#✨│ai-help
1 messages · Page 21 of 1
spaces in filenames
Ayo? @frozen quest level 1 !!! 
make sure there are no spaces in any folder/file names where RVC does anything
that includes :
the folder where RVC is
Any dataset files
Okay, thank you!
Ayo? @coral galleon level 1 !!! 
It still doesnt work
I'm not sure, sorry
Hey @mystic ferry
Hey
Oh
doesn't sound like the model a lot, can you help with that?
Uhh I mean ig
Thank you man
Could you hold up like 5 min?
sure
Want to hop in vc or over text
Ayo? @mystic ferry level 6 !!! 
Also just saying I'm not very knowledgeable but I'll gladly help
Applio RVC work and I have an error during preprocess. path to RVC is: "D:\rvc\applio\Applio-RVC-Fork", path to dataset is: "d:\datasety\mydataset". When i tried to preprocess using web i get an error that The system cannot find the path. using cli and after pasting this: Myname d:\datasety\mydataset 40k 16 i get this error: TypeError: preprocess_dataset() missing 1 required positional argument: 'dataset_path' i tried to move dataset to datasets dir in RVC folder. In web interface after reloading dataset list it discover it, but after clicking "preprocess data" button i get the same error. I tried to change \ to /, cover path in < > or " but always the same error. What could i do yet?
Yeah, just search and test them out
hey guys i just downloaded the app and i run the command window and no reaction
any helpers
what did u download??
How to Sound Like an Anime Girl With This New AI Voice Changer
AI Tools Search
•
940.483 Aufrufe
im trying to set up
Ayo? @brittle wing level 1 !!! 
ohh voice changing i see
i wanna sound like aatrox or goku
mmvc
mhh dunno what that is sorry
u could try okada
not really rn haha
heres a guide for it
its pretty good
i also suggest to u to go in this channel #🔍│help-w-okada u should find someone that can help u
@copper pebble how can i remove the delay
go here im not good at voice changin
sorry
@brittle wing
@copper pebble ok i got all done how can i make it now on dc
.
Ayo? @copper pebble level 26 !!! 
wrong person to ask sorry bud
doesnt it say it her?? @brittle wing
why does it say error?
hello, what is the meaning of D_XXXX.pth and G_XXXX.pth?
How do I continue training a model? I know I have to paste the D_XXXX.pth and G_XXXX.pth from the RVC-Backup but what to do next?
Do I have to redo the "Process Data" and "Feature Extraction" if I continue training a pretrained model?
you have to start the new training and specify the old D_XXXX.pth and G_XXXX.pth from the previous training
I knew that already but >Do I have to redo the "Process Data" and "Feature Extraction" if I continue training a pretrained model?
I didnt know about that
Ayo? @odd flower level 2 !!! 
Can I just specify the old D_XXXX.pth and G_XXXX.pth and hit the "train Model' right?
yeah
Without having to redo the "Process Data" and "Feature Extraction"?
Does anyone have a guide here on how to train ai models with my own gpu?
How long should it usually take to pitch extract with rmvpe on a 6 minute dataset
less than 1 min
With cpu or gpu
gpu
Mine was using both at the same time what does that mean
How do you even train without local wtf
are you using colab, if yes them please share the notebook
I can't made model. I always get error
set([name.split(".")[0] for name in os.listdir(gt_wavs_dir)])
FileNotFoundError: [Errno 2] No such file or directory: '/content/Retrieval-based-Voice-Conversion-WebUI/logs/sunflower/0_gt_wavs'
What i need to do?
u doing on colab or locally??
On colab
Ayo? @vague eagle level 1 !!! 
And colab delete all again
did u upload the zipped folder with ur dataset in the "rvc disconnected" folder on drive??
Yes
did u named ur zip like the folder??
i mean in the cell folder
give me a sec pls
anyone?
is it the scout one??
another
that
then write here "Sunflower voicemodel.zip"
send me a screen pls
try to reupload the zip in the drive folder naming it without spaces
try also to write it without spaces
Ayo? @vague eagle level 2 !!! 
why they can't find folder?
do this idk maybe its a bug
maybe it's not zipped correctly
cuz sometimes when I handle zip files for other purposes I find the actual folder needed is inside the zip, so the program can't ready it properly
And... yea... you right
me??
He
yh i was thinking abt that too
i trusted u hahah
Making minecraft server experience really helped me with this
People don't know how to zip their datapacks properly
On windows you select the content inside, not the whole folder smh
change acc
I've already done this several times
try to not use all the 4hrs of runtime
they give me less than 10 minutes
lmao
i was training few days ago with no problem, today when I wanted to train there are a whole folder missing or its not finding it, can anyone help or maybe send the /mute folder?
i did everything correctly as far as I know, no space in folder path, processed my data, extracted feature
Process Process-1: Traceback (most recent call last): File "multiprocessing\process.py", line 315, in _bootstrap File "multiprocessing\process.py", line 108, in run File "D:\MangioRVC\Mangio-RVC-v23.7.0\train_nsf_sim_cache_sid_load_pretrain.py", line 110, in run train_dataset = TextAudioLoaderMultiNSFsid(hps.data.training_files, hps.data) File "D:\MangioRVC\Mangio-RVC-v23.7.0\train\data_utils.py", line 27, in __init__ self._filter() File "D:\MangioRVC\Mangio-RVC-v23.7.0\train\data_utils.py", line 41, in _filter lengths.append(os.path.getsize(audiopath) // (3 * self.hop_length)) File "genericpath.py", line 50, in getsize FileNotFoundError: [WinError 3] The system cannot find the path specified: 'D:\\MangioRVC\\Mangio-RVC-v23.7.0/logs/mute/0_gt_wavs/mute40k.wav'
I think that is because of the path file
'D:\MangioRVC\Mangio-RVC-v23.7.0/logs/mute/0_gt_wavs/mute40k.wav'`
Try searching that folder on your computer
Because for file paths Windows uses " \ " and not "/" to my understanding
funny thing is i dont even have the "mute" folder in logs, wonder wtf is happening lol
Did you search on D: or E:?
yo im currently trying to do a cover locally but I dont know how to mkae this work on cpu ( Have a radeon gpu)
"No supported Nvidia GPU found, use CPU instead
CUDA is not available. Make sure you have an NVIDIA GPU and CUDA installed.
Use Language: en_US"
Are you using Applio?
a what?
Ayo? @hybrid heart level 1 !!! 
I jsut have local model. I start the go-web bat and then i try to change the inferencing voice and it just puts everything in queue for all eternity
Yeah I checked there is nothing
i dont have the mute folder, or that is not a folder that anyone can send over and I can replace?
Try using this one
If you keep getting an error lmk!
Make sure you don't close the black command window until you are done using RVC
yo I have a question
Do you have hidden folders on?
i want to train to around 700 epochs
Try turning on that option
but whenever I train I run out of gpu time
It's on the settings of the file explorer
it has intel in it I have an amd cpu does it matter?
?
Are you google collab?
like this?
Yes
That version is for both AMD and Intel
Can u help me @molten pecan
Yes
yeah it was on
ah ok
If you still can't find it then I think either the folder is missing, you need to make one, or Applio is just broken like that 😭
@molten pecan help
i have mangiorvc
Yeah. https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/RVC-beta.7z download this one
Download it, unzip it and click on go-web to open it
Are you trying to make/train a model?
Or just inference?
Will it train to 700 epochs?
That one is unlimited
yeah trying to make a model, with a 12 minutes dataset
Ayo? @thin bronze level 2 !!! 
Does it have tensirboard
I don't think you even need internet to use it
I don't know about that, I don't make models
Yeah it's local
I don’t have a gpu or a good machine
Training requires gpu
Search on Google
na it don’t work tho
-collab
RVC Guides (How to Make AI Cover)
Translation by country
-commands
honestly maybe I should just try reinstalling my rvc?
Yeah go ahead
If it's still happening use my link
But weird, dont know how I was able to train models before this 3-4 days ago, and i did not delete anything and now its giving me this error
Mangio and Applio are weird like that
No idea why
I guess AI technology is still kinda primitive
errored out again 2023-11-07 21:33:00 | INFO | infer.modules.vc.pipeline | Loading rmvpe model,assets/rmvpe/rmvpe.pt
2023-11-07 21:33:01.2065584 [W:onnxruntime:, session_state.cc:1169 onnxruntime::VerifyEachNodeIsAssignedToAnEp] Some nodes were not assigned to the preferred execution providers which may or may not have an negative impact on performance. e.g. ORT explicitly assigns shape related ops to CPU to improve perf.
2023-11-07 21:33:01.2118916 [W:onnxruntime:, session_state.cc:1171 onnxruntime::VerifyEachNodeIsAssignedToAnEp] Rerunning with verbose output on a non-minimal build will show node assignments.
C:\Users\alexh\Desktop\RVC1006AMD_Intel>pause
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/RVC-beta.7z use this one then
You are the fourth person that got an error after using the AMD version
I guess that version is more broken
And the beta version is more stable
I don't often get errors in it
zamn
how long should this take? im on "loading rmvpe model"
The first time I opened it took me like 10 minutes
But that's because I have a really slow PC
aha I see I have a ryzen 9 so it should be faster ig
Yeah it should be faster for you
works well thanks
someone help :c on applio
what do i use to make an ai model?
Colab is good if you don’t have a really powerful pc
What’s the best “Available Models” I should download for UVR5?
To get the cleanest vocals possible from a song?
oh okay I see, thanks for your work btw ❤️
someone help with this pls
Hi, I have a prob, I received a 30minutes audio clean and I want to create a model with that with a .pth and .index to use in w-okada but I watched like 8 hours (seriously) of tuto and its impossible for me, I understand nothing, can someone help me please ?
Ayo? @worldly epoch level 1 !!! 
no lmao
please
desesperate
I have a record of 30min (podcast and I cleaned it)
I just want a model to put in in -okada ye
sure take your time ofc, maybe you should consider uploading a working older version so people can use that until you finish the new one
I have the record of 30min which I received from a podcast and I want to convert it in an file .zip contain the .pth and the .index, What I understand its I need to train it but Im not sure, I tried a lot of things but everything didnt work maybe I didnt do it great
What link do I use/press in info idk since there’s 20
What I have to do with 30mins audio ? That's my question, to get the voice of the character

did you find any fix?
I used the mangio fork instead
What is mangio fork and rvc
so you did not installed and used this RVC V2? https://docs.aihub.wtf/vocal-cloning-guide/rvc-v2
Also why is my shit saying queue for so long? it was 800 secs too before
same prob
i have a 3090 so it should not take this long
It kept giving me errors no matter what I did, so I stopped trying to use that specific one
Ok so how does it work? Text to speech Orr voice changer
It does audio conversion, both from files and realtime
i guess I will have to use the fork one too dont know whats wrong with this queue thing on RVC V2
No idea What that means but ok
Ayo? @fierce crown level 2 !!! 
It takes voice files and uses the AI models to turn them into another voice. It can also be used to change your voice in real time, for use on voice chats for example.
Ahh ok
still would be nice to know a fix for this queue thing
Ayo? @thin bronze level 3 !!! 
Question, Crepe and Mangio Crepe what are the differences, and which one is beter?🙏🏼thanks
usually its rmvpe for clear and good quality datasets and mangio crepe
Thank you!
#@title GPU Check
import torch
ngpu = torch.cuda.device_count()
gpu_infos = []
mem = []
if_gpu_ok = False
if torch.cuda.is_available() or ngpu != 0:
for i in range(ngpu):
gpu_name = torch.cuda.get_device_name(i)
if any(
value in gpu_name.upper()
for value in ["10", "16", "20", "30", "40", "A2", "A3", "A4", "P4", "A50", "500", "A60", "70", "80", "90", "M4", "T4", "TITAN"]
):
if_gpu_ok = True
print("Compatible GPU detected: %s" % gpu_name)
gpu_infos.append("%s\t%s" % (i, gpu_name))
mem.append(int(torch.cuda.get_device_properties(i).total_memory / 1024 / 1024 / 1024 + 0.4))
if if_gpu_ok and len(gpu_infos) > 0:
gpu_info = "\n".join(gpu_infos)
else:
raise Exception("No GPU detected; training cannot continue. Please change your runtime type to a GPU.")
gpus = "-".join(i[0] for i in gpu_infos)
how fix
So in other words, Mangio Crepe generally speaking better than Crepe?
this was in GPU check
yes "crepe" is an older version of mangio crepe
Wait I need help I tried to transfer a Weights.gg zip file into the model converter for Google Collab v2 and it didn't work
It said error module not foujd
Is there a way around this
Please help
Ayo? @sharp island level 5 !!! 
what's your GPU
How do I check
is this colab ?
Sí
any idea on this? @sand iris
¯_(ツ)_/¯
How do I fix my res? It only goes up
will pcie generation change rvc performance
like
ive got shit for gen 2
for 3 3060s
will the performance drop to gen 2 be more than say
10%?
any tool to make a voice into instrumental?
you want to make a voice into an instrumental?
hi hi
Guys hi, how do I make the voice very similar? I do with 300 Echo settings, but the voice is not very similar. Need more? How many files to upload approximately and for how many seconds?
Can anyone help me set up a voice changer?
how much voice files are enough for rvc traning
yo
should i repeat a soundbite of him speaking clearly for like 6 sec or should i cut out the stutters in audacity
i have 7 minutes as a dataset
i need answers people
cut out the sttuters maybe
i'm new to traingin too
hmm aight
Please refrain from posting about this topic. Pedophilia is not allowed on the server and is a permanent ban offense. AI Hub does not condone training NSFW content and NSFW content is not allowed on this server. This is a verbal warning.
my guy i'm not the pedophile
i dont diddle little kids ok
fine whatever ill delete the messages good god

What did they post 
dms?
true i would be a scared mf too
not sure if this is the right channel,but my voice changer doesn't work in fortnite. i have my device that others hear set to virtual audio cable but that doesn't seem to work. it works in discord just fine
thx
I can't send pics ugh
But I keep getting an error
In rejekts easy rvc
When I inference
It keep getting an error at step 16 to 18
About the output audio
And idk what to do about it
Please help lmao
why
Ayo? @paper wave level 4 !!! 
Tag.me.if you do
Idk man
It won't let me
Ig I need to tall more or smt
Please send me an easy inference collab if anything
Ayo? @hushed steppe level 1 !!! 
maybe you want to use a computer
I've never had issue
I've been using rvc for a year
I'm good, just not sure why I'm getting this error
Whats Mangio
Do u pay for Colab pro?
No
Doesn’t google colab restrict rvc?
Ayo? @iron sandal level 2 !!! 
I use a 30min soundtrack, yesterday I managed with a colab to do something not too bad, the only problem is that the colab closes after a while, I put 10 epochs in the config (about 10min treatment) the result is not bad but I would like to do more, let’s say a hundred for test but the colab closes (because sessions have limited times) there is a solution? apart from paying $50 for premium, importing into a local thing or something like
hey can i get some help sorting through guides of how to start making models myself?
i see the "create a model with rvc disconnect" but it talks like i already know how to prepare dataset files, im completely new to making models and need help bad so help would be appreciated
since when does Rvc 2 works so bad
im just trying to launch all the stuff but i keep getting errors saying like ''this code is not allowed on free tier''
to do a dataset u will need one/or more files of ur character speaking/talking/singing.. u can use UVR 5 for the vocal extraction.. then puts the wavs in a folder.. zip it and named it as u want
watchu trying to do??
when i launch this page
sometimes i random get a time out
with a error saying something like ''your running code that isnt allowed on free trial''
THIS
i think ur using an old colab
ok thanks
i will try it later after my school stuff
can u tag me in a couple seconds
i will leave this channel
so the ping stays
i will remember for later
ok im leaving this chat now
thanks
Hi.
My Python doesn't find 'soundfile' module to run GUI. What should I do?
alright done with making the dataset, now what should i do?
install it
im not wrong QWQ
now u can follow the guide
which guide sorry, theres many
lemme link it for ya
heres it
u can also dm me for any problems
How? It doesn't work by "pip install soundfile"
thank you very much, and absolutely will yeah
np
Would you help me in the way of installing?
im not that good in those things (locals)
Thank you for your response
Ayo? @soft remnant level 1 !!! 
u trying to train locally??
this tells me how many seconds it took doing such during voice conversion. what does each thing mean
i know infer means inference
but what does npy and f0 do
someone got hacked
@moderator
I just extracted the zip file and add 'hubert_base.pt' in it (the main file is the root file mentioned in official guide on Github, right?)
If I haven't done anything wrong, what's the next move?
wt r u using?
wht r u suing
rvc-beta local
isn't there a repo that will to live voice changes, voice change files an allow u to train ur own voice models all in one ui
yesn't
https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/tree/main
Has both training and inference on the same repo
and voice change in a separate script
But the voice change part doesn't necessarily work
yes i have that one ... it works pretty good but the live voice change doesnt necessarily work
complains about the sound device index being out of bounds ? Yeah
yes
okay
I need to train a pip model and annoy people that complains me for picking a character
real quick how do i know the total epochs of my dataset?
like how can i check
cos i’m using disconnected and it’s asking me but i’m not sure
i’m using the rvc v2 disconnected google colab to make my own model
in one section it says to set the total epochs, and the guide says to set it based on the dataset length, but i’m not sure of the length of my dataset
you don't really need to set it based on length
but anyways
do you still have the dataset zip/folder in ur pc
yup i do, like with the audio files
Ayo? @worldly light level 2 !!! 
So, if it's the folder, select all of them (Ctrl + A, I believe), right click and check properties
alright done that
There should be a part where it tells you how long all the audios are, so that's how long your dataset is
again, this varies a lot and it depends on a couple of factors
but I'd say, yeah, go for it in a first session
Just make sure to export models once it's done
and train index
alright awesome thank you so much i’ll try that, and yeah for sure i’ll make sure to do them
np ;)
can someone help me with voicemeeter
apprantly from my friend perspective it sound very robotic and he can hear himself
Can someone give me an easy inference things
activate the echo and sup2 options
wait im abit confused
so is the audio it transfer that other people can hear the output from the app or it have its own other output?
cuz rn if i were to use just the app output its fine but if i try output manually it detect my own voice not the one in app
only the output from the app
ahh
if it doesnt then ur input/output settings are wrong
so the audio is come from the on u put on gain out right?
Ayo? @neon mortar level 1 !!! 
in other word a one way correct
what
is it from that out audio that it extract
or this have seperate u can deactivate out
i think i figured it out
Sound goes through VoiceMeeter
Just a volume slider
so like the input voicemeter output voicemeter for that app only and other audio is aux input to aux output is it not correct
input in other programs should be the output of wokada
so the voicemeeter is input on discord, and in wokada its output
yeah
one last question i have for anyone, i’ve just finished training and it worked, so on google drive should i download the D and G .pth files too? or do i not need them
find a folder called weights and download your model from there
Help mee!
no idea on local
Press any key to continue
that doesnt help...
Try downloading this one
It's another version of RVC
Much more stable in my experience
Uhm I don't know what you are talking about
Have you loaded any models?
yes i downloaded a markiplier model to test. If i am correct i put the MarkiplierV2.pth in the weights folder and the Other folder in the file in the logs folder
Mhm
Does your PC use Nvidia?
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/RVC1006Nvidia.7z here's RVC for Nvidia
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/RVC1006AMD_Intel.7z and here's AMD_Intel
Maybe if you use one them it will work
I don't know how to fix that 😭
@fierce crown
i use nvidia and thats why i used the "RVC1006NVIDIA" one
and that one doesnt work
And this one?
Did it do so?
havent tried it trying to try and should i? I am trying the one you mentioned first
this
Yeah the beta one is much more stable than the others in my experience
What i mean is should i use the one i just mentioned or the one i already was
Ayo? @fierce crown level 3 !!! 
If one is not working use the other one
That's what I am saying
Try and see which one does
oke doke:)
one last question. Could it not work because i installed an used the Easy gui installer
i cant get it to work it always says an import error
That's odd
Guys, what is acceptable amount of epochs for 20 minutes of really good audio?
Yeah try uninstalling Easy GUI
Also what is an Epoch?
Try asking on #📑│making-models
don't use easy gui, and btw try use mangio or applio for inference that has mangio-crepe option besides rmvpe
could you give me a link
and whats best mangio or applio
whats best mangio or applipo
both are good, you can just try
What does the ML stand for?
In your bio
Machine learning prob?
machine learning (simply the AI we know)
hey yall
so when i load a model and then go to load the config into so-vits-svc it seems to crash every time no matter the .json file
does anyone have any recommendations on how to solve
@half cove 😭😭😭 send the gif
Who had the gif?
-svc
oh boy how clueless am i
Embed failure! Give me all of your points now
fuck you discord with the temporary link storage
😭
whats the best method yall have found for doing this because im stuck using so-vits and it seems like people have more new versions w better UI
rvc2
hahah
where can i find
nah im a producer of 11 years im abt to make some heat LOL
Ayo? @tawdry island level 1 !!! 
watchu need bro
use models / train a model
currently i installed so-vits via anaconda on python 3.10 and when i load a .json cofig file it crashes
AAAAAA
wtf that is
what r u trying to do??
-svc
FUCKING EMBEDS
im aiming to take my own vocals and use a model on them
just dont know the best method
open to other ideas tho
thats cover (inference)
i sing etc
thank you!
np
okay last question how to i use a model i have? just paste the link to the site i downloaded it
lets say this one for example
@copper pebble my knight in shining armour
Future | Dataset length: 10mins Sample Rate: 42K Trained With RVC V2. Trained on [RMVPE] WHEN YOU USE THESE MODELS YOU MUST CREDIT ME YouTube Link (MUST US | RVC Model, free download. RVCv2, SO-VITS-SVC, RVMPE, and more. Official AIHub Discord. | Artist, RVC v2
try this one
copying links from weights.gg its a lil bit complicated
youre so awesome so i see the "choose the model window" however the dropdown doesnt show any options. do i click the "download voice models" tab and then enter the link to DL?
which one?!? https://github.com/Mangio621/Mangio-RVC-Fork/releases or https://github.com/Mangio621/Mangio-RVC-Fork wait isnt these the same thing
if u didnt arleady yes
okay i got it DL but there is no option to drag or import the model file, thats the only point im hung up on
u just need the download link of the model u want to use
the dropdown in the "1 choose the model" doesnt really seem to react when i click on it
ahhh
and put that into the "choose the model" field'
?
no
im sorry for some reason im struggling usually im not this slow lol
Ayo? @tawdry island level 2 !!! 
what
then go back to "Inference"> refresh>choose your model
can you help me you seem to know alot
and then choose it in the menu
aaaa not again
oh no did i use wrong link
what links work best
suo?
im not dead set on using this model im open to others
hugginface ones
no no
okay is there a link to browse theirs easily, it looks like this site contains a lot of stuff lol
u can upload it on ur hf account
wait a momnt
or anything else'
man
youre awesome
Thanks
i need help with this i have the file "Infer-web" so idk why its saying it cant find it
go to hugginface> click on your avatar> +models> name ur model> choose OpenRail as license> files and version> upload the zip> scroll down> commit changes and wait
"upload the zip of the model that u want"
It's been going on for a while already...
It's annoying
do u know this??
that should work
will i see it now as an option in my dropdown or do i now have to use my upload link?>
Yeah??
.
Hf??
Nope
it will show a red box in files version as soon it finish the uploading
This is the reason why I use gradio
fr
Applio or Mangio are a little buggy and complicated to install
With gradio you just unzip the file and you are good to go
True tried to install it on another pc still broken :/
Idk why it's so unstable 😭
okay so its uploaded, but now how to i get that into the model dropdown from earlier
https://huggingface.co/zenophaze/future2/tree/main try it with this link
u mean uploaded
?
the issue is it wont download even tho thats my upload link
it uploaded sucessfully
sometimes when i train a model it disconnects and reconnnects at the 11th epoch
try this
even tho i think its another Future's model
what is an epoch
damn same thing i think
the file name shouldn't have spaces (escaped as %20) and/or any special characters
i gave him that dw
hey julia thanks for helping too
some models of mine works tho
what models do you have
how to open a voicechanger
You are welcome!
its converting rn imma test it
Is your banner what I think it is
you guys are a really nice community literally like you deserve a salary for this if it works lol
i give up:(
okay
rickie or julia do you guys have any music background or what do you use this stuff for?
i guess its killer for memes too
NO WAY
I am just really passionate about music and right now my new goal is to learn how to make studio quality AI covers ^-^
Congrats!!!
Still learning!
No way that's insane 😭
i could legit answer any qyuestions at all yall have been so fuckign nice
i change it
Ayo? @daring oak level 1 !!! 
Oh my god where do I even start
idts
gonna get off guys
gotta touch some grass
What's the best reverb plugin for FL out there?
It's okay if it's a paid one
I didn't like the valhalla one 😭
THANK
Ayo? @tawdry island level 3 !!! 
you
how can i make a custom voice? like, i wanna to make my friend voice
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- 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
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.
Oh my god where do I even start
@molten pecan literally DM anything i am happy to answer anytime
what?
Ayo? @magic dagger level 1 !!! 
Oh awesome! I'll consider it then
Thanks!
There's a guide on #1159513888199540817 on how to make a voice model
iirc
where is on #1159513888199540817 ? and ty @molten pecan
Good luck!
tyyyy
ping me or dm me for any questions/problems
Can I do that too?
yh
How do I treat my dad's diabetes?
imma send what yall enabled me to make with your pure essence of knowledge if thats allowed
pizza
rickie thank you as well dude
i really do appreciate yall
made the beat from scratch and just freestyled this
used the thing
yall are goats fr
ty
guys
how did you make that link i sent earlier
able to be downloaded
for example i uploaded another model
but it gives me the error when i copy my link
so how did you copy it differently @copper pebble @molten pecan
I search for the model on huggingface and go to Files
And I copy the URL
right click/hold on the red box and then copy link
ur jesus christ of the internet
nah he is debatable youre like some real figure thats godly
both of u
also @half cove wym you thought i was using an acapella?
does anyone know how to split isolated vocals to have a audio file per singer
With UVR5
It's free
For the clearest results, first remove reverb and then isolated vocals
In UVR5 there are two algorithms for isolated vocals
hp5 karaoke is for harmonies and hp6 is for backup
@rich bobcat what's the difference between crepe and mangio-crepe?
but would uvr5 be better then just using audiostrip's VB-Splitz V2 or VB-Splitz V3
No idea. Never heard of it
Disable your windows firewall
didn't work T_T
@molten pecan can you help me? :<
Sorry I don't know how to use Applio
oh okay
I think that means that your coputer is not leting the code open the url
computer*
try to restart your pc
i do it ;/
the first time I turned it on it worked and suddenly it stopped
Happens to me
im really sure is not, but its kinda like chance
sometimes works, sometimes not
i dont really have the solution for that
mangio's rvc supports amd cards?
Ayo? @naive fossil level 4 !!! 
actually having the same problem
when it was working the first time lines kept popping up
now it just says all keys matched successfully
is that normal?
cause now nothing works
https://gyazo.com/d04cabfe9f020f1f217b09eb73659428 mine's just this
oh
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- 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
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.
HIii one question
I need help, RVC not detected my gpu
my gpu is a rx 6700
any solution ? 😦
I use the cpu but the work its too hard for he, in moments i have lag in my pc and i think, in long time maybe can demage the cpu
srry my english is not the best
My cpu is a Ryzen 7 5700g
Anyone can help me with this Apollio error? FileNotFoundError: [Errno 2] No such file or directory: 'assets/pretrained/D32k.pth'
Cannot make models on AMD GPUs
NoooOooO
have any alternative for rvc ? :c
Ayo? @thick peak level 2 !!! 
idk if i should be asking here but, is it normal that a rvc huggingface stops working on everything after you stop using in a few hours? (incase anyone curious im using Ilaria RVC)
althrough idk if it happends because im duplicating/cloning to use it privately
**Method 1 🔎: ** The search accuracy is 100% 💕
Upload the <t:1688381843>
Server uploaded: AI HUB Brazil
Channel uploaded: AI HUB Brazil Post
Profile Owner: lh658
LH658
lh658
740
BR 🇧🇷
RVC V2
88 MB
N/A
100
0
skill issue
I've never used applio, but it looks like you just need to download some pretrained models. There should be dlmodels.bat file in ./lib/tools. Just run it and it should download all required models automatically
Ayo? @near flame level 2 !!! 
can anyone help with mangio?
Ayo? @fresh shadow level 1 !!! 
Do you have some error or just need help with installation?
Hola buenas
una consulta, no encuentro la forma para poder agregar el modelo y el index
Aprieto para poder agregar el el modelo y el index pero no me responde
**Method 1 🔎: ** The search accuracy is 100% 💕
Upload the <t:1688266860>
Server uploaded: AI HUB Brazil
Channel uploaded: AI HUB Brazil Post
Profile Owner: dimitri446
Dimitri
dimitri446
686
ES 🇪🇸
RVC v2
136 MB
N/A
310
0
Translation:
One question, I can't find a way to add the model and index
I press to add model and index but it doesn't respond```
You're welcome people who don't know spanish
I'm pretty sure gTranlate would have butchered this
Yo I'm confused. How does the All TF2 mercs (RVC v2) model work exactly? How do i pick a specific mercenary that i wanna use?
@acoustic thorn really sorry for the ping But i see that you have made the Mangio Rvc fork and i got this Error so i cant use it! I really hope you can help. Thanks!
It’s not an error but it’s just not working?
It says all keys match successfully but nothing is coming out of it
no idea I think you get a model from somewehre you like and then its like a voice changer
Ayo? @fierce crown level 4 !!! 
I’ve had it work the first time but when I restarted my pc it just stopped for some reason
Hey sorry do you know a good like tutorial or something to help use these models> Im new lol
Depends on what you’re using
Ive dowload a voice model from this discord I just dont know how to use it.
What error do you have i have a Like infer-web thingy
I’m not even sure if it’s an error, after I press start it says keys matched successfully on the prompt
The first time I got it to work I’d get non stop lines
Now it’s just that
What promt and what model are yoyu usinng ( Mangio Applio or another?)
I was using mangio rvc
https://gyazo.com/d04cabfe9f020f1f217b09eb73659428 but this is all i get now
Did you have any prvious error I have this error on 2 computer and it doesnt make sence
sense*
i had a bsod before this
I am trying to contact Kalomaze but they aint answering🥲
like is it even an error?
Ayo? @fresh shadow level 2 !!! 
I guess it aint opening Gradio or anything. 2 helpers dont know what to do to help me so i have no idea what to do now
are we having the same issue?
is that on the directory folder?
idk its the Go-web application
you trying to train?
nope just trying to use voice models for fun
if you're trying to open the voice changer you open go-realtime-gui. bat file
and make sure to put the .pth file on the weights
and indenx file on logs
gimme a sec
did you try using the bat file?
yesz
i'm stumped too
@fierce crown if u just trying to use models why dont u do it with ilaria rvc??
can i ask for help too?
hope i can help
it doesnt work and its slow as far as i tried it
https://gyazo.com/d04cabfe9f020f1f217b09eb73659428 i got it to work once but the second time this is all i get
and nothing comes out of it
did u duplicate the space??/ splitted the audio in more parts??
is it supposed to just end in all keys matched successfully?
yeah
.
same question
i did it still takes like 350 seconds
??
why r we talking abt epochs now??
Oh yeah and also i didnt understand how to use a custom model
what is an epoch
you could try okada
hi. new here. is there a way to create RVC models if I don't have a NVIDIA GPU? I'm on Mac
i think i saw a google doc based one not sure if that would help
sorry google drive
Maybe i used the wrong one? Could i get a link please?
watchu want to do???
the google colab doesn't work anymore
not sure if you can then
?????????
i got it to work
wtf
my brain
not sure how i got it to work again
thanks guys
lol
yeah i have no idea what to do and where
Hi
u wanna do covers using models? / u wanna create models? / u wanna change ur voice realtime?
all 3 i guess
an epoch is a kind of unit of measurement for patterns, it depends on the length of the audio
Also for training can I use one I’ve downloaded to add more into it or?
Ayo? @fresh shadow level 3 !!! 
i am new to using hugging face so how do i install it prepackaged
Do I start from scratch
I’m trying to train a gura voice file to be able to speak Japanese
Oh sorry thought you were talking to me
Since it sounds awkward transitioning from English to jp
oke
Would like to use it as a practice dummy
I was wondering since there are ways to make them sing if I could use that as a dataset like making her do japanese covers and using that for the dataset?
Sorry my English isn’t good
I see! Thank you!
i'll give it a try. thanks!
Think it works now! Thank you soo much spent way to many hours on this. Sorry if i am really stupid, I am new to this stuff. Thank you again!
Just wondering when pressing refresh voice list How long does it normally take is it normal it take over 200 seconds?
Ayo? @fierce crown level 5 !!! 
Takes around 2 seconds for me
ahahaha 850 seconds normal-
hi
anyone help me?
i have 18 minute sample voice
How many epoch numbers should I set for training?
this is a good tab
i usually do like 350 epochs for 18mins
Aye @copper pebble would appreciate your expertise on this:
This may sound stupid af but I recently came back to this. I downloaded a model that has a .Index file extension.
I'm used to the original so-vits-svc-fork which has .pth model & .json config.
The models I downloaded only have .pth & .index.
Wut do? Thanks!
thats so outdated that they made a meme with it bro
Damn 🤣
goat ty will look into it 🐐 🐐 🐐
u can ping me here for any questions
thanks
is it valid even if it is in 1 sound file?
ofc
i might save this for when i train more models
i need help with errors
screenshot pls
sent
im trying to make a model and it has lots of static and just sounds bad any reccomendations
Ayo? @hasty sigil level 1 !!! 
dm pls
do my samples for training need to be mp3 or wav ?
Can't send images in here
I haven't done AI voices in a while after the server was originally deleted, what would you recommend for pitch extraction algorithms?
(Pm, harvest, dio, crepe,crepe-tiny,mangio-crepe,mangio-crepe-tiny,rmvpe or rrmvpe+?)
Ayo? @tough drift level 1 !!! 
these settings is good?
I see
Alright thanks, I forgot which ones were the good ones
wav
sample rate 40k/ save frequency 20/
u using the old one btw
tysm
dunno if that is safe
i would say to ping me for helps but its late night im my country so ik about to get off
ok :)
oke and wich one first the train or the index?
hi
train
vouce changing or covers??
Ayo? @obtuse stream level 2 !!! 
now good?
-
for voice changing ask here #🔍│help-w-okada
-
for cover heres a simple guide with a simple hf space
https://discord.com/channels/1159260121998827560/1167844377046040769