im so autistic
#✨│ai-help
analog obsidian
oh also directml i believe
violet heron
Acts like accent
sharp field
lusty kindle
repost
it used to not be like choppy until i put it on 112
but now it is when its on 192 ?? 😭
i am so confused
and like 256 takes foreverrr
feral saffron
-colab
feral saffron
-guide
AI HUB Docs
brittle mulch
which models would you recommend for sunset shimmer and trixie lulamoon from mlp? I am willing to experiment but just wondering if the ones in the server are worth testing
rare gobletBOT
Ayo? @brittle mulch level 2 !!! 
remote gale
everytime i do asio it crashes can someone help
zealous shuttle
welp....guess the colab didn't support m4a format
tiny kelp
where do i put the model link in here man
void dome
guys
rare gobletBOT
Ayo? @void dome level 1 !!!
low shard
where do i put the model link in here man
In ilaria rvc zero, there's the model loader tab
open stag
Colab
glacial pollen
ckpt is just a " checkpoint " but that's not what rvc is programmed to support
rvc uses .pth being " PyTorch "
proud elbow
it seems like a GPT sovits model, you should read the model description carefully
proper shale
yes, just don't stop on a saving epoch
brittle wing
i dont have microphone to use voice changer..............
patent pasture
**i dont have microphone to use voice changer..............**
download droidcam
wispy lodge
How did you even intend to use a voice changer if you don't have a mic!?
proper shale
if you're on local just close cmd
patent pasture
I tried to use Applio, there I turned on "overtraining_detector" (overtraining_threshold = 100) and started training but the training ended on 135 epoch and best epoch was 35. It doesn't suit me. How to do to training didn't stop on small epochs but also saved best epoch?
proud elbow
**i dont have microphone to use voice changer..............**
what source do you want to use if it's not a mic?
proud elbow
I tried to use Applio, there I turned on "overtraining_detector" (overtraining_t...
I barely used that feature, just observe the tensorboard graph
rare gobletBOT
Ayo? @proud elbow level 46 !!!
patent pasture
I tried to use Applio, there I turned on "overtraining_detector" (overtraining_t...
I turned off overtraining_detector but in this case it doesn't save the best epoch
proper shale
yeah no, unless you close on a saving step
proper shale
I tried to use Applio, there I turned on "overtraining_detector" (overtraining_t...
keep training, and disable that feature
the best epoch is, most of the time, the recent one
also, testing is essential
patent pasture
idk, I can train model on 300 epoch and after training I will continue training but with enabled feature to Applio could save model with the best loss/g/total value. Or is it the worst idea?
proper shale
idk, I can train model on 300 epoch and after training I will continue training ...
it isn't necessarily the best model, just the best generator point really
so no, I'd say save all the epochs you can and test every 50 (50, 100, 150, etc)
or every 10 if you're patient ig
it normally says "Saving D/G_2333333" - just wait for the epoch after that one to stop cmd
What's your save frequency
So you know that every 50 epochs it saves...
So if you're on like, epoch 300, it's saving
So you stop on epoch 301
Basically
Yep
Yes
yep
just don't preprocess or feature extract again
then just train with same batch size, you can change saving frequency and total epochs
fathom whale
i dont know where to post this plese forgive UwU
i need help my cover got hate speech blocked was it because of the symbol on his arm or did the ai say something wrong. when i was making anther cover the auto cations in clip champ were writing random stuff please help UwU AND ALSO THE VIDEO WAS REMOVED AFTER 5-4 HOURS
i have made covers using him in past with no block
quaint dock
fathom whale
i am asking for help with youtube i will delete in 5 min plz no ban
for yt plz wait i get an ai cover hate spcash ban and cant send img in general
sorry for pinging you i wanted to say that he is using the rvc web ui so he should be asking for help here right ?
thank you for no ban i got advise : )
fathom whale
this
unique blade
what happend am i gonna be banned
unique blade
ah
native ice
on colab for whatever reason my index and path aren't showing up in the files when it says they were completed?
gg runtime disconnected 
stable dagger
Is there any disadvantages if i disable "save every weight"?
proper shale
on colab for whatever reason my index and path aren't showing up in the files wh...
did you export model as well?
proper shale
Is there any disadvantages if i disable "save every weight"?
no testing until the final epoch
which is terrible tbh
native ice
Well I would have but like I said there were no paths or index files in the model log folder
proper shale
maybe something went wrong with feature extraction? 
native ice
It says everything completed well
rare gobletBOT
Ayo? @native ice level 3 !!!
proper shale
well honestly idk
proper shale
I'm going back to local <:skullsob:1159372531992645662>
yeah do it
unique blade
somebody help me plz
proper shale
Don't use RVC GUI, use other local stuff instead
https://docs.aihub.wtf/ RVC > Local section
vale osprey
-colab
stark wadi
how come my output sounds vaguely Japanese even though it is English? Sorry if that's a weird question. Like certain words.
steep hinge
So what are ya'lls luck with RVC voice generation? Personally I can't make it sound non-robotic regardless of what I try
rare gobletBOT
Ayo? @steep hinge level 1 !!!
violet heron
So what are ya'lls luck with RVC voice generation? Personally I can't make it so...
Are you using RMVPE
proper shale
how come my output sounds vaguely Japanese even though it is English? Sorry if t...
perhaps the dataset is the issue?
or maybe the audio you're using?
steep hinge
Are you using RMVPE
yes
odd shale
i dont know where to post this plese forgive UwU
----
i need help my cover got h...
You're seriously trying to make austrian painter covers?
fathom whale
You're seriously trying to make austrian painter covers?
it was the symbol the other one is fine : )
patent pasture
what does sync_graph from Applio do? I read the description but didn't get anything
fair prism
do yalls know any ai models in UVR5 or smth else that seperates 2 or more different voices from one audio source?
brittle wing
can someone send me the rvc google collab
rare gobletBOT
Ayo? @brittle wing level 2 !!!
gusty scroll
Hello, I've not been able to make Ilaria RVC work at all today. I only use it for tts inference. But today, during the conversion process it started to consistently give me the message: "Error. GPU Aborted Process"
After many tries it also gave me the message : "You have exceeded your GPU quota (8s left vs 60s requested). Sign up on Hugging face to get more quotas or retry in 35 minutes.
rare gobletBOT
Ayo? @gusty scroll level 1 !!!
low shard
Hello, I've not been able to make Ilaria RVC work at all today. I only use it fo...
That's because of the hugging face space running on zerogpu, there's a quota per account (this is a normal thing for all spaces running on zerogpu), ud get more quota if u make an hf account
low shard
can someone send me the rvc google collab
-colab
low shard
check https://docs.aihub.wtf for guides
fleet jay
gusty scroll
That's because of the hugging face space running on zerogpu, there's a quota per...
So it's just a one limited usage? I mean, I've not used it for a week and it gave me "aborted process" as soon as I used it today. And only after multiple failed attempts it gave me the "get more quotas message". Also, I'm pretty sure I already was on my hf account while using it.
low shard
So it's just a one limited usage? I mean, I've not used it for a week and it gav...
No its not one limited usage, how long was the file?
btw you can also use alt acs for more zerogpu quota
gusty scroll
No its not one limited usage, how long was the file?
The file I converted a week ago? 30 minutes
Now I was trying to convert the same file to test different parameters
feral saffron
-colab
feral saffron
-guide
low shard
The file I converted a week ago? 30 minutes
Now I was trying to convert the same...
30 mins? that's alot and also the reason why it consumed your zerogpu quota, id suggest you to try with an alt acc and usually use shorter audio lenght files, this is a problem of zerogpu quota not of the space
gusty scroll
30 mins? that's alot and also the reason why it consumed your zerogpu quota, id ...
The strange thing is that it didn't immediately give me the quota related message at first. Just "process aborted", so I'm not sure if it's just the quota.
Anyways I managed to do it locally (sadly it took some time due to my pc). Will the quota eventually "recharge itself" ?
low shard
The strange thing is that it didn't immediately give me the quota related messag...
Yes the quota recharges itself with some hours
about the process aborted, it might be because the file is too long for a zerogpu environment
gusty scroll
Yes the quota recharges itself with some hours
Well I haven't been using it for a week. So it might have counted the aborted procedures as well
low shard
Well I haven't been using it for a week. So it might have counted the aborted pr...
it might counted as aborted process because the audio was too long
gusty scroll
about the process aborted, it might be because the file is too long for a zerogp...
Yep, I understood. I meant it aborted the process because the audio was too long and then still counted it for the "quota consumption". That's how I can explain what happened.
Do you know if there is a practical way to use tts inference locally? The Ilaria rvc is so easy to use for that...
proven hill
Yep, I understood. I meant it aborted the process because the audio was too lon...
why you need to infer a 30 minite audio anyway
gusty scroll
why you need to infer a 30 minite audio anyway
Ehm, I was actually just testing if it was usable to make audiobooks with a more natural and pleasant voice.
rare gobletBOT
Ayo? @gusty scroll level 2 !!!
gusty scroll
why you need to infer a 30 minite audio anyway
Thanks for making it anyway. As I said, it's very easy to use for tts!
proven hill
Thanks for making it anyway. As I said, it's very easy to use for tts!
thank you!
gusty scroll
thank you!
Is there something similar to use locally?
proven hill
Is there something similar to use locally?
ilaria rvc mainline!
gusty scroll
ilaria rvc mainline!
Oh cool! I suppose it's litterally the same thing, but would use my CPU/GPU. Is there a suggested system requirement to use it?
(I'd ideally like to infer tens of hours <.< eventually)
proven hill
Oh cool! I suppose it's litterally the same thing, but would use my CPU/GPU. Is ...
what gpu do you have?
gusty scroll
what gpu do you have?
mmh That's not a proper separate gpu. It's an integrated graphic card (don't know if it counts xD), it's listed as Intel® UHD Graphics 630.
gusty scroll
what gpu do you have?
Well the CPU did just infer those 30 mins in a little more than 30 mins. Running at 70%. So that's a start. And I do plan on getting a proper GPU eventually
proven hill
Well the CPU did just infer those 30 mins in a little more than 30 mins. Running...
wait are you already using it locally?
outer zinc
Hi, Is pitch guidance optional? All my datasets are singing ones
gusty scroll
wait are you already using it locally?
No, not Ilaria mainline, I manually used edge tts and then RVC locally before talking to you.
proven hill
Hi, Is pitch guidance optional? All my datasets are singing ones
yes optional since its broken anyway
outer zinc
yes optional since its broken anyway
Thank you
gusty scroll
make sense make sense
Was I about to do something bad by using the Mainline locally? Is it outside of my pc's possibilities? By the way I'm Italian too xD
proven hill
Was I about to do something bad by using the Mainline locally? Is it outside of ...
serix? mi fa piacere
se sei italianx mi puoi scrivere in dm
gusty scroll
Ah ok!
rich saddle
Not sure if this is the right channel but I want to know how to make Peter griffin say things
feral saffron
-colab
feral saffron
@low shard Hey man sorry to bother you but can you help me with something?
shrewd nebula
how to save progress to train more later?
empty raptor
If i am getting this straight, if I pitch down the vocals by -4 I need the instrumental to pitched down by -4 also?
zealous shuttle
should i fill the g & d?
plain fulcrum
Wow I came here to ask a question about this exact thing what are the odds😭
I was ganna ask if it's possible to take a model that's already been made and make it better with my own dataset
plain fulcrum
If you are training then yes
Is this an answer to me or the other guy
empty raptor
Is this an answer to me or the other guy
Yea
rare gobletBOT
Ayo? @empty raptor level 4 !!!
plain fulcrum
This might sound dumb but is it as simple as just adding more data on top of the pre-made model?
Cus I found this plankton model from weights.gg but I need it to be able to sing better
zealous shuttle
If you are training then yes
With random number or specific?
empty raptor
Hey what's the website most use to isolate vocals, dums etc?
https://x-minus.pro/ai is my go to
tame mica
Hey what's the website most use to isolate vocals, dums etc?
mvsep
craggy brook
I will make a commercial video for FFP. For this, I am looking for a voice-over artist who can beautifully sing an advertisement line with a 1-minute intermittent spoken text. If anyone wants to do it, you can send me a DM and I will give you the file text (I will name it as credit).
next iris
wheres the download for the voiceai
rare gobletBOT
Ayo? @next iris level 1 !!!
craggy brook
I will make a commercial video for FFP. For this, I am looking for a voice-over ...
I could do it myself, but I'm not very confident in my voice and I don't have a microphone that can produce a good sound.
empty raptor
Use ilaria instead
low shard
<@911742715019001897> Hey man sorry to bother you but can you help me with somet...
Sorry was sleeping
low shard
like this isnt what the youtube videos are showing
Dont use yt tuts, they are alot outdated, check https://docs.aihub.wtf/
low shard
That is outdated btw
gusty kestrelBOT
@wet heart, I have found 6 results that match your search!
Requested by .kllla
sage leaf
I want to use this program in vr chat games, but need another special program? If there's a separate way, I'd appreciate it if you could tell me
craggy brook
I will make a commercial video for FFP. For this, I am looking for a voice-over ...
yessss???
gusty kestrelBOT
@uncut marten, I have found 1 results that match your search!
Requested by gameon__
proper shale
could sb help me i dont know which rvc to use
What do you wanna do?
For normal audio voice conversions (AI covers and other things), check https://docs.aihub.wtf/ - I'd especially recommend checking the RVC > Cloud section in case you don't have a good GPU.
In case you want realtime voice changing, read this guide for W-Okada: https://rentry.co/VoiceChangerGuide
wooden trout
i want to prank some friends and sound like a girl
pastel oak
second link then
wooden trout
thx
rare gobletBOT
Ayo? @wooden trout level 1 !!!
proper shale
Just use Okada
pastel oak
all rvc v2 models work on real time
glacial pollen
In that case, use the realtime rvc's voice changer
https://github.com/codename0og/rvc-realtime-voice-changer
Instructions on installation included
And here's basics for you to understand what does what
Each of the ui elements explained
Sure thing, gluck!
rare gobletBOT
Ayo? @brittle wing level 3 !!!
tame mica
-docs
urban cradle
how do I do this
low shard
how do I do this
what do u wanna do
urban cradle
what do u wanna do
Cover of a song
urban cradle
what do u wanna do
I am on phone too idk if I can do it on phone
low shard
Cover of a song
low shard
I am on phone too idk if I can do it on phone
Yep you can, you can make ai covers online on your phone
you can't do it locally but can on cloud, our docs has everything you need to know
urban cradle
Yep you can, you can make ai covers online on your phone
Yeah it’s to complicated for me
low shard
Yeah it’s to complicated for me
you can also use ilaria rvc zero which is the fastest free online, here's The Guide, you ofc still need to separate the vocals and instrumentals of the song (which is explained in the docs i sent) and get a download link of an rvc model
you just need to read the guides and you will be all fine dw
urban cradle
That’s gonna be a lot but thank you lol
low shard
You are replying to a bot, if you want to use the rvc model in realtime voice changer for calls, check the wokada guides below
-rt
azure marshBOT
-rt
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
low shard
That’s gonna be a lot but thank you lol
i mean there's a program for you that separates the song, explained in the docs, anyways yw and for help you can ask here
yw
vale osprey
-colab
feral saffron
Sorry was sleeping
Hey it’s okay I figured it out. Thank you anyways. : D
patent pasture
what does sync_graph from Applio do? I read the description but didn't get anyth...
.
low shard
Hey it’s okay I figured it out. Thank you anyways. : D
that's good to hear
empty raptor
Can ilaia do tts
proper shale
Can ilaia do tts
There should be a TTS tab in Ilaria RVC Zero, if that's what you mean
But honestly I'd rather generate TTS in ElevenLabs and convert the audio it gives with RVC instead...
empty raptor
There should be a TTS tab in Ilaria RVC Zero, if that's what you mean
But hones...
But 11labs are paid services right
proper shale
But 11labs are paid services right
You can generate 10k characters of TTS for free
Which should be enough
Ofc their voice clonin stuff is paid
But it also isn't that good
empty raptor
Thanks
proper shale
You're welcome :)
glacial pollen
But it also isn't that good
you'd be surprised tbh
It's the best speech cloning we have currently and no xtts or such can match it
perfect style and intonation replication
like, I've seen crazy stuff made with it
and it's me saying it
proper shale
I've actually heard someone saying it didn't recreate a certain voice style
Imma try to find the convo - wait
glacial pollen
Maybe there are some cases but even famous polish voice actors' accents and styles are retained with nearly 100% accuracy
there's always some generalization error possibility in AI, that's normal but I seriously tell you
It'll take a lil bit for open source software to reach that kind of style replication levels
recently tested most of open source tts we have and none comes even close sadly + tons of em lack documentation on the training ( advance aspects ) itself
proper shale
Not saying EL is bad for cloning since it's probably been used in big shit like THE FINALS
But it's kinda hit or miss really. At least from what I've heard.
glacial pollen
But to have results of a " mickey mouse "
that sounds like some user error to me, haven't heard of any AI framework causing it
not even certain experimental transformers
proper shale
glacial pollen
I mean, maybe there was some critical error under the hood at the time 🤔
welp, guess we'll never know
proper shale
¯_(ツ)_/¯
glacial pollen
off topic but, seeing gpt-sovits still not having finetune for jp or eng
is just making me mad
proper shale
glacial pollen
like, how long it's been
proper shale
we need that fr
glacial pollen
2-3 months?
proper shale
id say like
glacial pollen
like, cmon
glacial pollen
ye
so around 3 months, ye
Imo releasing something like that for broader audience and yet, focusing on chinese aspects of it
is super foolish
Pretty bad way of advertising it imo
proper shale
it was like in march
probably earlier - either february or january xd
glacial pollen
well damn
glacial pollen
hmmmm
oh yea, I quietly hope we'll eventually have a really good open source speech cloning ( including style and all that's necessary ) but with a neat ui, similar to how SynthV handles it
with absolutely 100% live control
that'd be a legit game changer
glacial pollen
Hooooo, now this is something
No wonders I couldn't find it as it's not tagged properly
@proper shale Thanks a ton man 🙏
proper shale
No probs :)
edgy monolith
-local
ionic kernel
-colabs
blissful blade
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
chex 0.1.86 requires numpy>=1.24.1, but you have numpy 1.23.5 which is incompatible.
pandas-stubs 2.0.3.230814 requires numpy>=1.25.0; python_version >= "3.9", but you have numpy 1.23.5 which is incompatible.
torchtext 0.18.0 requires torch>=2.3.0, but you have torch 2.0.1 which is incompatible.
torchvision 0.18.0+cu121 requires torch==2.3.0, but you have torch 2.0.1 which is incompatible.
Successfully installed all packages!
can anyone help me with this ?
sacred roost
could someone help me just i feel like every time i talk the end of my sentence is cut off
analog obsidian
could someone help me just i feel like every time i talk the end of my sentence ...
are u using okada? #🔍│help-w-okada use this channel
sacred roost
im using realtime voice changer idk
analog obsidian
im using realtime voice changer idk
use this pic as reference
try sample length 1.5 and extra infer 2
and use rmvpe
sacred roost
i only have rmvpe_onnx it doesnt let me choose the others
rare gobletBOT
Ayo? @sacred roost level 1 !!!
sacred roost
and im not sure where the other sample length setting is?
analog obsidian
and im not sure where the other sample length setting is?
oh
let me guess
you have an amd gpu
right?
analog obsidian
i only have rmvpe_onnx it doesnt let me choose the others
if you're on nvidia realtime_gui you're supposed to get normal rmvpe
not onnx
sacred roost
idk i have a rtx3070
analog obsidian
idk i have a rtx3070
realtime_gui nvidia should look like this
(ignore settings lol)
can u confirm me u are using this
sacred roost
oh im dumb nvm
analog obsidian
blissful blade
ERROR: pip's dependency resolver does not currently take into account all the pa...
anyone know fix to this Im using Realtime Voice changer on colab
blissful blade
thats work thx
rare gobletBOT
Ayo? @blissful blade level 1 !!!
prime quail
does anybody have a mew
knotty moth
does anybody have a mew
your mew is in another castle
rare gobletBOT
Ayo? @prime quail level 1 !!!
prime quail
well how do i level up then
prime quail
im banned.
tidal trail
prime quail
bro give me a mew
tidal trail
i do not have a mew
prime quail
your name is like mew without the o
prime quail
im banned
and no one responds
i want a mew so bad
this is what desperation looks like
tidal trail
this channel is for help with rvc problems
prime quail
i have an rvc problem
knotty moth
what is it? does ur mew voice model sound weird?
rare gobletBOT
Ayo? @knotty moth level 3 !!!
prime quail
and its that i dont have a mew in my pokedex
lean bane
so, ive been getting this "Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass" error for a few days now, and i just got back from work and not a single bit of training was done. how do i fix this error?
neat merlin
so im using illaria RVC (hugging face) i put the model already and when i try to convert it keeps saying error is there a way to fix it?
rare gobletBOT
Ayo? @neat merlin level 1 !!!
neat merlin
and it shows a "0.0" red color on the index n pth
quartz valley
okay, Not sure if this the correct chat channel, just ping me
Using Applio's TTS, is there a way to emphisized emotion in the text input?
tidal trail
okay, Not sure if this the correct chat channel, just ping me
Using Applio's TT...
edge tts do not have emotion
you can use elevenlabs tts
and then convert the tts audio in applio
quartz valley
thanks
quartz valley
you can use elevenlabs tts
how do you use ElevenLabs's tts emotion?
rare gobletBOT
Ayo? @quartz valley level 1 !!!
tidal trail
how do you use ElevenLabs's tts emotion?
ehh tbh idk, i never use elevenlabs tts
pallid hare
How do I make so that the voice has the intensity and sounds more like the second audio, but still with the pitch/key of the first one?
As a reference, use Numb as an example
I think it gets clearer around the second chorus, 2:00 in the song
Also, the model I'm using is this one
https://weights.gg/models/clm72s5lu0ny1cctcpbfayzpb
lean briar
is there a way to use the default models from w-okada in rvc? they're in onyx format but i found a .pth for the tsukuyomi model from huggingface and when i try to use it the client freezes
prompt says these errors:
TypeError: init() takes 19 positional arguments but 20 were given
AttributeError: 'RVC' object has no attribute 'tgt_sr'
knotty moth
@solemn salmon regarding the Errno 13
solemn salmon
I believe I checked group permissions for myself for Apollo
empty raptor
I believe I checked group permissions for myself for Apollo
By mistake?
solemn salmon
No, I mean I looked and everything said I was in the clear
knotty moth
I believe I checked group permissions for myself for Apollo
try moving the entire applio to another D:\ or something
empty raptor
No, I mean I looked and everything said I was in the clear
Oh my bad
solemn salmon
try moving the entire applio to another D:\ or something
I was told Applio and other apps like this will break if they’re not on C: drive.
knotty moth
I was told Applio and other apps like this will break if they’re not on C: drive...
mine just works wdym
solemn salmon
Alr, then I’ll try it later today
rare gobletBOT
Ayo? @solemn salmon level 4 !!!
solemn salmon
I just still don’t understand why the program would error only after a restart
Except mainline
knotty moth
I was told Applio and other apps like this will break if they’re not on C: drive...
well I mean any root folder including C:
knotty moth
My go to rvc tool is actually Kits.ai
don't waste ur money on that app, while there are free RVC local and colab/spaces
empty raptor
don't waste ur money on that app, while there are free RVC local and colab/space...
I hack used some software before picking Kits.ai and both of google colab/spaces in my opinion is great but it can only store one model at a time
Unlike Kits.ai i store at this moment 4 of 7 bts members ai voices
low shard
My go to rvc tool is actually Kits.ai
kits.ai sucks
paywalled rvc
there's guides on how to use rvc for free https://docs.aihub.wtf
low shard
I hack used some software before picking Kits.ai and both of google colab/space...
wdym store only 1 model at a time? after you train the voice model, for example in google colab, it saves to google drive
empty raptor
paywalled rvc
I use ilaria
low shard
I use ilaria
ye ilaria is way better, it's an rvc fork
empty raptor
I am technical in software manipulation but it sometimes outputs an errorv
low shard
I am technical in software manipulation but it sometimes outputs an errorv
u mean the ilaria rvc zero?
knotty moth
havent had any issues on the latest ilaria
empty raptor
havent had any issues on the latest ilaria
Can u send me the lonkn
empty raptor
u mean the ilaria rvc zero?
Yes
low shard
Yes
the errors are either bc the audio is too long (u cant do very long inference on zerogpu) or bc u finished ur zerogpu quota (it recharges with time, and u can js use an alt acc)
soo yea
empty raptor
the errors are either bc the audio is too long (u cant do very long inference on...
Cab it do 3mins sondw
knotty moth
Can u send me the lonkn
empty raptor
Songs
low shard
Cab it do 3mins sondw
ye, just watch out for the zerogpu quota
when u use it, it ofc uses the zerogpu quota of ur acc
its a thing of every zerogpu hf space
sour spire
-colab
brittle wing
Why does overclocking my memory GPU result in this error
My gpu is stable during long gaming sessions
patent pasture
what does sync_graph from Applio do? I read the description but didn't get anyth...
.
empty raptor
ye, just watch out for the zerogpu quota
What about if there is some harmony
lavish oasis
Anyone can help me?
empty raptor
Anyone can help me?
On what
lean bane
so, ive been getting this "Variable._execution_engine.run_backward( # Calls int...
I think i figured it out, i cant use a vpn while training
proper shale
why is the voice changer not picking upo my microphone
check command prompt
wispy lodge
vrchat+rvc with rtx 3070
Easily
tawny needle
quest pro+virtual desktop
rare gobletBOT
Ayo? @tawny needle level 2 !!!
tawny needle
should i limit vrchat frame?
wispy lodge
Well, in VR mode it may be more difficult. You might need to reduce settings like VR resolution or framerate
rare gobletBOT
Ayo? @wispy lodge level 28 !!!
tawny needle
thank you. i will try reducing resolution.
glacial pollen
My gpu is stable during long gaming sessions
Because Machine learning is way more intensive than gaming on memory
both mining and machine learning
sour trout
hello, my CMD says "wait web server .... 120 and counting" ... the UI is white
proper shale
hello, my CMD says "wait web server .... 120 and counting" ... the UI is white
delete stored_setting.json and restart okada
sour trout
it worked, thanks
proper shale
you're welcome :)
tidal trail
he left
rare gobletBOT
Ayo? @tidal trail level 7 !!!
pearl turret
is amd rvc now fixed?
dire spoke
(the new) Ilaria RVC isn't working at all for me. Importing voice models (with any method) doesn't work at all for me, and even if I just don't do anything and refresh the 'select a voice model' bit it returns an error. Using base 64 btw
rare gobletBOT
Ayo? @dire spoke level 1 !!!
glacial pollen
is amd rvc now fixed?
Not sure about the mainline and dml in general but, perhaps it's somewhat fixed or improved in Fumiama's version of mainline:
https://github.com/fumiama/Retrieval-based-Voice-Conversion-WebUI
Wouldn't hurt to check
But I think you'd still have to resort to linux and Rocm
but yea, give it a try I suppose
pearl turret
thank you sir
ripe girder
rotund prawn
okada channel
split valve
how do i make it so discord can hear the voice changer
proper shale
how do i make it so discord can hear the voice changer
virtual audio cable
https://rentry.co/VoiceChangerGuide explains it
edgy lion
Need help with an error i keep recieving using rvc
tidal trail
Need help with an error i keep recieving using rvc
what error?
edgy lion
what error?
I'm using pinokio, do you know how i can copy the error text?
tidal trail
I'm using pinokio, do you know how i can copy the error text?
copy the txt ig, or send a pic of the error
edgy lion
yeah copying doesnt seem to wokr like normal i will screenshot it one sec
rare gobletBOT
Ayo? @edgy lion level 1 !!!
tidal trail
i think there is something missing
edgy lion
this is what it says when i first launch without running anything
tidal trail
Idk about Pinokio, maybe installing it again or hope someone helps you.
It seems that you have 2 libraries that are causing you problems or something was not installed correctly
edgy lion
okay i will try and install manually rather than using pinokio and see if that fixes it
thanks
pallid hare
How do I make so that the voice has the intensity and sounds more like the second audio, but still with the pitch/key of the first one?
As a reference, use Numb as an example
I think it gets clearer around the second chorus, 2:00 in the song
Also, the model I'm using is this one
https://weights.gg/models/clm72s5lu0ny1cctcpbfayzpb
I'm using Mangio RVC, mangio-crepe option. I don't know if there are other RVC models able to change the voice intensity or pitch without changing the tone, but I couldn't find any.
low shard
What about if there is some harmony
u should try to clean that out
cloud idol
anyone know how to add Formant shifting to RVC pinokio WebUI?
solemn salmon
During a model training I got this error Fail to allocate bitmap. Do I need to worry about it, and can I just resume training?
blissful blade
how I can fix this ?
Cannot connect to GPU backend
You cannot currently connect to a GPU due to usage limits in Colab. Learn more
To get more access to GPUs, consider purchasing Colab compute units with Pay As You Go.
odd shale
how I can fix this ?
Cannot connect to GPU backend
You cannot currently connect...
You'll probably have to switch between your google accounts to keep using the colab.
blissful blade
oh do I need to switch my google drive as well ?
odd shale
oh do I need to switch my google drive as well ?
Which colab are you using?
blissful blade
blissful blade
Which colab page i mean.
HIna modified Realtime
odd shale
HIna modified Realtime
Yeah, try using that colab with another account.
another google account.
blissful blade
ok thx
edgy monolith
-rvc
jolly onyx
-colab
rare gobletBOT
Ayo? @jolly onyx level 2 !!!
wet light
my RVC GUI always says it can't find the path specified, how can I fix it to make it run?
"Running with the runtime Python, Please wait.
The system cannot find the path specified.
Press any key to continue . . ."
proper shale
"Running with the runtime Python, Please wait.
The system cannot find the path s...
anything before that?
and how's your path to rvc look like
violet heron
my RVC GUI always says it can't find the path specified, how can I fix it to mak...
RVC GUI Zia outdated
-local
pallid hare
How do I make so that the voice has the intensity and sounds more like the secon...
Oh well, I guess not then.
restive quest
there is an rules to use rvc or what?
restive quest
What do you want to do
I want to use voices with RVC mode
but It's not work to me idk why
I tried a lot to search what's problem or how to fix that
but usless
violet heron
I want to use voices with RVC mode
In realtime or for covers
restive quest
In realtime or for covers
wdym?
rare gobletBOT
Ayo? @restive quest level 2 !!!
brittle wing
Because Machine learning is way more intensive than gaming on memory
Is it possible to make it run stable with my memory and gpu core overclocked
I think it would somewhat give a performance gain
restive quest
Do you want to use the voices in real time or to create covers
I think realtime?
violet heron
I think realtime?
-realtime
azure marshBOT
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
violet heron
First link should have everything
restive quest
"live"?
violet heron
First link should have everything
Downloads, settings, bug fixes
polar wyvern
What is better to use locally, ilaria, mainline or APPLIO?
polar wyvern
-local
lavish oasis
On what
I use Voice.ai when I upload the custom file it wont work
It says you cannot upload this files
To upload we need .zip files right?
glacial pollen
Is it possible to make it run stable with my memory and gpu core overclocked
Hard to tell but personally I wouldn't bother. The gains would be absolutely minimal
simple overclock won't bump your core counts or speed up your memory's bus
you know wut
I actually am investigating something
analog obsidian
I actually am investigating something
you fixed fp16?
edit: well, not really fix, but adding the option to train in fp16
glacial pollen
Imma test one thing
I have one suspicion, a thing I might have overlooked when enforcing fp32 some time ago in dev test of my fork
analog obsidian
analog obsidian
haha
glacial pollen
Guess what happened
glacial pollen
So basically, script works in a way where you have:
configs
inuse configs
configs are the templates
and they are only rewritten with fp32 ( as in, setting fp16 to false ) if gpu isn't capable of it
if it is capable tho, it leaves them as they are ( and potentially copies to inuse if they are missing )
in my case, they were not missing
but I previously forced them to be set to false ( cause I was testing fp32 )
that way I posted it on github
🤦
analog obsidian
glacial pollen
basically, just edit both configs and inuse configs with " true " in fp16 run
and that's it
analog obsidian
gonna try it
glacial pollen
yes
analog obsidian
yes
it seems is in fp16 now
config.json looks like this
but its normal im getting a bit slower speeds? in mainline i remember every epoch in this model was around 18 seconds but here is every 23-26 seconds
glacial pollen
ur gpu? @analog obsidian
analog obsidian
ur gpu? <@775545133448953856>
4060
glacial pollen
10 min set, 16 batch, 14-16 secs for me ( fp16 as well )
3060
and I have some gpu tasks in the background ( video decoding
So it's probs something with your own system's case
Cause you see, it's not uniform per hardware and it's actually dataset dependent really
analog obsidian
wait
is that the 3060 12gb or 8gb?
for some reason in mainline max i can use batch 8, but in applio allows me up to 12
i have 8gb of vram
analog obsidian
which variant you have?
4060 8gb
glacial pollen
welp, that's all clear now
your gpu's memory bus is only 128 bit
mine's 192
and 12 gig on top of that
analog obsidian
glacial pollen
so
I wouldn't care about few sec difference anyways
you get more perks in return
✨
analog obsidian
yeah haha its worth xD
knotty moth
I use Voice.ai when I upload the custom file it wont work
that app is terrible, better use the voice changer in the guide #🔍│help-w-okada message
ornate gulch
hi
vale osprey
-colab
lavish oasis
that app is terrible, better use the voice changer in the guide https://discord....
For the low end pc which one? I already installed but the delay..
ember dust
Hi everyone - does RVC just clone the higher formants like F4 and above? I ask because it seems like the singer's accent comes through and not my model's. In other words, does RVC only clone over a certain frequency?
tawdry hound
I basically installed one from a video about how to get the villager voice
it was on this video https://www.youtube.com/watch?v=OHJuaJukhCI&t
And it had me download something
i need to open the start_http
I opened it and when i did it would flash on my screen for a second
then disappear
rare gobletBOT
Ayo? @tawdry hound level 1 !!!
tidal trail
it was on this video https://www.youtube.com/watch?v=OHJuaJukhCI&t
Do not watch video tutorials
tawdry hound
Do not watch video tutorials
Why?
tidal trail
Read this https://rentry.co/VoiceChangerGuide
tidal trail
Oh
If you any problem with that ask in https://discord.com/channels/1159260121998827560/1159290161683767298
thin stump
Hi everyone - does RVC just clone the higher formants like F4 and above? I ask b...
and not my model
you mean the dataset? I think RVC tries to learn what you feed it and assigns it to a pitch. If you had a singer like Tiny tim who specializes in falsetto but regular singing as well, that gets mixed in if that makes sense
tawdry hound
If you any problem with that ask in https://discord.com/channels/115926012199882...
I know u told me to ask quesitons there but aint no one telling me so which one do i download
thin stump
> and not my model
you mean the dataset? I think RVC tries to learn what you fee...
I mean the falsetto gets assigned to the wrong parts sometimes
analog obsidian
Hi everyone - does RVC just clone the higher formants like F4 and above? I ask b...
yes, rvc has pitch range limits, og pretrain goes max to A5, after that it artifacts, klm pretrain can increase this a bit (around the C6-A6 range)
i don't know the technical explanation sorry but, yes, rvc can't do any pitch like synth v/vocaloid
when the model reaches a pitch it doesn't understand things go weird
tho when the model doesn't pronounce things like the original voice, its either undertrained or index value is too low (try around 0.5-0.7 index value)
icy ibex
Hello👋🏻, is it possible to adjust the volume of two individuals separately in an audio track?
empty raptor
Hello guys, how are we?, I just wanna know if I can isolate backing vocals from the main lyrics that a singer is singing
I removed the reverb but there is a backing vocal with the mix of the main chorus or bridge and
scenic copper
how do i download the software to use the voice models
empty raptor
how do i download the software to use the voice models
Ilaria or Gcolab! is my two backup options
empty raptor
Do you want me to show?!
lavish oasis
Do you want me to show?!
Sure
empty raptor
mind you i am doing it on a phone that’s why it didn’t work
lavish oasis
So the file will automatically there when we click download??
empty raptor
So the file will automatically there when we click download??
Both files would highly likely to be there when you use Discord ai hub like i shown
empty raptor
So the file will automatically there when we click download??
Working?
knotty moth
go to inference page then click refresh model
brittle wing
Hi, there are several apps exist for voice change. Which one is best?
violet heron
Hi, there are several apps exist for voice change. Which one is best?
-realtime
azure marshBOT
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
violet heron
Yes, realtime
https://rentry.co/VoiceChangerGuide try out w-okadas voice changer
It’s free, open source, and can use models from #1175430844685484042
brittle wing
Thx for the information, i will be try
rare gobletBOT
Ayo? @brittle wing level 1 !!!
violet heron
<https://rentry.co/VoiceChangerGuide> try out w-okadas voice changer
The guide here should have settings, downloads, and bug fixes
empty raptor
does anyone know how to remove backing vocals within a vocal only audio
knotty moth
does anyone know how to remove backing vocals within a vocal only audio
UVR BVE
knotty moth
Sane place yeah?
empty raptor
Do i need to download it
knotty moth
mvsep.com and x-minus.pro also has that model
knotty moth
Do i need to download it
yes, it is local installation, or consider the above for cloud alternative
knotty moth
empty raptor
Ty
shrewd nebula
applio vs rvc disconnected which is better
rare gobletBOT
Ayo? @shrewd nebula level 3 !!!
quasi thicket
Hello anyone knows the best realistic voice changer here?
empty raptor
applio vs rvc disconnected which is better
I don’t like using applio
shrewd nebula
well i use it cuz its interface is simple
i tried using rvc disconnected and my model results were awful af
the epochs were trained fast af, like 1 epoch was trained in 3 seconds i dont understand why.
thin stump
the epochs were trained fast af, like 1 epoch was trained in 3 seconds i dont un...
ye rvc disconnected is out of sync
and cant be fixed since it's been abandoned, just mostly bug fixes here and there
shrewd nebula
which you use?
thin stump
I switched to mainline colab, or ilaria rvc colab
shrewd nebula
ilaria
thin stump
they're almost the same thing. Applio is the only one that has mangio crepe
shrewd nebula
i see , thx
thin stump
i see , thx
most simple gui i've seen imo
im not sure if it have graph syncing which I go into depth here https://rentry.co/RVC-Mainline-Colab
but applio probably does it by default
empty raptor
What Aggressiveness value should i use
pale raft
are there any free things which i could use to clone a voice which sounds decent. Im broke asf.
i have no idea how to do any of this ngl
thin stump
i have no idea how to do any of this ngl
check dms?
knotty moth
What Aggressiveness value should i use
leave the default value mostly
empty raptor
leave the default value mostly
Will do
polar ember
Is someone able to vc with me to help me get my system working?
im quite lost i have everything downloaded and running but its not working
after further research it looks like my gpu is unsupported is there a versioun i can use with a GTX780?
knotty moth
Is someone able to vc with me to help me get my system working?
im quite lost i ...
don't think there's anything using PyTorch that supports old gpus
pastel oak
780 is very old not sure if its worth trying. what application you trying anyway
fluid thicket
I've been scouring for couple hours trying to solve all these errors and this made me happy. I agree Mangio simply works. First of probably many errors I have yet to solve is opening the program and on the top of the cmd prompt it says winerror bla bla something about permissions (though GUI does open).
Well I mean it works but specifically the training portion has been an absolute headache and doesn't work.
empty raptor
leave the default value mostly
Thanks it worked flawlessly
polar ember
780 is very old not sure if its worth trying. what application you trying anyway
i will do anything
i will reinstall any appication you think would work
pastel oak
i will do anything
that doesnt answer the question, there are applications for different purposes
polar ember
ive gotten 1 of them to work it was beatrice on MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a
it worked but none of the other voices did and i cant upload my own
also i can vc and screen share to try and solve any issues
pastel oak
its funny you still didnt answer if its for voice changing, if its for ai art etc.
just teasing lol
you can try emojikages fork, its optimized for worse gpus but no promise itll work
polar ember
its funny you still didnt answer if its for voice changing, if its for ai art et...
sorry voice changing
rvc
pastel oak
ive gotten 1 of them to work it was beatrice on MMVCServerSIO_win_onnxgpu-cuda_v...
https://github.com/deiteris/voice-changer/releases/tag/b2083
download
voice-changer-windows-amd64-cuda.zip.001
voice-changer-windows-amd64-cuda.zip.002
then check out the installation description
https://github.com/deiteris/voice-changer/blob/master-custom/README.md
pastel oak
wispy lodge
Modded version will not work with GPUs prior to GTX 900 series
pastel oak
ok so i just need to buy a new graphics card?
ye
fluid thicket
Error: Unable to connect to Rich Presence. [WinError 5] Access is denied
Theme Loaded: Applio
Running on local URL:
To create a public link, set share=True in launch().
Starting preprocess with 8 cores...
[WinError 123] The filename, directory name, or volume label syntax is incorrect: '"C:\AI\RVC-Training\Berry\Dataset"'
Preprocess completed in 0.00 seconds.
Is this fine orrr? My brain is exploading. Its a 23 minute dataset doing V2 and 48000
Also when I continue on from there, the dataset and file location is the same as I have always done. I haven't been doing anything different.
"Starting extraction with 8 cores and rmvpe...
F0 Extraction: 0it [00:00, ?it/s]
F0 extraction completed in 0.03 seconds.
Starting feature extraction...
An error occurred in the feature extraction, make sure you have provided the audios correctly."
empty raptor
Can mvsep do instrumental separation
knotty moth
Can mvsep do instrumental separation
there are some models like demucs ft that separate drums and bass
empty raptor
there are some models like demucs ft that separate drums and bass
Ty
knotty moth
Ty
UVR beta roformer patch also has BS roformer 1053 that separates bass & drum (together in a stem) from others
empty raptor
Ty
quasi talon
I'm so confused is this the channel on how to add voices and use the voice changer?
empty raptor
I'm so confused is this the channel on how to add voices and use the voice chang...
No this channel is purely for people who has some questions about different things about rvc
Are you in need for some ai models #1175430844685484042 is the channel you might need
knotty moth
I'm so confused is this the channel on how to add voices and use the voice chang...
basically anything using RVC technology, i.e. realtime & non-realtime voice changer, voice model training, etc
polar ember
ye
Im just confused because i got it working with 1 voice i was wondering if i could replicate the process on a different voice
pastel oak
Im just confused because i got it working with 1 voice i was wondering if i coul...
you have a rare issue involving only working beatrice, probably a reinstall was needed
you reinstall by deleting "pretrain" and "model_dir" folders and then run start_http again
polar ember
Im just wondering if there are any other models like beatrice becuse the other models give me this error: CUDA error: no kernel image is available for execution on the device
rare gobletBOT
Ayo? @polar ember level 2 !!!
polar ember
wait im getting it to work by using cpu but its quite laggy
pastel oak
wait im getting it to work by using cpu but its quite laggy
cpu is a bad idea
polar ember
why?
polar ember
yeah but i cant getting it working at all with gpu
pastel oak
whats the point in 20 second delay voice changer tho
might aswell do inferencing instead, which is putting a recorded audio in it and it lets you download a file with the model you selected
polar ember
CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
this is the error i get with gpu
is there any other voices like beatirce im confused why that one works
wispy lodge
You can get it to work on CPU but it may be not great since it heavily depends on CPU
polar ember
ok but why dose beatrice work?
wispy lodge
Because that's a different architecture, not RVC. It works on CPU but has only few models afaik
knotty moth
CUDA error: no kernel image is available for execution on the device
CUDA kernel...
make sure you're using a supported Nvidia gpu and updated driver recommended
polar ember
make sure you're using a supported Nvidia gpu and updated driver recommended
im now
not
wispy lodge
beatirce works on my gpu
Beatrice works only on CPU. GPU choice doesn't matter for it
knotty moth
ok well when i get a new pc ill be back
I'd recommend RTX 3060 12 GB or higher ones for better gaming performance
polar ember
Ye ill probably get a 4090
knotty moth
if you're patient enough, wait for some 5090 anonuncement and release in the end of this year
Core 12600 or Ryzen 5600 are generally sufficient, but if you're looking for high-end one, there are Ryzen 9000-series coming later in end of July
https://www.techspot.com/news/103620-early-amd-ryzen-9000-zen-5-listings-suggest.html
turbid flax
any help is appreciated!
polar ember
if you're patient enough, wait for some 5090 anonuncement and release in the end...
I will wait for that then buy the 4090
pastel oak
please delete the last one if it has copyright in it
turbid flax
can i upload it without instrumental
rare gobletBOT
Ayo? @turbid flax level 2 !!!
pastel oak
yes
turbid flax
so ive been trying to make a cover but for some reason the audio is really wack and idk why.
the model ive used works fine on everything else, its just this one song (although the distortion was in one part of another song somewhen)
other models have no problems with the song, but when i cover the cover it still gets wack.
does anyone know why this occurs or even how to fix it?
(i have used applio)
the following audios are: output with another model, output with the model i wanted, normal output with that model.
proper shale
so ive been trying to make a cover but for some reason the audio is really wack ...
mess with pitch
feels like the higher pitches are screwing up the model
stuck pewter
File "/content/Ilaria-RVC/app.py", line 2, in <module>
from mega import Mega
ModuleNotFoundError: No module named 'mega' how can i fix itt pls
stuck pewter
in wich page is the fix
proper shale
File "/content/Ilaria-RVC/app.py", line 2, in <module>
from mega import Mega...
use ila rvc zero instead as ilaria pointed out
stuck pewter
where is the link to ila rvc zero?? sorry guyss
zealous shuttle
how should i fill this?
can someone give the example?
especially with manual epoch number
stuck pewter
ichika help me
rare gobletBOT
Ayo? @stuck pewter level 1 !!!
stuck pewter
Ayo? <@1115301972246020227> level **1** !!! <:lfg:1159355870119993496>
yo chill i was just asking for help
zealous shuttle
ichika help me
sup?
proper shale
where is the link to ila rvc zero?? sorry guyss <:boohooh:1176674698629750975>
but if you don't want to look at it: https://huggingface.co/spaces/TheStinger/Ilaria_RVC
astral pasture
im new to this
how do i make mine generally configured to sound less robotic thanks.
proper shale
how do i make mine generally configured to sound less robotic thanks.
what r you using specifically
astral pasture
trying to get a girl voice to work
proper shale
mess with tune, index & extra
astral pasture
I have a relatively deep voice and have australia accent
proper shale
tune = pitch
index = applies new accent
extra = gives more time for gpu to get better results, so more realistic output
low shard
in wich page is the fix
You were using an outdated colab, use the newest ilaria rvc zero link they sent
astral pasture
extra iss maxed
rare gobletBOT
Ayo? @astral pasture level 1 !!!
astral pasture
Thanks for advice I will mess with index
analog obsidian
Thanks for advice I will mess with index
Don’t use index for realtime, increases cpu usage like crazy and its actually better having that off
astral pasture
I am using gpu i got a 2080ti
analog obsidian
Use pitch 12 or 14
knotty moth
are there any good paid models? that would make it better?
check in #1191429836321849435
analog obsidian
<@775545133448953856>
#1191429836321849435
or use #1159289738314919936 and remember to post it as a paid request
astral pasture
Would some models not work well because of my natural voice though?
analog obsidian
Would some models not work well because of my natural voice though?
Regular e-girl voices should be able to do male-female conversions
use tune 12 and above and hear the results
Sometimes u might need more than tune 12
If the model sound bad in realtime, its either settings being bad or the model is bad
And also rvc models in realtime works best if you mimic the original voice way to speak/accent
So for example e-girls models, they sound better if you actually talk like a girl
zealous shuttle
<https://docs.aihub.wtf/rvc/cloud/rvc-disconnected/>
i know the rest, but what i confused is manual epoch number
proper shale
Would some models not work well because of my natural voice though?
Probably not
turbid flax
mess with pitch <:nails:1159569314848972891>
i'll try thanks
proper shale
i'll try thanks
Alrighty - let me know how that goes
turbid flax
Alrighty - let me know how that goes
it surely makes it better, when the song gets louder. but it has still a lot of noise. interestingly, i put the original audio twice in a row and it isnt 100% the same so there is probably more to it than just pitch. ill try to find a solution though
proper shale
turbid flax
i guess ill just leave it at that
empty raptor
Guys help
Why does the audio cut pff certain lyrics , even though I used Dereverb
Thanks in advance
river wren
Guys my audios all choppy i jus wanna sound like kanye fr
empty raptor
Guys my audios all choppy i jus wanna sound like kanye fr
me too, but my problem is with dereverb or inverted audio
I swear can someone help me so bad again
the input file? yea
would you suggest me to use mvsep and use urc backing mode?
knotty moth
would you suggest me to use mvsep and use urc backing mode?
same models as there
empty raptor
i used that mode to made the audio i gave
the one you shown me
the audio i gave is the one you got from me
i dont have
collab have that
i guess that i have to use a accustic version instead
empty raptor
How big is UVR to download
rare gobletBOT
Ayo? @empty raptor level 8 !!!
compact violet
Trying to train on applio colab, but it gives me this error when I preprosess dataset
Starting preprocess with 2 cores...
[Errno 20] Not a directory: '/content/drive/MyDrive/HIGH.flac'
Preprocess completed in 0.00 seconds.
What should I do ?
compact violet
What should I do ?
oh never mind I solved it by putting the flac file in a folder and copy the folder's path instead of the flac file's path 😅
vale osprey
-colab
twin jay
Anyone got andrew tate voice model?
How to make those AI voice models have better quality?
rare gobletBOT
Ayo? @twin jay level 1 !!!
twin jay
I downloaded andrew tate voice model and it doesn't really sound like him
proper shale
How to make those AI voice models have better quality?
mess with pitch stuff.. if the pitch is quite different the model won't convert well
tardy badger
Hey guys, where can I get tenserboard download?
jaunty cave
how can i fix with voice changer cleint demo that i hear myself in dc and then the ai?
proper shale
Hey guys, where can I get tenserboard download?
proper shale
how can i fix with voice changer cleint demo that i hear myself in dc and then t...
hm?
i suppose your output stuff might be wrong but im not sure if i understood
tulip adder
KLM
earnest sparrow
Getting this model trained for the first time, got this error. Any fixes?
AssertionError Traceback (most recent call last)
<ipython-input-16-52e2e92e6faf> in <cell line: 9>()
7
8
----> 9 assert 'pretrain_type' in locals(), "You need to download a pretrain! Please run the "Download Pretrained Model" cell before continuing."
10
11
AssertionError: You need to download a pretrain! Please run the "Download Pretrained Model" cell before continuing.
proven hill
the solutions is to read the error
earnest sparrow
But I haven't pretrained a model, and steps close to that don't do anything.
proven hill
choose one, download it, do it again
rare gobletBOT
Ayo? @earnest sparrow level 2 !!!
earnest sparrow
choose one, download it, do it again
But I'm making a model of a family member. I can't just load in any old model, can I?
obtuse rapids
how to seperate different charachter audios into their own audio files
mighty nimbus
why do the ai vocals sound robotic and out of vocal range?
solar plover
depends on the audio you use for outputand what was used to train it and stuff
snow rose
Yo @violet heron , so about those custom lyrics i still have no clue. I literally now seen someone using alastor voice with his accent and word building in sentence at the start of video? How is even something similar made
mighty nimbus
depends on the audio you use for outputand what was used to train it and stuff
ive noticed that crepe worked better than the others. is it the best one to use?
rare gobletBOT
Ayo? @mighty nimbus level 1 !!!
solar plover
ive noticed that crepe worked better than the others. is it the best one to use?
personally i use the uhhh new one
harvest
tardy badger
why does it take so much time to train my 7 minute dataset? i have a 3060 ti and its like 2 hours already
is this normal or not
with the rvc
mighty nimbus
personally i use the uhhh new one
i tried with harvest at first it was trash lol
solar plover
i tried with harvest at first it was trash lol
violet heron
i tried with harvest at first it was trash lol
Use RMVPE then
snow rose
Has anyone used applio rvc? is it any good?
tidal trail
Has anyone used applio rvc? is it any good?
It has some extra additions like tts and other options apart from its friendly interface
analog obsidian
Has anyone used applio rvc? is it any good?
is the most easy fork to use and have a lot of useful plugins
brittle wing
can we rvcgui on linux?
craggy patrol
can't you just retrain from the d and g files but check the weights button
in the training page is there a spot for you to copy and paste the d and g file paths?
see the bottom left
craggy patrol
not gonna work out correctly
isn't it just like starting the training again ??
rare gobletBOT
Ayo? @craggy patrol level 1 !!!
proper shale
isn't it just like starting the training again ??
nope
it alr does that automatically
btw, that's not necessarily your best model
i think there's another way but i am not sure how to do it
all ik is that there's a possibility to generate the weight .pth from the d/g
rare gobletBOT
Ayo? @fierce cave level 4 !!!
proper shale
might be on the "ONNX/PTH processing" tab or whatever it's called on the GUI
knotty moth
Has anyone used applio rvc? is it any good?
it's completely free with tons of features and better than any other paywalled garbages
snow rose
it's completely free with tons of features and better than any other paywalled g...
So now which one is better, Mangio rvc or applio rvc?
knotty moth
So now which one is better, Mangio rvc or applio rvc?
Mangio is old already but quite stable
tidal trail
So now which one is better, Mangio rvc or applio rvc?
mangio is outdated
snow rose
mangio is outdated
0_0
tidal trail
use applio or mainline
snow rose
to this day im using mangio
terse turret
help help somebody help!!
zinc herald
i keep getting this
ember shore
Hey guys I just found a model website called "weights", is it safe?
thin stump
Hey guys I just found a model website called "weights", is it safe?
1 year worth of server logs and still being called "unsafe" is crazy 
empty raptor
To those people who have urv how big (storage wise ) is it
empty raptor
Hey guys I just found a model website called "weights", is it safe?
Yes I downloaded one ai model from there no worries about it
But you have have have to make a account
rare gobletBOT
Ayo? @brittle wing level 3 !!!
knotty moth
if you accidentally canceled the Drive mount prompt or it timed out, try running the cell again, or did it fail for some reason?
rich tiger
Who can help generate Travis Scott's voice? small chorus
grave kestrel
do i need to install tensorflow via pip to run applio's or any rvc training?
pastel oak
do i need to install tensorflow via pip to run applio's or any rvc training?
therws rvc mainline prepackaged where you dont have to run any commands and can get started right away
applio local never tried out
pastel oak
Who can help generate Travis Scott's voice? small chorus
its very easy to infer vocals, check out the latest pinned message and open the colab + guide
rich tiger
its very easy to infer vocals, check out the latest pinned message and open the ...
I don't have the necessary equipment for this now (computer), can you help me?
brittle wing
the vc doesnt work
grave kestrel
therws rvc mainline prepackaged where you dont have to run any commands and can ...
which rvc mainline are you talking about? is it this one?
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/RVC1006Nvidia.7z
glad zealot
those are the same thing
grave kestrel
should i follow the instruction envronment in the git RVC project because those have requirements to install (a lot) other things but in RVC1006Nvidia walkthrough i only need to install tensorvenv.bat
pastel oak
theyre the same thing, the first is prepackaged the second isnt
I uploaded a more up to date prepackage if you want it - theres not really any difference afaik but maybe small
https://huggingface.co/Shadicti/RVCLatest/blob/main/RVCNvidiaLatest.zip
pastel oak
should i follow the instruction envronment in the git RVC project because those ...
if you know how to install dependencies and like things to be manual, then go for it
else download what i sent you + tensorvenv.bat
grave kestrel
i see thanks
i ran the tensorvenv.bat but i got this error:
WARNING: The script virtualenv.exe is installed in ..\Desktop\RVC1006Nvidia\runtime\Scripts' which is not on PATH.
Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.
should i add the Scripts folder to the environment path?
reef trellis
hello ! I'm Jul ,and i would you demand ,why ,the appplication ,didn't want to record my voice ,when i clicked on the "start" button , i only had a bit sound that is reflected
thank you
empty raptor
Hi guys, I know I am repeating myself, but I still can't remove the backing vocals from my audios despite trying all of your suggestions, including using different models on MVSEP and UVR-BVE-4B_SN_441001-1. I haven't had any luck so far. The only success I've had is with acoustic versions of my favorite songs, but those don't have the full energy I need for the edit I'm planning. Can you guys help me in some way, shape, or form?
pastel oak
i ran the tensorvenv.bat but i got this error:
WARNING: The script virtualenv.ex...
i mean all this shouldnt happen if you downloaded rvc1006 for the first time and put tensor inside the rvc1006 folder (not in any other folder)
i dunno about that sry
pastel oak
hello ! I'm Jul ,and i would you demand ,why ,the appplication ,didn't want to r...
what is the full name of your GPU
pastel oak
Hi guys, I know I am repeating myself, but I still can't remove the backing voca...
At this point theres just so much you can do, not every vocal is perfectly splittable
empty raptor
At this point theres just so much you can do, not every vocal is perfectly split...
How small or big is UVR 5
odd shale
Hi guys, I know I am repeating myself, but I still can't remove the backing voca...
I'd suggest you to stick with acoustic versions, since not all songs are perfectly splittable.
Sorry.
empty raptor
I'd suggest you to stick with acoustic versions, since not all songs are perfect...
Ty
upbeat bloom
Why sometimes i hear strange noises?
pastel oak
Why sometimes i hear strange noises?
your mic is probably picking up everything you do, putting your hand on your table, someone in the background slamming the door etc
upbeat bloom
your mic is probably picking up everything you do, putting your hand on your tab...
how can i fix it?
pastel oak
how can i fix it?
do you use w-okada (mmvcserversio) or RVC gui
upbeat bloom
rvc
pastel oak
You can try enabling input and output noise reduction, but rvc is very limited in the noise suppression stuff
upbeat bloom
noise reduction on discord?
rare gobletBOT
Ayo? @upbeat bloom level 1 !!!
upbeat bloom
where?
upbeat bloom
i got a feeling you have w-okada and not rvc gui if you cant see it
i have this
upbeat bloom
oh
pastel oak
enable sup2
and i recommend f0 det: rmvpe_onnx
also your output has to be CABLE, not monitor. monitor can be your headphones to hear yourself
upbeat bloom
also your output has to be CABLE, not monitor. monitor can be your headphones to...
on output i need to put cable ?
pastel oak
yes
upbeat bloom
i put it
pastel oak
good
upbeat bloom
i need to do something else or is ok?
pastel oak
is ok
upbeat bloom
thank you 😄
obtuse glen
hey dpes anyone know what is the state of the art for stem extraction with music?
im trying to make a drum model
jolly onyx
I want to remove the noise in AI voices as best as possible (those little "tv static" sounds you hear when pronouncing some syllables), should I change any of these?
viral birch
Where i can make covers? (sorry for the question, im am an stater)
viral birch
Where i can make covers? (sorry for the question, im am an stater)
Covers on a android phone
jolly onyx
Where i can make covers? (sorry for the question, im am an stater)
type "-colab"
viral birch
-colab
viral birch
Thanks
jolly onyx
I don't know how it works on android, but it should work the same
your time on google colab is limited, so prepare everything before proceeding
I'd recommend the Ilaria RVC by thestingerx
it's really easy to mess around with
viral birch
Ok, thanks
jolly onyx
these buttons will be there
press the download one, wait until it finishes, and then click on the run one
it'll give you a gradio live link
that's where you wanna go
odd shale
I'd recommend the Ilaria RVC by thestingerx
Ilaria RVC Zero or Applio Colab
cyan cobalt
anyone know how I can use my amd gpu to use rvc?
rare gobletBOT
Ayo? @cyan cobalt level 1 !!!
cyan cobalt
anyone know how I can use my amd gpu to use rvc?
even tho its not supported
I kinda don't have money to buy a nvidia gpu currently
and it hurts my soul to have 12gb of vram and not be able to use it 😭
odd shale
im trying to make a drum model
I'm not sure but you could use Demucs.
odd shale
and it hurts my soul to have 12gb of vram and not be able to use it 😭
Just use Applio Colab if all you want to do is use models.
obtuse glen
I'm not sure but you could use Demucs.
well i used demucs but i wanted to know if there was something even better
cyan cobalt
Just use Applio Colab if all you want to do is use models.
so amd is useless in local terms?
odd shale
well i used demucs but i wanted to know if there was something even better
I'm not sure if there are some drum isolation models on X Minus pro (it's paid tho)
odd shale
so amd is useless in local terms?
No, i mean, i'm suggesting you a faster solution.
obtuse glen
I'm not sure if there are some drum isolation models on X Minus pro (it's paid t...
send link?
cyan cobalt
No, i mean, i'm suggesting you a faster solution.
thanks
odd shale
send link?
There is a Mel Roformer drums model but i'm not sure if you can use it for free.
obtuse glen
There is a Mel Roformer drums model but i'm not sure if you can use it for free.
is it open source somewhere
perhaps to get it running on my own machine
odd shale
perhaps to get it running on my own machine
Nope.
obtuse glen
huh