don't mind that, it's normal
#✨│ai-help
sand iris
tacit lintel
hey guys is it just me or #Step 3. Start the GUI, then open the public URL. It's gonna look like this doesnt show any links to urls
finite galleon
Why the hell is it spamming it lol
Usually when i get that it just give me the min and max
tacit lintel
Also why does it infinitely loads when converting when it’s local downloaded as couldn’t get the cloud one to work
Mangia crepe ^
azure marshBOT
RVC Guides (How to Make AI Cover)
low shard
hey guys is it just me or #Step 3. Start the GUI, then open the public URL. It's...
what google colab u using
hidden nimbus
I want to use rvc again after using it once. Traceback (most recent call last):
File "/content/Ilaria-RVC/app.py", line 1, in <module>
import subprocess, torch, os, traceback, sys, warnings, shutil, numpy as np
ModuleNotFoundError: No module named 'torch' Is there a reason why this error occurs?
proven hill
I want to use rvc again after using it once. Traceback (most recent call last):
...
try going in runtime, stop and delete runtime and refresh the page
pallid cedar
There is only one slowdown in one step, to reach 180. Sometimes I need to restart Colab. But that's the only thing that causes slowdown
proven hill
wdym to reach 180?
sonic garden
-colab
proven hill
Thank you so much, its fixed
no problem!
pallid cedar
wdym to reach 180?
I believe it is a part where some items are downloaded in the first stage, one of the downloads
proven hill
I believe it is a part where some items are downloaded in the first stage, one o...
oh, thats strange
tight wind
After generating an index on Applio the start training button gets grayed out
leaden mountain
Hello! help please. I decided to create my voice model but I can't finish because the pth extension file is not created, what should I do?
brittle wing
no problem!
Will test Ilaria RVC to make sure it’s working for me today
brittle wing
what do yall think is a decent model for 15$?
rare gobletBOT
Ayo? @brittle wing level 1 !!! 
dusty bramble
I think it is, tho I think I learned how to fix it
I think my training data was maybe too loud or had a tad bit of reverb
Just had to remove that
brittle wing
Ilaria RVC colab works fine
Local URL dosen't work for me but public URL does with gradio
half cove
Local URL dosen't work for me but public URL does with gradio
If you mean local url on ila rvc ofc that won't work cuz it's not local 
brittle wing
If you mean local url on ila rvc ofc that won't work cuz it's not local <:skulls...
i meant these
brittle wing
OOOOOOHHH RIGHT
half cove
It's on your PC
half cove
Happens
leaden mountain
Hello! help please. I decided to create my voice model but I can't finish beca...
i try to search answer, but fail. help pls
fast shadow
I usually made ai covers from mobile but now it won't let me start the web gui?
What can I do?
low shard
I usually made ai covers from mobile but now it won't let me start the web gui?
What google colab are you using
glad zealot
i try to search answer, but fail. help pls
how are you making the model? what program are you using
rare gobletBOT
Ayo? @glad zealot level 65 !!!
glad zealot
ayo
leaden mountain
how are you making the model? what program are you using
Gratio
rare gobletBOT
Ayo? @leaden mountain level 1 !!!
leaden mountain
no like the name of the program, is it RVC webui,mangio,applio?
rvc, i can send screenshot
glad zealot
ya send it
leaden mountain
proper shale
pitch extraction didn't work
thats why
just make sure you got some space left there
leaden mountain
i have more than 150 gb free
glad zealot
why not just put the adios to the dataset folder instead of uploading it like that?
leaden mountain
why not just put the adios to the dataset folder instead of uploading it like th...
idk how do that
glad zealot
where gpu select
training preference maybe?
torn bluff
How long is it supposed to take for your GPU data to go back to normal? I heard 12 hours but it's been longer then that. And no, I don't want to create a new Google account.
proper shale
How long is it supposed to take for your GPU data to go back to normal? I heard ...
12 to 24 hours, mainly 24 hours
but i've heard from somebody it took 2 days?
just one person tho
torn bluff
12 to 24 hours, mainly 24 hours
It's mainly taken 12 hours for me. Even if I FINALLY train a model, What is the next step after when I finally get it to do vocal covers of other songs? Is that still a rvc thing? Or do i use some program or website to do that?
proper shale
It's mainly taken 12 hours for me. Even if I FINALLY train a model, What is the ...
yeah you gotta use an rvc fork (or a colab) to use the model
torn bluff
So per song it will be an additional 12 to 24 hours?
proper shale
So per song it will be an additional 12 to 24 hours?
nah
you can do a bunch of songs in like, 4 hours (that's roughly the colab limit daily)
torn bluff
Okay, that's really good to hear. phew.
I have tried rvc Disconnected V2 like 4 different times with multiple different issues or mistakes ive made so we are looking at 5 days now where I haven't trained my model yet. The furthest I got last night was with @spare shale . But dummy me used a corrupted wav file after the hour plus of help they gave me.
But this is with EasyGui which i got further then ever with
spare shale
???
torn bluff
So hopefully when my GPU comes back I can finally do it correctly this time and get to finally making vocal covers.
torn bluff
I hope my GPU comes back before like 1 am EST time for me. I just want to finally get this done and over with lol
Wait was it 1 am my time when i ran out of GPU? now i think about it it might have been around 10 pm
grizzled forum
is there a doc somewhere on how to train a model locally?
what do you reccomend to train a model? a colab or local app?
grizzled forum
proper shale
what do you reccomend to train a model? a colab or local app?
if you have a good gpu go for local
proper shale
works just fine for me...
https://docs.aihub.wtf/rvc/local/mainline/ read this especially
if you want local
grizzled forum
tysm
proper shale
yw!
humble ember
why is the pitch changing on my ai vocals if its at 0
azure marshBOT
RVC Guides (How to Make AI Cover)
torn bluff
@spare shale You around?
spare shale
aaa yep
torn bluff
Perfect. My GPU is back.
torn bluff
Im about to start over again. Im about to Create Training Files. which folder do I go to upload the actual WAV file?
spare shale
- Install RVC (in colab)
torn bluff
Oh i have to do that again. okay. will do
spare shale
sowwy
You must do this Every time you want to do this you need to Download RVC (colab)
torn bluff
No problem. Thanks for the heads up
spare shale
No problem!
torn bluff
I'll ping you once the install is done
spare shale
OK
torn bluff
@spare shale Done
rare gobletBOT
Ayo? @torn bluff level 8 !!!
outer gull
It's all over his model <@1076931317209505894>
ok
spare shale
Next you should Save your Wav Audio to the Dataset Folder
spare shale
- Write the name and press the button
torn bluff
Okay it's running
thin stump
<:nails:1159569314848972891> ok
honestly that should be a pinned post if people bother to look at them. They don't know how dirty it can be
spare shale
model name i mean
torn bluff
Success
spare shale
- Write Epochs (250 or 500) and Turn on OV2
torn bluff
Should I do 500 or will it be too much for my GPU time limit?
spare shale
Should I do 500 or will it be too much for my GPU time limit?
Okay.. Let's make it 250 😅
spare shale
Okay.. Let's make it 250 😅
And then again but 500 If you don’t mind
torn bluff
So 500 then?
spare shale
No first 250
torn bluff
I got the right WAV file this time btw lol i checked
okay 250 it is.
Hit the play button?
spare shale
Yes everything seems fine
spare shale
Hit the play button?
Yes
torn bluff
I'll ping you once the Epochs are done
outer gull
honestly that should be a pinned post if people bother to look at them. They don...
true i dont see much about noise checking info around
dusty bramble
I'll ping you once the Epochs are done
all fun and games until u get disconnect 
torn bluff
*all fun and games until u get disconnect* <a:sonicglare:1202692067755823104>
Don't jinx it. It worked fine last night.
dusty bramble
Don't jinx it. It worked fine last night.
My bad lol
Good luck tho o7
torn bluff
Thanks o7
craggy willow
Yo i got a question
if i stopped training a model so i can continue on another colab session, could i add more data to the dataset and continue training or do i have to retrain from the start if i want to add more data?
outer gull
have to retrain from the start cuz before start the training you did the feature extraction process which is generated from your initial dataset
spare shale
Im about 180 out of 250 epochs
I think no
torn bluff
50 more epochs to go
spare shale
now 200?
spare shale
Yes I'm waiting don't worry
torn bluff
250 epochs. Done.
spare shale
Do you want your Model to be 500 Epochs?
torn bluff
I feel if I do that now, i could be screwed for another 12 hours. I think i should just train at 250 for now and see how it goes and do another 250 tomorrow or something
spare shale
Ok then find all the Files:
#✨│ai-help message
And paste them in here:
#✨│ai-help message
torn bluff
What about .pth file?
brave nova
hello, where can I find google colab to make a model in gui
spare shale
What about .pth file?
Yes I mean pth and index
spare shale
hello, where can I find google colab to make a model in gui
-colab
torn bluff
@spare shale They are both under logs?
torn bluff
So where do i get pth?
spare shale
torn bluff
WEIGHTS thats it thanks
wheat plover
question: why should I set "target_sample_rate" to 40k?
should I be using 48k? is there a difference?
torn bluff
@spare shale Got it. check error too?
spare shale
yep
torn bluff
What about input?
spare shale
Also you can change the audio
spare shale
What about input?
spare shale
i chose someguy.mp3 again
No problem
torn bluff
Looks good?
spare shale
Yes
torn bluff
hit the button?
spare shale
(Don't forget to Download Files by Double Click)
spare shale
hit the button?
yes
torn bluff
Okay so it's done. The voice sample audio is just a dude talking like last time. Is that normal?
spare shale
Okay so it's done. The voice sample audio is just a dude talking like last time....
Send me what the AI did
spare shale
Check your wav 😅
torn bluff
yeah
yeah its the right one
To be honest it does sound a bit like Corey Taylor if he did a interview
torn bluff
It's got clean singing in the WAV i put together too
But sure ill do -10 pitch
oh no whats the default index rate
I accidently touched it
spare shale
?
ionic roost
Anyone else try to run viva la vida and get bad results at the chorus section?
torn bluff
This
ionic roost
I've been trying to figure out how to fix it and it's kinda getting on my nerves that none of my models can do it but other people have posted videos where their models did
torn bluff
i accidentally messed around with it now i don't know what the default rate is
torn bluff
I almost liked it the way it was because Corey does sound kinda like that when he talks
thin stump
Could be meant for singing only
torn bluff
Not this one the last one
So I can't train scream and growl vocals?
I had both singing, screaming/growling in the wav
I know it can be done because I have seen YouTubers upload their own Slipknot (Corey Taylor vocals) covers
thin stump
So I can't train scream and growl vocals?
That's why people make different variations of models between singing only and talking. You can share how he sounds if he sings
winged plover
Is it possible to do an inference on the processor?
torn bluff
That's why people make different variations of models between singing only and t...
So what you are saying is I have to make a wav of just singing and a wav of him just screaming?
spare shale
So what you are saying is I have to make a wav of just singing and a wav of him ...
xD
thin stump
help
spare shale
@torn bluff?
spare shale
I don’t want to be clever, but such a model exists
torn bluff
But i need specific scream/growl vocals from The End So Far. His vocal style changes per album
I can't find any from that album
spare shale
I don’t want to be clever, but such a model exists
Forget my words
torn bluff
I don’t want to be clever, but such a model exists
Yes for specific album cycles. None for the style from the album that I want. His vocals change per album
spare shale
Yes for specific album cycles. None for the style from the album that I want. Hi...
Yes you repeated it 2 times
torn bluff
So because he talked in that 3 second clip. Am I screwed?
Is it not possible it will still work with screaming vocals and singing from randomguy.mp3?
spare shale
So because he talked in that 3 second clip. Am I screwed?
I think the problem is that it's a small post. (i mean wav)
try:
torn bluff
Small WAV? It's almost 6 minutes though. How long does the WAV need to be?
okay ill try fnaf.mp3
rare gobletBOT
Ayo? @torn bluff level 9 !!!
thin stump
So because he talked in that 3 second clip. Am I screwed?
Ok, just going to clarify really quick. If you want a model for speaking only (like his interviews), you don't include screams or shouting. But since you're making a singing model, screams are fine
torn bluff
No. I WANT SCREAMING/GROWLING
thin stump
yes and growls
torn bluff
Well too late, im disconnected
torn bluff
Another wasted day
torn bluff
It's not your fault
im just pissed, ill have to wait ANOTHER 12 hours.. (not getting a second google account). It's like ill never be able to train my AI voice
spare shale
im just pissed, ill have to wait ANOTHER 12 hours.. (not getting a second google...
I can try to make you a Voice Model
torn bluff
It literally doesn't make sense that the example clip is Corey talking when the WAV is FULL of screams/growls and cleans
torn bluff
[Hug](https://media.discordapp.net/stickers/1213202219508174848.png?size=160&nam...
You would be my hero if you did
spare shale
You would be my hero if you did
It's not difficult for me
The main thing is to send wav
spare shale
Okay would the WAV file i made suffice? Is 5:45 long enough?
I think so, but if it’s not difficult for you, you can send even more about the Voice
torn bluff
Okay how long do you want?
I can throw more together in Audacity and remove the reverb
spare shale
Okay how long do you want?
No problem how long it will last Just send a video or Recording Where he Speak
torn bluff
But I don't want speaking. I just want his screams etc
spare shale
I can throw more together in Audacity and remove the reverb
I can do this too 😅
spare shale
But I don't want speaking. I just want his screams etc
Hm..
Then I'm not sure that it will be a good model at all
torn bluff
Screams/growls and clean singing. Im trying to use the vocals off of that album to do a cover of the first albm
Why wouldnt it be?
thin stump
Oh that makes sense. it's a niche
spare shale
Screams/growls and clean singing. Im trying to use the vocals off of that album ...
Can you send the vocals to me in dm
torn bluff
Yeah ill send you my 5:45 WAV file in dms
spare shale
Yeah ill send you my 5:45 WAV file in dms
No problem
Or leave it here, it doesn't matter
spare shale
I don't care 😅
spare shale
Can you send me the original? (YT)
torn bluff
It's a collection of samples from 3 songs. Do you want me to send you "The End So Far" album playlist?
spare shale
AI Doesn't Understand This Scream And Thinks It's Noise (That's My Thought)
torn bluff
i chose these samples hand picked because he also has certain voice infictions that i want there too. like a whine at the end of his vocals etc
spare shale
- back vocal
torn bluff
That is what doesn't make sense to me because there is PLENTY of AI Slipknot (Corey Taylor) covers on YouTube with plenty of screams in it
Ill send you "The End So Far" playlist (the specific album i need vocals off of) and ill send a example of a AI cover already done on YouTube from Slipknot
https://www.youtube.com/watch?v=p8zamUoteG4&list=PLvXUqfNLb-tmW0ZXdHqSzWQlWi_ilGZ-_ Don't worry about the first and last song.
Here is a example of AI cover done from Slipknot with plenty of screams in it done well https://www.youtube.com/watch?v=VcyShyeFaOk
spare shale
For example:
I'm listening to this https://www.youtube.com/watch?v=VcyShyeFaOk and I THINK there were only Screams without a Simple Voice or Change in Tone
spare shale
Here is a example of AI cover done from Slipknot with plenty of screams in it do...
We have 1 Brain xD
torn bluff
I don't understand the logic to be honest with you. How can you train a specific album's scream vocals with only a simple voice to make a cover with a different albums vocals on a song from a different album?
Psychosocial as you shown is from a album called "All Hope Is Gone". The IOWA voice is from... well the album "IOWA"
thin stump
This was the from the old ai hub server before it was deleted. It's a single model https://weights.gg/models/clm73351q1hzscctcfhgnvcia
spare shale
@~@?!
torn bluff
@thin stump That example you presented is vocals from other albums that are not from "The End So Far". That's why I needed to train my own
spare shale
This was the from the old ai hub server before it was deleted. It's a single mod...
He means that in every song his tone and cry change
thin stump
gotcha, just trying to show his workflow
spare shale
-colab
torn bluff
Literally every Slipknot album Corey's vocals change, that's all im pointing out.
@spare shale Are you able to help me achieve this?
spare shale
<@892429048771399680> Are you able to help me achieve this?
I'll try but I'm sorry I don't Promise you (because if you read my nickname I'm Stupid + Lazy)
torn bluff
I'll try but I'm sorry I don't Promise you (because if you read my nickname I'm ...
Fair enough. I don't expect anything. The fact you are trying is awesome of you.
spare shale
???
torn bluff
Yeah?
spare shale
It's good that today is Friday I can't sleep the whole night xD
torn bluff
Oh yeah you are hours ahead of me, my guess is either England or Germany
indigo garden
ilaria rcv on huggin face gives me an error every time after 9 seconds, got fixes?
torn bluff
You put ??? thats why i replies with "yeah"
spare shale
Oh yeah you are hours ahead of me, my guess is either England or Germany
Czech
torn bluff
That would explain it. Must be early morning there. Sorry
spare shale
That would explain it. Must be early morning there. Sorry
Everything is fine! It's 9:16PM / 21:16
indigo garden
ilaria rcv on huggin face gives me an error every time after 9 seconds, got fixe...
can you see the logs when you get an error?
torn bluff
So yeah if you could use the WAV file because it's got multiple vocal style examples or try to work with the album playlist i sent. I need 2 different types of clean singing, multiple examples of screaming/growling and some cries at the end of vocals, some of his deep voice parts too and whispering. Can you handle that @spare shale ? lol
brittle wing
how do i setup the voice ai thing to use for discord?
spare shale
can you see the logs when you get an error?
You need to Post a screenshot so that other Helpers can help
indigo garden
You need to Post a screenshot so that other Helpers can help
screenshot of what? i cant find the logs lol
spare shale
So yeah if you could use the WAV file because it's got multiple vocal style exam...
A lot of text
spare shale
screenshot of what? i cant find the logs lol
Screenshot Errors
torn bluff
@spare shale to make it easier, use my WAV file at least and add anymore you find in the album
fathom reef
Real quick, what's the best AI voice training application that I could use?
indigo garden
Screenshot Errors
it's just this, i dont think i'll give much information
spare shale
it's just this, i dont think i'll give much information
If you made spaces
If you haven't done it then do it now
fathom reef
https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw
Really? Then why can't I download my voice model after training both the model and index?
torn bluff
Ping me when you fail or succeed. Thanks a lot for helping @spare shale
spare shale
<@892429048771399680> to make it easier, use my WAV file at least and add anymor...
I'll try but it's better not to believe me
indigo garden
If you made spaces
If you haven't done it then do it now
sorry whats spaces?
yeah i cant find logs
spare shale
Really? Then why can't I download my voice model after training both the model a...
Go to colab and find the Folder Icon
fathom reef
Go to colab and find the Folder Icon
And where would I find the files?
spare shale
sorry whats spaces?
indigo garden
yeah thanks i was doing that
spare shale
Ping me when you fail or succeed. Thanks a lot for helping <@892429048771399680>
Forgive me in advance If I can't do this
spare shale
The first one?
open Folder RVC
rare gobletBOT
Ayo? @brittle wing level 1 !!!
spare shale
Works
fathom reef
open Folder RVC
That doesn't tell me anything.
spare shale
That doesn't tell me anything.
Unhide the colab
fathom reef
The... What??
spare shale
i mean send screenshot*
fathom reef
pth
I'm talking about the files in the third screenshot. Which file do I download? I assume the first one.
spare shale
Send a full screenshot to Kodab
spare shale
I'm talking about the files in the third screenshot. Which file do I download? I...
Find where only your name and pth are written
fathom reef
So it is the first one.
spare shale
So it *is* the first one.
Yes
fathom reef
Gotcha. Thanks.
spare shale
Stop spamming
spare shale
Gotcha. Thanks.
Now you can save it to Google Drive or HF
easyGUI Works
Because now @fathom reef Uses easyGUI
- Now I will use it too
Everything works Next you need to download the files
torn bluff
Forgive me in advance If I can't do this
Just try please
spare shale
Just try please
Yes Good no Problem
torn bluff
I have a question. The colab you sent me. Does it only work with voices and not singing vocals?
I know it can be done because as I've said it's been done with various screaming bands. I'm wondering if there is a specific colab that can train those types of vocals?
spare shale
I have a question. The colab you sent me. Does it only work with voices and not ...
Voice and singing
torn bluff
Yeah then im confused lol
spare shale
why
the audio has 4 mins
torn bluff
maybe screams easyGUI can't handle. It might have to be another colab
spare shale
mb
spare shale
Is it HF or colab?
quartz aurora
hf
spare shale
quartz aurora
right after i imported the audio look
oh thanks
should i use gradio or cloudflared
spare shale
gradio
quartz aurora
ok
quartz aurora
but why isnt the HF version working?
spare shale
HF Works but very long Files will take a long time to load (Example: I waited 120 Seconds for the Sound to change voice from 60 Seconds)
quartz aurora
i am clicking refresh on the audio and it doesnt do nothing whyyy?😭
spare shale
click
spare shale
Click on Select audit file
quartz aurora
does anyone have a collab version of voice training that doesnt dissconect you?
please i need it very much
@spare shale do you?
spare shale
Google Colab dissconect all users
Only Locally can work well
But you need a good PC
quartz aurora
yes but i heard there is a version where you have to do chaptchas,but wont dissconect you
i would try local but i dont have nvidia,i am on amd
spare shale
i would try local but i dont have nvidia,i am on amd
amd Seems to work too, but slow
proven hill
ilaria rcv on huggin face gives me an error every time after 9 seconds, got fixe...
use colab
quartz aurora
which one is for amd?
spare shale
for amd?
Sorry, but I don’t have such Information (Perhaps I’m confusing something because I don’t understand Local)
What how?
spare shale
Then I'm sorry, I guess I'm confusing this with SD
quartz aurora
so isnt there anything that i can do so the collab will use chaptchas but wont dissconect me?
spare shale
so isnt there anything that i can do so the collab will use chaptchas but wont d...
Sorry I don't know that
quartz aurora
ok,then what is the latest training collab?
spare shale
-colab
quartz aurora
1. Model training -> https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8P...
yes model training
oh ok,thank you
spare shale
quartz aurora
so when you train voices,you do it on collab?
spare shale
I have enough of this for 24 hours
spare shale
so when you train voices,you do it on collab?
Yes
And the creation of Covers is also in colab*
quartz aurora
and how much do you run it,to make it sure you dont dissconect
spare shale
Sorry mistake
spare shale
and how much do you run it,to make it sure you dont dissconect
I just don't leave my laptop 😅
spare shale
yes,but i tried to run it for 4 hours and it simply dissconected
Hmm.. I don't know why..
Everything works fine for me
quartz aurora
and do i always have to keep the window open?
spare shale
For example: To create a Voice Model it is 5-7 Hours and no Shutdown
quartz aurora
or can i just leave it in the background?
spare shale
and do i always have to keep the window open?
I think yes?
quartz aurora
ohh,so i cant play while it is loading
spare shale
Close the mute folder and open it with the name of your Model
spare shale
ohh,so i cant play while it is loading
Yes
I'm just watching anime xD
I have Anime open on Colab 2 Screen
quartz aurora
like split screen?
spare shale
like split screen?
Yes
quartz aurora
ohh
spare shale
Or I can just connect 2 Screen
quartz aurora
ok,thank you
spare shale
Or I connect to the TV (In the picture I show picture in picture)
rare gobletBOT
Ayo? @brittle wing level 2 !!!
hot lagoon
Just a general question, but what epoch do you guys usually start for training then keep going going if it’s not overtrain
spare shale
Click twice
spare shale
Just a general question, but what epoch do you guys usually start for training t...
wdym?
hot lagoon
I’m just curious, that’s all
spare shale
Also Just press 2 times
Find the pth file and double click
hot lagoon
I think it's possible to retrain but I don't know how
I think retrain is just you putting in the same setting and updating the epoch if I’m not mistaken
spare shale
I think retrain is just you putting in the same setting and updating the epoch i...
I think yes
But I'm sorry, I really don't know
It’s better to ask such Questions to Professionals
hot lagoon
By professionals, you mean model, masters
spare shale
Yes or Other staff because I'm lazy and stupid
=-=
Tell me it's a joke
Please read My Message above
hot lagoon
Any updates from the career of UVR 5 on updates
round nimbus
i have a problem if i go in discord vc the half of the time they dont hear me and the other half they hear me wierd
hot lagoon
btw I don't know how to use it 😅
You’re kidding you’re a helper and you don’t know about tenser board that’s kind of cringe 
hot lagoon
Must be really easy than to get into helper
hot lagoon
No worries I was just testing your knowledge
spare shale
No worries I was just testing your knowledge
I'm Very Dumb
tacit lintel
what google colab u using
Was mangia
half cove
Must be really easy than to get into helper <:nails:1159569314848972891>
depends if you help, are trusted etc
poyo was accepted due to recommendation iirc
also depends who sees your application tbh
low shard
Was mangia
locally u mean ?
tacit lintel
yes
rare gobletBOT
Ayo? @tacit lintel level 1 !!!
outer stone
Anyone know why the download model function isn't working for this notebook anymore?
https://colab.research.google.com/drive/1TU-kkQWVf-PLO_hSa2QCMZS1XF5xVHqs#scrollTo=F6zbsuNs6xZF
half cove
Anyone know why the download model function isn't working for this notebook anym...
old colab
rare gobletBOT
Ayo? @outer stone level 1 !!!
half cove
-colab
outer stone
Anyone know why the download model function isn't working for this notebook anym...
Is there a colab with the mangio rvc fork? I.e RVCv2 training.
Looking to use the same interface as this one of possible
brittle wing
Is Applio 3.0.8 buggy for you guys? I either get moments of errors or crashes my pc?
Whenever I start training
azure marshBOT
RVC Guides (How to Make AI Cover)
brittle wing
I got an ffmpeg.exe Application Error, whenever I clicked on "Extract Features".
viscid moss
I got an ffmpeg.exe Application Error, whenever I clicked on "Extract Features".
@brittle wing
brittle wing
one moment
brittle wing
I tried again, and it didn't pop up again? That was weird, if it happens again or any other errors. Ill screen shot it for you next time.
spare shale
Virtual cabel -> use plz: #🔍│help-w-okada
brittle wing
During training just now, it just restarts/crashes my PC and I gotta do the training all over again?
Maybe Im using way to much GPU?
I have RTX 2080 Ti, 11 GB VRAM it put 11 batch size as default for me.
proud elbow
I have RTX 2080 Ti, 11 GB VRAM it put 11 batch size as default for me.
use batch size 8 (or somewhat 4 for short dataset)
the cpu threads should be set just 2 or 4
brittle wing
the cpu threads should be set just 2 or 4
I don't see CPU threads option in Applio training?
I have Ryzen 9 5900x 12 Core
proud elbow
I have Ryzen 9 5900x 12 Core
did u just manage to do feature extract?
brittle wing
did u just manage to do feature extract?
Yes, it went through great again.
proud elbow
I mean if he has tried it again
forest sluice
Must be really easy than to get into helper <:nails:1159569314848972891>
I think bro's more of an SD guy
forest sluice
Because it automatically binds to max CPUs
latent maple
anyone know why it does this
@forest sluice
latent maple
@half cove
half cove
I'm not a w-okada helper
Try asking in #🔍│help-w-okada
And try not to ping random staff members pls
forest sluice
Try asking in <#1159290161683767298>
He already asked there btw
latent maple
the error im getting is
half cove
He already asked there btw
Then be patient 🗿
latent maple
[Voice Changer] Web Server Launch Exception, Directory '/private/var/folders/fn/r20d45911tj453lpkflw2y0h0000gn/T/AppTranslocation/6EAD6A15-8EDB-4AD4-A1E3-5A7CE3A406B1/d/model_dir' does not exist
forest sluice
You relaunched the app right?
Still, I've never used a Mac in my life, so I can't provide much help
latent maple
You relaunched the app right?
yea
latent maple
Still, I've never used a Mac in my life, so I can't provide much help
whos on rn that could likely help
hot lagoon
I think bro's more of an SD guy
Sd guy?
carmine ether
downloadlink to AI Voice software?
shut wave
where do i downlaod the ai hub?
rare gobletBOT
Ayo? @shut wave level 1 !!!
shut wave
onde baixa o ai hub?
carmine ether
which AI hub
forest sluice
downloadlink to AI Voice software?
This is for AI Cover stuffs, if that's what you want
azure marshBOT
RVC Guides (How to Make AI Cover)
brittle wing
@crimson yew
compact spire
@rigid mango hey was wondering how to upload your voice, since when i try to upload it there is no index file and when i try to upload it with only the pth file it says 100% but it doesnt show on my rvc list
solar viper
I want to ask that I downloaded one of the RVC model here and put it into voice.ai app but it always been unstable and lagged. Is there anyway to fix it? (Sorry if this isn't supposed to be asked here 
)
sand iris
I want to ask that I downloaded one of the RVC model here and put it into voice....
We don't use voice.ai here
It's a paid software offering a rather bad service, when there are free options around
low shard
yea
solar viper
We don't use voice.ai here <:nails:1159569314848972891>
Ah, sorry then.
sand iris
Not only free as in price, but also as in freedom (the code is available to be modified however we want lol)
sand iris
The UI isn't quite as nice as voice.ai, but it functions the exact same. Voice in -> modified voice out
solar viper
I see. I will try. 
rare gobletBOT
Ayo? @solar viper level 1 !!!
solar viper
Lmao, that scared the s out of me.
urban turret
What's the best models for Ultimate Vocal Remover?
shut stratus
help
shut stratus
let me know so i can upgrade
sand iris
yeah, unfortunately it's too old to work on recent CUDA versions, which is what the voice changer uses to compute on GPU
shut stratus
yeah, unfortunately it's too old to work on recent CUDA versions, which is what ...
okay ill try on another device thanks for the help
rare gobletBOT
Ayo? @shut stratus level 2 !!!
sand iris
<@310840858767523846>
You would need to have a lot of audio from you speaking
without background noises
little lily
i do
rare gobletBOT
Ayo? @little lily level 1 !!!
sand iris
this should cover most of it
I don't do much model training, I'm just here to help with install issues
azure marshBOT
RVC Guides (How to Make AI Cover)
quartz aurora
hello, why does this appear?
sand iris
hello, why does this appear?
the path isn't the correct one
little lily
how about mine
sand iris
how do i fix this
¯_(ツ)_/¯
sand iris
If it keeps saying 0% it's currently working
quartz aurora
the path isn't the correct one
how to fix
sand iris
weird that tensorboard shows nothing... maybe try the refresh button ?
sand iris
how to fix
find the correct folder path, I'm not sure what it is
but 1) you gave it an audio file
and 2) it definitely isn't under /audio directly, there are folders before it
sand iris
how long does it take
idk
little lily
it said 0% for like 10 min
sand iris
yeah the percentage display isn't wired up properly
but as long as it keeps showing new lines, and tensorboard updates, it's fine
rare gobletBOT
Ayo? @sand iris level 81 !!!
little lily
ok
quartz aurora
so i am training on easy gui,and all the epochs arent saving on gd,if i wanna train for 5 hours without getting dissconected due to gpu runtime what can i do?
@sand iris
from my experiance,after 2 hours i get dissconected
little lily
my model wont show up on the easy gui when i downloaded the model intp it
rare gobletBOT
Ayo? @little lily level 2 !!!
little lily
@sand iris
raven burrow
-rvc
strange jackal
any of you guys know about how to use advanced mangio rvc fork to train voice model? im having some issues
polar cobalt
uhm i mean how many times do i need to wait for the ai acapella generate, i generate from acapella 3:17s?
dense marsh
I used a covergen thing and it says the song saved to the output /content/aicovers etc and idk how to access that on mobile
low shard
I used a covergen thing and it says the song saved to the output /content/aicove...
You are using the google colab right
tawdry current
"Running with the runtime python
cannot find the specified path
press any key to continue . . ."
can someone help
pressed a key and the cmd dissapears
vapid trellis
Hello everyone, has anyone encountered a problem in the Voice Changer client that when you say the word "Trail" is played?
onyx wraith
yo wassup i got a google colab+ subscription and wondering whats the best hardware accelaration option to use with rvc when making voice models? i can choose between A100 GPU, V100 GPU, T4 GPU
half cove
my model wont show up on the easy gui when i downloaded the model intp it
what colab
half cove
any of you guys know about how to use advanced mangio rvc fork to train voice mo...
elaborate
half cove
"Running with the runtime python
cannot find the specified path
press any key t...
how did you download it and what did you use to unzip it
half cove
yo wassup i got a google colab+ subscription and wondering whats the best hardwa...
a100 would be fastest
cinder grotto
Hello Good I have a problem I installed the MMVCServerSIO application and when I want to test an artificial intelligence voice I do not hear the voice I do not know if I'm doing it wrong or I do not know what to do
I've been trying for a while and I can't get it to work
verbal basin
brittle wing
how do u fix the echo
indigo garden
how do u fix the echo
use uvr with deecho
brittle wing
idk what that is
The application to search for different AI models does not work for me
indigo garden
idk what that is
ultimate vocal remover, it is a programmù
tame mica
how do you fix these?
i usually use a de-esser for these iirc
bold canyon
is there a app on mobile/desktop for using RVC?
daring raft
hey guys, i am used to do this on paperspace, so i am new to colab thing
When I want to save model. Do i need to stop training or i can just run cells for saving model while training
heady mirage
is that normal with the error
rare gobletBOT
Ayo? @heady mirage level 2 !!!
spare shale
@torn bluff Ask me, but I have the same OPTION as you (Once again, I’m sorry that I couldn’t Help you)
rare gobletBOT
Ayo? @spare shale level 39 !!!
torn bluff
<@572552916116701217> Ask me, but I have the same OPTION as you (Once again, I’m...
That's okay. I'm sure I'll find a way to do this eventually. If others on YouTube done it, I know it's possible. Just got to find out how.
Thank you for trying for me though. Much appreciated.
spare shale
That's okay. I'm sure I'll find a way to do this eventually. If others on YouTub...
You can try to take all the Songs from the Album and Remove Music and Send to AI (Remove Noises and Back Vocals)
azure marshBOT
RVC Guides (How to Make AI Cover)
fossil pasture
Is this an overfitting of the model?
tropic trail
-colab
fickle latch
heady mirage
is that normal with the error
can someone help me?
woeful nexus
I think the tutorials got taken down from google docs so idk what to do, I don't know how to look for the right settings or anything. I recorded my own voice to convert into a character's and I'm running it locally RVC1006Nivida(I don't think the models have a proper link for google colab, maybe I'm just not very smart) . Everything's sounding demonic and I don't know why. What tutorials/other options can I get into?
proper shale
I think the tutorials got taken down from google docs so idk what to do, I don't...
just see mainline part
woeful nexus
oh okay, thank you! I really appreciate all the help, even if my questions might seem a bit silly. It's still so new to me haha
proper shale
oh okay, thank you! I really appreciate all the help, even if my questions might...
it's alright :)
old oracle
-colab
green moon
I'm trying to use huggingface RVC V2 to create an AI cover and I entered my voice model and song in succesfully but it keeps processing forever whenever I hit convert does anyone know a way around that?
im using a macbook air do i need specific specs for this to work its my first time
spare shale
@green moon try: https://colab.research.google.com/drive/1u1brjK8IZt647UsbZuGYfW29oFM2I4tk?usp=sharing#scrollTo=9aMm_NgjUH4T
green moon
<@805930251120672770> try: https://colab.research.google.com/drive/1u1brjK8IZt6...
thank you i'll give it a try
iron cobalt
How long does model training take using Applio?
I'm trying to make my own AI voice model but no success. I'm more focused on using local RVC model making forks rather than using Google colab
I hate using Google Colab bcs I can't use GPU so long
old oracle
-colab
iron cobalt
-local
azure marshBOT
iron cobalt
-train
rare gobletBOT
Ayo? @iron cobalt level 4 !!!
neat olive
When i import a model into the collab it says "JSONDecodeError: Expecting value: line 1 column 1 (char 0)" no matter what model i use! I don't know what to do.
charred citrus
can someone tell me if my voice changer is working or not
tender loom
how can i make my voice smoother it keeps stuttering and lowkey sounds robotic
low shard
send link
that is outdated
it doesnt work anymore
check our docs https://docs.aihub.wtf
fast shadow
Ohhh Thanks!
rare gobletBOT
Ayo? @fast shadow level 2 !!!
brittle wing
For anyone that uses Applio: When I try to voice blend models that i myself haven't trained, it errors out. is there a way to fix this?
fast shadow
I'll try this!
low shard
yw
brittle wing
For anyone that uses Applio: When I try to voice blend models that i myself have...
Ohhhh i think i fixed it: it only works if the models are made using the same Pretrain
mellow flax
-help
fallow sparrow
someone please tell me how to make an ai song cover, i am making a lowtiergod cover of stronger than you and almost every single website costs money, someone please tell me the most easiest way of making it come true
rare gobletBOT
Ayo? @fallow sparrow level 1 !!!
fallow sparrow
sorry if i kinda sound dramatic, sorry
thin stump
sorry if i kinda sound dramatic, sorry
how good is your pc? Could recommend something super simple
and lightweight
fallow sparrow
i am using a chromebook, should have mentioned that, sorry
thin stump
i am using a chromebook, should have mentioned that, sorry
shame, try this though https://docs.aihub.wtf/rvc/cloud/ilaria-rvc/
fallow sparrow
shame, try this though https://docs.aihub.wtf/rvc/cloud/ilaria-rvc/
i'll try
charred nimbus
Hello everyone. I'm newbie and try to use Mangio. But I got something problem. When I press Train Model. the epoch is not start.
It said File "genericpath.py", line 50, in getsize
FileNotFoundError: [WinError 3] The system cannot find the path specified: 'C:\Users\ARRMNOK\Desktop\Mangio-RVC-v23.7.0_INFER_TRAIN\Mangio-RVC-v23.7.0/logs/mute/0_gt_wavs/mute48k.wav'
and Output information in browser is Chinese language
I tried to find answer in google but I can't find it. Thank you very much for helping.
lament eagle
is there like a way to seperate 2 people talking
viscid pagoda
I got my .pth file, do I need anything else from the training folder thing?
do I only need the .pth?
rare gobletBOT
Ayo? @viscid pagoda level 1 !!!
azure marshBOT
RVC Guides (How to Make AI Cover)
finite galleon
is there like a way to seperate 2 people talking
That is overlapped or like on a podcast?
lament eagle
Yeah
iron cobalt
-local
azure marshBOT
old oracle
-colab
spare shale
Why this problem while downloading model
Use the new Version
This is an old version, use the new one
Train: https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw
AI COvers: https://colab.research.google.com/drive/1u1brjK8IZt647UsbZuGYfW29oFM2I4tk?usp=sharing#scrollTo=9aMm_NgjUH4T
dense drift
This is an old version, use the new one
Train: https://colab.research.google.com...
how to use the train collab?
spare shale
Please Read my Message written 3-4 Days ago I Completely Wrote How and What
If you have any other questions about training the Model, write here
verbal oasis
@spare shale
https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw
Is this the colab ink from where to start the process?
spare shale
This is an old version, use the new one
Train: https://colab.research.google.com...
YOYO
spare shale
<@892429048771399680>
https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK...
Yes
spare shale
i cannot search for a model
muted rampart
can someone send me a link for a rick sanchez and morty smith model?
spare shale
can someone send me a link for a rick sanchez and morty smith model?
You can find Voice Models Here: https://www.weights.gg/
or here: #🔍│find-models / #1175430844685484042
muted rampart
yeah i tried searching them in voice models chanel but nothing comes up, i think somethings up with my computer becouse i could do it a few hours ago
rare gobletBOT
Ayo? @muted rampart level 1 !!!
spare shale
Is there a optimal video length for a training process?
optimal video?
muted rampart
can someone help me bruh i dont wanna give wheights acsess to my info and my mail tf
muted rampart
[smile](https://media.discordapp.net/stickers/1179133143463235645.png?size=160&n...
what is that supposed to mean
brittle wing
how open a rvc on applio? I dont make it
spare shale
muted rampart
rick sanchez: https://www.weights.gg/ru/models/cltco1njr0349z1db9tnw8z44
2: http...
i found it on wheights but to get a link for a collab i need to authorize wheights to get my email and my info
i dont need a google link bro i need a link for the collab
why does this have to be so complicated
muted rampart
generation cover?
yeah
muted rampart
#1177806958120685660 message i tried putting this link into the collab and it does not work
rare gobletBOT
Ayo? @muted rampart level 2 !!!
muted rampart
and now i dont know what to do
spare shale
send the screenshot
muted rampart
dense drift
why is this error coming?
spare shale
Have you completed 1 Button?
muted rampart
Have you completed 1 Button?
yeah
spare shale
why is this error coming?
spare shale
yea but why?
find the folder icon and try to create it yourself
brittle wing
https://docs.aihub.wtf/rvc/local/applio/
why unzip 55 minutes
dense drift
find the folder icon and try to create it yourself
how?
spare shale
why unzip 55 minutes
Because it will download to your PC. This is no longer my problem, but your PC’s.
or what do you mean?
brittle wing
Because it will download to your PC. This is no longer my problem, but your PC’s...
My pc have m2 holder
rare gobletBOT
Ayo? @brittle wing level 1 !!!
spare shale
=-=
spare shale
Use Another colab
muted rampart
Use Another colab
i only know about this one
spare shale
Sorry @brittle wing But I don't know how it works on PC I only use colab
spare shale
https://colab.research.google.com/drive/1u1brjK8IZt647UsbZuGYfW29oFM2I4tk?usp=sh...
!!!!!!!!!!!!!!!!!!!!
brittle wing
Sorry <@456226577798135808> But I don't know how it works on PC I only use colab
Oh. ok
muted rampart
!!!!!!!!!!!!!!!!!!!!
thanks imma try and if i cant get it to work imma come again
dense drift
dense drift
Send a full Screenshot
here
spare shale
Click on this and find this path:
dense drift
which one?
spare shale
dense drift
ok and then?
dense drift
the asstets folder or one of theese?
spare shale
name your model
dense drift
rename it
dense drift
spare shale
Create a folder with the name of your model
dense drift
ok i did
spare shale
Try again on the site
dense drift
should i put it inside the weights folder
spare shale
- Change here to dataset
dense drift
ok i did
spare shale
/content/dataset
dense drift
yes
spare shale
Now check if it works or not
dense drift
Now check if it works or not
now it says this: RuntimeError: The expanded size of the tensor (12800) must match the existing size (6720) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [6720]
dense drift
Sorry, but I would recommend that you do this without an Interface
what is an interface?
spare shale
Without this site
little creek
Is there a modified version of this notebook: https://github.com/w-okada/voice-changer/blob/master/Realtime_Voice_Changer_on_Colab.ipynb that will automatically download and install a model & index from hugging face? https://huggingface.co/RinkaEmina/RVC_Sharing/blob/main/Emilia_(Rie_Takahashi)_V2_Crepe.zip
dense drift
training model
rare gobletBOT
Ayo? @dense drift level 7 !!!
cold raptor
training model
-rvc
dense drift
but this error is always coming: The expanded size of the tensor (12800) must match the existing size (5760) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [5760]
cold raptor
In which cell ?
dense drift
when i try to train epoches
its the: TRAIN YOUR MODEL - You can also resume training here
cold raptor
Send link to this colab
dense drift
dense drift
send me link pls
dense drift
This is easy gui not Rvc disconnected
pls
normal hazel
-help
abstract kestrel
Is there anyone here who can write and vocalize?
azure marshBOT
RVC Guides (How to Make AI Cover)
dense drift
i found it
thick dagger
how i can fix my ai cover sound have echo
azure marshBOT
Suggestion for @mellow flax
mellow flax
-overtrain
azure marshBOT
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
mellow flax
-rvc
half cove
elaborate
fallow ferry
when ıclick
start_http.bat
its not opening
pls help me
who you help me
when ı click start_http.bat its not working
rare gobletBOT
Ayo? @fallow ferry level 2 !!!
peak urchin
who you help me
when ı click start_http.bat its not working
did you see the terminal
showing up any obvious error
peak urchin
did it popup or no
fallow ferry
when ı click its like loading but its and its closing
fallow ferry
how can ı solve this
peak urchin
you using RVC or w-okada voice changer-
oh yeah w-okada. RVC one doesnt have that name
anyway
fallow ferry
w okada
peak urchin
did everything are in their same position as after-extracting.
low shard
for wokada help maybe its better to ask in #🔍│help-w-okada or #1192011222023950368
fallow ferry
oh ı am sorry ı didnt see there
low shard
its fine dw
vagrant meadow
I'm doing sound modeling. But the voice of the character I chose was removed. What should I do?
odd shale
I'm doing sound modeling. But the voice of the character I chose was removed. ...
Which was the model you were looking for?
You know the name of the character?
verbal oasis
I am a super newbie, can someone help?
I wanted to get used to this and try to train a model for the first time.
I dowloaded the audio froma youtube video of a 1 minute monologue of a guy and followed the steps.
At the last step I get error
What am I missing?
After I click on Train Model, it takes time to elaborate and then it prints out a chinese titled .log file.
AfterI click donwload model I get errors
turbid yew
Hey, my mic keeps on capturing litlte noises and idk why... is there a software or something i could use to reduce them?
glad zealot
I am a super newbie, can someone help?
I wanted to get used to this and try to t...
why do you have different models for D and G
verbal oasis
why do you have different models for D and G
for D and G?
glad zealot
also the pretrain for G doesnt work, use the super fix
verbal oasis
so those 2 should match?
glad zealot
yes
glad zealot
yup
verbal oasis
getting Error
glad zealot
can you screenshot the console?
verbal oasis
I now click on Train Model and show you what log I get
before clicking on download model
verbal oasis
can you screenshot the console?
sure gimme a sec
glad zealot
ah you need to wait before the download model
glad zealot
it doesnt train for just a sec, it trains for minutes, even hours depending on your audio
rare gobletBOT
Ayo? @verbal oasis level 3 !!!
glad zealot
yup that one
it said it trained 150 epochs tho wut
can you check the folder on the side in colab?
look in RVC then logs
glad zealot
yup can you check in My-Voice?
verbal oasis
glad zealot
you can download the added file
thats the index that the model uses
also in the RVC folder, look for either weights or assets folder
glad zealot
no the one that starts with added
verbal oasis
glad zealot
yup those are you models
verbal oasis
all those?
glad zealot
you can probably just get the one with the biggest number if you want and test them somehow
just 1
verbal oasis
so I can put the pth and the index into a zip file
glad zealot
yup
verbal oasis
and share in in Huggingface?