Me and a friend wanna make a prank we saw on tiktok on another friend, although we installed the stuff we found on youtube tutorials she told me there was not an rvc model for the character we are looking for... so im guessing i have to train it, or pay someone to train it for me. Eather way its compleatly fine i just need to know how to get it done! So far i only got some audios that som1 told me i needed and thats it... can someone lend me a hand on this? we'd like to have it done before next friday if posible so im literally taking any advice :c
PLEASE and ty
#Idk how to train a rvc model...
1 messages · Page 1 of 1 (latest)
we found on youtube tutorials
here's the issue
every single youtube tutorial about RVC is outdated asf
whatever you installed, like mangio rvc, just delete it, it's as good as using windows xp in 2025
^
no one makes updated video tutorials since RVC trend died
or pay someone to train it for me
That's not allowed, we don't endorse any paid commissions anymore, don't trust to get a paid model from anyone on this server
Elaborate:
- your pc gpu
- your operative system
AI is a complex and intensive task, it needs a good pc
noticed that
ouh... okay
i think my pc is good enough? it has a 4050 i think and a ryzen5, idk much about computers but the ai i made the test with didnt explode my house...
yep, the trend died and AI changes fast
You can check your pc gpu on Windows via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU
so should we deleate what we already have?
It''s better u check
absolutely, forget everything they teached you like crepe or whatever
do you have any other GPU?
the other one says NVIDIA Geforce rtx 4050
AMD radeon graphics is just the integrated graphics of your CPU, which is pretty weak
could you check it's dedicated memory? is it like 6gb?
show a screenshot of your task manager
also, before making that model
You can search rvc ai voice models at:
- #1175430844685484042
- In #🔍│find-models , Do /find with @simple sage
- https://weights.com/ (login required)
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- Suggested Models for Realtime Voice Changing (Wokada)
- https://voice-models.com/
- https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)
if there isnt one, you can:
- make it yourself with our docs guides
- Be aware that we don't allow any model request, so don't fall for any "pay me 20 dollars and i will make the model for you" dm
:wave: @harsh echo, How can I help?
Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image
scroll down
oh nvm
lmao
@silent tide ould you please check if the model exists?
search in every site I gave you
be sure to use other common names that character might be called
You can train locally, but not having at least 8gb vram makes you disadvantages because you can't train it on batch size 8, which is needed for training any models with 30 or more minutes of dataset
you can still use it, you can inference and train locally, just not on 8 batch size
what's your friends pc gpu?
also its way better you check before trying to make it urself
got it
oooooh got it
ahh my shes got a pc that comes to the moon and back
i dont know the specs by memory but i do know it is the top of the line
It's way better you tell me her pc gpu, last time someone told me something like that, they had a gtx 1070 (released like 9 years ago)
lmao
yeah, i can ask her!
alr let me know
let me also know if you find that someone already made that model, would realy make your life easier
since making models can take a lot of hours
RVC and wokada aren't 1 click programs at all, Open Source AI is pretty complex
not so far
yeah...
be ready for a lot of tech nerdy stuff then, you will have to clean the dataset, and train it yourself (which that laone can take a lot of time), and monitor how it goes with the tensorboard
should we wait maybe for when you tell your friend's pc gpu?
yeahh! we should indeed cuz that does sound like exploting my humble laptop
still, im looking for it
kinda having a hard time couse i forgot the name of it
im just waiting for her to answer
maybe try reverse searching the character if you got a picture
or give hints to chatgpt maybe he can guess it
in the mean time, what program is it that we need to have installed? does it make a difference?
great idea
so how does it work... becouse last time it took us a 30 minute video and a lot of clicking
is it gonna take another 30 mins
found the rvc!
thanks for the links
it will take more than that to train a model, can take hours to actually finish a model
or do you mean just reading the guide?
5 mins lmao
oh! thank god
prob even easier than what you see on youtube
we were using
since they use 2023 outdated software
lemme guess, mangio fork?
MMVC something
idk what that is
i just got the tutorial and made everything it sed
MMVCServerSIO? that's not even for training models.. that's just the codename of Wokada, a program for realtime voice changing..
RVC = Retrieval-based-Voice-Conversion, the best Few Shots Speech To Speech AI Models (on v2), Inferences (use models) pre-recorded audio (ai covers) and train (make) models. Technically, Mainline RVC does have a go-realtime.bat (aka RVC-GUI), but it's pretty messy and outdated so it's extremely not suggested for realtime.
Wokada = uses RVC for realtime inference. There's 2 main versions, Original made by Wok, and the most suggested one is Deiteris Fork (modified version)
There's different programs depending on what you want to do
oh okay
do you want realtime or train a model?
Applio (RVC Fork) is good for that yes
to do that, you'd need Wokada Deiteris Fork
youtube tuts use an over year old version of original wokada with shitty quality and performance
-realtime
💻 Local Realtime RVC
Guides for Programs that use RVC Models in Realtime for Calls/Games
• Wokada Deiteris Fork
Most suggested. GUIDE
• Original Wokada
ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE
• **RVC GUI Mainline Realtime**
Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated
1st link, you gotta read this one too
yes
for training models and inference (use models) on pre-recorded audios
okay, okay
im sorry got lost reading the doc
yes we do have the cable thingy
but i did not got where am i supposed to download the newer version
vb audio cable?
that one is outdated too.. it causes issues on windows
please just forget you even searched for rvc and wokada on youtube 😭
ouhh
all info you got is off 2023 lmao
what kinda issues...
randomly stops working out of no where, its unstable
many users reported that on windows
yeah it might work fine, but then randomly not work the next day, its like playing russian roulette
oh! i see...
just uninstall it from windows app settings and get vac lite instead https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/#virtual-audio-cable
Last update: May 5, 2025
yeah we got the vb audio cable
guessing both you and your friend are on Nvidia GPUS: https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
Last update: May 5, 2025
no not vb audio cable, uninstall it, get vac lite
vb audio cable causes issues as i told you
the vac lite i gave you works better
from the windows app settings, like any other installed program on ur pc
Last update: May 5, 2025
ty ty