so you are gonna need a decent GPU if you wanna use the mainline one
#✨│ai-help
molten pecan
fringe linden
Ah shit here i got download RVC 
rare gobletBOT
Ayo? @fringe linden level 2 !!! 
molten pecan
Ah shit here i got download RVC <:eye_but_fast:853319286658957362>
it's not like realtime is far better than okada. I'm sure okada is decent enough
but if you are a perfectionist then yeah feel free to use mainline :3
fringe linden
it's not like realtime is far better than okada. I'm sure okada is decent enough
nws i'll give it a try
molten pecan
i think that was the name of the file
molten pecan
nws i'll give it a try
alright have fun! idk much about the voice changer so ask on #🔍│help-w-okada
fringe linden
tyy for the help
molten pecan
yw!
viral geyser
hello i am new on this ai stuff, i've been trying to find a good ai voice changer so i can use on discord, does anyone know where i can find one? (i'm on a macbook device)
molten pecan
hello i am new on this ai stuff, i've been trying to find a good ai voice change...
oh shit idk if you can run it on a Mac
ask on #🔍│help-w-okada
viral geyser
alrr tyty
half cove
RVC 6 wen?
molten pecan
SVC when?
fossil glen
half cove
iron wagon
how do i train an ai voice model
molten pecan
how do i train an ai voice model
what RVC are you using?
@half cove how to make ai cover
molten pecan
https://github.com/Tiger14n/RVC-GUI/releases/
phew, that's outdated
use Mainline RVC
-rvc
molten pecan
iron wagon
thanks
rare gobletBOT
Ayo? @iron wagon level 1 !!!
azure marshBOT
RVC Guides (How to Make AI Cover)
last token
Hey, someone have something for the voice don't have delay?
my voice stay stable but with much delay
lucid dagger
ima guess mabey 2 hour mby 3
viscid moss
Hey, someone have something for the voice don't have delay?
my voice stay stable...
Maybe bad settings. Do u have a GPU, right?
rancid pine
My message got deleted. Strange
viscid moss
My message got deleted. Strange
QWQ Chan delete edited messages
rancid pine
I see
What I was trying to say is that before I did the extra UVR steps, my model was already good enough and that I was trying to fix the pitch cracks
rare gobletBOT
Ayo? @rancid pine level 3 !!!
rancid pine
It was already isolated from the BGM
rancid pine
I am trying to restart the training process but I am getting this error
cobalt eagle
im a bit new to this, I dont know which one is the best, as at first I had only heard about the first rvc thing but now apparently there are forks and people are saying to use mangio? im not sure if it makes a difference and I dont know which fork is the best, does anyone know?
true vortex
what? the message was deleted
humble swan
i try to download a rvc but when i go to the link it said invalid username or password
molten pecan
@true vortex
molten pecan
i try to download a rvc but when i go to the link it said invalid username or pa...
Which RVC exactly?
What's the full name?
true vortex
not necessarily
depends on the size of the data
tiny totem
hello
Google notebook, I am training a voice but in the Index Training part it gives me an error
rare gobletBOT
Ayo? @tiny totem level 1 !!!
tiny totem
how can I solve it?
fluid topaz
-colab
tall rain
My epochs take 11 minutes to do each. After calculating, because I'm doing 300, it's going to take 3 days to finish. I can't afford to let it run for 3 days. How do I make One Click Training go by quicker?
azure marshBOT
RVC Guides (How to Make AI Cover)
proper shale
My epochs take 11 minutes to do each. After calculating, because I'm doing 300, ...
you may have screwed up by clicking on that
you should've trained index first and then train the model, not one click training.
proud elbow
My epochs take 11 minutes to do each. After calculating, because I'm doing 300, ...
it's either of these:
- the dataset is too long (not an issue if that's you want, but personally I won't use >2 hours dataset even with code tweak to remove the minibatch kmeans part)
- the batch size is too low or too high for your GPU vram (especially more than 16)
- there are some running games or apps that use VRAM and/or GPU process
rancid pine
What I was trying to say is that before I did the extra UVR steps, my model was ...
How do I fix the cracks? Do I need to fully restart and start from scratch, compiling another dataset?
If so, that's going to be a hard pill to swallow. Considering I've gathered around an hour of audio..
acoustic garnet
How do I fix the cracks? Do I need to fully restart and start from scratch, comp...
if cracks a present in your final model, first make sure the input sample is 100% clean and not the one causing the cracks. Consider testing an earlier epoch of the model, which sometimes gives smoother results. You may need to look into the dataset to see if there is any bad bits that the model has learnt, trim them out and try training again.
acoustic garnet
im a bit new to this, I dont know which one is the best, as at first I had only ...
Consider just using the original mainline rvc. Its more than usuable since it has rmvpe which is condidered better than mangio-crepe anyways.
-rvc
proper shale
-local
azure marshBOT
acoustic garnet
yes that one
rancid pine
if cracks a present in your final model, first make sure the input sample is 100...
Are there any other solutions? I'm pretty sure it's because the audio has sudden cuts, to get rid of that I'd have to do extensive clipping and reviewing all the audio rips one by one which is.. not very efficient
I tried an earlier Epoch of the model, and it wasn't more smooth, just retained more of the individuality of the voice
acoustic garnet
Not really. Sudden cuts are kind of just a bad dataset. consider adding fade in and outs to redice the effects of the cuts
proper shale
you wouldn't really need an hour for most models
you can just get 10-20 mins and it'll be fine
rancid pine
Would that ensure less pitch cracks?
proper shale
Would that ensure less pitch cracks?
well not necessarily, this could just make your life easier if you want to check if the dataset was fully clean
rancid pine
Good point
proper shale
maybe this is just the audio that you're converting that's the issue, or the pitch extractor if you're using anything other than rvmpe and mangio-crepe
stray ermine
need help
i have ai voiuce changer
it works sometimes but sometimes it says wait web server
but when it works i doesn't let me hear me own voice
rancid pine
How awful is my graph
Just to gauge how much I messed up before I attempt to make a better dataset
rancid pine
I was hoping it wasn't loss g total, but I definitely fucked up
proper shale
I was hoping it wasn't loss g total, but I definitely fucked up
oh that's just resuming gone wrong
but damn that graph flatlined
rancid pine
oh that's just resuming gone wrong
Can you give me an example?
It took me two days for 200 Epoch, and I'm assuming 48k steps?
proper shale
Can you give me an example?
the graph going back to a previous step horizontally is when d/g hasn't saved properly
haven't*
fringe linden
When i try to load another model beside default
proud elbow
How awful is my graph
you probably either resumed training with different settings such as batch size, or enabled "reload last checkpoint on mode collapse" option in Applio
copper pebble
hey guys, any link to train the RVC model?
heres the guide
u can find the link in it
azure marshBOT
RVC Guides (How to Make AI Cover)
grand wind
what rvc should i use for low end laptop?
fringe linden
Where can i download models for RVC
(Note that any other model i try to load results in this error #✨│ai-help message )
copper pebble
what rvc should i use for low end laptop?
training?
proper shale
Where can i download models for RVC
just add them manually to the folders
pth in weights
make a folder in logs with the model's for the index
distant wagon
Is there any way to adjust the tone of voice? D:
glad zealot
what rvc should i use for low end laptop?
If training use mainline on kaggle.. I don't have a copy paste link tho.. if just doing covers you can use colab covergen
fluid matrix
(Note that any other model i try to load results in this error https://discord.c...
usually means you have to reinstall python, try version 3.10
rancid pine
you probably either resumed training with different settings such as batch size,...
The former, yes
molten pecan
just add them manually to the folders
pth in weights
make a folder in logs with...
That last part
Does that only apply for mainline?
fringe linden
usually means you have to reinstall python, try version 3.10
Avec installing python, reopening the software and selecting another model, (i guess only when i select RVC-2 models ? idk)
i get this error
fluid matrix
Avec installing python, reopening the software and selecting another model, (i g...
have you updated to the latest gui.py? or installed all the necessary packages?
rare gobletBOT
Ayo? @fluid matrix level 49 !!!
fluid matrix
or just try to reinstall everything again, tbh thats way easier
fluid matrix
is it not loading the model at all? or does it just not convert
i think this is usually a file format issue, like its not being read or something. could be wrong
fringe linden
Yeah i guess so too
rare gobletBOT
Ayo? @fringe linden level 3 !!!
fluid matrix
you said loading model, so i assumed it doesnt even registed the .pth files. but i think you meant that it wont convert the audio right?
you selected a model, wanted to convert and got a error?
fringe linden
Im doing the realtime conversion
dapper forge
hmm is it just logging bug or is it gonna affect the model?
the graph just jumped step 17200 to 67000
dapper forge
hmm
I started new training session after 300 epochs with smaller batch size, is it may be a problem?
rare gobletBOT
Ayo? @dapper forge level 2 !!!
dapper forge
also do the batch size affect training result somehow?
glad zealot
No clue you can try the generated models tho
glad zealot
also do the batch size affect training result somehow?
I think so.. but no idea what it does.. I keep forgetting
Higher better most of the time .. at least what I do is just set it to 6-10
dapper forge
Higher better most of the time .. at least what I do is just set it to 6-10
oh thanks
i'm asking because recently i retrained the model with 32 batch size and it worked like a poop
in comparison with previous try with batch size = 8
mint geode
I got a cudnn_ops_infer64_8.dll error, do I have to get Cuda or do I have an alternative where I dont ?
radiant whale
https://discord.com/channels/1159260121998827560/1160926856548200531
THIS LINK https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS#scrollTo=Yj-_npuA_HRB won't let me to finish the train
rare gobletBOT
Ayo? @radiant whale level 1 !!!
proper shale
THIS LINK https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP...
Does it say anything?
radiant whale
Does it say anything?
its say the tab error
TBH they just want me to pay to COLLAB
that's it
proper shale
its say the tab error
"tab error"? Be a bit more specific
brittle wing
300 epochs for a 10 minuite model....
radiant whale
"tab error"? Be a bit more specific
idk how to explain, the error notification pop up
radiant whale
"tab error"? Be a bit more specific
and now this
proper shale
and now this
Oh you used too much GPU
Now you need an alt Google account to do stuff ig
radiant whale
Oh you used too much GPU
because it's take much time to finish it
radiant whale
Oh you used too much GPU
restart again?
proper shale
because it's take much time to finish it
You don't need to train until the end
Just train 100 epochs one day, stop training, then do it again next day
radiant whale
Just train 100 epochs one day, stop training, then do it again next day
does colab will save my data?
radiant whale
You don't need to train until the end
can we train the model using hugginface?
anyone know how to train the model in kits ai?
proper shale
can we train the model using hugginface?
nope
proper shale
does colab will save my data?
if you exported the model using the export model step, yes
hearty pawn
Hey everyone!
Anyone knows the best way to deploy RVC on a server with pretrained model for realtime usage in a game vc?
I recently jumped into this rabbit hole so very very confused.
radiant whale
if you exported the model using the export model step, yes
😭
rare gobletBOT
Ayo? @radiant whale level 2 !!!
tame mica
🗣️🗣️🗣️
hearty pawn
Hey everyone!
Anyone knows the best way to deploy RVC on a server with pretraine...
or if anyone knows a cheap API service that could directly be used without stressing about deployment and scaling
we're building this, to fuel creativity: https://twitter.com/spatialweeb/status/1724344367467663866
brittle wing
I have a model, how can I use it(with text to speech not speech to speech)?
cold raptor
hmm
I started new training session after 300 epochs with smaller batch size, is ...
YES
dont do that
ashen knot
hi, does anyone know if theres a fix to being disconnected from colab everytime???
proper shale
hi, does anyone know if theres a fix to being disconnected from colab everytime?...
use a newer colab
-colab
proper shale
I have a model, how can I use it(with text to speech not speech to speech)?
Only audio files, but you can use services like Elevenlabs to generate audio samples
cold raptor
why?
models need consistent dat when being trained thats why u choose same batch size when u are resuming training
shut remnant
feral bluff
Does anyone know how to remove the vocals from a song without removing the chorus?
warped ledge
what do do if the voice kinda gets cut?
blissful pike
what should I do if my voice repeats itself as an echo after I have spoken a phrase?
dapper forge
what should I do if my voice repeats itself as an echo after I have spoken a phr...
check yout windows mic settings, it might be listening from mic is on (i don't know how excatly that option called in english)
azure marshBOT
RVC Guides (How to Make AI Cover)
fathom anchor
what do i set crepe hop to, help i just downloaded rvc gui
chilly crag
what is the pitch extraction
it says smth like Manigio - Crepe
i found crepe but what is manigo?
low shard
what is the pitch extraction
pitch extraction is a techie way of figuring out the high and low notes in a sound, for the best pitch extraction id suggest you rvmpe
chilly crag
pitch extraction is a techie way of figuring out the high and low notes in a sou...
What do you think is best for this one?
low shard
What do you think is best for this one?
ohh your using a model, well in that post thats one of the training info, you choose a pitch extraction when you train a model,but when you use rvc (make ai covers), you js need to use it, and be careful abt the pitch you put,sometimes some models may need to lower or high it up,that depends also on the voice of the audio you want to change
chilly crag
ohh your using a model, well in that post thats one of the training info, you ch...
i see, im tryna figure this out lol but still not 💯 % sure
low shard
i see, im tryna figure this out lol but still not 💯 % sure
if you want you can ask me if you got more questions
chilly crag
if you want you can ask me if you got more questions
okay
rare gobletBOT
Ayo? @chilly crag level 1 !!!
low shard
you make ai covers using ilaria rvc ? or u do it locally?
chilly crag
you make ai covers using ilaria rvc ? or u do it locally?
ilaria rvc?
uhm
not sure 
im using this "voice changer client demo"
its from a yt video
should i try Ilaria rvc?
low shard
im using this "voice changer client demo"
what ?
chilly crag
yeah ill try that
low shard
are you trying to have a realtime voice changer or making an ai cover?
chilly crag
realtime
low shard
ohhhhh
chilly crag
which one is better you think?
chilly crag
w-okada..
copper pebble
yh w-okada goated
chilly crag
alright
low shard
w-okada..
https://discord.com/channels/1159260121998827560/1159620272316502096 these are the guides, please if you got any problems ask for it in #🔍│help-w-okada
chilly crag
ok thanks 😄
low shard
your welcome!
grand surge
I'm being told I have no CUDA GPUs available when trying to generate for some reason. it was working fine a moment ago (I'm using Hina Mod AICoverGen)
low shard
I'm being told I have no CUDA GPUs available when trying to generate for some re...
make sure you're connected to a gpu runtime. you can check this by going to "Runtime" -> "Change runtime type" and make sure 'GPU' is selected under 'Hardware accelerator'. it could be u runned out of the free gpu daily limit of colab
grand surge
make sure you're connected to a gpu runtime. you can check this by going to "Run...
are these colab solutions hosted locally or remotely? if it's remote i'd see why you'd have a gpu limit, and i would also want to do it locally. if not then ig i'm screwed
copper pebble
are these colab solutions hosted locally or remotely? if it's remote i'd see why...
- on cloud
- money
low shard
are these colab solutions hosted locally or remotely? if it's remote i'd see why...
all the stuff you run on google colab is hosted remotely on google's servers. that's why there's a limit on gpu usage, 'cause they gotta share the resources with everyone. also yes theres a local download of rvc (the ai program to make ai covers)
grand surge
can yall just point me to where i can see how i do this locally
rare gobletBOT
Ayo? @grand surge level 1 !!!
copper pebble
@low shard ur honor
icy thicket
Why does my voice have noise from the program? help
chilly crag
i do not sound like spongebob at all
my stuff like too choppy and deep for some reason
even after changing the tune
proven hill
Why does my voice have noise from the program? help
wdym=
proven hill
i do not sound like spongebob at all <a:PepeSob:1141189968740503643>
maybe its the model youre using
amber igloo
Hi, is it normal when I generate an output that the resulting audio mostly creates some gibberish or cut words? If it's a clear speak it's always quite good, but with amateur audios or other recordings sometimes it's impossible to understand
rmvpe local
This for example
cinder wyvern
is there a way to put multiple voice model zip files into one larger zip file? i have a bunch of voices and it takes ages to import them all into the colab every time i want to use it
proper shale
Hi, is it normal when I generate an output that the resulting audio mostly creat...
well yeah because that's how the AI was trained on, clean stuff
u can try adobe enhance for those amateur stuff
amber igloo
So it's not changing some Mangio settings that will fix it, right?
proper shale
So it's not changing some Mangio settings that will fix it, right?
Nope
amber igloo
Thanks
dapper forge
does index training depends on existing model or i can train index with just dataset?
proper shale
does index training depends on existing model or i can train index with just dat...
If you have the preprocessed files and pitch extracted the dataset, yeah
You can
dapper forge
oh thanks
brittle wing
can models also be used for text to speech?
copper pebble
can models also be used for text to speech?
just use the tts and then convert it using the model ya want
wary pecan
Anyone else here unlucky like me and rvc just wont work? Here's the issue I reported on Github https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/issues/1637
I don't know what to do apart from chuck my laptop outside and just play my xbox instead😆😅, anyone got any ideas?
modern oxide
delete the space between the words
deft edge
delete the space between the words
where
modern oxide
in the name
deft edge
@modern oxide erroe make sure your data was preproccesed what does this mean
modern oxide
did you press preprocessing section again?
deft edge
did you press preprocessing section again?
yes but it said success
where's the audio file it is in wav
modern oxide
ah
you put a zip file in dataset folder, you need to put .flac .wav or .mp3 files in there it's not like rvc v2 disconnected
flint seal
ok, so i have this issue where i can use rvc with this link: https://huggingface.co/spaces/Clebersla/RVC_V2_Huggingface_Version but when i attempt to upload a file it does not work, i am only able to record an audio and not upload one. is there a certain format i must use?
azure marshBOT
RVC Guides (How to Make AI Cover)
rare gobletBOT
Ayo? @pliant steppe level 3 !!!
flint seal
?
round fog
wanton knot
i dont undertsnad the authentication
i installed the app (mac)?
i have the authentication code
what do i do
tutorial?
mighty parcel
you have non utf-8 characters in file names
you know, being korean
just name stuff in eng
are there spaces in any file names?
@pliant steppe what does the cmd window say, the black window with the text
cmd 창에 뭐라고 적혀 있나요? 텍스트가 있는 검은 창
mighty parcel
'traceback'이 표시된 부분으로 스크롤하세요.
검은 창문에
'추적'이라고 적힌 부분을 보여주세요
'traceback'
저 창에 있어요, 두 번 스크롤해서 지나쳤어요
위로 가다
그냥 천천히 내려가
또 다른 텍스트 오류야
파일 이름 또는 파일 경로에 'utf-8'에 대한 잘못된 문자가 있습니다.
'rvc'는 중국 개발자가 만든 것인데 다른 언어에 대한 검사도 제대로 되어 있지 않아서 그렇습니다.
한국어가 아닌 루트 폴더에 넣어보세요
네 그냥 C드라이브에 폴더 만들고 영어로 이름 짓으세요
그리고 그것은 작동해야
그렇지 않은 경우 'logs' 폴더 경로도 utf-8이어야 합니다.
작동하지 않으면
이것은 내 로그 폴더 경로의 예입니다.
응, 아직도 한국어 텍스트 오류야
아니요
C 드라이브에 폴더를 만드는 것이 좋습니다
예를 들어, 데이터 세트를 어디에 두는지
하지만 거기에 rvc 폴더를 넣어
그러면 경로에 한국어 문자가 없습니다.
아니요
이상합니다. C 폴더의 루트에 rvc가 있다는 점을 고려하면 작동할 것입니다.
작동해야 하는데 그렇지 않습니다.
아 알겠습니다, 그럼 다시 시도해 보세요
우선 작동에만 집중하자
좋게 만드는 대신
알았어 작동 중이야
예
오류는 괜찮습니다
오류는 배치를 20으로 설정했기 때문에 발생합니다.
8을 사용
그 외에는 가셔도 좋을 것 같아요
예를 들어, 문제가 없어야 합니다
구글은 번역을 잘 못해요
응 괜찮아
진행되고 있다는 뜻이다
예, 'rvc' 훈련 속도는 'gpu' 속도와 데이터세트 길이에 따라 달라집니다.
하지만 '3060'은 느리지 않아요
'코 랩'이 무슨 뜻이에요?
검은 창을 클릭하세요
그런 다음 마우스로 검은색 창을 클릭하세요.
그런 다음 Enter를 클릭하십시오
검은 창을 클릭하지 마세요
클릭하면 일시 정지됩니다
하얀 상자 보이시죠?
흰색 상자는 클릭했다는 의미입니다.
그리고 일시중지됐어
그러니 Enter를 누르세요
텍스트 상자를 클릭하지 마세요.
어쨌든 가는 게 좋을 것 같아
나 다시 오수로 갈게
vapid gust
yo
why can’t my model replicate the proper accent on my input audio
im using stems to create the dataset but im not getting the right voice i wanted
tame mica
어쨌든 가는 게 좋을 것 같아
felt Korean era??? 
vapid gust
Can someone help me?
azure marshBOT
RVC Guides (How to Make AI Cover)
solid light
how do i download rvc?
manic swift
Are there any settings that i can use to decrease delay? I have a pretty beefy system but even at the max rvc still has a major delay is this just something that i have to wait for the ai to get better
or is there hardware changes that can increase the speed
low shard
how do i download rvc?
heres how to download locally rvc https://rentry.co/TrainingVoiceModels
quasi falcon
Can someone please tell me how to fix the error in the file "run_easiergui.bat"?
brittle wing
Can someone please tell me how to fix the error in the file "run_easiergui.bat"?
i dont know
proper shale
Can someone please tell me how to fix the error in the file "run_easiergui.bat"?
Replace this file with this version from Mangio tweaks
quasi falcon
Replace this file with this version from Mangio tweaks
Which file exactly now?
Ah, I think I have it.
rare gobletBOT
Ayo? @quasi falcon level 2 !!!
quasi falcon
Oh, now I'm getting this error:
brittle wing
how do i fix this?
glad zealot
how do i fix this?
verify your ngrok account
brittle wing
verify your ngrok account
how do i do that?
rare gobletBOT
Ayo? @brittle wing level 5 !!!
brittle wing
Ayo? <@456226577798135808> level **5** !!! <:lfg:1159355870119993496>
LESSS GOOOO!
brittle wing
go to the ngrok website and verify it
link?
brittle wing
literally on the eroor
nvm i found it i just don't know how to verfiy
burnt dew
У кого есть марат или адидас?
distant wagon
uhhh, it was fine yesterday, but why did the voice change today?
quasi falcon
Oh, now I'm getting this error:
Can someone please help me with my problem?
copper pebble
Can someone please help me with my problem?
i think thats bit outdated
ya wanna train or inference?
whats ur gpu?
solid light
how do i start RVC? i pressed start but it only delivers my voice
rare gobletBOT
Ayo? @solid light level 1 !!!
shy ingot
Helloo, i have a problem with the voice changer crashing for some reason i don't know, but idk if this is the right channel to ask for help in this kind of issue, any help?
glad zealot
uhhh, it was fine yesterday, but why did the voice change today?
uhhh try switching to another model and switch back?
warm needle
Hello guys, i have a question, i am back to rvc, id like to make more ai covers, but i was tryna find link for rvc and seems like there is different versions, which one is the best version to use for ai voice covers?
quasi falcon
ya wanna train or inference?
Both
quasi falcon
whats ur gpu?
RTX 3080
azure marshBOT
RVC Guides (How to Make AI Cover)
astral jungle
Yoooo
quasi falcon
name has spaces
I've already changed that, but it doesn't change anything.
quasi falcon
https://discord.com/channels/1159260121998827560/1159618293594861579
I'm a bit confused as to exactly what this is for.
copper pebble
train and inf
proper shale
I've already changed that, but it doesn't change anything.
Just make sure to remove the spaces ig
quasi falcon
train and inf
What is the inference for? What can you do with it?
copper pebble
inference (covers)
quasi falcon
inference (covers)
Can I make these ur ur ur sounds from the Freddy Fazbear meme?
copper pebble
like i said yesterday u need the model of that
just the ur ur
u could train it if it doesnt exists
distant wagon
uhhh try switching to another model and switch back?
not work
quasi falcon
u could train it if it doesnt exists
Yes, but even if I train it, it still says the words that are sung. Because you can only train a model with the pitch of the person , or have I misunderstood?
proper shale
Yes, but even if I train it, it still says the words that are sung. Because you ...
well not really, it trains it based on the sounds emitted too so it'll just sound like distorted ur urs
ig
molten pecan
how do you set it up? it depends on which version you are using. each one has its own way to install
which one should you choose? it depends on your needs
let's explain
the three main version are Mainline RVC, Applio, and Mangio
just casually or a bit more profesionally?
follow the section only
it's pretty easy
you mean storage or RAM?
for storage, I remember it was bigger than 5gb
idk if 5 or 7
quasi falcon
Why does the download take so long?
molten pecan
Why does the download take so long?
did you duplicate the space?
it's highly recommended to avoid errors and improve performance
quasi falcon
did you duplicate the space?
How do you mean?
molten pecan
How do you mean?
I explain it here https://rentry.org/ilarvc_inf_guide
quasi falcon
But I use it locally on my PC
molten pecan
I also higly recommend using that hugging face space instead. but pick whichever you like :3
ohhh I thought that was HF
did you try rebooting RVC?
@copper pebble can they?
quasi falcon
ohhh I thought that was HF
Oh sorry, I forgot to mention it before.
rare gobletBOT
Ayo? @quasi falcon level 3 !!!
molten pecan
mainline RVC is just a folder, not an app that you install locally through an exe
quasi falcon
did you try rebooting RVC?
No, not yet
molten pecan
you can move it wherever you want
sand iris
it should work
quasi falcon
did you try rebooting RVC?
Can the Ai voice model also be inserted manually? Is there a folder for this?
molten pecan
Can the Ai voice model also be inserted manually? Is there a folder for this?
yes. do this:
quasi falcon
yes. do this:
What is the name of the folder?
molten pecan
What is the name of the folder?
- download the model
- open the model's folder
- open the internal files of RVC
- go to the
weightsfolder and put the model's ,pth file there - go to
logsfolder and put the .index file there
latent prism
how do I delete a model slot in rvc ?
molten pecan
how do I delete a model slot in rvc ?
local rvc?
latent prism
yes
astral jungle
Hi
latent prism
found it thanks
quasi falcon
1. download the model
2. open the model's folder
3. open the internal files of R...
Ah thank you!
molten pecan
Ah thank you!
yw!
quasi falcon
Because I only have problems with my local RVC. I would like to install the latest version of it so that I can perhaps remove such errors. Can someone please tell me where I can download the latest version?
molten pecan
what were you trying to do there?
unzip it?
are you using winrar?
hm then try using winrar
haven't seen that problem before
proud elbow
probably turn off antivirus & firewall and redownload
quasi falcon
Hello, can anyone tell me how I can rectify the error? It is with the file: run_easiergui.bat
rare gobletBOT
Ayo? @brittle wing level 4 !!!
old oracle
is there a way to make this -100?
molten pecan
Have you turned off firewall and whitelisted RVC?
normal sigil
hi um I’m really a beginner in making covers, so now that I already downloaded the voice, what do I do? I already have the vocals and instrumental separated btw.
tacit jetty
is it good when your learning rate is like this
sand iris
Google "how to whitelist a folder from the firewall"
... It's not having firewall issues there
Check your nvidia driver install/if CUDA is available
molten pecan
Check your nvidia driver install/if CUDA is available
Why wouldn't it be available?
sand iris
Because it's failing to load a library that makes use of some cuda libraries
And supposedly, when you try to load a dynamic library (dll), which has prerequisites, it can just throw a winerror 126 and do exactly this
tacit jetty
is it good when your learning rate is like this
Is this good fired
sand iris
I don't train models, I don't know about that 😅
molten pecan
And supposedly, when you try to load a dynamic library (dll), which has prerequi...
So what's the solution? Reboot CUDA? Is that a thing you can do?
I'm trying to learn
I'm still an apprentice
_<
sand iris
Make sure it's installed on the system side
tacit jetty
I don't train models, I don't know about that 😅
Alr julia do you know
molten pecan
is it good when your learning rate is like this
Uhhh that's very unusual
Usually is not as straight as that line
azure marshBOT
RVC Guides (How to Make AI Cover)
tacit jetty
Usually is not as straight as that line
That is the learning rate
molten pecan
I don't train models either but I have seen lots of ppl's tensors but they were never like dat
Check how the model sounds like. Maybe we shouldn't worry
tacit jetty
Im gonna ask someone that knows it but @molten pecan thx for trying to help
proud elbow
is it good when your learning rate is like this
it's normal by default in every RVC fork
old oracle
is there a way to make a ai without the "background vocals" or "music" part going to only -20 cause i wanna mix the vocal and those just bleed into the lead
tacit jetty
it's normal by default in every RVC fork
Desuwa thx for answering fast can you tell me if the grand_norm_d chart is looking good
proud elbow
Desuwa thx for answering fast can you tell me if the grand_norm_d chart is looki...
idk but only d/total, g/total, mel, fm, kl might give meaningful indicator
tacit jetty
idk but only d/total, g/total, mel, fm, kl might give meaningful indicator
Wich one is the most meaningful one
molten pecan
Desuwa thx for answering fast can you tell me if the grand_norm_d chart is looki...
You should ask in #📑│making-models
proud elbow
Wich one is the most meaningful one
-overtrain
azure marshBOT
Hey, @tacit jetty!
👇 Here are some resources to help you identify if your model is overtraining
All-In-One Guide on how to make a good model
This guide explains how the D and G files works and much more: https://rentry.org/RVC_making-models
Credits: LUSBERT 
Automated Overtraining Detection (AOD)
Will be available soon in #1159513888199540817
Credits: grvyscale
tacit jetty
-overtrain
Thx for the documentations how long do you usually train a model without looking at the graphs
balmy imp
A noob-ish question - do I need to provide high quality vocals for RVC? Unsure where I can get non-copyrighted vocals that sound good and would be a good target pitch
acoustic forge
Where should .Index files be placed?
rare gobletBOT
Ayo? @acoustic forge level 1 !!!
sweet trench
hello! i just install local mangio-rvc locally for the first time but it keeps saying error when i convert. may i ask what the issue could be? is it ok to only have .pth file in your weight folder or do you also need the index?
molten pecan
hello! i just install local mangio-rvc locally for the first time but it keeps s...
the .index file is not a requierement for a model to function. it includes features from the dataset it was trained with, like accent, speech manner, etc.
old oracle
-colab
molten pecan
it's extra information
it makes the model sounds better
did you try selecting another model?
if it keeps happening then seems like mangio is not working so just reboot it
sweet trench
thank youuu may i ask that by rebooting, did you mean uninstalling it and reinstalling?
molten pecan
thank youuu may i ask that by rebooting, did you mean uninstalling it and reinst...
ah just closing it and opening it
not reinstalling
good luck!
sweet trench
okayyy thank you very much 🥺🥺🥺
balmy imp
A noob-ish question - do I need to provide high quality vocals for RVC? Unsure w...
bumping
rare gobletBOT
Ayo? @sweet trench level 1 !!!
molten pecan
A noob-ish question - do I need to provide high quality vocals for RVC? Unsure w...
yes of course, the higher the quality of the audio sample, the better the output quality
sometimes artists release the official vocals, so try googling the name of the song followed by "official vocals" or something like that
if not, you can extract the vocals yourself with UVR or MVSEP
molten pecan
make sure the audio is in WAV or FLAC and not MP3 too
MP3 by default compresses the original quality so it's lighter
warm needle
Guys, what can i use to seperate vocals and instrumental from the song? I used to use some ai software to do that but I don't remember what was it
rare gobletBOT
Ayo? @warm needle level 2 !!!
thin stump
Guys, what can i use to seperate vocals and instrumental from the song? I used t...
Use ensemble mode and combine the recommended models together then you can apply denoise or dereverb-echo afterwards
warm needle
ooo that was it! Thanks alot!
proud elbow
bumping
tbh streams/podcasts are more preferrable to soundtracks (even though you manage to get 'official' acapella you still need to separate back vocals, harmonies, vocoders, etc). the youtube's lossy 128 kbps AAC compression isn't even really a hindrance if you can do audio separation, cleaning, and denoising properly.
thin stump
ooo that was it! Thanks alot!
np, it's not always perfect. You should set the window size to 1024 or more
warm needle
I have a problem, I used to use this program couple months ago but now i get an error Error Occured Make sure you've chosen the correct GPU
I made sure and selected my video card
the same one i've been using all the time
hot pumice
does anyone have a recommended gui for training and using rvc models? and if so is there a tutorial for it too?
rare gobletBOT
Ayo? @hot pumice level 1 !!!
shrewd axle
can someone send me Thomas chick's ai models, please?
⠀⠀⠀⠀⠀⠀⠀⠀⢀⣤⣶⣿⣿⣿⣷⣶⣄⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⣀⣀⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⢀⣴⣿⣿⣟⡯⣷⣫⢟⡿⣿⣷⣄⠀⠀⠀⠀⠀⠀⠀⠀⠀⣠⣴⣶⣿⣿⣿⣿⣿⣿⣷⣄⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⣴⣿⣿⣟⣳⠾⣽⠶⣏⡿⣞⣽⣻⣿⣧⠀⠀⠀⠀⠀⢀⣴⣿⣿⣿⢻⣏⣟⡾⣵⣫⣟⡽⣿⣷⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⢀⣾⣿⣟⡷⢾⣭⣟⣳⠿⣭⢷⣛⣶⣛⡾⣿⣇⠀⠀⠀⣰⣿⣿⡟⣧⢿⣹⡞⣧⣟⠾⣵⢾⣹⢯⣿⣿⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⢀⣾⣿⣟⢾⣹⢷⡳⡾⣭⢿⣹⠾⣝⡶⢯⣽⣻⣿⡀⠀⢰⣿⣿⣳⢻⡽⣞⣧⢟⣳⡞⣿⣹⢾⣹⢾⣹⣿⡇⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⣾⣿⣟⡞⣯⡽⣞⣽⡳⢯⣏⣷⣻⣭⢟⣻⠶⣯⣿⡇⠀⣿⣿⣳⡽⣫⢷⣻⡼⢯⣳⢿⣱⢯⢷⣫⢯⠷⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⣸⣿⣿⢺⡽⣳⡽⣞⢧⣟⠿⣼⣣⢷⡞⣯⣏⡿⣵⣿⣿⢸⣿⣯⠷⣽⣫⢷⡳⢯⡟⣧⢿⣹⡞⣯⣳⢯⠿⣽⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⣿⣿⣽⢫⡷⣏⡷⢯⣛⡾⣛⣧⡟⣧⣟⣳⠾⣝⣾⣿⣷⣿⣿⣭⢟⣧⣟⣳⢟⣯⡽⣛⣾⢣⣟⣳⣭⡟⣯⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⢸⣿⡿⣞⢯⢷⣫⡽⣏⣯⢷⣻⡼⣽⢳⡾⣭⢿⣹⢾⣿⣿⣿⡷⢯⣞⢷⣺⠽⣞⣧⠿⣝⣞⡟⣞⣧⢷⡻⣵⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⢸⣿⡿⣭⣟⢾⣳⡽⣫⣞⠷⣽⣺⣭⠷⣏⡷⣏⣟⡾⣿⣿⣿⣻⢻⡼⣏⣷⡻⣽⢮⢿⣹⢮⣟⡽⢮⡷⣛⣷⣿⡇⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⢸⣿⣟⣳⡞⣯⢶⣛⣷⡹⣟⣧⢷⣫⠿⣭⢷⣛⣮⣽⣻⢿⡯⢷⣏⣟⡾⣱⢿⣱⢯⣯⣳⠿⣜⣻⡽⣳⢯⣿⣿⠃⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⢸⣿⣯⣗⣻⣭⢯⣛⣶⢻⣳⡞⣯⣳⢿⣹⠾⣝⡾⣖⡯⣷⣛⣯⢞⣧⢿⣹⢯⡽⣾⣱⢯⣻⣝⣧⠿⣭⣿⣿⣿⣤⣤⣀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⢸⣿⣷⣫⠷⣞⣯⡽⣞⢯⣳⡽⣳⣽⣾⣿⣿⣿⣷⣯⣷⡳⢯⣞⣻⡼⢯⣷⣿⣿⣿⣿⣿⣷⣞⣞⢿⣱⡟⣿⣟⡿⣿⢿⣿⣷⣦⡀⠀⠀⠀⠀⠀
⢸⣿⡷⣏⡿⣳⠾⣵⢯⣻⡵⣻⣿⡿⠋⠁⠀⠀⠙⢿⣿⣽⣻⡼⣳⣻⣿⡟⠉⠀⣀⡀⠈⢻⣿⣿⡞⣧⣟⣳⠾⣝⡾⣫⣽⢻⡿⣿⣶⡀⠀⠀⠀
⢸⣿⡿⣭⢷⣏⡿⣝⡾⣣⢿⣽⣿⡇⠀⢠⣿⣷⠆⢸⣿⣷⣳⣻⡵⣿⣿⠀⠀⠰⣿⣿⠀⠀⣿⣿⣽⡳⡾⣭⠿⣭⢷⣻⡼⢯⣽⣛⣿⣿⣆⠀⠀
⢸⣿⣿⣹⡞⡾⣝⡾⡽⣽⣣⢿⣿⣧⢄⠀⠙⠋⢀⣼⣿⣯⣗⣳⡽⣻⣿⣷⡠⠀⡈⠁⢀⣤⣿⣿⣳⡽⣏⣷⣻⣭⢷⣳⣛⣯⢶⣛⣞⣿⣿⡄⠀
⣿⣿⣷⣣⢿⣝⣯⠾⣝⣧⢿⣹⢿⣿⣷⣶⣶⣶⣿⣿⣟⡶⣏⣷⢻⣵⣻⢿⣿⣾⣶⣿⣿⣿⢿⣏⡷⡽⢾⣱⠷⣞⣯⢞⡽⣮⢟⡽⣞⢾⣿⡇⠀
⣿⣿⣭⡟⣾⣹⡽⢶⣻⣝⡾⣝⣮⢟⡿⣿⢿⡿⣟⣯⣽⢳⣽⢫⡾⣝⣯⢞⣿⣻⢿⡿⣻⢟⣮⣽⢳⢷⡺⣟⢶⢯⡽⣞⢷⣹⡞⡷⣽⢾⣿⠇⠀
⢿⣿⣷⢻⡵⢯⣞⡯⢷⡞⣽⡞⣽⢾⣹⣞⣳⣯⣽⢶⣫⡟⣞⡯⢷⣻⡼⣛⣶⣿⣷⣻⡽⣞⣳⡞⣯⢯⢷⣏⡿⣞⡽⣞⢯⣳⣻⢽⣾⣿⡟⠀⠀
⠘⢿⣿⣯⣞⠿⣼⢻⡽⣞⣳⣻⠽⣞⣳⣞⣳⢿⣿⣯⣷⣻⣝⡾⣏⡷⣽⣿⣿⡿⣟⣧⢟⣞⣧⢿⣹⡞⣯⣞⣳⢯⡽⣞⢯⣳⣽⣿⡿⠋⠀⠀⠀
⠀⠈⠻⣿⣾⣿⣭⡟⣾⡝⣧⡟⣯⣏⡷⢾⣭⢿⣹⢿⣿⣿⣾⣷⣿⣿⣿⡿⣟⣽⡳⣞⣻⢞⣮⢯⣗⣻⣵⣾⣽⣾⣽⣾⣿⡿⠟⠋⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠙⠻⢿⣿⣷⣿⣧⣟⣳⠾⣝⣯⡞⣷⣫⣟⣼⣯⣟⣿⣛⣯⢯⡽⣞⣧⠿⣭⢷⡻⣮⢷⣫⢷⣿⣿⡿⣿⣿⣏⠁⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠈⠉⠛⠻⠿⠿⣿⣿⣷⣿⣿⣿⣿⣿⣿⣿⣾⣹⢮⢯⢷⣻⡼⣻⠽⣞⣽⢳⡯⣽⣻⣿⣷⣟⡴⣛⣿⣷⣄⢀⣴⣾⣿⣿⡄⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢀⣀⣀⠀⠀⢀⣿⡿⣭⢻⣿⡷⣏⣯⢟⣞⣧⢿⣹⣻⣝⡞⣯⢷⣫⣽⣿⠟⢿⣷⣝⣦⢻⣿⣿⡿⣯⣿⣿⠃⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⣴⣿⡿⣿⣷⣦⣾⣿⠽⣼⣿⣿⣿⣽⣚⣯⣞⢾⣭⢷⣳⠾⣝⣯⡞⣷⣿⡟⠀⠀⢹⣿⣿⣣⠞⣭⣿⣿⣿⣷⣦⡀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠈⢿⣿⣷⣥⣟⡻⡿⣏⣿⣿⠏⠘⣿⣷⣻⡖⣯⣻⡼⣳⣏⡿⣽⣲⢟⣿⡿⠁⠀⣼⣿⣟⣳⣧⡟⣶⣧⣿⣜⣹⣿⣷
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢠⣿⣿⣿⣯⣳⣙⢶⣿⣿⣤⣀⠘⢿⣷⣟⣳⠷⣽⣳⠾⣽⠶⣯⣿⡟⠁⠀⠀⢿⣿⣾⣿⣿⣿⣹⣿⠻⠿⠿⠿⠋
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢠⣿⣿⡹⣶⣿⣿⢾⣿⣾⣝⣿⣿⡆⢸⣿⣿⣿⣿⣧⣯⣿⣷⣿⣿⣿⡇⠀⠀⠀⠀⠉⠉⠘⢿⣷⣿⣿⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠻⣿⣿⡿⣿⣟⣾⣿⠟⠿⠿⠿⠃⣸⣿⣏⢿⣿⡿⠿⠿⣿⣟⢧⣿⣷⠀⠀⠀⠀⠀⠀⠀⠈⠙⠋⠁⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠻⢿⣿⠿⠀⠀⠀⠀⠀⣿⣿⣎⢿⣿⡇⠀⠀⣿⡿⣎⣽⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⣀⣀⣀⣿⡿⡜⢾⣿⡇⠀⠀⣿⣿⠲⣽⣿⣶⣶⣶⣦⣄⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢀⣴⣶⣿⣿⢿⡿⡿⢿⡟⣭⢛⣿⡇⠀⠀⣿⣿⣙⠶⡹⢏⡿⣹⢛⣿⣷⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢸⣿⣿⡱⢎⣧⢳⣙⠧⣞⣜⣻⣿⠇⠀⠀⣿⡿⣬⢛⣭⢳⣚⣥⣻⣼⣿⠃⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠈⠻⢿⣿⣿⣶⣿⣾⣿⣾⣿⠿⠃⠀⠀⠀⠻⢿⣿⣿⣿⣿⠿⠿⠿⠛⠁⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
fluid topaz
sand iris
can you not
shrewd axle
k
brittle wing
where is the google colab link
neat peak
hey guys, when i installing google colab in python (pip install google-colab) to be able to use a local run time , i keep getting that error , i searched a lot, used chatGPT, used google AI, no solution woked, so if anyone can help with error tell me what to do ro share any link please
here's the error below and the python notes. hits and everything:
error: subprocess-exited-with-error
× python setup.py egg_info did not run successfully.
│ exit code: 1
╰─> [15 lines of output]
C:\Users\Yahia\AppData\Local\Temp\pip-install-6ic5a5lt\pandas_0a941bd0b3354a84bc2cccc61655b8c5\setup.py:12: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
import pkg_resources
C:\Users\Yahia\AppData\Local\Programs\Python\Python39\lib\site-packages\setuptools_init_.py:80: _DeprecatedInstaller: setuptools.installer and fetch_build_eggs are deprecated.
!!
********************************************************************************
Requirements should be satisfied by a PEP 517 installer.
If you are using pip, you can try `pip install --use-pep517`.
********************************************************************************
!!
dist.fetch_build_eggs(dist.setup_requires)
error in pandas setup command: 'install_requires' must be a string or list of strings containing valid project/version requirement specifiers; Expected end or semicolon (after version specifier)
pytz >= 2011k
~~~~~~~^
[end of output]
note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed
× Encountered error while generating package metadata.
╰─> See above for output.
note: This is an issue with the package mentioned above, not pip.
hint: See above for details.
sand iris
hey guys, when i installing google colab in python (pip install google-colab) to...
Not sure that's supposed to work like that
seeing how the latest release for the Google-colab python package is from 2019
things have changed a lot since then, if that's the way it's supposed to work, it should have seen newer releases since then
https://github.com/googlecolab/colabtools/issues/266
tldr: port forwarding
brittle wing
anyone know how to fix this FileNotFoundError Traceback (most recent call last)
<ipython-input-16-2b1fae8bf49c> in <cell line: 16>()
16 if ezmode:
17 if tunnel == "cloudflared":
---> 18 for line in fileinput.FileInput(f'Easier{weeee}.py', inplace=True):
19 if line.strip() == 'app.queue(concurrency_count=511, max_size=1022).launch(share=True, quiet=True)':
20 # replace the line with the edited version
neat peak
https://github.com/googlecolab/colabtools/issues/266
tldr: port forwarding
I used the same settings posted on Google colab website
golden sundial
how do i fix the error
-102 ERR_CONNECTION_REFUSED http://localhost:18888/
Preparing your application. Please wait...
(node:248) electron: Failed to load URL: http://localhost:18888/ with error: ERR_CONNECTION_REFUSED
its just saying it on repeat
when i try and do start_http
neat peak
Mind sharing them here ?
sand iris
Sure,
https://research.google.com/colaboratory/local-runtimes.html
Not sure if the localized french version is different, but I don't see them mentioning the package you're trying to install
Thing is, you're gonna need port forwarding if you haven't got that handled already
neat peak
Yeah they didn't mention pip install google-colab,, i didn't do it at the beginning but Google ai who told me to install it
when i first tried i runned the cell on Google colab website Normaly
The error after running Google colab on localhost i got "ModuleNotFoundError: no module named 'google.colab'
So Google Ai prompt to help and said try 'pip install google-colab'
neat peak
Not sure if the localized french version is different, but I don't see them ment...
Anyways is there's a better method but without GPU, Using CPU?
sand iris
You're trying to create new voice models or do covers/voice conversion?
vague sage
whats the most realistic girl voice you guys think?
neat peak
You're trying to create new voice models or do covers/voice conversion?
New voice model
sand iris
New voice model
You can use the "RVC Disconnected" or "EasyGUI" notebooks listed here :
-colab
neat peak
You can use the "RVC Disconnected" or "EasyGUI" notebooks listed here :
It's same Google colab, it works with me actually and i did Created an model earlier with it but the only problem with Google colab is sometimes it gets "Runtime disconnected" it happens a lot actually, so creating the model is stopping of course and all the time and internet has been spent for nothing when i get disconnected and i gotta start over again because there's no option to continue from where you've left that's why i was trying to set a local runtime or find anything with a stable runtime or option to continue from where you've left in case of runtime disconnection
sand iris
Oh yeah. not sure how to work around runtime disconnections, but I'm afraid there aren't other options than that, or running locally with a supported GPU
neat peak
Oh yeah. not sure how to work around runtime disconnections, but I'm afraid ther...
Okay thank you a lot i appreciate your answers 😊
rare gobletBOT
Ayo? @neat peak level 2 !!!
visual juniper
hey wondering if anyone can help me figure out how to import an use them in the RBVC WebUI
proper shale
hey wondering if anyone can help me figure out how to import an use them in the ...
just add the .pth to the weights folder and, for the index, create a folder in logs with your model's name.
visual juniper
just add the .pth to the weights folder and, for the index, create a folder in l...
So i usually enter the Model name and epoch in the "restore pth from google cloud drive" cell and run it. So i am not sure How this works?
proper shale
So i usually enter the Model name and epoch in the "restore pth from google clou...
hm? what colab is that
I don't think I've seen that one before
visual juniper
hm? what colab is that
Retreval Based Voice Conversion WebUI
rare gobletBOT
Ayo? @visual juniper level 1 !!!
visual juniper
proper shale
I'd use Ilaria RVC, because there's a high chance this colab has been restricted
-colab
proper shale
There's also a guide for it on #1159513888199540817
humble knot
hrlp
help
i confused
the voice changer really glitchy
i have amd 6600 xt
chunk 512
extra 8192
quasi falcon
u could train it if it doesnt exists
I have now trained an "ur ur ur" model, but he still only sings the songs.
half cove
the voice changer really glitchy
alpine sequoia
excuse me, i have the rvc that gives the voice live, how can i feed some vocals (or my system input that is playing an acepella) into the rvc engine?
cinder wyvern
is there a way to put multiple voice model zip files into one larger zip file? i have a bunch of voices and it takes ages to import them all into the colab every time i want to use it
ancient stratus
now I know this has probly been asked before but what does "Search Feature Ratio" do specifically?
azure marshBOT
RVC Guides (How to Make AI Cover)
wanton knot
what do i do after iv'e verified my ngrok account, i have my number and such
molten pecan
now I know this has probly been asked before but what does "Search Feature Ratio...
Search feature ratio
This affects how much the of the .index will be applied to the audio after inferencing. The higher the value, the more accent and speech manner of the voice model will be applied to the output audio.
For example, if the model's dataset has background noise, reducing this value will lessen that noise in the output.
Lowering this value can fix artifacting issues on some models.
Lowering it to 0 won't make the output not sound like the model, you can lower it as much as you want
if you wanna know what the settings do :3
molten pecan
is there a way to put multiple voice model zip files into one larger zip file? i...
yeah you can. make an empty zip folder and put the .zip models there
though I don't think colab will be able to read them if you put more than one zip in it
cinder wyvern
i did that but it just comes up with an error whenever i try to import it into the colab
rare gobletBOT
Ayo? @cinder wyvern level 1 !!!
molten pecan
it can only read one by one
molten pecan
excuse me, i have the rvc that gives the voice live, how can i feed some vocals ...
let's explain this.
RVC is the name of the voice cloning AI software itself. with With this software you can make two apps, one for converting <audio files>, and the other for converting voice in <real-time> that gets picked up through a mic, as you speak
if you want to convert audio files, you'll need the software for converting audio files
you understand now?
brittle wing
i'm new to making (not using) voice models, how would I go about training a model
noble junco
hello guys i am kinda new to this rvc stuff so i cant seem to get the hang of it i intend to create a model of a very good clear john wick for my video game voice overs but they seem not to be clear enough and some words dont get pronounced properly plz what can i do to make it as good as it can be?
spark lark
uneven scroll
im covering a song where the singer gasps and instead of gasping my model just makes like metal screeches is there any way to stop this and make it normal
ornate bobcat
Hey so uh I was going to convert a bunch of voice lines from a video game using rvc to another voice for myself but the amount of voice lines is a lot...is there any way to convert multiple files at once so I don't have to constantly upload files for it to convert, and instead just have it run overnight or something?
finite galleon
Hey so uh I was going to convert a bunch of voice lines from a video game using ...
You will need to know some python for that
ornate bobcat
Ok, guess im gonna hold off on this for now lol, i dont know python and im gonna have to look into og rvc and if it can do rmvpe or something but probably not
Thanks
rare gobletBOT
Ayo? @craggy heath level 1 !!!
old oracle
-colab
stone wyvern
hugging face spaces are giving me errors
im considering using rvc locally but idk if my 980 ti will handle it
stone wyvern
ok
dull whale
@molten pecan It's only a white screen
molten pecan
it didn't work
Tap the three buttons and go to Settings
Go to Restart this Space
And restart it
And just wait like before
Make sure your internet is stable and you stay on the page while it loads
dull whale
Make sure your internet is stable and you stay on the page while it loads
yeah it's still the same
molten pecan
yeah it's still the same
Try in another browser then
molten pecan
would firefox or edge work?
Yeah but I recommend Brave
dull whale
oh, why?
molten pecan
it has a built-in AdBlocker and is a little more private overall
You can turn it off whenever you want dw
rare gobletBOT
Ayo? @dull whale level 2 !!!
dull whale
I'll use Brave
dull whale
You can turn it off whenever you want dw
if I cna't use weight.gg links, how do I use them in your rvc?
molten pecan
if I cna't use weight.gg links, how do I use them in your rvc?
I included what to do about that in the guide
Check the troubleshooting section
dull whale
alright
molten pecan
if I cna't use weight.gg links, how do I use them in your rvc?
Also it's not my RVC, it's @proven hill's RVC >_<
The most stable RVC that I know :3
And it includes a TTS too!
dull whale
oh whoops!
proven hill
hello, its tha me!
dull whale
Wowie
brittle wing
How do I know when the wordband cover bot is finished
little owl
hi, the g/total graph has been increasing for a while and the d/total graph is still decreasing, should i continue training?
shy zealotBOT
**Method 1 🔎: ** The search accuracy is 100% 💕
Powered by ModelAI.net, Made with Love ❤️
azure marshBOT
RVC Guides (How to Make AI Cover)
honest timber
hello 
like i cant, fr im spending too much time trying to find something that work.I once tried run it locally but its not something for me (it didnt work) so id like to find any way to make it on cloud (hugging face, colab, ...)
i have alredy tried all colab models here, i can make ai cover,but i cant train a voice
pls help me
copper patio
Hi
Hello teacher, I used mango creppe while modeling and the creppe value was 128. I changed it to 56. For a better result, what should I do for my pitch value while doing an AI cover? It says 64 is 120. Would it be better if I set the pitch value to 56 while doing an AI cover?
finite galleon
Hello teacher, I used mango creppe while modeling and the creppe value was 128. ...
Just do rmvpe
copper patio
@finite galleon okey
frozen radish
I dunno if this is the right channel but,
rare gobletBOT
Ayo? @frozen radish level 1 !!!
frozen radish
I'm trying to download a link containing Leaf from Eeveelution Squad's voice model but it says "Invalid Username or Password". https://huggingface.co/UCringeRemade/LeafES/resolve/main/leaf.zip
This doesn't happen with other voice models, which work fine for some reason. What am I doing wrong?
finite galleon
I'm trying to download a link containing Leaf from Eeveelution Squad's voice mod...
If you click it it will say repository not found which means the person that uploaded it deleted the file
So its unsuable
frozen radish
So its unsuable
That's a shame, darn.
But, thank you very much for letting me know.
stable delta
guys how do i get rvc?
brittle wing
all of the sudden i get this error.. any help? Traceback (most recent call last):
File "/Users/kai/Desktop/Retrieval-based-Voice-Conversion-WebUI-main/.venv/lib/python3.10/site-packages/gradio/routes.py", line 437, in run_predict
output = await app.get_blocks().process_api(
File "/Users/kai/Desktop/Retrieval-based-Voice-Conversion-WebUI-main/.venv/lib/python3.10/site-packages/gradio/blocks.py", line 1349, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "/Users/kai/Desktop/Retrieval-based-Voice-Conversion-WebUI-main/.venv/lib/python3.10/site-packages/gradio/blocks.py", line 1283, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "/Users/kai/Desktop/Retrieval-based-Voice-Conversion-WebUI-main/.venv/lib/python3.10/site-packages/gradio/components.py", line 2586, in postprocess
file_path = self.audio_to_temp_file(
File "/Users/kai/Desktop/Retrieval-based-Voice-Conversion-WebUI-main/.venv/lib/python3.10/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'
2023-12-23 11:46:12 | INFO | httpx | HTTP Request: POST http://localhost:7865/api/predict "HTTP/1.1 500 Internal Server Error"
2023-12-23 11:46:12 | INFO | httpx | HTTP Request: POST http://localhost:7865/reset "HTTP/1.1 200 OK"
@sand iris any help?
brittle wing
all of the sudden i get this error.. any help? Traceback (most recent call last)...
i dont know....
gusty kestrelBOT
shy zealotBOT
**Method 1 🔎: ** The search accuracy is 100% 💕
Powered by ModelAI.net, Made with Love ❤️
elfin sable
Yo how to set Epoch?
distant wagon
Yo how to set Epoch?
I'm not sure, but I see it's marked on the model side.
elfin sable
Nuh uh
rare gobletBOT
Ayo? @elfin sable level 1 !!!
elfin sable
wth is that
distant wagon
magic
sand iris
It's the number of iteration the model went through attempting to learn from the dataset audio (the samples it's trying to reproduce)
distant wagon
It's the number of iteration the model went through attempting to learn from the...
can you help me
D:
elfin sable
It's the number of iteration the model went through attempting to learn from the...
how to set the epoch?
sand iris
can you help me
ain't available atm
distant wagon
sand iris
how to set the epoch?
you training a model ?
distant wagon
I'm Come Back
uhh can you help me?
astral jungle
uhh can you help me?
What's going on
karmic forge
nice clean and clear
astral jungle
Is it because of the model?
distant wagon
Is it because of the model?
I don't know, it was fine yesterday
Or should I create another Spaces?
astral jungle
I don't know, it was fine yesterday
Can you send me the rvc link?
distant wagon
Can you send me the rvc link?
uh how D:
rare gobletBOT
Ayo? @astral jungle level 13 !!!
distant wagon
Send me the link Ilaria rvc
o
azure marshBOT
RVC Guides (How to Make AI Cover)
distant wagon
I can not access
what now
distant wagon
is that ilariarvc's link??
yes
copper pebble
tf is NPCTW
copper pebble
https://huggingface.co/spaces/TheStinger/Ilaria_RVC
this is the link
distant wagon
i mean
distant wagon
this
copper pebble
is ur duplicated space
distant wagon
do ya want it to say just hello or smt else?
I want it back to normal qwq
copper pebble
I want it back to normal qwq
the GUI?
like this?
if ure talking about the model check the settings of the GUI
distant wagon
if ure talking about the model check the settings of the GUI
where?
rare gobletBOT
Ayo? @distant wagon level 7 !!!
copper pebble
where?
u didnt change anything?
just uploaded ur model?
did it always sound like this or is just this time?
distant wagon
did it always sound like this or is just this time?
from yesterday
copper pebble
thats bit weird
distant wagon
ok
karmic forge
https://huggingface.co/spaces/TheStinger/Ilaria_RVC
hi is this good for real time?
how to use it ^^
copper pebble
hi is this good for real time?
now thats just for covers
use okada for vc realtime
u can find a guide here #1159513888199540817
:3
karmic forge
okada i tried app but my spec is bad for local and tried colab colab keeps disconnecting
rare gobletBOT
Ayo? @karmic forge level 2 !!!
copper pebble
okada i tried app but my spec is bad for local and tried colab colab keeps disco...
mh
karmic forge
copper pebble
pretty bad situation
karmic forge
never mind
kind rune
40k vs 48k what wins?
vapid gust
40k
copper pebble
40k vs 48k what wins?
deez nuts
finite galleon
what is the best ultimite vocal remover model?
Depends on the use
For vocals and instruments i say MDX23C InstVoc
For back vocals i say karaokee 2
For harmonies i say BVE
For noise cleaning i say UVR deecho or denoise
brittle wing
like the model without weird artifacts
i have a 3080 gpu
my gpu does not support MDX23C InstVoc
finite galleon
my gpu does not support MDX23C InstVoc
Use mvsep ig
brittle wing
For vocals and instruments i say MDX23C InstVoc
For back vocals i say karaokee 2...
what is karokee 2? what model shoud i use?
finite galleon
**what is karokee 2? what model shoud i use?**
What are you even trying to do with uvr?
Seperate instruments from vocals? Seperate back vocals? Remove noises?
finite galleon
yeah someting like that
Karaokee 2 model
I think the specific name is like uvr karaokee 44000 or something like that
brittle wing
what is easy gui?????
what is this fille
@copper pebble** anyone help me with this file????**
storm hatch
<@1065579511183310939>** anyone help me with this file????**
I guess GUI means it has simple UI interface (like discord you see right now) rather than cmd line looking.
brittle wing
I guess GUI means it has simple UI interface (like discord you see right now) ra...
what does it even look like?
storm hatch
gui is like the interface you see right now (example discord) no GUI looks like cmd.exe
glad zealot
yes
brittle wing
also is 6 CPU processes good?
glad zealot
yup
brittle wing
i like it when the chat is more active
karmic forge
yo
i downloaded MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.17b.zip this version for amd and my GPU is not working yet in app and RVC is lagging as hell but beatrice is working well anyone can help!
brittle wing
how can i make my voice sound more natural when doing inference covers???
glad zealot
use the same gender most of the time when doing covers
ex. the model is a female voice, you need to use a female original audio to get the best result without changing too much settings
sand iris
i downloaded MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.17b.zip this version fo...
refer to https://rentry.co/VoiceChangerGuide
karmic forge
refer to https://rentry.co/VoiceChangerGuide
i checked ^^
sand iris
There's an extra step to have GPU acceleration on custom models with voice-changer on amd/intel cards
karmic forge
There's an extra step to have GPU acceleration on custom models with voice-chang...
okk
karmic forge
There's an extra step to have GPU acceleration on custom models with voice-chang...
thanks so much
low owl
Is there any way to mute the background vocals in this Colab? I've used this for a while and its great but sometimes it like....thinks main vocals are also background vocals so I sometimes get issues with the output being both the main vocals and the AI Cover. Its weird.
glad zealot
ye you could set the background to be very low but it doesnt mute it sadly
low owl
ye you could set the background to be very low but it doesnt mute it sadly
Is there any alternative Colab for RVC? If there isnt thats fine, its just odd that this has this problem.
glad zealot
not that i know of
lone sundial
yo guys so uh i wanna make a song. i got the instrumental but i need a bot too sing it so is there anyway to like give the bot the lyrics then it can sing
river egret
yo guys so uh i wanna make a song. i got the instrumental but i need a bot too s...
You can record yourself (or get a clip of someone else) singing it then apply the voice to it.
lone sundial
You can record yourself (or get a clip of someone else) singing it then apply th...
Oh thank you but is there anyway I can just give only the lyrics?
rare gobletBOT
Ayo? @lone sundial level 1 !!!
river egret
Oh thank you but is there anyway I can just give only the lyrics?
Not that I know of in RVC. My next best idea would be to have a bot speak it (Like Google Translate) but it will be a very flat delivery.
lone sundial
Not that I know of in RVC. My next best idea would be to have a bot speak it (Li...
Ok thanks a lot🥹
river egret
In the Latest RVC Web UI I can't select an inferencing voice, the dropdown has no entries.
I have a weights folder filled with .pth models which should be populating the dropdown.
I can't even supply my own path because if I do, with something like weights/CoolVoice.pth then the console says:
TypeError: can only concatenate str (not "NoneType") to str
The index paths are detected correctly and show up in their dropdown.
Are my models in the wrong path?
brittle wing
is there an ai that can replicate mp3 files to a track using the same compositions? i alrready heard of jukbox.....
azure marshBOT
RVC Guides (How to Make AI Cover)
fallow ether
Can someone help me making a rvc model?
sly widget
hey, I've been making AI covers for a while, but after Colab kinda stopped working I couldn't figure out how to use any other tool
is it worth to look for an online alternative for Colab, or will my PC setup do the job?
GTX1650 4GB a lil overclocked and a Ryzen 5 5600G
proper shale
hey, I've been making AI covers for a while, but after Colab kinda stopped worki...
Colab still works, it's just other Colab notebooks now, ones without the GUI
sly widget
darn, I'm trash at command lines
rare gobletBOT
Ayo? @sly widget level 1 !!!
proper shale
darn, I'm trash at command lines
nope, not command lines either
sly widget
then what...
proper shale
its just little cells that you run
sly widget
awh
sly widget
what setup does the free Colab plan have, tho? seems efficient, but I was actually wondering about using my lil box machine
proper shale
RVC Disconnected & EasyGUI are both good ones,m
proper shale
what setup does the free Colab plan have, tho? seems efficient, but I was actual...
Has a Tesla T4 with 15GB VRAM
And 12GBs of RAM
sly widget
alright i'd never afford that
proper shale
Has a Tesla T4 with 15GB VRAM
VRAM is quite important for AI training so that's why I'd recommend you use the T4 instead of ur own GPU
With 4 gigs of VRAM you could probably only get batch size 3, which would be extremely slow, and possibly not even best quality
sly widget
yeah well I'm not too rich lmao
my PC was 700 bucks on a discount and I didn't even pay for it, got it borrowed from my bestfriend who did
that's not relevant tho, thanks for help
proper shale
my PC was 700 bucks on a discount and I didn't even pay for it, got it borrowed ...
that's pretty nice value tho
proud elbow
even batch size 3 uses 6 GB vram
proper shale
good processor and the GPU allows u to play stuff
I'm paying $1600 in my future PC with an RTX 3060 and an i5 lol
proper shale
that's not relevant tho, thanks for help
You're welcome! Is there anything else I could help you with, btw?
sly widget
wish I had an RTX, but I'm not much of a gamer now since music production is my thing
proper shale
wish I had an RTX, but I'm not much of a gamer now since music production is my ...
Oh that's cool
sly widget
You're welcome! Is there anything else I could help you with, btw?
no, thank you, I think I know everything I should
as I mentioned earlier, I've already made some AI things, it's just the colab page that threw me off
proper shale
Oh wait, if you wanna make AI covers there's also
AICoverGen and Ilaria RVC
The ones I recommended earlier were for training models
sly widget
I wanna both train and convert, so that's cool
but I remember my issues with training models
it'd take ages to make a good one and I'd always find a thing to mess up
one model took me so many attempts that I just called my other friend who's into it and asked him to do it
turned out pretty neat
strong spindle
what goes into a dataset folder is it just the audio?
sly widget
I think so, I mean, what're you gonna be training it from? pictures?
strong spindle
no for an ai model
proper shale
what goes into a dataset folder is it just the audio?
yeah ur audio
strong spindle
dem im slow
sly widget
Ilaria doesn't support .flac files?
I was trying to convert a vocal in .FLAC, but it always returned an error
converting to WAV helped, but it's dumb
I mean, I know flac isn't used that much, but it's more lightweight than wav and still contains all the information
copper pebble
I was trying to convert a vocal in .FLAC, but it always returned an error
conver...
its not that big deal
sly widget
something must be wrong, bc a 3-minute vocal track takes over 400 seconds now... even though it initially showed 8.3s
rare gobletBOT
Ayo? @sly widget level 2 !!!
sly widget
and it usually took me a minute or two on colab
copper pebble
something must be wrong, bc a 3-minute vocal track takes over 400 seconds now......
nah thats normal
hf has a slower gpu
sly widget
how long should it take then?
crystal agate
hey does anyone know how to fix glitchy audio on discord? whenever i use the voice changer it always kinda breaks on discord and makes my voice glitchy
proper shale
something must be wrong, bc a 3-minute vocal track takes over 400 seconds now......
should be done soon, if it errors lmk, you may have not duplicated the space and all
proper shale
hey does anyone know how to fix glitchy audio on discord? whenever i use the voi...
could be the virtual audio cable you're using, try another one
sly widget
model download returned an error, but the model list still shows it and the processing goes well... just bc i got the tab refreshed like stupid, its late and im mindless
crystal agate
could be the virtual audio cable you're using, try another one
any reccomendation?
sharp island
Hey is it ok that clone resposutories doesn’t go through
proper shale
any reccomendation?
maybe VAC by Muzychenko would be good
sharp island
How is it not finding my dataset
rare gobletBOT
Ayo? @sharp island level 6 !!!
sharp island
Before you ask I added both “.zip” and “.ZIP”
brittle wing
anyone knows why the runtime keeps disconnecting after 1-2 minutes
sly widget
will it take less time with an mp3 file, or does it only depend on the length?
proven hill
Free vtuber girl voice for the voice changer as a christmas gift!
https://discord.com/channels/1159260121998827560/1188208519107850310
upper wedge
If i click on train model with mangio rvc web it processes but i dont get a model file
can somewone help me
barren sonnet
I'm planning on making a hl2 mod using the Half-Life 2 male 07 voice in the models repository, but who exactly do I credit, or do I need to credit at all? I'm not against adding a credit, but I'm not sure if I should credit the server, the individual user, or the user and the server! Or is this type of use forbidden?
proper shale
I'm planning on making a hl2 mod using the Half-Life 2 male 07 voice in the mode...
crediting both user and server would be best!
barren sonnet
That's what I was planning on doing, thanks for confirmation!
brittle wing
best esemble mode for uvr?
proper shale
**best esemble mode for uvr?**
some people said that ensembling is kinda meh, it's better to do everything individually
proper shale
That's what I was planning on doing, thanks for confirmation!
yw ;)
uneven scroll
bruh is breathy voice cloning just impossible with rvc
always sounds like metal scratching
bold valley
Is it possible to convert a three-minute cover in Ilaria_RVC? Please respond.
sand plinth
i just got rvc and i set the output and input but nothing plays and it says the volume is 0. i also tried it with a file and it didnt work
I got it to work, i switched to cpu then back to gpu
random cliff
hey, i'm sorry but the guides\stickies are abit confusing. if i want to convert a wav file to another voice do i need to use ilaria rvc and is it only hosted on hugging face? anywhere to dl it to run locally?
rare gobletBOT
Ayo? @random cliff level 1 !!!
quasi falcon
Hello, can anyone tell me how I can continue to train a completely trained voice model?
proper shale
hey, i'm sorry but the guides\stickies are abit confusing. if i want to convert ...
u can use that or other local forks like applio, mangio and the original rvc
-local
azure marshBOT
proper shale
Hello, can anyone tell me how I can continue to train a completely trained voice...
just change total training epochs and set all the same settings
it should work right away
random cliff
is there a specific one that can use the trained models people post on this discord?
proper shale
is there a specific one that can use the trained models people post on this disc...
all of em can
random cliff
oh all right, tyty
proper shale
yw :)
uneven scroll
where can I find more detailed info on how to tweak my output with covergen RVC?
I mean specifically i want to figure out the use of index rate, filter rate, filter radius, protect
also yes i did look in guides and no i didnt find info.
pearl wind
Hello Guys! Can anyone change this song voice into Atif Aslam (Pakistani Singer) voice?
https://drive.google.com/file/d/1Qxl4IFoZMvJGOwL7vDd4tmjx_rfonBIf/view?usp=drivesdk
molten pecan
Hello Guys! Can anyone change this song voice into Atif Aslam (Pakistani Singer)...
you can request for models or AI covers in #1159289738314919936
for some $$$
half cove
🤑
half cove
🤑
peak arch
Help, why when I am on the GUI it says error when I am training the ai and when I refresh the page it says " No interface is running right now".
spare shale
azure marshBOT
RVC Guides (How to Make AI Cover)
old oracle
-colab
full imp
friends, anybody knows how to kinda automate or make faster the process of splitting a long audio into 6 second chunks with FL STUDIO?
I saw a dude doing it in a ai cover tutorial LONG AGO and I just can't find it
molten pecan
friends, anybody knows how to kinda automate or make faster the process of split...
wait why would you do that?
for making your dataset?
full imp
for making your dataset?
yes, I read I have to make it into chunks o 6 seconds each
molten pecan
yes, I read I have to make it into chunks o 6 seconds each
uhm you don't have to do that with modern version of RVC
full imp
oh really?
molten pecan
yeah
full imp
I can just put my long ahh audio?
molten pecan
use a 30 minute-long audio if you want
full imp
damn that's epic, thanks a lot
molten pecan
I can just put my long ahh audio?
just remember that it doesn't have to be THAT long
usually 30 minutes is the maximum for a very good model
full imp
just remember that it doesn't have to be THAT long
no no, it's not that long, but I thought it still needed to be like that
gracias 😄
fluid topaz
i was on epoch 41, and got this error, how to avoid this?
CUDA warning: an illegal memory access was encountered (function destroyEvent)
molten pecan
i was on epoch 41, and got this error, how to avoid this?
CUDA warning: an ille...
@proper shale
that's not the antivirus acting up, right?
fluid topaz
well im trying to train in my own gpu, it only has 12gig of memory, looks like it ran out of memory, anyone know how to clear the cache every 10 epoch or something? im using mainline rvc
molten pecan
i was on epoch 41, and got this error, how to avoid this?
CUDA warning: an ille...
@true vortex any idea?
@civic pike
proper shale
well im trying to train in my own gpu, it only has 12gig of memory, looks like i...
did you enable the cache dataset option?
proper shale
hm that was a weird error. never saw that before
lapis glade
hello^^ when i attempt to downlaod a voice model. i get the popup "may not be trustworthy" and clicking "download anyway" results in the popup closing. but thats it. nothing downloads. am i the only one with such an issue? or am i just really smooth brain? cuz that also makes sense lol.
full imp
Set "save_frequency" to 25 or 50
25 or 50, like which one, I don't really understand, and the (Optional) To know the right amount of epochs use a Tensorboard.
Ok, and if I don't understand how to do that, how do I guess the number? What number do I write? Since it's, optional
brittle wing
I followed ai tools search's video to make my own trained voice but it always shows an error for some reason
Been trying and trying for ages and about to give up
Can anyone help? :c
proper shale
```Set "save_frequency" to 25 or 50```
25 or 50, like which one, I don't really ...
save frequency at 25 should be fine ig
and you dont really need to guess the number, just use 500 and go for that
proper shale
woah what gpu do you have?
full imp
save frequency at 25 should be fine ig
and you dont really need to guess the nu...
thanks a lot!
proper shale
0 NVIDIA GeForce RTX 3050 Laptop GPU
not bad, but maybe you shouldnt use a lot of cpu threads
well in that case it also affects gpu too
i recommend using 2
brittle wing
okay i'm gonna set it to 2 threads and see how it goes
fluid topaz
anyone have link how to use tensorboard with rvc logs?