#✨│ai-help
1 messages · Page 83 of 1
still not working
C:\Windows\System32>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897
thats the error
wehn i open go bath
the voice chaanger works but not the rvc
@half cove can you help em out?
use 7zip
for some reason winrar and the normal windows one messes up
is the model that's been training for 2 hours lost now?
i have the wirnar folder
unzip it using 7zip
the rvc folder
doing it rjn
Ayo? @brittle wing level 3 !!! 
if ur using that one it will be called Mangio-RVC-v23.7.0_INFER_TRAIN.7z
unzip that using 7zip
then try it
k
no error
but when i opem it
it looks like this
it wont open
ohh i have to wait
To create a public link, set share=True in launch().
you dont need public link
C:\RVC\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897
Found GPU NVIDIA GeForce RTX 3070 Ti Laptop GPU
Set fp16_run to true in 32k.json
Set fp16_run to true in 40k.json
Set fp16_run to true in 48k.json
Use Language: en_US
Running on local URL: http://0.0.0.0:7897
To create a public link, set share=True in launch().
i have to wait or something bcs nothing is happening
http://0.0.0.0:7897/
@brittle wing copy that URL and paste it in your browser
not working
I give up
what does it show 
xd
i assume it is
yep 
always try keep training under like 2h and save every 30 min
good advice
Ayo? @rain bronze level 2 !!! 
it's better to have more control, so ig mangio/mainline would be better so you can control the filtering of the vocals first, then mix the vocals correctly later
https://aihubdocs.github.io/en/ check the Ilaria RVC usage guide
Last update: Feb 10, 2024
some songs just break for some reasons
it maybe not as perfect as manually doing separation & dereverb using UVR/MVSEP and denoising & cleaning it
or that if you don't have a good PC yea
yeah it isn't
you can't really fix it tbh
it's a good way to make quick covers tho
yw :)
What do u do once u download a model???
Ayo? @dreamy cove level 1 !!! 
Hello, does anyone know how can i create a friends voice in ai with huggingface or something like that? i can't seem to find something on youtube/google
RVC Guides (How to Make AI Cover)
Translation by country
ya can't train models directly in huggingface, but you can do that in colab, just check https://aihubdocs.github.io/en
Last update: Feb 10, 2024
whats a voicechanger that will let me use my own models
one channel under this
why does it keep asking me i need a nvidia gpu, when i made a ai cover before..
Where can you show your covers?
sure, i made it on Gradio
i can send you the photo
Uhhh what's a good rvc model maker for windows
And is there a guide and shit like that
Anyone know how to fix this error?
AttributeError: 'NoneType' object has no attribute 'tobytes'
This error comes up when I try vocal inference
Ayo? @humble zephyr level 1 !!! 
I'm still getting this error even though I reinstalled RVC V2 fresh
Traceback (most recent call last):
File "C:\Users\james\OneDrive\Desktop\bg3-ai-voicetoolkit-main\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\routes.py", line 437, in run_predict
output = await app.get_blocks().process_api(
File "C:\Users\james\OneDrive\Desktop\bg3-ai-voicetoolkit-main\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\blocks.py", line 1349, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "C:\Users\james\OneDrive\Desktop\bg3-ai-voicetoolkit-main\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\blocks.py", line 1283, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "C:\Users\james\OneDrive\Desktop\bg3-ai-voicetoolkit-main\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\components.py", line 2586, in postprocess
file_path = self.audio_to_temp_file(
File "C:\Users\james\OneDrive\Desktop\bg3-ai-voicetoolkit-main\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\components.py", line 360, in audio_to_temp_file
temp_dir = Path(dir) / self.hash_bytes(data.tobytes())
AttributeError: 'NoneType' object has no attribute 'tobytes'
It keeps mentioning Gradio... I just noticed
@red kayak
Have you encountered this error before?
move out of onedrive pls
How?
we should really sunset mangio
only causes issues and the code is somehow worse than the og rvc project
Wtf
How do I stop windows from putting every fucking thing on my desktop into onedrive
How do you use jman one its not working
okay well look
drop the mainline folder here rq pls
Does the mainline rvc have rmvpe?
Ayo? @ionic canopy level 5 !!! 
it does
and delete mangio
make sure u run it on ur desktop and that all the folders are on the C drive
and not on ur onedrive
Onedrive is located in my c drive tho.
it does
Found jt.
Couldn’t tell because of the chinese lol
Traceback (most recent call last):
File "C:\Users\james\OneDrive\Desktop\RVC1006Nvidia\infer\modules\vc\modules.py", line 250, in vc_multi
os.path.join(dir_path, name) for name in os.listdir(dir_path)
NotADirectoryError: [WinError 267] The directory name is invalid: 'C:\Users\james\OneDrive\Desktop\1_1_1_1_Ed Sheeran - Photograph (Lyrics)(Vocals)(No Reverb)(No Echo)(Vocals).wav'
why is the directory name invalid?
give it the folder that holds the audio files
not the audio file itself
Ooooh
(+ there's spaces in that filename, it will freak out)
I’m an idiot
I see
Let me fix that rq
avoid spaces in any of the files/folder names at all costs lol
What are the differences between pretain ov2 to rin_e3
broo my rvc is so bad whenever I talk in Fortnite idk how to fix it😓
does anyone know it to make ur rvc faster?
What do I do if the new pre-trainings only let me train at 40k but my model is 48k
-colab
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
Weird. I don't hear anything in my voice model
I think it's broken
Here is my model
Can someone test it out to see if it is really broken?
how do I make my rvc gui faster?
All the voice models don't work except beatrice, can someone help?
hey
Can someone help me?
i have the error
Error downloading file: Request failed with status code 404
in applio
when installing
I would like some assistance 🙏
Send cmd logs in #🔍│help-w-okada
check https://docs.applio.org
see if theres error fixes for that
Where can i get my cmd logs?
command line, tab, whatever its called
just send screenshots of it there
when I use the rvc voice It uses my cpu power and not my gpu
if you're on AMD, you gotta export to onnx
and even then, it will still use more cpu because idk
DirectML version is goofy
ohh okay
https://rentry.co/VoiceChangerGuide#uploading-models-amdintel-gpus-and-fixing-its-laggy-issue
in case you need to know how to export to onnx
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful Links - You need to follow this if you are on an AMD or INTEL ARC graphics card
Frequently Asked Questions
Jump To The Important Sections
Other Links
Download NVIDIA
Download AMD/INTEL GPUs
Download Mac
Online Hosted Alt...
can u run rvc on a macbook
?
Ayo? @haughty gazelle level 3 !!! 
im using rvc v2
ah nvm
it might be an AMD issue
cuz RVC doesn't work with AMD cards currently, it uses cpu
for training, no
for inferencing yeah
how do u run it
i download the files but then it says i cant open it or it doesn’t let me click the file
you probably downloaded the windows version then, bat is a windows file only
idk how to do it
RVC Guides (How to Make AI Cover)
Translation by country
-local
- 🍏 Applio, 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
- GPT-SoVITS, by RVC-Boss 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.
Since the update, rejekts' GUIless Colab doesn't seem to have the option to save models to Drive automatically anymore; rather, it just saves to the local Colab directory, which is a big problem if you suddenly get disconnected cause it deletes all the files as well. How do I make sure I can continue training easily even when I get disconnected?
Meanwhile you can use rvcdisconnected instead
That one also doesn't have auto-save to Drive (at least from what I've been using)
I just play it safe and stop training every checkpoint and save to drive manually (although I skip the first 30 minutes of colab usage to not waste too much time, then I start saving every checkpoint whether it was every 5 of 10 epochs)
Dead kaggle updated something so need to fix it again.. also no internet for me atm
Is it possible to use RVC in a way you don't need the Internet or the server to use it ? Like, run it 100% locally?
Yes
RVC Guides (How to Make AI Cover)
Translation by country
does anyone know how to fix this error? i was told to use retrieval based voice conversion instead of okada for these particular voices i was given but when i click "start audio conversion" i get the error and the window freezes
PING ME
corrupted models
also are you doing real time or file conversion ?
w-okada is just more reliable if it's for real time
@sand iris
Corrupted models: I’ll have to talk with the owner about that cuz they gave me the voice model files and they said they never had problems.
Real time VS file conversion: how can I tell which one I’m doing? I’m unsure on which one is used by default cuz I haven’t been able to change the settings.
train ai model?
do you want to voice-changer, or make AI covers / do speaker conversion on audio files ?
Voice changer
okada would be better here
regarding the error, It could be the "pipeline" models that are corrupted, and not necessarily the voice model you downloaded
how do i upscale a bad quality video to 4k?
Alright well the thing about that is I tried 2 of the 3 models I was given and when I uploaded them in Okada they kept appearing as “blank”
Borked?
does anyone know
We mostly do voice changing in here, not sure who has done that with videos here
either videos or photos
Ayo? @mossy grail level 1 !!! 
rosy why the eggman logo?
What’s a borked model mean?
broken
Ah right sorry 😅
Cuz it’s my version of a skull 💀
if it gets you that "Failed reading central directory" error, it's unless not a proper model file, or it's somehow not downloaded fully
How could it not be fully downloaded?
When you download something wouldn’t that mean it gets downloaded properly?
If it didn’t then shouldn’t the download fail all together?

my bad is it against guidelines?
I’ll inform the owner of the voice models
Yeah it is, nothing personal 😅
Is there a guide or a place I can see how ? Or show what I need to do?
I mean it depends on how you downloaded it. If it was through a web browser, then yeah you shouldn't have the file unless it was downloaded funny
oh ok lol
other ways don't guarantee anything
(we get plenty of those on okada since it sometimes fails mid-download)
Well google chrome did mention something about the files being suspicious after downloading or something like that
eh that's usual
I do have Norton tho which protects me from dodgy files and stuff
It didn’t take action
it sees a .zip file which has never been downloaded before and freaks out
Yeah I got a .zip file
which contains a .index and a .pth right ?
Just .pth files no .index apparently which I should’ve asked about
It was 3 .pth files
Typo
Lemme check
also do any of them happen to end with _D.pth and _G.pth ?
Also while I do that I will also mention that the RVC GUI asks me for an .index file in order for the voice changer to work
ah. Well w-okada makes it optional
Yeah
No
Just names
Usually when it gives this error, the files are significantly smaller than the usual model files
which are roughly ~56MB
but here it seems fine... I guess try other models and see if those work
just to 100% make sure
Alright I’ll do that now
Ayo? @simple lichen level 6 !!! 

@sand iris ok i tried my other voices and they seem to work so i guess the files i got are messed up somehow but the voice doesn't sound good at all by either sounding nothing like it's supposed to and it's cutting off early and repeating stuff
it doesn't have some of these problems in Okada but even when i use that the voices don't sound the way they should either which has given me issues for months
tried messing with the pitch setting ?
i thought female to female was 0
sometimes tweaking a little can help
depends on your vocal range
if it sounds nothing like it though, maybe it's input mic quality / RVC being bad / the voice model being bad
ok i redownloaded the files one by one and they upload in okada properly now
so that's a start
i dont think the voice model is bad cuz he gave me a short sample of how they all sound. i am using a headset mic tho which i've been informed is a bad idea cuz i'm on a laptop
noise suppression can definitely help a lot
background static noise seems to make RVC freak out
right
i have one headset mic that picks up background noise and another one not so much but i am upgrading to a desktop for streaming anyway so i can get a proper mic then
i just hope it'll all work the same as the audio samples cuz i've been trying to make the ai voices work properly since like august last year and is really bugging me
i can send u one of their audio samples and one with how i currently sound to give u an idea
did u want that?
output 3.wav is mine from okada
1 (2).wav is the sample from the person who gave me the voices
other things i notice is the way the voice said "speak" with that K is something i can't seem to do when i use the voice changer so i guess that's cuz i'm using a headset mic
Which folder I need to download for Mac someone can send ?
hey does anyone know how to isolate two separate voices in mvsep and what model to use?
skip the first 30 minutes, as in you don't start saving stuff until it's trained for 30 minutes?
which software you using ?
MMVCServerSIO
it only lets me use "beatrice" and i cant use the other default ones
ok. can you screenshot the command prompt that's opened with MMVCServerSIO ?
great
yeah all good, got the error msg now
you'll need to delete the folders that are called pretrain and model_dir and then relaunch
it should redownload everything it needs
if it errors again, let me know, for some it just never downloads
okay let me try it
@sand iris I have these files if you’re still able to give advice
oh oops, completely phased out and forgot
😅
what are your settings currently on ?
might be because it's struggling to process in real time
No worries
I believe I was on the same settings that the server bot mentioned
How do I bring up that command again?
⠀
Settings for Nvidia GPUs 
F0 Det.: rmvpe (suggested for all series)
RTX 40-series: 80-96 chunk | +16384 extra
RTX 30-series: 96-112 chunk | +16384 extra
RTX 20-series: 112-128 chunk | +16384 extra
GTX 16-series: 128-192 chunk | +8192 extra
GTX 10-series: 128-192 chunk | +8192 extra
Advanced Settings
Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low
⠀
nvidia or AMD by the way ?
and it's set to use your GPU right ?
Yes
(as in you see the GPU name on okada)
should be good on that then
fired i fixed it
some users mistakenly download the amd/intel version
but my voice is soooo laggy
Yeah I download the cuda version right? Not directML?
yeah that's the right one
maybe try the settings shown above ?
To update it do I have to re-download the latest version?
Oof
Haven't been any improvemetns lately though
Ok
I don't think it's worth anything
Alright
for the time being
So I guess my sample doesn’t sound like the other one cuz of my headset mic and laptop maybe?
oh that plays a big role yeah
if you're using a poor mic and speakers, it'll sound bad and possibly echo because it's re-processing the audio that's leaking
I just don’t wanna find that I finally have my new desktop and mic etc. and it still doesn’t work properly
Any mic’s you recommend for this?
¯_(ツ)_/¯

it worked broooooo
Ayo? @long sphinx level 2 !!! 
thank u sm
No, 30 minutes is the shortest GPU session from colab after doing searches, it's basically a general safe spot, although sometimes, I do start saving every checkpoint a little before 30 minutes
so what's your process here? I don't exactly follow
Usually it's about an hour though before colab kicks me out
how do I check how long I've been running the GPU for
Oh there's no way at least from what I can find, you have to remember the exact time the first cell starts running
During training, you can check the time of when the first epoch of the session finished (probably in UTC time) but since it takes a few minutes to set up everything else even when just resuming, I just keep track of when I first ran the cell
if I recall correctly you can hover over the cell and it'll tell you how long you've been running it for
RVC Guides (How to Make AI Cover)
Translation by country
why does the "startHttp.command" not send me directly to run the system? it just gives me a bunch of options for apps instead :/
that's for the mac version... if you're on windows you got the wrong one
what should i do if the model i'm trying to train is gonna need well over 1k epochs, but colab doesn't go that far? should i be trying gradio? free user here, for context
Ayo? @hardy hare level 2 !!! 
ive been using rvc disconnected, but my past few attempts to even train this model to 1k have failed due to usage limits or because i stepped away at just the wrong time.
guys i need some help cuz i dunno where to start im so sorry
is there any video i should see if i want to use character voices?
India???
When it comes to datasets, is it okay to have some audio samples significantly longer than others, or should they all be around the same length?
yeah that's ok
as long as they aren't like
completely different 💀
but that's more about the audio contents and not about the length
Cause I found an audio sampling that has a bunch of voicelines clean and with almost zero silence in between them that stretches on for 5 minutes.
Meanwhile other samples I have are individual files between 1 and 10 seconds long cherrypicked from various sources
Yeah sure that's fine
You should try that then
Hmm? Try cutting the 5 minutes of audio into individual samples?
Nah, I just said try training those samples all together
Ah alright, I'll give it a try thanks
You're welcome! Let me know if it sounds good 🙏
⠀
Download for Nvidia GPUs 
Version 17b cuda
Download for AMD GPUs 
Version 17b directml
Download for Intel GPUs 
Version 17b directml
Download for Mac 
Version 17b Mac
⠀
⠀
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.
⠀
You can stop training anytime and continue where you left
@proper shale how does the acc switching work?
just make sure the D/G files have been properly saved before stopping training
honestly idk
never tried it
AMD version
that's the only ones that can use your gpu
yeah, but what about the model needing more than 1k epochs? ive also never made acc switching work.
You can increase the epochs amount when you re-train
@viscid moss teach him about acc switching in RVC-D when you wake up pls
...yeah
😭
retrain? through the same colab, or?
Yeah in RVC-D
yeah same colab, you can resume training
⠀
Download for Nvidia GPUs 
Version 17b cuda
Download for AMD GPUs 
Version 17b directml
Download for Intel GPUs 
Version 17b directml
Download for Mac 
Version 17b Mac
⠀
there it is
Does anyone have any idea why the rvc files don't work for me in the rvc client?
The only thing that works is the one pre-configured Beatrice file voice
Ayo? @bronze folio level 1 !!! 
you're welcome :)
Check cmd, send screenshots in #🔍│help-w-okada
Okay thank you ^^
-colab
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
anyone know how to fix the voice modules not working
does my accent affect it or something
im british
-colab
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
damn when i use models they sound completly different from the test samples. sometimes they arent even understandable. is there something i can do or?
⠀
Download for Nvidia GPUs 
Version 17b cuda
Download for AMD GPUs 
Version 17b directml
Download for Intel GPUs 
Version 17b directml
Download for Mac 
Version 17b Mac
⠀
RVC Guides (How to Make AI Cover)
Translation by country
Trancate is like, if it hears silence after "300" ms (that's what it's set to) it will start silence. That's probably what it is
Crossfade I'm not sure
i fix this.
pip install gradio==3.14.0
what is wrong if the g/total_loss graph in tensorboard is more ore less a horizontal line after 2k steps while training
idk why the voice dont sound like the charecters for me lol
let it cook more, if it went up you need to stop training
depends on your dataset and the song, try different songs and see how it goes
data set?
okay i will give it a try
gl
Hey do anyone know how create a voice cover with his own voice?
Last update: Feb 10, 2024
Here's everything you need to know
Good Day, someone know how to fix that?
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/gradio/queueing.py", line 407, in call_prediction
output = await route_utils.call_process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/route_utils.py", line 226, in call_process_api
output = await app.get_blocks().process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1559, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1447, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "/usr/local/lib/python3.10/dist-packages/gradio/components/audio.py", line 341, in postprocess
file_path = self.audio_to_temp_file(
File "/usr/local/lib/python3.10/dist-packages/gradio/components/base.py", line 336, in audio_to_temp_file
temp_dir = Path(self.DEFAULT_TEMP_DIR) / self.hash_bytes(data.tobytes())
AttributeError: 'NoneType' object has no attribute 'tobytes'
Are you on colab ?
update on applio links on nowebui colab, doesn't work is there a way it can be updated?
@pallid patio from which website are you getting the links from?
When I opened the RVC that you recommended to me, it asked me to get the link on huggingface, so I tried it.
Can you send me the link of the model?
I was trying out Shawn Mendes and Joshua Bassett. The links:
https://huggingface.co/Brzusiu/shawnmendes
https://huggingface.co/VittorX/JoshuaBassett
I got the links in the bot search tab here on the server.
You missed a step
The guide said you need to copy the download link
Go to the Files tab and hold the download button
Then copy the address
That's how it should look like
Ayo? @pallid patio level 2 !!! 
Now I understand, thanks again!
yw! Have fun
hmmm failed to fetch
It keeps giving the error screen. Like, the model is already downloaded, and the file with just the vocals (in mp3) too. I put "-12" since it's a female vocal going to a male voice and it still doesn't work.
is the audio selected?
in the dropdown bar
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
It is selected in both "Choose the audio file" and "Choose the index file".
my vocie changer ia not wokring
did you close the Colab tab?
is it still connected to the GPU?
I left everything open for fear it would stop working. Did it have to close?
no one is there
How do I know this? I don't know how to check.
can someone elp me in HELP w okada
nope, it must remain open & connected to the GPU
check the upper right corner
does it say something like "T4"?
yes
please read this #✨│ai-help message
don't beg for help
Is it allowed to upload photos here? I think I found T4.
yeah you can
try closing and opening the Gradio
oof if it's says connecting it's not a good signal
try doing this
It removed the files I had placed and now the voice model doesn't want to download. When I press connect the T4 it says it is not possible to connect.
for how long have you been connected to the Colab?
4 hours?
don't spam please
they prob don't know how to help
i keep trying to open
When I restart the page, it disconnects? I pressed start I think 3 times as the page restarted because I opened discord to ask for help.
can yall help
when you restart which page exactly?
The colab page
@bleak radish Come on, stop flooding. He already gave you the instructions. Filling in messages here won't achieve anything.
@bleak radish you'll get a bigger punishment next time. please behave
is it just me or are sapphire mutes taking a bit too long
before it was instant
like for spamming
for how many hours has the Colab been open?
If we don't count when it restarts, about 4 hours. If when it restarts it counts again, just over an hour. The page restarted a few times, so I'm not sure.
weird. try deleting history and try again
or use a different google account
or use AICoverGen
Delete Google history normally?
Ayo? @pallid patio level 3 !!! 
yup
Ookay, I'm going to change the account, wait.
thx sm
What's this?
Another Colab for inference
you can learn more on the docs page
I tried after deleting the history, and now everything is bugged. I'm almost giving up on making AI Covers. 🥲
Just search on Google?
did you run out of runtime?
I don't know. I haven't done any covers yet. Do they count just from being open?
Possibly. The Colab's been running for four hours
The usage time starts running from the moment you execute a cell
then prob
it will say cannot connect to gpu runtime type shi
use an alt acc then boom
Wow, that's terrible... I could count by time used. I was learning, it was open for a long time. When do they refresh the runtime?
or wait like 12h
as soon as u connect
Oh, makes sense...
when i use the ai cover gen why is there no public link
-colab
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
-local
- 🍏 Applio, 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
- GPT-SoVITS, by RVC-Boss 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.
Anyone know if there's a way I can make ai covers without extracting vocals? sort of similar to the NoWebUI but without the new links?












Do you guys know a solid sounding female rvc model?
RVC Guides (How to Make AI Cover)
Translation by country
How can I create my own audio file? The audio is saved as wav and I can't convert it to a zip file and add it to rvc :(
ok hi
so first thing, I would suggest two things
- Try downloading a different model, to see if it's not a link problem
- Rename the .zip file without spaces or special characters
im just confused since i used ilaria yesterday
and it was fine
i have the ui open it just doesnt download the model from the link
https://aihubdocs.github.io/en/rvc/cloud/inference/ilaria-rvc/ check the Troubleshooting page
Last update: Feb 10, 2024
that might help
thanks
Ayo? @verbal basin level 3 !!! 
yeah nothing there explains the error
can you try running ilaria and downloading it?
did you try closing & opening the Gradio?
or rebooting the Colab
yes
did you rename the zip?
I don't think that will work but go ahead.
You could try uploading the model to HF
Last update: Feb 10, 2024
its not downloading those either i just tried 😭
do you know if anyone else is having this problem?
I NEED HELP!!!
nope
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
i need download link
but for what 
For change ma voice
you mean making a voice model? or changing it in real-time?
lord have mercy
in real time
ask in #🔍│help-w-okada
I see these type of people everyday. I'm going insane
Hi i have 100w Rtx3060 Laptop gpu (6gb), do you guys have any advice "batch size" for training?
use 8
tysm ❤️
yw! remember that if the dataset is short (around 2 mins or less) use 4 instead
its 25 min
alr good. https://aihubdocs.github.io/en/vocal-isolation--datasets/datasets/ here's a dataset guide if you need it :3
Last update: Jan 9, 2024
I remember one being like "hello, it dont work" in help forum, or like someone "can i ask some question abt voices", a staff did the !howtoask and he replied with "thx it helped"😭
yup I saw it 😭
Ayone have a really good girl voice?
is that the same Ai hub that was shutdown before ?
I want to train a new AI model, but everytime I use the usual link, the interface is different and I can't do the public URL anymore. How do I do it now?
Ayo? @zealous ingot level 1 !!! 
-colab
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
anyone know why i keep getting crazy glitch beeping after using UVR to de-reverb
Ayo? @cunning topaz level 2 !!! 
no matter how many times i try
nah this is a new one. we made one from scratch
i mean is it the same staff
which link exactly?
some are new, some left
are you using the HF space? or the local version?
What's the best Hugging Face you guys use?
local on my laptop
are you using the latest version?
we don't recommend using Hugging Face spaces for RVC nowadays
they are obsolete
use instead the Google Colab spaces
yeah pretty sure, v5.6 2/12/24
https://aihubdocs.github.io/en/ check the Ilaria RVC or AICoverGen page :3
are you doing GPU conversion?
nope my laptop does not support it
hm I see. It's more convenient using MVSEP for vocal isolation
https://aihubdocs.github.io/en/vocal-isolation--datasets/vocal-isolation/ here's a usage guide
Last update: Feb 11, 2024
ill check it out. thanks
ah idk about EasyGUI. Try using RVC-D for now
Last update: Feb 9, 2024
yw! good luck
I'll give it a shot. I'll let you know about any future difficulty.
alr good luck!
yes
We still have old guys like Maki, guys like Kalomaze is no longer much active on this server
oh no, now it seems that google has made it so that a user session is required to download a file from google drive, causing gdown to break.
unlucky
the good thing is that it can be solved by putting a &at=usertokenoridk next to the link.
moment traductor
who knows who wants to volunteer to have a bunch of people use their account to make a request
Hi
Why when i Activate voice changer doesn't change in discord?
I selected my microphone and no changes
Reminder: this is not a help channel for Okada/Voice Changer.
Ask in #🔍│help-w-okada
thanks
and @spare shale I use a Laptop, without microphone (I use the laptop integrated microphone)
In any case, you need him
Ok, thanks u
Hey boss, I opened the VLC through the collab but none of the dropdowns are showing up, I'll download the model but nothing shows when I go to select it
gdown is dead again
What is a gdown
Ayo? @simple mango level 2 !!! 
a dependency that I think your colab uses to be able to download models from google drive
idk
Ok so it should be good like tmmrw?
"tmmrw" ? explain to me what it is, I speak Spanish
Manana
I don't understand anything
anyway, if you are not trying to download models from google drive, that would be another problem.
¿Ilaria debería trabajar mañana?
ah
En fin
Si traduces esto, solo te dire que si llegaras a descargar modelos de google drive mediante el colab, tendrias que poner tus cookies en un .json y ya
now translate it into English
Ok, eso no tenía mucho sentido, I don't speak enough Spanish to continue but I can try to dump my drive file into a .json or I'll just try to work on it tomorrow.
I took like one spanish class in high school lmao
can anyone tell me how to use the voice changer in discord?
like how exactly i can use the voice changer on it
como hago que el programa en cuestion funcione en discord?
oh alr
Ayo? @harsh pilot level 1 !!! 
btw i solve the issue
🎙️ Realtime Voice Changer (Local)
Description: This one has more options and is somewhat more convenient to use than Real-time RVC.
Link: Rentry
Credits: ravencutie21
Description: More accurate than W-Okada's, but requires a decent GPU.
Link: Google Docs
Credits: teas.hi
Description: Text to speech with RVC
Link: Rentry
RVC Guides (How to Make AI Cover)
Translation by country
Same issue, I can't even see my downloaded models in the select at all
What does it mean?
Ayo? @haughty moss level 3 !!! 
see your rvc folder name ? It contains a space (RVC tool) rename it to something without any spaces and it'll detect your files and won't tell you it has nothing to do
yeah
All the optional options are not showing up for me
Just have clon repo and start server using ngrock
I got no connect to google drive option
Ayo? @granite zephyr level 1 !!! 







So i got it working but after 7 or so minutes the runtime disconnects and voice changer stops working
which colab ?
Is there a way to split 2 or 3 backing vocals in harmony
pls someone that could help me
I always download a model that in the sample audio sounds really good and when I put it on the program the quality it's much much worst
but not the cuality of the ai, the quality of the audio
like if the ai was talking wwith a 3 euros headphones mic

how do i use this
Ayo? @exotic seal level 1 !!! 
you go to the pinned messages and you have a lot of guides
kk
can i upload a wav file here? trying to figure out if there's a setting I could change to make the results not screech/fail to hit high pitch notes
hey guys is there any way to fix the robotic voice?
RVC Guides (How to Make AI Cover)
Translation by country
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)
🎙️ Realtime Voice Changer (Local)
Description: This one has more options and is somewhat more convenient to use than Real-time RVC.
Link: Rentry
Credits: ravencutie21
Description: More accurate than W-Okada's, but requires a decent GPU.
Link: Google Docs
Credits: teas.hi
Description: Text to speech with RVC
Link: Rentry
Yes, it is.
You must use clean data and make sure to delete all noise and background music
god damn i need help I cant use RVC voicechanger
I think i can do something.
?
Do you wanna use local W-Okada?
not sure what is that Ig?
im been trying to get it working but not sure what is that anyways
W-Okada is another type of voice changer xd
there is no bgm tho i used my raw voice with no background noise
Ayo? @broken totem level 1 !!! 
https://rentry.co/VoiceChangerGuide
should help
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful Links - You need to follow this if you are on an AMD or INTEL ARC graphics card
Frequently Asked Questions
Jump To The Important Sections
Other Links
Download NVIDIA
Download AMD/INTEL GPUs
Download Mac
Online Hosted Alt...
idk which one IM using Xd
Are you sure?
Maybe you didn't overtrain it?
could also be extremely undertrained (low epochs)
i downloaded model from here
ah
i guess thats the reason
Nope, maybe you overtrained or undertrained it.
i mean... test other pitch values
bro is not training a model
anyways i'll look at it a bit more thanks for ideas
yw :)
well I got error that in cmd that said pipeline is not initialized when Im using RVC voice it
Ayo? @onyx plume level 1 !!! 
So most likely something went wrong with the initial install, when you open W-Okada for the first time. Try deleting the pretrain and model_dir folders (inside MMVCServerSIO) and restarting. It should re-do that initial installation and should work after that. Just make sure to keep the cmd open until the installation ends...
Being honest i got kinda confused with Shadows' issue.
It's ok
But he wanted to train a model?
Or he asked why his model was robotic?
Because if your model sounds robotic it can be for one or various reasons
AMD version
huh
this is the AMD version of the voice changer
That's w okada bro.
my bad Im conufuse with the name and it didnt show that name
But that voice changer works properly for you?
seems not really it did fix that error but Im not hearing myself
if you have an NVIDIA gpu you messed up
so Im trying to look at it
oof I just downloaded the NVIDIA
⠀
Download for Nvidia GPUs 
Version 17b cuda
Download for AMD GPUs 
Version 17b directml
Download for Intel GPUs 
Version 17b directml
Download for Mac 
Version 17b Mac
⠀
these r the sorted versions
So what would happen to my pc if that did happen
i mean... nothing much
W-Okada wouldn't work properly tho
good to know
You almost give me death Xd
Ayo? @onyx plume level 2 !!! 
well I will get the first one
so the pythorn and pytorch are still same?
k
you did not need to download python/pytorch for W-Okada
good to know
But you actually need a good VRAM and GPU for W-Okada
Idk how much but it ranges from 8 to 10 GB
hm i see
bro... no 💀
i mean like, if you wanna play games at the same time, yeah
12GB VRAM
but for like, normal use? 6GB VRAM is great already
You can run it just fine on a 2GB card... if it doesn't do anything else at the same time
MJ says if you got 6GB VRam then you're good to go.
At least if you don't plan on using Okada while playing AAA games
yw! :)
now I have to find good voice Xd
You can use the #1175430844685484042 channel.
There are plenty of good models, but you must test yourself.
yeah thats what im trying looking good one
Altho some models simply won't work well on W-Okada due to it's voices.
I suggest you to test character models.
Scruffygamer has done some HQ models of Hazbin Hotel, maybe you can test those.
I guess
you probably didnt select something properly.. or something
dunno how to troubleshoot applio problems tho ngl
why my rvc so laggy
Are you using local or Colab?
If you're using Colab, which one are you using?
Hello I am having trouble with RVC
'Cannot retrieve the public link of the file. You may need to change
the permission to 'Anyone with the link', or have had many accesses.'
'Error: Could not find a valid .pth file within the extracted zip.'
I am trying to use the doofensmirtz ai cover
again... im not sure what it could be, but when that happens its normally something that wasnt selected properly
I know why this happens.
ok
yeah the zip download link is google drive... google drive links dont work anymore
you gotta upload to Huggingface
oh
I suggest you to download the model you tried to use.
Upload it to your HF account.
https://aihubdocs.github.io/en has this
Last update: Feb 10, 2024
tutorial
on how to upload to hf
thank you
Ayo? @inland iron level 1 !!! 
yw :)
hi leo my favorite tutel representative
u should be happy for being tutel representative
Yesterday i got motivated to go back to helper role because someone got amazed by a girl model i gave to him.
damn
nice
i get motivated to help because i like helping people
Helping people is good too.
I asked Litsa and Simpl to give me back my helper role, i got it back but with 2 weeks of probation.
I don't really mind that.
What matters is to come stronger.
yeah!
🐢
does anyone have a link for the google collab?
Ayo? @silent lynx level 1 !!! 
Ilaria RVC?
ok thanks
https://aihubdocs.github.io/en/ check the Ilaria RVC page
Last update: Feb 10, 2024
For a usage guide
didn't properly select something
alr ty
This has something to do with gradio.
Unless you're using an outdated colab
does llaria rvc saves to gg drive?
How can I fix it then, because I opened the colab, and ran everything correctly.
that's the problem
i really don't know
no
So i have to install it once again
yeah, every new session
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
where can i find the okada rvc with gg drive save?
that thing is laggy as hell
altho id not recommend using that drive save (it might mess up)
its worse than the outdated okada
💀
Oh no!!
when will it reload the usage?
like, tomorrow
its like 12/24 hours so yea
FOR ONLY
Ayo? @barren vault level 3 !!! 
30 MINUTES OF USING
nah it limits out after 4 hours
if u were using the funny restricted ones tho maybe that screwed up with the usage
i haven't used up 4 hours
well... whatever it was
i think tomorrow it might be fixed
might
wait maybe try starting the colab again and check runtime type
i cant making ai cover song because it say no NVIDIA Driver from my system
but my laptop GPU is Intel UHD Graphics 620 what should i do
no its fine
you need nvidia gpu
yeah uh... you just don't make ai covers locally ig
use a colab or something
hello
ello
I dont know why but when I try to download voice model on ilaria rvc colab and press refresh, I dont see my model in list
maybe the link didn't work or something
if it's a huggingface link make sure you removed the "?download=true" part
what is the minimum gpu to use the final version of rvc?
if it's a drive link, it won't work
like... gtx 1050ti for inferencing at least
2060 for training
you can use cpu for inferencing tho
but it'll be slow
terribly slow


it's a decent card but... AMD compatibility is awful
Use RVC Disconnected if you want to train then
RX 580 is literally the king of sudden dying
Works very well for me =3
Nahh, i'm not sure that will work for training.
yea, it was drive link, but I definitely remember it worked yesterday
for basically everything related to RVC
It's better to use RVC disconnected.
yeah well, the google download stuff died today
you should use Huggingface links, always
what about GTX 1060
hm that works
ohhh, thanks for saving my time
ight bet
will be awful for training tho... veeery slower than a 20 or 30 series
I don't wanna train
yw :)
for inferencing yeah, it'll be alright
then just wait?
yeah you'd have to wait till tomorrow
Unless you got an Gmail alt account
must my model in hugging face be public?
-colab
- Applio, by IA Hispano Google 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
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- 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
- GPT-SoVITS (demo), by litagin Huggingface Spaces
Ayo? @brittle wing level 1 !!! 
yes
Ayo? @proper shale level 88 !!! 
anybody having problems with huging spaces rvc 2
tbh you shouldn't be using Huggingface spaces stuff rn
just way slower compared to colab
You should use colabs like Ilaria RVC, HF spaces are outdated
Doing one on Atrioc
A streamer, i'm sure 500 is enough. Trained 2 other models on 300 seems okay
uhhhhh
i closed the cmd that was running go-web.bat when i got finished training
did i fuck up?
RVC Guides (How to Make AI Cover)
Translation by country
wait wtf, why wasnt it saving checkpoints
omg i didnt have this set to yes
