#✨│ai-help
1 messages · Page 112 of 1
That's actually really useful oml
Ayo? @fluid zephyr level 10 !!! 
feel free to tag me if you encounter any issues or anything else, im avaliable and happy to help
Thank you! Will do 😄
it is!
of cours it not perfect and everything actually depends on a lot of factors but it gives a good idea to newcomers!
this is an exampl, hope you like the feature!
Can i train models locally with AMD graphics card ?
Like, will it be slower than google collab thing ?
sadly, no.
training requires CUDA cores wich AMD dosnt have.
you can the colab of ilaria rvc mainline OR rvc disconnected
Dam, okay, thanks
no problem!
depending on the card i think you can use linux and ROCm
but... convoluted and honestly it's better to use colab at that point
its an option but how mch slow would it be
don't think it'd be slow, just a convoluted process ig
but im not sure about speed tho
Sounds hard, i'd rather use colab
yeah
I can't drop the WAV file for train model i didn't see a message box for the drop file in Ilaria rvc !! how to plz help me (sry for my gramma)
hey i want, to make ai covers but when i open the go-web.bat its not warking,please help
also dont look at my pfp, i lost in a bet
we are working on this rn!
how do i open the rvc?
good evening, would he have a huggingface, which allow you to train a voice and make a model?
impossible, theres not enough resources on huggingface to do that
wich one
oh okay, thanks
no problem!
mangio
go web
Do you have a Google colab that allows you to do this, because all my old collabs no longer work to train the voices..
i did it
you can try the beta version of ilaria rvc
whats the error
Could I have the link please?
i press a key and it does nothing
do you have python installed?
no
there you go
why do i need it?
because it use python
ok
oh, it doesn't work on phone.. 😦
so i just need pyton and it will work?
yes it does, use he colab version
technically
ok
I can have it please I can't find it..
its the colab link
-rvc
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
How To Make an AI Cover With Ilaria RVC
Link: Rentry
Credits: 👽 Julia (ailen2091)
@proven hill
?
where can I find the training helper?
extra tab
-colab
- Applio, by IA Hispano Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
?
the extra tab in the new ilaria rvc
how do i install ilaria in mac?
Ayo? @lost quartz level 1 !!! 
@vale birch make sure you have saved files like these in your Drive before
if you have to start over, well do that and don't forget to save files like those via the save model cell
oke
@proven hill I have a problem, when I do train model, it gives me an error, even though I did everything in order
nvm found it
hi, which one is better? RVC disconnected or RVC CLI ??
test and see for yourself tbh, but rvc disconnected seems to be pretty nice for colab
which software can i use to train voices on a AMD grapichs card?
what does cmd say?
@proven hill @half cove how long is training helper supposed to take
depends on audio length but not that long
let me see what they use for calculating audio length rq
1 sec
i think it was changed but i dont think it should take that long
if you closed cmd that might be why
Ayo? @calm jackal level 2 !!! 
try a shorter audio if that doesnt work show what cmd says
I didn't
what does the ui cell say
but if I try a shorter audio the prediction might not be accurate
is it ok if my computer gives me the blue screen with my rvc
not for the helper i just want to see if thats the problem
I'm using the Easy GUI google colab
idk. I am running on colab
ok so prob the way it calculates audio or sample rate
one is super slow
tell me how long ur audio is?
ill just tel you manually
rn
how to input voice ai?
use normal pretrains and 32k
But I don't think it was actually doing anything in this one
cause it doesn't say queue 1/1
it shows on the run ui cell
that seems ok
i assume you reset it
if you put your other audio does it not work still?
I'll try again
which rvc are you using?
it shows no errors 
but timed out
weird
i assume this might be a public gradio moment
@half cove
yea i saw that and asked this
Do you mean on the colab
yes
what was the thing called i train with
so what can we do bout this
i trained before so i have minimum experience
true
Just wait
guys
im doing a fresh install of ila rvc rn cuz mine has random shit
what is the training gui called
from updates
hm?
lol
.
local
you can use mainline, ila rvc, applio
i think it was applio
-links
- 🍏 Applio, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team GitHub
- GPT-SoVITS, by RVC-Boss GitHub
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
i have new gpu and im bored
xd
-colab
- Applio, by IA Hispano Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
the software for removing music
I don't know
@half cove
hm?
read
uvr for local mvsep for online
what
got it
DEPRECATION: omegaconf 2.0.6 has a non-standard dependency specifier PyYAML>=5.1.*. pip 24.0 will enforce this behaviour change. A possible replacement is to upgrade to a newer version of omegaconf or contact the author to suggest that they release a version with a conforming dependency specifiers. Discussion can be found at https://github.com/pypa/pip/issues/12063
just download the main download link
no
i forgot your downloading applio
@half coveit cannot find path
Failed to train index: [WinError 3] Das System kann den angegebenen Pfad nicht finden: 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\3_feature768'
Ayo? @brisk quarry level 3 !!! 
idk
just use tensorboard if you want good models
hm
what does it do
what did you try?
so
yeah but it stopped going up after 170 on tb and I trained till like 250. But I suppose there was still more potential left in it
did you go straight to train model?
yes
yes
dont
there are other options
yes
where do training and INDEX file save? in which folder
doesnt explain what it does
in simple terms has a graph that you use to tell if your model is overtrained
how do i know if it is
if shit goes up
we have it in our guides
leme find
Last update: Feb 10, 2024
UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
grad.sizes() = [64, 1, 4], strides() = [4, 1, 1] | 0/14 [00:00<?, ?it/s]
bucket_view.sizes() = [64, 1, 4], strides() = [4, 4, 1] (Triggered internally at C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\torch\csrc\distributed\c10d\reducer.cpp:334.)
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass | 0/14 [00:00<?, ?it/s]
@half covetensorboard dont work
lots of errors
is that good or bad
yes
what epoch u at?
well it didnt do much
its kinda slow
yep
y
thats how ai is
how long per epoch what gpu what batch size how long dataset
you would need more
overtrain tho
how to open Voice Changer I did everything right, I restarted the computer, but I do not know how to open Voice Changer
start_https.bat

Isn't it the http not the https
idk what u expected with 40 epoch 😭
both work but one is web
forgot which one which
I have already downloaded it, but how to open it after restarting the PC
1 sec
i recommend 8
i have 12gb vram
????
yep http is the better one for most users
in your folder that you unzipped there is a file called start_http.bat
thank you very much
Ayo? @pseudo pier level 1 !!! 
Ayo? @brisk quarry level 4 !!! 
@half coveis that good or bad
just check tensorboard
click refresh
i do
yea wait
when loss/g/total goes up
then stop
also
i think applio has a auto detect overtraining
thing
you can make it auto refresh every 30 seconds
in training settings
gpu dependent
yep but usually
yea
but it didnt save
it should have in logs
tf do i do with logs if i need the model
thats where shit saved
download python
Which version?
90*
guys when the training is done, where do i find the files?
Ayo? @limpid plinth level 1 !!! 
what fork u using
Applio in the browser
3.10
hm?
uh
3.10
so colab or local?
colabù
what happend to fp16
prob set to false
how do i set to true
in the code
Downloaded it, same error.
did you click this?
colab
Just did now.
in colab file manager as program_ml then logs
or if you clicked the auto drive option in extra it would save it in drive for you
@half cove
what zip did you download 
No, but it won't even let me update it either...
@brittle wing 
i mean it doesnt use runtime for the assistant i think
But this is like... kind of important?
wait no
ooo
this should fix the updating bit
its a pain with directories with other one so ill do later
Is there someone else who can help with that besides you?
I... kinda wanted to use this now.
@half coveyou said its taking hours
real
@half coveRuntimeError: The expanded size of the tensor (12800) must match the existing size (0) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [0]
bro said nuhuh
tf did u do
nvm
i forgot to enter dataset
hehe
@half coveidk what i did
i get error again
Loaded checkpoint 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\D_3100.pth' (epoch 100)
Loaded checkpoint 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\G_3100.pth' (epoch 100)
C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\site-packages\torch\autograd_init_.py:251: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
grad.sizes() = [64, 1, 4], strides() = [4, 1, 1]
bucket_view.sizes() = [64, 1, 4], strides() = [4, 4, 1] (Triggered internally at C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\torch\csrc\distributed\c10d\reducer.cpp:334.)
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
Process Process-1:
Traceback (most recent call last):
File "C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\multiprocessing\process.py", line 315, in _bootstrap
self.run()
File "C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\multiprocessing\process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\train\train.py", line 264, in run
train_and_evaluate(
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\train\train.py", line 449, in train_and_evaluate
wave = commons.slice_segments(
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\lib\infer_pack\commons.py", line 49, in slice_segments
ret[i] = x[i, :, idx_str:idx_end]
RuntimeError: The expanded size of the tensor (12800) must match the existing size (0) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [0]
Saved index file 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\added_IVF701_Flat_nprobe_1_v2.index'

Ayo? @brisk quarry level 5 !!! 
i have no idea
@half covehelp
idfk what is wrong
Traceback (most recent call last):
File "C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\multiprocessing\process.py", line 315, in _bootstrap
self.run()
File "C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\multiprocessing\process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\train\train.py", line 264, in run
train_and_evaluate(
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\train\train.py", line 449, in train_and_evaluate
wave = commons.slice_segments(
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\lib\infer_pack\commons.py", line 49, in slice_segments
ret[i] = x[i, :, idx_str:idx_end]
RuntimeError: The expanded size of the tensor (12800) must match the existing size (0) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [0]
Saved index file 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\added_IVF1143_Flat_nprobe_1_v2.index'
what are you doing to get that error
nothing else than before
training?
length
check ur audios
for what
if 1 is like suuuper small
because
just add it to the other audio
blud has 5 ms audiofile
iirc 4 sec should be fine
ill remove it
try
Starting training...
Loaded checkpoint 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\D_3100.pth' (epoch 100)
Loaded checkpoint 'C:\Users\Lukag\Desktop\Applio-3.0.9\logs\armstrong\G_3100.pth' (epoch 100)
C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\site-packages\torch\autograd_init_.py:251: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
grad.sizes() = [64, 1, 4], strides() = [4, 1, 1]
bucket_view.sizes() = [64, 1, 4], strides() = [4, 4, 1] (Triggered internally at C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\torch\csrc\distributed\c10d\reducer.cpp:334.)
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
Process Process-1:
Traceback (most recent call last):
File "C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\multiprocessing\process.py", line 315, in _bootstrap
self.run()
File "C:\Users\Lukag\Desktop\Applio-3.0.9\env\lib\multiprocessing\process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\train\train.py", line 264, in run
train_and_evaluate(
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\train\train.py", line 449, in train_and_evaluate
wave = commons.slice_segments(
File "C:\Users\Lukag\Desktop\Applio-3.0.9\rvc\lib\infer_pack\commons.py", line 49, in slice_segments
ret[i] = x[i, :, idx_str:idx_end]
RuntimeError: The expanded size of the tensor (12800) must match the existing size (12200) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [12200]
whats the difference between every pitch extraction algorithm
just use rmvpe
i know but i wana know what each one does
no
but removing it didnt help
try again
🗿
im confiused
removing 1 more audio fixed it
nvm
it crashed again
Does anyone have a working Google Colab page?
Ayo? @scenic pagoda level 1 !!! 
Suggestions for @scenic pagoda
- Applio, by IA Hispano Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
yesterday i started training my voice model, but i run out of gpu's time so it stopped
Is there any guide on how to use it? I see many links but I don't know how it works, it doesn't look the same as before, and I don't want to train a model, it's just to use the ones I have
today i went again to train it and it again from the beginning, why'd that happen?
time???????
colab
ok
best valorant jett ai voice
no
yes
well I'll investigate on my own
yes, check out our docs https://docs.aihub.wtf you will find guides to the google colabs and each of them what they do (for example, one for training, another for using the models etc) inside the cloud folder inside of the rvc folder of the docs
Last update: Mar 10, 2024
i personally use ilaria rvc
i made this before

markiplier
why is it stuck at epoch 34
how to do it better
Ayo? @brisk quarry level 6 !!! 
RVC Guides (How to Make AI Cover)
Translation by country
Any updates?
Fun…
So uh… can you handle the stuff needed to get the pretrained assistant thing working? Sorry, I just want to be prepared…
hello, guys when i upload a custom voice, it says file not found, idk why. and no i didnt extract the file. can anyone help me?
I hope that’s cool.
just do pip install requests in cmd
But through where? Where do I initiate the cmd?
right click windows icon click on run then write cmd
then ok
So the normal directory that opens when I open cmd is okay?
I don’t have to open it through another directory, yes?
It works!

hi guys i use rvc and it sounds a little bit choppy and bad... theres also a lot of delay. Is there anything I can do to maybe make this a little better?
how do i know how much echos and batch_size i should set ?
what gpu u have and with epochs just do 500 and check if it overtrained using tensorboard
amd rx570
thats what i do
then use 8
and what about echos ? can i check using tensorboard on colab ?
after making a new model i get this error again for no reason
RuntimeError: The expanded size of the tensor (12800) must match the existing size (0) at non-singleton dimension 1. Target sizes: [1, 12800]. Tensor sizes: [0]
hii!
yo
where i download the real time voice changer
Ayo? @worn prawn level 2 !!! 
which is the best out of these?
Ilaria RVC
Applio Colab
RVC Disconnected
AICoverGen
Paperspace
not here
none
everything sucks
😭
Ayo? @lean valve level 1 !!! 
is there anything better than those?
no
where
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
but if u really had to choose, which one would be the one?
for colab training try the new ila rvc rvc disconnected or the applio one ig
anyone who knows a software that works with AMD grp card ?
I tried using cloud but no luck
what
amd just suck
realtime or normal?
tyy
sorry I'm new
what's that
realtime is voicechanger
no, I just have to train the voice (index and rdp or smt like that)
using prezip
it dont use runtime python
anyway i just made em download python and requests
all fine
whats the point of prezip then
prezip but u gotta download requirements
amazin
didnt have python either
the script doesnt use runtime python
music detected
no
probably
check rules
no?
if you wanna send copyrighted song use youtube
the server gets in trouble if members send copyrighted audios
cuz aint hosted here
yes it does...
what if i use a file host
uhhhhhhh

here
if it has no instrumentals its fine
Ayo? @broken lagoon level 1 !!! 
is my gpu not supposed to heat
what do i do if it says "no-feature-todo"
I've been out of the voice ai loop for a while, is rvc 2 still the best?
how so
i don't know what any of it means
i don't get how the dataset path is incorrect given that it was able to detect my dataset
yes i know
im saying i don't know what anything on the cmd means
scribbled out my name
is it my GPU?
I got a question?
How do you connect to Xbox with ai voice or is that possible
oh
Ayo? @broken lagoon level 2 !!! 
which
alright
is this what it should say? my computer is lagging like fucking crazy rn so im a bit worried
idk
AMD Ryzen 7 5825U with Radeon Graphics
i think
that's what i found as my processor
r u using the radeon or do u have a nvida card or smth
It works with amd if you really want it
so my computer had a heart attack lmao
ty
can i ask questions about the Ilaria rvc here too ⁉️
What is the best sounding pretrain thing in Ilaria rvc?
whats your dataset length?
I haven't made one yet, It's been a while since I have made one. Which are best for which lengths?
was this on colab?
under 8min try ov2 for the rest try normal
and 90% of the time use 32k sample rate
you mean the huggingface AI voice conversion site?
Ayo? @fiery raptor level 3 !!! 
iirc ila rvc isnt on huggingface spaces
im asking if your using local or colab
whats the diff
Anyone got the latest easyGUI for models? I had a Google notebook link from a few months ago but it doesn't seem to work anymore
local on your computer
-colab
Suggestions for @dim saddle
- Applio, by IA Hispano Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
oh the google one
huh
Thanks!
I use this little thingy
One more question, is UVR still the best for vocal isolation?
for pc yep
this a little too advanced for me
a new beta came out for viperx bs roformer
what the colab?
yeah
ill work on a guide
but
run the first cell
then the second
it will give you a link
am i even in ther ight place
click that and opens ui
screenshot the page
so what do i run the google drive backup?
thats mostly for training
if your only making covers then no
let me set that as off on default
1 sec
ight
don
you quick asl
so download completed?
oh
Sorry, now last question, what is the best model on uvr?
mdx23 but there is a new beta for viperx bsroformer models that are super good (even better than mdx23c)
Thank you so much for the help. Haven't messed with rvc since rvc 2 first came out
Ayo? @glass cypress level 2 !!! 
i mean your not downloading anything but googles computers are

oo lala
alright finished installing
now what
click on the run cell
So uuh I've tried easyGUI, Applio and Illaria but neither lets me upload my voice models? There's only a text box with an arrow that does nothing... 😭
Ayo? @dim saddle level 1 !!! 
this thingy?
yes
hm?
This is where I'm supposed to upload my .pth models? Because it's not letting me do anything in here other than typing
there is a download models tab
download the model
then click refresh and it will be in the dropdown
wait it told me sum about a link that expires
Oh... thanks 😅
dw
click the gradio public link
ayyy
Ayo? @fiery raptor level 4 !!! 
now someone should make a Ilaria RVC kaggle version
shouldnt be too hard
one last thing. how do I put voices in this RVC
maybe ill do that
there is a download model tab
colab or local?
im running this on colab
Ayo? @glad veldt level 4 !!! 
there is no audio folder in there
maybe its not on the github
leme check
yep
ok so
make a folder in the ila rvc one called audios
should do the same thing
maybe ill make it create it on startup
i dont know if its really that needed though
I mean for me it times out when i try to upload my audio file using the "drop in" method
times out while converting specifically
what do you mean by times out?
was this when clicking convert or uploading the audio?
convert
and the run cell was running the entire time?
yep
seems like gradio public has a limit
i might add Cloudflare or ngrok
if thats the case
think it's most likely the case
okay I did that
yeah but it's not displaying the audio file
i see another problem
it's not in the menu
audios doesnt have a refresh button
rerun the gui
should refresh
maybe ill fix that
okay now it's insta erroring
without the timeout message
the audio file is there
what?
I selected it
tf
hit convert


fuck
this shit has caused me so much pain https://github.com/TheStingerX/Ilaria-RVC-Mainline/pull/38
ima wait till the main is fixed
i aint touching it
did uploading the audio using dropdown have the same error?
as the
uploading audio
what error did the other have
using dropdown this happens with me:
this happens when directly uploading
just tested and it doesnt block me from uploading big files
dw about the spam of this is a test i just needed a audio
oh wait
it did
balls
well
basically
gradio seems to have a limit
and
audio dropdown is fucked
so
fuck
faojladkjfslkjaf
i'm having an issue registering with weights.gg
you could ask in #1212837348463607890
the weights.gg devs check that channel sometimes
river is the owner and beatsboy is a dev iirc
RVC Guides (How to Make AI Cover)
Translation by country
hey when i paste the link (the voice model) in the google doc it gives me an error, what do i do?
@half cove when using ilaria mainline colab I also get the jsondecode error, what do I do?
How the actual fuck
somebody help me pls lol
hey when i paste the link (the voice model) in the google doc it gives me an error, what do i do?
Ayo? @mint rivet level 1 !!! 
Might be the pretrains json
Then
If its training
Idk how to fix it tho
I installed the right json
I'm pretty sure
I chose one pretrain then it chose to install all of them for some reason
CAN SOMEONE HELP
What's the full error
Old colab
Suggestions for @mint rivet
- Applio, by IA Hispano Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
wtf do i do then
the same error ppl get when using an old colab
except its longer
a longer error
Send
well I can't send it now becuz I started over, ig I could see if the error pops up again, if it does I'll ping you
Okie
use the updated colabs inside the cloud folder inside the rvc folder of our docs https://docs.aihub.wtf
Last update: Mar 10, 2024
so basically i upload it and then i put the song link and i get the cover
bro wym the name 😭 🙏
like what do i put here
Ayo? @mint rivet level 2 !!! 
@half cove got the error again, here's what it looks like
this is from ilaria rvc mainline colab
I tested them and both the overtrained & the 320 epochs one sound good not overtrained
Will save on notes
alr
I'll check code tomorrow
oke
the non overtrained model will sound better
Well it might not preform with high notes as good
Honestly they're not very different
But use whatever you want
If you want I can DM you samples
Are you sure cause it's meant for rapping mostly
I cant tell you because I cant hear your model
Can I DM you the samples or send them there.
You could send them here
I'm converting them into MP3.
Anyways the model is very realistic.
Done
Alr
It even has his accent in "heard" his raspy voice everything...
Which one
Ayo? @brittle wing level 8 !!! 
Here
Which model out of the two.
it keeps showing "ModuleNotFoundError" can anyone help me?
I dont see the MP3 files
....
These
I sent em
They're here I sent em
hey guys i just joined the server
can someone please send me the collab link
how do istart creating ai music for fun?
- Applio, by IA Hispano Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Applio, by IA Hispano Huggingface Spaces
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HFv2, by r3gm Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
the rvc collab to clone voices i having been here since the first server got closed xD
Ayo? @regal hearth level 1 !!! 
What was it called
THANK U SO MUCH
Ah okay
Your first instinct was to spam in #model-maker? 🤔 You need to get RVC here https://docs.aihub.wtf/essentials/how-to-make-ai-cover/
It's outdated + RVC is way better
Can you help me I can't decide-
I trained a model yesterday someone told me it's a little overtrained so I saved it at lesser epochs but the overtrained one sounds better
overtrained on the g/total? I just pick the step count closest to the dip even if it's considered overtraining on the guide
hi
like if my dip is at 13,300 steps and the one closest is 13,330 then I would use that rather than 12,800. (These are example numbers)
i am new to ai-cover, if i follow i need to download rvc to make them myself ?
Ayo? @wide sleet level 1 !!! 
Read the guide here. It will teach you how to make covers and where to download models https://docs.aihub.wtf/essentials/how-to-make-ai-cover/
ok thank you
hey guys on weights the website once the ai starts to suceed and sing can it improve?
Ayo? @pseudo lark level 1 !!! 
whenever I try to install mangio fork by opening the bat it says error code 0x80072efe
Am I the only one whose page we use for AI models fails?
Ayo? @cyan sparrow level 2 !!! 
why are the icons just loading forever
hi there im trying to install tensor board locally, but for this attached google doc tutorial the discord links dont work https://docs.google.com/document/d/1ihvqWnSyabvZ9jX6BsOxEY1FpplwjY3Jse3she_VO6I/edit
i finished training my model and now i have these files in my google drive folder. i tried copying the links for every one of them but none of them seem to work when trying to use the model
thank u soooo much for real, you saved me from a mental breakdown 💀😭
anyone needs help? im free
What do you use to clean datasets?
I use audacity and RX 10
What plugins do you use to clean datasets? (if any)
I also use RX
I'm just trying to find out common tools that people use
lucky you get to use RX
what specific? like spectral denoise and declick, deplosive, etc?
I use denoise de ess and deplosive and sometimes declick if there are clicks in the audio
AiCoverGen (Modified by Hina)
RVC Guides (How to Make AI Cover)
Translation by country
it works now thou
i only use audacity
Hello, no matter which option I choose in the rvc application, it uses the CPU, not the GPU, and that's why it lags. Can you help me?
My gpu AMD
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/gradio/routes.py", line 437, in run_predict
output = await app.get_blocks().process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1346, in process_api
result = await self.call_function(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1074, in call_function
prediction = await anyio.to_thread.run_sync(
File "/usr/local/lib/python3.10/dist-packages/anyio/to_thread.py", line 33, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 877, in run_sync_in_worker_thread
return await future
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 807, in run
result = context.run(func, *args)
File "/content/Ilaria-RVC-Mainline/infer-web.py", line 753, in click_train
config.json_config[config_path],
KeyError: 'v2/OV2-40k.json'
Ilaria colab error how to fix?
whats rvc?
how do i enable fp16
It wasn't very overtrained
Now I dm you the samples & you tell me.
Okay I'll send them again
Maybe cause you arent Yes
They have instrumentals please delete
It must be raw output from RVC
Just listen to them quickly a and I'll delete them
Okay I'll remove the music
And send the raw output
👍
You shouldve just ran the vocals through RVC again instead of isolating the RVC vocals from the song
But they sound very similar
I guess the top one sounds better?
Honestly I used a webui that mixes everything but I still.knoe how to mix & add reverb
Oh
How?
Because it may sound a little less robotic?
The removal artifacts and reverb really make it difficult to tell
The non-ovetrained one
Yeah
Wait
Ask others possibly of what they think to not get such a one-sided opinion
But I would recommend to remove reverb and noise from the dataset and try to remove backing vocals if any
guys is applio broken?
im getting this error when trying to run it : ImportError: cannot import name 'Option' from 'typer' (unknown location)
make: *** [Makefile:20: run-applio] Error 1
Also I usually do 200 to 250 to 280 epochs for example this I dont stop at the final epoch i stop at a model near the near lowest steps not the lowest cause that kinda results in a robotic sounding model imo
Ayo? @rugged stirrup level 19 !!! 
My dataset was good I did everything correctly.
Now listen to the samples and tell me.
I know how to make a dataset very well


