#✨│ai-help
1 messages · Page 45 of 1
😭😭😭??
its almost at 3k visits
🤫
@glad zealot
HINA!!!!!!!!!!!!!!!!!!!!!!!!
wanna help me out on a super ultra secret project :-3
mew
is dat a yes or a nah
why not
check dms !!
everytime i try to use easygui i get this
Please install tensorboardX: pip install tensorboardX
Traceback (most recent call last):
:3
tensorboard being STOOPID.
something about numpy module not found, how to fix? please respond, i like to use this when making new voices
Close W-Okada, delete stored_setting.json, restart
pls help how to fix
how i close w_okada
close the cmd window
maybe bcuz python is in program files?
hm
@sand iris 💀
I mean yeah this ain't using the runtime python
ofc it is why wouldnt be?
didnt have any trouble running before, never bitched about numpy
bout that chief
where do i enter that
cause i tried that by opening up a python prompt
Oh yeah no that's almost it
but it says that it doesnt know what pip is
you want command prompt to deal with python packages
ya
ok and how would i punch that in? just pip install numpy?
Yup, should be good now
Yeah it throws a bunch of errors and warnings, but the important part is "installed numpy"
ffs
same thing again for this one and subesquent bitchings?
Yeah we're doing this again !
pip install protobuf
OR, alternatively, what could work in only one step is pip install -U tensorboard
without the google. ?
Ayo? @short pecan level 4 !!! 
shut bot
Which will update tensorboard & everything it depends on
yeah without it
the pip package name doesn't always match the python code
ok its working differently now
its not spitting a bajillion lines now about modules
yknow what?
@true vortex this is what I meant by modules 😭 I heard people having problems with it
think it mightve been micro~~ cocks~~soft fault
Though I don't know the context
skill issue
/s
bc windows update tends to fuck with shit, so it mightve diddled with the python modules
What do they mean by modules?
Python is a programming language right ?
@molten pecan
Rather than shoving a billion things into said programming language, there's modules
which are just bundles of code you can choose to use
modules are basically scripts that python programs can call to and use for whatever purpose
some of those modules come by default, others have to be installed externally, via the pip utility
^
also tensor is working now thanks @sand iris ! ^o^
👍
what is the collab link to the easy gui rvc ?
you press on #1159513888199540817 on the colabs thingy
#📑│making-models sory
someone knows how to switch voices with those kind of ai bank voices?
Any idea why it doesn't find the wav files?
The zip file might have separate folders which each member maybe
its all grouped
in a single file
That's crazy. I don't think it's supposed to be like that
Did you check the reviews?
yeah
Ayo? @cyan tusk level 3 !!! 
@proper shale what do you think?
A model with four different voices in it
200 epochs
Uhh that feels like a little too much. Plus if he said "each" then there should be four models
And it says 200 epochs here
Idk what's going on
oh
why my mic repet the what friend say
i dowload the wrong one
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
one of those should work
you need put into .zip
didn't worked
screnshoot
I tried everything, upload through collab, through drive, zip, not zip,...
1 audio file, many, mp3 files, wav files....
try click it multiple then run again
run preprocess data?
yes
same result
hold up

use this folder(not zip file) then change you google drive language to english, and change name folder sample astuoni.wav du.wav, etc also folder data set name in gdrive, no capital alphabet
work?
Can't change the language
Its still french
It gets french on the google search part
But not for the drive,....
do you try change folder name with no capital alphabet
Will do it rn, without changing the language then

Nah, still same problem
Thats how I named the files

check on this gdrive, do you have this folder
Worked
Yeah I do have it
any arabic vc ?
Can I ping you for my training, caus I have problem in training:
- It stops training for no apparent reason
- And when I try to resume it it says that
Thank you ❤️
last error?
@fossil glen any arabic vc ?
no try use google translate
you want resume?
Hey guys! Lovin this supportive community so far. I was wondering if anyone here knew how one could import the hugging face voice conversion RVC model into the unity game engine? Does anyone have any experience doing this? Thank you in advance!
Look, I will go through the whole process, I have the same error for 6 days in a row. May I ping you once I get to the specific case?
Bless you btw
you need disconnect then run it again from 1.Preprocess Data 2, 3, to 4.Train Model, step by step
Thank you ❤️
this run only 1 times
also thanks, i just try figure out whay my models get to many errors or glitch, cause you problems, i check again my dataset the name its have symbols, but .... when i proccess a data set its run without errors 

i need help guys so im opening the RVC gui and its just open a wierd black screen
RVC in to Unity??? There is no way to do that
rlly? damn
Ayo? @inland jolt level 1 !!! 
can i ask why tho??
like what about it is incompatible with a game editor like unity
can someone help me??
what's the issue
python
- super GPU intensive
- the app wont open
you wouldn't make a game like that run easily
Which app?
the ai changer
There's a lot of forks of RVC
W-Okada, correct?
How to troubleshoot
- Give context: Throwing your error/issue alone will not help. Tell what step are you on, what were you doing before issue, your settings, your terminal, et cetera.
- Be relevant: You come to ask for help, not to complain. You can't ask for RVC help in voice changer help and vice versa.
- Be polite: DONT DEMAND FOR HELP OR I WILL FUCKING MURDER YOU. Ask politely, the server was not made for you.
- Don't be an attention whore: If people are ignoring your help, kindly ping them. If they are ignoring you, either they don't know how to solve it, they are busy, or you are an asshole.
WITH ALL DUE RESPECT, DO NOT BE AN ASSHOLE AND HAVE COMMON SENSE.
too mean
I see, thank you for the input 🙂
np ;) glad to help
if you have any more questions let us know
Follow up question though, isn't there a setting for the hugging face RVC that allows for less GPU intense performance?
the yt said to open start_http
"pitch extraction algorithm"
you can use dio or harvest
to use less GPU
I see
buuuuut it'd be slow
are those specific rvc's?
especially harvest, which is cpu only
just pitch exteactors
Ayo? @thin stump level 5 !!! 
extractors*
i see
And what happens once you open it?
Does it say anything or just not open at all?
nothing
So my question is, say I just wanted to import the rvc to unity even though it would be slow
how would I go about doing that, by say downloading the ONNX file?
Sorry if this question is crudely asked, I have little experience with AI models
screnshoot
it wont open at all
I don't even think you can, at least running directly
Maybe an API?
Check the pins in #🔍│help-w-okada, they might help
k thx
I see, I'm trying to avoid using any sort of APIs as I want to see if I can host the entire AI voice changer RVC program locally
LOL
I'll look more into it and see if its possible
-rvc
Suggestion for @violet mica
Translation by country
?
Ayo? @violet mica level 1 !!! 
-colab
Suggestion for @violet mica
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
im trying to download it but its show the black scrren i showed you
RVC Disconnected, Kit Lemonfoot Google Colab
for 48khz (good result)
easyGUI, by rejects Google Colab
only 40khz (normal result) @violet mica
https://docs.aihub.wtf/vocal-cloning-guide/rvc-v2
if you want use local pc read properly and do it step by step
actually, 48k wouldn't make anything better depending on the dataset
bro im just saying how to download the app.
safest bet is 32k
-local
Suggestions for @violet mica
- 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 Huggingface
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.
im even more confuse than before
use google translate
bru
read and just do it do it
bruh what?
forget it im just not gonna do ai
alrighty then
I've shown you the links
@violet mica use this https://github.com/Tiger14n/RVC-GUI/releases/
You should probably go to the folder you installed your fork, type cmd, hit enter
i just already download it
Then write the bat file that needs to be opened
That’s heavily outdated
Yeah
But either way, if you wanna open it, do this
It doesn’t even have RVMPE
And this
its says in frnech
Running with the system Python.
Python n'a pas ÚtÚ trouvÚ; exÚcutez sans arguments pour installer Ó partir du Microsoft Store, ou dÚsactivez ce raccourci dans ParamÞtres > GÚrer les alias d'exÚcution d'application.
Appuyez sur une touche pour continuer...
that what it's says bruh
but its says that they did not find python
Don’t use that one
yeah use Mangio
It’s outdated and doesn’t even have RVMPE
-local
- Applio-RVC-Fork, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team Huggingface
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.
it's right there
Mangio
Ayo? @violet mica level 2 !!! 
Y'know what, just use this then https://colab.research.google.com/github/hinabl/AICoverGen-Colab/blob/main/Hina_Mod_AICoverGen_colab.ipynb
You don't need to install anything with this
not really, but it's very straightforward
well shit
run all the steps
i dont know what to do with this
then wait 3 mins
and it'll give you 2 links, open the bottom one
after that it's simple lol
download model in there, copy the link of the song, select ur model, ez cover
no but like i have the page but idk where to go where to click
Just run all the steps
It really isn't that hard

I can rest assured knowing my grandma is able to make AI covers so you can too!
Believe in yoursef
i dont know i fi h ave to install something or just
drop the files and choose what i want for the voices
cuz i dont know where to put it
Look, with the Colab link I sent you, you don't need to install anything.
aight then where do i drop the files?
Just run the steps accordingly, open the Gradio link, and there you'll be able to use the YT links to make covers
You can upload the full songs if you have them installed too
It's probably a French equivalent then
There should be Runtime somewhere then
It's the fifth option right below the name of the Colab
(for me it's in Portuguese so I won't send a screenshot and confuse you even further lol)
Just click on "Run all"
yea just wait around 3 mins
aight after what am i doing
need to drop the song with both instrumental and vocals
separate?
everything in onr
Ayo? @violet mica level 3 !!! 
nah, the original song
oh
that's what i meant
aight ig
sorry if it got confusing 💀
and then u have to download ur model, there should be a section for it, easy stuff
cuz the video isaw the guy said tograbs the song separeatly
speratly
bruh
yk what i mean
yeye
i did one
see this is an ai cover maker
so uh it does everything in one
while on local forks you need to process the vocals only cuz thats how rvc works
oh ok but this one i just have to put the music and vocal?
just the song yea
can i train models with 3050ti laptop 4gb vram?
its been more than 3 mins i still not have the thing ur telling me
it'll be slow but it's possible
have you scrolled down to the last cell
u can send screenshots, i think
lol
yea but idk what to do
Click the public link

download model tab
where
ooor upload model tab, where u upload the zip
it's there
(but it's the zip)
like a Huggingface link
alrighty so use the Upload model section
usually takes a bit
Ayo? @acoustic garnet level 1 !!! 
aight i was about to say it's a bit long
ok i hav eboth now
do i press generate now?
u have everything together, Goku's voice and the instrumental
yea takes a bit
you're welcome :)
@proper shale trying the colab site now, thanks for the help!
couldnt make any ai cover songs with other colab links and gui rvc .bat whatever app that runs a cmd x32
@violet heron im getting a different messgage now in the go-web cmd prompt
Can I upload audio files to this app and just convert instead of doing real time voice? If not - is there a downloadable app that can do that (not the web browser ones?)
fukkkkk it was all perfect but something doesnt work, it detected one vocal only and changed that, the other it tought are background vocals
Ayo? @fleet stirrup level 1 !!! 
is there something to fix this? not many options I see, it completely changed the background vocals with my model but forgot the main vocals lol
the other problem is that I can't upload a file, nothing happens, can only link a youtube link
it oesnt change half my vocals bruh
Okay I just explored the server. I am officially lost. First of all how do I install RVC locally, HuggingFace seems like a headache, and how do I train models?
If someone can link a tutorial it would be nice
ive got an old rv installed
rvc gui.bat file stuff
they said its kidna outdated, would probably get the job done but it has errors for me at the end stating stuff with red text
also crashed my pc completely once
your best bet is most probably these google colab oublic links it seems, they get the job done too, although a bit goofy
I've got a pretty beefy computer though, isn't it good if there is more available RAM for the model?
I dont even know what ure talking about king
might be better to run stuff locally on your pc
Idk either 😭 🤣
I don't know how these stuff work at all
Alr, is there a download link somewhere?
the google colab links work, with different models, you also just have to link it, dont need to download it
nah i dont know any, there was amesssage tho ill search it for u 1 min
Thanks man!
-local
Suggestions for @azure marsh
- 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 Huggingface
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.
Ayo? @fleet stirrup level 2 !!! 
Do all those have a different style or do they all do the same thing
I thought making freaking ai cover songs is a simple thing
or not that complicated
turns out im sitting for 3 hours in front of my pc and could barely made a song
Lol yeah. I just want to see if I can make some realistic string instruments with RVC
wym by different style or what lol
idk if the ai can imitate strings or other instruments
wait for a mod bruh
Like the different types of "forks". Like does Applio RVC do the job differently than AICoverGen?
I asked someone in general and they said someone made a litteral chair model
So I think strings are possible :/
tbh i linked it now but idont even know whats a fork
everything is possible its ai
I think its like a branch of the main project
the previous guy was asking about changing the acapellas human sound
Like a fork in the road. No clue 💀
Like changing the timbre?
I thought its just an .exe file runs an executable shit and I just render a song
simply making an ai song bro, changing the singer
with a human model
Ohh yeah I get you
you download a model or link a model link and the ai will change the acapella to the model's sound
so Kanye can sing Edith Piaf
or whatever anything
I've done some Fooocus stable diffusion stuff, so I knew it wasn't going to be a simple one-file solution but I didn't realise I have to do THIS much stuff
Ayo? @latent frigate level 2 !!! 
Yeah yeah I understand
Can anyone make a proper Chika Takami model by combining these .pths somehow and pasting it here so that I don't have trouble and if possible can you ping me when done you'll find it in weights methinks https://huggingface.co/trioskosmos/rvc_models/tree/main I don't understand how to do it and there's a cover I really wanted to make 
question what you use to make covers
tryna make covers now
but google colabs arent that good
searching for other colabs
or i need to try something else
this didnt work for me today
ive seen it in a video
i could upload a song and i couldnt change the audio to my song, even when i refreshed
and then i timed out of course
maybe it was too busy?
it was randomboy.mp3 and randomgirl.mp3 only, the basic stuff
couldnt use my literal uploaded stuff
should i pay for colab pro it would help?
again
i refreshed and the audio is not there
@urban rapids It works for you?
or maybe i should use mp3? guy in the video used wav
Did you ever click convert
convert what
bro check the left side
im talking about my audi not showing up
what ijust uploaded
I have the model but just 2 audio
randomboy and randomgirl
fordrake vocals (wav) is uploaded
yet it doesnt show up
theres the audio he uploaded on the left side
fucking not working for me and I dont get why
its working for you? @urban rapids
i even try to scroll up down
ill try with mp3 real quick
Click convert anyways
So I am trying to ownloa a model for an ai cover I want to make and it keeps doing a vharacter named Viego instead of the character I want. What am I doing wrong?
https://rentry.org/ilarvc_inf_guide or use my updated guide :b
~ Guide made by Julia ~
~ Improved by Alex and Squishy ~
Note: It works on mobile too! If your device is slow and can't run RVC locally, don't worry. This software doesn't use your device's resources. All you need is a stable internet connection.
READ THE GUIDE TO THE END. It will be worth it!
~T...
download for huggingface*
clicked convert anyways and of course it rendered with randomguy.mp3
Ayo? @fleet stirrup level 3 !!! 
a 3 second test
Oh btw did anyone see this I've been trying since yesterday
Ayo? @urban rapids level 2 !!! 
Roger and ty
Try with my guide then
i cant use huggingface links bruh
i just get this when i try to copy the link address of a model "https://huggingface.co/datasets/shrek101202303/ai/resolve/main/Garnet (Steven Universe) - Weights.gg Model.zip?download=true"
and not the correct link
In case you wanna help you can go to the thread too https://discord.com/channels/1159260121998827560/1184335899903938641 no pressure
Yeah it's not supposed to say the whole download thing it should just be .zip at the end
how do i fix this
doesnt even work now doesnt connect to T4 at all
link should look like this king
right click and copy link on discord
not copy text or else
copy link
yeah ik, but im following the tutorial copying the download link address on my hugging face
idk king why it does that
Wdym by correct link? What are you trying to do?
Yeah sure :3
Tysm 
okay so im uploading models from weights.gg to my huggingface dataset to use them and i dont get this
^
this shit dont work idk how you guys managed to get it working
he gets a goofy link with =true (20% stuff
instead of a link like this
his links look like this
Well first of all it's best to remove the spaces
do i just put weights.gg models' links like huggingface models or do i have to do something else?
Putting spaces in the model file name might give you errors
it automatically puts them in thou, i know i can remove them but thats really tedious for every model tbh
putting the links only doesn't work
what do i do?
Seems like the link is private
And not public
-huggingface
English guide
Link: https://rentry.org/fdg_guide_new
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
It should tell you somewhere how to set it as public
"HTTP Error 401: Unauthorized" 💀
Ayo? @civic onyx level 4 !!! 
Where?
Where?
so i have to make my dataset public?
Yeah
Also, are you making a model of Garnet from scratch?
You do know there are already plenty of models of her, right?
no no, i took it from weights.gg and put it in my huggingface dataset files to use it on a colab
actually, is there a way to make the audio not automatically only detect vocals?
Ayo? @brittle wing level 1 !!! 
i think an option like that is making my stuff sound all muffled
nevermind the only one online is spanish
Wait what are you trying to do?
Inference?
anndd this happened when i pressed "make it public"
"{"error":"Sorry, we can't find the page you are looking for."}"
nice site running huggingface
(sarcasm)
can you remove the spaces of the zip file?
Does it have any?
tried that already
Still getting an error?
yeah
even after making it public
its still being stubborn
You can rename zip files from huggingface?
okay so the 20% thing finally worked
after i made it public
didnt matter if it had the 20 thing in it
Ah alright
would of been nice for the tutorial to tell me that i needed to have my dataset be public lol
Hello, so i downloaded a dataset and im wondering how am i supposed to train it
kinda need help
i already have the rvc downloaded as well as the gui to train the voice but idk how to do it
Should I give up on an AI cover if all the currently uploaded ones on youtube have robotic voice cracks? I tried UVR but this is a song with harmonies. -Guide didn't help much
Don't give up already!!!
I know the solution!
Is your audio in lossless quality in the first place?
oh no, I guess we're all working with mp3s
since it came from youtube and it's not really an official song (Red flags)
You can still download it in flac although it is compressed beforehand
Do it eitherway
Use a "youtube to wav" website
And remove the instrumental
And then make sure you remove the reverb and then the harmonies. In that order
Huh, i'll try that. I think what I got wrong was keeping the reverb in
im sorry when i download models , is it like a dataset or do i have to make one?
im so confused
A dataset and a voice model are two different things
if i want to train a voice how am i supposed to proceed
like i downloaded a model
i want to train it with my voice
Dataset: collection of audios that are used to make a model
Voice model: AI model that is built over the dataset
ohhh makes sense
so i can directly train a model correct?
Ayo? @leaden meteor level 1 !!! 
Uhh if you want to make a voice model of yourself, you need to make it from scratch. You can't combine it with a preexisting model
what i meant is that
basically
i downloaded a voice model
and i want to make it more realistic
because it sounds like ai
Mhm
That could be due to plenty of reasons. Maybe it's not the model's fault itself
To train a model you always need a dataset. It can't train by itself if it has no files to take its information from
Mhm
If you can reach out to however made the model, ask for the dataset they used
Fingers crossed
🤞
Yw! For more help go to #📑│making-models or #1159290066074607626
alright
i'm training a model right now using original rvc, and i set it to 404 epochs. is this overtraining it? because each epoch takes about 6-9 seconds, and very rarely up to a minute.
Ayo? @brittle wing level 2 !!! 
yey.
-local
- Applio-RVC-Fork, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team Huggingface
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.
-colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
How can i fix it
İndex -1 is out of bounds for axis 0 with size
Runtime disconnected
Your runtime has been disconnected due to executing code that is disallowed in our free of charge tier. Colab subsidizes millions of users and prioritizes interactive programming sessions while disallowing some types of usage as outlined in the FAQ. If you believe this message is in error, file an appeal. Please include any relevant context about your usage.
Your compute unit balance is 0. Purchase more
To connect to a new runtime, click the connect button below.
Ayo? @half raptor level 1 !!! 
help
bruuu not again
Ayo? @runic lynx level 1 !!! 
hina modified again
this string pops up everytime i reload the google colab
it takes up the queue of my dependencies
No results found for the search exo baekhyun...
How do I make the rvc make only taking 50% of the model voice?
Ayo? @brittle wing level 2 !!! 
Its strange caus it says that the folder is empty + that is is a success
it happens with certain songs, tbh I'd go the manual route, filter out the vocals with UVR/MVSep and process them through Ilaria RVC
Btw it seems to work for me and I can resume now ❤️
is there any usable way to train rvc on AMD
or an alternative to RVC that works with amd
anyone know how to export something like this as multiple files in audacity
ah pain
You can on linux, but if you don't know how to use it and/or don't have an install of it available, it's gonna be more of a bother than anything
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
RVC Disconnected
easyGUI
Those are the training ones
Thank you!
The UI... Isn't as pretty as locally installed options, but they do work if you don't train for more than three hours in a single session
Yeah I only need it for something short
Do you think one is better than the other? I was told RVC Disconnected was broke
¯_(ツ)_/¯

how do i fix this shit
Install the shit in requirements.txt
where is that
yes
im asking for help rn
personally i use mainline on kaggle(cloud)
Ayo? @brittle wing level 2 !!! 
Yeah I still get that problem it stops training for no reason
And when I try to resume I can't
find a colab that is working
Its Easy Gui, it is supposed to work
- I was told RVC Disconnected is broke
No like a other one that have "The Mangio-RVC-Fork "
Any link to suggest?
*** it is?***
Well I was told its broken and gives bad results
so you are saying i need all of that downloading nonsense
idk
i have a slower pc than a turtle
Who told you that?
A guy in the server
RVC disconnected uses mangio rvc fork for training
It shouldnt be worse in any way
It all depends on the settings and dataset
So If I get bad results it is because of the dataset or the settings I put in
With RVC Disconnected
Yeah
Do you use these settings?
Target sample rate to 40k
all the time? even If I have that much Hz in my dataset?
48k leads to ringing and noise if the dataset isnt 48khz
okay
If you have less pick 32k
Got it
And you use rvmpe?
Or another one? Personally?
Ye
what is this
A problem I get on easy gui, for no reason it stops training, and when I resume it I get another error
Idk what is it due to
ya got a screen?
Here @copper pebble
I'm having issues on train model step in colab version it opens some kind dashboard and does nothing
TensorBoard? Have you clicked on the refresh
button
that it has
Ye i did just shows this: "No dashboards are active for the current data set.
Probable causes:
You haven’t written any data to your event files.
TensorBoard can’t find your event files."
hey do someone know if a local solution for mac m1 is available?
what colab r u using
mangio fork works for mac i think
also mainline rvc maybe?
-local
- Applio-RVC-Fork, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team Huggingface
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.
try mangio
Model training
What's the name of it
Ayo? @brittle wing level 1 !!! 
oh that's old
-colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
mangio didnt worked on my mac.. because it didnt load while choosing a pht file
use RVC Disconnected
which fork is the best one for mac?
i had issues with mangio..
huh?
maybe you put the model in the wrong place?
yeah i installed it and after choosing the pth file it was loading infinitely
under weights folder
oh
what would you recommend? applio?
where is it?
-local
- Applio-RVC-Fork, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team Huggingface
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.
original rvc
cant find mainline?
Ayo? @peak tendon level 1 !!! 
ah ok thanks i wll try. is there any guide for mac installation?
uh
nope
sadly
?
Works now, ty. I was trying to make fortnite amie voice lol.
anyone know how to download voices from weights.gg?
np ;)
you need to be logged in
No one is using a Mac and a fork ? Damn
Yeah, gotta own apple hardware for that lol
But the experience shows it is not possible ??
?
are the dataset for training a model suppose to contain speaking audio clips, singing audio clips, or both? I am having trouble with creating good apacellas for singing audio clips of my target voice
who can help, I’m sending you a voice, and you can make a model out of it, I just can’t do it no matter how many times I try
yo anyone know how to upload a zip file to the RVC v2 disconnected
both should be fine
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
^^^
hmm ill check it
why easyGUI doesn't work

I'm making a model but it doesn't appear
so last night i was training a model for like, two hours, and i went to bed before it finished, and an hour later my computer went to sleep.
the thing stopped at epoch 395 (i set it to 404) and i thought i remembered that you could resume training even if it stopped before it finished.
am i correct? and if so, how do i do this? (i'm using original rvc)
looks like it's just restarting from epoch 1, but i guess i can wait lol
can i use all talking audio clips, or is at least some singing audio clips recommended to be included in the dataset?
FYI, your reply is not pinging the person you are replying to i believe you have that feature off
help me with with this message
@sand iris help me with this message i have a cuda gpu
What's the GPU ?$
Ayo? @sand iris level 49 !!! 
I have a nivida 3080 gpu
you don't have to ping me twice, I coud just not be available
how to find a server with a specific language?
anyone help me with this error message?
sorry
You mean voice models ?
Did it work before ?
Yes, mis typing servers 😭
https://weights.gg/ might have but I'm unsure
Sorry to do it to you, but have you tried rebooting ? I've had a friend get dumb errors after his system went to sleep mode
does that even work?
are you following a guide that uses python venv?
Ayo? @azure shoal level 4 !!! 
i dont know
C:\TRAIN\runtime\lib\site-packages\torch\nn\modules\module.py
Do you count prebuilt runtimes as venvs ?
I'd recommend using just part of them if you're making a singing model
And yes that's intentional, I don't like pinging peeps haha
Been a while since I did that stuff
i just restarted the pc
theres a 'CPU THREAD' setting somewhere right ? Anything other than 2 causes issues on that if I remember correctly
whats that
whats a "CPU THREAD"
can you screenshot the training UI ?
i used 12 cpu threads
yeah, other than 2 causes issues I've been told
i still got the same error message
You might want to check your nvidia drivers
if they are up to date
I really don't understand why it's bugging this way though
i opended eveything fine and evrything was working but when i closed it and am trying to reopen it the client is just showing a blank screen how do i fix that?
whats the reccomended batch size for training?
Ayo? @brittle wing level 12 !!! 
i still get the cuda out of memory error
That's weird, I've always used that colab and it works, it just happens with that model, it used to work but then it gave an error, but it works well with other models
Ayo? @young halo level 2 !!! 
they are up to date
i just checked them
The weird thing is that it tells you there's no cuda GPU available, but then tries to use it
You are using "Mangio-RVC-Fork" right?
Yeah....
I really don't know what's up with it
Mangio hasn't been worked on for months at this point, but the existing releases just shouldn't break
are there download links?
That's over 6 months old and has been restricted by Colab so..
If you have Colab Pro it works
Hey! I was trying to train a model with a dataset, but at the middle of the training the server disconnects, so i was wondering what can I do to fix it or if you could help me making the model
I don't have and it still works for me, I'm just having issues with that model, the other ones work just fine
Here's the "main" RVC, which is what Mangio was created from
oh maybe it could be the ?download=true part
or perhaps the model was privated
which I think is very possible
I'm going to try the other colabs that you sent me, thanks!
Now do you guys know what can I do about this?
alrighty, I'd recommend ilaria rvc for covers n stuff and RVC Disconnected for training
that would solve this issue
Thanks!
how do i do this... @sand iris
any good and fast inference methods other than ilaria? ilaria claims to be faster but it takes literally 4 times the og file time to actually do something
preferably on colab since its faster, locally i have a lot of issues (not the brightest, i admit)
(and with ilaria it takes a great amount of time to just end up giving me an error)
You did duplicate the space beforehand, right?
Oh it's giving out errors ?
We've had someone else report that too
yes but even in the duplicated space it gives me errors or just takes a huge amount of time
When does it give you an error?
after finishing the "inference" which never actually finishes
it just processes for a long time and then "error"
Ayo? @frozen sandal level 2 !!! 
How long is the audio?
like 3 minutes
a lot of work lol
if i have to do all that to get a simple inference, it's not worth it
Wdym it's super simple 😭
Unless you are using windows xp or sum
yeah but with colab I didn't had that issue, i'd sometimes inference 5 or 6 minute audio in literally less than a minute
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
any info on how to get thru this? and no i'm not executing it on admin, i'm literally just double clicking
it's for local inferencing with applio
@proper shale
Are there other reasons why people choose Mangio over Applio besides hybrid training?
What voice?
An audio file?
Or do you want to change your voice in real-time while you speak?
What is optimal amount of time to train a model? Is it safe to assume the more you train the better?
Ayo? @brittle wing level 1 !!! 
~ Guide made by Julia ~
~ Improved by Alex and Squishy ~
Note: It works on mobile too! If your device is slow and can't run RVC locally, don't worry. This software doesn't use your device's resources. All you need is a stable internet connection.
READ THE GUIDE TO THE END. It will be worth it!
~T...
My guide
It depends
Are you talking about the amount of epochs or the dataset?
I'm assuming I can't use a 1030 to train a model without it taking more than a day to complete 1 model?
Ayo? @crimson wraith level 1 !!! 
Oh boy. Unfortunately you won't be able to train on that at all
You'd need at the very least 6GBs of VRAM
would the colab pro subscription suffice?
i also have a macbook with an m2
not sure if it can train - seems like maybe not.
It would yes. Just know that there are colab notebooks that can run on the free tier
Unsure about that, we don't have many people running mac, and while training could theoretically work, I don't think it's implemented
thanks. this stuff is pretty wild.
can someone explain me what crepe hop is?
@molten pecan since you made a guide about illaria
i'm getting this error when updating the duplicated space for a better GPU at hugging face
Is there a red text saying "Error"? Otherwise ignore it. While it's building it might give some errors but still work fine
Try restarting the space
as it says in the last line FileNotFoundError: [Errno 2] No such file or directory: 'configs/32k.json'
factory reset
Click the three dots, Settings and scroll until you find Restart this space
i did both already, both restart and factory reset didn't work
i get the same error on both methods
I never used the paid feature so I don't know. If I was you I would try from a different HF account or in another device
I think the space just wasn't meant for any GPUs to be used 😭
CPU isn't that bad
@proven hill you think so?

