show the error
#✨│ai-help
delicate whale
how do you take a screen shot lol
low shard
how do you take a screen shot lol
- press stamp/print depending on your keyboard
or - windows+shift+s
cloud finch
skibidi
delicate whale
- press stamp/print depending on your keyboard
or
- windows+shift+s
here it is
cloud finch
did i already show u the screenshot that the upload failed lmao
no 😭
simple ore
how do you manage to f this up
delicate whale
I created a new folder in my C drive and then extracted that folder their
low shard
so what do i do?
put it in a path without spaces
just remove the spaces where the program is
delicate whale
what do you mean remove the spaces
low shard
So now i run it back again?
after renaming the folder,yes
eternal aurora
how hard is this?
is this apparently hard?
simple ore
for ipad generation it seems so
delicate whale
after renaming the folder,yes
says this nowwwwwww
simple ore
holy shit, finally
low shard
says this nowwwwwww
did it open ur browser automatically
delicate whale
yuh
delicate whale
it runs locally, just uses a web browser interface
where do i add my voice
simple ore
looks like its doing well it is at its 1.25k marjk
click the resize button again
simple ore
are you training without a pretrain?
eternal aurora
are you training without a pretrain?
me?
simple ore
yes
simple ore
why
eternal aurora
i didnt know what model to use ngl
simple ore
use default then
eternal aurora
i have to download them to get them on the list
simple ore
on the training tab
truly ipad generation
you're training a model, why you're looking at inference tab?
eternal aurora
eternal aurora
you're training a model, why you're looking at inference tab?
mb
simple ore
down to training in advanced options
eternal aurora
simple ore
do not uncheck
simple ore
was it ticked when you started the training?
eternal aurora
yh
delicate whale
https://docs.aihub.gg/rvc/local/applio/#inference-
how can you add just the file that I downloaded
eternal aurora
oh
simple ore
how can you add just the file that I downloaded
put the model files into /logs/ folder
low shard
how can you add just the file that I downloaded
wdym
it's explained in the guide how to add models
delicate whale
it says link where am i gonna get the link
eternal aurora
i went into my game files and got the audio like that and got like 400 voice clips totalling up to 28 mins and then i combined the audios it cuz uploading it through the batch file link thing was being stupid
simple ore
you dont need to create a dataset or upload files if you run it locally
only if you're using UI Colab
eternal aurora
yeah im using it on ui not local
delicate whale
it's explained in the guide how to add models
Whats the best settings for it?
low shard
Whats the best settings for it?
you usually just need to play around with the search feature ratio and pitch that's all
it all depends on how good is the trained model
fading barn
Anyone on linux can help me set up the a virtual audio cable so I can use okada on other programs? 🙂
dark cove
this is my first time using this. how can i stop my wokada from stuttering?
delicate whale
wdym
Hey so if were using a virtual cable how do we hear the audio of a video for example?
eternal aurora
Hey so if were using a virtual cable how do we hear the audio of a video for exa...
you dont need on at all times
delicate whale
what if were streaming for example?
low shard
this is my first time using this. how can i stop my wokada from stuttering?
elaborate:
- your pc gpu
- what u want to do
- what tutorial link did u use
- a screenshot of ur wokada
low shard
Hey so if were using a virtual cable how do we hear the audio of a video for exa...
wdym? the output in twitch settings will still be ur headphones, the input will be line 1
eternal aurora
what if were streaming for example?
the vac is only used on okada
delicate whale
wdym? the output in twitch settings will still be ur headphones, the input will ...
for example I have right now Line 1 virtual cable audio on and if i go to youtube I cant hear the audio their
eternal aurora
for example I have right now Line 1 virtual cable audio on and if i go to youtub...
change ur audio in settings
low shard
for example I have right now Line 1 virtual cable audio on and if i go to youtub...
do ctrl+windows+b and show a screenshot
dark cove
elaborate:
- your pc gpu
- what u want to do
- what tutorial link did u use
- a ...
my laptop uses a rtx 3050. i want to change my voice in real time. i used a youtube tutorial from ai search.
delicate whale
do ctrl+windows+b and show a screenshot
Nothing pops up when I do it
dark cove
i cant send screenshots
low shard
my laptop uses a rtx 3050. i want to change my voice in real time. i used a yout...
rip, video tutorials are very outdated, that's the issue
a year old video tutorial is prehistorical in ai field
all youtube tuts use an old version of original wokada
and vb audio cable which causes issues
uninstall all you got off youtube
-realtime
patent trellis
low shard
read the 1st link, wokada deiteris fork
dark cove
alright thanks man
low shard
Nothing pops up when I do it
try ctrl+windows+v
delicate whale
try ctrl+windows+v
Can I have the output on RealTime Voice Changer Client on Line 1 Virtual Audio Cable and the output of my pc for my headphones??
low shard
alright thanks man
yw lmk
low shard
Can I have the output on RealTime Voice Changer Client on Line 1 Virtual Audio C...
yes, that's what im trying to help you get, try that combo so i can see the output device you're using
delicate whale
I just tried it
low shard
I just tried it
did it show something on the right bottom of ur pc?
delicate whale
no
low shard
is it windows 10 or 11?
delicate whale
10
low shard
10
oh
go to settings > search "audio" > click the 1st option > then select ur speakers as the output device
that should fix it
eternal aurora
is this done?
simple ore
Anyone on linux can help me set up the a virtual audio cable so I can use okada ...
you can do it with default PulseAudio
simple ore
is this done?
not even close
delicate whale
that should fix it
I changed my audio from virtual cable, to my headset on my PC and when I open the app it goes to my headset on my PC if I change it on Realtime Voice Changer Client back into Line 1 virtual cable for the output on the website and for my actual audio I change it to my headset for my monitor itself would it operate the same?
low shard
I changed my audio from virtual cable, to my headset on my PC and when I open th...
it should work
delicate whale
okay prec8 it, just a question how did you know my file name?
eternal aurora
not even close
ugh i may need to leave it overnight or sum shit
low shard
okay prec8 it, just a question how did you know my file name?
wdym?
low shard
ugh i may need to leave it overnight or sum shit
training takes hours
delicate whale
wdym?
Remember when you showed the screen shot and it said my file name when I had no spaces?
fading barn
you can do it with default PulseAudio
How? I've tried a few tutorials but nothing is really working 😛 The deiteris guide just says to google it which also doesn't help a ton
low shard
Remember when you showed the screen shot and it said my file name when I had no ...
because it shows the path of the file in cmd (the black window u saw)
it was literally showed in ur screenshot since there was an error lol
i actually did a screenshot of ur own screenshot that time to show u the error lol
delicate whale
I thought bro was a hacker lol
eternal aurora
training takes hours
Yeah I can tell lol I’m on my 2 hrs mark
low shard
I thought bro was a hacker lol
u just needed to check ur own screenshot 😭
u can go back and read it if u want
delicate whale
I trust you bro lmaoo
low shard
Yeah I can tell lol I’m on my 2 hrs mark
goodluck 🔥
eternal aurora
goodluck 🔥
yep i defo need it, when its done what do i do?
stop training on applio?
then export?
low shard
well u would actually need to check the tensorboard to check how's the training going
eternal aurora
well u would actually need to check the tensorboard to check how's the training ...
yhh ik but lets say i checked tensor and its done what do i do/
low shard
yhh ik but lets say i checked tensor and its done what do i do/
u stop training and get the best epoch based on the lowest point step
low shard
bro the Applio is so hard to control
how
eternal aurora
u stop training and get the best epoch based on the lowest point step
epoch is the y axis?
eternal aurora
u stop training and get the best epoch based on the lowest point step
how do i get the best epoch?
reef flax
icl my w-okada keeps delaying for some reason even though i have an 4060ti and it can run pretty fast mostly but now for some reason it delays a lot more.
to like 220ms
delicate whale
Im trying to make it record so I can make a video
reef flax
even though it can handle 120ms
simple ore
why did you select the entire assets/audio for batch infer?
reef flax
what am i supposed to do to fix that beacuse its annoying asf
delicate whale
why did you select the entire assets/audio for batch infer?
how do you put a model into it and actually use the damn thing
reef flax
what am i supposed to do to fix that beacuse its annoying asf
wokada deiteris fork btw.
ts pmo bro why is this suddenly so ass
eternal aurora
how do you put a model into it and actually use the damn thing
go to download
eternal aurora
how do you put a model into it and actually use the damn thing
scroll all the way down once ur on the download tab,then go into download pretrain model
reef flax
no way yall r still here
simple ore
how do you put a model into it and actually use the damn thing
as I said, put the model into /logs/, click refresh, pick it from the drop-down
eternal aurora
as I said, put the model into /logs/, click refresh, pick it from the drop-down
he needs to download pretrained models first from the download tab
delicate whale
Could yall do a tutorial
Like genuinely if someone downloads the app what should they do and etc
eternal aurora
have u figured it out?
delicate whale
Nope
delicate whale
as I said, put the model into /logs/, click refresh, pick it from the drop-down
Whattt
reef flax
updating my gpu bc maybe thats the issuetbh
violet rose
I have a problem, I had a crash on my pc and the okada launched in “security” mode but deleted all my voices. How can I get it back to the way it was before?
delicate whale
For example I download a model RVC here and then I copy paste that file into Logs and then refresh it and then put it into Appoil?
violet rose
all .pth transform on .safetensors
delicate whale
u talking to me or wha
violet rose
if u can help me yes
low shard
WOKADA has been ragebaiting me
wdym
reef flax
wdym
i just got it fixed with you and it reponsded faster finally, like it SHOULD.
now its back to the delaying thing.
for some reason
i dont know if its even using my GPU correctly.
but i never had this problem and it just randomly came
and ive tried so many things.
updated my gpu. deleted the old driver, bc i replaced the driver thing like u said.
delicate whale
What is Wokada?
low shard
What is Wokada?
the realtime voice changer program
low shard
but i never had this problem and it just randomly came
does it like go idle and cut at the start of your phrases?
reef flax
it must be like an glitch or something by the program itself because this is crazy
delicate whale
the realtime voice changer program
it’s different then the two you recommended me?
reef flax
does it like go idle and cut at the start of your phrases?
yes that ifmy phrases even COME.
as of right now no phrases are workin
low shard
it’s different then the two you recommended me?
no, it's literally the first program i recommended u for realtime lol
reef flax
like i genuinely feel like a dumbass
delicate whale
no, it's literally the first program i recommended u for realtime lol
ohh shiii😭🤣🤣
eternal aurora
@low shard imma leave my pc on overnight now is there anything i should do?
low shard
yes that ifmy phrases even COME.
share a screenshot of ur wokada and tell what program ur using it with
low shard
<@911742715019001897> imma leave my pc on overnight now is there anything i shou...
on kaggle?
low shard
cause im not using another other program
like if ur using it on discord or a game
reef flax
like if ur using it on discord or a game
just discord
not even discord tbh
discord is running in the background
just it it self.
low shard
yep
you should be fine, just be aware that you could risk being disconnected for inactivity but it's rare that happens on kaggle
low shard
not even discord tbh
are you using wokada for using models on pre-recorded audios or smt? wokada is meant for realtime like discord vc
eternal aurora
you *should* be fine, just be aware that you could risk being disconnected for i...
ah okay wish me luck ig for tmrw and thanks for urs and @simple ore help
low shard
yw
reef flax
are you using wokada for using models on pre-recorded audios or smt? wokada is m...
im using
wokada models
not prerecorded
like realtime
this is not my first time it being ass
this is my first time it being ass.
low shard
not prerecorded
I was just asking if you were gonna use wokada for realtime usage or for pre-reocrded audios
reef flax
like right now, its speaking live again. but after a while, it will die out again, and suddenly delay. if i open it again in abit, do you know why these issues could happen?
reef flax
I was just asking if you were gonna use wokada for realtime usage or for pre-reo...
ohh yeah i use realtime mostly.
i mess around with friends a lot with it and its fun
rn its running as it should
but if i closed it and tried again im sure it wont
low shard
like right now, its speaking live again. but after a while, it will die out agai...
mmm, close the program, run the force gpu clocks then restart the program and show a screenshot of ur entire wokada
reef flax
mmm, close the program, run the force gpu clocks then restart the program and sh...
see now i restarted it and it has the delay again
even though i know it can run fast
it always has
ts pmo
low shard
nice image model, set chunk to 192
does it go like 'idle' everytime u put it on background?
reef flax
nice image model, set chunk to 192
Lmaooo
reef flax
nice image model, set chunk to 192
ive tried that didn’t work
it’s not that the chunk is the issue it’s the fact that it just gives up on loading like it always does for some reason
It’s just literally sometimes it wants to work like it used to but mostly it’s just this shit
it delays like crazy ior just freezes up
I deleted it at first thinking it got outdated.
low shard
does it go like 'idle' everytime u put it on background?
@reef flax could u answer this precisely
reef flax
my bad man
what do you mean with idle exactly?
like it just stops or doesn’t perform as well?
low shard
like it just stops or doesn’t perform as well?
yes
reef flax
not every time I put it on background
Every time I start it in general, unless. I’m lucky
delicate whale
<@839962384902258698> could u answer this precisely
You said if I needed to record something like a yt video I should use Applio, so how do you operate it and add the voice and get the system running?
low shard
not every time I put it on background
maybe after a while?
reef flax
maybe.
reef flax
Either way it does fall under that category I’m assuming
reef flax
did u run the force gpu clocks?
I did
reef flax
Also.
low shard
you mean that's updated too?
reef flax
Yes
low shard
does your gpu have hiccups/issues at any other program like 3d rendering or games?
reef flax
No I replayed resident evil 4 remake and village
It went great
GPU is pretty recently put in too
delicate whale
No I replayed resident evil 4 remake and village
just try to re downlload it
and delete all the other files
low shard
GPU is pretty recently put in too
that's pretty weird since i have your same gpu lmao
reef flax
that's pretty weird since i have your same gpu lmao
right?
i can run alot of games
just fine
delicate whale
Restart your PC
reef flax
i already did multiple times gng.
delicate whale
Bro tried everything lmao
low shard
right?
@crude flame do you may have an idea?
it seems like his gpu goes idle after a bit the program stays in background, even if he has no other program than discord opened and already tried force gpu clocks
reef flax
i used to run the w-okada voice changer i think this is deiteris fork with no issues. same gpu, same settings. now suddenly its super delayed or i hear nothing at all. gpu is active (30–40% load), so the model runs, but no voice comes out. and if it does? it delays like crazy.
short explanation so your buddy doesnt have to scroll up and go crazy.
delicate whale
how do you put a model voice youve downloaded from the server?@low shard
reef flax
idk for sure
but i think voice model
. it literally says select the voic emodel to use for the conversion..
💔
low shard
how do you put a model voice youve downloaded from the server?<@9117427150190018...
go to the download tab
delicate whale
I tried it but doesnt work I think theres another way for Applio
delicate whale
go to the download tab
It says google drive link in the Guide
What does it mean by that
low shard
It says google drive link in the Guide
the type of model download link
else if u already downloaded it on ur pc, just extract it and upload the pth then .index
reef flax
yeah the delay is still crazy
do yk the immense ragebaiting
w-okada is doing to me rn
delicate whale
else if u already downloaded it on ur pc, just extract it and upload the pth the...
I click on upload Audio??
reef flax
https://tenor.com/view/mewing-bird-gif-16835999897918525830
this bro
im the tiger
w-okada the monkey (wise and ragebaiting)
low shard
I click on upload Audio??
in the download tab, there should be an upload model part
delicate whale
in the download tab, there should be an upload model part
Dont work tryna join the VC and check it out
low shard
yk ur cooked when even my guy nick088 dont know
my suspects are that ur gpu goes idle after the program goes in background for a bit, but usually force gpu clocks totally fixes that
low shard
Dont work tryna join the VC and check it out
!give-media-perms 1h @delicate whale
u can just send screenshots
reef flax
my suspects are that ur gpu goes idle after the program goes in background for a...
i might force bro with my fists
no but im force gpu'ing a million time srn
low shard
click the Download tab
delicate whale
Just did
reef flax
i added it to NVIDIA Control Panel
specifically in 3D settings > program settings, ykyk.
hoping maybe this might work?
delicate whale
Bro got unlimited Ideas lol
simple ore
I tried it but doesnt work I think theres another way for Applio
I fking told you twice
simple ore
- put the models into logs, 2) click Refresh on inference, 3) select the model from the dropdown
inner trout
Whats your guys's go to for ai tts? Applio still?
reef flax
icl i think its still idling ive tried almost everything doesnt work
delicate whale
1) put the models into logs, 2) click Refresh on inference, 3) select the model ...
I put the models there and it wasnt appearing, this shiiii is soooo fking harddddddd
delicate whale
Can we please join a VC call and actually help me out
reef flax
😭😭😭😭😭😭😭😭😭😭
delicate whale
I SWEAR THIS SHI HARD ASL
simple ore
i'm fking done... i'm gonna open a bottle of vodka and just drink it all
reef flax
i'm fking done... i'm gonna open a bottle of vodka and just drink it all
im sorry man
im actually sorry for you
delicate whale
i'm fking done... i'm gonna open a bottle of vodka and just drink it all
PLEASEEEEEEEEEEEEEE HELP MEEEEE
delicate whale
Buddy quit yapping and figure out the shi w ur GPU
reef flax
already did everything anormal person could do , even the dev didnt know im just waiting maybe theyll come back with something
u on the other hand are just making their lives a living hell
use common sense before you ask for help
simple ore
PLEASEEEEEEEEEEEEEE HELP MEEEEE
did you unzip the downloaded model or you're expecting applio to load .zip for some reasonm?
delicate whale
yes
simple ore
yes yes or yes no?
delicate whale
YES YES YES
echo meadow
how i upload
my private model
in this huggingface website, so the aicovergen can download the model
reef flax
icl w-okada ragebaiting me like crazy ive tried everything and this dumbass shit still delaying for no fucking reason bro praying on the application's downfall, have a good day
delicate whale
HOW DOES APPLIO WORK TO PUT IN YOUR MODELLLLLLLLLLL
simple ore
in this huggingface website, so the aicovergen can download the model
on huggingface make a new model
delicate whale
on huggingface make a new model
I putted the model link then clicked on Download Model
simple ore
how i upload
then from the model card
simple ore
I putted the model link then clicked on Download Model
I've explained 3 times
I can not explain any more
delicate whale
Whats Next Come on Dont be like that
simple ore
I can not help you if you can not do a basic computer stuff
delicate whale
Just one more time thats it
I been trying to figure it out for the past 2 hours its frustrating please
and I finally did it
simple ore
I been trying to figure it out for the past 2 hours its frustrating please
explain in your own words
torn seal
anyone know how i can remove some light backround talking in a clean acapella
simple ore
elaborate
delicate whale
then Refresh it?
inner trout
Put pth and index models into /models
Refreshing does not make them appear, have tried a restart on applio. Ideas?
torn seal
anyone know of a model that can remove backround talking on a acapella of some rapping
simple ore
Put pth and index models into /models
Refreshing does not make them appear, have...
just no
inner trout
just no?
simple ore
logs, not rvc/models
simple ore
that's how it is
torn seal
lmao why would it ever be in a logs folder
lol your a noob thats how it is in rvc
simple ore
rvc/models is for internal stuff
torn seal
you got spoild on aplio
inner trout
Sounds like some developer made a dumb choice and it's never been changed
torn seal
just stop
inner trout
Oh well ill modify it
analog obsidian
torn seal
be greatful the developer created this
inner trout
ok and ill be greatful that im going to change it because why the hell would you put premade models in a folder where logs should lie
its not that deep
analog obsidian
tbh it also felt weird to me applio stores models in the logs folder, mainline has its own dedicated folder for that
torn seal
lol it is when your learning rvc. do your own research if you don't like it.
inner trout
again some developer made a stupid decision and no one decided to change it
no big deal
ill change it myself
torn seal
logs is for saving all the trained data such as index. the wav. ect
inner trout
okay just leave and quit talking about it if you can't understand my simple point
inner trout
i dont care what the standard is, the naming makes no sense.
delicate whale
I dont have hop length and the if......... one on their @low shard
this is the picture of the guide and it has it
reef flax
bro
i literally tried every signle thing
my w-okada is still annoying
can someone lockin and help please
💔
torn seal
sorry your wokanda is broke
echo meadow
on huggingface make a new model
where is that
torn seal
your welcome
torn seal
you know with how tech is its easy to figure out stuff without peopole
reef flax
you know with how tech is its easy to figure out stuff without peopole
yeah well
ive tried 20 times
without people
ive tried multiple things
went on here and he didnt even know what to do
gang
torn seal
such as?
torn seal
did you try googling it?
reef flax
i updated my gpu
wdym which one
i have chatgpt plus
the newest model
and the smartest
reef flax
grok is good
wdym with grok gang
torn seal
grok by x
reef flax
grok ai?
torn seal
yep
reef flax
yea okay
delicate whale
I FINALLY DID ITTTTTTTTTTTTTT
torn seal
deep research helps
delicate whale
OMGGGGGGG
torn seal
i used grok to help me fix why sovits svc wasn't installed properly and it help me fix it so..
delicate whale
ONCE YOU GET THE AUDIO FROM APOLLEI YOU PUT IT IN REALTIME VOICE CHANGER CLIENT?
???
torn seal
its for cloning and infer voice
reef flax
ur idea
ur AI
came iwth everything IVE ALREADY TRIED.
im just gonna post on github
last resort.
delicate whale
its for cloning and infer voice
do you think it makes a bigger difference if your only using Voice Changer Demo AI?
elder coral
sour sun
Heyo again; quick question!
I'm experimenting with the experimental KLM6 Spin pretrain. I know for a fact I've 1. placed the embedder in the proper folder and selected it when training, and 2. opened up filelist.txt to replace /mute/ with /mute_spin/, but so far the early model sounds like... well, it sounds like whispers and noise. Is this normal, or is something not working properly?
viscid moss
how i get background vocals
Using MelKaraoke
viscid moss
where is that melkaroake
available on UVR5, UVR5 UI, MVSEP and MSST Colab
viscid moss
it was really the best ones
echo meadow
it isolated automatically and everything
viscid moss
pls check this
viscid moss
it depends if u want that locally or cloud
echo meadow
locally i think lol
viscid moss
locally i think lol
viscid moss
nice
viscid moss
echo meadow
let me try again
viscid moss
ok, just download, extract and run
echo meadow
ok, just download, extract and run
yup
Traceback (most recent call last):
File "E:\RVC-AI-Cover-Maker-UI-main\RVC-AI-Cover-Maker-UI-main\programs\applio_code\rvc\lib\tools\prerequisites_download.py", line 3, in <module>
from tqdm import tqdm
ModuleNotFoundError: No module named 'tqdm'
Traceback (most recent call last):
File "E:\RVC-AI-Cover-Maker-UI-main\RVC-AI-Cover-Maker-UI-main\main.py", line 1, in <module>
import gradio as gr
ModuleNotFoundError: No module named 'gradio'
An error occurred. Exiting...
Drücken Sie eine beliebige Taste . . .
xd
reef flax
hey so i found this while trying to run it
can someone tell me what this means, and if it has something to do with my problems >/ if i can fix it/how.
viscid moss
Traceback (most recent call last):
File "E:\RVC-AI-Cover-Maker-UI-main\RVC-AI-...
use ur C drive
reef flax
can someone tell me what this means, and if it has something to do with my probl...
right now im assuming its because my CUDA is maybe TOO up to date then w-okada can do right now gpu wise?
so im going to CUDA 12.2. someone tell me if im on the right track here or completely off track
maybe even flying off a bridge right now.
viscid moss
damn, i think that's the problem. Earlier a guy had the same problem
What about cloud then?
Colab
simple ore
or just did not install requirements properly
like did you just clone the stuff and trying to run it?
viscid moss
you're trying to install ancient requirements
Haven't updated the code tbh 
simple ore
did not do pip install -r requirements.txt?
viscid moss
like did you just clone the stuff and trying to run it?
it's using the precompiled version made by nick
echo meadow
https://colab.research.google.com/github/Eddycrack864/RVC-AI-Cover-Maker-UI/blob...
that look confuing
confusing
whats nko
simple ore
did @echo meadow download the precompiled?
echo meadow
ngrok*
viscid moss
did <@905524360436125746> download the precompiled?
yep
simple ore
why would it then break at missing requirements?
viscid moss
ngrok*
a tunnel to access the interface, u can use gradio instead. For ngrok u need a ngrok token
viscid moss
why would it then break at missing requirements?
i've no clue 😭
delicate whale
Do you guys think using VCD and then recording it in OBS is better than using Aplio?
simple ore
yep
well, no he did not
precompiled works fine
source needs install
@echo meadow download precompiled RVC-AI-Cover-Maker-UI-main for local use
echo meadow
a
simple ore
viscid moss
a
echo meadow
a
1 question
i used collab
where i find background vocals in this
like only this
file where its done
echo meadow
its like
this got all added in 1 song, but cant see them seperate
like i dont see instrumental
lead vocals
like the old one that you made
viscid moss
oh, but need to disable an option "Delete the audios after the conversion."
echo meadow
oh, but need to disable an option "Delete the audios after the conversion."
hmmm
where is that
viscid moss
advanced settings
echo meadow
yea
viscid moss
disable it
echo meadow
ok ill try again
viscid moss
ok
viscid moss
i see many folders lol
show me
viscid moss
now u can
echo meadow
viscid moss
karaoke
echo meadow
thanks
viscid moss
ur welcome
rotund vigil
hello! Im new to AI stuff and I just wanted to know some stuff about it. I have a laptop and I want to build my AI bot for Telegram, but I'm unable to do so, since it takes all my CPU and not loading a response. I suspect it to be due to my hardware being shit, but I just wanted to ask if thats the case
rotund vigil
hello! Im new to AI stuff and I just wanted to know some stuff about it. I have ...
My AI right now has a model of Tiny Llama 1.1v - this is like a last resort and it still doesn't work properly. My whole project is on venv (like when you do cmd stuff, it shows (venv) at the very left), everything is parsed and should be working as expected, but my AI just doesn't want to respond
ivory surge
hi! does anyone know an ai that seperates the leading and backing vocals?
olive cliff
hi! does anyone know an ai that seperates the leading and backing vocals?
You can use the MelBand Roformer Karaoke model (particularly, Becruily's model). That model can separate the leading and backing vocals.
ivory surge
You can use the MelBand Roformer Karaoke model (particularly, Becruily's model)....
Thank you!
sour sun
Heyo again; quick question!
I'm experimenting with the experimental KLM6 Spin p...
Figured out the problem, the embedder wasn't being regognized properly. I had placed the embedder folder in embedders_custom thinking it'd automatically select the proper files, but it wasn't actually doing anything 😅
maunally placing the files worked 👍
edgy saddle
when you have more epocs for training ai does that mean that the quality of the ai will be better
sour sun
when you have more epocs for training ai does that mean that the quality of the ...
Not necessarily. In simple terms it's possible to overtrain a model if you train it for too long (too many epochs)
delicate whale
Is there a free version of Voice Mod with unlimited free voices and soundboards
sour sun
It's kind of like baking a cake!
You want enough time in the oven so the model sounds like your voice (dataset)
However, you don't want to overbake it, because if you leave it cooking for too long it can't generate new sounds easily, and can start to sound robotic
edgy saddle
how many epochs would you suggest i use for 12 minutes worth of the certain voice
sour sun
So, each voice dataset is different. Two 12 minute datasets can potentially need different epochs
What most people use to figure out just how many epochs is through something called Tensorboard
edgy saddle
does the folder that i put the tensorboard bat into need to have a specific name
sour sun
Not that I know of, no
edgy saddle
because when i run the bat it says "system cannot find the path specified"
edgy saddle
because it doesnt have any spaces or special characters
the folder name is just "rvcs"
sour sun
And you've made sure the whole filepath also has no spaces or special characters?
edgy saddle
yeah
sour sun
Hmm
edgy saddle
like after the folder is open its just that
sour sun
Is the RVC folder also located on a filepath in your C drive?
edgy saddle
The rvc folder is in the download file path
crude flame
does the folder that i put the tensorboard bat into need to have a specific name
are you using applio?
if so you dont need to download that file in the docs
edgy saddle
Does applio have a feature that does something similar to tensorboard?
sour sun
The rvc folder is in the download file path
There we go that's probably the issue
When looking for specific files, RVC defaults to your C drive, so it's recommended to place your folders there (like on your desktop or something)
Applio is an offbranch of RVC that users here are regularly improving
crude flame
Does applio have a feature that does something similar to tensorboard?
yes
its called the tensorboard
just run
lusty whale
does anybody know why my w okada voice crackles and bugs in discord but not in like other games
simple ore
does anybody know why my w okada voice crackles and bugs in discord but not in l...
step 1 - disable hardware acceleration in discord so it does not offload work to GPU that is busy handing the voice changer
and browser too
outer wasp
Can i ask? How to train a voice model for RVC?
knotty moth
Can i ask? How to train a voice model for RVC?
the complete guide is there https://docs.aihub.gg/
lusty whale
step 1 - disable hardware acceleration in discord so it does not offload work to...
wait where is that
fluid topaz
i wonder if theres a realtime rvc client that able to apply vst plugin into the converted voice
fleet wharf
Does anyone know an ai that can increase the quality and clarity of audio?
river gulch
Now that I think so, is there a way for test already made models with index in a tensorboard way? I have some old models made on RVC that are actually good and have used in years, but I never test them with tensorboard when I was making them years ago. So it bring my curiosity now.
fluid topaz
Does anyone know an ai that can increase the quality and clarity of audio?
think its called audioSR
lunar geyser
I wonder how to make models 😔
brittle wing
tf is this
latent kettle
I wonder how to make models 😔
lunar geyser
https://docs.aihub.gg/essentials/how-to-make-voice-models/
Thank you 🫣 so much
latent kettle
If you face any problems ping me
brittle wing
If you face any problems ping me
can u help me bro
hallow thistle
can u help me bro
!howtoask
latent kettle
can u help me bro
What is the problem?
brittle wing
im on pretraining i was uploading the dataset
and randomly during that i faced this
and it stopped uploading
latent kettle
Well most of the collab notebooks are broken
brittle wing
damn so it's a problem on the notebook?
like its not something that i did wrong? can it be fixed?
latent kettle
brittle wing
https://discord.com/channels/1159260121998827560/1159380240271953940/13446873669...
thank you
if i have any other issues can i ping u?
brittle wing
damn thank u bro
is there like a tutorial or something for kaggle?
@latent kettle
latent kettle
is there like a tutorial or something for kaggle?
brittle wing
https://docs.aihub.gg/rvc/cloud/applio-kaggle/#applio-kaggle
im running the last cell rn but i have a question will this disconnect like google colab when im inactive?
brittle wing
alr thanks
if i got any more issues ima lyk cause i deadass dont know shit about training ai
latent kettle
if i got any more issues ima lyk cause i deadass dont know shit about training a...
Be aware about ngrok bandwidth
brittle wing
Be aware about ngrok bandwidth
tf is that
latent kettle
Its a limit of 1gb to upload or download data from connected agents
brittle wing
Its a limit of 1gb to upload or download data from connected agents
whats a connected agent bro
latent kettle
If you cross the limits you can't access your training session anymore
latent kettle
whats a connected agent bro
Here it is Your applio notebook
brittle wing
shit thats bad
latent kettle
shit thats bad
You can monitor it on ngrok web page
brittle wing
Here it is Your applio notebook
oh so for example if my data set is above 1 gig its pretty much over?
latent kettle
oh so for example if my data set is above 1 gig its pretty much over?
Yep
brittle wing
damn my dataset is 521 MB
it's a big wav file i used to do it like chopped up into like 70 different files but i was told it was better with a big wav with all of them on it
brittle wing
You can monitor it on ngrok web page
and where exactly can i monitor this
latent kettle
it's a big wav file i used to do it like chopped up into like 70 different files...
You can use FLAC format it's lossless and also saves memory.
latent kettle
and where exactly can i monitor this
Dm me a screenshot of your dashboard.
brittle wing
You can use FLAC format it's lossless and also saves memory.
hmm that's interesting i didnt know that
latent kettle
Yes
latent kettle
hmm that's interesting i didnt know that
You must read "how to create dataset "
brittle wing
You must read "how to create dataset "
i didn't read it honestly but i saw on a yt video that they used .wav format
also im a lil bit slow where can i find the dashboard
latent kettle
Most of the yt tutorials are outdated don't follow them
brittle wing
yeah i noticed just by the google colab thing
i appreciate u for helping me bro fr if it wasnt cause of u i'd still be stuck tryna figure out wtf the problem was
also bro one thing
latent kettle
brittle wing
i'm on the actual training now
im training a future model from around 40 mins worth of studio sessions how much epochs do u think i should set it to
eternal aurora
Most of the yt tutorials are outdated don't follow them
is it normal that i have this many path files after training?
latent kettle
is it normal that i have this many path files after training?
Yes. But it's eating kaggle storage which is also limited
eternal aurora
Yes. But it's eating kaggle storage which is also limited
yeah i can tell so which one is the file i need?
and how do i clear my storage?
brittle wing
yeah i can tell so which one is the file i need?
i think the one with the highest epochs if u set it to the prevent overtraining mode
normally it would save the best epoch i think
eternal aurora
i think the one with the highest epochs if u set it to the prevent overtraining ...
is the epoch the number before pth?
brittle wing
@latent kettle how much total epochs u recommend me to get for this model
brittle wing
is the epoch the number before pth?
notice how it says 240e, 250e etc thats the amount of epochs
latent kettle
is the epoch the number before pth?
With modelname_e
latent kettle
<@1174561195102056459> how much total epochs u recommend me to get for this mode...
What's the dataset length and batch size
eternal aurora
notice how it says 240e, 250e etc thats the amount of epochs
seems like my highest epoch is 430
brittle wing
What's the dataset length and batch size
dataset length is 50 minutes and batch size is set to 15 by default idk if i should change it
eternal aurora
With modelname_e
yeah i see it so my highest is 430 is that the one i should download?
brittle wing
yeah i see it so my highest is 430 is that the one i should download?
i'd say try that one since it's probably the best if there's no overtraining
the reason why u have so many of them is because u set it to save every 10 epochs im pretty sure
eternal aurora
i'd say try that one since it's probably the best if there's no overtraining
yeah i might do that, how do i download it?
eternal aurora
the reason why u have so many of them is because u set it to save every 10 epoch...
yh i did and i left for 7 hrs
brittle wing
hmmm tbh i dont really remember how to get the index file 😭
eternal aurora
hmmm tbh i dont really remember how to get the index file 😭
its in the same file as pth
latent kettle
dataset length is 50 minutes and batch size is set to 15 by default idk if i sho...
Broh... you should look at tensor board
eternal aurora
i jus downloaded it
brittle wing
Broh... you should look at tensor board
dude i dont know any of that graph shit LMFAO
tensor board is the graph where u check overtraining and such right
?
latent kettle
yh i did and i left for 7 hrs
You can check tensor board and download the best. 7 hours must be over fitted
latent kettle
tensor board is the graph where u check overtraining and such right
Yes
brittle wing
Yes
alr so i set the save every epoch to 50, batch size at 15 rn is that good
i wanna have a decent amount of max epochs but idk if 600 total epochs is too overkill
that'd mean 30 pth files
that could be risky what do u think
latent kettle
latent kettle
i wanna have a decent amount of max epochs but idk if 600 total epochs is too ov...
Epochs are not equal to quality
brittle wing
Epochs are not equal to quality
ik the dataset gotta be clean too
it is
i mean its studio sessions i cleaned them up a lil bit as well
eternal aurora
You can check tensor board and download the best. 7 hours must be over fitted
Oh okay how would Ik which one is the best?
brittle wing
it's raw vocals of future
latent kettle
Oh okay how would Ik which one is the best?
brittle wing
wait what if i enable overtraining detector @latent kettle
latent kettle
wait what if i enable overtraining detector <@1174561195102056459>
That's also good but I never used it
brittle wing
what does it do? i assume it automatically stops it
latent kettle
Oh okay how would Ik which one is the best?
Make sure to read advance guide too
brittle wing
@latent kettle are these the correct settings?
😭
i just dont wanna fuck it up my previous models were bad like weird typa bad they didnt even sound like the person i trained them off of
latent kettle
<@1174561195102056459> are these the correct settings?
Remove save only latest and save every weights. Reduce batch size to 8 or 10
Set max epochs to 300 for now
brittle wing
okay thank you
latent kettle
Enable fresh training
eternal aurora
Make sure to read advance guide too
Okay
brittle wing
Enable fresh training
anything else?
latent kettle
Enable fresh training
Only for now (don't do that when you are going to resume training)
latent kettle
anything else?
Not sure
latent kettle
so its ready for training now right?
Show me all settings
brittle wing
brittle wing
Show me all settings
so are these ready?
latent kettle
Read the guides to observe tensor board
latent kettle
anything else that i gotta know?
Everything is inside the guide good luck
brittle wing
Everything is inside the guide good luck
okay thank you
can i ping u if i need anything else?
latent kettle
okay thank you
can i ping u if i need anything else?
You can but there will a chance for late reply
brittle wing
You can but there will a chance for late reply
okay
thank you
dude @latent kettle i dont think theres any info about tensorboard on kagglio on the guides
how do i get it to work i dont see it anywhere normally on colab it will open automatically not here tho
eternal aurora
how do i get it to work i dont see it anywhere normally on colab it will open au...
select scalars, then click the gear icon and select Reload data
Nd select ur model near the left
brittle wing
on kagglio or ngrok?
ohh
nahh i get what u mean but i dont see that thing on my kagglio
eternal aurora
nahh i get what u mean but i dont see that thing on my kagglio
Wait press the tensorboard link
brittle wing
Wait press the tensorboard link
ohhh
eternal aurora
Nwrs and just do the bit I told you
brittle wing
Nwrs and just do the bit I told you
yup just did it
btw if i let it run nothing wrong gonna happen right?
im asking since u let it run for 7 hrs
😭
its 4:11 am bruh
i wanna sleep
eternal aurora
btw if i let it run nothing wrong gonna happen right?
Hopefully not it didn’t happen to me it just stopped ngl
eternal aurora
im asking since u let it run for 7 hrs
Ahaha
brittle wing
ight ima gts now
brittle wing
I’m at the gym rn
bro im meant to be at the gym at 11 am
eternal aurora
Have a good one
brittle wing
😭
brittle wing
Have a good one
aye u too man
eternal aurora
bro im meant to be at the gym at 11 am
7hrs of sleep is not
Enough 😭😭
low shard
be honest with me should i just GIVE UP
nahh, did Razer reply?
latent kettle
aye u too man
Sorry but I thought you got you answer
simple ore
dataset length is 50 minutes and batch size is set to 15 by default idk if i sho...
batch 15 is ridiculously high, change it to 8 for 50 min set
latent kettle
batch 15 is ridiculously high, change it to 8 for 50 min set
I recommend 10
simple ore
it can be experimented, 4, 6, 8, maybe 10, it is something to experiment and compare the result
reef flax
nahh, did Razer reply?
no.
low shard
no.
maybe @pastel oak or @analog obsidian could help too, or you could retry the force gpu clocks
reef flax
maybe <@965636299509882980> or <@775545133448953856> could help too, or you coul...
i tried 2.0.77.beta and it works but its not as good as the previous one, and i just want the previous one to work again.
reef flax
maybe <@965636299509882980> or <@775545133448953856> could help too, or you coul...
this is what im getting by the way
knotty moth
i tried 2.0.77.beta and it works but its not as good as the previous one, and i ...
try the fork version
reef flax
okay
reef flax
try the fork version
i think this is just the same i already have
eternal aurora
knotty moth
does it work anyway?
reef flax
does it work anyway?
it works just not as well as its supposed to
its delaying, and lagging.
which is not supposed to happen because it used to work fine and fast.
now its giving this lil warning
knotty moth
which is not supposed to happen because it used to work fine and fast.
how about another model? or perhaps the chunk settings? or perhaps some GPU issue?
reef flax
icl i tried everything
another model i already tried
chunk settings ive played arouind it
an dmy gpu is new..
like recently bought
knotty moth
not RTX 50-series for sure?
reef flax
4060ti
knotty moth
have you tried different driver version? perhaps Dec 2024?
knotty moth
becuase ive tried a different cuda toolkit wahtever
does it lag on certain games?
reef flax
does it lag on certain games?
if ur talking about my gpu
no ive said thsi before i played re4 remake and most of the other games smoothly
knotty moth
no ive said thsi before i played re4 remake and most of the other games smoothly
ye it runs smoothly alone but it is rather demanding
knotty moth
try lowering graphics settings and cap fps to 60
reef flax
do you mean my hertz or something
reef flax
try lowering graphics settings and cap fps to 60
i dont mean im using it while playing a game.
im using it while playing absolutely nothing other than having disord run
ive had it run perfectly before, while no teven playing games.
now its just delaying
and it gives me this warning,
im not playing anything.
knotty moth
it makes sense if for running discord
reef flax
uhhh
dont make sense to me because it used to work perfectly fine
im confused
what might be the issue
or are you lost too
lol
knotty moth
you mean this thing doesn't appear in your previous version?
reef flax
previous versions, and the most recent one like 2.0.77 beta
does work well
and how its supposed to
but the fork one, which runs models more smoothly for me
suddenly doesnt work anymore.
knotty moth
so you'd want to roll back to the 18a?
reef flax
so you'd want to roll back to the 18a?
i just want my fork to work well ...
again.
like it used to
i dont know if theres an misunderstanding here
knotty moth
so within a VM container?
knotty moth
I don't recall this kind of interface even on the og 18a
reef flax
this is on the most recent one from w-okada
icl ill just show youi
its this one
which is not the one i want fixed tbh because that one just works, i want hte fork one or at least understand whts going wrong because it used to work well.
reef flax
https://huggingface.co/wok000/vcclient000/blob/main/vcclient_win_cuda_2.0.77-bet...
this one doesnt run through chrome and is just an application by itself, it just doesnt run models as smooth as it should
reef flax
which is not the one i want fixed tbh because that one just works, i want hte fo...
the one i had the w-okada deiteris fork one
runs it smooth and i like it more, but now its suddenly delaying.
knotty moth
https://huggingface.co/wok000/vcclient000/blob/main/vcclient_win_cuda_2.0.77-bet...
we here don't support and guarantee it work
reef flax
runs it smooth and i like it more, but now its suddenly delaying.
im trying to get the delay to be fixed..
the warning i get is from fork
knotty moth
get what i mean
https://discord.com/channels/1159260121998827560/1297207135469305866 the guide here does cover on the fork version
not the 2.x beta
reef flax
paus elet me explain, the fork version ive already ran it before on the same settings and it worked just fine. the guide there probably tells me what settings to do best.
it just sudednly doesnt work anymore, doesnt matter what my settings are
but okay sure.
knotty moth
but okay sure.
if ts seems puzzled, the last resort may be to clean install windows and Nvidia driver & cuda toolkit stuffs
reef flax
if ts seems puzzled, the last resort may be to clean install windows and Nvidia ...
icl ts pmio
lol
any other way than just cleaning my windows completely and having to download everything i needed all over again..
reef flax
https://tenor.com/view/vro-pmo-ngl-gif-10268498521266426722
w-okada ragebaiting me
since yesterday
lunar geyser
https://tenor.com/view/vro-pmo-ngl-gif-10268498521266426722
Lmao what does this mean the only word I understood is the pmo
eternal aurora
@low shard it looks like its working it seems to be off a bit but its still good ngl
its a male voice what should the tune value be?
latent kettle
its a male voice what should the tune value be?
What do you mean by tune value
eternal aurora
mb i meant pitch
latent kettle
If you are converting female to male you should go with -ve pitch values example-3 ,-5 or vice versa
eternal aurora
If you are converting female to male you should go with -ve pitch values example...
okay ty
ngl my thing sounds good im surpised it works
latent kettle
Which model do you used
latent kettle
No, i mean you let it train for 7hrs
eternal aurora
yes
latent kettle
And it created multiple pth
eternal aurora
yes
latent kettle
So which one are you using
latent kettle
You can check tensor board for lowest Loss values.
eternal aurora
eternal aurora
You can check tensor board for lowest Loss values.
how do i do that?
latent kettle
Do you used over fitting monitor?
eternal aurora
whats that?
latent kettle
That thing automatically detect over fitting
eternal aurora
ngl im not too sure is there a way to check?
latent kettle
There will be a sweet spot where your model will perform its best. There could be chance that your model is over trained
What batch size do you used and what was your dataset length @eternal aurora
eternal aurora
There will be a sweet spot where your model will perform its best. There could ...
ohhhh
eternal aurora
What batch size do you used and what was your dataset length <@71350938911349153...
gimme a sec
reef flax
if ts seems puzzled, the last resort may be to clean install windows and Nvidia ...
i guess bruh
eternal aurora
What batch size do you used and what was your dataset length <@71350938911349153...
my shit is bugging
@latent kettle my batch size is 10
and my length is 28 mins
what do i look for?
latent kettle
theres a bunch of rvc which one do I download?
Applio
latent kettle
Can you please read the guide?
eternal aurora
Can you please read the guide?
i have
eternal aurora
Can you please read the guide?
i genuinely cant tell if i have low points or not
hallow thistle
whats the difference between that one and deiteris' optimized W-Okada?
Applio is the regular RVC program, a program that can do AI cover and train voice model. W-Okada is realtime voice changer program that mainly use RVC voice model to change your voice in real time.
!howtoask
silk niche
Is possible to use 2 gpus to decrease the delay? 3060 12gb and 5060ti 16gb
silk niche
Is possible to use 2 gpus to decrease the delay? 3060 12gb and 5060ti 16gb
In w okada
hallow thistle
Is possible to use 2 gpus to decrease the delay? 3060 12gb and 5060ti 16gb
One W-Okada program can only use one GPU at a time.
low shard
Is possible to use 2 gpus to decrease the delay? 3060 12gb and 5060ti 16gb
there isn't multi gpu support unfortunately
@wispy lodge that might not be a bad idea (if possible) btw
lusty mason
Applio is the regular RVC program, a program that can do AI cover and train voic...
so for realtime should i do deiteris' optimized W-Okada or the original one, i have a 4070 ti?
low shard
Is possible to use 2 gpus to decrease the delay? 3060 12gb and 5060ti 16gb
also, ofcourse the 50 serie would give less delay
low shard
so for realtime should i do deiteris' optimized W-Okada or the original one, i h...
you should use the wokada deiteris fork
don't use video tutorials
-realtime
hallow thistle
so for realtime should i do deiteris' optimized W-Okada or the original one, i h...
That's a confusing question I've ever seen. Yes, you should use Detris' fork W-Okada for realtime.
I'm not sure how one would attempt to run W-Okada with two different GPUs at a time. Plus, only Detris' fork W-Okada that made for RTX 50 could only work with RTX 50.
silk niche
Okeyy thanks
latent kettle
i genuinely cant tell if i have low points or not
Have you increased smoothing
lusty mason
That's a confusing question I've ever seen. Yes, you should use Detris' fork W-O...
okay and if I wanted a ai voice to say typed out text like TTS I would use applio
hallow thistle
Also, it's unlikely for these two different NVIDIA GPUs of different generations to combine themselves and work together. Each one would work independently from other. But you can set W-Okada to use RTX 3060 and one of your games to use RTX 5060 Ti.
hallow thistle
okay and if I wanted a ai voice to say typed out text like TTS I would use appli...
Applio has a built-in TTS feature. 