#✨│ai-help
1 messages · Page 66 of 1
it's rvc disconnected 😭
also the lowest value is the best point right
am i correct?
ye
thanks!, any clue how should i extract that point on rvc disconnected 😥
as i already said - no clue
OK, look at /content/Mangio-RVC-Fork/weights
Using the file explorer at the sidebar
hey so i managed to install 12.1 cuda from lastnight but the problem still persists, i even have the right pytorch and everything, all required dependencies and i tried reinstalling about 12 times over the sum of a week
i always get the same error
return F.batch_norm(
File "C:\Users\chrus\Desktop\Applio-RVC-Fork\env\lib\site-packages\torch\nn\functional.py", line 2478, in batch_norm
return torch.batch_norm(
RuntimeError: Expected tensor for argument #1 'input' to have the same type as tensor for argument #2 'weight'; but type torch.cuda.FloatTensor does not equal torch.cuda.HalfTensor (while checking arguments for cudnn_batch_norm)
alright, I think I've found an actual solution:
- Download this file, put it in Applio directory https://github.com/IAHispano/Applio/commit/b6371eda4e3b4a6055ed48e9f39cb68339de842e.patch
which folder
Your applio folder
OK, so @royal jay how do you download Applio normally? zip file, or git clone?
i tried doing it manually, windows installer, exe installer, zip file, glit cloned, github desktop, tried different browsers
Ayo? @royal jay level 1 !!! 
hmm, maybe try downloading this branch https://github.com/sonphantrung/Applio/tree/rmvpe-fix
which directory?
-rvc
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
I just wanted to try
okay i'll see if it works, then if it doesnt ill check the rvmpe_gui fix that guy did thanks
@forest sluicebtw after cloning, what to do after system cannot find path specified when opening go.bat
Hmm, now that I think what I'm going to tell you may be too complicated, so I recommend deleting the folder, then download this. It's just full Applio, but fixed (I hope so).
okay sounds good, after downloading it automatically says after opening go.bat
im thinking i have to set my env and reinstall dependencies and such off the rip
@forest sluicealso btw should you have miniconda installed?
i deleted it before trying any of this because i thought it was causing malfunctions but apparently it is indeed needed
yes
see idk how to get it all to work, because it says i need miniconda installed but it interferes with my python
install and use runtime rather than conda, unless ur using AMD GPU I have no idea 
so after cloning, install nvidia runtime and i should be good?
it should work
Or, run go-applio-manager.bat
unless you've actually run that script...
yeah that's what i did i was tryna reinstall dependencies
even after it says the system cannot find the path specified, im just gonna let it launch and see if itll work with the rvmpe_fix fork u sent me
Ayo? @royal jay level 2 !!! 
thank you so much for your time and responses
this thing has been annoying me for a week straight but i still haven't given up
tried reinstalling windows and everything
Unfortunately, it still does it, so i'll try pasting the file in the directory now
AND after pasting the patch u sent me in the directory, it still persists
im thinking i may have to include it in addition to some code or some sort
im going to try the branch u sent me
but i have forked it instead so now its doing this again, but however ill see if it fixes it
even after applying this guys fix, it still does this but with a changed code
@forest sluicenot sure if you're still here, but just tagging in case u can still help by any chance
EGIRL AND OTHER GIRL VOICE MODELS HERE!! 
E-Girl, Mommy, V-Tuber, Anime girl and much more!
Wondering what models?
You find them here: https://ko-fi.com/ilariaowo/shop 
INCLUDING FREE ONES!! 
I also do commissions if you want aspecific voice! 
I accept the following payment methods:
- Ko-Fi

- PayPal

- Revolut

- Steam Gift Cards

- Amazon Gift Cards

- Throne

Grab yours today!!! 
Do you accept robux 
how to use rvc v2 in real time?
@royal jay Hey! I found the source of the error, here's how to fix:
- Go to
lib/modules/train/extract/extract_f0_rmvpe.py - Go to line 20
- replace
is_half = sys.argv[5]withis_half = bool(sys.argv[5])
@rigid light
RVC Guides (How to Make AI Cover)
Translation by country
what should i choose here
I saw, thank you
32k
how do i get rmvpe its on onnx ? it says n/a
What's your GPU ?
if it's an AMD/Intel one, yes your choices are going to be reduced to just those that can actually run on those GPUs
yeah im on amd
jpw dp o convert it?
the voice is like choppy
did you "export to onnx" an re-import the .onnx file ?
I have found 1 results for the search kuplinov...
hi
each time i'm trying to download a model it says "success" but when i go to choose it in inference it just isn't here
did anyone else have this problem before
Ayo? @vague haven level 5 !!! 
hey guys
if you post your problems in #1192011222023950368 you're going to get responded to a lot quicker
well idk whats wrong with man lag a lot when i open it
idk what to do
am downgliding it again
MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.15.zip that one right?
That should be fine, just keep in mind you have to export your downloaded models to onnx
Since you're using the AMD version
i mean the thing lag a lot
cant do thing
on downloaded models this happens cuz its using the cpu instead of the gpu - so to use the gpu and make the thing more responsive you need to export to onnx
no not the models the app
they should check Ilaria's models lol
Ye
either way, if ya do these things it might work
what the best one to download i mena the app
What's your GPU
Ayo? @barren acorn level 2 !!! 
i have gtx 1650 4gb, have reccomendation?
i have gtx 1660 6gb
yeah u got the wrong version my dude
this is the right one
read the https://rentry.co/VoiceChangerGuide too
In llaria rvc, whenever i put my model and the audio file i want to make an AI cover, and when i try to click "convert", it just keeps processing, example:
is there a way to fix this?
@proper shale what app do downloads for rtx3070ti 64gb
Yeah this takes a while, wait for it
alright
Ayo? @vast kraken level 2 !!! 
The RTX 3070 Ti does not have 64GBs... but oh well, read the guide I sent you
It has already reached "516" and it didnt even finish.
it just keeps going
it cant stop
It takes a while because it's inferencing on the CPU.
whenever i put in a model then refresh the model doesnt show up
was the download link a google drive link
yup
it said success then i refreshed by clicking the refresh button to select the model and it aint their
where are those? and how do i find a model im looking for
Ayo? @shadow hinge level 1 !!! 
prolly in #1175430844685484042 , if you wanna search u can try using the bots in #🔍│find-models
the ilaria RVC Huggingface doesn't work. it keeps on timing out 😦 any ideas/suggestions/fixes?
in what way
Did you duplicate the space
I myself dont use ila rvc so 
i think i did, downloaded it, refreshed and selected it
could it be a problem with the file name?
or with the extension? maybe i have to use an mp3?
Ayo? @edgy sky level 2 !!! 
don't believe that's the problem
not this either
I downloaded the model, what i gotta do next?
huh i really dont know what could be causing this
-rvc
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
same i dont know why that would happen 😭
should i try changing the pitch extracting algorythm?
Maybe try reduplicating the space?
nah that aint it
i'll try with mangio-crepe, let's see what happens
check the logs too
doesn't mention any error
Wtf is causing the error lmfao
i'll try on another browser maybe?
Yeah no errors here... hm.
What browser r u currently using
I've seen people make AI covers where the AI seemingly ignores the background vocals, or covers that are a duet of two singers. How exactly does that work?
firefox
It extracts the back vocals using a uvr model
Try Chrome just for this
now on Chrome it doesn't even let me download the model ☠️
The process of ai covers is like song > uvr to get vocals and instruments > uvr to get back vocals and main > rvc inference > smack them back together
-
You'd extract the main vocals using a Karaoke vocal extractor, convert that
-
You'd convert all the vocals with the two different models and just trim the parts that are for each singer
does the space have to be public?
Nope not really
Oh alright
Model's already downloaded tho
So u can click refresh
mix manually, with instrumental stem, reverb & effects, etc
Thanks you two
You're welcome :)
oh ok let me try
I tried to mix it in the pun
still nothing on Chrome 😦
So what's the best current colab for pro users to train a model using a 23 minute completely dry hq vocals dataset & how many epochs should the model be?
You can try using the Applio Colab if you have Colab Pro.
https://colab.research.google.com/drive/157pUQep6txJOYModYFqvz_5OJajeh7Ii
I'd recommend going for 500 epochs and checking the graphs for every model.
Hey! how can i create a ai cover with my friend voice?
You'd have to gather audios of your friend speaking, train a model, then do the AI cover
Alrdy have speaking because he have a yt channel 😄 how can i train a model?
Check the #1159513888199540817, specifically Create a model with RVC Disconnected
Oh okey
Is there any to properly install voice models from weights.gg I always have this issue where I install them but when I open the zip it just says “model.pth” “model.index” but when I install them on the gui the index doesn’t even appear
Ayo? @sturdy sleet level 1 !!! 
Thanks, got a nice easy to follow guide to help those who're new to training models?
Oh nevermind apparently Applio's Colab was deleted 😭
You should try RVC Disconnected then: https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS
Here's the guide for it:
https://docs.google.com/document/d/1XuxQYiqEhYrdYeCZRRLrmV_ciMKo0bV-jTCGHu_-5Cc/edit?usp=drivesdk
If you have any issues lmk
how do i use this?
📚 AI HUB Docs
https://docs.aihub.wtf/v/spain
❔ Perguntas Frecuentes
https://docs.aihub.wtf/v/spain/faq
🔗 Servidor IA Hispano
• Guía de Uso del Colab AICoverGen
Link: Google Docs
Créditos: Eddy (Spanish Helper)
• Cómo Hacer un Cover de IA Usando Ilaria RVC
Link: Rentry
Créditos: Julia (ailen2091)
RVC Guides (How to Make AI Cover)
Translation by country
Insert your HuggingFace token at the "hftoken:" , then insert the last one like this: Username/ModelNameHere
Then run the cell
And tick the "create_new_repo" option.
i just downloaded the json, index, pth, put in .zip folder, add new model, put the .zip into the hugginface model and then done
@molten pecan como haré modelo rvc para que no tenga tanto peso de archivo gb en drive
Qué archivos pusiste en el .zip?
Solo necesitas 2 en total. El .pth final y el .index llamado "added_"
Ayo? @blissful pecan level 2 !!! 
@molten pecan
Please post issues in #1192011222023950368
Okey
https://docs.applio.org/installation/requirements/ i really have to install all this shit?
how do I know if my model is rvc v1 or rvc v2? I made it in august 2023 and I can't remember
nvm I can't remember half of the stuff I need to enter to get access
guess I can't upload this model
Can you send a screenshot of the files inside your model's .zip file?
I know which are v1 or v2
I don't have the rights to post images
Ayo? @thin plume level 1 !!! 
v2
okay, but I also don't remember the epochs, which training method I used...
which sucks, wish I wrote that down somewhere
Do you have the training files somewhere
yes
Do you remember if you trained until the total training epochs number?
If so, you can check the .json files and see what pitch extraction and the total epoch settings were
the file date seems to be later than 70 epochs, but earlier than 80 epochs so I'm guessing 70 epochs? it reached the desired level very quickly and everything after sounded worse
oh I do have a json file
Possibly 75 epochs? Hmm
If your saving frequency was every 5 epochs that's a possibility
It seems to have saved after every 10
Hi, I'm training a model with a three minute long dataset with OV2 Super on 200 epochs. I was wondering if the valleys on the graph were mode collapses, and, if they are, how I could take steps toward remedying that. My current batch size is eight.
What does "Saving every epochs" mean?
Yeah they are mode collapses
Saving frequency
So you save every x epochs
wait I found it! I used rmvpe 😄
rmvpe for life 🗣️
@ionic fossil I guess try testing at a lower batch size (4) and see how that works out
Else it's most likely your dataset
Thank you for the help. If it ends up being my dataset, what in a dataset would cause the mode collapse?
Either badly trimmed silences or bad quality data
lol I have no idea I was probably just following a tutorial
XD
lol
By the way, everyone, please use #1192011222023950368 for your issues - this allows new users to search for their issues easily.
Disabled talking here so that we can test out the new format. ^
Help me
Why do you think I’m here?
nuh uh
dont ruin help-rvc my beloved
-colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
RVC Guides (How to Make AI Cover)
Translation by country
Does anyone have the images or tutorial for hugging face because I don't remember if I was supposed to mark it or not.
Mark what?
I don't even remember because I already made the model but the second option was the first one before uploading the dataset.
I do not remenber the name.
Can Someone help me with a hard issue i have with the install of this AIcoverGen???
I will give more details and images in my DM
PLS HELP
hey guys, what tts do yall recommend?
Hihi fam, wanna ask is there a guide that allow us to do inferencing locally with RVC? the guide from the FAQ doesn't work for me tho.
I need to ask a question, I have already used my gpu usage limit, is it possible to use easygui without connecting the gpu and it works normally?
-rvc
Suggestion for @rough lichen
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
No. If you want to train your model you can't!
if you only infer, sure, go rmvpe. Otherwise, don't use CPU-only for training
Ayo? @forest sluice level 24 !!! 
So I'm using hugging face, right? I did what it said and refreshed it so I could choose the model. But when I click on it, nothing happens. I've refreshed it several times and tried two different models. What am I doing wrong?
Oh wait LMFAO I figured it out
Ayo? @brittle wing level 2 !!! 
Never mind I don't know what I'm doing 😭😭
RVC Guides (How to Make AI Cover)
Translation by country
Hello everyone
So first try did um...poorly. Does anyone have any tips?
It's supposed to be Kyojuro Rengoku but it doesn't sound like him at all for one and two its like...terrible quality
Hi so I'm using Mangio RVC and whenever I try to do anything everything just gets stuck in a queue that says 1/1 for all of eternity
If anyone could help I'd appreciate it🙏
I think it is fixed in #1192011222023950368
hello guys,can i get a help with something?
Yes you can
My voice just breaks in discord idk why...can someone tell me the solution?
Idk why this is happening! You can ask in #1192011222023950368
ok,one more thing...why cant i use crepe or harvest or any other mode other than rmvpe_onnx? its showing N/A beside those options
Ayo? @merry river level 1 !!! 
excuse me, im trying to use rubick voice while playing dota but seems like its not working. anyone knows the settings for dota in-game?
total newbie here, where on earth do I even begin to start learning what tools exist and what they can do?
Yes! Ask your quarries
11labs
welcome
then I guess I'll start with: what tools exist and what can they do?
a project I'm persuing is a grox cosplay that will involve some voice changing on the fly
so I'll be making a voice model and doing voice changing, as real time as possible but understandable if delay comes with the territory
Ayo? @trim pagoda level 1 !!! 
no it usually depends on the settings and on ur pc build
rtx 3080, ryzen 5900x
u can both train locally and online
well RVC is the main tool u can use the fork "applio" that i mainly suggest for local training
oh, I should specify, free, local tools
can that do more or less realtime voice changing?
all of our tools are free
thats just for training
for voice changing u can use okada
ah, I see
brilliant, thank you!
feel free to ping me for any other questions
ok so I managed to get rid of the queue thing but now when I try to convert it says "error" and in the cmd thing it says AttributeError: 'NoneType' object has no attribute 'tobytes'
@copper pebble is it ok if I also ping you for this I'm losinf my mind
u doing inference right ?
ya
Ayo? @vital maple level 1 !!! 
u sure u putted the right file path?
yes
do not put the path directly in the audio but just in its folder
what😭
how did u set the audio path??
you mean the audio I'm trying to convert?
I just put it in the audio folder
is it not meant to be there?
whats the name of the folder?
audios
it's m4a
try to convert it into wav
mhhh
what happen
this thing
is that the easy colab?
can u send me the link
cuz im pretty sure it is outdated and its not working anymore
oh no its the hina
well try to delete the runtime and redo the cells
if it doesnt work its maybe cuz its actually having problems
i did that 2 times
well its probably having problems
@winter seal i'd suggest to wait for the problem to be fixed or use ilaria rvc or the eddy's one
is it normal for the outdated google collab to work wonders and the newer one saying it has 30k ms 
which one do you mean by newer
the older one is banned, giving a "Your runtime has been disconnected due to executing code that is disallowed in our free-of-charge tier." and ive always been redirected to smth else in #🔍│help-w-okada . the older, banned, is pretty fine, maybe about 5 ms. the second one, almost always though, ends up piling like hell in the ms and i cant figure out why
yh i got that but can u send me the link of boths
1 that bans me: https://colab.research.google.com/github/w-okada/voice-changer/blob/master/Realtime_Voice_Changer_on_Colab.ipynb
1 that loves 30k ms: https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb
(and i am now just realizing, perhaps i am in the wrong channel)
Ayo? @quick tendon level 1 !!! 
well ure not the first one saying that
and yes
maybe in the other channel u can find somebody that knows smt more
abt it
my very bad, ill move my stuff to the proper channel 🫡
np
ty
you make a dataset, basically a voice recording of you voice without any background noise
then you train it with an RVC software
Ah okay, so i record my voice and is there a link to a RVC software to train it?
Ayo? @brisk gull level 1 !!! 
well yeah you can both train locally and online
it depends, if ya got a good gpu u can train locally otherwise go for the online methods
Amazing, thanks a lot Rickie. Is there a link to the online method?
Sorry for all these questions
dw heres the guide for the online method u can find the link inside of it
oh ye weird i didnt touch the code for the app itself
maybe u also want to read this a bit https://rentry.org/RVC_making-models
THIS WAS ALL TESTED ON THE Mangio-RVC-Fork WITH THE F-RVC-exp CONTINUATION FORK OF Mangio-RVC-Fork
(From what I know, each fork logs differently for TensorBoard)
Things you may need for this guide:
Audacity
iZotope RX Audio Editor
Adobe Audition
Spek
TL;DR is at the bottom of the page
Mangio logs...
what could it be?
Is it joever?
no clue lemme try loading my old applio colab if it also does the same thing
lemme try ngrok first
RVC Guides (How to Make AI Cover)
Translation by country
ngrok still works
gradio prob
When training of model is complete i need to grab logs from rvcLogs.zip , right? in collab
can anyone help with me creating a AOI TODO voice
I already have a clean voice clip can anyone create a link for me please
kek gradio works again
its a hard coded fix i made but at least it work.. for now
k champ
can anyone help Me?
lemme go add it
can ya test it real quick? i added it on the notebook already
sure
whats up
yo
Ayo? @muted star level 1 !!! 
I'm trying to create an AOI TODO AI voice link
but my mango thing is bugging
so I was hoping someone can help create it with me
I already know how to do it
@copper pebble
whats the bug
and I have a clean voice clip
I procees the data for it to not show anything on the output
can u send me a screenshot pls
convert it to wav
mew works?
you didnt use fallback mode..
hi
makes it waaay faaster to install
yh i know
why can't I send a picture?
ig 2nd cell just takes forever
need to be lvl1 or 5 i forgot
is going down, but there are sudden dips and I would like to show them to you
Ayo? @slow aspen level 1 !!! 
send a screen here
ok so it's not MODE COLLAPSE? xd
I'm new so I don't know if this is Mode Collapse
same
:<
not a mode collapse... or at least it doesn't seem so
wait actually it might be
oh
ehhhhh
@copper pebble yo
can someone just creat a link or something
@proper shale could u pls help him im going for a break
could be batch size tho, since g is looking rather flat
ty
ill try my man
tysm
hmm
Create a link for what exactly
I will wait and see what happens next
⠀
HuggingFace Spaces 🤗
⠀
Ilaria RVC
EasyGUI port with some improvements, by Ilaria.
RVC-HFv2
Applio port, by r3gm.
AICoverGen
AICoverGen port, by r3gm.
Advanced RVC Inference
Extended version of the GUI with advanced settings, r3gm.
⠀
No one will, sharing public links are very risky, just use the ones above
Ayo? @slow aspen level 2 !!! 
Yeah
Click the third button in the bottom left of the pic
To center the graph
I have trained up to 300 epochs, is there any point in continuing to train?
Stop training until you are very sure you are overtraining
oh okey thaks :p
When you suspect it, lower the smoothing value a bit (not lower than 0.7) and place your mouse over the lowest point in the graph
It will tell you the step number
Then use the .pth that was closest to before the overtraining point
You got it?
for this?
i got a model link from weights.gg and wheni paste it on rvc it says that its been an error
what do i do?
like
English guide
Link: https://rentry.org/fdg_guide_newer
Credits: FDG
Spanish Guide
Link: https://rentry.org/AI_Hub_guide_esp
Credits: Julia (ailen2091)
Hindi guide
Link: https://rentry.org/AI_hub_guide_hi
Credits: Enes
Turkish guide
Link: https://rentry.org/AI_hub_guide_tr
Credits: Enes
Yeah
The .pth will be automatically saved with this format
modelname_epochnumber_stepnumber.pth
Example: julia_e120_s340.pth
Thx ❤️
So if the lowest point was 350, find the .pth closest to before it. Like 345 or 340
Don't worry if its not exactly the lowest point
You won't hear any difference quality-wise
Does anyone have the TTS link?
Hey someone knows about cluster computing if yes kindly DM me i have to ask something!
Suggestion for @echo meadow
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
So I used hugging face with my model but it sounded nothing like the character I was trying to make and it sounded like... really bad too. Does anyone have any tips for beginners?
I can help you!
Did you duplicate the space beforehand?
How long was the sample?
Oh that might be my problem. Should I cut the song up into smaller samples? That's my bad
Yeah Rickie sent you an outdated guide it seems. here's the new one along with much more info and error fixes
Thank you so much!!!
Ping me if anything
Guys, look Ai kits has changed a bit
@half cove applio update looks fire 🔥
Back down where?
cuz its still not overtrained 
Oh my fucking god
Can you
Start from scratch but with a 20 sec dataset?
ok 😭
Thank you
RVC Guides (How to Make AI Cover)
Translation by country
how do i like make my voice as a model?
Hey des anyone have a solution why i cant open Tensorboard? everytime i try to open it i get the error "The system cant find the file " ive tried everything that i found online, ive even read the guide that was linked to install tensorboard. Sadly the files the person linked in the guide for the tensorboard are no longe online for download
did you install tensor board
i have problem that like google colab have a thing code but idk how to do it
wait nvm
yes i did, i even uninstalled it and installed it again.
did you run that in your RVC folder
yeah i did.
yes i did and become this êrror "Fatal error in launcher: Unable to create process" system cant find mentioned file
I have no idea
what version of rvc r u using
like mangio, applio ect
bro when im trying to convert my song it just kept error
Ayo? @sick canopy level 1 !!! 
I use Mangio Rvc
try put this in your rvc folder and run
thank you so much it worked now


yo can someone actually help me out
please
i want to deploy my rvc model and turn it into an api
Hey guys its my first day doing this. If I'm creating a dataset do I have to create a bunch of .wav files all under 10 seconds? Or can I just have a singular 10 minute long .wav file of someone talking
one file works fine
rvc cuts it into multiple anyway
okok perfect man thanks
why does my ai sometimes not pickup my voice and when it does its choppy and laggy?
why does it make my voice echo?
Ayo? @astral rune level 1 !!! 
you can have a singular .wav
u using okada?
can someone help me with some errors ive been getting when trying to change a voice, i get this message: Warning
You may be executing code that is disallowed which may terminate your runtime without warning. Colab prioritizes interactive notebook compute and disallows some types of usage when executing code without compute units as outlined in the FAQ.
im using this guys tutorial https://youtu.be/gtjCmJ_6d0k?si=qkDbCqb0M23Agc3K
Okay so I came accross an error during training and most people online are saying you can no longer train online and must do it locally. I have a 3070 ti so I should be fine to run it locally but im not sure where the install is. Is it somewhere in this discord?
you came? 
Ayo? @warm notch level 1 !!! 
-local
Suggestions for @warm notch
- Applio-RVC-Fork, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team GitHub
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
download original rvc i think
ok let me send you the guide rq
dont say sorry for asking help in help-chat 😭
i need some help, i want do the easiest way to implement rcv on a python code, i just need to get an audio.mp3 file, use a model.pth and output a audio.mp3, using python. i dont want to use easy gui because i will automate and run locally a conversor. can someone help me do this please?
i already searched and tried this whole afternoon to setup rvc pipeline and i couldnt.
(ab)use the command line they have
I'm like 90% sure they just invoke a python script via subprocess & a bunch of args
Or maybe it's for training only ?
i was thinking about this but i saw the code and im just confused
i tried following this tutorial
https://youtu.be/MckT7z7W_qM?t=200
but when i download the rvc as a packet to use its functions it download as a archive on my project
Ayo? @grave socket level 1 !!! 
the links to download the bat files don't work :(
hey i need help
thanks
@half cove is there a reason it keeps spamming the same thing after installation and nothing opens
it shouldnt 
what does it say
it's still doing the same thing
still spamming the same thing
Idk what to do
and now
everytime I launch tensorboard
for anything
it spams that
and Idk why
lemme try something
nvm
keep downloading google links and they wont show up no mater how many times i reload this acursed thing
was that for me or the guy above me?
pardon
oh what the fuck
but ive tried links from this discord too
where do i find a frank sinatra ai?
i cant even tell if that was a sarcastic awnser, im so confused by what to do
blanc?
are you still here?
im fucking losing it man
ive been taking models from there too....
Ayo? @torpid night level 2 !!! 
they wont work either
even after clicking the refresh button on the gui?
i didnt just press it, i borderline assaulted it, refreshed the page after some time too
it refuses to work
im trying this one now
but idk if this makes any diference
wait @torpid night try ilaria rvc again but with this link: https://huggingface.co/StakenS/sinatra/resolve/main/sinatra (1).zip
you pasted the entire thing
You didn’t duplicate the space
ooh right
i thought i only had to do that once?
after you go on your profile and open from there
could you please join me in a vc to get this over with because im very stupid
WAI
T
THE LINK U GAVE ME WORKED
GIVE ME A SEC TO TEST IT
i did once but closed the tab where i did
is that the problem maybe?
I THINK ITS GONNA WORK
nope, ima make it private or whats the word
gonna use a reduced file size, praying rn
wait does it have to be bigger than 22.2 mb?
no that wouldnt make sense either.....
i got this from the site itself...
what the hell is wrong with this acursed AI
it wont work....
made a new sesion thing, its private, its not loading, its stuck like this, refrshing did nothing, what the fuck do i do
why does my rvc look like this
Uhh
ModuleNotFoundError: No module named 'rvc.lib'
try wait for the precompile
how
how the fuck
where did you find this
You got it fixed?
yes
RVC Guides (How to Make AI Cover)
Translation by country
I keep getting this error saying Dll load failed.
What Colab notebook/fork r u using
Im using Voice Changer Client
Ayo? @gray solstice level 1 !!! 
When I try to load applio v3 it gives me```
ModuleNotFoundError: No module named 'rvc.lib'
Ayo? @fathom remnant level 7 !!! 
How do i get the Voice Changer?
redownload it
Ok
I get an error when loading applio v3
ModuleNotFoundError: No module named 'gradio'
when did u download it

the file?
when did u download the file
yeah
literally just did like 2 minutes ago
using applio 3.0.1
Does anyone know how I can put my models in the Ilaria RVC?
Idk how to uninstall it tho
delete the file
delete
which one
Ayo? @desert niche level 5 !!! 
Yeah
yea
okay
@inner cloak Have you tried the BVE model on uvr? Theres also spectral editing on audacity but that's to remove it from the track
@half cove would u say rvc v2 local is equally as good as applio or no, I did have rvc v2 local installed but I deleted it, and forgot to save my models but no matter I can just re-download them
@thin stump I did and both INSTRUMENTAL and Vocals do not fix the part. I tried just ensembling the song rather than the vocals, but only UVR HP 6 gets it out. The vocals aren't really high quality
it just came out 😭
i mean v3 is applio main now
yea
im gunna test speeds tomorrow
o okay
the quality will be the same on all forks
yeah harmony extractions aren't that great. I have to manually adjust the volume
alr
might reinstall rvc v2 local
-links
- Applio-RVC-Fork, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team GitHub
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
the inferencing was so fast, probably becuz of my nvidia gpu
I'm just trying to eq it in FL Studio and see if it fixes it. Should I try BVE via ensemble or in the VR Arch method? @thin stump
I did it in VR arch. That was my last resort and it got a couple of it out compared to HP_5 or 6
Let me see how it goes. Thank u.
Tried installing Applio v3 for a short while now, every time I attempt to run it after installing, this happens.
the rvc is slow
I got that error too
twice
Yes, and?
?
What am I to do? What is causing it?
Ayo? @real arch level 2 !!! 
Where are models stored in the new applio update?
try downloading from github here ig
Thanks.
i dont nderstand whats causing this
it worked last time
but now failed twice in a row?
You didn’t duplicate the space
cant get this to work
Can you link the colab your using?
i have to do that after every conversion?
No
You didn’t even duplicate it
i have done all the steps i dont understand
could you assist me threw screensharing cause ive tried to do that 4 times
i know where to press but it says i already made one
Tried it and... same exact result.
so i make another
If that doesn't work you can use the colab #📰│dev-updates message or wait for the precompile
then it worked once
Then it should be on your profile
now its faling again
Check your profile on HuggingFace
Click Ilaira RVC
it seems your keep using the original one
Wait for them to release the precompile
Not the duplicated one
Click one
Thanks anyway. I did the local client precisely because of some issues I had with the Colab, so I'll just be waiting it out.
Oof
Change it to public
Run the cell below it
sorry if this is a stupid question but the last time i did this i was using google colab with mega.nz files. is there anywhere in the server with like a more recent tutorial
Think I found my problem, my dumb self didn't install Git...
Oh lol
Yes
And it now seems to be working... yeah, turned out my problem was me. Thanks again for assisting.
Nice 
How else are you going to use it
So I tried the Ov2Super40k, and it worked great on a small epoch. Set it to 5K epoch and it finished seemingly fine. Except when I run inference, the audio generated has nothing.... completely silent. Wasn't like that for the 150 step epoch. Checked and anything above 375 epoch generates a silent audio file.
made a new acc
confirmed email
and came back here
and its still blank
what the fuck does this site want from me
what should my chunk be since i got a NVIDIA Geforce GTX 1600 SUPER
Ayo? @onyx brook level 3 !!! 
it worked again
Ayo? @torpid night level 4 !!! 
am I experiancing the mandala effect?
I swear i have had a madoka model much better than the one on weights (jp)
i thought i got it from here but the one in the database is just the weights one
i see
glad I have a copy somewhere, but it's only madoka. the weights one really is bad
i downloaded it and unless i used it wrong it doesnt seem to work great
Ayo? @serene frost level 1 !!! 
the original one I downloaded worked really well, I just didnt have a gpu that could handle wokada
i could literally say a line and sound like her instantly even if it was choppy
I cant get the new one to do that
it just sounds like madoka ate a bunch of cigarettes in monotone
does anyone know how to use the voice in discord what should my inputs and outputs be on?
huh
i may be an idiot, i thought i could get away with not installing discord on this thing. maybe I can set up a pass through?
do you know how to get the voice working through discord @serene frost ?
you just need to redirect the the output audio to an input device and then change ur input device on discord to that
idk how to do that cause I've never used the realtime voice changer but theres probably something like that
now that you mention it i probably could just buy a cord yeah
Ayo? @marble forge level 4 !!! 
like ong im not trying to rickroll you i'm actually serious
@proven hill music
i found it like 1 year ago
That is copyrighted
ok
ok here how would i make any video look this good with ai?: https://www.youtube.com/watch?v=o-YBDTqX_ZU
is bored
Ayo? @jaunty mason level 1 !!! 
I was scrolling through the help forums and just wanted to check that version 10 for python is the way to go?
For uhm. Okada
Can I directly download an RVC model from weights gg into AICoverGenWebUI? I tried the to copy the link but it doesn't seem to like the self-hosted weights link is there a place on the site to see the direct hugging face link?
doesn't work
I'll just stick with rvc v2 local
until a stable build is available
ig
or not
Idk
like i said im not trying to rick roll you i am actually serious @proven linden
what should my chunk be since i got a NVIDIA Geforce GTX 1660 SUPER
how i do like extract like some songs like emaple
"blah blah"
"yee ha"
but theres another guy singing in the background is there anyway to do it?
or should i say just like song name judgement from yakuza 0 that they have nishiki in the background
those are background vocals, you want to get rid of with UVR using hp_5, hp_6 or BVE. And people already done a lot of judgement ai covers
how do i get hp?
and do i need to use essemble mode??
In UVR, download models in the options
is it name like hp uvr?
Ayo? @sick canopy level 2 !!! 
or is it karoake?
oh wait let me look
oh ok
5_Karoake and 6_Karoake https://docs.aihub.wtf/vocal-cloning-guide/vocal-and-music-separation/uvr
I don't use ensemble when i'm going through the steps to dereverb > deecho normal > denoise
so use the mdx?
Yes, MDX23 and MDX voc ft. I think adding anymore like Kim Vocals 2 adds noise. At least that's what I've been told
or not?
Here's my typical ensemble
idk but i just mdx but dont see any karoake so maybe ill use essemble
also forgot to say is background vocal is instrumental or vocals?
7_HP2 UVR
MDX-NET kim vocal 2
MDX-NET MDX23C -instvochq
MDX-NET inst HQ 3 or Inst main and main
MDX-NET karaoke 2
MDX-NET Voc_FT
Demucs:v4 | htdemucs_ft
the background vocal is a vocal
whats the process method?
ensemble mode, 1024 segment size, default overlap. I think Lusbert said 512 segment size but I can't tell if he's being sarcastic. You can check where he mentions it
thanks man
yw, good luck
and just in case, for hp_5 and 6, you need to have no options on
i see thanks again
also do you have any idea like how to make your voice to ai tho?
#1159289738314919936 if you don't want to make it yourself. It's like $10 for the usual rate
you provide the speaking sample to the model master
i wanted to do myself
Suggestion for @sick canopy
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
RVC colab
which one is it is the full ai voice model one?
or Applio if you have a gpu but it's not functioning right now
thats how people make voice ai?
If you ask questions, people are always going to refer you to the guides here. #1159514067187277865 is always fun to watch
Yes
oh ok
wait what about hugging face?
HuggingFace Spaces?
yea
why the rvc colab is kinda took long to make yourself ai clone voice?
Do you need help with them or wondering what they are
Training takes time
what they are
Code mainly AI ran on HuggingFaces hardware
CANNOT train
There are some RVC HuggingFace spaces
Really slow tho
oh i see
Ayo? @sick canopy level 3 !!! 
what about rvc
RVC is the program used to make AI Covers
It’s completely open source and free
also whats the website where they made their own voice?
RVC
There is a google colab notebook for making RVC voices too
oh ok thanks man
hey mate uhh i think the background vocal i still can hear it even i change into 5
I didn't say it was perfect
anything where they can do it?
how do I make it so only 1 voice goes through
I do recommend splicing it together using different aggression settings. Because that's how I did it in audacity
huh?
Ayo? @proven cave level 3 !!! 
Use hp_5, hp_6, or BVE in UVR so the background vocals is separated
I was replying to Pokpong
so theres nothing can do with the background vocals?
If people can get the harmonies extracted in their covers, i'm sure it's possible. Just try a bit harder
different combinations etc.
so hp 5 and 6 cant do anything?
No, it DOES do something. Really depends on the song
but i just want the background vocals
i mean i said how do i remove the background vocals
but now i want it tho
some1 help please
You should get 2 audio files when you use this. One without the background vocals, one with background vocals only
@proven cave I think what i'm talking about relates to your problem
im on okada
i do the same thing i pick vocals only now it just one guy singing instead of a background vocal
Does anyone know how to fix the delay for my real-time voice changer client?
Just gotta mess around with
trial and error
i just try it now it just only a guy singing with no background vocals
what song is it? DM me
rvc boss syndrome is real
I got this error when running run-applio.bat
any fix? or what am I doing wrong?
RVC Guides (How to Make AI Cover)
Translation by country
uhh hey if anyone can help me out with this? i haven't ran into this before if you could please help me out with this i would greatly appreciate it
it's a fresh install too so
yo my mic is acting robotic
my mic is fine but
the ai voice soundin like a fckin robot
Does anyone know a thing about vobile? I got traffic from them but no copyright
Ayo? @thin stump level 9 !!! 
Any way to get RVC to sound less robotic?
nevermind, they only helped add a copyright claim to the video for the claimant since it hasn't been identified for weeks



