser that is up...
do you think mediafire would work
low shard
do you think mediafire would work
I think it's in your end, if it was server side it wouldn't have just worked for me and the other people I just helped
It would be better if you try the things I told you
leaden saffron
everytime it fails it turn sinto this file type
low shard
everytime it fails it turn sinto this file type
Delete those files, try the things I told you please
torpid vortex
aight i got it but i think it doesnt work in discord
and i speak but only hear my voice after like 10 seconds
leaden saffron
after tring in chrome then edge i tried firefox an it finally wORKED
low shard
after tring in chrome then edge i tried firefox an it finally wORKED
Yup, it was on your end then
Well u can unzip it and continue with the guide then
low shard
aight i got it but i think it doesnt work in discord
!howtoask
low shard
and i speak but only hear my voice after like 10 seconds
I told you to not use that one, and that you need to elaborate, please do that #✨│ai-help message
leaden saffron
Yup, it was on your end then
alright thank you so much
low shard
alright thank you so much
Let me know how it goes
leaden saffron
Let me know how it goes
i should deleTYE THE OLD FOLER FOrmj the other one i wa using right?
oops accidental caps
low shard
i should deleTYE THE OLD FOLER FOrmj the other one i wa using right?
The old original wokada and vb audio cable? Yes
tiny ridge
Quick question function call in Ai context is just asking open ai question then based on it respond we excute some code then based on result of this code we ask new question and get final result ?
leaden saffron
@low shard
'''2025-08-14 09:44:09,438 INFO [WeightDownloader] Loading weights.
2025-08-14 09:44:09,932 INFO [Downloader] Verified pretrain/crepe_onnx_full.onnx
2025-08-14 09:44:09,936 INFO [Downloader] Verified pretrain/crepe_onnx_tiny.onnx
2025-08-14 09:44:09,989 INFO [Downloader] Verified pretrain/crepe_full.pth
2025-08-14 09:44:09,992 INFO [Downloader] Verified pretrain/crepe_tiny.pth
2025-08-14 09:44:10,310 INFO [Downloader] Verified pretrain/rmvpe.pt
2025-08-14 09:44:10,538 INFO [Downloader] Verified pretrain/fcpe.pt
2025-08-14 09:44:10,563 INFO [Downloader] Verified pretrain/fcpe.onnx
2025-08-14 09:44:11,150 ERROR [WeightDownloader] Failed to download or verify pretrain/content_vec_500.onnx
2025-08-14 09:44:11,150 ERROR [WeightDownloader] 'pretrain/content_vec_500.onnx failed to pass hash verification check. Got 99cbfd8c7be5b32af2d208d8569f5cdd, expected ab288ca5b540a4a15909a40edf875d1e'
NoneType: None
2025-08-14 09:44:11,151 ERROR [WeightDownloader] Failed to download or verify pretrain/rmvpe.onnx
2025-08-14 09:44:11,151 ERROR [WeightDownloader] 'pretrain/rmvpe.onnx failed to pass hash verification check. Got 9461af137f188837d82c21af61f3f570, expected 9c6d7712f84d487ae781b0d7435c269b'
NoneType: None
Traceback (most recent call last):
File "client.py", line 22, in <module>
File "asyncio\runners.py", line 194, in run
File "asyncio\runners.py", line 118, in run
File "asyncio\base_events.py", line 687, in run_until_complete
File "main.py", line 90, in main
File "downloader\WeightDownloader.py", line 88, in downloadWeight
Exceptions.PretrainDownloadException: 'Failed to download pretrain models.'
Press Enter to continue...'''
i down.oaded everythign now this is what shows whne i open it
low shard
<@911742715019001897>
'''2025-08-14 09:44:09,438 INFO [WeightDownloader] L...
your connection seems unstable, check https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#how-to-fix-failed-to-download-or-verify to fix it
hallow thistle
<@911742715019001897>
'''2025-08-14 09:44:09,438 INFO [WeightDownloader] L...
If you see this error while downloading "pretrain models" within W-Okada terminal, it indicates that your internet was slow and that it was timed out. Follow this guide section for solution. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#how-to-fix-failed-to-download-or-verify
leaden saffron
your connection seems unstable, check https://docs.aihub.gg/realtime-voice-chang...
thats the thing though everything downlaoded
all the downloads finished
hallow thistle
So is this solved?
leaden saffron
So is this solved?
im gonna try his thing he senthopefully al works
hallow thistle
Quick question function call in Ai context is just asking open ai question then ...
What is this supposed to mean?
prime scarab
what dose mean
Documents\Ai voice\dist\main>main.exe cui --https false --no_cui True
hallow thistle
what dose mean
Documents\Ai voice\dist\main>main.exe cui --https false --no_cui...
What program are you trying to run at this moment? Sounds like you're trying a realtime voice changer program.
!howtoask
prime scarab
What program are you trying to run at this moment? Sounds like you're trying a r...
yes i am trying to get the real time voice changer
hallow thistle
yes i am trying to get the real time voice changer
What do you use W-Okada the realtime voice changer for? Trolling your friends or catfishing? And what is your PC GPU?
prime scarab
What do you use W-Okada the realtime voice changer for? Trolling your friends or...
and i use the realtime voice changer client i use it to sound like youtubers to have fun whit my self and friends and my pc windos 10 and gpu i think is 2080 nvidia
hallow thistle
and i use the realtime voice changer client i use it to sound like youtubers to ...
Download the better W-Okada version from there and follow this guide. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
prime scarab
Download the better W-Okada version from there and follow this guide. https://do...
thank you :)
guestion where do i download it i cant find it for windows ;-;
hallow thistle
thank you :)
Let me know if you have finished installing and running the program.
There is for Windows.
prime scarab
thanks :)
brittle wing
Hi i was working on a small project its a simple Text to speech running in directml and requiers Onnx models, is converting a pth model into onnx possible?
hallow thistle
Some people wondering how I get one tab on Google Chrome to alive. I have a secret. I used an equalizer extension on a tab to keep it active even if I switched to another tab. The thing is an audio playback sound needs to be present first before you enable the extension on that tab, and then I switch to "Google Colab" link real quick.
hallow thistle
Hi i was working on a small project its a simple Text to speech running in direc...
Which program are you looking for? And what is your PC GPU?
!howtoask
brittle wing
Which program are you looking for? And what is your PC GPU?
I made my own program with some help from an ai agent, right now
am on windows 11
gpu : radeon 7900xtx
suprisingly my TTS works but it only has a en_US-lessac-medium.onnx model working for it so i was wondering if i can convert the models we have here into onnx
viral mason
hi there, do you know how to make a song but using the voice models as the voca...
I'm not entirely sure what to use for making covers besides weights.gg, u can ask a mod or helper for that
hallow thistle
hi there, do you know how to make a song but using the voice models as the voca...
Both "RVC voice model" and RVC program don't make the whole song track like Suno/Udio. Instead, in an RVC program, if you upload an audio file with a voice model applied, you click convert, the program would process and then becomes into what it known as "AI cover" or AI converted vocal.
What do you use RVC for? And what is your PC GPU? In case if you have one so there's a local option.
viral mason
Basically you're putting a voice overtop an existing voice of an already existing song
leaden saffron
@low shard i got it now im not sure how toactivate the voice changer i pres start and it doesnt changte my voice
hallow thistle
<@911742715019001897> i got it now im not sure how toactivate the voice changer ...
Since this specific AMD Radeon RX GPU is the mobile one, here's your settings:
Chunk: 128 ms, although you can increase this if your perf number at top left is red and audio is unstable
Extra: 2.7 s
GPU: AMD Radeon RX 6550M
On W-Okada, the green "start" button is where you activate the program to start working.
leaden saffron
Since this specific AMD Radeon RX GPU is the mobile one, here's your settings:
C...
i did that but the voice didnt change'
hallow thistle
i did that but the voice didnt change'
If it still not working, make sure to check your microphone and try and wait again.
What do you use W-Okada for? Trolling your friends or catfishing someone? And what is your PC GPU?
!howtoask
leaden saffron
If it still not working, make sure to check your microphone and try and wait aga...
stilll not working
hallow thistle
stilll not working
Have you restarted the program?
leaden saffron
Have you restarted the program?
yeah i just did
lofty parrot
Hi, I am new to this server and new to AI. with so many AI models or platforms out there I am trying to figure out which one might be best suited for what I want and I don’t want to waste my money taking on a subscription and have a turn out to be useless to me. Is there anyone here that might be able to have a conversation with me so that I can get the correct one picked out for me that will meet all my needs?
hallow thistle
Hi, I am new to this server and new to AI. with so many AI models or platforms ...
!howtoask
hallow thistle
May I ask, what AI program are you looking and use it for? AI cover or realtime voice changer?
lofty parrot
right now, I have been using the free plan for ChatGPT, but only for the conversation and research. I am looking to find a model that can help me do video, including deep fakes and also for static images and also for programming, and I don’t do any of those, so I’m looking for something that can be used from a conversational functionality
I know that ChatGPT on some of its various models one of them is not video and it also won’t do deep fakes, and I know that because I’ve asked it to describe the various models it has and what they can do and what they cannot do
hallow thistle
ChatGPT model itself doesn't do deepfake and other video related features. Dall-E and Sora, both from OpenAI, are known to generate image and video respectively. Other alternatives are also possible; some of which are free and open source but need you to set up by yourself, but also some are paid and easy to use.
lofty parrot
my situation is that I know little to nothing about AI and besides ChatGPT and mid journey those are the only names I know and I don’t even know what they do how they do it what their limitations are so I’m looking to educate myself on all of the various ones that are out there and figure out what they each do
lofty parrot
I am working on a couple of possibilities for doing a few different YouTube channels, and I will need both static images for the creation of memes and also to help me design things like background graphics or even frames for when I insert content. I will also be having some projects where I will need some programs to be created for me and I don’t do programming at all. I wouldn’t be wanting complete impersonation as far as deep fakes are I just would want it to be willing to create the images of the people I would be wanting to use for my memes both static and video and I know that a lot of AI programs have been created or programmed to not allow the usage of tons of public people
do you know the names of the programs or models that people have used to do deep fakes that have not been gutted to refuse to create the visual image of these public or famous people? Part of my problem I don’t even know the names of these various AI models or platforms.
simple ore
you dont need a deepfake to create a random person's picture
lofty parrot
it wouldn’t be just random people I would be wanting to use in my videos or my static memes. It would be political people and entertainment people.
perfect example of what I’m talking about is a while ago like last year I tried to get a Meem created on the concept of wonder woman both as the comic book type of drawing and also the Lynda Carter image along with people like Joe Biden, and other political figures being in circled in her golden Lasso, and instead of creating Joe Biden it kept creating Donald Trump
also, like stated previously, I only know the names and existence of two AI models or platforms like I said I only know ChatGPT and mid journey I don’t know the names of any others to even begin researching what they are and what they can do and what they cannot do
Hello
low shard
!howtoask
quartz dawn
wait @low shard can you give me an alternative to vb cable so im able to try it out?
low shard
wait <@911742715019001897> can you give me an alternative to vb cable so im able...
you'd need to change the over year old version of original wokada too, not just vb audio cable
it's better you elaborate as said in #✨│ai-help message
there's different versions depending on the gpu for example
lofty parrot
@simple ore did I lose you? Or did I say something wrong?
simple ore
i'm busy with other things
lofty parrot
oh, I just wasn’t sure if I lost you or something. You have to pardon me. I’m new to this server.
Uhhh what dis she say? I only speak English…LOL
next breach
What i have to do now and best settings for?
low shard
What i have to do now and best settings for?
high perf value, you're running something else other than the voice changer which can affect the settings
are you trying to do e girl trolling? or roleplay in games?
sleek ridge
Hi everyone.
I'm currently running a channel in which I'm narrating some webnovel series (currently only 1) And have been looking for a way to change my voice with an app for the girl characters... Been just trying to physically change my voice, it works... but... it hurts lol.
I see that some people showing images of their recordings on Audacity, so that already answered my first question.
My Current Desktop specs are
i5-9400F @2.90
Ram 32 gb
Nvidia RTX 2060, 6gb.
With all these versions, I'm wondering which one would be best suited for my needs.
I'd greatly appreciate the help.
Thankyou
dusty rune
steady pivot
how do i sett it up to fivem
opal garden
how do i fix the audio cable saying i cant redownload it
arctic cave
hello guys i have an issue where if i use ai Voice changer client demo and link it to discord using VB output and input in discord people hear my original voice,and then the changed voice and people also hear themselfves too so please can anyone help me and tell me how to fix this issue?
low shard
how do i sett it up to fivem
setup what? this is a general ai server, please elaborate
!howtoask
low shard
how do i fix the audio cable saying i cant redownload it
redownload it? could u please elaborate?
egirl/eboy trolling? or like spongebob?
low shard
hello guys i have an issue where if i use ai Voice changer client demo and link ...
dont use video tutorials, vb audio cable can give issues on windows, video use an outdated oriignal version of wokada
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
- what you want to do? There isn't a program that does everything, there's a program for each thing:
- AI Covers
- Train RVC Models
- E girl trolling
- TTS
- Roleplay in VC
- Roleplay in Games
- etc
- what tutorial link are you using
- a screenshot of the program
low shard
I forgot to say <@911742715019001897> I got it working!
thx for letting me know, and yw!
low shard
that is catfishing and illegal. You have been warned and won't get help.
leaden saffron
thx for letting me know, and yw!
what is the index thing for
pls dont catfish its just corny in general
be yourself and have fun w the voice vchanger
low shard
what is the index thing for
in rvc context:
- pth files: contain the voice
- added index files: contain the accent
- metadata.json file: it's just some extra info about the model download link if you downloaded it off weights.com, it's not needed and won't impact the actual model at all
TL;DR:
index = trained accent
crystal shoal
i dont get what epochs are and how i can set it
pastel oak
i dont get what epochs are and how i can set it
Epochs are set during the models training phase, it's the amount of cycles the training phase goes through. If you are using a model you don't have to bother with it, if you are training a model (which it doesn't sound like it since you'd immediately see the option), it's another thing
low shard
i dont get what epochs are and how i can set it
they don't mean quality btw
scarlet wedge
what does "zero gpu worker error" means in ilaria rvc space in huggingface?
what should i do
wheat pewter
Radeon 7800xt
my b sent too soon
radeon 7800xt, windows 11, strange tunnel effect present with the voice that randomly appears
ive been told its like speaking into a cup
small sorrel
how do you view the tensorboard in rvc v2 disconnected to see how many epochs you need?
*i need
low shard
what does "zero gpu worker error" means in ilaria rvc space in huggingface?
can you show a screenshot of the issue?
@viscid moss
low shard
radeon 7800xt, windows 11, strange tunnel effect present with the voice that ran...
can you also share a screenshot of the program, and the tutorial link?
low shard
how do you view the tensorboard in rvc v2 disconnected to see how many epochs yo...
rvcidsconnected is outdated #📰│dev-updates message
what's ur pc gpu?
small sorrel
i have a nvida gtx 1660 super
small sorrel
what's ur pc gpu?
yeah
royal gull
Can someone help me with an issue still? I am using Vonovox for Vtubing, but with games like Peak, and Marvel Rivals, the changer seems to "break" or glitch to where my mic cuts out with almost every word, just want to know how to fix it.
Currently operating:
RTX 2070 Super
AMD Ryzen 3700x
Mic used: Shure M7
simple ore
Can someone help me with an issue still? I am using Vonovox for Vtubing, but wit...
get a better/second gpu
tune game settings down so gpu is free
smoky arrow
you can change the sample rate with audacity to match 32k
hi i might need some help regarding how to make ai covers
low shard
i have a nvida gtx 1660 super
it can train but it's not suggested
Train (make) RVC Models on cloud:
- Prepare the Dataset
- Setup RVC:
Choose a cloud way to use RVC,
- Google Colabs (max 4 hours of daily T4 16gb gpu not granted for free, but easy to use, there's a paid tier):
- Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus, either T4x2 16gb each or P100 16gb, only free):
- Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly, Free Studios run 24/7 but require restart every 4 hours. There's a paid tier):
- Be sure to know about the tensorboard
Google Colab = Easier but risk of getting disconnected
Kaggle = Harder but more gpu time
If you want the easiest way and for free, is using https://weights.com/ which uses RVC
RVC Inference (use models) on pre-recorded audio on Cloud
You can use either:
- Weights.com: Easiest Possible Ever Automatic
- Ilaria RVC Zero: Fast and free on cloud
- Applio UI Colab: RVC Fork with some extra features like TTS
- RVC AI Cover Maker UI: Automatically Separates the Vocals and Instrumentals, converts the voice and mixes them back
better to use cloud
low shard
hi i might need some help regarding how to make ai covers
Sure!
Could you tell your pc gpu? I'm guessing you're on windows
smoky arrow
rtx 4080 and yes im on windows
so ive read the docs, there are 2, how to make ai covers and ai cover makers
are they differents?
so rn im trying the how to make ai cover
so after i convert the music to vocal and instrument, then change some settings on audicity
need to convert it on voiceconversion
after i read the instruction, after i convert it, below it there is training instrustions
after the download output
low shard
so ive read the docs, there are 2, how to make ai covers and ai cover makers
the how to make ai cover is an essential page explaining how to make it yourself
the aicovermaker fork is an automatic process which simplifies the whole thing
id suggest the 2nd if u want only ai covers